Navigating Language in the Early Islamic World: Multilingualism and Language Change in the First Centuries of Islam 2503603017, 9782503603018

Citation preview

Navigating Language in the Early Islamic World

Interdisciplinary Studies in the Middle Ages and the Renaissance

Publications of the Marco Institute

Volume 2

Editorial Board under the auspices of the Marco Institute for Medi­eval and Renaissance Studies, University of Tennessee, Knoxville Gregor Kalas (General Editor), Justine Andrews, Manuela Ceballos, Heather Hirschfeld, Robin Jensen, Roy Liuzza, Anne-Hélène Miller, James T. Palmer, Alison Vacca

Navigating Language in the Early Islamic World Multilingualism and Language Change in the First Centuries of Islam

Edited by

Antoine Borrut, Manuela Ceballos, and Alison M. Vacca

F

British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library.

© 2024, Brepols Publishers n.v., Turnhout, Belgium. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise without the prior permission of the publisher. ISBN: 978-2-503-60301-8 e-ISBN: 978-2-503-60302-5 DOI: 10.1484/M.ISMAR-EB.5.131893 Printed in the EU on acid-free paper. D/2024/0095/80

Table of Contents List of Illustrations

7

Acknowledgements 11 Introduction: Multilingualism in the Early Islamic World Antoine Borrut, Manuela Ceballos, and Alison M. Vacca 13 Section One Languages of the Umayyad Caliphate The Early Islamic Empire’s Policy of Multilingual Governance Petra M. Sijpesteijn 43 The Translation of the Dīwān and the Making of the Marwanid ‘Language Reform’: Secretarial Agency, Economic Incentives, and Regional Dynamics in the Umayyad State Marie Legendre 89 Towards an Arabic Cosmopolis: Culture and Power in Early Islam Antoine Borrut 167 Section Two Multilingualism, Empires, and Local Elites The Arabs and Northern Languages and Scripts before Islam Muriel Debié 195 Triliteral Coins and Political Authority along a Contentious Frontier: Between Arabic, Bactrian, and Pahlavi in Late Antique Khurāsān Robert Haug 259

6

ta b l e of con ten ts

Arabic and the Public Performance of Power  in Early Medi­eval Armenia Alison M. Vacca 281 From Bactrian to Arabic: Changes in Seals and Sealing Practices as Observed in the Documents from Bactria Judith A. Lerner 307 Section Three Languages in Contact and Shared Spaces Navigating Persian: The Travels and Tribulations of Middle Iranian Languages Khodadad Rezakhani 329 Once Again, the Twin Histories of Arabic and Aramaic (with a Focus on Syriac) Aaron Michael Butts 365 Sharing the Written Space: Contact and Interaction between Arabic and Other Cultures and Scripts Arianna D’Ottone 403 ʿAbbāsid Book Culture and Ninth-Century

Jewish Sectarianism Fred Astren 425

Index 467

List of Illustrations Figure 0.1. Muʿāwiya’s Greek inscription at Hammat Gader. From Whitcomb, ‘Notes for an Archaeo­logy of Mu’awiya’.

16

Figure 0.2. Latin solidus minted in Ifrīqiya during the reign of al-Walīd b. ʿAbd al-Malik after the supposed language reforms of his father. 

17

Figure 0.3. Bilingual Arabic-Uyghur coin minted under Abaqa b. Hulegu. 

22

Figure 0.4. Bilingual Arabic-Sanskrit coin minted under Maḥmūd of Ghazna, showing the shahāda in Arabic and its loose Sanskrit translation. 

28

Figure 1.1. An Arabic-Greek protocol dating to 98–99/716–17.

57

Figure 1.2. Eighth-century Greek list of prisoners with their place of origin and the crime they committed.

58

Figure 1.3. Arabic protocol dating to the financial governor­ship of ʿĪsa b. Yūnus (in office 227–29/ 841–44) found in the Fayyūm.

59

Figure 1.4. Unpublished Arabic letter mentioning land measurement in relation to land taxes.

60

Figure 1.5. Coptic request for information on workers and trees from the pagarch of Ihnās/ Heracleopolis that Rāshid b. Khālid issued in 718 ce.

63

Figure 3.1. Muʿāwiya’s dam near Ṭāʾif, 58 ah/677–78 ce.

174

Figure 4.1. Distribution of Palaeo-Arabic inscriptions.

198

8

l i s t of i l lustr ation s

Figure 4.2. P.Euphrate 18 (P.Euphr. inv. 19).

207

Figure 4.3. Codex parisino-petropolitanus, Hand E; 6:  76–91.

208

Figure 5.1. Arab-Hephthalite coin of Zhulād Gōzgān from Ambīr, British Museum. 69 ah/ 688 ce.

262

Figure 6.1. Coin of ʿAbd al-Malik minted in Dabīl/Duin. American Numismatic Society NS 1988.132.

288

Figure 6.2. Eighth-century Arabic inscriptions in the church of Zuart‘noc‘, after Xač‘atryan.

290

Figure 6.3. Eighth-century Arabic inscription on the church of Aruč.

291

Figure 6.4. Smbat Bagratuni on the walls of the monastery at Hałbat.

296

Figure 7.1. Map of the Hindu Kush and Northern Afghanistan, showing the places mentioned in the Bactrian documents.

308

Figure 7.2. Document J (517 ce), purchase contract.

310

Figure 7.3. Double document examples from Egypt.

312

Figure 7.4. Avroman Document II.

313

Figure 7.5. Document J sealings.

315

Figure 7.6. Document Uu (722 ce/104 ah), undertaking to keep the peace.

316

Figure 7.7. Letter from the Tabaristan Archive (Doc. Tab. 28; 8th century ce).

318

list of illustrations

Figure 7.8. Document Y (771 ce/154–55 ah), judicial declaration.

319

Figure 7.9. Document 31 (763 ce/146 ah), emancipation of a slave.

320

Figure 9.1. Namāra inscription.

368

Figure 9.2. Ḥimà-Sud PalAr 1.

371

Figure 9.3. Ḥimà-Sud PalAr 8.

372

Figure 10.1. BBAW/GCS, Akz.-Nr. 481/136, photo­graphed in 1901. Third/ninth century. 

404

Figure 10.2 BBAW/GCS, Akz.-Nr. 481/137, photo­graphed in 1901. Third/ninth century. 

405

Figure 10.3. BBAW/GCS, Akz.-Nr. 481/136, detail of line 11, left-hand folio — photo­graphed in 1901. 

406

Figure 10.4. BBAW/GCS, Akz.-Nr. 481/136, detail of line 9, left-hand folio — photo­graphed in 1901. 

407

Figure 10.5. Details from Berlin-Brandenburg Academy of Sciences and Humanities, BBAW/GCS, Akz.-Nr. 481/136–37.

409

Figure 10.6. BnF, MS gr. 497, fol. 1v, Arabic note in the upper margin.

414

9

Acknowledgements The editors would like to thank the faculty and staff at the University of Tennessee’s Marco Institute for Medi­eval and Renaissance Studies, especially Gregor Kalas, Jay Rubenstein, and Katie Hodges-Kluck, who provided essential logistical and financial support for the conference that led to this volume. Their assistance has been crucial to the completion and publication of the book. At Brepols, Guy Carney’s guidance and input has been invaluable. We are grateful for his patience. We would also thank the Department of Religious Studies, the Humanities Center, the Office of Research of the University of Tennessee, and the College of Arts and Sciences for the subvention that covered costs associated with the production of Navigating Language in the Early Islamic World. We also appreciate the additional support from Fesjian Publication Funds at Columbia University. Finally, we are indebted to Hanna Siurua for her meticulous copy-editing and ability to spot errors in multiple languages. We are deeply saddened by the sudden passing of Kate Mertes who had agreed to prepare the index to this volume. Kate’s extensive contributions to the field of indexing will be remembered, and we extend our condolences to her family and friends. The editors are profoundly thankful to Rebecca McCorkle for stepping in under such difficult circumstances.

Antoine Borrut, Manuela Ceballos, and  Alison M. Vacca

Introduction Multilingualism in the Early Islamic World

Traditional narratives of language change in the early Islamic period struggle to account systematically for the overwhelming quantity and variety of multilingual evidence provided by papyri, coins, and inscriptions. Without a proper integration of multilingualism, a deceptive sense of linear change emerges, feeding into a teleo­ logical search for the origins of an imagined exclusively Arabic-speaking Islamic world. However, the story of Arabicization is complicated and evades such easy categorization because of the immense diversity of the early Islamic world. The fate of Arabic was inextricably connected to the many languages of the Islamic world as well as to the variety of social, economic, political, and religious practices of the many communities living under the early Caliphate. If some communities chose to adopt and adapt Arabic as a political statement in the fashion of the Umayyads, others recognized Arabic first and foremost as a religious language or an economic tool. The reception of Arabic in the Umayyad and early ʿAbbasid Caliphates cannot be succinctly explained through neat and tidy stories. Historians of the early Islamic world have traditionally located Arabic at the heart of their scholarly enquiry, which has encouraged the growth of academic (philo­logical) silos. Although Persian and Turkish have also held positions of prestige, the marriage of Arabic, Persian, and Turkish divides the ‘Islamic’ languages from those of the non-Muslims who constituted the majority of the people living under the early Caliphate. The field of early Islamic history is changing to accommodate a new vision of the Umayyad and early ʿAbbasid periods by integrating Arabic with sources in Greek, Syriac, Coptic, Armenian, Bactrian, and Sogdian, among others. This shift is not about recognizing that Arabic cannot stand as the only language of Islamic history; generations of historians have turned to non-Arabic

Antoine Borrut is Associate Professor of History at the University of Maryland, College Park. Manuela Ceballos is Assistant Professor of Religious Studies at the University of Tennessee, Knoxville. Alison M. Vacca is Gevork M. Avedissian Associate Professor of Armenian History and Civilization at Columbia University. Navigating Language in the Early Islamic World: Multilingualism and Language Change in the First Cen­turies of Islam, ed. by Antoine Borrut, Manuela Ceballos, and Alison M. Vacca, ISMAR 2 (Turnhout: Brepols, 2024) pp. 13–40 10.1484/M.ISMAR-EB.5.134623

FHG

14

a n to in e bo r r u t, m an u e l a c e b al lo s, and  ali so n m. vacca

sources with fruitful lines of enquiry.1 Rather, the mental remapping concerns the relationship between sources in a variety of languages. Modern Islamicists now recognize that the many sources produced by non-Muslims in languages other than Arabic are in fact internal to Umayyad and ʿAbbasid history, rather than outsider perspectives serving to corroborate or complicate the Arabic.2 In other words, the lack of engagement with multilingualism has concrete implications with regard to the definition and articulation of corpora, not to mention historians’ ability to embrace new paradigms. The situation proves especially paradoxical in light of the integration of nascent Islam into a late antique framework,3 an area of study in which multilingualism has been a salient focus over the past few decades.4 And yet despite significant recent shifts, early Islam is still primarily conceived in Arabic.5 The cosmopolitan world of early Islam was multilingual, and as a result, the study of the early Islamic world must also accommodate multiple languages as a matter of course. The wide variety of existing sources poses a seemingly insurmountable challenge to modern historians’ attempts to tell the particularly unwieldy history of language change.6 This present volume contributes to the ongoing discussions about and against Arabicization in the broader context of language change by adopting the

1 There are too many such works for an exhaustive list, but of particular importance are Conrad, ‘Theophanes and the Arabic Historical Tradition’, p. 15; Conrad, ‘The Conquest of Arwād’; Hoyland, Seeing Islam as Others Saw It. 2 Borrut, Entre mémoire et pouvoir, pp. 137 ff. Although he originally resorted to non-Muslim sources to ‘step outside’ of the ‘Islamic tradition’, Hoyland eventually concluded that ‘this approach is not really valid, since the two bodies of material are much more intertwined than had previously been thought’. See his ‘Reflections on the Identity of the Arabian Conquerors of the Seventh-Century Middle East’, p. 78. Hoyland made his new methodo­logy clear in a recent book (In God’s Path, pp. 2–3): ‘I do not want to champion non-Muslim sources over Muslim sources; indeed, it is my argument that the division is a false one. Muslims and non-Muslims inhabited the same world, interacted with one another and even read one another’s writings. In this book the distinction I make is simply between earlier and later sources, and I favor the former over the latter irrespective of the religious affiliation of their author’. 3 Borrut, ‘An Islamic Late Antiquity?’. 4 The topic has generated a considerable amount of scholarship. See, inter alia, Smelik, Rabbis, Language and Translation in Late Antiquity; Mullen and James ed., Multilingualism in the GraecoRoman Worlds. 5 Egypt has been a notable exception, given the unique preservation of papyri; see, in particular, Papaconstantinou ed., The Multilingual Experience in Egypt, from the Ptolemies to the Abbasids. On the symbolic role of Arabic in the multilingual setting of early Islamic Egypt, see Garosi, ‘Imperial Arabic’; Papaconstantinou, ‘Arabic’; and the various contributions in Cotton and others eds, From Hellenism to Islam. 6 This challenge is not restricted to Islamic studies and has also been lamented by specialists of the medi­eval West, who have urged future generations to work ‘across different languages’ and to ‘understand the dissemination and use of different languages in the Middle Ages’ (Gaunt, ‘Can the Middle Ages Be Postcolonial?’, p. 172). This programmatic effort was taken up by Suzanne Conklin Akbari and Karla Mallette in an important volume, A Sea of Languages (see, in particular, the introduction by Akbari). Akbari and Mallette advocate for combining the ‘postcolonial, the cosmopolitan, and, above all, the philo­logical’ in the footsteps of the late María Rosa Menocal (d. 2012) and her pioneering work.

i nt ro d u ct i o n

specific lens of multilingualism. This introduction starts with an explanation of the many ways in which early Islamicists have thought about the practice of Arabicization over time, highlighting some of the publications that have dominated the conversation while acknowledging that scholars working on the late antique and early modern periods have fostered productive conversations about multilingualism that might inform future approaches to the early Islamic period. It then provides an overview of the three main nodes that connect the papers in this volume. The first section approaches the relationship between language change and Empire by examining policies of language management under the Umayyad and early ʿAbbasid Caliphates. The second section turns to the role of multilingual elites who navigated their positions within the Umayyad and early ʿAbbasid empires through the deliberate use of multiple languages. And finally, the third section explores the idea of languages in contact with papers investigating the intertwined histories of languages in multilingual spaces. The introduction closes by recognizing the lacunae in this volume as productive opportunities to continue the conversation about multilingualism and Arabicization in the early Islamic world.

Current Conversations about Arabicization This volume situates conversations about Arabic as a single — albeit important — thread in the broader story of multilingualism. We differentiate Arabization (an association with Arabness) from Arabicization (the spread and development of the language of Arabic).7 In proffering this definition, we envision Arabicization as part of a multidirectional process of language change at the hands of an immense variety of people across social, religious, ethnic, and political divides. Three main themes of Arabicization dominate traditional narratives of Islamic origins, and all of them have recently come under close reconsideration. The first one is linguistic and examines the emergence of the Arabic language itself and the subsequent formation of a discrete Arabic script.8 Scholars of early Islam and the Arabic language have recently developed current understandings of the history of Arabic, recognizing that the incredible diversity of pre- and early Islamic Arabic complicates any presumption of clear-cut narratives of language change.9





7 The benefits of distinguishing between ‘Arabization’ and ‘Arabicization’ have been recently stressed by Van Bladel, ‘Arabicization, Islamization, and the Colonies of the Conquerors’, p. 92 n. 13. Arabization has its own long and distinguished biblio­graphy. On Arabness, see Macdonald, ‘Arabs, Arabias, and Arabic before Late Antiquity’; Retsö, ‘The Earliest Arabs’; and Retsö, The Arabs in Antiquity. This last book should be read in light of Robin, ‘Les arabes vus de Ḥimyar’; Fisher, Arabs and Empires before Islam; Webb, Imagining the Arabs. See also the valuable discussion of Aillet, ‘Islamisation et arabisation dans le monde musulman médiéval’. 8 On the development of the Arabic script, see, most notably, Gruendler, The Development of the Arabic Scripts; Larcher, ‘Al-lugha al-fuṣḥâ’ and L’invention de la luġa al-fuṣḥā; Macdonald ed., The Development of Arabic as a Written Language, particularly the chapters by Macdonald (pp. 5–28) and Nehmé (pp. 47–88); Nehmé, ‘Aramaic or Arabic?’. 9 Al-Jallad, ‘What is Ancient North Arabian?’. See also Macdonald, ‘Reflections on the Linguistic Map

15

16

a n to in e bo r r u t, m an u e l a c e b al lo s, and  ali so n m. vacca

Figure 0.1. Muʿāwiya’s Greek inscription at Hammat Gader. From Whitcomb, ‘Notes for an Archaeo­logy of Mu’awiya’. Reproduced with permission.

This research is currently progressing at breakneck speed, as early Islamic Arabic inscriptions are discovered across Jordan and Arabia at a faster pace than scholarly publication can match. This situation means that scholars first see some inscriptions on social media. Such chaotic access further complicates grasping the burgeoning compilation of sources that forms the basis of modern understandings of the development of the Arabic language and script.10 The second theme in modern publications about Arabicization is political, focusing on Umayyad policies that allegedly made Arabic the official administrative language of the Caliphate. Whereas Muʿāwiya could identify himself in either Greek, Middle Persian, or Arabic on both stones and coins (see Fig. 0.1),11 everything seems to have radically changed with ʿAbd al-Malik in the context of the Marwanid restoration after the second fitna (60–72/680–92), as illustrated by the new style of Pre-Islamic Arabia’; Stein, ‘The “Himyaritic” Language in Pre-Islamic Yemen’; Iema, Al-Jallad, Macdonald, and Nehmé, ‘Provincia Arabia’, esp. pp. 395 ff. 10 See, in this volume, the contributions of Muriel Debié and Aaron Butts. See also Anthony, ‘Early Arabo-Islamic Epi­graphy and the Positivist Fallacy’; al-Ghabbān and Hoyland, ‘The Inscription of Zuhayr, the Oldest Islamic Inscription’; Al-Jallad and Sidky, ‘A Paleo-Arabic Inscription on a Route North of Ṭāʾif ’’; Hoyland, ‘Two New Arabic Inscriptions’; Imbert, ‘L’islam des pierres’; Imbert, ‘Le Coran des pierres’; Imbert, ‘Graffiti arabes de Cnide et de Kos’; Robin, al-Ghabbān, and al-Saʿīd, ‘Inscriptions antiques de la region de Najrān’; Robin, Najrān en Arabie et ses martyrs chrétiens; al-Shdaifat, Al-Jallad, al-Salameen, and Harahsheh, ‘An Early Christian Arabic Graffito Mentioning “Yazīd the King”’. 11 Hirschfeld and Solar, ‘The Roman Thermae at Ḥammat Gader’; Miles, ‘Early Islamic Inscriptions near Ṭāʾif in the Ḥijāz’; al-Rāshid, ‘Sadd al-Khanaq’; Walker, A Catalogue of the Muhammadan Coins in the British Museum, pp. 25–26.

i nt ro d u ct i o n Figure 0.2. Latin solidus minted in Ifrīqiya during the reign of al-Walīd b. ʿAbd al-Malik after the supposed language reforms of his father. The David Collection, Copenhagen. Photo by Pernille Klemp. Reproduced with permission.12

of coins and milestones. In Chase Robinson’s words, ‘the Marwanids were linguistic imperialists’ and it is because of them ‘that Arabic became the lingua franca of the Middle East, the speed of this change depending much on geo­graphy’.13 The traditional vision of a straightforward, top-down Arabicization implemented by the Marwanids has also come under sustained scrutiny. Documents from the Umayyad chancery in Egypt tell quite a different story and suggest that the Umayyads deliberately facilitated a multilingual administration.14 This argument is supported by the continued use of other languages on Umayyad and even ʿAbbasid coins produced well after ʿAbd al-Malik’s purported policies of Arabicization, such as the Latin solidi struck in Ifrīqiya (Fig. 0.2) and the bilingual Arabic-Latin dinars of Umayyad Andalus. In the Iranian sphere, coins were minted with Middle Persian inscriptions into the ʿAbbasid period, well past the Umayyad Arabicization reforms.15 The third major theme of Arabicization in the narratives of early Islam is both political and literary: the famous ‘translation movement’, during which texts in Greek and Middle Persian made their way into Arabic. Dimitri Gutas understood the translation movement as a strictly ʿAbbasid-era phenomenon and argued that it would have been simply unthinkable in Umayyad times. He thus concluded that ‘had the ʿAbbasid dynasty not come into power and moved the capital to Baghdad, there would have been no Graeco-Arabic translation movement in Damascus’.16 In Baghdad, the ʿAbbasids’ programme Hellenized the Arabs but also Arabicized Greek and thus rendered Greek unnecessary.17 Like the first two themes, this vision of the translation movement has been contested. The documents of the Qubbat al-Khazna found in Damascus demonstrate that Greek did not taper off immediately despite the spread of Arabic.18 George Saliba has made a strong case for the Umayyad-era 12 This coin reads in abbreviated form ‘DeuS ETeRNuS DeuS MaGNuS DeuS OmnIum Deus’ (God the Eternal, God the Mighty, God the Omniscient) along the margin, and ‘SOMNium Creator’ (God the Creator of all) in the centre. On the reverse, it reads in abbreviated form ‘IN Nomine DomINi MiSeRiCordis SoLidus FERitus IN AFRiCa’ (in the name of the Lord, the Merciful, solidus made in Africa). 13 Robinson, ʿAbd al-Malik, p. 126. 14 See the contributions by Marie Legendre and Petra Sijpesteijn in this volume. 15 Malek, The Dābūyid Ispahbads and Early ʿAbbāsid Governors of Tabaristān. 16 Gutas, Greek Thought, Arabic Culture, pp. 18–19. 17 Wasserstein, ‘Why Did Arabic Succeed Where Greek Failed?’. 18 D’Ottone Rambach, Hirschler, and Vollandt ed., The Damascus Fragments; D’Ottone, ‘Manu­scripts as Mirrors of a Multilingual and Multicultural Society’. See also Antoine Borrut’s chapter in this volume.

17

18

a n to in e bo r r u t, m an u e l a c e b al lo s, and  ali so n m. vacca

beginnings of the translation movement, arguing that a translation boom in early ʿAbbasid Baghdad required a pre-existing scholarly milieu.19 Moreover, the depictions of several of the Greek muses (philosophy, history, and poetry) on the walls of Quṣayr ʿAmra, as well as the actual Greek inscriptions on the Six Kings mural, confirm an interest in pre-existing intellectual traditions and Hellenism in particular, not to mention the practice of a ‘thoughtful bilingualism’ (bilinguisme réfléchi).20 This view is further supported by the voluminous quantity of reports associating the Umayyad prince Khālid b. Yazīd with pre-Islamic scientific traditions and depicting him as the patron of some of the earliest translations into Arabic.21 All these elements invite us to rethink Gutas’s premises, to seriously entertain the possibility of Umayyad-era translations, and to reconsider cultural life under the first dynasty of Islam more broadly.22 The rationale for nesting the study of Arabicization and language change in the early Islamic Near East into the context of multilingualism stems in part from the interplay of three influential publications. In 2003, David Wasserstein published an article entitled ‘Why Did Arabic Succeed Where Greek Failed? Language Change in the Near East after Muhammad’. In this article, taking the documents found at Nessana as the basis,23 Wasserstein explores the implications of a (probably) Arabic- and Syriac-speaking community that produced documents in Greek and traces how Arabic began to supplant Greek from the Umayyad period onwards. He places this community into the broader context of Arabicization, which unfurled differently in the various provinces of the Islamic world: Arabic spread largely unchallenged in ex-Byzantine territories, where it was adopted by people of all classes as opposed to the ‘fairly thin ruling crust’ that had embraced Greek before it. Wasserstein associates the shift to Arabic-speaking masses with the development of an Islamic world traversed by travellers who relied on the ubiquity of a shared common language, as well as with the settlement of Arabs across the Near East both before and after the rise of Islam. After the foundation of Baghdad — which he 19 Saliba, Islamic Science and the Making of the European Renaissance. An ʿAbbasid chrono­logy for the translation movement has endured in most scholarship, even when Saliba’s points are acknowledged; see, for instance, Yücesoy, ‘Translation as Self-Consciousness’. See also Morrison, ‘Islamic Astronomy’, pp. 116–17. 20 Imbert, ‘Le prince al-Walīd et son bain’, p. 347; Fowden, Quṣayr ʿAmra, esp. chs 9 and 10. We should also keep in mind the importance of the Sasanian elements in Quṣayr ʿAmra, which serve as a clear reminder not to envision the Umayyads as Byzantium reinterpreted. 21 Modern scholarship has been highly dismissive of Khālid b. Yazīd in the footsteps of Ruska, Arabische Alchemisten and especially Ullmann, ‘Ḫālid ibn Yazīd und die Alchemie’. See, however, Saliba, Islamic Science and the Making of the European Renaissance, pp. 45–51; Hoyland, Seeing Islam as Others Saw It, pp. 233–34. A more substantial discussion on Khālid will be included in Borrut’s forthcoming book dedicated to early Islamic-era astro­logical histories and tentatively entitled Aligning Heaven and Earth: Astrology and History in Early Islam. 22 A point already made in the late 1930s in Mackensen, ‘Supplementary Notes to “Arabic Books and Libraries in the Umayyad Period”’. See also Eche, Les bibliothèques arabes publiques et semi-publiques, pp. 11–20. 23 The Nessana documents have prompted an interesting set of relevant discussions; see notably Stroumsa, ‘Greek and Arabic in Nessana’.

i nt ro d u ct i o n

presents as a resoundingly Arab city where Arabic predominated — the translation movement continued the transition from Greek to Arabic.24 Wasserstein’s article inspired a number of queries and responses,25 three of which deserve particular attention for the purposes of this volume. Robert Hoyland identified his 2004 article ‘Language and Identity: The Twin Histories of Arabic and Aramaic (and: Why Did Aramaic Succeed Where Greek Failed?)’ as a ‘footnote’ to Wasserstein’s article. In it, Hoyland explores the interplay of Aramaic and Arabic over the long course of history from the third century ce into the early Islamic period. He follows the development of Arabic to its role as Reichssprache and Kultursprache (as Wasserstein did), adding its significance as Religionssprache. Hoyland points out that in this context, the question should not be solely about the failure of Greek and the success of Arabic but should rather acknowledge that the story of Aramaic was intertwined with both. He suggests that the survival of Aramaic was due to a number of factors, among which were its ubiquity as a spoken language (as opposed to a learned language such as Greek), the relations between Syrian Christians and the early Muslims, and Christian participation in the cultural and political activities of the Caliphate, such as the translation movement.26 Shifting the focus to Egypt, Arietta Papaconstantinou responded to both Wasserstein’s and Hoyland’s articles in her 2012 chapter ‘Why Did Coptic Fail Where Aramaic Succeeded? Linguistic Developments in Egypt and the Near East after the Arab Conquest’. She begins by questioning what it means for a language to ‘fail’, comparing the case of Aramaic with that of Coptic. She reviews the economic and cultural reasons proposed in modern literature for the demise of Coptic but suggests that the comparison with Aramaic offers a fruitful way forward. Coptic claimed some of the same attributes that Hoyland ascribes to Aramaic, but Papaconstantinou draws attention to the difference between institutional uses of language for initiatives such as the translation movement and a language’s common acceptance by the masses. She identifies a number of possible explanations for the Arabicization, including the social mobility of Christian administrators and the economic opportunities offered to enterprising Christians. Manu­script evidence points to the dual relevance of Coptic and Arabic in the Coptic Church and in private correspondence.27 Most recently, Kevin van Bladel criticized the lack of convincing explanations for language shift in early Islam, rejecting in particular the notion that ‘language shift is caused by changes in identity’. Instead, he forcefully argued that ‘the most basic and important factor in the shift of a population’s use of one language to another is contact between speakers of the two languages’.28 Such a focus on social and demo­graphic factors led him to insist on the centrality of the amṣār, and on

24 25 26 27 28

Wasserstein, ‘Why Did Arabic Succeed Where Greek Failed?’. See, for example, the chapter by Aaron Butts in this volume. Hoyland, ‘Language and Identity’. Papaconstantinou, ‘Why Did Coptic Fail Where Aramaic Succeeded?’. Van Bladel, ‘Arabicization, Islamization, and the Colonies of the Conquerors’, pp. 96–97.

19

20

a n to in e bo r r u t, m an u e l a c e b al lo s, and  ali so n m. vacca

settlement and migration patterns more broadly, to document the asynchronous rhythms of language change throughout the huge spaces inherited from the conquests. Read together, the articles by Wasserstein, Hoyland, Papaconstantinou, and van Bladel demonstrate the many complicated factors that contributed to the process of Arabicization in various regions of the early Islamic world and, in particular, the regional specificities that informed language change. Language change defies easy summaries because of the variety of case studies, each with details specific to a particular time or place that render it resistant to the blanket generalization of its conclusions. This realization is particularly important given that Arabic emerged as a Reichssprache, Kultursprache, and Religionssprache in the Iranian cultural sphere but never took hold in the vast majority of Iranian communities, even if the Arabic script was adopted and selectively adapted. Languages such as Persian, Sogdian, Bactrian, Syriac, Armenian, and Georgian continued apace alongside Arabic in the early Islamic period. Building on this set of conversations, then, the goal of the present volume is not to overturn the three themes of extant scholarly literature on Arabicization outlined above but to expand on them by embracing the complicated range of case studies in order to place the process of language change within a broader context of multilingualism in the early Islamic world. Approaching early Islam more systematically through the lens of multilingualism is long overdue, especially if one recognizes that such a multilingual setting has been taken for granted in other disciplines. Looking westward, Islamicists might consider the productive conversations generated in the study of the Mediterranean. Two particularly significant interventions, which tell the history of Arabic and Latin in a comparative historical framework, deserve brief mention here. Although both draw on the work of linguists, they situate themselves in the broader setting of social and intellectual history. In 2012 Benoît Grévin published Le parchemin des cieux: essai sur le Moyen Âge du langage, which explores the ways in which medi­eval thinkers conceptualized language, focusing specifically on Latin and Arabic and deeply intertwined intellectual histories of those two languages. He proposes three registers of language: prestigious and privileged langues référentielles (including his main examples, Latin and Arabic — both saturated with religious meaning), elite langues courtoises (such as French and Persian), and popular langues vulgaires. These distinctions allow him to complicate simplistic summaries of language use in favour of recognizing the distinct realms in which people chose to prioritize one language over another (for example, in teaching or giving sermons). Grévin also considers the idea of a langue véhiculaire, which serves as a connective lingua franca.29 Karla Mallette has recently continued the study of Arabic and Latin in concert in her Lives of the Great Languages: Arabic and Latin in the Medi­eval Mediterranean, published in 2021. She begins with a provocative question: ‘What bait did the cosmopolitan languages dangle to lure generations of writers away from the bosom of the mother tongue?’. By means of well-chosen examples, Mallette investigates how and why

29 Grévin, Le parchemin des cieux. On lingua francas in the Mediterranean more broadly, see Wansbrough, Lingua Franca in the Mediterranean and Dakhlia, Lingua franca.

i nt ro d u ct i o n

individuals chose to write in a cosmopolitan language, revealing the cultural settings that encouraged or challenged the use of Arabic and Latin.30 Both of these books offer inventive and compelling ways to grapple with the interplay of the enormous stories of Arabic and Latin; both were composed by scholars more familiar with European than Islamic history, throwing down the gauntlet to Islamicists to respond.31 Another model to inspire and inform early Islamicists in the study of multilingualism and language is to look forward chrono­logically. Since later periods fall outside of the remit of this volume, suffice it to say that the Mongol period clearly stands out in light of the Rasulid Hexaglot, a glossary produced in fourteenth-century Arabia that employed the Arabic script to represent vocabulary in Persian, Turkic, Byzantine Greek, Armenian, and Mongol; the Hexaglot thus demonstrates the networks that crossed Eurasia under the Mongols, mirroring the exchanges on the political and cultural levels.32 One could also contemplate the case of the Ilkhanid ruler Ghāzān Khān (r. 694–703/1295–1304), who was credited by Rashīd al-Dīn (d. 718/1318) with formidable linguistic skills ranging from his native Mongolian tongue to (at least some level of ) Arabic, Persian, Hindi, Kashmiri, Tibetan, ‘Khitāʾī’,33 Frankish,34 and other languages.35 The Mongol chancery was also famously multilingual,36 as evidenced not only by written accounts of the Mongols’ rule but also by material evidence such as an Arabic-Uyghur coin minted in Tiflīs (Fig. 0.3). The explicitly Christian message of the bilingual coin pairs neatly with the mix of inscriptions in Mongolian, Syriac, Armenian, and Arabic in the monastery of Mar Behnam near Mosul. Ethel Sara Wolper has framed the study of the monastery as a practice in translation, referring to the ways in which Muslim and Christian

30 Mallette, Lives of the Great Languages. 31 Other fruitful studies fostered in European history include Classen ed., Multilingualism in the Middle Ages and Early Modern Age; Kappler and Thiolier-Méjean ed., Le plurilinguisme au moyen âge; Tornow, Abendland und Morgenland im Spiegel ihrer Sprachen. 32 Golden ed., The King’s Dictionary, p. 1. See also Golden, ‘The Byzantine Greek Elements in the Rasulid Hexaglot’. 33 Most likely Kitan, the language of the Kitai people whose name was rendered by Muslims as Khitā or Khatā, a term used to refer to northern China (hence the name Cathay, which is how the region was known to the medi­eval West). Kitan remains a largely undeciphered ‘para-mongolic’, ‘Altaic’ language, on which see Kane, The Kitan Language and Script. We see no reason to translate Khitāʾī as ‘Chinese’, as done by Boyle in ‘Dynastic and Political History of the Il-Khāns’, p. 397, and again in his entry ‘G̱ ẖāzān’. 34 On the perception of Romance languages in the Levant, see Aslanov, ‘Quand les langues romanes se confondent’. 35 Rashīd al-Dīn, Jāmiʿ al-tawārīkh, pp. 1337–38; Golden, The King’s Dictionary, p. 5. Some scholars have expressed scepticism at Rashīd al-Dīn’s ‘implausibly long list of Ghāzān’s talents’ ( Jackson, ‘Ghāzān Khān Maḥmūd’). 36 Golden, The King’s Dictionary, p. 5, citing Juvaynī’s Taʾrīkh-i Jahān-gushā, p. 607: ‘They are attended by scribes of every kind for Persian, Uighur, Khitayan, Tibetan, Tangut, etc., so that to whatever place a decree has to be written it may be issued in the language and script of that people’. The Mamluk chancery was equally multilingual; see, for instance, Bauden, ‘Les relations diplomatiques entre les sultans mamlouks circassiens et les autres pouvoirs du Dār al-islām’; Bauden and Dekkiche ed., Mamluk Cairo, a Crossroads for Embassies.

21

22

a n to in e bo r r u t, m an u e l a c e b al lo s, and  ali so n m. vacca

Figure 0.3. Bilingual Arabic-Uyghur coin minted under Abaqa b. Hulegu. The David Collection, Copenhagen. Photo by Pernille Klemp. Reproduced with permission.

beliefs found common expressions in a shared space where the Mongols sought to inscribe themselves as guardians.37 Research on language change and exchange in the early modern period has also addressed questions of translation, the learning of languages, and the relationship between languages and peoples through the lens of multilingualism. This approach has often led to a re-evaluation of traditional narratives about the history of languages in particular regions. For example, Claire Gilbert has argued that despite anti-Arabic legislation and persecution of Arabic-speaking minorities, Arabic played a crucial cultural, social, and administrative role in early modern Spain.38 Helen Pfeifer has explored the centrality of literary salons (majālis) as venues for the interaction of elites across linguistic divides in the Mamluk Sultanate and the Ottoman empire during the sixteenth century.39 Kristina Richardson has produced evidence of sixteenth-century Ottoman Turkish and Ottoman Arabic sign systems.40 These are but a few recent works, all of which focus on the relationships between Arabic and other languages, but the topic has been (and will likely continue to be) a fertile area of study. Much of this recent scholarship is, in turn, indebted to the booming study of cosmopolitanism, a term that has multiple genealogies and definitions.41 Despite the conceptual difficulties of definitively describing cosmopolitanism as a concept and a practice, it has become an important theoretical perspective in Islamic studies, drawing on the work of scholars such as Dipesh Chakrabarty, Carol Breckenridge, Homi Bhabha, and Sheldon Pollock, Kwame Anthony Appiah, and Bruce Lawrence, particularly (though not necessarily) in studies focusing on

37 Wolper, ‘Khidr and the Politics of Translation in Mosul’. 38 Gilbert, In Good Faith. 39 Pfeifer, ‘Encounter after the Conquest’ and Empire of Salons; Mauder, In the Sultan’s Salon. 40 Richardson, ‘New Evidence for Early Modern Ottoman Arabic and Turkish Sign Systems’. 41 Pollock, Bhabha, Breckenridge, and Chakrabarty, ‘Cosmopolitanisms’, p. 1.

i nt ro d u ct i o n

the contemporary world.42 In premodern contexts, perhaps the most significant theoretical intervention has been Pollock’s on the ‘Sanskrit cosmopolis’.43 Clearly, the rich scholarship that could be only briefly addressed above suggests the formidable untapped potential of multilingualism and cosmopolitanism to shed fresh light on nascent Islam and the early Caliphate. And on top of new theoretical approaches, what the narratives of Arabicization and of the translation movement discussed above have in common is that they suggest deliberate cultural choices, reflecting state- and society-sponsored imperial policies, that deserve further investigation and new interpretative frameworks.

Navigating Language: The Interventions of this Volume This volume collects papers about multilingualism, language change, and the role of Arabic in the early Islamic world. It is the result of a symposium, ‘Navigating Language in the Early Islamic World’, held at the University of Tennessee in 2018, hosted by the Marco Institute for Medi­eval and Renaissance Studies. The participants were historians and art historians, rather than philo­logists or linguists. The goal of the symposium was to explore the social and cultural history of language in a multidisciplinary conversation that bridged the Mediterranean region and the Iranian cultural sphere. Our enquiries also addressed the ways in which language change reflected processes that predated Islam, in effect asking whether the rise of Islam changed the spread and restrictions of languages in territories formerly part of the Roman or Sasanian Empires. In response to the discussions in Knoxville, we invited several others to submit chapters to extend the conversation into areas that were not covered in the symposium itself. Three permeable thematic nodes emerged, each building on the others. The first section of this volume, entitled ‘Languages of the Umayyad Caliphate’, collates papers about Umayyad imperial strategies of language change. In the first chapter, Petra Sijpesteijn reads documents produced from Egypt to Khurasan against ʿAbbasid-era Arabic histories to argue that the Umayyads continued to employ languages other than Arabic well past the period associated with ʿAbd al-Malik’s reforms. She contends that the Umayyads ran a multilingual administration for both practical and ideo­logical reasons, focusing in part on the central role of scribal education in determining language change over time. In the next chapter, Marie

42 Breckenridge, Pollock, Bhabha, and Chakrabarty ed., Cosmopolitanism; Appiah, Cosmopolitanism; Lawrence, Islamic Cosmopolitan Spirit. A recent example of the influence of Appiah’s version of cosmopolitanism in Islamic art across time and space is Belli Bose ed., Intersections. 43 Pollock, The Language of the Gods in the World of Men. For the ways in which Pollock’s intervention informs the study of early Islam, see the chapter by Borrut in this volume. Pollock’s magnum opus has also proved particularly influential for East Asianists, who have coined the concept of a ‘Sino­graphic cosmopolis’; see King, ‘Introduction’, p. 2, and the volumes in Brill’s book series Language, Writing and Literary Culture in the Sino­graphic Cosmopolis, edited by Ross King, David Lurie, and Marion Eggert.

23

24

a n to in e bo r r u t, m an u e l a c e b al lo s, and  ali so n m. vacca

Legendre systematically and exhaustively tallies the accounts of the translation of the Umayyad dīwān to dismantle the idea of a single, sweeping language reform ordered by the caliph. By delving into the specific details of how these authors perceived the dīwān — including the technical skills associated with the work of its translation — Legendre locates the diffuse process of Arabicization firmly in the hands of the scribes rather than those of the caliph. Finally, Antoine Borrut’s chapter offers a theoretical prism for understanding imperial strategies on language. Building on recent scholarship on cosmopolitanism, he postulates the emergence of an early Islamic-era ‘Arabic cosmopolis’. He also sheds light on the tensions between competing strategies that reveal the maintenance of cultural boundaries through language management in early Islam. Borrut thus provide a framework for thinking about the articulation of language, culture, and power and for analysing the ‘politics of cultural difference’.44 Taken together, these contributions force us to rethink the agreed-upon narrative of Arabicization to prioritize a resolutely multilingual and multiscriptural scenario. After the first section has established the multilingual and cosmopolitan nature of the early caliphal administration and clearly located the role of Arabic in this system, the second section, ‘Multilingualism, Empires, and Local Elites’, turns to the ways in which local elites across the caliphate participated in the process of language change. Muriel Debié’s chapter begins with an explanation of the development of Arabic — and the Arabic script, in particular — out of multilingual Arabia and Syria, suggesting connections to Syriac. Relying on references in written works and inscriptions, she assesses the familiarity of various languages in Arabia in Late Antiquity. Arabs actively participated in and negotiated imperial rule in multiple languages in the pre-Islamic period. By exploring multilingual contacts in armies and church hierarchies, for example, Debié shows that Arabs were integral players in the multilingual written (not just oral) world of Late Antiquity. Next, Robert Haug’s chapter investigates the coins minted along the eastern frontier in Arabic, Bactrian, and Middle Persian during the second fitna, coins that carved out claims in the interstices of both Empire and, perhaps more practically, circulation zones. The use of Arabic on these coins could be interpreted to signal its prestige as the language of Empire, rather than to demonstrate caliphal control over the regions where the coins were minted. By combining the languages of the Umayyad Caliphate and the Sasanian Empire into the same (very small) space, the Hephthalite ruler catapulted Bactrian to the status of a language of Empire. The following chapter, by Alison Vacca, similarly examines the appropriation of Arabic as a political symbol by non-Arabs on the edges of the Caliphate. Focusing on Armenia, her chapter reviews the evidence for the use of Arabic in the multilingual province and concludes that the choice to employ Arabic in certain circumstances reflected the language’s status as a symbol of political power: after Armenia gained its independence in the ninth century ce, Arabic became a public signifier of the prestige of Armenian royalty. The final paper in this section is Judith A. Lerner’s 44 Lavan, Payne, and Weisweiler, ‘Cosmopolitan Politics’, p. 9.

i nt ro d u ct i o n

examination of seals on Bactrian and Arabic documents from Greater Khurasan. She introduces the forms of documents and sealing practices from ancient times and uses these as the backdrop to reveal the new protocols and seal designs of early Islamic Khurasan. Lerner also outlines the changes from Hellenistic practices to Sasanian formats and images, which were brought to Khurasan via Iran by Muslim elites. Like the scribes discussed in the first section of this volume, the elites of the second section demonstrate the variety of ways in which language choices responded to the needs and goals of local elites as they negotiated their own positions in relation to the Caliphate. The third and final section of this volume, ‘Languages in Contact and Shared Spaces’, does not wholly step away from questions of power but grapples more directly with the encounters between languages on the page, in the markets, and at work. Khodadad Rezakhani’s chapter bridges sections two and three by addressing both the story of local elites and the concurrent development of Middle Iranian languages. He investigates the story of New Persian as the language of the Samanid court while tracing how the languages of the East (Bactrian and Sogdian) have informed its development. In doing so, he disputes the presumption of a linear evolution from Middle Persian to New Persian and proposes a rival model of two contemporary languages developing along different paths. In the process, he also integrates the study of Parthian into that of both Middle Persian and New Persian. The next chapter, by Aaron Butts, similarly challenges the idea of unidirectional language change by complicating the narrative about the process through which Syriac-speaking Christians in the Near East adopted Arabic. He traces the use of Arabic among Christian communities before Islam and their continued use of Syriac well after the rise of Islam. Butts delves the use of Arabic in al-Ḥīra and the story of Ḥunayn b. Isḥāq, in particular, to demonstrate that multiple forms of both Arabic and Syriac were current in this period, further problematizing the misconception of an easy ‘transition’ from Syriac to Arabic. The third chapter in this section is Arianna D’Ottone Rambach’s investigation into two Christian manu­scripts. The first is a Christian Arabic manu­script from the Qubbat al-Khazna in Damascus in which the Arabic letter forms reveal similarity to the Hebrew script. This discovery undermines simplistic assumptions about the cultural and linguistic context of scribal education and, consequently, monasteries as a site of manu­script production. The second example is a Greek homiliary with an Arabic marginal note that illustrates multilingual transfer across the Mediterranean. Finally, the volume turns to the participation of Jews in the flourishing book culture of ʿAbbasid Baghdad. Fred Astren situates early ʿAbbasid-era Jewish sectarianism in the triple context of urbanization, economic growth, and Arabic literary production. Especially in Baghdad, Jews had access to a burgeoning collection of books and conversations as well as linguistic and literary techniques that transcended confessional divides and helped shape new ways of reading. Appositely, newly emerging Jewish readerships adopted novel approaches to interpreting ancient Hebrew and to writing contemporaneous Hebrew and Arabic. The consequence was new ways of Jewish belonging that mirrored the contentious boundary-making and identity formation of the Muslim milieu. The four chapters of this section thus provide a sense of the

25

26

a n to in e bo r r u t, m an u e l a c e b al lo s, and  ali so n m. vacca

complicated interactions of Arabic, Hebrew, Persian, and Syriac, languages that inhabited the same communities and developed in dialogue with one another. The papers in this volume make no claim to comprehensive coverage of language change and multilingualism or the place of Arabic in relation to those fields during the early Islamic period. Rather, this volume coalesced around the three aforementioned themes as part of the first round of conversations. It makes sense to begin with the acknowledgement that several themes, provinces, and languages remain outside the scope of this volume but nevertheless should be part of the conversation moving forward. Clear lacunae include languages with their own established traditions of study, such as Tamazight45 and Latin,46 and although Coptic47 and Greek48 appear in several of the papers here, there is significantly more work to be done on both. Furthermore, provinces such as al-Andalus49 and Sicily,50 and the ‘Islamic West’ more broadly,51 deserve more attention than space allows in a single volume. Above and beyond these concerns, three thematic areas emerge as particularly important for further work: religion, trade, and identity. The very idea of multilingualism carries religious weight. The story of Babel in the Hebrew Bible confers negative connotations on language change, presenting the cacophony of many languages as something that must be overcome through miraculous multilingualism rather than human ingenuity.52 However, Christians

45 On Tamazight in the early Islamic world, see Kossmann, The Arabic Influence on Northern Berber, particularly ch. 3 on the pre-Islamic and early Islamic periods; Meouak, La langue berbère au Maghreb médiéval. 46 On Latin in the early Islamic world, see the previously cited Grévin, Le parchemin des cieux; Mallette, Lives of the Great Languages, as well as Aillet, Les Mozarabes; Internullo and D’Ottone Rambach, ‘One Script for Two Languages’; König, Latin and Arabic; König, ‘The Unkempt Heritage’; Lancel, ‘La fin et la survie de la latinité en Afrique du Nord’. 47 On Coptic in the early Islamic world, see Papaconstantinou’s ‘Why Did Coptic Fail Where Aramaic Succeeded?’, discussed above, as well as Boud’hors, ‘Degrés d’arabisation’; Papaconstantinou, ‘“They Shall Speak the Arabic Language and Take Pride in It”’; Richter, ‘Greek, Coptic and the “Language of the Hijra”’. 48 On Greek in the early Islamic world, see Johnson, Languages and Cultures of Eastern Christianity: Greek; Mango, ‘Greek Culture in Palestine after the Arab Conquest’; Mavroudi, ‘Greek Language and Education under Early Islam’. 49 On multilingualism in al-Andalus, see Hamilton and Silleras-Fernandez eds, Iberian Babel: Translation and Multilingualism in the Medi­eval and the Early Modern Mediterranean. On linguistic diversity and/ or Arabic in al-Andalus, see some of the sources under Latin above, as well as part two of Akbari and Mallette, A Sea of Languages, but also Aillet, Les Mozarabes; Brann ed., Languages of Power in Medi­eval Spain; Gallego, ‘The Languages of Medi­eval Iberia and their Religious Dimension’; Miller, Jewish Multiglossia; Navarro García and Sarr ed., Arabización, islamización y resistencias en al-Andalus y el Maghreb; Wasserstein, ‘The Language Situation in al-Andalus’. 50 Dolezalek, Arabic Script on Christian Kings; Johns, Arabic Administration in Norman Sicily; Jamil and Johns, ‘Signs of the Times’; Mallette, The Kingdom of Sicily, 1100–1250; Metcalfe, Muslims and Christians in Norman Sicily; Miller, ‘Muslim Poets under a Christian King’; Nef, Conquérir et gouverner la Sicile islamique aux xie et xiie siècles. 51 See for instance the relevant contributions in Valérian ed., Islamisation et arabisation de l’Occident musulman médiéval (viie–xiie s.). 52 Papaconstantinou, ‘Introduction’. On the late reception of the Babel motif in Europe, see Grévin, Le parchemin des cieux, 13–14 n. 121.

i nt ro d u ct i o n

living in the early Islamic world selectively recast the proliferation of languages as it suited their purposes. An eighth-century Christian polemic conveyed in the guise of a letter from the Roman emperor Leo the Isaurian to the Umayyad caliph ʿUmar b. ʿAbd al-ʿAzīz constructs linguistic diversity as a way of protecting against the falsification of scripture: Եւ արդ ըստ քո ասացուածիդ գրէ թէ միոյ կամ զերկուց ազգի փոխէր ոք զգրեանն արդեւք. իսկ զիարդ և զայլոց ևս ազգաց զորս յոյժ հեռաւորս գիտէք ի մէնջ և ի միմեանց աւտար լեզուաւք և սովորութեամբք [47r].

And now, as you said, maybe someone in one or two of these peoples really changed the Scriptures, but how do you know about other peoples, who [are] very distant from us and with languages and customs that are foreign to each other?53 In other words, scripture could not be successfully falsified because no one could corrupt the texts in all of the languages of the world. This argument recasts the negative connotations of the Babel story in an effort to refute Muslim accusations of taḥrīf. A broader study of the religious connotations of language change would track such varied interpretations of multilingualism. This volume also does not venture into the role of Arabic as lingua sacra in the context of a multilingual Islamic world. The tremendous significance of Arabic in Islamic circles seems largely self-evident, given the famous claim that the Qurʾānic revelation descended in ‘clear Arabic language’ (‫ ;وهذا لسان عربي مبين‬Q 16:103).54 However, recent scholarship has underscored the various ways in which Muslims translated both the Qurʾān and, more broadly, their beliefs into many languages across the Islamic world. The tenth-century Sogdian author Narshakhī famously tells of Umayyad-era Muslims teaching converts how to pray in both Persian and Sogdian: ‫و مردمان بخارا به اول اسالم در نماز قرآن بپارسی خواندندی و عربی نتوانستندی آموختن و‬ ‫ مردی بود در پس ایشان که بانگ {زدی} بکنیتا نکینت و چون سجده‬،‫چون وفت رکوع شدی‬ 55.‫خواستندی {کردی} بانک کردی نکونیا نکونی‬

And the people of Bukhara, at the beginning of [their conversion to] Islam, recited the Quranic [verses] in their prayers in Persian and could not learn Arabic, and when it was time for rukūʿ, there was a man behind them who would shout ‘BKNITA NKINT’, and when they wanted to do sujūd, [he] would shout ‘NKUNYA NKUNY’.

53 La Porta and Vacca, An Armenian Futūḥ Narrative. 54 On the difficulty of interpreting this famous Qurʾānic expression, see Gilliot and Larcher, ‘Language and Style’, pp. 113–15. 55 On this passage and the complications on rendering it into English, see Rezakhani in this volume, whose translation we follow here.

27

28

a n to in e bo r r u t, m an u e l a c e b al lo s, and  ali so n m. vacca

Figure 0.4. Bilingual Arabic-Sanskrit coin minted under Maḥmūd of Ghazna, showing the shahāda in Arabic and its loose Sanskrit translation. The David Collection, Copenhagen. Photo by Pernille Klemp. Reproduced with permission.

Narshakhī’s text, though written in the tenth century ce, was heavily revised in subsequent centuries, so this tale is quite late to narrate Umayyad history. The extant interlinear Arabic-Persian codices of the Qurʾān similarly date from a later period, the earliest from the eleventh century.56 However, Travis Zadeh has convincingly demonstrated that early Muslims accommodated the translation of scripture without denying the inimitability of the text.57 To ensure the transmission of Islamic beliefs, the tenets, at least, had to be accessible in multiple languages. The Umayyads actively participated in the public display of Islamic beliefs in languages other than Arabic as a way of expressing their own legitimacy as caliphs, as we saw above with the coin displaying the statement of faith in Latin; we also find that Zubayrid coins proclaimed the statement of faith in Middle Persian and, much later, Ghaznavid coins did the same in Sanskrit (Fig. 0.4).58 Furthermore, the story of the role of Arabic in religious settings or in the making of religious claims must account not only for the linguistic diversity of the early Islamic world but also for confessional diversity in an empire in which Muslims were long the minority. The use of Arabic proliferated among Christians from an early period, not just in daily life but in religious contexts, as well.59 It is important to note that the Arabicization of Christian communities was largely restricted to those in the formerly Byzantine provinces of Syria, Palestine, and Egypt. In the Iranian cultural sphere, Christians retained languages such as Armenian, Georgian,

56 Finbarr Barry Flood, ‘Eclecticism and Regionalism’. 57 Zadeh, The Vernacular Qurʾan. 58 On the Zubayrid coins, see Heidemann, ‘The Evolving Representation of the Early Islamic Empire and its Religion on Coin Imagery’ and Bacharach, ‘Signs of Sovereignty: the Shahāda, Qurʾanic Verses, and the Coinage of ʿAbd al-Malik’. On Ghaznavid coins with the statement of faith, see Flood, Objects of Translation, p. 41. 59 Griffith, Arabic Christianity in the Monasteries of Ninth-Century Palestine; Griffith, ‘From Aramaic to Arabic’; Leeming, ‘The Adoption of Arabic as a Liturgical Language by the Palestinian Melkites’.

i nt ro d u ct i o n

and Syriac in ecclesiastical settings as well as in their personal lives. For Jews, the story is similarly complicated. If we take the references to yahūdiyya in Arabic texts as denoting Judaeo-Arabic (which is very far from certain), some Jews adopted Arabic even before the rise of Islam.60 Beginning in the eighth or ninth century ce, papyri attest to the use of Judaeo-Arabic,61 a type of Arabic specifically legible to Jews. Shortly thereafter, we find Judaeo-Arabic translations of the Hebrew scriptures and, after that, works on theo­logy, law, and exegesis — all of which suggest the use of Arabic in religious contexts.62 However, just as Christian adaptation of Arabic depended on location, so, too, did some Jewish communities find other ways to communicate. Like the religious connotations of multilingualism among Muslims, Christians, and Jews, the significance of trade and travel similarly deserve additional attention in a broader conversation about multilingualism in the early Islamic world. Merchants and scholars communicated as they travelled across the breadth of the Caliphate and beyond, with the result that geo­graphical treatises and travelogues are fundamental to understanding ʿAbbasid history. Many of these texts refer to the languages spoken in different regions and provide anecdotes about how the travellers communicated. Nāṣir-i Khusraw, for example, uses the multilingualism of Akhlāṭ/ Xlat‘ to explain the city’s name as being based on the Arabic root kh-l-t, meaning ‘mixed’: ‘In the city of Akhlāṭ [Xlat‘] they speak three languages, Arabic, Persian, and Armenian. It is my supposition that this is why they named the town Akhlāṭ’ (‫ وظن من آن بود كه اخالط بدين‬.‫ تازى وپارسى وارمنى‬:‫ودر اين شهر اخالط به سه زبان سخن گویند‬ ‫)سبب نام آن شهر نهاده اند‬.63 Travellers of the Umayyad and ʿAbbasid periods similarly remarked on multilingualism in the Islamic world; for example, the Korean pilgrim Huichao asserted that ‘the people speak various languages’ in Iraq in the early eighth century ce.64 For merchants, in particular, the ability to engage with people as they travelled across and beyond the Caliphate was a noted advantage. In later periods, particularly in West Africa, scholars have traced the movements of merchants by the proliferation of Arabic manu­scripts.65 Such processes are more difficult to track in the Umayyad and early ʿAbbasid periods, but the role of Arabic and the importance of multilingualism appear in written descriptions of merchants. For example, Ibn Khurradādhbih notes the languages spoken by members of a famous group of 60 On the languages of the Jews in pre-Islamic Arabia, see the chapter by Debié in this volume. See also Newby, ‘Observations about an Early Judaeo-Arabic’. 61 Blau and Hopkins, ‘Judaeo-Arabic Papyri’. As far as Christian communities are concerned, an early case of Garshuni has been potentially documented, too; see Binggeli, Briquel Chatonnet, and Desreumaux, ‘Un cas très ancien de garshouni?’. 62 Hary, Multiglossia in Judeo-Arabic, p. 76. Hary’s chapter 4 lays out the history of Jewish languages and the periodization of Judaeo-Arabic, and the biblio­graphy lists a significant number of relevant studies. See also Joshua Blau’s discussion of the scholarship on the use of Hebrew in Judaeo-Arabic texts in Blau, ‘The Status of Arabic as Used by Jews in the Middle Ages’. 63 Nāṣir-i Khusraw, Book of Travels, p. 6. 64 George, ‘Direct Sea Trade between Early Islamic Iraq and Tang China’, p. 594. 65 Krätli and Lydon ed., The Trans-Saharan Book Trade.

29

30

a n to in e bo r r u t, m an u e l a c e b al lo s, and  ali so n m. vacca

Jewish merchants called the Rādhāniyya in his account of ‘the itinerary of the Jewish merchants al-Rādhāniyya, who speak Arabic, Persian, Greek, Frankish, Romance, and Slavonic, for they travel from west to east and from east to west by land and by sea’ ( ‫مسلك التجار اليهود الراذانية الذين يتكلّمون بالعربية والفارسية والرومية واالفرنجية واالندلسية‬ ً ‫برا ً وبحرا‬ َّ ‫)والصقلبية وانهم يسافرون من المشرق الى المغرب ومن المغرب الى المشرق‬.66 Above and beyond merchants’ role in the spread of Arabic and their ability to navigate vast multilingual spaces, the centrality of merchants to the story of Arabicization and multilingual encounters prompts a question about the formative role of class in the process of language change. Furthermore, the survival of documents in the Cairo Geniza offers an incredible opportunity to study how trade proliferated in a variety of languages. Finally, a third major theme that deserves more consideration is the relationship between language and identity. The field of Islamic history has made significant headway into the study of ethnic conceptualizations, questioning what made someone Arab,67 Persian,68 Amazigh,69 Kurd,70 or Armenian,71 to name just a few examples. Ethnic identity was associated with descent. Although the identity of one’s ancestors was negotiable,72 such a change was intergenerational. By contrast, language could be acquired or discarded through personal decision. Particularly in the case of Arabic, the use of a learned language informed an individual’s social standing and the circles that the individual could navigate. However, the complicated study of shuʿūbiyya (ethnic particularism) demonstrates that Arabicization (the spread of the Arabic language) did not automatically indicate Arabization (association with Arab identity).73 As the ninth-century litterateur Ibn Qutayba noted with some incredulity: ‘Were it the case that a person relinquishes his genealogy when he learns how to speak a language different from the language of his people, then any Easterner [lit. non-Arab] who speaks Arabic must necessarily be reckoned an Arab!’ (‫ولو كان ك ّل من تعلّم لسانا ً غير لسان قومه ونطق به خارجا ً من نسبهم لوجب أن يكون ك ّل من نطق‬ 66 Ibn Khurradādhbih, Kitāb al-Masālik wa-l-mamālik, p. 153, mentioned in Fred Astren’s chapter in this volume. For additional studies on the Rādhāniyya, see Gil, ‘The Rādhānite Merchants and the Land of the Rādhān’; Rabinowitz, Jewish Merchant Adventurers. See also the forthcoming work by Simcha Gross on this topic. 67 See above, n. 7. 68 Cooperson, ‘“Arabs” and “Iranians”’; Savant, ‘“Persians” in Early Islam’. 69 Rouighi, Inventing the Berbers; Valérian ed., Les Berbères entre Maghreb et Mashreq. 70 Asatryan, ‘Prolegomena to the Study of the Kurds’; James, ‘Arab Ethnonyms’; James, ‘Le “territoire tribal des Kurdes” et l’aire iraqienne’; James, Genèse du Kurdistan. 71 Garsoïan and Martin-Hisard, ‘Unity and Diversity in Medi­eval Caucasia’; Redgate, ‘Myth and Reality’; Vacca, ‘Conflict and Community in the Medi­eval Caucasus’; Zekiyan, ‘Le croisement des cultures dans les régions limitrophes de Géorgie, d’Arménie et de Byzance’. 72 The Kurdishness of the Rawwadids stand out as a good case study of this phenomenon. 73 Sarah Savant has argued convincingly that there were people who used shuʿūbī as an insult but no one who wore it as a badge of honour. As such, we might question whether there actually was a movement called shuʿūbiyya. For our purposes, the very fact that shuʿūbiyya was an object of conversation is a useful data point for thinking about the role of Arabic in creating a sense of the cultured or the Arab. See Savant, ‘Naming Shuʿūbīs’; her biblio­graphy offers a great introduction to the study of shuʿūbiyya in both Iran and al-Andalus.

i nt ro d u ct i o n

ً ‫)بالعربية من العجم عربيّا‬.74 In similar ways, Armenians living in the early Islamic world

wrote about other Armenians who no longer understood the Armenian language but nevertheless remained associated with Armenian identity.75 But at the same time, Caucasian Albanians employed the Armenian language to immortalize their own distinct history in ways that were definitively not Armenian.76 Language was a tool for inscribing communal identity, but it also had the capacity to collapse borders between communities or to redefine perceptions of communal bonds.77 The aspects of religion, travel, and identity certainly deserve additional attention as the study of language change in the early Islamic world continues. This volume contributes to our knowledge about the history of language in the early Islamic world, with a particular focus on the political dimensions of language choice, local elite strategies, and shared spaces of multiple languages, informed by some of the theoretical insights into multilingualism and cosmopolitanism developed in neighbouring fields. Taken together, all these elements invite us to think about culture and power in early Islam and about how the first generations to live under caliphal rule were ‘navigating language in the early Islamic world’.

74 Ibn Qutaybah, The Excellence of the Arabs, pp. 30–31. 75 See La Porta, ‘Captive Identities’. 76 On Movsēs Dasxuranc‘i’s history of the Albanians, which was written in Armenian, see HowardJohnston, ‘Caucasian Albania and its Historian’ and Zuckerman, ‘The Khazars and Byzantium’; on the complicated history of language use in Caucasian Albania, see Gadjiev, ‘The Role and Place of Middle Persian Language and Writing in Caucasian Albania’; Gadjiev, ‘On the History of Writing in Caucasian Albania’. 77 A useful case study for Coptic is Papaconstantinou, ‘“They Shall Speak the Arabic Language and Take Pride in It”’.

31

32

a n to in e bo r r u t, m an u e l a c e b al lo s, and  ali so n m. vacca

Biblio­graphy Primary Sources Ibn Khurradādhbih, Abū al-Qāsim ʿUbayd Allāh, Kitāb al-Masālik wa-l-mamālik, ed. by M. J. de Goeje (Leiden: Brill, 1889) Ibn Qutaybah, The Excellence of the Arabs, ed. by James Montgomery and Peter Webb, trans. by Sarah Bowen Savant and Peter Webb (New York: New York University Press, 2017) Juvaynī, Taʾrīkh-i Jahān-gushā, ed. Muḥammad M. Qazvīnī, iii (London: Luzac & Co, 1912–1937), trans. by John A. Boyle, The History of the World-Conqueror, ii (Manchester: Manchester University Press, 1958) Nāṣir-i Khusraw, Book of Travels, ed. and trans. by Wheeler Thackston (Albany: State University of New York Press, 1985) Rashīd al-Dīn, Jāmiʿ al-tawārīkh, ed. by Muḥammad Rūshan and Musṭafā Mūsavī, ii (Tehran: Markaz-i Pizhūhishī-i Mīrāth-i Maktūb, 1994) Secondary Works Aillet, Cyrille, Les Mozarabes. Islamisation, arabisation et christianisme en péninsule Ibérique (ixe–xiie siècle) (Madrid: Bibliothèque de la Casa de Velázquez, 2010) —— , ‘Islamisation et arabisation dans le monde musulman médiéval: une introduction au cas de l’Occident musulman (viie–xiie siècle)’, in Islamisation et arabisation de l’Occident musulman médiéval (viie–xiie siècle), ed. by Dominique Valérian (Paris: Éditions de la Sorbonne, 2011), pp. 7–34 Akbari, Suzanne Conklin, and Karla Mallette, ed., A Sea of Languages: Rethinking the Arabic Role in Medi­eval Literary History (Toronto: University of Toronto Press, 2013) Al-Jallad, Ahmad, ‘What is Ancient North Arabian?’, in Re-engaging Comparative Semitic and Arabic Studies, ed. by Daniel Birnstiel and Naʿama Pat-El (Wiesbaden: Harrassowitz, 2018), pp. 1–43 Al-Jallad, Ahmad, and Hythem Sidky, ‘A Paleo-Arabic Inscription on a Route North of Ṭāʾif ’, Arabian Archaeo­logy and Epi­graphy, 33.1 (2021), 1–14 Anthony, Sean, ‘Early Arabo-Islamic Epi­graphy and the Positivist Fallacy: A Brief Communication’, Al-ʿUṣūr al-Wusṭā, 26 (2018), 201–07 Appiah, Kwame Anthony, Cosmopolitanism: Ethics in a World of Strangers (New York: W. W. Norton, 2006) Asatryan, Garnik, ‘Prolegomena to the Study of the Kurds’, Iran and the Caucasus, 13 (2009), 1–58 Aslanov, Cyril, ‘Quand les langues romanes se confondent…: la Romania vue d’ailleurs’, Language et société, 99 (2002), 9–52 Bacharach, Jere, ‘Signs of Sovereignty: The Shahāda, Qurʾanic Verses, and the Coinage of ʿAbd al-Malik’, Muqarnas, 27 (2010), 1–30 Bauden, Frédéric, ‘Les relations diplomatiques entre les sultans mamlouks circassiens et les autres pouvoirs du Dār al-islām: l’apport du ms. ar. 4440 (BNF, Paris)’, Annales islamo­ logiques, 41 (2007), 1–29

i nt ro d u ct i o n

Bauden, Fréderic, and Malika Dekkiche, ed., Mamluk Cairo, a Crossroads for Embassies: Studies on Diplomacy and Diplomatics (Leiden: Brill, 2019) Belli Bose, Melia, ed., Intersections: Islamic Art and Cosmopolitanism (Gainesville: University Press of Florida, 2021) Binggeli, André, Françoise Briquel Chatonnet, and Alain Desreumaux, ‘Un cas très ancien de garshouni? Quelques réflexions sur le manuscrit British Library Add 14644’, in Loquentes linguis: studi linguistici e orientale in onore di Fabrizio A. Pennacchietti, ed. by Pier Giorgio Borbone, Alessandro Mengozzi, and Mauro Tosco (Wiesbaden: Harrassowitz, 2006), pp. 141–47 Blau, Joshua, ‘The Status of Arabic as Used by Jews in the Middle Ages’, Journal of Jewish Studies, 10 (1959), 15–23 Blau, Joshua, and Simon Hopkins, ‘Judaeo-Arabic Papyri: Collected, Edited, Translated and Analysed’, Jerusalem Studies in Arabic and Islam, 9 (1987), 9–27 Borrut, Antoine, Entre mémoire et pouvoir: l’espace syrien sous les derniers Omeyyades et les premiers Abbassides (v. 72–193/692–809) (Leiden: Brill, 2011) —— , ‘An Islamic Late Antiquity? Problems and Perspectives’, Journal of Late Antique, Islamic and Byzantine Studies (forthcoming) Boud’hors, Anne, ‘Degrés d’arabisation dans l’Égypte du viiie siècle: CPR II 228 revisité’, in Mélanges Jean Gascou. Textes et études papyro­logiques, ed. by Jean-Luc Fournet and Arietta Papaconstantinou (Paris: Association des ami du Centre d’histoire et civilisation de Byzance, 2016), pp. 71–89 Boyle, John A., ‘Dynastic and Political History of the Il-Khāns’, in The Cambridge History of Iran, ed. by John A. Boyle, 8 vols (Cambridge: Cambridge University Press, 1968), v, pp. 303–421 —— , ‘G̱ ẖāzān’, in Encyclopaedia of Islam, 2nd edn, ed. by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs (Leiden: Brill, online edn) [accessed 4 January 2022] Brann, Ross, ed., Languages of Power in Medi­eval Spain (Bethesda: CDL Press, 1997) Breckenridge, Carol A., Sheldon Pollock, Homi K. Bhabha, and Dipesh Chakrabarty, ed., Cosmopolitanism (Durham, NC: Duke University Press, 2002) Classen, Albrecht, ed., Multilingualism in the Middle Ages and Early Modern Age: Com­ munication and Miscommunication in the Premodern World (Berlin: De Gruyter, 2016) Conrad, Lawrence, ‘Theophanes and the Arabic Historical Tradition: Some Indications of Intercultural Transmission’, Byzantinische Forschungen, 15 (1990), 1–44 —— , ‘The Conquest of Arwād: A Source-Critical Study in the Historio­graphy of the Early Medi­eval Near East’, in The Byzantine and Early Islamic Near East, i: Problems in the Literary Source Material, ed. by Averil Cameron and Lawrence Conrad (Princeton, NJ: Darwin Press, 1992), pp. 317–401 Cooperson, Michael, ‘“Arabs” and “Iranians”: The Uses of Ethnicity in the Early Abbasid Period’, in Islamic Cultures, Islamic Contexts: Essays in Honor of Professor Patricia Crone, ed. by Behnam Sadeghi, Asad Q. Ahmed, Robert Hoyland, and Adam Silverstein (Leiden: Brill, 2014), pp. 364–87 Cotton, Hannah M., Robert G. Hoyland, Jonathan J. Price, and David J. Wasserstein, eds, From Hellenism to Islam: Cultural and Linguistic Change in the Roman Near East (Cambridge: Cambridge University Press, 2009)

33

34

a n to in e bo r r u t, m an u e l a c e b al lo s, and  ali so n m. vacca

Dakhlia, Jocelyne, Lingua franca: histoire d’une langue métisse en Méditerranée (Arles: Actes Sud, 2008) Dolezalek, Isabelle, Arabic Script on Christian Kings: Textile Inscriptions on Royal Garments from Norman Sicily (Berlin: De Gruyter, 2017) D’Ottone, Arianna, ‘Manu­scripts as Mirrors of a Multilingual and Multicultural Society: The Case of the Damascus Find’, in Negotiating Co-existence: Communities, Cultures and Convivencia in Byzantine Society, ed. by Barbara Crostini and Sergio La Porta (Trier: Wissenschaftlicher Verlag Trier, 2013), pp. 63–88 D’Ottone Rambach, Arianna, Konrad Hirschler, and Ronny Vollandt, ed., The Damascus Fragments: Towards a History of the Qubbat al-Khazna Corpus of Manu­scripts and Documents (Beirut: Ergon, 2020) Eche, Youssef, Les bibliothèques arabes publiques et semi-publiques en Mésopotamie, en Syrie et en Égypte au Moyen Age (Damascus: Institut Français de Damas, 1967) Fisher, Greg, ed., Arabs and Empires before Islam (Oxford: Oxford University Press, 2015) Flood, Finbarr Barry, Objects of Translation: Material Culture and Medi­eval Hindu-Muslim Encounter (Princeton, NJ: Princeton University Press, 2009) —— , ‘Eclecticism and Regionalism: The Gwalior Qurʾan and the Ghurid Legacy to PostMongol Art’, in Le Coran de Gwalior: polysémie d’un manuscrit à peintures, ed. by Éloïse Brac de la Perrière and Monique Burési (Paris: De Boccard, 2016), pp. 153–69 Fowden, Garth, Quṣayr ʿAmra: Art and the Umayyad Elite in Late Antique Syria (Berkeley: University of California Press, 2004) Gadjiev, Murtazali S., ‘On the History of Writing in Caucasian Albania’, in Written Culture in Daghestan, ed. by Moshe Gammer (Helsinki: Annales Academiae Scientiarum Fennicae, 2015), pp. 41–54 —— , ‘The Role and Place of Middle Persian Language and Writing in Caucasian Albania’, Oriens, 5 (2021), 59–70 Gallego, María Ángeles, ‘The Languages of Medi­eval Iberia and their Religious Dimension’, Medi­eval Encounters, 9.1 (2003), 107–39 Garosi, Eugenio, ‘Imperial Arabic: Some Notes on Visual Symbolism’, in Christians and Muslims in Early Islamic Egypt, ed. by Lajos Berkes (Durham, NC: American Society of Papyro­logists, 2022), pp. 13–31 Garsoïan, Nina, and Bernadette Martin-Hisard, ‘Unity and Diversity in Medi­eval Caucasia, 4th–11th Centuries’, in Languages and Cultures of Eastern Christianity: Georgian, ed. by Stephen Rapp and Paul Crego (Burlington, VT: Ashgate, 2012), pp. 49–96 Gaunt, Simon, ‘Can the Middle Ages Be Postcolonial?’, Comparative Literature, 61.2 (2009), 160–76 George, Alain, ‘Direct Sea Trade between Early Islamic Iraq and Tang China: From the Exchange of Goods to the Transmission of Ideas’, Journal of the Royal Asiatic Society, 3rd ser., 25.4 (2015), 579–624 al-Ghabbān, ʿAlī Ibrāhīm, and Robert Hoyland, ‘The Inscription of Zuhayr, the Oldest Islamic Inscription (24 ah / ad 644–5), the Rise of the Arabic Script and the Nature of the Early Islamic State’, Arabian Archaeo­logy and Epi­graphy, 19 (2008), 210–37 Gil, Moshe, ‘The Rādhānite Merchants and the Land of the Rādhān’, Journal of the Economic and Social History of the Orient, 17.3 (1967), 299–328 Gilbert, Claire, In Good Faith: Arabic Translation and Translators in Early Modern Spain (Philadelphia: University of Pennsylvania Press, 2020)

i nt ro d u ct i o n

Gilliot, Claude, and Pierre Larcher, ‘Language and Style’, in Encyclopaedia of the Qurʾān, ed. by Jane Dammen McAuliffe, iii (Leiden: Brill, 2003), pp. 109–35 Golden, Peter B., ‘The Byzantine Greek Elements in the Rasulid Hexaglot’, Archivum Eurasiae Medii Aevi, 5 (1985), 41–166 —— , ed., The King’s Dictionary: The Rasūlid Hexaglot — Fourteenth Century Vocabularies in Arabic, Persian, Turkic, Greek, Armenian and Mongol (Leiden: Brill, 2000) Grévin, Benoît, Le parchemin des cieux: essai sur le Moyen Âge du langage (Paris: Seuil, 2012) Griffith, Sydney, Arabic Christianity in the Monasteries of Ninth-Century Palestine (Aldershot: Ashgate, 1992) —— , ‘From Aramaic to Arabic: The Languages of the Monasteries of Palestine in the Byzantine and Early Islamic Periods’, Dumbarton Oaks Papers, 51 (1997), 11–31 Gruendler, Beatrice, The Development of the Arabic Scripts: From the Nabatean Era to the First Islamic Century According to Dated Texts (Atlanta: Scholars Press, 1993) Gutas, Dimitri, Greek Thought, Arabic Culture: The Graeco-Arabic Translation Movement in Baghdad and Early ʿAbbāsid Society (2nd–4th/8th–10th Centuries) (London: Routledge, 1998) Hamilton, Michelle M., and Nuria Silleras-Fernandez, eds, Iberian Babel: Translation and Multilingualism in the Medi­eval and the Early Modern Mediterranean (Leiden: Brill, 2022) Hary, Benjamin H., Multiglossia in Judeo-Arabic with an Edition, Translation, and Grammatical Study of the Cairene Purim Scroll (Leiden: Brill, 1992) Heidemann, Stefan, ‘The Evolving Representation of the Early Islamic Empire and its Religion on Coin Imagery’, in The Qurʾān in Context: Historical and Literary Investigations into the Qurʾānic Milieu, ed. by Angelika Neuwirth, Nicolai Sinai, and Michael Marx (Leiden: Brill, 2010), pp. 149–95 Hirschfeld, Yizhar, and Giora Solar, ‘The Roman Thermae at Ḥammat Gader: Preliminary Report of Three Seasons of Excavations’, Israel Exploration Journal, 31 (1981), 197–219 Howard-Johnston, James, ‘Caucasian Albania and its Historian’, in From Albania to Arrān: the East Caucasus between the Ancient and Islamic Worlds (ca. 330 bce–1000 ce), ed. by Robert Hoyland (Piscataway, NJ: Gorgias Press, 2020), pp. 351–70 Hoyland, Robert, Seeing Islam as Others Saw It: A Survey and Evaluation of Christian, Jewish, and Zoroastrian Writings on Early Islam (Princeton, NJ: Darwin Press, 1997) —— , ‘Language and Identity: The Twin Histories of Arabic and Aramaic (and: Why Did Aramaic Succeed Where Greek Failed?)’, Scripta Classica Israelica, 23 (2004), 183–99 —— , In God’s Path: The Arab Conquests and the First Islamic Empire (Oxford: Oxford University Press, 2015) —— , ‘Reflections on the Identity of the Arabian Conquerors of the Seventh-Century Middle East’, Al-ʿUṣūr al-Wusṭā, 25 (2017), 113–40 —— , ‘Two New Arabic Inscriptions: Arabian Castles and Christianity in the Umayyad Period’, in To the Madbar and Back Again: Studies in the Languages, Archaeo­logy, and Cultures of Arabia Dedicated to Michael C. A. Macdonald, ed. by Laïla Nehmé and Ahmad Al-Jallad (Leiden: Brill, 2017), pp. 327–38 Iema, Z. T., Ahmad Al-Jallad, Michael C. A. Macdonald, and Laïla Nehmé, ‘Provincia Arabia: Nabataea, the Emergence of Arabic as a Written Language, and GraecoArabica’, in Arabs and Empires before Islam, ed. by Greg Fisher (Oxford: Oxford University Press, 2015), pp. 373–433

35

36

a n to in e bo r r u t, m an u e l a c e b al lo s, and  ali so n m. vacca

Imbert, Frédéric, ‘L’islam des pierres: l’expression de la foi dans les graffiti arabes des premiers siècles’, Revue des mondes musulmans et de la Méditerranée, 129 (2011), 57–77 —— , ‘Le Coran des pierres: statistiques épi­graphiques et premières analyses’, in Le Coran: nouvelles approches, ed. by Mehdi Azaiez and Sabrina Mervin (Paris: CNRS, 2013), pp. 99–124 —— , ‘Graffiti arabes de Cnide et de Kos: premières traces épi­graphiques de la conquête musulmane en mer Egée’, Travaux et mémoires, 17 (2013), 731–58 —— , ‘Le prince al-Walīd et son bain: itinéraires épi­graphiques à Quṣayr ʿAmra’, Bulletin d’études orientales, 64 (2016), 321–63 Internullo, Dario, and Arianna D’Ottone Rambach, ‘One Script for Two Languages: Latin and Arabic in an Early Allo­graphic Papyrus’, in Palaeo­graphy between East and West: Proceedings of the Seminars on Arabic Palaeo­graphy at Sapienza University of Rome, ed. by Arianna D’Ottone Rambach (Rome: Fabrizio Serra, 2018), pp. 53–72 Jackson, Peter, ‘Ghāzān Khān Maḥmūd’, in Encyclopaedia of Islam, 3rd edn, ed. by Kate Fleet, Gudrun Krämer, Denis Matringe, John Nawas, and Everett Rowson (Leiden: Brill, online edn) [accessed 29 December 2021] James, Boris, Genèse du Kurdistan: les Kurdes dans l’Orient mamelouk et mongol (Paris: Éditions de la Sorbonne, 2021) —— , ‘Le “territoire tribal des Kurdes” et l’aire iraqienne (xe–xiiie siècles): esquisse des recompositions spatiales’, Revue des mondes musulmans et de la Méditerranée, 117–18 (2007), 106–26 —— , ‘Arab Ethnonyms (ʿAjam, ʿArab, Badū and Turk): The Kurdish Case as a Paradigm for Thinking about Differences in the Middle Ages’, Iranian Studies, 47.5 (2014), 683–712 Jamil, Nadia, and Jeremy Johns, ‘Signs of the Times: Arabic Signatures as a Measure of Acculturation in Norman Sicily’, Muqarnas, 21.1 (2004), 181–92 Johns, Jeremy, Arabic Administration in Norman Sicily: The Royal Dīwān (Cambridge: Cambridge University Press, 2002) Johnson, Scott Fitzgerald, ed., Languages and Cultures of Eastern Christianity: Greek (London: Routledge, 2015) Kane, Daniel, The Kitan Language and Script (Leiden: Brill, 2008) Kappler, Claire, and Suzanne Thiolier-Méjean, Le plurilinguisme au moyen âge: de Babel à la langue une (Paris: Harmattan, 2009) King, Ross, ‘Introduction’, in Infected Korean Language, Purity vs. Hybridity: From the Sino­graphic Cosmopolis to Japanese Colonialism to Global English, ed. by Koh Jongsok (Amherst, NY: Cambria Press, 2014), pp. 1–16 König, Daniel, ‘The Unkempt Heritage: On the Role of Latin in the Arabic-Islamic Sphere’, Arabica, 63.5 (2016), 419–93 —— , Latin and Arabic: Entangled Histories (Heidelberg: Heidelberg University Press, 2019) Kossmann, Maarten, The Arabic Influence on Northern Berber (Leiden: Brill, 2013) Krätli, Graziano, and Ghislaine Lydon, The Trans-Saharan Book Trade: Manu­script Culture, Arabic Literacy and Intellectual History in Muslim Africa (Leiden: Brill, 2011) Lancel, Serge, ‘La fin et la survie de la latinité en Afrique du Nord: état des questions’, Revue des études latines, 59 (1982), 269–97 La Porta, Sergio, and Alison M. Vacca, An Armenian Futūḥ Narrative: Łewond’s EighthCentury History of the Caliphate (Chicago: Institute for the Study of Ancient Cultures of the University of Chicago, forthcoming)

i nt ro d u ct i o n

La Porta, Sergio, ‘Captive Identities: Inscribing Armenianness from Sebēos to Ayrivanec‘i’, in War and Collective Identities in the Middle Ages: East, West, and Beyond, ed. by Yannis Stouraitis (Leeds: Arc Humanities Press, 2023), pp. 65–86 Larcher, Pierre, ‘Al-lugha al-fuṣḥâ: archéo­logie d’un concept “idéolinguistique”’, Revue des mondes musulmans et de la Méditerranée, 124 (2008), 263–78 —— , L’invention de la luġa al-fuṣḥā: une histoire de l’arabe par les textes (Louvain: Peeters, 2021) Lavan, Myles, Richard E. Payne, and John Weisweiler, ‘Cosmopolitan Politics: The Assimilation and Subordination of Elite Cultures’, in Cosmopolitanism and Empire: Universal Rulers, Local Elites, and Cultural Integration in the Ancient Near East and Mediterranean, ed. by Myles Lavan, Richard E. Payne, and John Weisweiler (Oxford: Oxford University Press, 2016), pp. 1–28 Lawrence, Bruce B., Islamic Cosmopolitan Spirit: Myth or Movement (Hoboken, NJ: WileyBlackwell, 2021) Leeming, Kate, ‘The Adoption of Arabic as a Liturgical Language by the Palestinian Melkites’, Aram, 15 (2003), 239–46 Macdonald, Michael C. A., ‘Reflections on the Linguistic Map of Pre-Islamic Arabia’, Arabian Archaeo­logy and Epi­graphy, 11 (2000), 28–79 —— , ‘Arabs, Arabias, and Arabic before Late Antiquity’, Topoi, 16.1 (2009), 277–332 —— , ed., The Development of Arabic as a Written Language (Oxford: Archaeopress, 2010) Mackensen, Ruth, ‘Supplementary Notes to “Arabic Books and Libraries in the Umayyad Period”’, American Journal of Semitic Languages and Literatures, 56.2 (1939), 149–57 Malek, Hodge Mehdi, The Dābūyid Ispahbads and Early ʿAbbāsid Governors of Tabaristān: History and Numismatics (London: Royal Numismatic Society, 2004) Mallette, Karla, The Kingdom of Sicily, 1100–1250: A Literary History (Philadelphia: University of Pennsylvania Press, 2005) —— , Lives of the Great Languages: Arabic and Latin in the Medi­eval Mediterranean (Chicago: University of Chicago Press, 2021) Mango, Cyril, ‘Greek Culture in Palestine after the Arab Conquest’, in Scritture, libri e testi (Spoleto: Centro italiano di studi sull’alto Medioevo, 1991), pp. 149–60 Mauder, Christian, In the Sultan’s Salon: Learning, Religion and Rulership at the Mamluk Court of Qāniṣawh al-Ghawrī (r. 1501–1516), 2 vols (Leiden: Brill, 2021) Mavroudi, Maria, ‘Greek Language and Education under Early Islam’, in Islamic Cultures, Islamic Contexts: Essays in Honor of Professor Patricia Crone, ed. by Behnam Sadeghi, Asad Q. Ahmed, Robert Hoyland, and Adam Silverstein (Leiden: Brill, 2014), pp. 295–342 Meouak, Mohamed, La langue berbère au Maghreb médiéval: textes, contextes, analyses (Leiden: Brill, 2016) Metcalfe, Alexander, Muslims and Christians in Norman Sicily: Arabic-Speakers and the End of Islam (London: Taylor and Francis, 2014) Miles, George C., ‘Early Islamic Inscriptions near Ṭāʾif in the Ḥijāz’, Journal of Near Eastern Studies, 7.4 (1948), 236–42 Miller, Elaine R., Jewish Multiglossia: Hebrew, Arabic, and Castilian in Medi­eval Spain (Newark, DE: Juan de la Cuesta, 2000) Miller, Nathaniel, ‘Muslim Poets under a Christian King: An Intertextual Reevaluation of Sicilian Arabic Literature under Roger II (1112–54)’, Mediterranean Studies, 27.2 (2019), 182–209

37

38

a n to in e bo r r u t, m an u e l a c e b al lo s, and  ali so n m. vacca

Morrison, Robert, ‘Islamic Astronomy’, in The Cambridge History of Medi­eval Science, ii: Medi­eval Science, ed. by David C. Lindberg and Michael H. Shank (Cambridge: Cambridge University Press, 2013), pp. 109–38 Mullen, Alex, and Patrick James, ed., Multilingualism in the Graeco-Roman Worlds (Cambridge: Cambridge University Press, 2012) Navarro García, Ángeles, and Bilal Sarr, ed., Arabización, islamización y resistencias en alAndalus y el Maghreb (Granada: Universidad de Granada, 2019) Nef, Annliese, Conquérir et gouverner la Sicile islamique aux xie et xiie siècles (Rome: École Française de Rome, 2011) Nehmé, Laïla, ‘Aramaic or Arabic? The Nabataeo-Arabic Script and the Language of the Inscriptions Written in this Script’, in Arabic in Context: Celebrating 400 years of Arabic at Leiden University, ed. by Ahmad Al-Jallad (Leiden: Brill, 2017), pp. 75–98 Newby, Gordon Darnell, ‘Observations about an Early Judaeo-Arabic’, Jewish Quarterly Review, 61 (1971), 212–21 Papaconstantinou, Arietta, ‘“They Shall Speak the Arabic Language and Take Pride in It”: Reconsidering the Fate of Coptic after the Arab Conquest’, Le muséon, 120.3–4 (2007), 273–99 —— , The Multilingual Experience in Egypt, from the Ptolemies to the Abbasids (Abingdon: Routledge, 2010) —— , ‘Why Did Coptic Fail Where Aramaic Succeeded? Linguistic Developments in Egypt and the Near East after the Arab Conquest’, in Multilingualism in the GraecoRoman Worlds, ed. by Alex Mullen and Patrick James (Cambridge: Cambridge University Press, 2012), pp. 58–76 —— , ‘Introduction’ in The Multilingual Experience in Egypt from the Ptolemies to the Abbasids, ed. by Arietta Papaconstantinou (London: Routledge, 2010), pp. 1–16 —— , ‘Arabic: Language of Empire, Language of Egypt’, in Egypt and Empire: The Formation of Religious Identity after Rome, ed. by Elisabeth R. O’Connell (Leuven: Peeters, 2022), pp. 293–309 Pfeifer, Helen, ‘Encounter after the Conquest: Literary Salons in Sixteenth-Century Ottoman Damascus’, International Journal of Middle East Studies, 47 (2015), 219–39 —— , Empire of Salons: Conquest and Community in Early Modern Ottoman Lands (Princeton, NJ: Princeton University Press, 2022) Pollock, Sheldon, The Language of the Gods in the World of Men: Sanskrit, Culture, and Power in Premodern India (Berkeley: University of California Press, 2006) Pollock, Sheldon, Homi K. Bhabha, Carol A. Breckenridge, and Dipesh Chakrabarty, ‘Cosmopolitanisms’, Public Culture, 12.3 (2000), 577–89 Rabinowitz, Louis, Jewish Merchant Adventurers: A Study of the Radanites (London: Edward Goldston, 1948) al-Rāshid, Saʿd b. ʿAbd al-ʿAzīz, ‘Sadd al-Khanaq: An Early Umayyad Dam near Medina, Saudi Arabia’, Proceedings of the Seminar for Arabian Studies, 38 (2008), 265–76 Redgate, Anne Elizabeth, ‘Myth and Reality: Armenian Identity in the Early Middle Ages’, National Identities, 9.4 (2007), 281–306 Retsö, Jan, ‘The Earliest Arabs’, Orientalia Suecana, 38–39 (1990), 131–39 —— , The Arabs in Antiquity: Their History from the Assyrians to the Umayyads (London: RoutledgeCurzon, 2003)

i nt ro d u ct i o n

Richardson, Kristina, ‘New Evidence for Early Modern Ottoman Arabic and Turkish Sign Systems’, in Sign Language Studies, 17.2 (2017), 172–92 Richter, Tonio Sebastian, ‘Greek, Coptic and the “Language of the Hijra”: The Rise and Decline of the Coptic Language in Late Antique and Medi­eval Egypt’, in From Hellenism to Islam: Cultural and Linguistic Change in the Roman Near East, ed. by Hannah M. Cotton, Robert G. Hoyland, Jonathan J. Price, and David J. Wasserstein (Cambridge: Cambridge University Press, 2009), pp. 401–46 Robin, Christian Julien, ‘Les arabes vus de Ḥimyar’, Topoi, 14 (2006), 121–37 —— , Najrān en Arabie et ses martyrs chrétiens (forthcoming). Robin, Christian, ʿAlī Ibrāhīm al-Ghabbān, and Saʿīd Fāyiz al-Saʿīd, ‘Inscriptions antiques de la region de Najrān (Arabie Séoudite méridionale): nouveaux jalons pour l’histoire de l’écriture, de la langue et du calendrier arabe’, Comptes rendus des séances de l’Académie des inscriptions et belles-lettres, 3 (2014), 1033–28 Robinson, Chase F., ʿAbd al-Malik (Oxford: Oneworld, 2005) Rouighi, Ramzi, Inventing the Berbers: History and Ideo­logy in the Maghrib (Philadelphia: University of Pennsylvania Press, 2019) Ruska, Julius, Arabische Alchemisten, i: Chālid ibn Jazīd ibn Muʿāwija (Heidelberg: Carl Winter, 1924) Saliba, George, Islamic Science and the Making of the European Renaissance (Cambridge, MA: MIT Press, 2007) Savant, Sarah Bowen, ‘Naming Shuʿūbīs’, in Essays in Islamic Philo­logy, History, and Philosophy: A Festschrift in Celebration and Honor of Professor Ahmad Mahdavi Damghani’s 90th Birthday (Berlin: De Gruyter, 2016), pp. 166–84 —— , ‘“Persians” in Early Islam’, Annales islamo­logiques, 42 (2008), 73–91 al-Shdaifat, Younis, Ahmad Al-Jallad, Zeyad al-Salameen, and Rafe Harahsheh, ‘An Early Christian Arabic Graffito Mentioning “Yazīd the King”’, Arabian Archaeo­logy and Epi­ graphy, 28.2 (2017), 315–24 Smelik, Willem F., Rabbis, Language and Translation in Late Antiquity (Cambridge: Cambridge University Press, 2013) Stein, Peter, ‘The “Himyaritic” Language in Pre-Islamic Yemen: A Critical Re-evaluation’, Semitica et classica, 1 (2008), 203–12 Stroumsa, Rachel, ‘Greek and Arabic in Nessana’, in Documents and the History of the Early Islamic World: 3rd Conference of the International Study for Arabic Papyro­logy, ed. by Alexander T. Schubert and Petra Sijpesteijn (Leiden: Brill, 2015), pp. 143–57 Tornow, Siegfried, Abendland und Morgenland im Spiegel ihrer Sprachen: ein kulturhistorischer Vergleich (Wiesbaden: Harrassowitz, 2009) Ullmann, Manfred, ‘Ḫālid ibn Yazīd und die Alchemie: eine Legende’, Der Islam, 55 (1978), 181–218 Vacca, Alison M., ‘Conflict and Community in the Medi­eval Caucasus’, Al-ʿUṣūr al-Wusṭā, 25 (2017), 66–112 Valérian, Dominique, ed., Islamisation et arabisation de l’Occident musulman médiéval (viie– xiie siècle) (Paris: Éditions de la Sorbonne, 2011) —— , ed., Les Berbères entre Maghreb et Mashreq (viie–xve siècle) (Madrid: Casa de Velázquez, 2021)

39

40

a n to in e bo r r u t, m an u e l a c e b al lo s, and  ali so n m. vacca

Van Bladel, Kevin, ‘Arabicization, Islamization, and the Colonies of the Conquerors’, in Late Antique Responses to the Arab Conquests, ed. by Josephine van den Bent, Floris van den Eijnde, and Johan Weststeijn (Leiden: Brill, 2022), pp. 89–119 Walker, John, A Catalogue of the Muhammadan Coins in the British Museum, i: ArabSassanian Coins (London: British Museum, 1941) Wansbrough, John E., Lingua Franca in the Mediterranean (Richmond: Curzon, 1996) Wasserstein, David, ‘The Language Situation in al-Andalus’, in Studies on the Muwashshah and the Kharja, ed. by Alan Jones and Richard Hitchcock (Oxford: Oxford University Press, 1991), pp. 1–15 —— , ‘Why Did Arabic Succeed Where Greek Failed? Language Change in the Near East after Muhammad’, Scripta Classica Israelica, 22 (2003), 257–72 Webb, Peter, Imagining the Arabs: Arab Identity and the Rise of Islam (Edinburgh: Edinburgh University Press, 2016) Wolper, Ethel Sara, ‘Khidr and the Politics of Translation in Mosul: Mar Behnam, St George and Khidr Ilyas’, in Sacred Precincts: The Religious Architecture of Non-Muslim Communities across the Islamic World, ed. by Gharipour Mohammad (Leiden: Brill, 2014), pp. 379–92 Yücesoy, Hayrettin, ‘Translation as Self-Consciousness: Ancient Sciences, Antediluvian Wisdom, and the ʿAbbāsid Translation Movement’, Journal of World History, 20.4 (2009), 523–57 Zadeh, Travis, The Vernacular Qurʾan: Translation and the Rise of Persian Exegesis (Oxford: Oxford University Press, 2012) Zekiyan, Boghos Levon, ‘Le croisement des cultures dans les régions limitrophes de Géorgie, d’Arménie et de Byzance’, Annali di Ca’ Foscari, 25.3 (1986), 81–89 Zuckerman, Constantine, ‘The Khazars and Byzantium: The First Encounter’, in The World of the Khazars, ed. by Peter B. Golden and others (Leiden: Brill, 2007), pp. 404–17

Section One

Languages of the Umayyad Caliphate

Petra M. Sijpesteijn

The Early Islamic Empire’s Policy of Multilingual Governance Siempre la lengua fue compañera del imperio Antonio de Nebrija, Gramática de la lengua española, 14921

Explaining the process of Arabicization remains one of the main challenges in the study of the history of the early Muslim empire. Arabic was used immediately as a symbol of rule and authority in the areas conquered by the Muslims, and the use of Arabic increased over time. At the same time, the Muslim administration continued to produce and interact with documents in Coptic, Greek, Pahlavi, Latin, and Sogdian even centuries after the conquest. Other historical processes, such as personnel changes in the administration and the development of governance infrastructures, impacted language use and interfered with later accounts recording linguistic transformations in the caliphate’s chancery and beyond. Using documents from Egypt to Khurasan produced by the administrations of the Umayyads and the early ʿAbbāsids — that is to say, from the 640s to the mid ninth century ce — with a focus on the languages that these documents were written in and how these languages relate to each other, together with historical accounts and theoretical insights from empire studies, this chapter will examine the question of language use and language policy in the early Islamic empire as an element of the greater discussion about Arabicization.

Changing the Language of the Dīwān Let us begin with the state of the field concerning the spread of the Arabic language after the great Arab conquests of the mid-seventh and eighth centuries ce. How have scholars explained the fact that large parts of the Muslim world, where a multitude of languages were in use at the time of the Arab conquests, eventually

1 ‘Language was always the companion of empire’; Nebrija, Gramática de la lengua española, p. 109. An English translation of the introduction can be found in Armillas-Tiseyra, ‘On Language and Empire’, p. 202. I would like to thank Farah Bazzi for these references.

Petra M. Sijpesteijn ([email protected]) is professor of Arabic at Leiden University. She uses material, documentary and literary sources to study the workings of the early Islamic empire. Navigating Language in the Early Islamic World: Multilingualism and Language Change in the First Cen­turies of Islam, ed. by Antoine Borrut, Manuela Ceballos, and Alison M. Vacca, ISMAR 2 (Turnhout: Brepols, 2024) pp. 43–87 10.1484/M.ISMAR-EB.5.134624

FHG

44

p e tr a  m . s ij pe s t e i j n

became Arabic-speaking? Like the medi­eval historians whose works they use, scholars today have searched for a singular decision, decree, or event that sparked the linguistic change to Arabic.2 Indeed, the literary sources contain such reports, and every story of the empire’s Arabicization starts with them. The anecdotes are set in the time of the Marwānid caliphs, famous for their long and successful reigns during which they introduced plentiful governmental innovations. The accounts describe the translation of the dīwāns, or state registers, from local languages into Arabic. In the context of these anecdotes, dīwān has a restricted meaning. It denotes lists with information about taxable landed properties; the names of the fiscally liable owners or cultivators; and factual information on the dimensions, conditions of irrigation, location, and state of cultivation of these properties, including the kinds of crops sown, progress in the crop’s growth, and the area of land that was being cultivated. Finally, and — from the point of view of the authorities — most importantly, the dīwān contained a calculation of the taxes due on the basis of all this information. These ledgers were updated on an annual basis using information sent to the central offices by surveyors and assessors who visited the countryside and by local administrators based in the cultivators’ communities.3 Such fiscal registers had already been in use in the pre-Islamic period. After the conquests, not much changed in the kinds of information recorded or the ways in which the necessary information was gathered and organized, including the languages in which this was done, although the Arab administration in Egypt definitely supervised its officials more closely and produced more administrative documentation than its direct predecessors had done.4 Other kinds of lists produced for the caliph’s administration were referred to with the same word; these included the supposedly first dīwān to be newly established, namely, the register of fighters and others who migrated to garrisons in the newly annexed territories and the stipends to which they were entitled. Stipends were calculated on the basis of individuals’ positions in the army and their Muslim status. This was the so-called dīwān al-jund, and it had always been kept in Arabic.5 By extension, dīwān was also used to refer to







2 The most thorough discussion and evaluation to date of the narrative reports on the so-called language decree is Marie Legendre’s chapter in this volume, and it should be read in tandem with this one. A useful overview of the treatment of the so-called change of the state registers to Arabic in contemporary scholarship is also offered by al-Qāḍī, ‘The Names of Estates in State Registers’, p. 255 n. 1. Since then, publications by Nancy Khalek (‘Some Notes on the Representation of Non-Muslim Officials in al-Ǧahšiyārī’s Kitāb al-Wuzarāʾ wa-l-Kuttāb’) and Luke Yarbrough (‘Did ʿUmar b. ʿAzīz Issue an Edict Concerning Non-Muslim Officials?’) have touched on the topic as well. 3 For this process in Egyptian papyri and narratives, see Frantz-Murphy, ‘Land Tenure in Egypt in the First Five Centuries of Islamic Rule’. For the situation in Iraq, see Morony, Iraq after the Muslim Conquest, pp. 99–106. 4 For the increase in the quantity of administrative documents produced under the Arab rulers of Egypt, see Sijpesteijn, Shaping a Muslim State, pp. 74–75. Al-Maqrīzī (d. 845/1442) confirms in his Khiṭaṭ, i, p. 264: ‘First this dīwān [i.e. dīwān al-kharāj] was recorded in Damascus and Iraq in the Islamic period according to how it had been done before Islam’ (‫دون هذا الديوان في اإلسالم بدمشق‬ ّ ‫وأول ما‬ ‫)والعراق على ما كان عليه قبل اإلسالم‬. 5 Duri, ‘Dīwān i. The caliphate’.

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

the offices where these lists and registers were kept, such as the ‘tax office’ (dīwān al-kharāj), as well as other government offices such as the ‘dīwān of letters’ (dīwān al-rasāʾil), ‘of the seal’ (dīwān al-khatam), ‘of the postal service’ (dīwān al-barīd), and ‘of decorative textiles’ (dīwān al-ṭīrāz). All these offices obviously produced copious documentation to record the activities that they supervised or organized, but these were not necessarily (or mainly) in the form of the lists of the original registers. Whereas the anecdotes set in the Marwānid period clearly describe the translation of specific registers — that is, lists of information related to the land taxes, hitherto recorded in one of the (non-Arabic) administrative languages of the empire — into Arabic,6 later authors reporting these accounts as well as modern scholars basing their analyses on them sometimes confuse the restricted meaning of dīwān as a register with its meaning as an administrative office or even the chancery as a whole, which had become the term’s dominant connotation by the time these reports were written down. The result is the misinterpretation in many current studies as well as in some historical sources of the single act of translating fiscal ledgers as a decision to transform the multilingual administration of the immediate post-conquest caliphate into a monolingual Arabic chancery.7 Before connecting the anecdotes about the translation of the dīwān in historical narrative sources to contemporary documentary evidence that sheds new light on





6 This is evident, for example, from the fact that a specific sum of money and a designated period were allocated to the translation. The annual fiscal income of wilāyat al-Urdunn was granted to ʿAbd al-Malik’s secretary in Damascus for translating the registers in one year (al-Balādhūrī, Futūḥ al-buldān, p. 193), and the caliph named a specified period during which Ṣāliḥ b. ʿAbd al-Raḥmān was to translate the Iraqi registers (al-Balādhūrī, Futūḥ al-buldān, p. 301; Ibn al-Nadīm, al-Fihrist, ii, p. 141). Further, Ibn al-Nadīm treats the translation of the dīwān (‫‘ ث ّم نقل الديوان‬then the dīwān was translated’) as an event that immediately followed and was a logical consequence of the translation of a number of scientific works from Greek and Coptic into Arabic on the orders of Khālid b. Yazīd b. Muʿāwiya (d. 85/704): ‘And he ordered them to translate the books on science from the Greek and Coptic languages into Arabic’ (‫;)وأمرهم بنقل الكتب في الصنعة من اللسان اليوناني والقبطي إلى العربي‬ Ibn al-Nadīm, al-Fihrist, ii, pp. 139–40. This passage, and its consequences for our understanding of the so-called scientific translation movement, is discussed extensively in Saliba, Islamic Science and the Making of the European Renaissance, pp. 46–47. That it concerned the translation of lists or texts is also confirmed by references to the (trial) translation of some ‘lines’ (asṭur) as a first stage in the process and the described difficulty of translating numerals (see below). The interpretation of the dīwān as ledgers or accounts related to agricultural taxation is followed in some secondary studies (e.g. Morony, Iraq after the Muslim Conquest, pp. 53–54; al-Qāḍī, ‘The Names of Estates in State Registers’, p. 256; Saliba, Islamic Science and the Making of the European Renaissance, pp. 53–54). 7 For example, Philip Khuri Hitti’s translation (The Origins of the Islamic State, p. 301), ‘Greek remained the language of the state registers, until the reign of ʿAbd al-Malik ibn Marwān who in the year 81 ordered it changed’; the original Arabic is better translated as follows: ‘The dīwān remained in Greek (bi-l-rūmiyya) until ʿAbd al-Malik […] ordered its translation (bi-naqlihi)’. See also the mixing of lists and the language of the administration in general in the anonymous Chronicle of 1234: ‘Walid, the king of the Ṭayyāyê [= Arabs], ordered that in his chancery, i.e., the treasury, which these Ṭayyāyê call the dīwān, one should not write in Greek but in the Arabic language, because up to that time the ledgers of the kings of the Ṭayyāyê were in Greek’ (translated in Theophanes, The Chronicle of Theophanes Confessor, pp. 524, 596, Hoyland, Theophilus of Edessa’s Chronicle, p. 200 and Palmer, The Seventh Century in West-Syrian Chronicles, p. 209). For its impact on current historical discourse, see below.

45

46

p e tr a  m . s ij pe s t e i j n

these accounts, let us take a look at the various versions that appear in the historical texts. In her contribution to this volume, Marie Legendre offers a systematic overview and evaluation of the different accounts in the historical sources. My treatment, by contrast, generally follows the extensive presentations in al-Balādhurī’s (d. 279/892) Futūḥ al-buldān and al-Jahshiyārī’s (d. 331/942) Kitāb al-Wuzarāʾ wa-l-kuttāb with occasional references to other sources when they contain relevant deviations or additional information. The first person to whom measures to change the administration’s language to Arabic are ascribed is the famous governor of Iraq, al-Ḥajjāj (in office 75–95/694–713).8 Zādhānfarrūkh (d. 82/701), head of the Persian dīwān, was under the false impression that he and his fellow Persian-speaking secretaries were the only ones able to make the calculations in Persian that were needed for the fiscal accounts of the dīwān.9 But then al-Ḥajjāj’s scribe Ṣāliḥ b.  ʿAbd al-Raḥmān (d. 99/717) — a mawlā from eastern Iran who was bilingual in Arabic and Persian — showed that converting the writings of the dīwān from Persian into Arabic would cause him no trouble at all by translating a couple of lines. In response, Zādhānfarrūkh told his fellow secretaries: ‘Find a different occupation!’.10 The governor subsequently ordered Ṣāliḥ to translate the whole dīwān into Arabic in the year 78/697.11 Zādhānfarrūkh’s worry that he and his fellow secretaries would lose their jobs was surely justified, as it is claimed that Iraq’s subsequent secretaries were all Ṣāliḥ’s students.12

8 Al-Balādhurī, Futūḥ al-buldān, pp. 300–01; al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, pp. 33–34. Nancy Khalek (‘Some Notes on the Representation of Non-Muslim Officials in al-Ǧahšiyārī’s Kitāb al-Wuzarāʾ wa-l-Kuttāb’, pp. 510–15) has compared al-Jahshiyārī’s account with those of other historians, arguing that he was especially keen to connect Zādhānfarrūkh’s arrogance and haughtiness with Ṣāliḥ’s translation as an explanation for the downfall of the Persian scribal class. For an overview of the sources on the Arabicization of the dīwān in Syria and Iraq, see Sprengling, ‘From Persian to Arabic’, pp. 194–96. 9 When al-Ḥajjāj’s scribe comes to apo­logize to Zādhānfarrūkh for having been appointed over him, the latter utters the unforgettable response: ‘Don’t worry; he [the governor] needs me more than I need him’ (‫ي منّي إليه‬ ّ ‫ ;)ال تفعل فانّه أحوج إل‬al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 33. Al-Balādhūrī and Ibn al-Nadīm quote a similar retort from Zādhānfarrūkh when Ṣāliḥ expresses concern that with his own appointment the position of his mentor, Zādhānfarrūkh, has become superfluous: ‘Don’t believe ّ ‫ ;)ال‬al-Balādhurī, that; his need of me is greater than my need of him’ ( ‫ي منّي إليه‬ ّ ‫تظن ذلك هو أحوج إل‬ Futūḥ al-buldān, p. 300; Ibn al-Nadīm, al-Fihrist, ii, p. 140. 10 Al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 33 : ‫التمسوا مسكنا غير هذا‬. Cf. al-Ṣūlī (d. 335/947), Adab al-kuttāb, p. 192, where the more likely reading maksaban, livelihood, replaces maskanan, dwelling. Khalek discusses which sources contain Zadhānfarrūkh’s conversation with his colleagues (‘Some Notes on the Representation of Non-Muslim Officials in al-Ǧahšiyārī’s Kitāb al-Wuzarāʾ wa-l-Kuttāb’, p. 511). 11 ‘And al-Ḥajjāj ordered Ṣāliḥ to translate the registers into Arabic’ (‫وأمر الحجاج صالحا نقل الدواوين إلى‬ ‫ ;)العربيّة‬al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 33. The question is whether this happened while Zadhānfarrūkh was still alive or after he had died. Al-Balādhūrī and Ibn al-Nadīm place the event after Zadhānfarrūkh’s death during the revolt of Ibn al-Ashʿath (81–85/700–04). According to al-Jahshiyārī, it happened while Zadhānfarrūkh was still alive. Al-Ṣūlī is indecisive on this issue. 12 ‘The majority of the scribes in Iraq were Ṣāliḥ’s students’ (‫;)وكان عا ّمة كتّاب العراق تالمذة صالح‬ al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, pp. 33–34. See also al-Ṣūlī, Adab al-kuttāb, p. 192: ‘The scribes of the two Iraqs were all his servants and students’ (‫)وكان كتّاب العراقين كلّهم غلمانه وتالميذه‬. In Ibn

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

The next dīwān to allegedly switch to Arabic was that of Syria. In this account, the caliph ʿAbd al-Malik (r. 65–86/685–705) himself plays a role. The report is rather fantastical, ascribing the caliph’s decision to end the use of Greek in the administration to a disrespectful Christian Greek scribe at the Umayyad court’s chancery: after running out of ink, the scribe reportedly urinated in his inkwell to be able to continue to write a little while longer.13 Al-Balādhurī, a major source on the event, presents this incident as the immediate cause that triggered drastic administrative reforms, but other Arabic historians give also another reason. According to them, ʿAbd al-Malik was looking for a way to get rid of the Syrian scribe Sarjūn b. Manṣūr al-Rūmī (d. c. 80/700) who was in charge of the Greek dīwān (or all dīwāns) — either the registers themselves or the office that managed them.14 The caliph’s secretary Sulaymān b. Saʿd al-Khushanī offered a solution: change the language in which the registers are kept to Arabic and Sarjūn will be out of a job.15 It took Sulaymān a year to transfer all the registers in the dīwān to Arabic from their original Greek, but by 81/700 the job was done.16 When Sarjūn was confronted with the results of Sulaymān’s labours, he, like his Iraqi colleague, turned to the other Christian secretaries and sadly told them that they had better look for another way to make a living as their work in the chancery was finished.17

al-Nadīm’s Fihrist Zadhānfarrūkh’s son curses Ṣāliḥ: ‘May God uproot your descendants from this world as you have uprooted Persian’ (‫ ;)قطع هللا أصلك من الدنيا كما قطعت أصل الفارسيّة‬Ibn al-Nadīm, al-Fihrist, ii, p. 141, translated in Saliba, Islamic Science and the Making of the European Renaissance, p. 45. 13 Al-Balādhurī, Futūḥ al-buldān, p. 193. 14 Yarbrough, Friends of the Emir, pp. 73–74; Sprengling, ‘From Persian to Arabic’, p. 212. Compare the context of the event in al-Jahshiyārī (al-Wuzarāʾ wa-l-kuttāb, p. 35): ‘One day ʿAbd al-Malik asked him [Sarjūn] something, and he dragged his feet about it and was lax in dealing with it, so he [ʿAbd al-Malik] repeated it and urged him to take care of it, but he saw [only] neglect and disregard’ (‫فأمره‬ ‫)عبد الملك يوما بشيء فتثقال عنه وتواني فيه فعاد لطلبه وحثّه فيه فرأى منه تقريطا وتقصيرا‬. However, Khalek (‘Some Notes on the Representation of Non-Muslim Officials in al-Ǧahšiyārī’s Kitāb al-Wuzarāʾ wa-lKuttāb’, pp. 512–14) has shown that al-Jahshiyārī was biased against non-Muslim administrators such as Sarjūn. Indeed, al-Balādhurī and Ibn ʿAbd al-Rabbihi (d. 328/940), who also report the episode, do not mention ʿAbd al-Malik’s annoyance. Al-Ṣūlī (d. 335/947) and Ibn al-Nadīm do mention that ʿAbd al-Malik found Sarjūn lax (‫)رأى عبد الملك منه توانيا‬, which ‘angered’ the caliph, according Ibn al-Nadīm; thereafter, Sulaymān transferred the agricultural dīwāns into Arabic (Ibn al-Nadīm, al-Fihrist, translated in Saliba, Islamic Science and the Making of the European Renaissance, p. 46). According to al-Ṣūlī (Adab al-kuttāb, pp. 192–93), the caliph confessed to Sulaymān: ‘I cannot bear Sarjūn’s negligence (saḥb); can you not think of a device against him?’ (‫ )فما عندك حيلة في أمره‬Sulaymān responded: ‘I do; I will translate the accounts from Greek to Arabic’. Khalek claims erroneously that al-Ṣūlī’s version does not mention ʿAbd al-Malik’s annoyance. Theophanes Confessor (d. 818), perhaps unsurprisingly, asserts that Sarjūn ‘stood on close terms with Abimelech’ in The Chronicle of Theophanes Confessor, p. 510. The much later historian Ibn ʿAsākir (d. 571/1176) has ʿAbd al-Malik voice a more general vexation: ‘The control that the Christians exercise over the Muslims’ affairs has annoyed me’ (translation cited in Yarbrough, Friends of the Emir, p. 74). 15 However, Sarjūn, like his father, was probably also competent in Arabic and Aramaic besides Greek (Griffith, ‘The Manṣūr Family and Saint John of Damascus’, p. 30). 16 Wadād al-Qāḍī has traced this change in the names of estates recorded in the fiscal register (‘The Names of Estates in State Registers’). 17 Al-Balādhurī, Futūḥ al-buldān, p. 193; al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 35. All the same Sarjūn’s

47

48

p e tr a  m . s ij pe s t e i j n

The Egyptian historians Ibn ʿAbd al-Ḥakam (d. 257/871) and al-Kindī (d. 350/961) describe the linguistic transition of the dīwāns in Egypt. As in Iraq, the change in Egypt was ascribed to the governor. ʿAbd Allāh b.  ʿAbd al-Malik (in office 85–90/704–08 or 709) had been appointed Egypt’s governor by ʿAbd al-Malik’s son and successor al-Walīd (r. 86–96/705–15) with explicit instructions to replace all of the previous governor’s appointees.18 The transition of the local dīwāns into Arabic took place in the year 87/705–06.19 It is unlikely that the original dīwān had in fact been kept in Coptic, as al-Kindī claims.20 Even though Coptic was introduced as a chancery language under the Arabs relatively early on, Greek was still the main administrative language at this time (see below). The transfer coincided with a replacement of the two Christians who headed the dīwān, Athanasios from Edessa and Isaac, an Egyptian, with a certain Ibn Yarbūʿ al-Fazārī from Ḥims, whose name points to Arab descent or affiliation as a mawlā.21 Both Syriac and Greek chronicles place the administrative linguistic changeover in al-Walīd’s caliphate.22 These sources, too, contend that the transition in Syria resulted in the loss of jobs for Christian Greek-writing scribes.23 Ibn al-Nadīm (d. c. 385/995) and al-Maqrīzī (d. 845/1442) mention that some claim it was caliph Hishām (r. 105–25/724–43) who initiated the translation of the dīwān in Syria.24

son, the famous John of Damascus (d. before 136/754), initially also worked in the Umayyad administration in Damascus, albeit not at such a high position as his father and grandfather had had. 18 For the rivalry between the Marwānid caliphs and the Banū ʿUmar, see Sijpesteijn, ‘An Early Umayyad Papyrus Invitation for the Hajj’. 19 Al-Kindī, al-Wulāt wa-l-quḍāt, pp. 58–59: ‘And he [the caliph al-Walīd] appointed ʿAbd Allāh b. ʿAbd al-Malik over the registers, and they were copied in Arabic, whereas before they had been written in Coptic’ (‫ ;)وأمر عبد هللا بن عبد الملك بالدواوين فنسخت بالعربيّة وكان قبل ذلك تكتب بالقبطيّة‬Ibn ʿAbd al-Ḥakam, Futūḥ Miṣr, p. 122: ‘And he was the first to translate the registers into Arabic, when they had been in the local language [i.e. Coptic]’ (‫)وهو أول من نقل الدواوين إلى العربيّة وإنّما كانت بالعجميّة‬. Cf. History of the Patriarchs of the Coptic Church of Alexandria, iii, pp. 48, 54; al-Maqrīzī, Khiṭaṭ, i, pp. 264–65. 20 As does al-Maqrīzī, Khiṭaṭ, i, p. 264. 21 On Athanasios from Edessa, see Debié, ‘Christians in the Service of the Caliph’ and the Chronicle of 1234, pp. 202–04. Ibn Yarbūʿ’s name points to the Banū Yarbūʿ, a subtribe of Tamīm, and his nisba, al-Fazārī, refers to the Banū Fazāra, which originated in Najd. Andrew Marsham has argued that the tribal nisbas of Marwānid scribes indicate mawlā affiliations rather than ethnic Arabness (Rituals of Islamic Monarchy, pp. 160–61). 22 The Andalusi author Ibn ʿAbd al-Rabbihi mentions in his ʿIqd al-farīd both ʿAbd al-Malik and al-Walīd as the caliph under whom the fiscal registers were transferred into Arabic (quoted in Khalek, ‘Some Notes on the Representation of Non-Muslim Officials in al-Ǧahšiyārī’s Kitāb al-Wuzarāʾ wa-l-Kuttāb’, p. 511). Al-Maqrīzī’s addition that ‘the first to translate the dīwāns from Persian into Arabic was a certain al-Walīd b. Hishām b. Quḥzum’ (d. 220/835) seems to be based on an onomastic mix-up (Khiṭaṭ, i, p. 264). 23 Theophanes, in The Chronicle of Theophanes Confessor, pp. 524, 596, places the dismissal of Christian scribes in the years 706–07 and 758–59. However, he adds that they were rehired immediately as they were the only ones who could handle numerals. The episode is also mentioned by Michael the Syrian, by the Chronicle of 1234 (for references, see the edition of Theophanes), and arguably already by Theophilus of Edessa (d. 785) in Theophilus of Edessa’s Chronicle, pp. 199–200. 24 Translated in Saliba, Islamic Science and the Making of the European Renaissance, p. 46; al-Maqrīzī, Khiṭaṭ, i, p. 265.

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

It was also during Hishām’s caliphate, in the year 124/741–42, that the Khurasani chancery reportedly transitioned from Persian to Arabic. At the time, ‘most scribes [in the Umayyad chancery] were Zoroastrians and the accounts were in Persian’.25 Having decided that this situation was untenable, the governor of Iraq, Yūsuf b. ʿUmar al-Thaqafī (in office 120–26/738–44), wrote to Naṣr b. Sayyār, governor of Khurasan (in office 121–31/738–48): ‘No polytheist can hold a function in governance or administration’.26 Here, too, the implication is that a change in the language used in the dīwān would automatically lead to a reorganization of personnel. In response, Naṣr decided to transfer the records from Persian to Arabic.27 All in all, despite their differences, the reports about the language change in the Umayyad administrative offices have some common elements. They all place the event around the 80s/700s during the caliphate of ʿAbd al-Malik or that of his successor al-Walīd; the state registers were translated from local languages into Arabic, and henceforth produced in Arabic, on the ruler’s orders, resulting in a loss of jobs for the secretaries and scribes who had held appointments in the chancery. However, the three phenomena described in these accounts — the translation of the agricultural fiscal registers, the eventual dominance of Arabic in the administration as a whole, and the shift from local Christian to Arab and Muslim administrators — were, in fact, related but differentiated historical events and processes. They became conflated in some of the later accounts into one integrated sequence of events or are even presented as one, and this view is followed in some of the historical studies based on these sources. Besides the observation that the historical sources describe three distinct historical developments, rather than just one, there is a conceptual problem with considering these events as taking place at the same time and as being directly dependent on each other.

Rulers and Decrees Scholars have seen the decision to use Arabic for the state registers as a central feature of Marwānid administrative measures, especially those attributed to caliph ʿAbd al-Malik b. Marwān, and assumed that this decision underpinned the Marwānids’ Muslim and thus Arabic state-building policies. As Chase Robinson has put it: If social and economic forces were promoting Arabic from the 640s onwards, it was only in ʿAbd al-Malik’s reign that the political elite resolved to transform the language of God into the language of empire. […] In fact, the decision was radical. Christians and Jews alike had gotten along fine as linguistic schizophrenics, reading, writing and speaking a variety of languages and 25 ‫ ;وكان أكثر كتّاب خراسان إذ ذاك مجوس وكانت الحسبابات بالفارسيّة‬al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 64. 26 ‫ ;يأمره أن ال يستعين بأخد من أهل الشرك في أعماله وكتابته‬al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 65. 27 ‘The first one to translate the writings from Persian to Arabic in Khurāsān was the secretary Isḥāq b. Ṭulayq of the Banū Nashhal’ (‫وكان ّأول من نقل الكتابة من الفارسيّة إلى العربيّة بخراسان إسحاق بن طليق الكاتب‬ ‫ ;)رجل من بني نشهل‬al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 65.

49

50

p e tr a  m . s ij pe s t e i j n

scripts, translating (as it suited them) scripture from one language to the next. Muslims were altogether more ambitious; they were hardly the first linguistic imperialists, but they were the first to insist that the language spoken by God and those delegated by Him (caliphs, governors, commanders, etc.) should be the language of the mundane job of ruling. Robinson also argued that the speed with which the change to Arabic subsequently spread throughout the empire was largely geo­graphically determined: ‘The closer to the centre, the faster and more profound the change’.28 The idea is thus that the caliph ordered an absolute ideo­logically motivated switch to Arabic of the administration as a whole, but that this order was implemented fully and quickly only in Syria, whereas areas further away from the caliphal court initially maintained the use of local languages.29 The implication is that in theory the chanceries and all the documents they generated should have been in Arabic after the caliph’s order, but that in practice the change was slow to take effect and local administrators continued as they had always done to use their ‘own’ languages. This is why, even centuries later, the empire’s linguistic landscape remained diverse, not uniformly Arabic. Accordingly, documents in Greek, Coptic, Sogdian, Bactrian, and Pahlavi continued to be produced in Muslim chanceries throughout the empire for up to a century and a half after the decision to switch to Arabic was supposedly made. There are a couple of obvious problems with this model and its underlying assumptions. First, it does not fit our understanding of the functioning of the Umayyad state. Concrete policy instructions from the caliph would hardly be ignored by lower administrators or fail to reach officials located in the provinces for a century or more.30 In fact, the caliphal court maintained continuous contact 28 Robinson, ʿAbd al-Malik, pp. 125–26. Writing more recently, Lucian Reinfandt (‘Petosiris the Scribe’, p. 144) concurs: ‘The need to unify the administration of the provinces under one general chancery language, which was to be Arabic, also became pressing only around this time. This major move was made by the Umayyad caliph ‘Abdalmalik ibn Marwān with his famous decree from c. 705 ordering the provincial chanceries to change their languages from Greek to Arabic (in Syria and Egypt) and from Pahlavi to Arabic (as in Mesopotamia)’. Similar views are voiced by Andreas Kaplony in ‘Die Arabisierung der frühislamischen Verwaltung Syrien-Palästinas und Ägyptens’, p. 396 (‘Stellen die literarischen Quellen die Arabisierung und Islamisierung als Vollzug eines Befehls von ʿAbd al-Malik [685–705] dar’); by Maged Mikhail in From Byzantine to Islamic Egypt, p. 90 (‘Since the Arabization of the bureaucracy in the early eighth century’); and by Khalid Blankinship in The End of the Jihad State, p. 28 (‘Also, the caliph [ʿAbd al-Malik] decreed the replacement of Greek with Arabic as the official chancery language in Syria and Egypt’). 29 Robinson even explains the appearance of Persian in the tenth century ce by arguing that Persia’s distance from Damascus meant that the Arabic language policy of the empire had not reached it by that time as it had other areas (ʿAbd al-Malik, p. 126). According to Kaplony, the greater presence of Arabic speakers in Syria-Palestine than in Egypt explained the faster Arabicization of the administration in Syria-Palestine compared to Egypt (‘Die Arabisierung der frühislamischen Verwaltung Syrien-Palästinas und Ägyptens’, p. 400). 30 Such orders stand in contrast to symbolic orders and claims of control, which helped establish hegemony in low-power states that could not exercise complete control within their realms but which were not (intended to be) universally implemented (Richardson, ‘Before Things Worked’, pp. 17–21).

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

with the provinces through the sending and receiving of representatives, regular correspondence, and other forms of communication. The members of provincial elites maintained relations with the caliph’s court through institutions such as the wafd, a regular delegation of provincial Arab notables to the court.31 These notables used their contacts at court in times of crisis to obtain a safe haven or to intercede on behalf of themselves or others.32 The Umayyad court also exerted direct control over the provinces through its gubernatorial appointees. Qāḍīs were likewise appointed to the provinces from the capital.33 Umayyad princes serving as governors provided a direct channel of communication between the centre and the periphery.34 Within Syria, the caliph himself moved through different

In the case of the caliphate, such measures can be associated with Islamicizing (symbolic) policies. For example the caliph Yazīd b. ʿAbd al-Malik (r. 101–05/720–24) prohibited the use of images in 103/721; the ban was lifted three years later by Hishām (r. 105–25/724–43) having had little real effect, especially in the private realm, where, as Hishām declared, people are free to act as they want. ʿAbd al-Malik’s order to slaughter all pigs lead to incidental symbolic destruction of pig herds but not in the disappearance of pigs from the caliphate altogether (as Kruk, ‘The Saddest Beast’, observes, pigs remained present in medical and other scientific domains in the Muslim empire throughout the medi­eval period). For both of these cases, see Mikhail, From Byzantine to Islamic Egypt, pp. 117–18. Such symbolic orders were especially prone to being retrojected in history to the rule of especially pious caliphs, whose orders were deemed especially effective. See, for example, the debate about the so-called shurūṭ of ʿUmar (Cohen, ‘What Was the Pact of ʿUmar?’; Yarbrough, Friends of the Emir, pp. 102–03). ʿAbd Allāh b. ʿAbd al-Malik, the above-mentioned governor of Egypt who ordered the translation of the Egyptian dīwāns, was also supposedly the first one to forbid the wearing of the hooded cloak, the burnus, by (presumably) Coptic priests (Ibn ʿAbd al-Ḥakam, Futūḥ Miṣr, p. 122). Al-Mutawakkil’s (r. 232–47/847–61) decree prohibiting the employment of non-Muslims in high administrative posts except in the caliph’s own chancery was not driven by any anti-Muslim feelings on the part of the caliph; rather, it was a political measure to appease ʿulamāʾ and was dropped soon after (Yarbrough, Friends of the Emir, pp. 88–109, especially p. 107 n. 89). Obviously, the decree had not been directed at and did not affect the employment of non-Muslims in lower administrative offices. Such orders contrast starkly with the kind of administrative measures (fiscal measures, such as land surveys, censuses, registration and taxation of land, translations of fiscal registers, coin reforms) discussed here, which were implemented uniformly throughout the caliphate. See also the implementation of the miḥna under the caliphs al-Maʾmūn (r. 189–218/813–33), al-Muʿtaṣim (r. 218–27/833–42), and al-Wāthiq (r. 227–32/842–47), during which opponents of the miḥna’s central doctrine were effectively banished from public life. Qāḍīs, for example, barred such opponents from serving as legal witnesses in the different provinces as well as the caliphal centre (Tillier, L’invention du cadi, pp. 300–02). Cf. Yarbrough, Friends of the Emir, pp. 89–90. Compare also the impact of imperial legal policies on Egyptian procedural and public legal practice in the late and Byzantine Roman empire (Beaucamp, ‘Byzantine Egypt and Imperial Law,’ pp. 275–85). 31 For the wafd, see Tillier, ‘Représenter la province auprès du pouvoir impérial’ and Lev, The Adminis­ tration of Justice in Medi­eval Egypt, pp. 17–19. 32 For amāns provided by the caliph to rebels from the provinces, see Hagemann, ‘Was Muṭarrif b. al-Mughīra al-Thaqafī a Khārijite?’ and Sijpesteijn, ‘Closing Ranks’. 33 For qāḍī appointments, see Tillier, L’invention du cadi. 34 ʿAbd al-Malik’s brother and heir apparent, ʿAbd al-ʿAzīz, was appointed governor of Egypt by their father Marwān (Sijpesteijn, ‘An Early Umayyad Papyrus Invitation for the Hajj’). Other members of the caliph’s family were also appointed governors of Egypt (Sijpesteijn, ‘Egypt’s Connections in the Early Caliphate’). In Fars, Muʿāwiya used the appointment of his half-brother Ziyād b. Abī Sufyān (in office 45–53/665–73) to gain closer control over the province. Ziyād’s son ʿUbayd Allāh subsequently

51

52

p e tr a  m . s ij pe s t e i j n

‘capitals’ with his entourage in a form of itinerant kingship.35 In addition, the caliph frequently sent representatives both openly and in secret to check on governors and local elites.36 All in all, contacts between the caliphal court and the provincial administrations were frequent, regular, and elaborate, allowing quick and accurate communication of caliphal orders and supervision of their execution. Indeed, the caliphal administration had no problem with implementing other administrative reforms on an empire-wide scale and at a fast pace: land surveys and population censuses famously ordered by the Marwānid caliphs were carried out from al-Andalus to the Sawād and Armenia.37 The well-known coin reforms enacted first by al-Muʿāwiya in the 60s/680s and again by ʿAbd al-Malik and al-Walīd in the early eighth century ce were implemented across the empire. Muʿāwiya’s monetary reform, which introduced a new icono­graphy, extended from Syria and Egypt in the west to Iraq and Iran in the east and left traces in the form of coins, dice, and textual references despite being eventually abandoned.38 The later, more lasting coin reforms were applied across the empire, as is shown by the minting in remote provinces of post-reform coins — without images and with Arabic religious phrases — that matched those issued in the centre of the empire. Examples include the dirhams struck in ah 78 in Armenia, Azerbaijan, and Shaqq al-Taymara in the province of Isfahan, those issued in ah 79 in Anbir in Juzjan in the province of Khurasan, and the two silver dirhams struck in ah 81 and 90 in Oman.39 Even though the coin reform underwent several distinct phases and extended over a long

followed in his father’s footsteps as governor of Fars (in office 55–64/674–84). ʿAbd al-Malik appointed a relative, Khālid b. ʿAbd Allāh, governor of Basra (in office 71–73/690–93), later replacing him with his brother Bishr as governor of Kufa and Basra in 73/692 (Gundelfinger and Verkinderen, ‘The Governors of al-Shām and Fārs in the Early Islamic Empire’, pp. 276–78; al-Ṭabarī [d. 310/923], Taʾrīkh al-rusul wa-l-mulūk, ii, pp. 819 and 853). Muḥammad b. Marwān was governor of al-Jazīra and Armenia in 73–91/692–709. Maslama, ʿAbd al-Malik’s brother, served as governor of Qinnasrīn, was appointed governor of al-Jazīra, Armenia, and Azerbaijan in 91/710, and served as governor of Armenia 107–11/725–29 and again 112–14/730–32. Marwān b. Muḥammad also served as governor of Armenia and Azerbaijan from 114/732, 116/734, or 117/735 until he became caliph in 127/744 (Vacca, Non-Muslim Provinces under Early Islam, p. 151; Hawting, ‘Marwān II’; Rötter, ‘Maslama b. ʿAbd al-Malik b. Marwān’; Zetterstéen, ‘Muḥammad b. Marwān’). 35 Borrut, ‘Pouvoir mobile et construction de l’espace dans les premiers siècles de l’islam’. 36 For spies, see Silverstein, Postal Systems in the Pre-modern Islamic World. For delegations sent by the caliph to solicit opinions about the governor, see Hagemann, ‘Was Muṭarrif b. al-Mughīra al-Thaqafī a Khārijite?’. 37 Sijpesteijn, Shaping a Muslim State, pp. 94–95. For al-Andalus, see Manzano Moreno, ‘Conquest and Settlement’. For Armenia, see Vacca, Non-Muslim Provinces under Early Islam, pp. 195–200, especially 199–200. 38 Shaddel, ‘Monetary Reforms under the Sufyanids’, pp. 291–93. 39 Walder, Catalogue of the Arab-Byzantine and Post-Reform Umaiyad Coins. For the two post-reform coins struck in Oman, see Darley-Doran, History of Currency in the Sultanate of Oman. I would like to thank Peter Webb for this reference. For the post-reform coins struck in Armenia, Azerbaijan, and Isfahan, see Treadwell, ‘ʿAbd al-Malik’s Coinage Reforms’, pp. 374–75. In Ifrīqiya, coin types underwent significant changes, including the introduction of the shahāda in Latin, starting in 79/699–700, but Arabic legends appear on dirhams only in 97/715–16 in Ifrīqiya and in 103/721–22 in al-Andalus, and they follow Damascene models only from 114/732–33 onwards (Fenina, ‘L’arabisation

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

period, the models struck in Damascus were followed throughout the empire.40 The coins produced in the provinces reflected the changes introduced in and imposed by the capital of the caliphate in Damascus almost immediately. So why would there have been a disconnect between state-level policy and local practice in the case of the caliphate’s administrative language policy? The idea of such a discrepancy is even more remarkable when we consider that according to the anecdotes quoted above, in Iraq, Khurasan, and Egypt the decision to switch to Arabic was made by local governors. Moreover, as Marie Legendre argues in her chapter, these are the provinces that formed the core of Umayyad caliphal control. But even in the centres of provincial rule, documents in different languages continued to be produced for more than a century after the decision to switch to Arabic in the administration was supposedly made. Even if a change in the language of the administration would have involved more investment, organization, and infrastructure than did the striking of new coins (but not more than the implementation of regular land surveys and censuses), it does not make sense that it would have taken that long to accomplish.41 Producing letters, decrees, and other documentation in Arabic only would after all not have required much more than appointing sufficient scribes able to write in Arabic, no trivial skill, but also not one that could only be mastered after a century. When we take a closer look at the case of Egypt, where a wealth of extant sources in the form of papyri allow us to trace the so-called language switch in the chancery, the theory of an implementation lag becomes even more problematic to uphold. There are two reasons the implementation of caliphal measures in Egypt at this time was especially straightforward. First, during the Umayyad period, with the imperial capital in Damascus, Egypt was within easy reach of the caliphal court. The papyri support the observation that the caliph’s long arm reached Egypt not only in matters of public administration but even in more private matters such as participation in the caliph’s ḥajj caravan.42 Although coin specimens are du monnayage d’Ifrīqiya‘, pp. 149–52; Jonson, ‘A Numismatic History of the Early Islamic Precious Metal Coinage of North Africa and the Iberian Peninsula’, i, pp. 39–41). 40 Heidemann, ‘The Evolving Representation of the Early Islamic Empire and its Religion on Coin Imagery’. The innovations did not concern only the design but encompassed a new weight standard, with the old coins being withdrawn quickly (Treadwell, ‘ʿAbd al-Malik’s Coinage Reforms’, p. 358). 41 Sulaymān b. Saʿd al-Khushaynī allegedly asked ʿAbd al-Malik for one year’s worth of fiscal income from the province of al-Urdunn, 180,000 gold dinars, to cover the costs of translating the Syrian registers (Sprengling, ‘From Persian to Arabic’, p. 213). Sulaymān had served as governor of al-Urdunn before being made head of the dīwāns in Damascus. And the land surveys and censuses introduced by the Marwānids necessitated huge investments in human and infrastructural capital, which were realized. 42 The point stands even if some of the caliph’s interference in Egypt might have been prompted by a loss of control over this important province, which fell within the ruler’s direct sphere of influence. So, after a period of deliberate obstruction of the caliph’s policies under ʿAbd al-ʿAzīz’s governorship, the caliph ordered some members of the Egyptian elite to join his ḥajj caravan to Mecca in c. 101/719 (Sijpesteijn, ‘An Early Umayyad Papyrus Invitation for the Hajj’). A papyrus from the eighth century ce (late first or second century ah) records the presence in Egypt of the caliph and his armies, presumably after a period of lost control (Sijpesteijn, ‘Army Economics’). See also the orders for

53

54

p e tr a  m . s ij pe s t e i j n

lacking, glass weights suggest that even Egypt’s governor ʿAbd al-ʿAzīz (in office 65–86/684–705), whose rivalry with his brother ʿAbd al-Malik and the resulting obstruction of caliphal policies are well documented, was intimately involved in the centrally orchestrated monetary reforms.43 In short, more than any other province in the Umayyad caliphate, Egypt was closely integrated with Damascene policies. As mentioned earlier, this would have been even more the case under the governorship of ʿAbd Allāh b.  ʿAbd al-Malik, who was, after all, the caliph’s son, installed to forge a closer connection between the province and the caliphal court after the governorship of the aforementioned ʿAbd al-ʿAzīz; it was ʿAbd Allāh who allegedly implemented the language change. The second reason that Egypt was perfectly situated to implement caliphal administrative policies concerns the province’s geo­graphy. In Egypt, effective central control through the rapid dissemination and communication of governmental decisions was made possible by the Nile River. Once Egypt’s governor had decided to implement an administrative measure, it would have been easy and quick to impose the change from the capital, Fusṭāṭ, across the province.44 If the intention was indeed to switch the language of the dīwān to Arabic in as wholesale a manner as scholars have postulated on the basis of the historical reports, there is no reason the change would not have happened right away in Egypt, in particular. However the documentary evidence shows that it did not: Greek and Coptic were regularly used at all levels of the Egyptian administration long after the year 700 ce. My second objection concerns the pairing of the language switch with other administrative measures taken by ʿAbd al-Malik — the standardization of weights and, more specifically, the coin reforms.45 The coin reforms as reported in the historical sources are well attested in numismatic evidence from all over the empire. As is well known, ʿAbd al-Malik replaced images and phrases in Pahlavi, Greek, Latin, and other pre-Islamic languages on coins with Qurʾanic phrases in Arabic. The process of change was gradual, with progressive adjustments to the Sasanian, Byzantine,

building materials from the Upper Egyptian town of Ishqūh/Aphroditō for monumental Umayyad projects in Jerusalem and Damascus from the early second/eighth century (Morelli, ‘Legname, palazzi e moschee’). Finally, in a third/ninth-century letter, the sender, located in Egypt, requests information on the appointment of a new caliph and any news from Fusṭāṭ and Syria (Rāġib, ‘Lettres arabes II’, pp. 15–17). 43 Bacharach and Awad, ‘Rare Early Egyptian Islamic Coins and Coin Weights’, pp. 52–53. 44 Even given the regular back-and-forth between local administrators and the governor in Fusṭāṭ over the execution of administrative tasks, the degree of detail in the administrative instructions and supervision as recorded in the papyri found in Ishqūh/Aphroditō from the beginning of the eighth century ce (P.Lond. IV) is astounding. Editions of papyri are cited by the abbreviations used in the Checklist of Arabic Documents (https://www.naher-osten.lmu.de/isapchecklist) and the Checklist of Editions of Greek, Latin, Demotic, and Coptic Papyri, Ostraca, and Tablets (https://papyri.info/docs/ checklist). 45 See Robinson’s above-quoted conflation of the language reform with other centralizing and Islamicizing measures. See also Duri, ‘Dīwān i. The Caliphate’ (‘ʿAbd al-Malik initiated the policy of Arabization in the dīwāns, ṭirāz and currency’) and Kaplony, ‘Die Arabisierung der frühislamischen Verwaltung Syrien-Palästinas und Ägyptens’, p. 396.

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

and Roman models.46 The association of the coinage reforms with the change of the dīwān to Arabic is easily made: both the coins and the administration switched from Greek and Pahlavi to Arabic, and both changes supposedly reflect efficient centralized governance.47 But the coin reforms were not so much about making the coins Arabic as they were about making them Islamic, or rather Qurʾanic. The switch was inspired, as Luke Treadwell has convincingly argued, by the Marwānid caliphs’ desire to showcase their piety in response to the religious claims made by rebels such as ʿAbd Allāh b. al-Zubayr (d. 73/692), as expressed, inter alia, on his coins.48 The religious character of the new coins is confirmed by the reactions of religious scholars, who debated whether it was permissible to place Qurʾanic quotations on coins that might be handled by non-Muslims.49 Considering the coin reforms as a religious rather than a linguistic change also makes sense of the remarkable Latin shahāda inscribed on North African coins long before the inscriptions were switched to Arabic.50 Muʿāwiya’s short-lived decision to produce coins without crosses while otherwise following existing Byzantine models also points to a religious motive.51 Conversely, the translation of the records in the dīwān is never linked to religious arguments (such as that Arabic would be more religiously appropriate than Greek or Pahlavi) in the sources.52 That the coin reforms and the language reform of the dīwān fall in different categories is also supported by the fact that ʿAbd al-Malik’s coin reforms — which garners much more attention in our sources — is always treated in a separate section in our texts and is not associated with a switch to Arabic at all.53 In sum, the coin reforms were religiously motivated and were also interpreted as such. 46 Heidemann, ‘The Evolving Representation of the Early Islamic Empire and its Religion on Coin Imagery’. 47 See, for example, Treadwell, ‘ʿAbd al-Malik’s Coinage Reforms’, p. 358: ‘The coinage and language reforms supplied the administrative mechanisms required for the operations of centralised rule — effective communication across immense distances by means of monolingual bureaucracy and efficient distribution of resources from a single imperial centre’. 48 Treadwell, ‘ʿAbd al-Malik’s Coinage Reforms’, pp. 377–79. 49 Treadwell, ‘Qur’anic Inscriptions on the Coins of the Ahl al-Bayt’, pp. 52–55. 50 Fenina, ‘L’arabisation du monnayage d’Ifrīqiya’, p. 117; Treadwell, ‘ʿAbd al-Malik’s Coinage Reforms’, p. 362. 51 Shaddel, ‘Monetary Reforms under the Sufyanids’, pp. 265–66. For this episode, see also below, n. 53. 52 Even in historical accounts that connect ʿAbd al-Malik’s decision to strike his own coins, to his dislike of the designs of the protocols produced in Egypt, it was not the Greek language that the caliph took offense at but the fact that the sheets at the beginning of the papyrus rolls had the name of Jesus identified as God and a cross in the place of the basmala (‫األقباط تذكر المسيح في رؤوس الطوامير وتنسبه الى‬ ‫ وتجعّل الصليب مكان بسم هللا الرحمان الرحيم‬. . . ‫)الربوبية‬. It was those symbols and expressions he removed, but there is no mention that he forbade the use of Greek (al-Balādhurī, Futūḥ al-buldān, p. 240). 53 Theophanes’ remark (The Chronicle of Theophanes Confessor, pp. 509–10) that ʿAbd al-Malik’s reformed coins were unacceptable because they were ‘of a new kind that was never produced before’ as they did not have the ‘Roman imprint’ on them has generally been interpreted as referring to the change of the icono­graphy on the coins, not to the use of Arabic (Treadwell, ‘ʿAbd al-Malik’s Coinage Reforms’, p. 377). This is confirmed by the fact that in the late antique Roman and Sasanian empires the image on coins had become unchangeable, functioning as a symbol of the coins’ value (Heidemann, ‘The Standing Caliph Type’, p. 25). Earlier, the inhabitants of Syria had supposedly

55

56

p e tr a  m . s ij pe s t e i j n

My final argument concerns the mechanics of the early Muslim state’s administration or imperial rule. It draws on comparative empire studies, which studies the governance of populations of diverse linguistic, ethnic, and religious backgrounds over vast geo­graphical areas. This field has shown the fallaciousness of trying to ‘clean up’ the fuzziness of historical processes in such environments by replacing the complex interplay of trial and error and the back and forth of administrative developments that typically involve government bureaus and officials across the governmental hierarchy with the tidy assumption of a single order and a top-down command structure.54

Multiple Languages in the Administration The Arab Muslim administration continued to produce documents in languages other than Arabic long after the alleged change in the language of the chancery. Let us consider a few examples of these documents, starting in Egypt. The first document is a so-called protocol (Fig. 1.1).55 This is a sheet that was attached to the beginning of a papyrus roll, stating where and when, and under which ruler, it was produced.56 In this case, the caliph Sulaymān (r. 96–99/715–17) and the governor of Egypt ʿAbd al-Malik b. Rifaʿa (in office 96–99/715–17) are mentioned. The protocol is written in Greek and Arabic and contains more or less the same text in both languages. It was issued in ah 98–99, that is, 716–17 ce. This was seventeen years after the language change was supposedly initiated in Syria and twelve years after the same change was introduced in Egypt. The protocol was an official document produced under the direct supervision of the Muslim administration in Egypt, and a controllable one, too. The protocol was added to the papyrus roll in the papyrus ‘factory’, and the state had a monopoly over papyrus production. The production of papyrus, moreover, took place in the Nile Delta very close to the provincial capital of Fusṭāṭ.57 Even so, the protocol gives the same importance to Greek as it does to Arabic. The second example consists of three lists of prisoners held in a jail in the Fayyūm oasis (Fig. 1.2).58 The lists are in Greek and name the prisoners together with the refused to accept coins struck by Muʿāwiya’s administration from which the crosses had been removed (Maronite Chronicle, p. 32). See also the claim, mentioned above (n. 52), that ʿAbd al-Malik was inspired to strike his own coins so as not to be dependent on Byzantine ones with (potential) blasphemous depictions or expressions. The caliph and his officials, by contrast, did not express any concern about which language was used, only what was written on the coins. 54 See, for example, Duri, ‘Dīwān i. The Caliphate’: ‘The Arabization of the dīwāns was effected in the empire by stages’. Cf. the references cited above in n. 28. 55 This document is kept in the Austrian National Library under inventory number P.Vindob. AP 3976 and it is published in CPR III no. 65. It was found in the Fayyūm. 56 Grohmann, From the World of Arabic Papyri, pp. 32–39. 57 Malczycki, ‘The Papyrus Industry in the Early Islamic Era’. 58 Morelli, ‘Dalle prigioni dell’Arsinoite’. The lists are currently kept in the Austrian National Library Papyrus Collection under the inventory numbers P.Vindob. G 12030 + 25958; 19591 recto and 19591 verso.

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

57

Figure 1.1. An ArabicGreek protocol dating to 98–99/716–17 published by Adolf Grohmann as CPR III no. 65 and currently kept in the Austrian National Library P.Vindob. AP 3976. © Österreichische Nationalbibliothek, Papyrus­ sammlung. Reproduced with permission.

crimes they committed. Some of the men were in prison because they had stolen clothes, wine, or a bull, but three of them had been imprisoned because they had burnt the land-tax registers. These documents were produced by a government office connected to the prison less than a hundred kilometres from Fusṭāṭ in the eighth century ce. The Muslim administration in Egypt produced plenty of other Greek documents, such as administrative letters and decrees, ledgers related to the registration of people and agricultural property, and orders and receipts for taxes to be paid in money and kind as well as labour requisitioned by the Muslim state.59 These documents were generated by offices at all levels, from the governor in Fusṭāṭ to local district chanceries in Upper Egypt, throughout the eighth century, by Egyptian administrators, but also in the name of Arab Muslim officials.60 59 For examples of tax registers, see CPR XXII 17 (dated to 789–90 ce, Fayyūm/Arsinoite), 21 (796–97 ce), 23 (788 ce), 25 (742–43 ce), 35 (750–69 ce), and 45 (early eighth century); P.Lond. IV 1449 (711 ce, Ishqūh/Aphroditō). For examples of tax orders, see Gonis and Schenke, ‘Two Entagia from Cambridge’, p. 373 (726–27 ce, Anṣīna/Antinoopolis); CPR XIX 26 (718 or 733 ce, al-Ashmūnayn/Hermopolite); CPR XXII 7 (751–52 ce), 8 (729–30 ce), and 9 (729 ce); SPP VIII 1199–1200 (744 or 759 ce). See also the notary Paulos who produced legal settlement documents and other administrative papers (albeit not necessarily in a governance setting) in the first third of the eighth century in Ihnās/Heracleopolite (Kovarik, ‘Der Heraklopolitische Notar Paulos’). 60 Particularly famous are the Greek letters reportedly sent by Qurra b. Sharīk (in office 90–96/709–15) to Basileios, pagarch of the district of Ishqūh/Aphroditō in Upper Egypt (P.Lond. IV). See also the

58

p e tr a  m . s ij pe s t e i j n Figure 1.2. Eighth-century Greek list of prisoners with their place of origin and the crime they committed. Published by Federico Morelli, ‘Dalle prigioni dell’Arsinoite’, and now kept in the Austrian National Library P.Vindob. G 19591 recto. © Österreichische Nationalbibliothek, Papyrus­sammlung. Reproduced with permission.

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

Figure 1.3. Arabic protocol dating to the financial governorship of ʿĪsa b.  Yūnus (in office 227–29/841–44) found in the Fayyūm. It was published by Adolf Grohmann as CPR III no. 156 and is now kept at the Austrian National Library P.Vindob. AP 4343 recto. © Österreichische Nationalbibliothek, Papyrussammlung. Reproduced with permission.

The latest dated Greek administrative document was produced at the very end of the eighth century, in 796 or 797 ce.61 This papyrus does not contain more than a list of personal names, amounts of money, and the date according to the hijri calendar. However, according to Federico Morelli the many surviving eighth-century Greek papyri show that novel Greek formulae were being introduced in this period and that these formulae reveal an active command of the Greek language amongst the scribes who worked in the Muslim administration, possibly in interaction with Byzantine practices.62 Wherever the inspiration for new administrative practices in Greek came from, the important point is that Greek scribal and secretarial forms continued to develop long after the Muslim administration supposedly switched entirely to Arabic. Knowledge and application of the Greek language in the chancery offices continued even longer. The following document is written in Arabic and Greek (Fig. 1.3 and 1.4).63 On one side (Fig. 1.3) it contains another protocol text, this time written entirely in Arabic. It mentions Abū Jaʿfar Ashinās, who served as governor of Egypt from ah 219 to 230 (818–44 ce), and the director of finances ʿĪsā b. Yūnus (in office 227–29/841–44). This means that the protocol was issued under al-Muʿtasim (r. 218–27/833–42) or al-Wāthiq (r. 227–32/842–47). The back side of the protocol (Fig. 1.4) was used to write an Arabic letter related to a land survey (miṣāḥa).

entagia, tax orders, written in name of the pagarch Nājid b. Muslim (CPR XXII 8–9) in 729–30 ce and in the name of ʿAbd al-Malik b. Yazīd (CPR XXII 7) in 751–52 ce. From the pagarch Nājid b. Muslim a Greek administrative letter is also preserved, dating to 727–50 ce (CPR XXII 10). 61 CPR XXII 21. 62 Morelli, Documenti greci per la fiscalità e la amministrazione dell’Egitto arabo, p. 53. 63 This papyrus is currently kept in the Austrian National Library under inventory number P.Vindob. AP 4343. The Arabic text of the protocol which appears on the recto of the papyrus is published in CPR III, no. 156. An edition of the Arabic letter and the Greek phrases on the verso of the papyrus is being prepared by Federico Morelli and myself.

59

60

p e tr a  m . s ij pe s t e i j n Figure 1.4. Unpublished Arabic letter mentioning land measurement in relation to land taxes with two lines of Greek written in the left margin at a 90° angle to the Arabic text. The papyrus is kept in the Austrian National Library P.Vindob. AP 4343 v and is currently being prepared for publication by Federico Morelli and myself. © Österreichische Nationalbibliothek, Papyrussammlung. Reproduced with permission.

In the margin of the Arabic letter a line of Greek has been added, as well as some more words of official record-keeping related to land surveys. This document shows that in mid-ninth-century Egypt, that is to say two hundred years after the establishment of Arab rule, there were apparently scribes employed in Muslim administrative offices who were still actively using Greek, even if at this point solely for formulaic expressions not extensive texts, alongside Arabic.64 And it was not just Greek that was used by the Muslim administration in Egypt. The use of Coptic exploded in the eighth century ce, and this boom continued into ninth. The gradually increasing presence of Coptic in the public sphere was a process that had begun already before the arrival of the Arabs. Coptic legal documents, (fiscal) receipts, testaments, and petitions had started to circulate and to be produced in monasteries, on great agricultural estates, and in village communities in the sixth century ce.65 The adoption of Coptic in 64 Around the same time (210 ah/825 ce) a complete Greek document was still produced, however, in a monastic context probably in Fayyūm/Arsinoite or Ihnās/Heracleopolite (Berkes, ‘The Latest Iden­tified Greek Documentary Text from Egypt’, p. 244). 65 Fournet, The Rise of Coptic, pp. 76–144; Cromwell,

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

the chancery was, in that sense, just a matter of time. A seventh-century Coptic petition sent to a Byzantine official before the Arab conquest forms a literary and symbolic bridge between the private and public uses of Coptic.66 The earliest Coptic administrative documents produced in the administration date from the time of ʿAmr b. al-ʿĀṣ, who conquered Egypt and acted as its first governor (in office 21–25/642–45 and 38–43/658–64). Two declarations in Coptic confirm that all men of fourteen years and older from the village of Tjinela in the district of al-Ashmūnayn/Hermopolite have been registered for the purpose of some imposition by the Arab administration, fiscal or corvée labour.67 At this point, however, the use of Coptic for administrative purposes was still minimal and taking place at the village level in environments in which Coptic (and Greek) were spoken and written. Nevertheless, the Muslim administration definitely embraced if not actively supported the use of Coptic. Some of the larger Egyptian monastic centres were engaged in the purposeful training of Coptic- and Greek-writing scribes for administrative tasks, and the production of administrative documentation, for example, must have furthered the position of Coptic.68 Coptic scribes, too, launched new scribal practices as products of an active and ambitious Coptic scribal culture.69 The result was that official administrative documents produced in Coptic were an integral part of the Muslim administration of Egypt.70 See, for example, the eighth-century tax demand note originating in al-Ashmūnayn/Hermopolite in Middle Egypt that was issued by Yazīd son of ʿAbd al-Raḥmān, an Arab official, to Severus son of Bane. Yazīd’s Arabic seal appears at the bottom of the document. On the back are traces of an earlier official Arabic text in Kufic script.71 Another Coptic administrative document lists requisitioned workers and deliveries of trees issued in 718 ce by Rāshid son of Khālid, an Arab official appointed over the district of

Recording Village Life, pp. 10–13; Garel, Héritage et transmission dans le monde égyptien; Dekker, Episcopal Networks and Authority in Late Antique Egypt. 66 Delattre and Fournet, ‘La première pétition en copte’, pp. 180–82. 67 P.Lond.Copt. 1079. It is not clear whether the Coptic declarations are related to the Greek lists of names that appear on the other side. Cf. Papaconstantinou, ‘Administering the Early Islamic Empire’, pp. 61–63. 68 Palombo, ‘The Christian Clergy’s Islamic Local Government’; Berkes, ‘The Latest Documentary Greek Text from Egypt’. 69 Cromwell, Recording Village Life, ch. 6; Cromwell, ‘A Village Scribe on the Eve of Change’. 70 See the Coptic documents related to official business that Papas, pagarch of Edfu, produced and received in the 670s ce (Boud’hors, ‘Situating the Figure of Papas, Pagarch of Edfu at the End of the Seventh Century’, pp. 68–71). Jennifer Cromwell has published Coptic fiscal documents related to the pagarchy of Attias in the 690s: ‘Coptic Texts in the Archive of Flavius Attias’. And such Coptic administrative documents continued to be produced, like Greek ones, albeit in smaller numbers than Arabic ones, into the ninth century ce (Berkes, Delattre and Vanthieghem, ‘A Late Coptic Tax Receipt from the Egyptian National Library’; Berkes, Delattre and Vanthieghem, ‘A Ninth-Century Coptic Tax Refund Document. Reedition of CPR IV 197’; Berkes, ‘The Latest Documentary Greek Text from Egypt’). 71 P.Ryl.Copt. 117. Other eighth-century Coptic fiscal documents include Gonis and Schenke, ‘Two Entagia from Cambridge’, no. 2 (729–30 ce, Anṣīna/Antinoopolis); Schenke, ‘Rashid ibn Chaled and the Return of Overpayments’, pp. 202–09 716 or 731 ce, provenance al-Ashmūnayn/Hermopolite.

61

62

p e tr a  m . s ij pe s t e i j n

Ihnās/Heracleopolite. This document also contains an Arabic seal identifying the sender at the bottom of the document (Fig. 1.5).72 See also the Coptic version of a decree from the Egyptian governor Qurra b. Sharīk (in office 90–96/709–14) that Eline Scheerlinck has identified. She argues that the document was translated or paraphrased at the village level in accordance with an active policy of translating documents in order to communicate government orders to the local populations at whom they were directed.73 The quantity of such Coptic administrative material increased dramatically in the eighth century. The final document I wish to discuss is a bilingual Coptic-Arabic demand for taxes to be paid, which also has a Greek note at the top. Until the first decades of the eighth century the Muslim chancery produced bilingual Arabic-Greek documents on which the same text was twice, in Greek and in Arabic. These were typically protocols, fiscal orders, and receipts.74 But the Muslim chancery seems also to have issued Coptic-Arabic bilingual fiscal documents, although these are much less frequently attested than are Greek-Arabic papyri.75 The document contains an order from Hishām b. Ziyād, pagarch of Upper Ashmūn, to Ioannes son of Isidoros of the monastery of Apa Apollo in Bawīṭ to pay two dinars in poll tax for the year 136/753.76 The fiscal demand note is written in duplicate in Coptic and Arabic, although the two texts do not constitute literal translations of one another but rather a paraphrasing. The Greek note at the top presumably confirmed that the payment was completed. It fulfilled a practical function and was added by a specialist Greek-writing scribe who also knew enough Arabic and/or Coptic to check the amount listed in the demand note. The Coptic and Arabic texts, by contrast, serve a very different purpose, communicating ideo­logical rather than practical concerns. What this document shows above all is the self-evident intertwining of the three languages of Arabic, Coptic and Greek in Muslim administrative offices in Egypt in the middle of the eighth century ce, more than a century after the Arab takeover.77 72 Garel, ‘Une demande de recensement du pagarque Rāšid b. Ḫālid’. 73 Scheerlinck discusses this document (P.Ryl.Copt. 277) in chapter 5 of her ‘Solidarity and Self-Interest’. Cf. Richter, ‘Language Choice in the Qurra Dossier’, pp. 203–08, 216–18. 74 For protocols, see CPR III. For bilingual tax demand notes, see, for example, P.Ness. 60b (dated to 54/674), 61–63 (55/675), 64–65 (56/676), 66 (57/676–77), and 67 (70/689–90); P.Heid.Arab. I a–l (all dated to 91/709–10, Ishqūh/Aphroditō). For bilingual receipts, see, for example, SB 16 13018 (714 ce, al-Ashmūnayn/Hermopolite); SB 18 13771 (677 or 707 ce, Ihnās/Heracleopolite). Cf. Richter, ‘Language Choice in the Qurra Dossier’; Kaplony, ‘Die Arabisierung der frühislamischen Verwaltung Syrien-Palästinas und Ägyptens’; Sijpesteijn, ‘Arabic-Greek Archives’. 75 See the Coptic receipt with an Arabic subscript in P.Ryl.Copt. 214 quoted in Delattre and others, ‘Écrire en arabe et en copte’, p. 172 n. 2, and the Coptic tax receipt with a Coptic summary of the Arabic text published in Berkes and Vanthieghem, ‘A Late Coptic Tax Receipt from the Egyptian National Library’, p. 13 n. 2. I would like to thank Eline Scheerlinck for pointing me to this reference. 76 Clackson and Sijpesteijn, ‘A Mid-Eighth-Century Trilingual Tax Demand Related to the Monastery of Apa Apollo at Bawit’, pp. 102–19. 77 See also the scribal exercises in Coptic, Arabic, and Greek produced by one and the same scribe at the end of the eighth century ce (Berkes and Younes, ‘A Trilingual Scribe from Abbasid Egypt?’; Berkes, ‘Bemerkungen’, pp. 375–76), the Coptic receipt that is verified by Greek signatures and has an Arabic registration mark at the top (Delattre, ‘Langue et sources documentaires coptes’; I would

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

Figure 1.5. Coptic request for information on workers and trees from the pagarch of Ihnās/Heracleopolis that Rāshid b. Khālid issued in 718 ce. It was most recently published by Esther Garel ‘Une demande de recensement du pagarque Rāšid b. Ḫālid’ and is kept in the Austrian National Library P.Vindob. K 8313. © Österreichische Nationalbibliothek, Papyrussammlung. Reproduced with permission.

63

64

p e tr a  m . s ij pe s t e i j n

In other parts of the Islamic empire, different languages were used but the principle of multilingual administration was the same. In the year 721 ce, ʿAbd al-Raḥmān son of Ṣubḥ, who worked for the governor of Khurasan, Saʿīd b. ʿAbd al-ʿAzīz (in office 104/722–23), wrote a letter to Dhēwāshtīch, the ruler of Panjikent (r. 705?–22 ce). The latter had fled from the incoming Arab armies to Mount Mugh, where this letter was found.78 The letter was written on ʿAbd al-Raḥmān’s behalf by a Sogdian scribe, who plays with the Arabic and Sogdian languages to convey subtle and not so subtle messages about the hierarchy between the Arab ruler and the local one.79 First, the scribe begins the letter with the following address: ‘From ʿAbd al-Raḥmān b.  Ṣubḥ to Dhēwāshtīch, king of Sogdiana and ruler of Samarkand: praise be to God’. With this phrase, he deviates from the way in which Sogdian letters to Dhēwāshtīch were generally composed: in such letters the name of Dhēwāshtīch comes first in the address, and the sender calls himself ‘your humble servant’. By naming the Arab official and doing so before naming the addressee, Dhēwāshtīch, the scribe places the former above the latter in the hierarchy of power despite his respectful description of Dhēwāshtīch as king of Sogdiana and ruler of Samarkand. Second, the blessing ‘praise be to God’ was used when Muslim officials addressed non-Muslims.80 Muslims were greeted in a subtly different way: ‘I praise God for your sake’. The choice of invocation demonstrates awareness of the addressee’s religion. This document thus reveals two important things. It shows that the scribe had mastered both Arabic and Sogdian to such a degree that he could use them creatively to serve very specific goals. And it proves that the Muslim administration maintained a Sogdian scribe. On the other side of the correspondence, Dhēwāshtīch also had at his disposal scribes with knowledge of both Arabic and Sogdian. Most of his correspondence was obviously conducted in Sogdian, but a couple of years before the letter just discussed, between 717 and 719 ce, Dhēwāshtīch sent a petition written in Arabic to the then-governor of Khurasan and Sijistan, al-Jarrāḥ b. ʿAbd Allāh al-Ḥakamī (in office 99–100/718–19).81 As seen above, Khurasani Arab governors had at their disposal scribes who could handle Sogdian, and Dhēwāshtīch could thus have

like to thank Cecilia Palombo for pointing me to this reference), and the trilingual documentation of fiscal orders (Delattre and Vanthieghem, ‘Un ensemble archivistique trilingue à Strasbourg’). When Nājid b. Muslim was pagarch of the Fayyūm, his office, too, produced documents in Arabic, Coptic, and Greek (Berkes and Vanthieghem, ‘Notes on the Careers of Nāǧid b. Muslim and ʿAbd al-Malik b. Yazīd’, pp. 158–59) Cf. below, nn. 100 and 110. 78 Only a photo­graph of this document has been preserved: Livshits, Sogdian Epi­graphy of Central Asia and Semirech’e, p. 90. For the analysis of these documents on which the following is largely based, see Huseini, ‘Thinking in Arabic, Writing in Sogdian’. 79 How to interpret the tone of the letter has been debated. On the one hand, Ilya Yakubovich has argued that the letter conveys ‘friendly persuasion’ (‘Mugh 1.i. Revisited’); on the other, Vladimir Livshits sees in it ‘a stern reprimand’ (Sogdian Epi­graphy of Central Asia and Semirech’e, p. 92). Both quotations are taken from Yakubovich’s review of Livshits’s translated volume. 80 Grob, Documentary Arabic Private and Business Letters on Papyrus, p. 40; Garosi, ‘Cross-Cultural Parameters of Scribal Politesse in the Correspondence of Arab-Muslim Officials’. 81 Kratchkovsky and Kratchkovskaya, ‘Drevnejshij arabskij dokument iz Srednej Azii’.

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

composed his plea in that language. The fact that he did not do so demonstrates that the languages in these letters were not chosen for reasons of practical communication; instead, particular languages served particular goals. The use of Arabic in Dhēwāshtīch’s petition served obvious rhetorical purposes, signalling humility and accommodation, on the one hand, and rapprochement and affinity, on the other. The governor Saʿīd b. ʿAbd al-ʿAzīz had another Sogdian letter sent in his name. The document is very fragmentary because it was reused to line the scabbard of a dagger. It was also found on Mount Mugh and is now in the Hermitage Museum.82 In this epistle, Saʿīd b.  ʿAbd al-ʿAzīz addresses the chief (?) priest of Samarkand. The Arab Muslim administration in Samarkand thus produced documents in local languages executed by trained scribes who worked in the Muslim chancery. These example documents, which were produced in different parts of the Islamic empire and using different languages in the eighth and even ninth centuries ce show that there was no blanket decision to change the language of the administration to exclusively Arabic. These documents were not the result of local practices lingering on in remote places in active or passive contravention of orders from the centre. The Muslim empire, like empires before and after it, did not strive for a monolingual administration.83 This does not mean that Muslim rulers were unaware of the power of language as a tool of domination, governance, and communication. It also does not mean that they did not stimulate the use of Arabic in the administration and beyond. Both the use of Arabic and the use of other languages formed part of the rulers’ linguistic policy. But they also had to accommodate the practicalities of administering a large and diverse empire. For both of these reasons — linguistic ideo­logy and pragmatism — they actively promoted and facilitated a multilingual administration that produced bilingual and multilingual documentation, translations and redactions of decrees, correspondence, and other documents, and that was paid for and endorsed by the Arab administration.

Two Phases of Multilingualism The Muslim empire maintained a multi-tiered linguistic administration that began in the time of the great conquests and continued for centuries afterwards. This multilingualism is evident in letters, administrative correspondence, and fiscal orders for payment that the Muslim administration issued to non-Arabic-speaking 82 Livshits, Sogdian Epi­graphy of Central Asia and Semirech’e, pp. 37–39. 83 On multilingual empires and societies in general, see Papaconstantinou ‘Introduction’, pp. 1–2. For specific cases, see, on the Persian empire, Khurt, ‘State Communications in the Persian Empire’; on the Ptolemaic empire, Crespo, ‘The Linguistic Policy of the Ptolemaic Kingdom’; on the Greek world, Crespo, ‘Language Policies’; on the Roman Empire, Rochette, ‘Language Policies in the Roman Republic and Empire’; on the Sasanian Empire, Morony, Iraq after the Muslim Conquest, pp. 64–66; on the Ottoman empire, Darling, ‘Ottoman Turkish’; on the Habsburg empire, Hochedlinger, Mat’a, and Winkelbauer, Verwaltungsgeschichte; Gilbert, ‘Multilingual Administrative Dynasties in Habsburg Spain’, pp. 173–74, 186. Cf. Fournet, ‘The Multilingual Environment of Late Antique Egypt’ and Reinfandt and Tost, ‘Aller Anfang ist schwer’.

65

66

p e tr a  m . s ij pe s t e i j n

interlocutors. Arguably, these publicly disseminated writings were subject to obvious rhetorical concerns. But it is also evident in the registers kept by the Muslim administration for its own use. The fiscal ledgers and lists of landed and other properties, taxpayers, prisoners, and corvée labourers that contained all the practical information that the administration needed to run its day-to-day operations continued to be produced in local languages as well as in Arabic. This multilingual administration existed right from the moment the Arabs inaugurated their rule. It was based on ideo­logical as well as practical reasoning. In fact, the chancery was bilingual from the time of the caliph ʿUmar b. al-Khaṭṭāb (r. 13–23/634–44). Historical sources report that one register (dīwān) was kept in Arabic to keep track of those entitled to receive a stipend, while another dīwān, kept in Greek in Syria and in Persian in Iraq, was used to calculate the taxes due on people’s possessions, especially agricultural estates.84 This makes perfect sense. Registers recording landholdings and population numbers that had been in use in the pre-Islamic period were maintained and continued in their original languages, whereas the newly created register of Arab soldiers, their families, and their dependents was kept in Arabic. Other reports inform us that daily checks were made on deaths and births in the garrisons to update this Arabic register.85 Documents seemingly produced by or in relation to this dīwān are indeed written in Arabic. Two Arabic papyri dating to the first or second/seventh–eighth century contain names of Arab men with their addresses in Fusṭāṭ and dependents respectively, supposedly registrations related to the dīwān.86 Two other papyri listing Arab men and their mawālī in units seem to be similarly related to this register.87 Other Arabic documents record how stipends were distributed on the basis of enrolment in the dīwān.88 Conversely, documentation related to the agricultural fiscal registers continued to be kept in whatever language had been in use before the coming of the Arabs. Above I discussed several examples of the continued function of non-Arabic registers for fiscal purposes in the Muslim chancery in Egypt.89

84 ‘In Kufa and Basra there used to be two dīwāns, one of them in Arabic to keep count of people and their stipends, and this is the one that ʿUmar had already set up, and another one for the purpose of [recording] properties, which was in Persian. In Syria it was the same: one register was kept in Greek, the other in Arabic’ (‫ولم يزل بالكوفة والبصرة ديوانان أحدهما بالعربيّة إلحصاء الناس واعطياتهم وهذا الذي كان‬ ‫)عمر قد رسمه واألخر لوجوه األموال بالفارسيّة وكان بالشام مثل ذلك أحدهما بالروميّة واألخر بالعربيّة‬. Al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 33. The Persianized names of estates belonging to owners with Arabic names that appear in Arabic agricultural fiscal ledgers are a relic of these ledgers’ Middle Persian origins; al-Qāḍī, ‘The Names of Estates in State Registers’, pp. 273–74. 85 Sijpesteijn, ‘Army Economics’. 86 A list of Arab men and their addresses in Fusṭāṭ: Sijpesteijn, ‘A Seventh/Eighth-Century List of Companions from Fustat’. A list of households identified by names of men and women, and including numbers of household members: P.Ryl.Arab. I V 1. 87 P.Mil.Arab. 6; P.Mird. 33. Cf. Sijpesteijn, ‘Army Economics’, p. 256 n. 28; Qadi, ‘Death Dates in Umayyad Stipends Registers’, pp. 68–79. 88 Sijpesteijn, ‘Army Economics’. I am preparing for publication another hitherto unpublished Arabic papyrus letter dating to the first/seventh century that discusses the distribution of ʿaṭāʾ and rizq. 89 Some scholars have identified certain Greek lists on papyrus as being related to the dīwān of Arab

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

The simultaneous use of both Arabic and local languages to record information in the registers of the chancery was thus driven by practical considerations. There were also, however, ideo­logical reasons to maintain multiple languages in the dīwān. In the year 22/643, during the Muslim takeover of Egypt and only four years after the Arab armies initially arrived in Egypt, a bilingual Arabic-Greek receipt was issued for sixty-five sheep transferred to an Arab army unit on its way to Upper Egypt, officially for the purpose of collecting taxes.90 A second goal of this mission was surely to pacify and implement the conquest in the area south of the capital in Fusṭāṭ. This thesis is confirmed by the name of the person who issued the receipt: ʿAbd Allāh b. Jābir, the well-known general in charge of taking control of Upper Egypt. The Greek and Arabic parts of the papyrus were written by two different scribes, who are named at the bottom of the texts (the Greek part was written by the notary and deacon Johannes, the Arabic part by a man by the name of Ibn Ḥudayd). Although the contents of the two texts are comparable, they use different phrases and expressions, displaying two distinct documentary and even legal traditions.91 To give just one example, whereas the persons are identified by their names and patronymics in the Arabic text, the Greek text instead gives their titles, even Arabic titles, which are transcribed into Greek, with amīr, for example, appearing as amiras. So the new rulers actively used Arabic together with the Greek in their administrative documentation and official communication immediately after the conquest. Why were both languages used in this receipt preserved on papyrus? It has been argued that the Arabic version was meant to be shown to other Arab units passing through the district, who might otherwise have demanded additional contributions. The question is whether Arabic was in fact needed for this purpose. Would there not have been anyone in such a future Arab army or administrative unit who could read Greek? The Arab administration issued plenty of exclusively Greek receipts at this time for goods and services delivered to representatives of the Arab military (some were even issued in the name of ʿAmr b. al-ʿĀṣ who died ca. 42/663), and these were presumably to be used in similar circumstances — that is, to show to Arab officials or army commanders coming through and demanding further goods, accommodation, or other benefits.92 In other words, reliance on Greek did not constitute an obstacle to communication with representatives of the Arab regime. The reason that Arabic was included in the receipt — albeit in a subordinate position vis-à-vis the Greek text, which appears at the top of the papyrus — presumably lies in the role of ʿAbd Allāh b. Jābir. He was a highly placed member of the new ruling elite, the personification, one could say, of the new power

soldiers and settlers but I do not think these lists record anything but ‘regular’ taxes (Sijpesteijn, ‘Army Economics’, p. 27). Cf. Qāḍī, ‘Death Dates in Umayyad State Registers’, pp. 68, 75. 90 Grohmann, From the World of Arabic Papyri, pp. 113–15. 91 See the extensive discussion of this document in Sijpesteijn, Shaping a Muslim State, pp. 67–68. For a discussion of how Arabic papyrus formulae that diverge from local Egyptian legal traditions can be traced to pre-Islamic Arabian customs, see also Khan, ‘The Opening Formula and Witness Clauses in Arabic Legal Documents’. 92 SB 7/8, pp. 9748–60.

67

68

p e tr a  m . s ij pe s t e i j n

and dominance present in Egypt. And he was keen and able to express this position, inter alia through the language that represented the new ruling group: Arabic.93 The Greek text, on the other hand, made the receipt easier to use and accept by local administrators who were used to operating in Greek. One could thus argue that Greek served a more practical function here. Given the limited number of Arabs present in the Egyptian countryside, it simply did not make sense to issue a receipt in Arabic only. Although this argument of pragmatism makes very good sense, the very acknowledgement of practical concerns — making the life of the local administrators easier — also displays an ideo­logical motive. Immediately following the conquest of Egypt the Muslim administration made some drastic changes to the administrative organization of the province. Most importantly for our purposes, the duxes, heads of the five Egyptian administrative districts, were stripped of their military duties and partially divested of their fiscal responsibilities. An Arab army officer, amīr, was appointed alongside each dux to oversee the transfer of taxes from the districts to Fusṭāṭ.94 It is safe to assume that these amīrs operated in Arabic. Arabic was thus injected to local-level administration directly after the conquest, albeit within the limits of direct (oral?) communication, and the Egyptian population simply had to cope with the change. Interpreters and other bilingual officials could manage Arabic communications on a small scale but were obviously unable to effect a fully fledged administration in Arabic at this time. In other words, in some areas Arabic was used for ideo­logical reasons, to announce the presence of a new political reality, whereas in others Arabic was the language of operation of the administrators (such as the amīrs assisting duxes and pagarchs) installed by the new rulers, and somehow a workable way of communicating was created. Simultaneously, such Arabic-writing and -speaking local administrators could provide a seemingly unintrusive way for the new rulers to assert a presence (and some form of control) in the lower layers of the administration extending into the Egyptian countryside. Meanwhile, other languages were maintained both for practical reasons, because it was not possible to replace the entire existing administrative system with an Arabic one, and to make a gesture of goodwill to local interlocutors. The latter motive was especially important in the first phase of the establishment of Arab rule, when the Arabs’ presence was neither certain nor entirely secure.95 The Sogdian-language letter that the Arab governor of Khurasan, Saʿīd b. ʿAbd Allāh, sent to the local Sogdian ruler Dhēwāshtīch served two rhetorical

93 Garosi, ‘Cross-Cultural Parameters of Scribal Politesse in the Correspondence of Arab-Muslim Officials’, pp. 73–74. 94 Morelli, Documenti greci per la fiscalità e la amministrazione dell’Egitto arabo, introduction; Legendre, ‘Neither Byzantine nor Islamic?’. 95 Sijpesteijn, ‘New Rule over Old Structures’, pp. 186–90. Fenina (‘L’arabisation du monnayage d’Ifrīqiya’, pp. 151–52) makes the same point in relation to the peculiar form that the coin reform took in Ifrīqiya. Not wanting to endanger a delicate balance of political and military domination, the Arab conquerors changed the language on the coins only gradually, first introducing Islamic legends in Latin and in general changing the coins later than was done in the central provinces.

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

purposes. On the one hand, the use of Arabic formulae translated into Sogdian and the form of address used for the Sogdian ruler emphasized his subordinate and momentarily precarious position, holed up in a castle surrounded by Arab troops. On the other, the use of Sogdian was a deliberate sign of deference made with the diplomatic aim of winning Dhēwāshtīch over to the side of the Arabs. The other Sogdian letter mentioned above was a display of respectful homage to the chief Zoroastrian priest of Samarkand, whose high sociopolitical status was evident to the Umayyad governor. It fits well with the deference early Egyptian governors paid the Coptic patriarchs at a time when the Coptic Church’s support was still crucial.96 The same attempt at conveying political superiority through the use of Arabic while simultaneously showing respect to local powers lies behind the bilingual Arabic-Greek and Arabic-Coptic communications preserved on papyrus. The Arab rulers thus used multiple languages for external communications with foreign rulers such as Dhēwāshtīch, in the same way that the Sasanid chancery, for example, maintained a corps of multilingual scribes both for the purpose of producing letters in the languages of their correspondents abroad and for internal communication with their subject populations. The purpose of the different languages used in these duplicate documents was not so much to serve the practical objective of information conveyance — translators and interpreters with enough linguistic knowledge to transmit the contents of these writings would have been readily available — but rather to showcase a multilingually run administration in which languages were used for specific governance and communication purposes. Nevertheless, Arabic served as the language of power in these communications. In the earliest dated bilingual Arabic-Greek document from 22/643, discussed above, the Arabic text comes second after the Greek. This is also the order in which the two languages appear in the earliest bilingual protocols (see below). Soon thereafter, however, bilingual Greek-Arabic and Coptic-Arabic documents attest to a reversed order, with Arabic preceding the other language. Indeed, the systematic subordination of local languages to Arabic that is visible already in the earliest bilingual papyri eventually led to the domination of Arabic over all other languages in Egypt across all literary genres and linguistic domains, but this development took more than three centuries and was not a straightforward, ‘natural’ process.97 As Antoine Borrut discusses in his chapter in this volume, processes of linguistic subordination and assimilation are motivated by the presence of a language of political or cultural dominance. Surely this principle should be used to explain the eventual complete Arabicization of Egypt. Conversely, strong counterforces were simultaneously at work, and these should also be taken into account when explaining the spread of Arabic in this region specifically.98 96 Sijpesteijn, ‘New Rule over Old Structures’, pp. 188–90. 97 Papaconstantinou, ‘Introduction’, pp. 5–6. 98 Other regions experienced their own peculiar interactions with Arabic as the dominant language of power, demonstrating the varied linguistic landscape of the Islamic empire. Note, for example, the resilience of Berber in the Maghrib and the revival of Middle Persian in Iran but not in Iraq. See also Arietta Papaconstantinou’s warning against colonial perceptions of the relation between indigenous

69

70

p e tr a  m . s ij pe s t e i j n

In the first phase of Muslim rule in Egypt (c. 640–720 ce), the Arab authorities actively used multilingual documents to appeal to the different communities under their dominion. As Muslim rulers sought to establish lasting control over the areas they had conquered, maintaining stability through the continuity of existing practices in the languages that were in use was an important concern. This might have been reinforced by the cooperation between the Arab administration and monastic or clerical centres in schooling and production of scribal and clerical expertise.99 Once Muslim rule was more firmly established, the need to communicate with the population in multiple languages simultaneously might have subsided, but in the absence of a ‘monolinguistic administration decree’, a multilingual administration continued to be in place for at least another century. During this period, the names of scribes listed at the ends of documents, linguistic features, and scribal practices all point to the existence of separate scribal traditions. In other words, the various scribes who produced the Coptic, Greek, and Arabic texts did so within their respective scribal traditions. Bilingual documents continued to be produced into the eighth century ce, long after the first generations of Coptic and Greek scribes who had been in service at the time of the conquest had died. This means that the multiple scribal traditions were continually reproduced in successive generations of Muslim administrations. Later on, scribal trends and methods of writing converged with ever greater numbers of bilingual scribes being active in the administration, but — and this is the important point — the multilingual administration persisted.100 A multilingual administration requires the maintenance of a complex infrastructure: scribes need to be trained in multiple languages, not only in the practice of writing and using letters, numbers, and so on but also in the literary culture of the languages in which they operate.101 The creative skills of Greek and Coptic chancery

and incoming linguistic cultures and her call for a more integrated view of ‘language, ethnicity, religion, social status and political allegiance’ (‘Introduction’, p. 7). 99 See above, n. 68. 100 Garosi, ‘Cross-Cultural Parameters of Scribal Politesse in the Correspondence of Arab-Muslim Officials’, pp. 75–81; Berkes and Younes, ‘A Trilingual Scribe from Abbasid Egypt?’; Delattre and Vanthieghem, ‘Un ensemble archivistique trilingue à Strasbourg’. See also the application of the Islamic hijra calendar in eight- to ninth-century Greek and Coptic documents (Berkes, ‘The Latest Identified Greek Documentary Text from Egypt’). Cf. nn. 77 and 110. 101 See also George Saliba’s reconstruction of the reasons for and timing of translations of Middle Persian and Greek scientific and philosophical works into Arabic. Following Ibn al-Nadīm, Saliba locates the first translations in the Umayyad chancery and argues that the handbooks, instruction manuals, and reference works containing information on the arithmetical calculations and astronomical computations that were crucial for agricultural fiscal accounts were translated together with the accounts (Islamic Science and the Making of the European Renaissance, pp. 52–56). But even the (at first sight) complicated calculations in Egyptian papyrus fiscal accounts convey a straightforward and simple system, as Federico Morelli has shown (Documenti greci per la fiscalità e la amministrazione dell’Egitto arabo, pp. 145–46). In Egypt, surface measures used to calculate agricultural taxes most commonly used fractions based on 3 (i.e. 1/3, 1/6, 1/12, 1/24, 1/48), especially when the denominator was larger than ten. Other kinds of calculations, such as those concerning money, used fractions based on 2 (1/2, 1/4, 1/8, 1/16, 1/32, 1/64). Combinations were also possible, as fractions based on 2 can be

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

scribes in Islamic Egypt and Sogdian ones in Islamic Khurasan have already been noted. They were not simply applying frozen forms mechanically, without having actively mastered the relevant languages. Far into the eighth century, the employees in the chancery were trained to be fully capable secretaries who could make use of their linguistic skills in flexible and creative ways. Innovations continued to be introduced within each linguistic sphere.102 Documentary evidence illustrates how the scribes learned their trade. There are plenty of Greek and Coptic scribal exercises from the eighth century in which secretaries practice official epistolary and contractual formulae. It is not only the expressions they use that place these exercises in the context of the chancery but also the names of the officials that appear in the exercises.103 The trainee scribes copied out real letters, tax receipts, or other official administrative documents when practicing their writing skills. It seems that scribes were trained mostly in the private sphere, as opposed to academies or schools. Potential secretaries were initially apprenticed to their fathers and would then work with other scribes for further training.104 Even in more institutionalized settings, such as monasteries, which seem to have played a role in the training of scribes for the Muslim chancery, scribal education probably did not follow a fixed curriculum.105 However informal they were, these patterns of schooling and training maintained highly developed scribal practices, including innovative trends in the local languages. Indeed, the educational infrastructure within which scribes were trained, in combination with the limited presence of Arabic-speakers outside the larger settlements, kept local languages prominent in the Muslim chancery. There simply was no need or opportunity to add Arabic systematically to the training programme. Native scribes working in the central chanceries of the provincial capitals would have been the first to experience the need and opportunity to add Arabic to their palette. Beyond the exercise of basic linguistic writing and translation skills, George written as fractions based on 3 (for example, 1/2 can also be written as 1/3 + 1/6). More complicated calculations that required the kind of arithmetic work Saliba refers to might have been applied in other domains; an example, which Saliba, too, mentions, is the calculation of Islamic inheritance shares. See also Ibn Qutayba’s (d. 265/879) recommendations for essential background reading in his Adab al-kuttāb (Saliba, Islamic Science and the Making of the European Renaissance, pp. 54–55). 102 See above, nn. 62 and 69. 103 For the schooling of scribes in early Islamic Egypt, with references to Greek, Coptic, and Arabic scribal exercises, see Sijpesteijn, Shaping a Muslim State, pp. 131–32. Two papyri dating to the end of the eighth century ce contains writing exercises in Arabic, Coptic, and Greek from a chancery context (Berkes and Younes, ‘A Trilingual Scribe from Abbasid Egypt?’; Berkes, ‘Bemerkungen’, pp. 375–76). Cf. above, nn. 77 and 100. For the teaching of Greek within the Islamic empire, see also Mavroudi, ‘Greek Language and Education under Early Islam’. For the training of Greek scribes, see Berkes, ‘The Latest Identified Documentary Greek Text’, and for Coptic scribes, see Cromwell, Recording Village Life, ch. 6; Cromwell, ‘A Village Scribe on the Eve of Change’; Palombo ‘The Christian Clergy’s Islamic Local Government’, Ch. 3. 104 Cromwell, Recording Village Life; Palombo, ‘The Christian Clergy’s Islamic Local Government’. This seems to have been true in other scholarly domains as well; see Thomann, ‘From Lyrics by al-Fazārī to Lectures by al-Fārābī’. I would like to thank Antoine Borrut for this reference. 105 See above n. 68.

71

72

p e tr a  m . s ij pe s t e i j n

Saliba has shown that they would also have been engaging with Arabic literary and scientific texts as they participated in Arabic court culture.106 The scribes who, according to the stories quoted above, were responsible for the translation of the fiscal registers in the provincial capitals of Egypt, Iraq, and Syria are good examples of this trend. These locals had all established connections to the new Arab rulers through patronage (walāʾ) and were generally fully integrated and acculturated into Arab culture and society. The previously mentioned Ibn Yarbūʿ al-Fazārī from Ḥims, who was brought to Fusṭāṭ to replace Isaac and Athanasios in the Egyptian chancery, was a mawlā of the Banū al-Dhiyāl.107 Ṣāliḥ b.  ʿAbd al-Raḥmān from Sijistan was a (son of a) mawlā of the Banū Tamīm.108 Sulaymān b. Saʿd al-Khushanī was the son of an Arab Christian from Palestine and might himself have been Christian. However, his nisba points to his family’s integration into the Khushayn, a major subtribe of Quḍāʿa. Interestingly, whereas Zādhānfarrūkh and Sarjūn lost their positions as heads of the chanceries of Iraq and Syria, respectively, their sons, who were, no doubt, more acculturated to the Arab Muslim environment than their fathers had been, remained employed in the Muslim chancery, albeit not at such a high level.109 Their multilingualism enabled these scribes to experiment with the kind of hybrid documents that the Sogdian-Arabic scribe of governor Saʿīd b. ʿAbd al-ʿAzīz produced. Egyptian papyri similarly show how greeting and dating formulae, and other phrases based on Arabic expressions were introduced in Greek and Coptic documents produced in the Muslim chancery.110 The chanceries of smaller administrative districts likewise maintained multilingual administrations. In early eighth-century lists of expenses for the administration of the Upper Egyptian town of Ishqūh/Aphroditō, payments are recorded for Greek and Arabic scribes who obviously worked together in the pagarch’s offices but represented distinct linguistic skills.111 At this time, Muslim chanceries did not employ separate secretaries to produce full translations for bilingual documents; rather, secretaries specialized in specific languages handled documentation and communications in different languages not on the basis of any ideo­logical division of languages but as a matter of course.

106 See above, n. 101. 107 Al-Kindī, al-Wulāt wa-l-quḍāt, p. 59; History of the Patriarchs of the Coptic Church of Alexandria, iii, p. 48. According to al-Azdī’s (d. 321/933) Jamharat al-lugha, the Banū al-Dhiyāl belonged to the great Arabian tribe of the Banū Saʿd. For the continued employment of mawālī in the Umayyad chancery, see also Duri, ‘Dīwān i. The Caliphate’. 108 Al-Balādhurī, Futūḥ al-buldān, pp. 300–01. 109 Sprengling, ‘From Persian to Arabic’, p. 195. Zādhānfarrūkh may have continued working in the dīwān in Basra after Ṣāliḥ replaced him as head of the chancery. 110 For the dating formulae, see e.g. CPR XXII 8, 9 (both 111 ah/729–30 ce; provenance Ihnās/ Heracleopolite), 21 (180 ah/796–97 ce), 25 (125 ah/742–43 ce). For the greeting formulae, see Garosi, ‘Cross-Cultural Parameters of Scribal Politesse in the Correspondence of Arab-Muslim Officials’, pp. 75–81. In a Coptic tax receipt dated 273/886–87 the phrase ‘God willing’ (‫)إن شاء هللا‬ is expressed in Coptic (Berkes and Vanthieghem, ‘A Late Coptic Tax Receipt’, 15). 111 Sijpesteijn, Shaping a Muslim State, p. 236 n. 100. See also the references above to multilingual scribes operating in the Muslim administration in nn. 74–79, 100.

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

Personnel Rather than Language Change If ʿAbd al-Malik did not issue a decree to change the language of the caliphate’s chanceries to Arabic, what do the anecdotes in the historical sources tell us about the greater story of the Arabicization of the Islamicate empire? The accounts seem rather unambiguous. They speak of naqala, nasakha, ḥawwala, or jaʿala al-dīwān bi-l-ʿarabiyya, which all refer to switching, transcribing, or translating the fiscal registers to Arabic. The loss of jobs for local scribes associated with this process of changing the dīwāns linguistically also suggests a rather intrusive measure. Finally, the stories also contain references to the continued use of the local Greek and Pahlavi numeral systems and the supposed technical expertise involved in keeping the accounts and doing the calculations needed to determine fiscal dues. Both the translation story set in Iraq and that which takes place in Syria bring up the claim that Arabic does not permit the efficient writing of the numbers and fractions necessary for fiscal calculations.112 The story of a switch in the language of the administrative offices fits well the circumstances of the eighth and ninth centuries ce, when the historical reports were drawn up. By this time, Arabic had made huge headway everywhere in the empire; the administration functioned to a large degree in Arabic, and large parts of the population seem to have done so as well, at least in their day-to-day business. In other words, Arabic was very present and increasingly so. At the same time, Greek and Pahlavi numerals remain widely attested in documents produced in this period by Muslim chanceries, even in Arabic-language documents.113 Medi­eval historians attempting to explain these developments looked for a historical figure to whom this seemingly seminal transition could be ascribed (a well-known historio­graphical ruse). Not surprisingly, they arrived at the Marwānid caliphs, whose drastic administrative reforms and spectacular building programmes were, of course, well known. However, the same reports that ascribe the translation of the fiscal registers to the caliphs ʿAbd al-Malik, al-Walīd, and al-Hishām and their

112 Under the entry for the year 758–59 ce, Theophanes notes (The Chronicle of Theophanes Confessor, p. 596): ‘In this year the Arabs maliciously expelled the Christians from government chanceries for a short time, but were once again obliged to entrust the same duties to them because they were unable to write numbers’. The story of the translation to Arabic of the Iraqi dīwāns in Ibn al-Nadīm’s Fihrist describes Zādhānfarrūkh’s son questioning Ṣāliḥ about how he will render the fractions 1/10 and 1/20 in Arabic (relayed in al-Qāḍī, ‘The Names of Estates in State Registers’, p. 256 n. 3). 113 For Sasanian numerals, see Frye, ‘Sasanian Numbers and Silver Weights’. I would like to thank Khodadad Rezakhani for pointing me to this reference. For Pahlavi documents from the Islamic period using such numerals, see Weber, Ostraca, Papyri und Pergamente, pp. 227–28. Greek letters continued to be used long into the medi­eval period in Greek, Coptic, and Arabic papyri. For Greek numerals in Greek papyri from the Islamic period, see Morelli, Documenti greci per la fiscalità e la amministrazione dell’Egitto arabo. For Arabic documents with Greek numerals, see Werner Diem, Arabische Steuerquittungen des 8. bis 11. Jahrhunderts. So-called Indian numerals have still not been identified with certainty in the papyro­logical record (pace Grohmann, From the World of Arabic Papyri, p. 89). Arabic writers start using them from the ninth century onwards, especially for mathematical calculations.

73

74

p e tr a  m . s ij pe s t e i j n

governors in the provincial capitals also provide important elements of the real, more complex answer. There was no single, decisive ‘language decree’; rather, the authorities issued multiple proclamations to introduce innovative administrative practices, translation projects, and other measures to reform the administration throughout this formative period.114 In fact, the chronicles’ reports for the period from 77/697 to 141/759 record several instances that involved the translation of fiscal registers in various provinces — instances that modern scholars have turned into a single event incorrectly ascribed to one ruler or another (perhaps further described as having been executed in stages). The fact is that the reports in our histories refer not to one caliph’s decree or to a single watershed moment in the history of the administration of the caliphate but to a series of pronouncements made by a number of different administrators and rulers. Governors introduced administrative reforms seemingly on their own initiative, but they were surely also influenced by developments elsewhere in the empire. This does not mean that they were responding to some central order, rising like a wave in Damascus and flooding the provinces as it rolled outwards. Again the papyro­logical evidence nicely illustrates the point. Protocols, the papyrus sheets added to the beginnings of newly produced rolls, have already been invoked to demonstrate the absence of a decree to change the administrative language to Arabic. The protocols nevertheless reveal an unmistakable linguistic development, in which clearly discernible advances in the status of Arabic vis-à-vis other languages can surely be connected to administrative changes. However, these moments of change do not pivot around the caliphates of ʿAbd al-Malik or al-Walīd or the governate of ʿAbd Allāh b. ʿAbd al-Malik, which the literary accounts connect to the language switch. Immediately after the Arab conquest, the protocol texts continued to be produced exclusively in Greek, as they had been before the conquest, albeit now with the names of Arab caliphs and governors. The first change took place at the end of the seventh century ce, when the protocols started to contain the same text in Greek and Arabic. The oldest dated bilingual protocol dates to 75/694–95 and originates in Ansina/Antinoopolis.115 In it, the Arabic text is written interlinear with the Greek, but every time following the Greek in a manner similar to the order of texts in the earliest bilingual Arabic-Greek papyrus, which is dated to 22/643 (discussed above). Over time, this order is reversed, with the Arabic text taking up the primary place. But it took several decades, until 114/732, before the protocols were written exclusively in Arabic.116 That it was the well-known forceful Egyptian finance minister ʿUbayd Allāh b. Ḥabḥāb (d. after 123/741) under whose 114 Similarly, the treatment of non-Muslims differed from caliph to caliph (Yarbrough, ‘Muslim Rulers, Non-Muslim Subjects’, pp. 368–69, 372). Ziyād b. Abīhi, governor of Iraq for Muʿāwiya (in office 45–53/665–73), is also reported to have made substantial adjustments to the fiscal administration of the province (al-Balādhurī, Futūḥ al-buldān, p. 464; Sprengler, ‘From Persian to Arabic’, p. 185; Duri, ‘Dīwān i. The Caliphate’). 115 P.Gascou 27a. For further bilingual Greek-Arabic protocols, see, amongst others, CPR III 4, 6–8, 10, 23, 34, 36–38, 43, 46. 116 CPR III 108.

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

governorship Arabic protocols were introduced is probably no coincidence.117 Interestingly, the switch to exclusively Arabic protocols took place during the long caliphate of Hishām, a time that also saw the translation of the fiscal registers from Persian into Arabic in Khurasan and, according to some sources, in Syria. Nevertheless, the protocol texts, especially those produced exclusively in Arabic, show great internal variation — even among texts produced under the same ruler.118 The papyri also show that decisions made about the role of Arabic in government documents differed across genres. The decisions depended on the documents’ function, handling, degree of public exposure, and audience, and they varied over time and according to circumstances. Public correspondence, letters, decrees, and even fiscal demand notes and receipts were treated differently from documents produced for internal use, such as fiscal registers, lists, and calculations, which, in turn, differed from official records such as protocols. The documents illustrate variety in attitudes towards and treatment of Arabic. There is still another reason to think that the anecdotes recorded in the historical sources do not tell the story of an absolute switch from local languages to Arabic in the dīwān. Rather, what they describe is a different, probably at least as radical, change in the administration. According to al-Balādhurī, after ʿAbd al-Malik had given orders to switch the dīwān in Syria from Greek to Arabic, he called in the Greek clerk Sarjūn to inform him of the change. The latter then ‘left the caliph sad and annoyed, and meeting certain Greek clerks, he said to them: “Seek your livelihood in any other profession than this, for God has cut it off from you”’. The Greek chronicler Theophanes (d. 818) wrote in his history that the decision to change languages resulted in the dismissal of Christian notaries, who were, however, rehired quickly as the Arab scribes who replaced them were unable to handle Greek numerals.119 The account set in Iraq also has the Persian scribe Zādhānfarrūkh urge his colleagues to find another job. Even in the tenth-century account of al-Kindī, the event is discussed in the framework of a change of personnel in the Egyptian chancery. What these sources are referring to is a shift in the composition of the officials working in the Muslim administration. After an initial period of some fifty years during which local administrators had continued to man most offices in the new Muslim administration, especially at the local level, Muslim Arab officials belonging to the new ruling elite started to replace indigenous Christian ones around 80/700. This development forms the context of the reported lamentations about evil Arab administrators who mistreat indigenous scribes by cutting off their income. This was, after all, a real worry: the appointment of acculturated Arabic-speaking Muslim administrators in the lowest offices in the administration did in fact push local officials out.120 And

117 Khoury, ‘ʿUbayd Allāh b. Ḥabḥāb’; Sijpesteijn, Shaping a Muslim State, pp. 212–13. 118 Grohmann, From the World of Arabic Papyri, p. 38. 119 See above, n. 112, for this account and n. 14 for references to other Christian and Muslim historians recounting the same events. 120 We see the results of this process in the attempts made by local elites to identify ways and arenas

75

76

p e tr a  m . s ij pe s t e i j n

indeed, Arabic did eventually become dominant. But there is a crucial difference between the depiction of the change in the sources and what actually happened: the hiring of Muslim administrators in place of local ones led to the increased use of Arabic not because of a top-down decree that imposed the use of Arabic but simply because the administrators brought their primary language into office with them. Conversely, the Muslim administration did not impose Arabic as a roundabout way of getting rid of non-Muslim administrators.121 Coptic papyri show that the inhabitants of Egyptian monasteries made efforts to learn Arabic.122 Other papyri, too, document Greek and Coptic scribes’ efforts to master the language.123 In other words, an increase in acculturated Muslim, Arabic-speaking personnel in lower administrative offices led to an increase in the application of and demand for Arabic in the administration.124 This, in turn, led to an increase in the use of Arabic both inside and outside the dīwān.125 Another dimension should be kept in mind as well when considering the later reports. The story that they tell also functions as a form of Muslim polemic against non-Muslim administrators — a genre of which we have more examples especially from the ninth century ce. By that time, the Muslim empire had become a very different place compared to what it had been in the past century and a half. Thanks to progressive Arabicization (and Islamicization), it had become harder to distinguish one religious or ethnic group from another. Arabs had lost their privileged position through processes of acculturation and changes in the demo­graphic composition

outside the Muslim administration, such as mediation, in which to play a governing role; Sijpesteijn, ‘Loyal and Knowledgeable Supporters’. 121 A much later report presented by Ibn ʿAsākir on the translation of the Syrian dīwāns demonstrates this by recounting the events in the following order: Sarjūn dies; Sulaymān takes his position as the first Muslim to head all chancery offices; Sulaymān subsequently translates the dīwāns to Arabic. Ibn ʿAsākir’s report is cited by Yarbrough (Friends of the Emir, pp. 73–74), who emphasizes that ‘it contains no suggestion that Arabicization per se was calculated to exclude non-Muslims. Rather, the language change was a result of employing Muslims’ (p. 75). However, using this same passage, Yarbrough seems to suggest the opposite in a later publication (‘Muslim Rulers, Non-Muslim Subjects’, p. 368): ‘Arabic gradually became the official language of government administration, a policy that may have arisen partly from a desire to exclude Christians from influence’. Milka LevyRubin makes a similar claim (Non-Muslims in the Early Islamic Empire, p. 100): ‘His decree regarding the Arabization of the dīwāns seems to have also entailed an attempt to curtail the employment of non-Muslims in the administration’. Arabic chroniclers nevertheless also sometimes remember it as such: the ‘solution’ offered for the dominance of non-Muslim administrators in the Khurasani chancery is to switch its language to Arabic (see above nn. 25–26). 122 Palombo, ‘The Christian Clergy’s Islamic Local Government’, ch. 3. 123 See, for example, a small cheat sheet for a scribe needing to use Greek and Arabic in his work that lists the Greek letters corresponding to the numbers one to forty with the Arabic numeral letter equivalents (von Karabacek, Papyrus Erzherzog Rainer, p. 223). See also the references above, n. 103. 124 Sijpesteijn, Shaping a Muslim State, pp. 95–96, 111. 125 See also Saliba’s argument, discussed above (n. 101), that the translation of agricultural fiscal ledgers led to the translation of the scientific treatises needed to produce these documents. The use of Arabic increased amongst Egypt’s population in the ninth century, as shown, for example, by a papyrus dated c. 267/880–81 that contains Hebrew letters and their Arabic equivalents (von Karabacek, Papyrus Erzherzog Rainer, no. 829).

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

of the Muslim empire at large.126 In response, attempts were made to draw up and enforce boundaries between different religious groups and between Muslims and non-Muslims.127 Muslim rhetoric against non-Muslims presents non-Muslim administrators as disrespectful and even as undermining Islam. The stories quoted above surrounding the switch to Arabic in the administration seem to fit this trend of negative attitudes and suspicion vis-à-vis non-Muslims.128 This context explains why the anecdotes reporting an actual change in administrative personnel composition added a narrative layer about outrageously insolent non-Muslim administrators — a Syrian clerk urinating in an inkwell, Persian and Byzantine heads of the administration behaving haughtily and arrogantly, and Zoroastrian scribes in Khurasan embracing polytheism. The new Muslim rulers definitely used language and especially Arabic to express their dominance, privileged position, and imperial ambitions in explicit and open ways right from the moment of their arrival on the historical scene. They did so via all sorts of public writings, but also via documents produced in the dīwān. Greek, Coptic, and Sogdian texts from the pre-Islamic period did not start with an invocation, but documents in those languages dating to the earliest years of Muslim rule do include the invocation ‘In the name of God’ (en onomati tou theou/ par naam bgy damdane), which mirrors the basmala even on the visual level as it appears set off from the body of the letter at the top of the text. This fundamental change in administrative documents written in local languages was surely imposed by the new rulers. Similarly, the use of Arabic in bilingual documents functions as a clear symbol of the regime change that followed the conquests. In short, the Arab rulers were not passive governors, and their interventions also covered the language of administration and communication. They did implement invasive administrative changes directly after their arrival, but they did not impose a policy of monolingualism as a governmental measure by insisting on the exclusive use of Arabic in the dīwān to the detriment of all other languages. There was no caliphal decree by ʿAbd al-Malik, al-Walīd, or any other caliph ordering a linguistic switch across the entire administration.129 Rather, there were many individual decisions by governors, caliphs, and their agents aimed at improving the dīwān, and these decisions included some that increased the use of Arabic, especially as the number of Arab/Arabicized and Muslim/Islamicized administrators grew, though this process took different forms depending on the place and the genre of text.

126 Sijpesteijn, ‘Loyal and Knowledgeable Supporters’. 127 For a discussion on when such restrictions should be considered a coherent or ad hoc policy, see Levy-Rubin, Non-Muslims in the Early Islamic Empire, pp. 88–96; Levy-Rubin, ‘ʿUmar II’s Ghiyār Edict’; Yarbrough, ‘Did ʿUmar b. ʿAzīz Issue an Edict Concerning Non-Muslim Officials?’. 128 Yarbrough, Friends of the Emir, pp. 88–109; Levy-Rubin, Non-Muslims in the Early Islamic Empire, p. 102 n. 25; Khalek, ‘Some Notes on the Representation of Non-Muslim Officials in al-Ǧahšiyārī’s Kitāb al-Wuzarāʾ wa-l-Kuttāb’, p. 506. 129 Yarbrough discusses the claims concerning a caliphal decree in the Arabic chronicles and in scholarly interpretations in Friends of the Emir, p. 76 n. 86.

77

78

p e tr a  m . s ij pe s t e i j n

Instead of a wholesale decree of Arabicization, the Arabs opted for the active maintenance of a multilingual administration, which served them well. Rather than undermining the infrastructures that supported a multilingual administration, they maintained, invested in, and even stimulated them. This allowed local languages to continue to play a role in the Muslim administration, or even to play a larger role than they had previously, for centuries after the establishment of Arab rule.130 In taking this approach, the Arabs were applying sound governmental practice by making use of linguistic identity politics — a strategy that has, of course, characterized premodern and early modern multilingual empires throughout the world. Arabic did eventually become the language of the caliphate. It became dominant because it was the language of the conquerors, whose newly established rule and religion were expressed in Arabic. As conquered peoples joined them through conversion, assimilation, or association, there were inevitable linguistic consequences. But the idea of language as a necessary vehicle of political and cultural dominion, naturally associated with the forced imposition of monolingualism, would not appear until much later, with the nation state.131 Instead, an opposite trend, at least equally essential to empires, accompanied linguistic changes in the caliphate. This was the awareness that governing many different peoples, with diverse customs and practices, was best accomplished in a variety of languages. In practice, this meant that a multilingual administration was actively upheld in the caliphate. Had the caliph intended to impose Arabic throughout his realm, it would not have taken several centuries for his wish to be realized and there would not have been large areas where local languages remained dominant even after large-scale Arabicization had become a fact. Such a decree would also have precluded the trial and error, back and forth, and uneven progress that characterize the process of Arabicization in the sources. Indeed, the situation in the early Islamic empire matches that envisioned by Antonio de Nebrija, whose often quoted statement from 1492 — that language always accompanies empire — also appears at the beginning of this chapter. At first sight, his statement, connecting as it does language to power, might seem like an archetypal description of imperial (colonial) linguistic imposition. Magalí Armillas-Tiseyra argues, however, that Nebrija’s use of the term compañera to express the relation between empire and language is simultaneously suggestive and elusive, embodying ambitions and anxieties about this nexus. She contends that upon a reading of his entire introduction, ‘Nebrija’s claim that language is the companion of empire becomes less a triumphant assertion of the spread of language with empire (or empire with language) than the submission that their

130 Pace Bertold Spuler, who wrote concerning the naql al-dīwān (Iran in früh-islamischer Zeit, p. 245): ‘The Arabization of the administration was enforced so rigorously that in the course of the eighth and ninth centuries Arabic in fact became the language of the administration’. See the unexpected Coptic-Greek letter directed in the ninth century at an Arab official at the provincial level (Berkes, ‘On Arabisation’, pp. 415–17). 131 Mignolo, Local Histories/Global Designs, pp. 218–19, 256–57.

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

partnership is in fact changeable, provisional and above all unpredictable’.132 Indeed, the Spanish empire, which eventually caused major language shifts in Meso-America, relied on local languages besides Spanish for its administration and conversion for several generations.133 It was multilingual rather than monolingual policies that characterised the world in which the caliphate operated alongside other empires of the day. * * * This research was supported by the European Research Council under grant number 683194. Versions of this paper were presented at the Macro seminar at the University of Tennessee, the Near Eastern Studies department of Princeton University, and the Dar al-Athar al-Islamiyya in Kuwait. I would like to thank the participants for their feedback and comments. I have also greatly benefited from discussions with Abdullah AlHatlani, Alon Dar, Jeroen Duindam, Edmund Hayes, Reza Huseini, Marie Legendre, Bernhard Palme, Cecilia Palombo, Karen Radner, Khodadad Rezakhani, Eline Scheerlinck, Mehdy Shaddel, Alp Yenen, and Uri Yiftach. Finally, volume editors Antoine Borrut and Alison Vacca offered very useful commentary and pointed me to additional references. Any remaining mistakes are, of course, my own.

132 Armillas-Tiseyra, ‘On Language and Empire’, p. 199. 133 Mignolo, The Darker Side of the Renaissance, pp. 48–67.

79

80

p e tr a  m . s ij pe s t e i j n

Biblio­graphy Primary Sources al-Azdī, Jamharat al-lugha, ed. by Ramzī Munīr Baʿlbakkī (Beirut: Dār al-ʿilm li–l-malāyīn, 1987–1988) al-Balādhurī, Futūḥ al-buldān, ed. by Michael Jan de Goeje (Leiden: Brill, 1968) Chronicle of 1234, trans. by A. Palmer, The Seventh Century in the West-Syrian Chronicles, Translated Texts for Historians, 15 (Liverpool: Liverpool University Press, 1993) History of the Patriarchs of the Coptic Church of Alexandria, ed. and trans. by B. Evetts, iii: Agathon to Michael I (ad 766), Patro­logia Orientalis, 5 (1910), 3–215 Ibn ʿAbd al-Ḥakam, Futūḥ Miṣr: The History of the Conquest of Egypt, North Africa and Spain, ed. by Charles Torrey (New Haven, CT: Yale University Press, 1922) Ibn al-Nadīm, Kitāb al-Fihrist, ed. by Ayman Fuʾād Sayyid (London: Al-Furqān Islamic Heritage Foundation, 2014) al-Jahshiyārī, Kitāb al-Wuzarāʾ wa-l-kuttāb: Das Kitāb al-Wuzarāʾ wa-l-Kuttāb des Abū ʿAbdallāh Muḥammad Ibn ʿAbdūs al-Ğahšiyārī (Leipzig: Harrassowitz, 1345/1926) al-Kindī, Kitāb al-Wulāt wa-kitāb al-quḍāt, ed. by Rhuvon Guest (Leiden: Brill, 1912) al-Maqrīzī, Khiṭaṭ = al-Mawāʾiz wa-l-iʿtibār fī dhikr al-khiṭaṭ wa-l-āthār, ed. by Ayman Fuad Sayyid (London: Al-Furqān Islamic Heritage Foundation and Centre for the Study of Islamic Manu­scripts, 2013) Maronite Chronicle [664 ce], trans. by Andrew Palmer in The Seventh Century in the WestSyrian Chronicles, Translated Texts for Historians, 15 (Liverpool: Liverpool University Press, 1993) Nebrija, Antonio de, Gramática de la lengua Española, ed. by Antonio Quilis (Madrid: Centro de Estudios Ramón Areces, 1989) SB = Sammelbuch griechischer Urkunden aus Ägypten al-Ṣūlī, Adab al-kuttāb, ed. by Muḥammad Bahjat al-Atharī and Maḥmūd Shukrī al-Ālūsī (Baghdad: al-Maṭbaʿa al-ʿArabiyya; Cairo: al-Maṭbaʿa al-Salafiyya, 1341/1923) al-Ṭabarī, Taʾrīkh al-rusul wa-l-mulūk, ed. by Michael Jan de Goeje and others (Leiden: E. J. Brill, 1879–1901) Theophanes Confessor, The Chronicle of Theophanes Confessor: Byzantine and Near Eastern History ad 284–813, trans., introd., and comment. by Cyril Mango and Roger Scott with Geoffrey Greatrex (Oxford: Clarendon Press, 1997) Theophilus of Edessa, Theophilus of Edessa’s Chronicle and the Circulation of Historical Knowledge in Late Antiquity and Early Islam, trans. by Robert Hoyland, Translated Texts for Historians, 57 (Liverpool: Liverpool University Press, 2011) Secondary Works Armillas-Tiseyra, Magalí, ‘On Language and Empire: The Pro­logue to Grammar of the Castilian Language (1492)’, Publications of the Modern Language Association of America, 131 (2016), 197–208 Bacharach, Jere L., and Henri Amin Awad, ‘Rare Early Egyptian Islamic Coins and Coin Weights: The Awad Collection’, Journal of the American Research Center in Egypt, 18 (1981), 51–56

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

Beaucamp, Joëlle, ‘Byzantine Egypt and Imperial Law,’ in Egypt in the Byzantine World, ed. by Roger S. Bagnall (Cambridge: Cambridge University Press, 2007), pp. 271–87 Berkes, Lajos, ‘On Arabisation and Islamisation in Early Islamic Egypt. I: Prosopographic Notes on Muslim Officials’, Chronique d’Égypte, 113 (2018), 415–20 —— , ‘The Latest Identified Greek Documentary Text from Egypt: A Papyrus from 825 ad (SPP III2 577 Reconsidered)’, Zeitschrift für Papyro­logie und Epigrafik, 209 (2019), 242–44 —— , ‘Bemerkungen zu Verwaltungsdokumenten aus dem früharabischen Ägypten’, Archiv für Papyrusforschung und verwandte Gebiete, 68 (2022), 375–78 Berkes, Lajos, and Naïm Vanthieghem, ‘A Late Coptic Tax Receipt from the Egyptian National Library (P.Cair.Nat. Library inv. 3890)’, Journal of Coptic Studies, 21 (2019), 13–17 —— , ‘Notes on the Careers of Nāǧid b. Muslim and ʿAbd al-Malik b. Yazīd’, Chronique d’Égypte, 95 (2020), 154–61 Berkes, Lajos, and Khaled Younes, ‘A Trilingual Scribe from Abbasid Egypt? A Note on CPR XXII 17’, Archiv für Papyrusforschung und verwandte Gebiete, 58 (2012), 97–100 Berkes, Lajos, Alain Delattre, and Naïm Vanthieghem, ‘A Ninth-Century Coptic Tax Refund Document. Reedition of CPR IV 197’, Chronique d’Égypte, 96 (2021), 164–72 Blankinship, Khalid Yahya, The End of the Jihad State: The Reign of Hishām b. ʿAbd al-Malik and the Collapse of the Umayyads (New York: State University of New York Press, 1994) Borrut, Antoine, ‘Pouvoir mobile et construction de l’espace dans les premiers siècles de l’islam’, in Le gouvernement en déplacement: pouvoir et mobilité de l’Antiquité à nos jours, ed. by Sylvain Destephen, Josiane Barbier, and François Chausson (Rennes: Presses Universitaires de Rennes, 2019), pp. 243–67 Boud’hors, Anne, ‘Situating the Figure of Papas, Pagarch of Edfu at the End of the Seventh Century: The Contribution of the Coptic Documents’, in Living the End of Antiquity: Individual Histories from Byzantine to Islamic Egypt, ed. by Sabine Hübner and others (Berlin: De Gruyter, 2020), pp. 63–71 Clackson, Sarah, and Petra M. Sijpesteijn, ‘A Mid-Eighth-Century Trilingual Tax Demand Related to the Monastery of Apa Apollo at Bawit’, in Monastic Estates in Late Antique and Early Islamic Egypt: Ostraca, Papyri, and Essays in Memory of Sarah Clackson, ed. by Anne Boud’hors and others (Cincinnati: American Society of Papyro­logists, 2009), pp. 102–19 Cohen, Mark R., ‘What Was the Pact of ʿUmar? A Literary-Historical Study’, Jerusalem Studies in Arabic and Islam, 23 (1999), 100–57 Crespo, Emilio, ‘The Linguistic Policy of the Ptolemaic Kingdom’, in Φωνής χαρακτήρ εθνικός: Actes du ve congrès international de dialecto­logie grecque (Athènes, 28–30 septembre 2006), ed. by Miltiadis Hatzopoulos and Vassiliki Psilakakou (Athens: Boccard, 2007), pp. 35–49 —— , ‘Language Policies’, in Encyclopaedia of Ancient Greek Language and Linguistics, ed. by Georgios K. Giannakis (Leiden: Brill, 2013), http://dx.doi.org.ezproxy.leidenuniv. nl:2048/10.1163/2214-448X_eagll_COM_000048 [accessed 30 June 2021] Cromwell, Jennifer, ‘Coptic Texts in the Archive of Flavius Attias’, Zeitschrift für Papyro­logie und Epi­graphik, 184 (2013), 280–88 —— , Recording Village Life: A Coptic Scribe in Early Islamic Egypt (Ann Arbor: University of Michigan Press, 2017)

81

82

p e tr a  m . s ij pe s t e i j n

—— , ‘A Village Scribe on the Eve of Change’, in Living the End of Antiquity: Individual Histories from Byzantine to Islamic Egypt, ed. by Sabine Hübner and others (Berlin: De Gruyter, 2020), pp. 129–40 Darley-Doran, Robert E., History of Currency in the Sultanate of Oman (London: Spink, 1990) Darling, Linda T., ‘Ottoman Turkish: Written Language and Scribal Practice, 13th to 20th Centuries’, in Literacy in the Persianate World: Writing and the Social Order, ed. by Brian Spooner and William L. Hanaway (Philadelphia: University of Pennsylvania Press, 2012), pp. 171–95 Debié, Muriel, ‘Christians in the Service of the Caliph: Through the Looking Glass of Communal Identities’, in Christians and Others in the Umayyad State, ed. by Antoine Borrut and Fred M. Donner, Late Antique and Medi­eval Islamic Near East, 1 (Chicago: Oriental Institute of the University of Chicago, 2016), pp. 53–71 Dekker, Renate, Episcopal Networks and Authority in Late Antique Egypt: Bishops of the Theban Region at Work (Louvain: Peeters, 2018) Delattre, Alain, ‘Langue et sources documentaires coptes’, Annuaire de l’École pratique des hautes études (EPHE), Section des sciences historiques et philo­logiques, 149 (2018), 109–11 Delattre, Alain, and others, ‘Écrire en arabe et en copte: le cas de deux lettrés bilingues’, Chronique d’Égypte, 87 (2012), 170–88 Delattre, Alain, and Jean-Luc Fournet, ‘La première pétition en copte: révision de SB Kopt. IV 1709’, Chronique d’Égypte, 93 (2018), 167–82 Delattre, Alain, and Naïm Vanthieghem, ‘Un ensemble archivistique trilingue à Strasbourg: un protocole et deux ordres de réquisition de la fin du viie siècle’, in Mélanges Jean Gascou, ed. by Jean-Luc Fournet and Arietta Papaconstantinou, Travaux et mémoires, 20 (Paris: Association des Amis du Centre d’Histoire et Civilisation de Byzance, 2016), pp. 109–31 Diem, Werner, Arabische Steuerquittungen des 8. bis 11. Jahrhunderts aus der Heidelberger Papyrussammlung und anderen Sammlungen (Wiesbaden: Harrassowitz, 2008) Duri, A. A., ‘Dīwān i. The Caliphate’, in Encyclopaedia of Islam, 2nd edn, ed. by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs (Leiden: Brill, online edn) Fenina, Abdelhamid, ‘L’arabisation du monnayage d’Ifrīqiya: étapes et signification’, in Civilisation en transition (II): sociétés multilingues à travers l’histoire du moyen orient, ed. by Jean-Luc Fournet, Jean-Michel Mouton, and Jacques Paviot (Byblos, Lebanon: Centre International des Sciences de l’Homme, 2016), pp. 115–68 Fournet, Jean-Luc, ‘The Multilingual Environment of Late Antique Egypt: Greek, Latin, Coptic, and Persian Documentation’, in Oxford Handbook of Papyro­logy, ed. by Roger Bagnall (Oxford: Oxford University Press, 2009), pp. 418–51 —— , The Rise of Coptic: Egyptian versus Greek in Late Antiquity (Princeton, NJ: Princeton University Press, 2020) Frantz-Murphy, Gladys, ‘Land Tenure in Egypt in the First Five Centuries of Islamic Rule (Seventh–Twelfth Centuries ad)’, in Agriculture in Egypt from Pharaonic to Modern Times, ed. by Alan Bowman and Eugene Rogan, Proceedings of the British Academy, 96 (Oxford: Oxford University Press, 1999), pp. 237–66 Frye, Richard N., ‘Sasanian Numbers and Silver Weights’, Journal of the Royal Asiatic Society, 105 (1973), 2–11 Garel, Esther, ‘Une demande de recensement du pagarque Rāšid b. Ḫālid: CPR IV 1 revisité’, Chronique d’Égypte, 93 (2018), 187–99

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

—— , Héritage et transmission dans le monde égyptien (Cairo: Institut Français d’Archeo­logie Orientale, 2020) Garosi, Eugenio, ‘Cross-Cultural Parameters of Scribal Politesse in the Correspondence of Arab-Muslim Officials from Early Islamic Egypt’, in Living the End of Antiquity: Individual Histories from Byzantine to Islamic Egypt, ed. by Sabine Hübner and others (Berlin: De Gruyter, 2020), pp. 73–94 Gilbert, Claire, ‘Multilingual Administrative Dynasties in Habsburg Spain’, in Iberian Babel: Translation and Multilingualism in the Medi­eval and the Early Modern Mediterranean, ed. by Michelle M. Hamilton and Nuria Silleras-Fernandez, The Medi­eval and Early Modern Iberian World, 82 (Leiden: Brill, 2022), pp. 170–94 Gonis, Nicolaos, and Gesa Schenke, ‘Two Entagia from Cambridge’, Chronique d’Égypte, 88 (2013), 372–78 Griffith, Sidney H., ‘The Manṣūr Family and Saint John of Damascus: Christians and Muslims in Umayyad Times’, in Christians and Others in the Umayyad State, ed. by Antoine Borrut and Fred M. Donner, Late Antique and Medi­eval Near East, 1 (Chicago: Oriental Institute of the University of Chicago, 2016), pp. 29–51 Grob, Eva Mira, Documentary Arabic Private and Business Letters on Papyrus: Form and Function, Content and Context (Berlin: De Gruyter, 2010) Grohmann, Adolf, From the World of Arabic Papyri (Cairo: Ma’aref Press, 1952) Gundelfinger, Simon, and Peter Verkinderen, ‘The Governors of al-Shām and Fārs in the Early Islamic Empire: A Comparative Regional Perspective’, in Transregional and Regional Elites: Connecting the Early Islamic Empire, ed. by Hannah-Lena Hagemann and Stefan Heidemann, Studies in the History and Culture of the Middle East, 36 (Berlin: De Gruyter, 2020), pp. 255–330 Hagemann, Hannah-Lena, ‘Was Muṭarrif b. al-Mughīra al-Thaqafī a Khārijite? Rebellion in the Early Marwānid Period’, Al-ʿUṣūr al-Wusṭā, 30 (2022), 445–68 Hawting, Gerald R., ‘Marwān II’, in Encyclopaedia of Islam, 2nd edn, ed. by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs (Leiden: Brill, online edn) Heidemann, Stefan, ‘The Evolving Representation of the Early Islamic Empire and its Religion on Coin Imagery’, in The Qurʾān in Context: Historical and Literary Investigations into the Qurʾānic Milieu, ed. by Angelika Neuwirth, Nicolai Sinai, and Michael Marx (Leiden: Brill, 2010), pp. 149–95 —— , ‘The Standing Caliph Type: The Object on the Reverse’, in Coinage and History in the Seventh Century Near East, ed. by Andrew Oddy (London: Archetype Publications, 2010), pp. 23–34 Ḥitti, Philip Khûri. The Origins of the Islamic State (New York: Columbia University Press, 1916) Hochedlinger, Michael, Petr Mat’a, and Thomas Winkelbauer, eds, Verwaltungsgeschichte der Habsburgermonarchie in der Frühen Neuzeit, Vol. 1: Hof und Dynastie, Kaiser und Reich, Zentralverwaltungen, Kriegswesen und landesfürstliches Finanzwesen (Vienna: Böhlau Verlag, 2019) Huseini, Reza, ‘Thinking in Arabic, Writing in Sogdian: Arab-Sogdian Diplomatic Relations in the Early Eighth Century’, in From Samarqand to Toledo: Greek, Sogdian and Arabic Documents and Manu­scripts from the Islamicate World and Beyond, ed. by Andreas Kaplony and W. Matt Malczycki (Leiden: Brill, 2022), pp 67–87

83

84

p e tr a  m . s ij pe s t e i j n

Jonson, Trent, ‘A Numismatic History of the Early Islamic Precious Metal Coinage of North Africa and the Iberian Peninsula’, 2 vols (unpublished doctoral thesis, University of Oxford, 2014) Kaplony, Andreas, ‘Die Arabisierung der frühislamischen Verwaltung Syrien-Palästinas und Ägyptens im Spiegel der zweisprachigen griechisch-arabischen Dokumente (550–750): Ein Plädoyer für einen regionalen Ansatz’, in Denkraum Spätantike: Reflexionen von Antiken im Umfeld des Koran, ed. by Nora Schmidt, Nora K. Schmid, and Angelika Neuwirth, Episteme in Bewegung: Beiträge zu einer transdisziplinären Wissensgeschichte, 5 (Wiesbaden: Harrassowitz, 2016), pp. 387–404 Karabacek, Joseph von, Papyrus Erzherzog Rainer: Führer durch die Ausstellung (Vienna: Selbstverlag der Sammlung, 1894) Khalek, Nancy, ‘Some Notes on the Representation of Non-Muslim Officials in alǦahšiyārī’s (d. 331/942) Kitāb al-Wuzarāʾ wa-l-Kuttāb’, Arabica, 62 (2015), 503–20 Khan, Geoffrey, ‘The Opening Formula and Witness Clauses in Arabic Legal Documents from the Early Islamic Period’, Journal of the American Oriental Society, 139 (2019), 23–39 Khoury, R. Georges, ‘ʿUbayd Allāh b. Ḥabḥāb’, in Encyclopaedia of Islam, 2nd edn, ed. by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs (Leiden: Brill, online edn) Khurt, Amélie, ‘State Communications in the Persian Empire’, in State Correspondence in the Ancient World: From New Kingdom Egypt to the Roman Empire, ed. by Karen Radner (Oxford: Oxford University Press, 2014), pp. 112–40 Kovarik, Sophie, ‘Der herakleopolitische Notar Paulos: Ein Kompromiss aus dem umay­ yadischen Ägypten — der bisher späteste griechische Vertrag’, Tyche, 35 (2020), 55–70 Kratchkovsky, Ignati J., and V. A. Kratchkovskaya, ‘Drevnejshij arabskij dokument iz Srednej Azii’, in Izbrannye sochineniia, ed. by Ignati J. Kratchkovsky (Moscow: Izdatel’stvo Akademii nauk, 1955), i, pp. 182–212 Kruk, Remke, ‘The Saddest Beast? Notes on the Pig in Arabic Culture’, Quaderni di studi arabi, 14 (2019), 243–61 Legendre, Marie, ‘Neither Byzantine nor Islamic? The Duke of the Thebaid and the Formation of the Umayyad State’, Historical Research, 89 (2016), 3–18 Lev, Yaacov, The Administration of Justice in Medi­eval Egypt from the 7th to the 12th Century (Edinburgh: University of Edinburgh Press, 2021) Levy-Rubin, Milka, Non-Muslims in the Early Islamic Empire: From Surrender to Coexistence (Cambridge: Cambridge University Press, 2011) —— , ‘ʿUmar II’s Ghiyār Edict: Between Ideo­logy and Practice’, in Christians and Others in the Umayyad State, ed. by Antoine Borrut and Fred M. Donner, Late Antique and Medi­eval Islamic Near East, 1 (Chicago: Oriental Institute of the University of Chicago, 2016), pp. 157–72 Livshits, Vladimir A., Sogdian Epi­graphy of Central Asia and Semirech’e, ed. by Nicholas Sims-Williams, trans. by Tom Stableford, Corpus Inscriptionum Iranicarum, pt 2: Inscriptions of the Seleucid and Parthian Periods and of Eastern Iran and Central Asia, iii: Sogdian, 4 (London: School of Oriental and African Studies, 2015) Malczycki, W. Matt, ‘The Papyrus Industry in the Early Islamic Era’, Journal of the Economic and Social History of the Orient, 54 (2011), 185–202

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

Manzano Moreno, Eduardo, ‘Conquest and Settlement: What al-Andalus Can Tell Us about the Arab Expansion at the Time of the Umayyad Caliphate’, in The Umayyad World, ed. by Andrew Marsham (London: Routledge, 2021), pp. 314–31 Marsham, Andrew, Rituals of Islamic Monarchy: Accession and Succession in the First Muslim Empire (Edinburgh: Edinburgh University Press, 2009) Mavroudi, Maria, ‘Greek Language and Education under Early Islam’, in Islamic Cultures, Islamic Contexts: Essays in Honor of Professor Patricia Crone, ed. by Asad Ahmed, Behnam Sadeghi, Robert Hoyland, and Adam Silverstein (Leiden: Brill, 2015), pp. 295–342 Mignolo, Walter D., The Darker Side of the Renaissance: Literacy, Territoriality and Colonization, 2nd edn (Ann Arbor: University of Michigan Press, 2003) —— , Local Histories/Global Designs: Coloniality, Subaltern Knowledges, and Border Thinking, 2nd edn (Princeton, NJ: Princeton University Press, 2012) Mikhail, Maged S. A., From Byzantine to Islamic Egypt: Religion, Identity and Politics after the Arab Conquest (London: I. B. Tauris, 2014) Morelli, Federico, ‘Legname, palazzi e moschee: P.Vindob. G 31 e il contributo dell’Egitto alla prima architettura islamica’, Tyche, 13 (1998), 165–90 —— , Documenti greci per la fiscalità e la amministrazione dell’Egitto arabo, Corpus Papyrorum Raineri, 22: Griechische Texte, xv (Vienna: Brüder Hollinek, 2001) —— , ‘Dalle prigioni dell’Arsinoite’, in Gedenkschrift Ulrike Horak, ed. by Hermann Harrauer and Rosario Pintaudi (Florence: Gonnelli, 2004), pp. 186–95 Morony, Michael G., Iraq after the Muslim Conquest (Princeton, NJ: Princeton University Press, 1984) Palombo, Cecilia, ‘The Christian Clergy’s Islamic Local Government in Late Marwanid and Abbasid Egypt’ (unpublished doctoral dissertation, Princeton University, 2020) Papaconstantinou, Arietta, ‘Administering the Early Islamic Empire: Insights from the Papyri’, in Money, Power and Politics in Early Islamic Syria: A Review of Current Debates, ed. by John Haldon (London: Routledge, 2010), pp. 57–74 —— , ‘Introduction’, in The Multilingual Experience in Egypt, from the Ptolemies to the ʿAbbāsids, ed. by Arietta Papaconstantinou (Burlington, VT: Ashgate, 2010), pp. 1–16 al-Qāḍī, Wadād, ‘Death Dates in Umayyad Stipends Registeres (Dīwān al-ʿAṭāʾ)? The Testimony of the Papyri and the Literary Sources,’ in From Bāwīṭ to Marw. Documents from the Medi­eval Muslim World, ed. by Andreas Kaplony, Daniel Potthast and Cornelia Römer (Leiden: Brill, 2015), pp. 59–82 —— , ‘The Names of Estates in State Registers before and after the Arabization of the “Diwans”’, in Umayyad Legacies: Medi­eval Memories from Syria to Spain, ed. by Antoine Borrut and Paul M. Cobb (Leiden: Brill, 2010), pp. 255–80 Rāġib, Yūsuf, ‘Lettres arabes II’, Annales islamo­logiques, 16 (1980), 1–29 Reinfandt, Lucian, ‘Petosiris the Scribe’, in Living the End of Antiquity: Individual Histories from Byzantine to Islamic Egypt, ed. by Sabine Hübner and others (Berlin: De Gruyter, 2020), pp. 141–52 Reinfandt, Lucian, and Sven Tost, ‘Aller Anfang ist schwer …: Mehrsprachigkeit und ihre gesellschaftlichen Hintergründe im spätantiken Ägypten’, Österreichische Nationalbibliothek Forschungsblog, 20 April 2017, www.onb.ac.at/forschung/ forschungsblog/artikel/news/aller-anfang-ist-schwer/

85

86

p e tr a  m . s ij pe s t e i j n

Richardson, Seth, ‘Before Things Worked: A “Low-Power” Model of Early Mesopotamia’, in Ancient States and Infrastructural Power: Europe, Asia and America, ed. by Clifford Ando and Seth Richardson (Philadelphia: University of Pennsylvania Press, 2017), pp. 17–62 Richter, T. Sebastian, ‘Language Choice in the Qurra Dossier’, in The Multilingual Exper­ ience in Egypt, from the Ptolemies to the ʿAbbāsids, ed. by Arietta Papaconstantinou (Burlington, VT: Ashgate, 2010), pp. 189–220 Robinson, Chase, ʿAbd al-Malik (Oxford: Oneworld, 2005) Rochette, Bruno, ‘Language Policies in the Roman Republic and Empire’, in A Companion to the Latin Language, ed. by James Clackson (Chichester: Wiley-Blackwell, 2012), pp. 549–63 Rötter, Gernat, ‘Maslama b. ʿAbd al-Malik b. Marwān’, in Encyclopaedia of Islam, 2nd edn, ed. by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs (Leiden: Brill, online edn) Saliba, George, Islamic Science and the Making of the European Renaissance (Cambridge, MA: MIT press, 2007) Scheerlinck, Eline, ‘Solidarity and Self-interest: Protective Interventions by Local Elites in the Countryside of Early Islamic Egypt’ (unpublished doctoral dissertation, Leiden University, 2023) Schenke, Gesa, ‘Rashid ibn Chaled and the Return of Overpayments’, Chronique d’Égypte, 89 (2014), 202–09 Shaddel, Mehdy, ‘Monetary Reforms under the Sufyanids: The Papyro­logical Evidence’, Bulletin of the School of Oriental and African Studies, 84 (2021), 263–93 Sijpesteijn, Petra M., ‘New Rule over Old Structures: Egypt after the Muslim Conquest’, in Regime Change in the Ancient Near East and Egypt: From Sargon of Agade to the Seljuks, ed. by Harriet Crawford, Proceedings of the British Academy, 136 (London: British Academy Publications, 2007), pp. 183–202 —— , ‘A Seventh/Eighth-Century List of Companions from Fustat’, in Sixty-Five Papyro­ logical Texts Presented to Klaas A. Worp on the Occasion of his 65th Birthday, ed. by Francisca A. J. Hogendijk and Brian P. Muhs (Leiden: Brill, 2008), pp. 369–77 —— , ‘Arabic-Greek Archives’, in The Multilingual Experience in Egypt, from the Ptolemies to the ʿAbbāsids, ed. by Arietta Papaconstantinou (Burlington, VT: Ashgate, 2010), pp. 105–26 —— , ‘Army Economics: An Early Papyrus Letter Related to ʿAṭāʾ Payments’, in Histories of the Middle East: Studies in Middle Eastern Society, Economy and Law in Honor of A. L. Udovitch, ed. by Roxani Margariti, Adam Sabra, and Petra M. Sijpesteijn (Leiden: Brill, 2011), pp. 245–68 —— , Shaping a Muslim State: The World of a Mid-Eighth-Century Egyptian Official (Oxford: Oxford University Press, 2013) —— , ‘An Early Umayyad Papyrus Invitation for the Hajj’, Journal of Near Eastern Studies, 73 (2014), 179–90 —— , ‘Loyal and Knowledgeable Supporters: Integrating Egyptian Elites in Early Islamic Egypt’, in Empires and Communities in the Post-Roman and Islamic World (c. 400–1000 ce), ed. by Walter Pohl and others (Oxford: Oxford University Press, 2021), pp. 329–59 —— , ‘Egypt’s Connections in the Early Caliphate: Political, Economic and Cultural’, in Egypt and the Eastern Mediterranean World: From Constantinople to Baghdad,

the e a r ly i s l am i c e m p i r e ’s p o l i cy o f mu lt i li ngual gove rnance

500–1000 ce, ed. by Jelle Bruning, Janneke de Jong, and Petra M. Sijpesteijn (Cambridge: Cambridge University Press, 2022), pp. 238–71 —— , ‘Closing Ranks: Discipline and Loyalty in the Umayyad Army’, Al-ʿUṣūr al-Wusṭā, 30 (2022), 469–99 Silverstein, Adam, Postal Systems in the Pre-Modern Islamic World (Cambridge: Cambridge University Press, 2007) Sprengling, Martin, ‘From Persian to Arabic’, American Journal of Semitic Languages and Literatures, 56 (1939), 175–230, 325–36 Spuler, Bertold, Iran in früh-islamischer Zeit: Politik, Kultur, Verwaltung und öffentliches Leben zwischen der arabischen und der seldschukischen Eroberung, 633 bis 1055 (Wiesbaden: Steiner Verlag, 1952) Thomann, Johannes, ‘From Lyrics by al-Fazārī to Lectures by al-Fārābī: Teaching Astronomy in Baghdād (750–1000 c.e.)’, in The Place to Go: Contexts of Learning in Baghdād, 750–1000 c.e., ed. by Jens Scheiner and Damien Janos (Princeton, NJ: Darwin Press, 2014), pp. 503–25 Tillier, Matthieu, L’invention du cadi : la justice des musulmans, des juifs et des chrétiens aux premiers siècles de l’Islam (Paris: Publications de la Sorbonne, 2017) —— , ‘Représenter la province auprès du pouvoir impérial: les délégations (wufūd) égyptiennes aux trois premiers siècles de l’Islam’, Arabica, 67 (2020), 125–99 Treadwell, Luke, ‘ʿAbd al-Malik’s Coinage Reforms: The Role of the Damascus Mint’, Revue numismatique, 165 (2009), 357–81 —— , ‘Qur’anic Inscriptions on the Coins of the Ahl al-Bayt from the Second to Fourth Century ah / ‫’النقوش القرآنية على نقود المعدنية ألهل البيت ﻣﻦ القرن الثاني إلى القرن الرابع الهجري‬, Journal of Quranic Studies, 14 (2012), 47–71 Vacca, Alison, Non-Muslim Provinces under Early Islam: Islamic Rule and Iranian Legitimacy in Armenia and Caucasian Albania (Cambridge: Cambridge University Press, 2017) Walder, John, Catalogue of the Arab-Byzantine and Post-Reform Umaiyad Coins (London: Trustees of the British Museum, 1956) Weber, Dieter, Ostraca, Papyri und Pergamente, Corpus Inscriptionum Iranicarum, pt 3: Pahlavi Inscriptions, iv–v: Ostraca and Papyri, Texts 1 (London: School of Oriental and African Studies, 1992) Yakubovich, Ilya, ‘Mugh 1.i. Revisited’, Studia Iranica, 31 (2002), 231–53 —— , Review of Vladimir A. Livshits, Sogdian Epi­graphy of Central Asia and Semirech’e, Indo-Iranian Journal, 60 (2017), 413–26 Yarbrough, Luke, ‘Did ʿUmar b. ʿAzīz Issue an Edict Concerning Non-Muslim Officials?’, in Christians and Others in the Umayyad State, ed. by Antoine Borrut and Fred M. Donner, Late Antique and Medi­eval Islamic Near East, 1 (Chicago: Oriental Institute of the University of Chicago, 2016), pp. 173–206 —— , Friends of the Emir: Non-Muslim State Officials in Premodern Islamic Thought (Cambridge: Cambridge University Press, 2019) —— , ‘Muslim Rulers, Non-Muslim Subjects’, in Christian-Muslim Relations: A Biblio­ graphical History, xv: Thematic Essays (600–1600), ed. by Douglas Pratt and Charles L. Tieszen (Leiden: Brill, 2020), pp. 359–87 Zetterstéen, Karl Vilhelm, ‘Muḥammad b. Marwān’, in Encyclopaedia of Islam, 2nd edn, ed. by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs (Leiden: Brill, online edn)

87

Marie Legendre

The Translation of the Dīwān and the Making of the Marwanid ‘Language Reform’ Secretarial Agency, Economic Incentives, and Regional Dynamics in the Umayyad State

Introduction In the early 2010s, Wadād al-Qāḍī and Fred Donner both lamented that little attention had been paid to the ‘language reform’ or ‘decree’ of the caliph ʿAbd al-Malik (r. 65–86/685–705).1 This is especially true in comparison with the wealth of research on the contemporary reform of the gold coinage in 77/696. The most systematic work on the topic remains Martin Sprengling’s 1939 article, which George Miles called ‘a polemic article in which few Arabists, living or dead, come off unscathed’.2 Sprengling was indeed highly judgmental of the field of his time. His main focus was on Iraq, and he brought together the material available in his time on the career and family history of two key administrators of the dīwān of Basra: Zādhān (or Zādān) Farrūkh and Ṣāliḥ b.  ʿAbd al-Raḥmān.3 Yet, in the past eighty years or so, as the field of Islamic studies and its methodo­logies have evolved, we are able to ask new questions of the sources Sprengling examined, and new material on the topic has come to light. In that sense, there is a pressing need for an update. The present research is also deeply indebted to several milestones in the field since Sprengling’s article. First, a wealth of papyro­logical literature has been published on linguistic practices

1 Al-Qāḍī, ‘The Names of Estates in State Registers’, p. 255; Donner, The Articulation of Early Islamic State Structures, p. xxviii. 2 Sprengling, ‘From Persian to Arabic’; Miles, ‘A Byzantine Weight Validated by al-Walīd’, p. 3. 3 For Sprengling, the core of the Umayyad empire lay ‘in the Persian area’; ‘From Persian to Arabic’, p. 175. In his opinion, the narratives on Syria are ‘clearly secondary’ (p. 212), and he addresses only briefly the narratives on the dīwāns of Egypt and Khurasan (pp. 213–14).

Marie Legendre ([email protected]) is Senior Lecturer in Islamic History at the University of Edinburgh. She is the Principal Investigator of the ERC-funded project ‘The Finances of the Caliphate: Abbasid Fiscal Practice in Islamic Late Antiquity’. Navigating Language in the Early Islamic World: Multilingualism and Language Change in the First Cen­turies of Islam, ed. by Antoine Borrut, Manuela Ceballos, and Alison M. Vacca, ISMAR 2 (Turnhout: Brepols, 2024) pp. 89–165 10.1484/M.ISMAR-EB.5.134625

FHG

90

m a r i e l ege n d r e

in the Umayyad period that this contribution will build on.4 Additional research on the dīwān of Basra by al-Qāḍī has revealed something crucial. On the basis of al-Balādhurī’s (d. 279/892) Futūḥ al-buldān, she argued that what happened at the time of ʿAbd al-Malik was a translation of administrative registers, not a reform nor an overarching change of language in the administration.5 Roughly at the same time, George Saliba made a similar argument, using Ibn al-Nadīm’s (d. 385/995) Fihrist and, to a lesser extent, al-Jahshiyārī’s (d. 331/942) Kitāb al-Wuzarāʾ wa-l-kuttāb.6 According to Saliba, not only administrative registers were translated but also ‘some elementary texts or manuals that were used to train those who worked in the dīwān’. In his view, the translation of such manuals constituted the foundation of the ‘translation movement’.7 He also rightly noted that the translation of administrative documents went along with a change of personnel.8 Indeed, languages do not exist in isolation from the agents of linguistic practices. Milka Levy-Rubin went further by stating that ʿAbd al-Malik’s ‘decree regarding the Arabization of the dīwāns seems to have also entailed an attempt to curtail the employment of non-Muslims in the administration’.9 The debate was picked up by Luke Yarbrough, who focused on Ibn ʿAsākir’s (d. 571/1176) account of the translation of the dīwān of Syria and recognized in it early traces of attitudes towards the employment of non-Muslims in the administration before the development of a discourse on the topic in Islamic thought. In Yarbrough’s view, Ibn ʿAsākir’s text ‘contains no suggestion that Arabicization per se was calculated to exclude non-Muslims. Rather, the language change was a result of employing Muslims, which, in its turn, may have been occasioned by communal solidarity and rivalry’.10 Focusing on Egypt, Petra Sijpesteijn has also connected the translation of the dīwān with a change of personnel, non-Muslim secretaries being replaced by mawlās.11 Most recently, Kyle Longworth has shifted the debate away from the Muslim/non-Muslim dichotomy to argue that the so-called Arabization is better understood as the evolution of the administrative control of new local elites, as the secretaries involved in the 4 A turning point was the publication of Papaconstantinou ed., The Multilingual Experience in Egypt. 5 Al-Qāḍī, ‘The Names of Estates in State Registers’; see also al-Qāḍī, ‘Identity Formation of the Bureaucracy of the Early Islamic State’, pp. 143–44. The idea of a translation was first formulated by Ulrich Rebstock, though he did not make a case for it: Rebstock, ‘Observations on the Diwan al-Kharaj and the Assessment of Taxes in Umayyad Syria’. 6 Saliba, Islamic Science and the Making of the European Renaissance, pp. 45–72; see also Borrut, Entre mémoire et pouvoir, pp. 70–71. 7 Saliba, Islamic Science and the Making of the European Renaissance, pp. 54–58; the quotation is on p. 54. Saliba was reacting to Dimitri Gutas’s views on the topic; see Antoine Borrut’s contribution to this volume. 8 Saliba, Islamic Science and the Making of the European Renaissance, pp. 58–64. According to Saliba, Greek and Persian secretaries’ loss of the highest posts in the caliphal administration also provided an impulse to members of those linguistic communities to excel in other domains, another drive in the origins of the translation movement. 9 Levy-Rubin, Non-Muslims in the Early Islamic Empire, pp. 100, 211 n. 25. 10 Yarbrough, Friends of the Emir, pp. 73–81; the quotation is on p. 74. 11 Sijpesteijn, ‘Loyal and Knowledgeable Supporters’, pp. 343–44. Sijpesteijn builds on that point in her contribution to the present volume. See also Klasova, ‘Empire through Language’, pp. 124–25.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

translation of the dīwāns were mawlās of prominent local tribes rather than envoys sent by the central caliphal administration.12 Despite all this recent research, the idea that the translation of registers implied an overarching change of language in the ‘chancellery’, the ‘administration’, or ‘the bureaucracy’ remains pervasive but it is misleading.13 Understanding the translation of the dīwān as a ‘language reform’ or a ‘decree’ applied to the chancellery as a whole creates a puzzling contrast with the continuing use of Greek and Coptic in Egyptian administration long after the change in the language of the dīwān.14 As Sijpesteijn has noted, after the translation ‘it would take at least another century before Greek and Coptic disappeared from the administration’.15 Even though many historians recognize that there was no overnight change under ʿAbd al-Malik, the caliphate of his son and successor al-Walīd (r. 86–96/705–15) is often identified as the moment at which the supposed ‘reform’ was properly implemented.16 Another popular opinion is to see in this progressive change a gradual shift in the language of all written procedures of the Umayyad state from pre-Islamic languages to Arabic. Egypt being a ‘peripheral’ province, linguistic innovations were implemented there more gradually than they were at the centre.17 12 Longworth, ‘Islamic Bureaucrats in Late Antiquity’, ch. 3. 13 Hoyland interprets Ibn ʿAbd Rabbihi’s, al-Jahshiyārī’s, and Ibn al-Nadīm’s accounts of the translation of the dīwān as portraying an attempt ‘to replace Greek with Arabic as the language of the bureaucracy’; Hoyland, Seeing Islam as Others Saw It, p. 481 n. 89. Anthony claims that ‘accounts […] link the end of Sarjūn’s service with ʿAbd al-Malik’s ambitious enterprise to arabicize the record keeping of the Umayyad bureaucracy, called in Arabic the dīwān’; Anthony, ‘Fixing John Damascene’s Bio­graphy’, pp. 616–17. Whether ‘language policies’ are relevant for the study of the premodern world is discussed by Sofia Torallas Tovar in ‘Greek and Egyptian: Did linguistic policies exist in GraecoRoman Egypt?’ and by Arietta Papaconstantinou in ‘Arabic: Language of Empire and Language of Egypt’, p. 294, where she writes: ‘The very notion of linguistic policy is not adapted to the ancient and Medi­eval worlds. If anything came close to it, it was the combination of incentive and positive reinforcement naturally provided by the possibilities that official languages offered: allowing better careers and direct access to law and administration. Never did it involve, as has been the case in nation states, forceful imposition or prohibition, and this has generally been acknowledged by scholars’. 14 Kosei Morimoto concludes his article on the dīwān with this point, leaving the question open: ‘In a record concerning the allocation of land to the Arabs, Ibn ʿAbd al-Ḥakam states that the governor ʿAbd Allāh b. ʿAbd al-Malik was “the first person to change the dawāwīn into Arabic”: could it be that the dawāwīn here refer to the actual stipend-registers of the Arab tribes? If not, how should we interpret the fact that Greek was used well after that as the main administrative language even in documents relating to taxation?’, Morimoto, ‘The Dīwāns as Registers of the Arab Stipendaries in Early Islamic Egypt’, p. 365. 15 Sijpesteijn, Shaping a Muslim State, p. 104; Richter, ‘Language Choice in the Qurra Dossier’, pp. 189–220, at p. 216: ‘After the Arab conquest of Egypt in 641 ce, […] Greek apparently remained indispensable and continued to be used during at least the entire eighth century’; pace Cobb, ‘The Empire in Syria’, p. 242: ‘Greek continued to be used as the administrative language in Egyptian papyri until early in the reign of al-Walīd, in 706’. 16 Miles, ‘A Byzantine Weight Validated by al-Walīd’, pp. 2–3; Kameya, ‘From Qusṭāl to Jahbadh’, pp. 145–46; Mikhail, From Byzantine to Islamic Egypt, p. 116. 17 Robinson, ʿAbd al-Malik, p. 126: ‘that Arabic became the lingua franca of the Middle East, the speed of this change depending much on geo­graphy’; Anthony, ‘Fixing John Damascene’s Bio­graphy’, pp. 617–18: ‘Although certainly Arabization was real and a watershed event, the papyro­logical

91

92

m a r i e l ege n d r e

I argue in the following pages that this understanding of an innovative centre against a more passive periphery, slow at implementing the instructions received from above, is inadequate.18 It does not fit the picture we get from administrative documents dating to the Umayyad period and written in languages other than Arabic, as they display innovations of their own and evolved following a variety of dynamics. Whereas previous research has refocused the field on the change of personnel and the translation of administrative documents in various case studies, the present article will offer a reassessment of (1) what we know about the translation of the dīwān in Iraq, Syria, Egypt and Khurasan, who initiated it and (2) when they did so, (3) what was translated, (4) who completed the translation, (5) what were the economic consequences of the change of personnel and finally, (6) where the translation did not take place and why. I will begin by examining the transmission of narratives about the translation of the dīwān. Throughout this discussion, the secretaries (kātibs) will appear as the main agents of linguistic change, in the central administration as much as in the secondary chancelleries. I follow and expand on al-Qāḍī’s understanding of dīwān as register and reassess the meaning of ‘translation’ in that context. Overall, this article demonstrates that linguistic practices in the Umayyad administration were diverse and that they changed in response to dynamics that did not originate exclusively in the caliphal centre. They cannot be understood solely through the lens of ‘Arabization’, as innovations are clearly visible in other administrative languages.19

Translation Narratives and Agency First, a few remarks on the available literary source material about the translation of the dīwān are necessary. The transmitted information generally consists of two types: short reports about officials in charge of the dīwān that refer in passing to their role in the translation, or full narratives on the translation of administrative documents in the Marwanid period that describe how the events unfolded. In that sense, we can distinguish between incidental and direct references to the translation of the dīwān; the latter are called hereafter ‘translation narratives’.20 I have examined thirty-four individual sources, mainly written in Arabic, one written in Greek (Theophanes [d. 818]) and three in

evidence for the usage of Greek and Coptic in administrative documents late into the eighth century shows that such administrative measures were not enforced nearly as abruptly as the literary sources suggest’. Noth and Conrad, The Early Arabic Historical Tradition, p. 189, and Hawting, The First Dynasty of Islam, p. 63, refer to a ‘gradual shift’. 18 Similar points have been made about Islamic law and legal institutions by Brockopp in ‘The For­ma­ tion of Islamic Law’ and Tillier in ‘Les “premiers” cadis de Fusṭāṭ et les dynamiques régionales de l’innovation judiciaire’. I would like to thank Edward Zychowicz-Coghill for pointing this out to me. 19 This has also been recently argued in Cromwell, ‘Language Policy and the Administrative Framework of Early Islamic Egypt’ which I could consult shortly before the publication of the present volume. 20 Similarly to Robert Hoyland’s methodo­logy in Seeing Islam, pp. 5, 48.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

Syriac (Michael the Syrian [d. 1199], Barhebraeus [d. 1226], and the anonymous Chronicle of 1234). The dates of the sources range between the early ninth and early sixteenth centuries ce (third to tenth centuries ah); Theophanes’ Chrono­graphia is the earliest and al-Suyūṭī’s (d. 911/1505) Taʾrīkh al-khulafāʾ the latest medi­eval iteration of such narratives.21 The types of texts are diverse: works interested in secretaries and administration generally copy translation narratives.22 Incidental references are found in some geo­graphical and lexico­graphical works.23 The Greek and Syriac texts are all chrono­graphies. In the Islamic tradition, no early chrono­graphers other than Khalīfa b. Khayyāṭ (d. 240/854) refer to the translation of the dīwān.24 Perhaps the most famous early chrono­grapher, al-Ṭabarī (d. 310/923), makes no mention of this change. In his Taʾrīkh, he finds solely the change to the coinage worth mentioning among the remarkable events of the caliphate of ʿAbd al-Malik.25 Miskawayh’s (d. 421/1030) Tajārib al-umam is the earliest universal history in which the translation of the dīwān is integrated into the grand narrative of the early history of Islam and, over time, later authors follow suit.26 There are four regional focuses in the sources: Iraq, Syria, Egypt, and Khurasan. The Iraqi and Syrian translations are by far the best documented, featuring in twenty-four and twenty-one sources respectively, compared with five for Egypt and one for Khurasan. Below I examine the transmission history of each regional narrative individually along with the formation of certain motifs within the narratives and their presentation of historical actors. This work builds on a previous publication by Nancy Khalek, who has examined some of these texts with a focus on al-Jahshiyārī’s narrative choices.27

21 Theophanes, Chrono­graphia, p. 376; trans. Theophanes, Chronicle, p. 524; al-Suyūṭī, Taʾrīkh al-khulafāʾ, p. 258. 22 For example, al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, pp. 38–39, 40, 67; al-Ṣūlī, Adab al-kuttāb, p. 192. 23 See, for example, Ibn al-Faqīh, Kitāb al-buldān, p. 257, and Ibn Durayd (d. 321/933), Jamharat al-lugha, i, p. 324. 24 Khalīfa b. Khayyāṭ, Taʾrīkh, p. 30. 25 Al-Ṭabarī, Taʾrīkh al-rusul wa-l-mulūk, vi, p. 256. Al-Ṭabarī makes no mention of another of ʿAbd al-Malik’s achievements that has received great attention in modern research: the construction of the Dome of the Rock. There is no mention of the translation in al-Yaʿqūbī’s (d. after 295/908) Taʾrīkh nor in Eutychius of Alexandria’s (d. 940) Taʾrīkh al-majmūʿ ʿalā al-taḥqīq wa-l-taṣdīq. 26 Miskawayh, Tajārib al-umam, ii, pp. 386–88; the narrative appears in a section about the wuzarāʾ of the time of ʿAbd al-Malik, at the end of the chapter on his caliphate. The next universal history to include a report on the translation of the dīwān was written two centuries later: Ibn al-Athīr (d. 606/1210), al-Kāmil fī al-taʾrīkh, iii, p. 534. 27 Khalek, ‘Some Notes on the Representation of Non-Muslim Officials in al-Ǧahšiyārī’s Kitāb al-Wuzarāʾ wa-l-Kuttāb’.

93

94

m a r i e l ege n d r e

The Iraqi Narrative

A popular report about ʿAdī b. Zayd, the pre-Islamic poet from al-Ḥīra, claims that he was the first to write in Arabic in the Sasanian dīwān, which was supposedly in Ctesiphon (al-Madāʾin), but we are given no sense of what this dīwān was for or what he wrote in it.28 It is only after the Islamic conquest that concrete evidence of the administrative use of Arabic is available for Iraq in the context of the translation of the local dīwān. The transmission of reports on the authority of various authors can be identified, and they are listed in Table 2.1. Beyond the similarities that are visible across several sources, it is clear that each author was working with eclectic materials and had their own authorial input. Most Iraqi narratives (eleven sources) had access to a telling on the authority of the Basran akhbārī al-Madāʾinī (d. 228/843), as first visible in al-Balādhurī’s Futūḥ al-Buldān, who is also the only one who names his source.29 These eleven texts are by no means identical, but they share enough similarities to reveal that their authors all had access to al-Madāʾinī’s narrative in some shape or form, and in addition to other material. Al-Balādhurī does not rely solely on al-Madāʾinī’s report as he adds another, rather condensed report on the authority of Sahl b. Abī al-Ṣalt, and in this he is followed by Ibn ʿAsākir.30 In the texts of Ibn al-Nadīm, al-Māwardī (d. 450/1058), al-Nuwayrī (d. 733/1333), and al-Maqrīzī (d. 845/1442) in Kitāb al-Muqaffā al-kabīr, Sahl’s short report is integrated into al-Madāʾinī’s narrative without any reference to Sahl.31 The channels of transmission between these works are hard to reconstruct: either they had access to the same source, or all the authors writing later than al-Balādhurī worked with the Futūḥ al-buldān or with a source that had copied the translation narrative either from the Futūḥ or directly from al-Madāʾinī. The second group had access to a telling on the authority of another prominent Basran, al-Walīd b. Hishām al-Qaḥdhamī (d. 222/837–38), and that is according to al-Ṣūlī (d. 335/947) who is the only one to name him.32 Al-ʿAskarī (d. c. 1010) and al-Qalqashandī (d. 821/1418) in his Subḥ al-aʿshā both copy al-Ṣūlī’s final statement saying that the secretaries of the two Iraqs were all trained by Ṣāliḥ b. ʿAbd al-Raḥmān, the translator of the local dīwān.33 A similar report appears in al-Jahshiyārī’s text at the end of his translation narrative; he is the only one who elaborates on it as he lists a few of Ṣāliḥ’s students.34 However, it is not clear whether 28 See, for example, Ibn ʿAsākir, Taʾrīkh madīnat Dimashq, xl, p. 113: ‫فكان عدي أول من‌كتب‌بالعربية في‌ديوان كسرى‬. 29 Al-Balādhurī, Futūḥ al-buldān, pp. 300–01. My use of ‘telling’ follows Ricci, Islam Translated, p. 21. 30 Al-Balādhurī, Futūḥ al-buldān, p. 301. Ibn ‘Asākir, Ta’rīkh Dimashq, xxiii, p. 344. Sahl b. Abī al-Ṣalt was a student of Ḥasan al-Baṣrī (d. 110/728). Some say he was a Muʿtazilite: Juynboll, Muslim Tradition, p. 219. He is not a common authority for Iraqi history, but he appears in Ibn ʿAsākir’s Taʾrīkh in a report about the death date of the governor al-Ḥajjāj b. Yūsuf: Taʾrīkh madīnat Dimashq, xii, p. 199. 31 Ibn al-Nadīm, Kitāb al-Fihrist, p. 303; al-Māwardī, al-Aḥkām al-sulṭāniyya, pp. 301–02; al-Nuwayrī, Nihāyat al-arab, viii, pp. 198–200; al-Maqrīzī, Kitāb al-Muqaffā al-kabīr, iii, p. 129. 32 Al-Ṣūlī, Adab al-kuttāb, p. 192. 33 Al-ʿAskarī, Kitāb al-Awāʾil, p. 256; al-Qalqashandī, Subḥ al-aʿshā, i, p. 423. 34 Al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 39.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

al-Jahshiyārī copied this line from al-Qaḥdhamī or whether al-Ṣūlī actually got it from al-Jahshiyārī or the other way around. Indeed, al-Ṣūlī and al-Jahshiyārī are known to have interacted at the Abbasid court.35 Miskawayh’s narrative stands out, as he is the only one who added al-Madāʾinī’s dialogue on fractions (see below) to al-Jahshiyārī’s text.36 He also had access to al-Ṣūlī’s work, as indicated by the way in which he identifies Ṣāliḥ b. ʿAbd al-Raḥmān.37 These two first groups of narratives agree on the general course of events: the Persian secretary Zādhān Farrūkh b. Bīrī38 is replaced by Abū al-Walīd Ṣāliḥ b. ʿAbd al-Raḥmān al-Baṣrī, a mawlā of Tamīm, whose father was made prisoner during the conquest of Sijistan.39 Zādhān is said to have headed the dīwān of Iraq since the time of Ziyād b. Abīhi (in office c. 49–53/670–73), and Ṣāliḥ is presented as one of Zādhān’s assistants.40 The narratives of both groups revolve around dialogues between Zādhān and Ṣāliḥ, after which the latter starts the translation. This change is presented as being initiated among the secretaries. It is then completed following an order of the governor al-Ḥajjāj b. Yūsuf (in office 76–96/695–714). In the tellings on the authority of al-Madāʾinī, Zādhān says to Ṣāliḥ that al-Ḥajjāj is dependent on him because he is the only one who can do his accounting (‫)يكفيه حسابه‬. Ṣāliḥ then offers to start translating the dīwān into Arabic, assuming that this is something that al-Ḥajjāj would appreciate. After a while, Zādhān tells him to stop and to pretend that he is ill. When Zādhān dies during the revolt of Ibn al-Ashʿath (c. 80–84/699–704),41 Ṣāliḥ reports to al-Ḥajjāj that he had started the translation of the dīwān into Arabic and the governor orders him to complete it.42 As for tellings in the authority of al-Qaḥdhamī (found in the longest form in al-Ṣūlī’s text), Ṣāliḥ offers to translate the dīwān, and Zādhān agrees that he should start by translating one section. When he is presented with the result, Zādhān tells his fellow Persian secretaries to seek employment elsewhere. Subsequently, al-Ḥajjāj orders Ṣāliḥ to complete the translation.43 35 Yāqūt, Muʿjam al-udabāʾ, i, p. 275; ii, pp. 58, 166; iii, p. 146; Stasolla, ‘How a 10th Century Learned Man Reads History’, p. 225. 36 On Miskawayh’s slight rephrasing of al-Jahshiyārī’s description of the dīwān, see below, section 3. 37 The passage appears within a notice on Ṣāliḥ b. ʿAbd al-Raḥmān at the end of the section on ʿAbd al-Malik’s caliphate: Miskawayh, Tajārib al-umam, ii, pp. 386–88. 38 Sprengling, in ‘From Persian to Arabic’, pp. 177–80, 185–87, understands Bīrī to refer to Payruzān, who appears in narratives about the foundation of the dīwān of ʿUmar b. al-Khaṭṭāb. On the dīwān of ʿUmar, see below, section 3. 39 His full name is copied by al-Ṣūlī in Adab al-kuttāb, p. 192, as follows: ‫أبو الوليد صالح بن عبد الرحمن‬ ‫ وهو مولى بني مرة بن عبيد من بني سعيد بن زيد مناة بن تميم وكان من سبي سجستان‬،‫البصري‬. On his father, see Sprengling, ‘From Persian to Arabic’, pp. 191–92. 40 Sprengling, ‘From Persian to Arabic’, pp. 185–90, 195. 41 On the date of the revolt, see Bates and Shaddel, ‘Note on a Peculiar Arab-Sasanian Coinage of Ibn Al-Ash’ath’, pp. 1–2. 42 See, for instance, al-Balādhurī, Futūḥ al-buldān, pp. 300–01. 43 Pace Nancy Khalek, ‘Some Notes on the Representation of Non-Muslim Officials in al-Ǧahšiyārī’s Kitāb al-Wuzarāʾ wa-l-Kuttāb’, p. 512: ‘The private conversation between Zādān and Ṣāliḥ as found in Balāḏurī’s account is thus made into a public one instigated by al-Ḥaǧǧāǧ in al-Ṣūlī’s account’. Al-Ḥajjāj does not take part in the dialogue found in al-Ṣūlī’s text.

95

96

m a r i e l ege n d r e

As Khalek has argued, al-Jahshiyārī’s telling, which Miskawayh follows, stands out because in both his Iraqi and Syrian narratives he portrays the translation of the dīwān as a means to swiftly get rid of arrogant non-Muslim secretaries. Though Zādhān Farrūkh presents himself as irreplaceable in tellings on the authority of al-Madāʾinī as well, his claim is immaterial to the story since it is only after his death that the translation can be completed. Al-Jahshiyārī is the only author who clearly had access to al-Madāʾinī’s narrative but chose to disregard the information about Zādhān’s death during the revolt of Ibn al-Ashʿath, instead claiming that the translation took place before that in 78/697–98, immediately following the unfortunate dialogue. Khalek identifies ‘chrono­logical collapsing’ as a central feature of the Kitāb al-Wuzarāʾ wa-l-kuttāb.44 In this particular case, the introduction of a new — and uncorroborated — dating fits al-Jahshiyārī’s agenda of showing that ‘the arrogance of the non-Muslim Persian secretary causes the demise of Persian scribes as a group, and is contrasted with the humility and cleverness of the Muslim scribe who bests his superior’.45 The incertion of this new date clearly shows the kind of input that an author could have on the transmitted story, al-Jahshiyārī’s being the most visible. Al-Ṣūlī’s telling of the story on the authority of al-Qaḥdhamī agrees that the decision to undertake the translation arose from the dialogue between the two secretaries, though he does not say when this took place. This comparison further indicates that the two early fourth/tenth-century Baghdadi authors were working with similar materials but made distinctive authorial decisions. The third group of tellings comprises of incidental references to the translation, all in the form of lists of secretaries. These are the accounts of ʿAbd Allāh b. ʿAbd al-ʿAzīz al-Baghdādī (d. 255/869), Ibn ʿAbd Rabbihi (d. 328/940), and al-Qalqashandī in Subḥ al-aʿshā.46 None of these authors mentions his source, meaning they could be drawing on the same source or different ones or the two later authors could have relied on al-Baghdādī. In all these accounts, the translation is ascribed to the famous Qaḥdham, the grandfather of al-Qaḥdhamī, who is the authority on whom al-Ṣūlī relies. As shown by al-Qāḍī, all of the bio­graphical information we have on

44 Khalek, ‘Some Notes on the Representation of Non-Muslim Officials in al-Ǧahšiyārī’s Kitāb al-Wuzarāʾ wa-l-Kuttāb’, pp. 514–15: ‘This chrono­logical collapsing is a rather central aspect of al-Ǧahšiyārī’s chosen mode of representation, as events which are drawn out over time in other renditions of the same historicized incidents are artfully condensed into anecdotal exempla of a more episodic and illustrative nature’. Dominique Sourdel has also commented on incoherent datings in the Kitāb al-Wuzarāʾ wa-l-kuttāb; see his ‘La valeur littéraire et documentaire du “Livre des vizirs” d’al-Gahšiyari’, pp. 208–09. 45 Khalek, ‘Some Notes on the Representation of Non-Muslim Officials in al-Ǧahšiyārī’s Kitāb al-Wuzarāʾ wa-l-Kuttāb’, p. 511. The dates of the translation are discussed below pp. 109–14. 46 Sourdel, ‘Le “livre des secrétaires” de ʿAbdallāh al-Baġdādī’, p. 140; Ibn ʿAbd Rabbihi, al-ʿIqd al-farīd, iv, p. 252. This work draws mainly on Iraqi material, and it is a known source for al-Qalqashandī’s work; see Toral-Niehoff, ‘History in Adab Context’, pp. 63–65. Al-Qalqashandī refers to the translation in two different places in Subḥ al-aʿshā. In one, he transmits Ibn ʿAbd Rabbihi’s lists, and, in the other, he seems to rely on the report transmitted on the authority of al-Qaḥdhamī: al-Qalqashandī, Subḥ al-aʿshā, i, pp. 40, 423.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

Qaḥdham connects him to Basra, where he lived at least until the 770s or 780s. He was probably the grandson of a conquest captive from Isfahan who converted and became a mawlā of Abū Bakra al-Thaqafī.47 According to al-Jahshiyārī, he was trained by Ṣāliḥ b. ʿAbd al-Raḥmān, and he is otherwise known for his knowledge of Persian.48 He was also the teacher of both al-Madāʾinī and his own grandson al-Qaḥdhamī.49 However, in the report transmitted on the authority of al-Qaḥdhamī by al-Ṣūlī, Qaḥdham is not mentioned. Undoubtedly, had his grandfather played a prominent role in the translation of the dīwān, al-Qaḥdhamī would certainly have known and mentioned the fact, though it is possible that the mention was lost in al-Ṣūlī’s editing. Al-Maqrīzī, in his Khiṭaṭ, is the only one to note that although most agree that Ṣāliḥ b. ʿAbd al-Raḥmān was responsible for the translation of the dīwān, some hold that it was al-Qaḥdhamī.50 In all, out of the twenty-four texts containing a telling of the translation of the dīwān of Iraq, eighteen ascribe the translation to Ṣāliḥ b. ʿAbd al-Raḥmān. Twenty acknowledge the role of secretaries in the translation, and fourteen mention that the governor al-Ḥajjāj was involved in the decision. None of the sources written before the seventh/thirteenth century indicates that the caliph ʿAbd al-Malik had any involvement in the translation of the dīwān of Iraq, and he is mentioned only in incidental references to the translation. The only earlier authors to refer to him are al-Jahshiyārī, Ibn ʿAbd Rabbihi, and Miskawayh, who note that the dīwāns were translated into Arabic during the caliphate of ʿAbd al-Malik. Ibn ʿAbd Rabbihi adds that some say that the translation took place during the caliphate of ʿAbd al-Malik’s son al-Walīd.51 Only a few late references to the translation of the dīwān of Iraq attribute the initiation of the translation to ʿAbd al-Malik: Ibn al-Athīr’s (d. 606/1210) al-Kāmil fī al-taʾrīkh, Ibn al-Ṭiqṭaqā’s (d. 709/1309) al-Fakhrī fī al-ādāb al-sulṭāniyya wa-l-duwal al-islāmiyya, al-Qalqashandī’s Maʾāthir al-ināfa fī maʿālim al-khilāfa, and al-Suyūṭī’s (d. 911/1505) Taʾrīkh al-khulafāʾ.52 This means that the sources closest to the events focus on the local dynamics at play and do not support the argument that centralizing measures prompted the translation of the dīwān of Iraq.

47 According to Wadād al-Qāḍī, there is some uncertainly about his full name: Qaḥdham b. Sulaymān/ Abī Sulaymān/Sulaym b. Dhakwān. See al-Qāḍī, ‘The Names of Estates in State Registers’, p. 261. 48 Al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 39. Al-Qāḍī, ‘The Names of Estates in State Registers’, pp. 261–66. 49 Lucas, ‘Le pouvoir de la terre’, pp. 310–12. 50 Al-Maqrīzī, al-Khiṭaṭ, ed. by Sayyid, i, pp. 264–65. As shown above, al-Qaḥdhamī is one of the authorities on the topic, so he would have known had he been involved in the translation. 51 Ibn ʿAbd Rabbihi, al-ʿIqd al-farīd, v, p. 148. 52 Ibn al-Athīr, al-Kāmil fī al-taʾrīkh, iii, p. 534; Ibn al-Ṭiqṭaqā, al-Fakhrī fī al-ādāb al-sulṭāniyya, p. 123; al-Qalqashandī, Maʾāthir al-ināfa, iii, p. 345. Al-Suyūṭī claims he copied his report from al-ʿAskarī, but his text shares no similarities with the latter’s Kitāb al-Awāʾil; see al-Suyūṭī, Taʾrīkh al-khulafāʾ, p. 258.

97

98

m a r i e l ege n d r e

Table 2.1. Narratives about the translation of the Iraqi dīwān and their sources

TRANSLATION NARRATIVES

INCIDENTAL REFERENCES

On the authority of al-Madāʾinī (d. 228/843) Translation by Ṣāliḥ b. ʿAbd al-Raḥmān

On the authority of al-Qaḥdhamī (d. 222/837–38) Translation by Ṣāliḥ b. ʿAbd al-Raḥmān

al-Balādhurī (d. 279/892), Futūḥ al-buldān

al-Jahshiyārī (d. 331/942), Kitāb al-Wuzarāʾ wa-l-kuttāb (?)

al-Jahshiyārī (d. 331/942), Kitāb al-Wuzarāʾ wa-l-kuttāb

al-Ṣūlī (d. 335/947), Adab al-kuttāb

Ibn ʿAbd Rabbihi (d. 328/940), al-ʿIqd al-farīd

Ibn al-Faqīh (late third/ ninth century), Kitāb al-Buldān

Ibn al-Nadīm (d. 385/995), Kitāb al-Fihrist

al-ʿAskarī (d. c. 1010), Kitāb al-Awāʾil

al-Qalqashandī (d. 821/1418), Subḥ al-aʿshā fī ṣināʿat al-inshā

Ibn Durayd (d. 321/933), Jamharat al-lugha

al-Māwardī (d. 450/1058), al-Aḥkām al-sulṭāniyya wa-lwilāyāt al-dīniyya

al-Qalqashandī (d. 821/1418), Subḥ alaʿshā fī ṣināʿat al-inshā

 

No identifiable source Translation ordered by or in the time of ʿAbd al-Malik

al-Rāghib al-Iṣfahānī (d. c. 443/1060), Muḥāḍarāt al-udabāʾ wa-muḥāwarāt alshuʿarāʾ wa-l-bulaghāʾ

 

 

Ibn al-Athīr (d. 606/1210), al-Kāmil fī al-taʾrīkh

Ibn ʿAsākir (d. 571/1176), Taʾrīkh madīnat Dimashq al-Nuwayrī (d. 733/1333), Nihāyat al-arab fī funūn al-adab

Common source (?) Translation by Qaḥdham

No identifiable source Translation by Ṣāliḥ b. ʿAbd al-Raḥmān

al-Baghdādī (d. 255/869), Kitāb al-Mubarrad (d. 285/898), al-kuttāb wa-ṣifat al-Kāmil fī al-lugha wa-lal-dawāt wa-l-qalam adab wa-l-naḥw wa-l-taṣrīf wa-taṣrīfihā

Ibn al-Ṭiqṭaqā (d. 709/1309), al-Fakhrī fī al-ādāb al-sulṭāniyya wa-lduwal al-islāmiyya al-Qalqashandī (d. 821/1418), Maʾāthir alināfa fī maʿālim al-khilāfa

al-Dhahabī (d. 748/1348 or 753/1353), Taʾrīkh al-islām wa-wafayāt al-mashāhīr wa-l-aʿlām

al-Suyūṭī (d. 911/1505), Taʾrīkh al-khulafāʾ

Ibn Khaldūn (d. 808/1406), Taʾrīkh

 

al-Maqrīzī (d. 845/1442), alMawāʿiẓ wa-l-iʿtibār fī dhikr al-khiṭaṭ wa-l-āthār and Kitāb al-Muqaffā al-kabīr

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

The Syrian Narrative

Changes in the Syrian dīwān are documented in texts written by both Muslim and Christian authors (sixteen and five texts, respectively). In the latter group, Theophanes, Agapius of Manbij (wr. c. 940), Michael the Syrian (slightly summarized in Barhebraeus’s chrono­graphy), and the Chronicle of 1234 do not refer explicitly to a translation but, in broader terms, to a change of language from Greek into Arabic. These four tellings share many similarities, and there is considerable debate in the field on whether their common source is Theophilus of Edessa (d. 169/785). I will not attempt to settle this debate and will merely make an observation on the dating of the translation of the dīwān in these sources.53 All of them place the change of language at the time of the caliphate of al-Walīd b.  ʿAbd al-Malik, between 705 (Agapius of Manbij, who locates it ‘in the first year of his [al-Walīd’s] reign’) and 710–11 (Michael the Syrian).54 However, they disagree on the exact date, which does not inspire confidence that they knew when this took place. The different dates also show that even if their reports about the caliphate of al-Walīd look very similar, they were using varied materials or different dating logics.55 In each of the five reports, one finds an enumeration of the abominable deeds that the caliph al-Walīd perpetrated on the Christians of Damascus: the destruction of churches, the construction of the Umayyad mosque, and the change in the language of the dīwān.56 The caliph is the only authority appearing in these texts; there is no mention of secretaries. For Theophanes, al-Walīd was a ‘wretched man’; according to Michael the Syrian, he ‘hated the Christians and uprooted their churches’.57 The caliph also appears in a martyro­logy, in which he sentences Peter of Capitolias (d. 715) to death and is described as the ‘tyrant of the Arabs’.58

53 On the difficulty (or even necessity) of identifying a common source between Theophanes, Agapius, Michael the Syrian, and the Chronicle of 1234 and on the variety of texts that they are known to have reused, see Debié, ‘Theophanes’ “Oriental Source”’, pp. 377–82; Conterno, ‘Theophilos, “the More Likely Candidate”?’, pp. 393–400; Hilkens, The Anonymous Syriac Chronicle of 1234 and its Sources; Debié, L’écriture de l’histoire en syriaque, pp. 27–31. 54 Agapius, Kitāb al-ʿunwān, p. 498; Michael the Syrian, Chronique, i, xvii, pp. 451, 481. Theophanes gives the date as am 6199/706–07 ce in Chrono­graphia, p. 376; trans. Theophanes, Chronicle, p. 524. The Chron­icle of 1234, i, pp. 298–99, has 707–08. All four passages are translated in Hoyland, Theophilus of Edessa, pp. 199–200. Barhebraeus has 711 in his Chronicon, p. 115; trans. Barhebraeus, Chrono­graphy, i, p. 106. 55 See also below, pp. 114–24, for the different technical terms they use to refer to the dīwān. On the dubious chrono­logical choices made in the text of Theophanes, see Ostrogorsky, ‘Die Chrono­logie des Theophanes im 7. und 8. Jahrhundert’. 56 At the time of al-Walīd, ‘une véritable frénésie de construction s’empare alors du califat’; Borrut, Entre mémoire et pouvoir, p. 387. Borrut also refers to al-Muqaddasī’s (d. after 378/988) claim that the new mosque was meant to rival the many Syrian churches and to focus the attention of the Muslims on a place of worship of their own; see al-Muqaddasī, Aḥsan al-taqāsīm, p. 159. Al-Walīd is depicted in a very positive light by the Muslims of Syria for the same reasons he is despised by Christian writers; Hoyland, Seeing Islam as Others Saw It, p. 424 n. 119. 57 Tannous, The Making of the Medi­eval Middle East, p. 402. 58 Hoyland, Seeing Islam as Others Saw It, pp. 357–59; Shoemaker, A Prophet Has Appeared, pp. 215, 218 (translation at pp. 212–19).

99

1 00

m a r i e l ege n d r e

This is a picture shared by authors of other denominations, as al-Walīd is described as a ‘great oppressor and builder’ in the Jewish Apocalypse against the Umayyads.59 In contrast to the diverse Iraqi reports, the sixteen Islamic narratives about the translation of the dīwān of Syria show more consensus. All the sources concur that Abū Thābit Sulaymān b. Saʿd al-Khushanī/Khushaynī translated the dīwān from Greek into Arabic.60 There are more identifiable authorities after which those reports are transmitted (summarized in Table 2.2). The narrative copied by al-Balādhurī includes an entertaining anecdote about an unnamed Greek secretary who was sacked by ʿAbd al-Malik because he urinated in an inkwell. This dramatic event is said to have set off the translation of the dīwān al-Shām.61 Al-Māwardī and al-Nuwayrī agree with this course of events though only the former adds that the information comes from al-Madāʾinī (al-Balādhurī, al-Māwardī, and al-Nuwayrī also shared the same translation narrative about Iraq).62 These tellings are the only ones providing a date for the translation of the dīwān of Syria (81/700–01), unlike reports on the authority of al-Madāʾinī on Iraq, which did not include an explicit date. Ibn Khaldūn (d. 808/1406) also copies part of this narrative but he does not refer to the secretary’s misbehaviour.63 In these four texts (al-Balādhurī, al-Māwardī, al-Nuwayrī and Ibn Khaldūn), one Greek secretary is eventually mentioned: Sarjūn b. Manṣūr, who was first appointed over the dīwān by Muʿāwiya b. Abī Sufyān (governor from 18/639 then caliph 41–60/661–80).64 The focus here is not on his position in relation to the dīwān. His role in the anecdote is to lament to his fellow Greek secretaries that they are losing a prestigious position in the caliphal administration because of the translation.65 In most other tellings (ten out of the sixteen texts), the Greek secretary who misbehaves is Sarjūn himself. He is sacked for negligence and failure to follow ʿAbd al-Malik’s orders.66 Ibn al-Nadīm offers both this version of the story and 59 Lévi, ‘Une apocalypse judéo-arabe’, pp. 178–79: ‘Après lui [ʿAbd al-Malik] règnera Welid [b. ʿAbd al-Malik] fils de Merwan, qui répandra beaucoup de sang, exercera l’oppression et se construira une ville appelée de son nom’. This is presumably a text from the second half of the eighth century, as it ends with Marwān b. Muḥammad b. Marwān (r. 127–32/744–50); see Hoyland, Seeing Islam as Others Saw It, p. 317. 60 Khalīfa b. Khayyāṭ seems to be confused about Sulaymān’s identity, because when he first mentions Sulaymān, he describes him as a mawlā of Khushayn, a clan of Quḍāʿa, but later he connects Sulaymān to ʿĀmir b. Luʾayy. Khalīfa b. Khayyāṭ, Taʾrīkh, p. 187. However, the latter could be a copyist error, since the next administrator in the list on that page, who is in charge of the bureau of the seal, is ʿAmr b. al-Ḥārith, a mawlā of ʿĀmir b. Luʾayy. 61 Al-Balādhurī, Futūḥ al-buldān, p. 193. 62 Al-Māwardī, al-Aḥkām al-sulṭāniyya, p. 301; al-Nuwayrī, Nihāyat al-arab, viii, p. 198. 63 Ibn Khaldūn, Taʾrīkh, i, p. 303. 64 On Sarjūn b. Manṣūr and his identification as John of Damascus’s grandfather, not his father, see Anthony, ‘Fixing John Damascene’s Bio­graphy’, pp. 613–18. According to Sprengling, ‘From Persian to Arabic’, p. 182, ‘a conservative guess for the year of Sarjūn’s installation by Muʿāwiyah would be the year 30/650–1’. 65 See, for instance, al-Nuwayrī, Nihāyat al-arab, viii, p. 198: ،‫فدعى سرجون كاتبه فعرضه عليه فغ ّمه وخرج كئيبا‬ ‫ اطلبوا المعيشة من غير هذه الصناعة فقد قطعها هللا عنكم‬:‫ فقال لهم‬،‫فلقيه قوم من كتّاب الروم‬. 66 The exceptions are the accounts of Sibṭ b. al-Jawzī, al-Maqrīzī, and al-Qalqashandī, which are rather

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

another, according to which it was Sarjūn’s son Manṣūr b. Sarjūn (‘Manṣūr the younger’, as Sean Anthony calls him) who was fired, not Sarjūn himself, and that the dīwān was translated at the time of Hishām b. ʿAbd al-Malik (r. 105–25/724–43) by Sulaymān b. Saʿd.67 This is incompatible with what is otherwise known about Sulaymān’s career, as the last caliph he is known to have served is ʿUmar b.  ʿAbd al-ʿAzīz (r. 99–101/717–20).68 It is notable that no text of the Christian tradition makes reference to Sarjūn b. Manṣūr in relation to the change of language of the dīwān. Although there are some positive representations of Sarjūn in the Chronicle of 1234, one reason the Syriac Orthodox texts would not have wanted to focus on the fate of the Manṣūr family and on the Christian secretaries working for the Umayyads in Damascus is that those positions seem to have been monopolized by men those later sources would identify as Melkites.69 We could also assume that Syriac Orthodox writers, retrospectively, would rejoice that Melkites had lost their prestigious positions in the administration. However, in practice, the change in the language of the dīwān was not accompanied by the complete replacement of Christian administrators: even after the translation, Christians, including the Manṣūrs, still made up a significant proportion of the administrative staff.70 Coming back to the Islamic tradition, as al-Ṣūlī’s report on the translation of the dīwān of Syria directly follows his account of the translation in Iraq, with no interruption in the text nor a new isnād given, it is possible that both reports were transmitted on the authority of al-Qaḥdhamī.71 As in the case of Iraq, al-Ṣūlī’s telling of the events in Syria is similar to those of al-ʿAskarī and al-Qalqashandī in Subḥ al-aʿshā. The latter also used the report transmitted by al-Jahshiyārī and Ibn ʿAbd Rabbihi.72 Ibn ʿAsākir copied two accounts of the translation of the dīwān of Syria: one incidental report on the authority of Khalīfa b. Khayyāṭ and a long and unique translation narrative on the authority of Sulaymān b. Saʿd himself that we will come back to shortly.73 Al-Ṣafadī (d. 764/1363) used both of Ibn ʿAsākir’s reports to produce his telling. Sibṭ b. al-Jawzī’s (d. 654/1256) indirect reference to the translation is on the authority of the Syrian-based Abū al-Ḥusayn al-Rāzī

short and mention neither the secretary who was replaced in connection with the translation nor the reason for the replacement. Sibṭ b. al-Jawzī, Mirʾāt al-zamān, x, p. 269; al-Maqrīzī, al-Khiṭaṭ, ed. by Sayyid, i, p. 265; al-Qalqashandī, Subḥ al-aʿshā, i, p. 423; al-Qalqashandī, Maʾāthir al-ināfa, iii, p. 345. 67 Ibn al-Nadīm, Kitāb al-Fihrist, p. 303. In his second report, he agrees with most other sources that Sarjūn was replaced by Sulaymān at the time of ʿAbd al-Malik. Anthony, ‘Fixing John Damascene’s Bio­graphy’, pp. 618–22. 68 Ibn ʿAsākir, Taʾrīkh madīnat Dimashq, xxii, p. 317. 69 As noted by Muriel Debié, Sarjūn was ‘positively appreciated in the Syriac chronicles for being well known and highly considered by the Arabs’, with reference to the Chronicle of 1234: Muriel Debié, ‘Christians in the Service of the Caliph’, p. 62. 70 See below, pp. 124ff. 71 Al-Ṣūlī, Adab al-kuttāb, pp. 192–93. 72 Al-ʿAskarī, Kitāb al-Awāʾil, p. 256; al-Qalqashandī, Subḥ al-aʿshā, i, p. 423. 73 Ibn ʿAsākir, Taʾrīkh madīnat Dimashq, xxii, pp. 320–21.

101

102

m a r i e l ege n d r e

(d. 347/958).74 Whereas the texts of the Christian tradition all tether the change of language to Damascus, in the Islamic tradition the only account that specifies a location is Ibn ʿAsākir’s second telling on the authority of Sulaymān b. Saʿd: he reports a dialogue between Sulaymān and ʿAbd al-Malik about the replacement of Christian secretaries that took place in al-Ṣinnabra, a winter residence of the Umayyads in the jund of al-Urdunn. However, it is unclear whether the translation is supposed to have been initiated there.75 Overall, agency in initiating the translation is mostly presented as shared between a secretary who is also a mawlā and the highest authority in the province, in this case the caliph. Sibṭ b. al-Jawzī, al-Ṣafadī, and al-Maqrīzī in his Khiṭaṭ focus solely on Sulaymān’s translation with no reference to the caliph beyond the fact that the translation took place during his caliphate.76 Only in al-Qalqashandī’s Maʾāthir al-ināfa fī maʿālim al-khilāfa do we find a very short report on the translation focusing solely on ʿAbd al-Malik’s role.77 This means that in the long term, the memory of the secretaries’ role in the translation was more enduring in the Syrian narratives than it was in the Iraqi narratives. Beyond the variety of authorities in the transmission of the Islamic narratives concerning the Syrian dīwān, their description of the course of events is remarkably stable. Similarly to al-Jahshiyārī’s report on Iraq, these narratives condemn the behaviour and unreliability of the Greek secretaries. This is presented across most reports as the main reason behind the translation. However, the power dynamics between old and new secretaries are different. Sulaymān b. Saʿd is not portrayed as a subordinate of Sarjūn, in most tellings it is said that he was in charge of correspondence for ʿAbd al-Malik.78 Most sources that do not rely on al-Madāʾinī reproduce a dialogue between Sulaymān and ʿAbd al-Malik that prompted the decision to translate the dīwān. Only in Khalīfa b. Khayyāṭ’s incidental reference to the translation and in Ibn ʿAsākir’s second narrative, on the authority of Sulaymān b. Saʿd, must the translation wait for Sarjūn to die, a point to which we shall return below. Ibn ʿAsākir’s second telling is notable for another reason. The great majority of the reports pertaining to Iraq and Syria (and, as will be seen, to Egypt as well) do

74 Sibṭ b. al-Jawzī, Mirʾāt al-zamān, x, p. 269. Al-Rāzī’s Tasmiyat kuttāb umarāʾ Dimashq is a well-known source of Ibn ʿAsākir’s Taʾrīkh madīnat Dimashq, as evident, for instance, in the latter’s short bio­ graphy of Sarjūn b. Manṣūr (xx, p. 161) and in the section on Sulaymān b. Saʿd (xxii, p. 317). Al-Rāzī’s Tasmiya is also named as a source for al-Ṣafadī’s Syrian translation narrative in his Kitāb al-Wāfī bi-l-wafayāt, xv, p. 239. See also Conrad, Abū’l-Ḥusain al-Rāzī und seine Schriften and Scheiner, ‘Ibn ʿAsākir’s Virtual Library as Reflected in his Taʾrīkh madīnat Dimashq’, pp. 213–19. 75 Ibn ʿAsākir, Taʾrīkh madīnat Dimashq, xxii, p. 320. Al-Maqrīzī (in the Khiṭaṭ) also refers to Damascus when he explains how the dīwān was organized after the conquest, see below, pp. 114–24. On al-Ṣinnabra, see Borrut, Entre mémoire et pouvoir, pp. 397–411; Borrut, ‘Pouvoir mobile et construction de l’espace dans les premiers siècles de l’islam’, pp. 255–65. 76 Sibṭ b. al-Jawzī, Mirʾāt al-zamān, x, p. 269; al-Ṣafadī, Kitāb al-Wāfī bi-l-wafayāt, xv, p. 239; al-Maqrīzī, al-Khiṭaṭ, ed. by Sayyid, i, p. 265. 77 Al-Qalqashandī, Maʾāthir al-ināfa, iii, p. 345. 78 See, for example, Ibn al-Nadīm, Kitāb al-Fihrist, p. 303: ‫وكان على كتابة الرسائل‬.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

not make explicit reference to the religious affiliations of the secretaries beyond the common epithet for Sarjūn b. Manṣūr, ‘the Christian’ (‫)النصراني‬.79 Ibn ʿAsākir’s report on the authority of Sulaymān b. Saʿd and al-Jahshiyārī’s narrative on the dīwān of Khurasan, examined below, are exceptions. Ibn ʿAsākir writes that ʿAbd al-Malik told Sulaymān b. Saʿd that he was annoyed by the control Christians had over Muslims. When Sarjūn fell ill and was asked by ʿAbd al-Malik who could replace him, he answered: ‘If a Muslim, then Sulaymān b. Saʿd; and if a Christian, then a certain man from Baʿlabak’. After Sarjūn’s death, he was replaced by Sulaymān b. Saʿd, who was the first Muslim to be put in charge of the dīwān, and this is why he translated it.80 As mentioned above, Ibn ʿAsākir gives an isnād for this report that goes back to Sulaymān b. Saʿd himself. Though any isnād could in theory be forged, Yarbrough has argued that certain features of this narrative inspire confidence that it was transmitted ‘intact’.81 As will be seen below, the account contains archaic termino­logy that further supports an Umayyad context of production. Yarbrough also identifies in this narrative early evidence of an attitude against the employment of non-Muslims in the Islamic state, before the development of a prescriptive discourse on the topic after the mid-third/ninth century.82 Overall, we find varying opinions in the narratives about the translation of the dīwān of Syria, ranging from discomfort with the employment of non-Muslim secretaries in Ibn ʿAsākir’s narrative on the authority of Sulaymān b. Saʿd to outright condemnation of non-Muslim secretaries’ attitudes and behaviour in most others. A unique report on the translation of the dīwān of Khurasan is a further example of the discourse against the employment of non-Muslim secretaries.

79 See, for instance, al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 40. 80 Ibn ʿAsākir, Taʾrīkh madīnat Dimashq, xxii, pp. 320–21: ‫قال سليمان فخال بي عبد الملك فقال إن ما يلي النصارى من‬ ‫أمور المسلمين لم يزل يغيظني وإني لغالم أفد إلى معاوية ثم قال لقد أردت أن أذكر ذلك أيام مروان فذكر شيئا منعه منه ثم دعاني‬ ‫إلى أن يوليني عمل سرجون قال فهبت ذلك ولم أجبه إليه وذكرت بعض ما أتخوف أال أعرف عمله قال إني بعون هللا أوثق مني‬ ‫لك بعلمك فبينا هو يذكر ذاك إذ سمع تنحنح روح بن زنباع وكان ال يحجب فقال لي تنح فإن روحا ال يكتم شيئا قال ثم إنه قال لروح‬ ‫إني كلمت كاتب جندكم هذا وروح يومئذ على األردن فذكر له ما ذكر لي من أمر سرجون ثم دخل وتركني وروحا فأقبل علي‬ ‫روح يحثني أن أقبل ما عرض علي من ذلك حتى كان من قوله إن أمير المؤمنين قد اهتم من هذا بما تركه غيره من الخلفاء فإن‬ ‫أنت تركت أن تقبل ذلك تخوفت أن يدوم األمر على ما كان عليه من تولية النصارى قال واشتكى سرجون بعد ذلك مرضه الذي‬ ‫مات فيه فأرسل إليه عبد الملك من ترى لعملك الذي أنت فيه قال إن كان من المسلمين فسليمان بن سعد وإن كان من النصارى‬ .‫وحولها بالعربية‬ ّ ‫ففالن رجل من أهل بعلبك فمات سرجون وولى عبد الملك سليمان بن سعد فهو أول مسلم ولي الدواوين كلها‬ The passage is partially translated in Yarbrough, Friends of the Emir, pp. 73–74. 81 The passage ‘lacks a pervasive parabolic tone and contains several details that tether it to its setting, such as the names of Sulaymān, Rawḥ, the jund of al-Urdunn, and al-Ṣinnabrah. It also preserves several “blanks”—narrative junctures that call for detail, such as the reason that ʿAbd al-Malik did not confront Marwān or the name of the Christian from Baʿlabakk — that are left unfilled, suggesting that it was transmitted intact rather than embellished’. Yarbrough, Friends of the Emir, p. 74. On mechanisms of transmission in Ibn ʿAsākir’s Taʾrīkh madīnat Dimashq, see Zychowicz-Coghill, The First Arabic Annals, pp. 15–19. 82 Yarbrough, Friends of the Emir, pp. 88–109.

103

104

m a r i e l ege n d r e

Table 2.2. Narratives about the translation of the Syrian dīwān and their sources.

TRANSLATION NARRATIVES On the authority of al-Madāʾinī (d. 228/843)

On the authority of al-Qaḥdhamī (d. 222/837– 38) (?)

Common source (?)

INCIDENTAL REFERENCES On the Khalīfa authority of b. Khayyāṭ Sulaymān (d. 240/854), b. Saʿd Taʾrīkh (d. c. 101/720)

al-Balādhurī (d. 279/892), Futūḥ al-buldān

al-Jahshiyārī (d. 331/942), al-Ṣūlī Kitāb al(d. 335/947), Wuzarāʾ Adab al-kuttāb wa-l-kuttāb

al-Māwardī (d. 450/1058), al-Aḥkām al-sulṭāniyya wa-l-wilāyāt aldīniyya

Ibn ʿAbd al-ʿAskarī Rabbihi (d. c. 1010), (d. 328/940), Kitāb al-Awāʾil al-ʿIqd al-farīd

al-Ṣafadī (d. 764/1363), Kitāb al-Wāfī bi-l-wafayāt

al-Qalqashandī (d. 821/1418), Subḥ al-aʿshā fī ṣināʿat al-inshā

al-Qalqashandī (d. 821/1418), Maʾāthir alināfa fī maʿālim al-khilāfa

al-Nuwayrī (d. 733/1333), Nihāyat al-arab fī funūn al-adab

Ibn ʿAsākir (d. 571/1176), Taʾrīkh madīnat Dimashq

On the authority of Abū al-Ḥusayn al-Rāzī (d. 347/958)

Common source (?) Translation taking place under the caliph Hishām

Ibn ʿAsākir Sibṭ b. al-Jawzī (d. 571/1176), (d. 654/1256), Ibn al-Nadīm (d. 385/995), Taʾrīkh Mirʾāt madīnat al-zamān fī Kitāb al-Fihrist Dimashq taʾrīkh al-aʿyān al-Maqrīzī (d. 845/1442), al-Mawāʿiẓ wa-l-iʿtibār fī dhikr al-khiṭaṭ wa-l-āthār  

No identifiable source

Ibn Khaldūn (d. 808/1406), Taʾrīkh

 

The Khurasani Narrative

The translation of the Khurasani dīwān is addressed in a single source, al-Jahshiyārī’s Kitāb al-Wuzarāʾ wa-l-kuttāb.83 Since this work is arranged according to caliphates, the event is not treated together with the translations of the Iraqi and Syrian dīwāns. According to al-Jahshiyārī, the translation took place in 124/741–42, and his account presents Khurasan as a dependency of Iraq. The governor of Iraq, Yūsuf b. ʿUmar al-Thaqafī (in office 120–27/738–44) is said to have written to his appointee Naṣr b. Sayyār (in office 120–31/738–48) to order him not to employ anyone from the polytheists (ahl al-shirk) after which Naṣr decides to have the dīwān translated from Persian into Arabic.84 A certain Isḥāq b. Ṭulayb of the Banū Nahshal is assigned to the task. The change of personnel is here again explicit and the religious identity of the secretaries is presented as the motive behind the translation. Al-Jahshiyārī’s 83 Al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 67. 84 Elsewhere in the text referred as majūs, or Zoroastrians: ‫وكان أكثر كتّاب خراسان إذ ذاك مجوس وكانت‬ ‫الحسبانات بالفرسيّة‬. Al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 67.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

text resonates with other descriptions of the governorate of Yūsuf b. ʿUmar, who succeeded Khālid b. ʿAbd Allāh al-Qasrī (in office 105–20/723–38). Khālid’s governorate is known as a time in which non-Muslims dominated administrative positions in Iraq, especially visible in Kufa, something that angered the local community of Muslims and over time, also raising the ire of the caliph Hishām b. ʿAbd al-Malik.85 As Yarbrough has argued, in no other context in the first two centuries of Islam can we identify such animosity against the appointment of non-Muslims in the administration.86 This phenomenon did not escape the attention of al-Jahshiyārī: the bulk of his section on the caliphate of Hishām focuses on Iraq, discussing Khālid al-Qasrī, Yūsuf b. ʿUmar (including a section on the most famous secretary of the time, Qaḥdham), and Yūsuf ’s appointment of Naṣr b. Sayyār over Khurasan.87 Following Yarbrough, it is precisely in the context of opposition to Khālid al-Qasrī and his non-Muslim staff that al-Jahshiyārī’s report on the translation of the dīwān of Khurasan should be understood, as Yūsuf b.  ʿUmar removed the non-Muslim officials who had formed the backbone of Khālid’s administration. For the same reason, this report tells us little about the dīwān of Khurasan. The translation narrative might well have been used as a topos to elaborate on the context of Khālid al-Qasrī’s dismissal and on the far-reaching measures that followed it. However, with only one report available on the topic, it is hard to reach any definitive conclusions on the historicity of the translation of the Khurasani dīwān and I will leave this unresolved. The wider context of this period is also that of the years preceeding the disintegration of Umayyad control in the region in the face of the Third Fitna (125–32/743–50) and the rise of Abū Muslim (d. 137/755).88 The Egyptian Narrative

The final regional narrative concerns the translation of the dīwān of Egypt, and all five available texts on this event were written by Egyptian authors. They focus on the change of personnel more than they do on the translation. Athanasius bar Gūmōyē, the well-known Edessan administrator of the governor ʿAbd al-ʿAzīz b. Marwān (in office 65–86/685–705), was dismissed in 86/705 by the new governor, ʿAbd Allāh b. ʿAbd al-Malik (in office 85–90/705–09). He was replaced by Yarbūʿ al-Fazārī from Homs, a mawlā of the Banū al-Dhiyāl, about whom nothing is otherwise known.89 Both Athanasius and Yarbūʿ hailed from greater Syria, meaning that they were outsiders to the province of Egypt, much like the governors of the time. Undoubtedly, Yarbūʿ came to Egypt with ʿAbd Allāh b. ʿAbd al-Malik,

85 On the dismissal of Khālid al-Qasrī, see Lucas, ‘A State Letter from a Marwanid Caliph to his Governor of Iraq’. 86 Yarbrough, Friends of the Emir, pp. 58–61. 87 Al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, pp. 59–67. 88 Haug, The Eastern Frontier, pp. 149–52. 89 Bouderbala, ‘Ǧund Miṣr’, p. 201 n. 116.

105

1 06

m a r i e l ege n d r e

who had previously served as governor of Homs.90 Unlike Sarjūn b. Manṣūr and Zādhān Farrūkh, Athanasius was not a member of an established family local to the province in which he served. However, he might have been employed in the Egyptian administration for a lengthy period similar to the tenures of his Iraqi and Syrian counterparts. According to the Chronicle of 1234, Athanasius remained in Egypt for twenty-one years, in line with the length of the governorate of ʿAbd al-ʿAzīz b. Marwān.91 Al-Kindī (d. 350/969) concurs, mentioning Ashinās (‫)أشناس‬ only during the governorate of ʿAbd al-ʿAzīz b. Marwān.92 The History of the Patriarchs reports that Athanasius was appointed as one of the secretaries of ʿAbd al-Azīz by the latter’s father, Marwān b. al-Ḥakam (r. 64–65/684–85), but the text does not make clear where Athanasius was before his appointment.93 He might have already been in Egypt, like Zādhān Farrūkh, who was reintroduced in the administration of Basra after the Zubayrid interlude in Iraq.94 An obscure passage in Ibn ʿAbd al-Ḥakam’s (d. 257/871) Futūḥ Miṣr might imply this as well. The text refers to a certain Antanās (‫ )أنتناس‬who asks the caliph Muʿāwiya b. Abī Sufyān for a residence close to the dīwān, a request that the caliph approved.95 The favour in itself speaks to the status of this Antanās.96 The report says that he was in charge of the military and fiscal administration of Maslama (‫ — )صاحب الجند وخراج مسلمة‬that is, Maslama b. Mukhallad al-Anṣārī (in office 47–62/668–81), the longest-serving governor of Sufyanid Egypt. The residence in question was previously owned by Wardān, an Armenian mawlā of ʿAmr b. al-ʿĀṣ (in office 20–25/641–46 and 38–43/649–64), whom Ibn ʿAbd al-Ḥakam describes as in charge of the fiscal administration (‫ )صاحب الخراج‬for the conqueror and first governor of the province of Egypt.97 Supposedly, Athanasius replaced Wardān in his post at the same time as

90 Borrut, ‘ʿAbdallāh b. ʿAbd al-Malik b. Marwān’. 91 See below, p. 138, where the relevant passage is quoted. 92 Al-Kindī, Kitāb al-Wulāt, pp. 50, 59. 93 Alexandrinische Patriarchengeschichte von S. Marcus bis Michael I, p. 116; History of the Patriarchs, p. 12. 94 Sprengling, ‘From Persian to Arabic’, pp. 188–89. According to al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 44, Muṣʿab b. al-Zubayr was known to have employed a certain Sārzād for fiscal administration. 95 The residence was close to the mosque of ʿAmr b. al-ʿĀṣ, in the khiṭṭa of the ahl al-rāya. Ibn ʿAbd al-Ḥakam, Futūḥ Miṣr, p. 98: ‫ سأل معاوية أن يجعل له منزال قرب الديوان‬،‫أنتناس صاحب الجند وخراج مسلمة‬ ‫ يأمره أن يشترى له منزل وردان‬،‫فكتب معاوية إلى مسلمة بن مخلّد‬. Muʿāwiya is mentioned only by his first name, but it is likely that the reference is to the caliph and not to Muʿāwiya b. Ḥudayj al-Tujībī (d. 52/672), although the latter is known to have been in charge of the allocations of khiṭṭas, as Muʿāwiya b. Ḥudayj died around the same time as Wardān (see below). On Muʿāwiya b. Ḥudayj, see Bouderbala, ‘Ǧund Miṣr’, pp. 127–28. In addition, the Muʿāwiya in Ibn ʿAbd al-Ḥakam’s passage writes to Maslama and orders him to allocate the dār to Athanasius, which implies that he had authority over the governor of Egypt. I would like to thank Edward Zychowicz-Coghill for pointing this out to me. 96 This spelling of the name is not attested in any other source that I am aware of. As mentioned above, al-Kindī spells Athanasius’s name as ‫أشناس‬. In al-Jahshiyārī’s text, he is ‫ياناس بن خمايا‬: al-Wuzarāʾ wa-lkuttāb, pp. 34–35. In the History of the Patriarchs, he is ‫أتناسيوس‬: Alexandrinische Patriarchengeschichte, p. 116; History of the Patriarchs, p. 12. 97 His full name is reported as Abū ʿUbayd Allāh Wardān al-Rūmī; see Sijpesteijn, ‘A Seventh/EighthCentury List of Companions from Fusṭāṭ’, pp. 374–75. See also Ibn ʿAbd al-Ḥakam, Futūḥ Miṣr, pp. 33, 78, 74, 86, 93, 98, 100, 124, 136, 177.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

he asked to take over the latter’s residence, which was located conveniently close to the dīwān. In this report, then, the dīwān is associated with a building.98 Sobhi Bouderbala places this episode before Wardān’s death in 53/672–73, so Athanasius would have been appointed before that date.99 The only Athanasius who would warrant such treatment in the early Islamic administration of Egypt is Athanasius bar Gūmōyē. If this passage refers to the beginning of his career, he remained in the highest ranks of Umayyad administration in Egypt for more than thirty years, enjoying a longer career than any of the Umayyad governors of Egypt. Coming back to the translation narrative, Ibn ʿAbd al-Ḥakam, al-Kindī, and al-Maqrīzī in two of his works agree that the governor ʿAbd Allāh b. ʿAbd al-Malik ordered the translation; according to al-Kindī he did so only once he had been confirmed in this position by his brother, the caliph al-Walīd.100 Only al-Qalqashandī attributes the translation to ʿAbd al-ʿAzīz b. Marwān instead.101 In Ibn ʿAbd al-Ḥakam’s text, the dīwān was translated from ʿajamiyya (‘the language of the non-Arabs’ or ‘non-Arabic language’), which al-Kindī interprets to mean Coptic, the only ‘language of the non-Arabs’ that was used in Egypt at the time when he was writing in the fourth/tenth century. He is followed in this by al-Maqrīzī and al-Qalqashandī.102 However, in the late first/seventh and early second/eighth centuries, administration and accounting in Egypt were dominated by Greek, and the dīwān was undoubtedly in that language before its translation into Arabic.103 When reports about the translation of the dīwān are found in the form of awāʾil (accounts of ‘firsts’), Albrecht Noth and Lawrence Conrad deem them ‘untendentious’, concluding that they ‘may be considered to be objectively correct’.104 As our overview of the regional narratives comes to a close, I have made no attempt to eliminate some reports more than others as ahistorical. I will

98 Ibn ʿAbd al-Ḥakam’s text further indicates that when Athanasius died, the residence was given to ʿUmar b. Marwān b. al-Ḥakam, who is mainly known for having acted as interim governor of Egypt after ʿAbd al-Azīz b. Marwān passed away until his successor, ʿAbd Allāh b. ʿAbd al-Malik, reached the province: Ibn ʿAbd al-Ḥakam, Futūḥ Miṣr, p. 237. According to Ibn Yūnus, ʿUmar died in 115/733–34, but not much is otherwise known about him. Ibn Yūnus, Taʾrīkh, ii, p. 157. 99 Bouderbala, ‘Ǧund Miṣr’, p. 134. 100 Ibn ʿAbd al-Ḥakam, Futūh Misr, p. 122; al-Kindī, Kitāb al-Wulāt, pp. 58–59; al-Maqrīzī, al-Khiṭaṭ, ed. by Sayyid, i, p. 264; al-Maqrīzī, Kitāb al-Muqaffā al-kabīr, iv, p. 322; Sijpesteijn, ‘Multilingual Archives and Documents in Post-Conquest Egypt’, p. 106; Sijpesteijn, Shaping a Muslim State, p. 104. 101 Al-Qalqashandī, Subḥ al-aʿshā, i, p. 423. 102 And by a few modern historians; see, for example, al-Qāḍī, ‘The Names of Estates in State Registers’, p. 255; Dūrī, Early Islamic Institutions, pp. 161, 170. 103 See below, pp. 124ff.; Sprengling, ‘From Persian to Arabic’, p. 213. 104 Noth and Conrad, The Early Arabic Historical Tradition, p. 106, though when they examine lists of administrators, which tend to name the secretaries involved in the translation of the dīwān, they doubt their authenticity beyond the names of governors, pp. 101–02. References to the translations of the regional dīwāns are found in some sources in the form of awāʾil: Khalīfa b. Khayyāṭ, Taʾrīkh, p. 302 (Syria); Ibn ʿAbd al-Ḥakam, Futūh Misr, p. 122 (Egypt); al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 67 (Khurasan); al-ʿAskarī, Kitāb al-Awāʾil, p. 256 (Iraq and Syria); Ibn al-Athīr, al-Kāmil fī al-taʾrīkh, iii, p. 534 (Iraq); Ibn ʿAsākir, Taʾrīkh madīnat Dimashq, xxii, pp. 320–21 (Syria on the authority of Khalīfa b. Khayyāṭ); Sibṭ b. al-Jawzī, Mirʾāt al-zamān, x, p. 269; al-Ṣafadī, Kitāb al-Wāfī bi-l-wafayāt, xv, p. 239;

107

1 08

m a r i e l ege n d r e

address the historicity of the translation of the dīwān throughout the rest of this article. At this stage, we can say that some features of the narratives are obvious embellishments: anecdotes about administrators urinating in inkwells were composed for the intended audience, and references to the unreliability of Greek or Persian scribes also clearly fit a purpose. In the earliest reports, secretaries are presented as the key agents in how the translation of the dīwān is remembered, in dialogue with the highest regional authorities. The translation narratives penned between the third/ninth and sixth/twelfth centuries acknowledge secretarial agency in the shaping of the Marwanid administration especially in the cases of Iraq and Syria, for which the narratives are the most elaborate. One secretary, Ṣāliḥ b. ʿAbd al-Raḥmān, is even said to have started the translation of the dīwān on his own initiative in Iraq. These narratives paint a picture of the rise of a new class of mawlā secretaries and their productive relationship with their superior — the caliph or the governor.105 As Longworth has argued, these mawlās were for the most part locals who became affiliated with prominent tribes of the ruling elite; as such, they were representatives of new local elites.106 Over time, these figures became an integral part of the self-legitimizing discourse of the secretarial class, and the reports transmitted by Ibn ʿAsākir on the authority of Sulaymān b. Saʿd could be the earliest trace of that discourse.107 However, later sources, especially those that make only incidental reference to the translation, tend to simplify the course of events and to attribute the translation of the dīwāns of Iraq and Syria to ʿAbd al-Malik, an approach followed by several modern historians.108 The particular perspective of each work also needs to be taken into account. In Maʾāthir al-ināfa fī

105

106

107 108

al-Dhahabī, Taʾrīkh al-islām, vii, p. 110; al-Qalqashandī, Subḥ al-aʿshā, i, p. 423 (Iraq, Syria, and Egypt); al-Maqrīzī, al-Khiṭaṭ, ed. by Sayyid, i, p. 264 (Egypt and Iraq); al-Suyūṭī, Taʾrīkh al-khulafāʾ, p. 258 (Iraq). According to the most famous mawlā secretary of the Umayyad period, ʿAbd al-Ḥamīd al-Kātib, this relationship might not have been that easy to navigate on the part of the secretaries; see al-Qāḍī, ‘Identity Formation of the Bureaucracy of the Early Islamic State’, pp. 148–51. On ʿAbd al-Ḥamīd, see below, pp. 124ff. A few sources posit that some non-Muslim secretaries were also mawlās, yet only al-Ṭabarī claims that Sarjūn b. Manṣūr was a mawlā of Muʿāwiya b. Abī Sufyān, a piece of information he claims from Ibn al-Kalbī (d. 204/819); al-Ṭabarī, Taʾrīkh al-rusul wa-l-mulūk, v, p. 356. As mentioned above (at n. 97), Wardān, who headed the dīwān of Egypt before Athanasius bar Gūmōyē, is presented as a mawlā of ʿAmr b. al-ʿĀṣ. Longworth, ‘Islamic Bureaucrats in Late Antiquity’, p. 113: ‘Ṣāliḥ, and his notable students, were not Damascene members of the Umayyad political elite, but local elites already prominent in the region. Therefore, the literary sources provide no indication that the conversion was an endeavor at the haste of ʿAbd al-Malik for the capital’s direct control of the administration; nor is there any evidence of this with the actual makeup of the bureaucracy which remained local’. See also pp. 143–44: ‘The administrative reforms associated with al-Ḥajjāj and ʿAbd al-Malik were not ruptures opening up new careers for social climbers, rather they were the progressive takeover of regional administration by established elites’. As Yarbrough puts it: ‘The men who produced the discourse were primarily concerned to secure and advance their own positions in particular fields of competition, sometimes in relation to state power’; Yarbrough, Friends of the Emir, p. 7. See also below, pp. 129–33. For instance, Ibn al-Athīr mentions the translation in a rather condensed form: among awāʾils attributed to ʿAbd al-Malik, he remarks that the latter was the first to translate the dīwān (meaning the dīwān of Iraq) from Persian to Arabic. Ibn al-Athīr, al-Kāmil fī al-taʾrīkh, iii, p. 534.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

maʿālim al-khilāfa, al-Qalqashandī attributes the translation to ʿAbd al-Malik as his focus in that work is on caliphs, whereas in Subḥ al-aʿshā his focus is on secretaries, so he refers to Qaḥdham and Ṣāliḥ b. ʿAbd al-Raḥmān.109 It is noteworthy that the translation of the regional dīwāns is not mentioned in the bio­graphical accounts of ʿAbd al-Malik or al-Ḥajjāj b. Yūsuf in al-Balādhurī’s Ansāb al-ashrāf or in Ibn ʿAsākir’s Taʾrīkh madīnat Dimashq. From the perspective of these works, it is not part of the memorable record of these caliphs’ achievements. In that sense, an interpretation of the translation of the dīwān as a centralized decision made by ʿAbd al-Malik gives precedence to post-sixth/twelfth-century sources. That a number of late sources notably misspelled the names of secretaries is clear testimony to the secretaries’ oblivion or the late sources’ disinterest in their story.110 I will now examine the dating of the different translations to assess whether the translations of the various dīwāns can be connected to each other.

Dating the Translations and the Dynamics of Administrative Change The reports examined above give no sense that the translation of the dīwān was carried out at the same time across regions, and there is no clear consensus on the dates of the Iraqi and Syrian translations. According to al-Balādhurī’s Futūḥ (which follows al-Madāʾinī), the Syrian translation was the first to take place. It is the only one for which he provides an explicit date: 81/700–01.111 In his narrative on the Iraqi translation, he mentions only that it was completed after the death of Zādhān Farrūkh, who was killed during the revolt of Ibn al-Ashʿath (c. 80–84/699–704).112 According to Khalīfa b. Khayyāṭ, Zādhān Farrūkh died at the battle of Dayr al-Jamājim in 82/701–02.113 This would have been a propitious time to translate existing registers or even to create new ones, as a number of sources, including al-Balādhurī, claim that some registers were burned at the time of the battle.114

109 Al-Qalqashandī, Subḥ al-aʿshā, i, pp. 40, 423; and Maʾāthir al-ināfa, iii, p. 345. 110 However, I am relying only on editions for this point, as I have not been able to check the original manu­scripts. For instance, Qaḥdham’s name in al-Maqrīzī’s Khiṭaṭ is spelled as ‫ مخزوم‬in the Bulaq edition (i, p. 98) and as ‫ قُحْ ُزم‬in the edition of Ayman Fuʾād Sayyid (i, p. 264). Both editions have Marwān Shāh b. Zādhān (‫ )مروان شاه بن زادان‬instead of Mardānshāh (Bulaq ed., i, p. 98; ed. by Sayyid, i, p. 265). In Ibn Khaldūn, Tarīkh, i, p. 303, we find ‫ سرحون‬for Sarjūn. 111 Al-Balādhurī, Futūḥ al-buldān, p. 193, followed by al-Māwardī, al-Aḥkām al-sulṭāniyya, p. 301, and al-Nuwayrī, Nihāyat al-arab, viii, p. 198. 112 Al-Balādhurī, Futūḥ al-buldān, p. 300. 113 Khalīfa b. Khayyāṭ, Taʾrīkh, p. 288. He also reports (p. 144) that Zādhān was advising al-Ḥajjāj at the beginning of the revolt. 114 Al-Balādhurī, Futūḥ al-buldān, pp. 272–73: ‫فلما كانت وقعة‌الجماجم أحرق الناس‌الديوان فأخذ كل قوم ما‬ ‫يليهم‬. According to Noëmie Lucas, these were the registers of ṣawāfī lands: ‘Le pouvoir de la terre’, pp. 229–30. This is not the only account of registers burning in Iraq, though the purpose of the following anecdote is clearly to praise the outstanding abilities of Zādhān Farrūkh: al-Jahshiyārī reports that once the dīwān of Basra burned and Zādhān Farrūkh managed to reproduce it entirely

109

110

m a r i e l ege n d r e

Al-Jahshiyārī is the only dissenting voice, dating the Iraqi translation to 78/697–98. However, as seen above, this date is a deliberate addition that reflects the overall purpose of his work more than any historical reality.115 As for Syria, I showed above that the Christian narratives convey a certain perception of al-Walīd’s caliphate, especially with reference to Damascus, and they do not teach us much about the translation of the dīwān. The dates they provide appear particularly unreliable.116 Finally, al-Kindī’s Egyptian narrative connects the translation of the dīwān of Egypt to the accession of al-Walīd in 87/705,117 and al-Jahshiyārī’s report on the translation in Khurasan dates it to 124/741–42. The justification given for the translation of the dīwān can appear anecdotal: a competition between Iraqi secretaries, the firing of a negligent Sarjūn b. Manṣūr, or a secretary urinating in an inkwell. The latter is characterized by Noth and Conrad as ‘a pseudo-cause [that] replaces the genuine reasons behind a crucial historical development’.118 In the case of Iraq, the most common tellings, which are on the authority of al-Madāʾinī, claim that the turning point for the completion of the translation was the death of Zādhān Farrūkh. Al-Madāʾinī’s report is used in some form in nine different texts, though only seven of them include the passage on Zādhān’s death. In the three tellings transmitted on the authority of al-Qaḥdhamī, Zādhān appears to be still alive during the translation, though the account is much less detailed. As for Syria, Ibn ʿAsākir is the only source to agree with Khalīfa b. Khayyāṭ that Sarjūn b. Manṣūr was replaced only because he died.119 As seen above, Ibn ʿAsākir adds to Ibn Khayyāṭ report a narrative on the authority Sulaymān b. Saʿd, which stands out for its congruence with known features of the early second/eighth-century Umayyad context. If we accept that Zādhān Farrūkh and Sarjūn b. Manṣūr were replaced only because they passed away, it is even less likely that the translation was undertaken as a centralized enterprise prompted by a caliphal decree. Zādhān and Sarjūn were key personalities in the post-conquest provincial administration of Iraq and Syria, who had held their position since the time of the conquest or early Umayyad times. Indeed, both of them must have been quite old at the turn of the second/eighth century: Sprengling posits that Zādhān Farrūkh was around seventy years old by then and that Sarjūn b. Manṣūr had spent around fifty years in the Umayyad administration of Syria.120

from memory, omitting only one person. Al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 99; Sprengling, ‘From Persian to Arabic’, p. 187. 115 See above, n. 44. Sprengling follows al-Jahshiyārī’s dates in ‘From Persian to Arabic’, p. 211. 116 See above, pp. 99–104. 117 Al-Kindī, Kitāb al-Wulāt, pp. 58–59. 118 Noth and Conrad, The Early Arabic Historical Tradition, p. 189. Consequently, they doubt that the caliph ʿAbd al-Malik was ever involved in the process. 119 Ibn ʿAsākir, Taʾrīkh madīnat Dimashq, xxii, p. 321; Yarbrough, Friends of the Emir, pp. 73–75. As shown above, Ibn ʿAsākir and Khalīfa b. Khayyāṭ are followed in this respect by al-Ṣafadī, Kitāb al-Wāfī bi-l-wafayāt, xv, p. 239. 120 Sprengling, ‘From Persian to Arabic’, p. 189; for Sarjūn’s career see above, at n. 64.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

In an attempt to clear out the death date or dismissal date of Sarjūn b. Manṣūr, other reports prompt us to look conjointly at his fate and that of Athanasius bar Gūmōyē, though this actually complicates the picture. The Chronicle of 1234, transmitting an episode on the authority of the Syrian orthodox patriarch of Antioch Dionysius of Tell-Maḥrē (d. 845), indicates that when Athanasius was revoked from the Egyptian administration in 705, he travelled to Damascus. When he arrived, Sarjūn is said to have drawn attention to Athanasius’s wealth and to have accused him in front of the caliph ʿAbd al-Malik of appropriating the resources of Egypt for himself.121 It is not reported what position Sarjūn held at that time, and he is not mentioned in the Chronicle’s subsequent account of the translation of the dīwān, for which the Chronicle gives the unlikely date of 707–08.122 The aim of the Chronicle is not to present a detailed account of Umayyad administration but to make a point about the antagonistic relationship between the ‘Melkite’ Manṣūrs and the Syriac Orthodox bar Gūmōyēs. In the Chronicle of 1234, Athanasius’s role in Egypt is also praised because he was the ancestor of Dionysius of Tell-Maḥrē.123 For the same reason, the Chronicle (or Dionysius) is mistaken elsewhere about appointments in the Marwanid administration: it claims that ‘ʿAbd al-Malik was informed of his [Athanasius’s] reputation as an intelligent man, well trained in the scribal skills, and he summoned him to Damascus, where he made him the guardian of his younger brother ʿAbd al-ʿAziz, whom he had made the emir of Egypt while still a child’.124 The point of this report is to acclaim Athanasius’s abilities and the central position he occupied in Islamic administration. It is otherwise well known that ʿAbd al-ʿAzīz was appointed governor by his father, Marwān b. al-Ḥakam, not by his brother ʿAbd al-Malik, and that he was not a child when he became governor of Egypt.125 Consequently, it is plausible that the Chronicle is also mistaken in its recounting of the events following the death of ʿAbd al-ʿAzīz and that it was not Sarjūn who denounced Athanasius to the caliph but, possibly, his son Manṣūr b. Sarjūn, who is also known to have worked in the administration of Syria. As Anthony has shown, there is considerable confusion between the two across the sources and in the modern historio­graphy.126 According to most Islamic narratives, Sarjūn was no longer in office in 705, as he had either died or had been dismissed. However, the chrono­logy of Athanasius’s dismissal from Egypt in the Chronicle of 1234 is in agreement with that given in the available recensions of the History of the Patriarchs of Alexandria, primitive and late. Both hold that Athanasius was dismissed immediately after ʿAbd al-ʿAzīz’s death. He then travelled to Damascus, where he was arrested by ʿAbd al-Malik and dispossessed of his wealth, though Sarjūn does not appear 121 Debié, ‘Christians in the Service of the Caliph’, p. 63; Chronicle of 1234, i, p. 295. On Dionysius of Tell-Maḥrē’s work in the Chronicle of 1234, see Hilkens, The Anonymous Syriac Chronicle of 1234 and its Sources, pp. 267–79; see also below, pp. 136–39, for further discussion of this passage. 122 See above, pp. 99–104. 123 Debié, ‘Christians in the Service of the Caliph’, p. 55. 124 Translated in Wood, The Imam of the Christians, pp. 53–54. 125 Mabra, Princely Authority in the Early Marwānid State, p. 103. 126 Anthony, ‘Fixing John Damascene’s Bio­graphy’, pp. 618–22.

111

112

m a r i e l ege n d r e

in this text.127 The History of the Patriarchs makes no mention of the translation of the dīwān, while the text otherwise spares no words on the injustice of Marwanid administrative and fiscal policies.128 Another text, al-Jahshiyārī’s Kitāb al-Wuzarāʾ wa-l-kuttāb, agrees that Athanasius was dismissed while ʿAbd al-Malik was still alive. Athanasius appears there as Yanās b. Khumāyā from al-Ruhā (Edessa), and he is mentioned in only one khabar about the events following ʿAbd al-ʿAzīz’s death. According to the account, when ʿAbd al-Malik received news of the death, he sent a certain al-Ḍaḥḥāk b.  ʿAbd al-Raḥmān to Egypt to divide the wealth of Athanasius.129 From this wealth, jewels were brought before ʿAbd al-Malik for distribution, and he gave some of them back to Athanasius. This report indicates that the Edessan administrator was in Syria with the caliph at the time, but nothing in the text points explicitly to his dismissal. It is interesting to read this passage in conjunction with the Chronicle of 1234, as the comparison reveals that reports about ʿAbd al-Malik’s division of Athanasius’ wealth were cast in different ways by Dionysius of Tell-Maḥrē and by al-Jahshiyārī: for Dionysius, the division was undertaken at Sarjūn’s instigation and was to Athanasius’s disadvantage, whereas for al-Jahshiyārī it resulted in some benefits for Athanasius. In sum, the Chronicle of 1234 on the authority of Dionysius of Tell-Maḥrē, the History of the Patriarchs, and al-Jahshiyārī all agree that Athanasius was brought to ʿAbd al-Malik in Syria after ʿAbd al-ʿAzīz’s death, and the first two sources explain that the reason was that he had been dismissed from his post in Egypt. By contrast, most sources that report on the translation of the dīwān of Egypt (Ibn ʿAbd al-Ḥakam, al-Kindī, and al-Maqrīzī) claim that Athanasius was still in Egypt when the translation was undertaken on the orders of the new governor ʿAbd Allāh b. ʿAbd al-Malik. According to al-Kindī, this took place only after ʿAbd Allāh had been confirmed as governor by his brother, the new caliph al-Walīd. It is not surprising that sources do not agree on the exact timing of Athanasius’s dismissal — right after the death of ʿAbd al-ʿAzīz or right after that of ʿAbd al-Malik. The Marwanid brothers died only five months apart, the former in Jumādā i 86/ May 705 and the latter in Shawwāl 86/October 705.130 Still, being able to date the translation of the dīwān of Egypt with more certainty would be crucial. If,

127 Alexandrinische Patriarchengeschichte, p. 135; History of the Patriarchs, p. 54. My thanks to Perinne Pilette, who shared with me details of her forthcoming re-edition of the History of the Patriarchs, including several new manu­scripts, and confirmed that the translation of the dīwān is absent from all versions. 128 For instance, on the replacement of Christian dukes by Muslim officials, see Alexandrinische Patriarchengeschichte, p. 134; History of the Patriarchs, p. 52. In the History of the Patriarchs, p. 67, the ṣāḥib al-kharāj Usāma b. Zayd al-Tanūkhī (in office 96–99/714–17 and 102–04/720–23) is said to have asked for an assessment of all the districts to be written in Arabic upon his arrival in Egypt. 129 Al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, pp. 34–35. Al-Ḍaḥḥāk b. ʿAbd al-Raḥmān b ʿAzrab/ʿAzram al-Ashʿarī also completed a taʿdīl for ʿAbd al-Malik in the Jazīra (see below, pp. 144–46) and became governor of Damascus under ʿUmar b. ʿAbd al-ʿAzīz: al-Qāḍī, ‘Population Census and Land Surveys under the Umayyads’, pp. 367–68. 130 For ʿAbd al-ʿAzīz, see al-Kindī, Kitāb al-Wulāt, p. 43; for ʿAbd al-Malik, see Khalīfa b. Khayyāṭ, Taʾrīkh, p. 293.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

as claimed by al-Kindī, ʿAbd Allāh b.  ʿAbd al-Malik ordered the translation only after his father had died, it would confirm that ʿAbd al-Malik was not so keen on a centralized Arabic administration. This would also cast doubt on the idea that ʿAbd al-ʿAzīz was blocking the ‘reform programme’ of his brother.131 To come back to the dating of the translation of the other dīwāns, narratives about the dismissal of Athanasius do not allow to shed any light on Sarjūn’s fate. It remains uncertain whether the translation of the dīwān of Syria was prompted by Sarjūn’s death, as most Islamic narratives state that he was still alive when the dīwān was translated. However, this uncertainty does not prevent us from drawing some interesting conclusions. There is no obvious reason to dismiss al-Madāʾinī’s dating of the translation of the Syrian dīwān in 81/700–01 and that of the Iraqi dīwān to around 82/701–02, following Zādhān Farrūkh’s death. There is less variation in the dating of the translation in Egypt (87/705), and we lack comparative material to evaluate the date given by al-Jahshiyārī for Khurasan (124/741–42) beyond noting that aligns with other narratives about the governorates of Yūsuf b. ʿUmar and Naṣr b. Sayyār. If we accept this chrono­logy and the narratives attached to each date, the translation of the dīwān would have been caused by the dismissal or death of Sarjūn in Syria, by the death of Zādhān Farrūkh in Iraq, by the appointment of a new governor in Egypt, and by the backlash against Khalīd al-Qasrī’s personnel choices in Iraq and Khurasan after his dismissal. The dynamics of change do not need to be the same from province to province; indeed, it would be suspicious if they were. The idea of organic change is problematic only if we try to superimpose a preconceived idea of the Umayyad Empire as a highly centralized polity on the evidence.132 In fact, the narrative sketched here fits the known dynamics of the Umayyad state as a polycentric space in which each province introduced innovations differently.133 Those polycentric dynamics are evident when we examine, for instance, innovations on the coinage and the various attempts at creating an Islamic imagery before the introduction of the epi­graphic model by ʿAbd al-Malik.134 The key difference between the coinage reform and the translation of the dīwān is that in the case of the former there was eventually an overarching top-down reform that changed the visual of the gold and silver coinage in the provinces of the empire. This is a reform properly defined that aimed at introducting a uniform coinage for the empire. No such thing ever happened for the translation of the dīwān as there was no attempt to introduce a uniform dīwān. Each dīwāns was translated

131 As argued in Mabra, Princely Authority in the Early Marwānid State, pp. 83–118. 132 See, for instance, Bouderbala and Fenina, ‘Fiscalité et pratiques monétaires dans l’Égypte préfatimide’, p. 396 n. 15. 133 ʿAbd al-ʿAzīz Dūrī and Koseï Morimoto accept the uneven chrono­logy of the ‘arabization policies of the Caliph ʿAbd al-Malik’. See Dūrī, ‘Dīwān’, p. 324; Morimoto, ‘The Dīwāns as Registers of the Arab Stipendaries in Early Islamic Egypt’, p. 364. 134 Generally called ‘transitional coinage’ and dated 72–77/691–97; for Iraq, see Treadwell, ‘The “Orans” Drachms of Bishr ibn Marwān and the Figural Coinage of the Early Marwanid Period’, pp. 223–69; for Egypt, see Foss, Arab Byzantine Coins, pp. 100–05; for the Umayyad North, see Ilisch, ‘Muhammad Drachms and their Relation to Umayyad Syria and Northern Mesopotamia’, pp. 17–24.

113

114

m a r i e l ege n d r e

at different times and retained local features as visible in the shape of numerical notations (see below). This argument of decentralized dynamics for administrative innovations do not imply that the provinces were disconnected. The translations of the dīwāns of Syria, Iraq, and Egypt do appear to have taken place within half a decade of each other (c. 700–05). If we accept the dating offered by al-Madāʾinī/ al-Balādhurī, the translation of the dīwān at the caliphal centre in Syria took place first. Knowledge about that event certainly travelled, but the translation of each of the local dīwāns could take place only because local power dynamics between old and new administrative elites shifted and allowed the change to happen. Now that I have set the scene with the actors involved in the translation of the dīwān and the chrono­logy of this change, I will attempt to define what, exactly, was translated.

The Intricate Meaning of Dīwān Studying the genesis of the dīwāns of the various regions of the early Islamic empire would constitute substantial research in its own right.135 Only preliminary remarks are drawn here on the basis of the sources examined in this article because of their focus on the translation. Among them, three sources provide a definition of the dīwān that was translated, al-Jahshiyārī being the earliest to do so:136 ‫ وهذا الذي كان عمر‬137‫ولم يزل بالكوفة والبصرة ديوانان أحدهما بالعربيّة إلحصاء الناس وأعطياتهم‬ ‫قد رسمه واآلخر لوجوه األموال بالفارسيّة وكان بالشام مثل ذلك أحدهما بالروميّة واآلخر بالعربيّة‬ ‫فجرى األمر على ذلك إلى أيام عبد الملك بن مروان‬

135 Dīwān is a very versatile term in Arabic. Given its general meaning as a collection of texts, it can be applied to many types of written corpora, including adab, poetry, and administrative registers, and by extension to administrative offices. See Dūrī, ‘Dīwān’, p. 323. Past attempts to study administrative dīwāns include Puin, Der Dīwān von ʿUmar ibn al-Ḫaṭṭāb, a work of encyclopedic nature in which the author makes a rather unconvincing case that the dīwān has an Arab and Arabian origin going back to the time of the prophet Muḥammad. This was a reaction to Sprengling’s argument for a Persian origin of the dīwān in ‘From Persian to Arabic’, pp. 177–81. Ever since, Dūrī’s work has been the main contribution to the field, though it is quite positivist in its use of sources: Dūrī, Early Islamic Institutions, pp. 161–70. Finally, Kyle Longworth’s recent PhD dissertation, ‘Islamic Bureaucrats in Late Antiquity’, has some excellent passages on the early Islamic dīwāns, especially in the Umayyad period; see, for example, pp. 23–24: ‘Our sources typically use the term to refer to a set of responsibilities that transcended a particular administration or individual and credit the caliph under whom this transition for ad hoc to institution took place. In this way, the development of the dīwān from informal registers to developed bureaus mirrors many of the early stages of Umayyad state-building and the caliphate recognizing itself as a state composed of designated administrative structures with responsibilities that transcended the appointment of a particular individual or solving a specific issue’. 136 Miskawayh slightly alters al-Jahshiyārī’s description of the dīwān when he copies it: Miskawayh, Tajārib al-umam, ii, p. 387. 137 Miskawayh has ‫( إلحصاء الناس وأرزاقهم وأعطياته‬listing the beneficiaries, their sustenance [in kind] and their [cash] stipends).

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

There were always two dīwāns in Kufa and Basra: one of them was in Arabic, listing the beneficiaries138 and their stipends, and this is the one that was designed by ʿUmar. The other one was for the purpose of revenues, and it was in Persian. And it was similar in Syria: one of them was in Greek139 and the other in Arabic. It remained the same until the time of ʿAbd al-Malik b. Marwān.140 Al-Jahshiyārī then goes on to describe the translation of the Persian dīwān by Ṣāliḥ b. ʿAbd al-Raḥmān. Al-Ṣūlī has a similar account, though he uses different termino­logy: ‫ وديوان بالفارسية‬،‫كان بالبصرة والكوفة ديوانان إلعطاء الجند والمقاتلة والذرية بكتاب بالعربية‬ ‫وبالشام ديوان بالعربية لمثل ذلك وديوان بالرومية‬

There were two dīwāns in Basra and Kufa, [one] for the stipends of the jund, the fighters and the families,141 written in Arabic and a dīwān in Persian. In Syria, there was a similar dīwān in Arabic and a dīwān in Greek.142 As for Maqrīzī is his Khiṭaṭ:

‫ذكر ديوان الخراج واألموال‬

‫دون هذا الديوان في اإلسالم بدمشق والعراق على ما كان‬ ّ ‫وأول ما‬ ّ ،‫يقال لكتابة الخراج قلم التصريف‬ ،‫ وديوان مصر بالقبطية‬،‫ وديوان العراق بالفارسية‬،‫ وكان ديوان الشام بالرومية‬،‫عليه قبل اإلسالم‬ ‫فنقلت دواوين هذه األمصار إلى العربية‬

Reports on the dīwān for taxation and revenues The keeping of fiscal records is called qalam al-taṣrīf. In Islamic times, this dīwān was first kept in Damascus and Iraq in the way it had been done before. The dīwān of al-Shām was in Greek, the dīwān of Iraq in Persian, and the dīwān of Egypt in Coptic. The dīwāns of these amṣār were then translated into Arabic.143 The first two definitions agree that the early Islamic state would have had two dīwāns, one for the collection of fiscal revenues (dīwān al-kharāj) and one for the recipients of revenues (dīwān al-jund). Even when the term is used in the singular, it is often meant that each dīwān was comprised of several registers: possibly one per tribe for the dīwān al-jund as suggested by Koseï Morimoto,144 while the fiscal revenues 138 Lit. ‘for the census of the people’; however, al-nās here refers not to the population of the provinces as a whole but only to those who were part of the jund and received stipends (ʿaṭāʾ). 139 Literally ‘in the language of the Romans’. 140 Al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 38. 141 In the first/seventh century, the common term for dependents seems to have been simply ʿiyāl. See Sijpesteijn, ‘Army Economics’, pp. 250, 259; Tillier and Vanthieghem, ‘Recording Debts in Sufyānid Fusṭāṭ’, p. 148. 142 Al-Ṣūlī, Adab al-kuttāb, p. 192. 143 Al-Maqrīzī, al-Khiṭaṭ, ed. by Sayyid, i, p. 264. My translation differs slightly from Karl Stowasser’s, p. 394. 144 Morimoto, ‘The Dīwāns as Registers of the Arab Stipendaries in Early Islamic Egypt’, p. 360.

115

116

m a r i e l ege n d r e

could have been listed according to types of taxes or the various territorial divisions of the province.145 In al-Jahshiyārī’s words, the registers that were translated were ‘for various kinds of revenues’, while al-Ṣūlī does not describe the dīwān’s contents, and, for al-Maqrīzī, it is the ‘dīwān for taxation and revenues’. On the other hand, the registers of beneficiaries or dīwān al-jund, as a foundation of the paradigmatic caliph ʿUmar b. al-Khaṭṭāb (r. 13–23/634–44), could only be in Arabic. In their analyses of the different translation narratives, most scholars have agreed with this identification of the translated dīwān as fiscal registers. For Sprengling, they were ‘the official set of books, the budget and balance sheets’.146 Ulrich Rebstock describes them as ‘tax lists’.147 Al-Qāḍī focuses on information about registers of landed estates in al-Balādhurī’s section on Basra — registers that were essential for the assessment of the main money tax of the period, the land tax — and she explains that dīwān means ‘financial registers’.148 Saliba understands Ibn al-Nadīm’s and al-Jahshiyārī’s use of dīwān as denoting the administrative ‘accounting procedures’. As Ibn al-Nadīm puts it in the mouth of Ṣāliḥ b. ʿAbd al-Raḥmān, what needs to be done for al-Hajjāj is to convert the accounts into Arabic (‫)أحول الحساب إلى العربية‬.149 Saliba also rejects the idea that those accounts were part of the dīwān al-jund.150 However, all the source material referring to the dīwān al-kharāj/dīwān al-jund binary and to the alleged foundation of the latter by the second caliph ʿUmar were composed after the rise of the Abbasids. This termino­logy also appears to have been articulated rather late. The term dīwān was possibly part of the late Sasanian administrative vocabulary as it is used in some seventh-century Middle Persian and Armenian literature.151 However, the term does not appear on any documentary

145 Details of the organization of local Greek registers are known from Aphrodito, where ‘stages within the assessment procedure are evidenced by two sorts of lists called diastalmos “separation”, probably compiled at, or at least in collaboration with, the taxpayer communities and then forwarded to the pagarch, and merismos, “apportionment”, obviously drafted within the governor’s office. […] Also the majority of the Greek merismos accounts were codices, usually made up of quires of one double leaf (four pages). All the entries consisting of the names of taxpayers and the figures of their shares of certain duties (mostly poll tax, land tax, chrysika dēmosia, and embolē) are written in the minusculelike stylized Greek hand, with a few but remarkable Copticisms’. Richter, ‘Language Choice in the Qurra Dossier’, pp. 202, 204. 146 Sprengling, ‘From Persian to Arabic’, p. 196. 147 Rebstock, ‘Observations on the Diwan al-Kharaj and the Assessment of Taxes in Umayyad Syria’, p. 232. 148 Al-Qāḍī, ‘The Names of Estates in State Registers’, p. 255. 149 Ibn al-Nadīm, Kitāb al-Fihrist, p. 303. See below, section 4, on the use of particular verbs to refer to the translation. 150 Saliba, Islamic Science and the Making of the European Renaissance, pp. 53–54. See also Khalek, ‘Some Notes on the Representation of Non-Muslim Officials in al-Ǧahšiyārī’s Kitāb al-Wuzarāʾ wa-l-Kuttāb’, p. 506: ‘the translation of the dīwān/fiscal register from Persian and Greek into Arabic (in formerly Sasanian and Byzantine territories)’. 151 In Middle Persian, see Mādayān ī hazār dādestān (Book of a thousand judgements): Perikhanian, The Book of a Thousand Judgements, p. 354. In Armenian, see The Armenian History Attributed to Sebeos, trans. by Thomson, Howard-Johnston, and Greenwood, p. 6 n. 35, p. 49 n. 309. The term dīwān also appears on a Middle Persian seal of uncertain dating, which could well be a product of the early Islamic period: Gyselen, ‘Diwan et “trésorerie” sassanides’, p. 125, n. 12.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

sources preserved from the early Islamic administration before the beginning of the Abbasid period. The earliest mention of this term in an Arabic papyrus dates to 140/760–61.152 The same can be said of the term kharāj.153 Of the terms mentioned above, only jund, as well as ʿaṭāʾ (stipend) or ʿaṭāʾ al-jund, are attested in documentation dated to the Umayyad period,154 though Naïm Vanthieghem and Lev Weitz have argued that in the Sufyanid period ʿaṭāʾ was a ‘generic term for the tax being requisitioned’.155 In the following pages, when the dīwān al-kharāj and dīwān al-jund are mentioned it is because that is the termino­logy of the Arabic sources written in the Abbasid era and later. This does not mean that there were no fiscal or military registers in the preceding period, but we are unsure what they were called in Arabic. Even if the termino­logy of the sources is a product of their time, it is clear in all of the examined textual traditions that what was translated were registers or accounts (dīwān or ḥisāb in Arabic). Theophanes writes about ‘public registers of the state offices’ (τοὺς δημοσίους τῶν λογοθεσίων κώδικας),156 and Michael the Syrian has ‘scribes of the public accounts’ (kāṯōḇē d-hušbānē d-DYWMWSYWN (δημόσιον))157 and ‘registers’ (DPTRʾ (διφθέραι)).158 As for the Chronicle of 1234, al-Walīd is said to have changed the language of the ‘record office’ (bēṯ qarṭīsē),159 that is, the ‘public chancery (DYWMWSYWN (δημόσιον)), which the Arabs call dīwān’.160 While we accept that registers were translated, can we agree that there was a clear organization with a military and a fiscal dīwān before the time of the translation? Though he argues that the caliph ʿUmar b. al-Khaṭṭāb did create a dīwān, GerdRüdiger Puin contends that it was focused on the distribution of revenues to all beneficiaries, not only those connected to the military, and that the distinction

152 P.Berl.Arab. I 2, 4 (Fayyum, 140/760–61), mentioning the dīwān of Lower Egypt (dīwān asfal al-arḍ), and P.Ryl.Arab. I 5, 21 (181/797), republished in Younes, ‘New Governors Identified in Arabic Papyri’, pp. 15–20. 153 Legendre, ‘Caliphal Estates and State Policy over Landholding’, pp. 403–07. 154 For example, P.Heid.Arab. I 1, 5–10 (Aphrodito, 91/701): ‫فإنَّه قد ذهب من الزمن ما قد علمت وقد استأخرت‬ ‫للا‬ ّ ‫الجزية وحضرعطاء الجند وعطاء عيالهم وخروج الجيوش إن شاء‬. On ʿaṭāʾ, see Sijpesteijn, ‘Army Economics’, pp. 245–68. 155 Vanthieghem and Weitz, ‘A Companion of Muḥammad in the Oldest Egyptian Bilingual Entagion’, p. 213. 156 Theophanes, Chrono­graphia, p. 376; trans. Theophanes, Chronicle, p. 524. 157 Sokoloff, A Syriac Lexicon, pp. 307b–308a. 158 For διφθέραι in Syriac, see Sokoloff, A Syriac Lexicon, p. 316b. The passage is translated in Michael the Syrian, Chronique, p. 481. Daftar is rarely used in descriptions of early Islamic administrative practices; we find only daftar al-jund in Ibn Aʿtham, Kitāb al-Futūḥ, vi, p. 46; Kennedy, ‘The Financing of the Military in the Early Islamic State’, p. 375. 159 Bayt al-qarāṭīs does appear in Arabic texts, though it is not very common. For instance, in al-Ṭabarī’s account of the revolt of ʿAmr b. Saʿīd b. al-ʿĀṣ al-Ashdaq in Damascus, al-Walīd b. ʿAbd al-Malik is brought to the bayt al-qarāṭīs after being struck on the head and a certain Ibrāhīm b. ʿArabī is said to have been ṣāḥib al-dīwān: al-Ṭabarī, Taʾrīkh al-rusul wa-l-mulūk, vi, p. 144. However, bayt al-qarāṭīs does not appear in any of the Arabic translation narratives. 160 Chronicle of 1234, i, pp. 298–99. The varying termino­logy also shows that the authors of the so-called Theophilus circuit were all working with different materials.

117

118

m a r i e l ege n d r e

between a fiscal and a military dīwān appeared only later, though he does not say when.161 Moshe Sharon is one of the few scholars to have argued that narratives of the foundation of the dīwān al-jund by ʿUmar b. al-Khaṭṭāb were fabricated as early Islamic chroniclers attempted to present a coherent picture of the rise of Islam and to ascribe innovations to the Prophet or to the first caliphs. According to Sharon, ʿaṭāʾ was indeed paid to the tribes that participated in the early Islamic conquest, but the process was ‘far from an orderly affair’. However, it is not clear in his analysis when this disorderly affair eventually led to an orderly dīwān, though he seems to accept that one existed in some form under the Umayyads, having developed out of the initial ad hoc and tribally organized payments.162 There is, moreover, a striking contradiction between the consensus in the Islamic sources on the foundation of the dīwān al-jund by ʿUmar and the absence of any clear account of how the dīwān al-jund or its counterpart, the dīwān al-kharāj, would have worked in the decades following ʿUmar’s death. This discrepancy can be illustrated with the information available on the dīwān of Egypt. Only a few sources refer to the early Islamic administration of this province: two sources are preserved from pre-Fatimid Egypt: Ibn ʿAbd al-Ḥakam’s Futūḥ, al-Kindī’s Kitāb al-Wulāt wa-al-Qudāt and fragments of two Taʾrīkhs have been collected, at least partially, those of Ibn Yūnus (d. 347/958) and, most recently, of al-Layth b. Saʿd (d. 175/791).163 As he investigated the first three, Bouderbala managed to track only seven aṣḥāb dīwān al-jund or kuttāb dīwān al-jund between 43/663 and 150/767, a list from which he excludes Wardān and Athanasius bar Gūmōyē.164 This strikes as a poorly documented or poorly established office, especially since, in comparison, Bouderbala counted forty-four aṣḥāb al-shurṭa for the same period.165

161 Puin, Der Dīwān von ʿUmar ibn al-Ḫaṭṭāb, pp. 95–100. His focus is on sābiqa as an organizing principle of the dīwān of ʿUmar. 162 Sharon, ‘The Military Reforms of Abū Muslim’, pp. 115, 118–23. Hawting also acknowledges the obscure origin of the dīwān and posits that under the Umayyads the word meant ‘government department’. In his view, ‘the main business of the administration remained the assessment and collection of taxes, and in a general sense the diwan refers to the administration’; Hawting, The First Dynasty of Islam, p. 63. This understanding is closer to al-Māwardī’s definition (al-Aḥkām al-sulṭāniyya, p. 297 ): ‫ ومن يقوم بها من الجيوش والع َّمال‬،‫والديوان موضع لحفظ ما يتعلَّق بحقوق السلطنة من األعمال واألموال‬. I leave aside what he and others say about other governmental departments such as the dīwān al-rasāʾil and the dīwān al-khātam, as this lies beyond the scope of the present research. See Dūrī, Early Islamic Institutions, pp. 168–70; Longworth, ‘Islamic Bureaucrats in Late Antiquity’, ch. 1. 163 Zychowicz-Coghill, The First Arabic Annals. On the problematic issue of the reconstruction of lost sources on the basis of quotations, see pp. 5–7. 164 Among sources referring to the translation, only al-Mubarrad (d. 285/898) mentions the title ṣāḥib dawāwīn al-ʿIrāq (‫)كان صالح بن عبد الرحمن كاتب الحجاج وصاحب دواوين العراق‬: al-Mubarrad, al-Kāmil fī al-lugha, ii, p. 145. 165 Bouderbala, ‘Ǧund Miṣr’, pp. 319–26. The same is evident in al-Jahshiyārī’s work, according to Longworth: ‘Al-Jahshiyārī seems to reflect this idea that the dīwān al-jund was not necessarily a separate bureau. For the Umayyad period, an individual in charge of the dīwān al-jund is only referenced twice in al-Jahshiyārī’s Kitāb al-wuzarāʾ: first for ʿAmr b. Saʿd al-ʿĀṣ al-Ashdāq who served as the secretary of the dīwān al-jund for Muʿāwiya and second for ʿAbd al-Malik b. Muhammad

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

Officials in charge of the dīwān of Syria are better documented. Two of the sources discussed in the present article, Khalīfa b. Khayyāṭ’s Taʾrīkh and Ibn ʿAbd Rabbihi’s al-ʿIqd al-farīd, allow us to track the titles ascribed to administrators of the dīwān of Syria throughout their record of the Umayyad period, as they record this information systematically.166 Both authors worked with a template: Ibn Khayyāṭ provides the names of high administrators under the year of each caliph’s death, while Ibn ʿAbd Rabbihi lists them at the beginning of his account of each caliphate. Ibn Khayyāṭ’s work is one of the earliest preserved sources to record the names of officials and offices systematically in the form of lists.167 For the Umayyad period, he is most consistent about the caliphal administration in Syria.168 Under Muʿāwiya b. Abī Sufyān, Ibn Khayyāṭ says simply that Sarjūn b. Manṣūr was in charge of the dīwān and his (Muʿāwiya’s) affairs/government (‫وعلى الديوان وأمره كله ‌سرجون‬ ‫)‌بن‌منصور الرومي‬. Under Marwān b. al-Ḥakam, he describes Sarjūn as the caliph’s secretary (‫)كاتبه‬. It is only under ʿAbd al-Malik b. Marwān that he attributes to Sarjūn responsibility for fiscal and military administration (‫)الخراج‌والجند‬.169 After ʿAbd al-Malik, the termino­logy remains remarkably stable in Ibn Khayyāṭ’s text. Under all the subsequent Umayyad caliphs, he lists the administrators in charge of al-kharāj wa-l-jund, with a few variants under Yazīd b.  ʿAbd al-Malik (‘fiscal and military administration, and correspondence’, ‌‫)الخراج‌والجند والرسائل‬, Yazīd b. al-Walīd (‘fiscal and military administration and the [bureau of the] seal’, ‌‌‌‫)الخراج‌والجند والخاتم‬, and Marwān b. Muḥammad (‘military and fiscal administration, treasuries, and storehouses’, ‫)ديوان ‌الجند ‌والخراج وبيوت األموال والخزائن‬.170 As seen above, Ibn Khayyāṭ and Ibn ʿAbd Rabbihi were not working with the same materials when writing about the translation of the dīwāns of Syria and Iraq. However, they agree on when the division between kharāj and jund was introduced in the titles of officials in charge of the dīwān al-Shām. Ibn ʿAbd Rabbihi also transmits lists of officials but interestingly does not mention Sarjūn under Muʿāwiya, instead naming Saʿīd b. Anas al-Ghassānī as the latter’s secretary.171 He names Sarjūn for

166

167 168 169 170 171

b. al-Ḥajjāj b. Yūsuf during the caliphate of Walīd II’. Longworth, ‘Islamic Bureaucrats in Late Antiquity’, pp. 33–34; al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, pp. 60, 118. Other lists at our disposal are not necessarily focused on administrators. They do not provide systematic information on titles, making it difficult to identify trends across their record of the Umayyad period. For instance, al-Baghdādī includes Sarjūn b. Manṣūr in a list of secretaries, but he gives only two names in that section, Sarjūn and Ḥassān al-Nabaṭī, secretary to al-Ḥajjāj. According to the text, Sarjūn was in charge of al-kharāj wa-l-ḍiyāʿ (taxation and landed domains), which is a way of saying that he was not in charge of the jund. Sourdel, ‘Le “livre des secrétaires” de ʿAbdallāh al-Baġdādī’, p. 139. According to Tobias Andersson, his lists are more detailed and systematic than those provided by al-Ṭabarī and al-Yaʿqubī: Andersson, Early Sunnī Historio­graphy, pp. 187–93. In Iraq, Zādhān Farrūkh is listed as the kātib al-kharāj for both Ziyād b. Abīhi and al-Ḥajjāj: Khalīfa b. Khayyāṭ, Taʾrīkh, pp. 197, 313. Khalīfa b. Khayyāṭ, Taʾrīkh, pp. 218, 259, 302. Khalīfa b. Khayyāṭ, Taʾrīkh, pp. 317 (al-Walīd b. ʿAbd al-Malik), 325 (Sulaymān b. ʿAbd al-Malik), 331 (ʿUmar b. ʿAbd al-ʿAzīz), 343 (Yazīd b. ʿAbd al-Malik), 362 (Hishām b. ʿAbd al-Malik), 267 (al-Walīd b. Yazīd), 371 (Yazīd b. al-Walīd), 408 (Marwān b. Muḥammad b. Marwān). Most probably this is a reference to Saʿīd b. ʿAws al-Ghassānī, better known as ʿUbayd or ʿUbayd

119

1 20

m a r i e l ege n d r e

the first time as the secretary of Yazīd b. Muʿāwiya (‫وكاتب يزيد بن معاوية ‌سرجون ‌بن‬ ‫)‌منصور‬172 or the secretary in charge of all his affairs/government (‫وكاتبه وصاحب‬ ‫)أمره ‌سرجون ‌بن ‌منصور‬.173 Under Marwān b. al-Ḥakam, Sarjūn is again presented as a secretary (‫)وكاتبه سرجون بن منصور الرومي‬,174 and it is only under ʿAbd al-Malik that Ibn ʿAbd Rabbihi describes him as responsible for the kharāj and jund (‫وكاتبه‬ ‫)على الخراج والجند ‌سرجون ‌بن ‌منصور الرومي‬.175 Similarly to Ibn Khayyāṭ’s work, the title attached to administrators of the dīwān remains stable in Ibn ʿAbd Rabbihi’s accounts of the subsequent Umayyad caliphs, with a few variations under Yazīd b. al-Walīd (‘fiscal and military administration, the smaller bureau of the seal, and the ḥaras’, ‫)على الخراج والجند والخاتم الصغير والحرس‬176 and Marwān b. Muḥammad, for whom Ibn ʿAbd Rabbihi lists a secretary (ʿAbd al-Ḥamīd b. Yaḥyā, to whom we will return) and another official in charge of military administration ( ‫)ديوان‌الجند على‬.177 According to both Ibn Khayyāṭ’s and Ibn ʿAbd Rabbihi’s lists of administrators, it was only during the caliphate of ʿAbd al-Malik that the separation of taxation (kharāj) and military administration (jund) was formally introduced for the dīwān in Syria, though both registers were headed at the time by a single official, Sarjūn b. Manṣūr.178 Until that time, both sources describe him simply as the caliph’s secretary (kātib) or as responsible for ‘the dīwān’. Under ʿAbd al-Malik, Sarjūn b. Manṣūr was put in charge of the new dīwān formalising fiscal and military administration and he remained in this position until he died or was replaced by Sulaymān b. Saʿd. This means that the administrative separation between kharāj and jund took place before the translation. The context of the Second Fitna (60–72/680–92) could have prompted this change, especially for the military registers, since the Fitna caused the Umayyads to lose the support of a large part of the empire’s tribal elites. An orderly system of payment of tribes must have been a pressing necessity in order to cement allegiances once unity was restored.179 The implementation of a division between taxation and military administration in the Marwanid period is also visible in fragments attributed to the Taʾrīkh of al-Layth b. Saʿd, which focuses on Egypt. The text employs a notably different termino­logy. The relevant information appears not in the titles of secretaries but in those of governors. On the basis of al-Layth’s Taʾrīkh, Edward Zychowicz-Coghill has shown that in Egypt, Umayyad caliphs appointed a governor over financial affairs alongside a military governor between 96/714 and 124/742. The financial governor is said to have been in charge of the people of the province (‫)على أهل األرض‬. In later Allāh. On this administrator, see Longworth, ‘Islamic Bureaucrats in Late Antiquity’, pp. 53–55. 172 Ibn ʿAbd Rabbihi, al-ʿIqd al-farīd, iv, p. 247. 173 Ibn ʿAbd Rabbihi, al-ʿIqd al-farīd, v, p. 124. 174 Ibn ʿAbd Rabbihi, al-ʿIqd al-farīd, v, p. 147. 175 Ibn ʿAbd Rabbihi, al-ʿIqd al-farīd, v, p. 148. 176 Often translated as ‘bodyguard unit’; see Perlman, ‘The Bodyguard of the Caliphs during the Umayyad and Early Abbasid Periods’. 177 Ibn ʿAbd Rabbihi, al-ʿIqd al-farīd, v, pp. 179 (ʿUmar b. ʿAbd al-ʿAzīz), 188 (Yazīd b. ʿAbd al-Malik), 191 (Hishām b. ʿAbd al-Malik), 208 (Yazīd b. al-Walīd), 212–13 (Marwān b. Muḥammad b. Marwān). 178 Al-Balādhurī agrees: Ansāb al-ashrāf, vii, p. 222. 179 On the Second Fitna, see Borrut, Entre mémoire et pouvoir, pp. 67–71; Robinson, ʿAbd al-Malik, pp. 31–48.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

works such as al-Kindī’s, this position is called ṣāḥib al-kharāj, but this termino­logy is not found in what can be reconstructed of al-Layth’s earlier Taʾrīkh. Al-Layth’s termino­logy (‫ )أهل األرض‬is also used in the letters of the Marwanid governor Qurra b. Sharīk to refer to the taxable population of Egypt.180 Judging from the fragments of the Taʾrīkh, the practice of appointing two governors came to an end in 124/742, when Ḥafṣ b. al-Walīd was named the sole governor over Arabs and non-Arabs alike (‫)وجمع لحفص ‌عربها ‌وعجمها‬.181 As Zychowicz-Coghill argues, ahl al-arḍ (the people of the province) can here be considered synonymous with ʿajamuhā (the non-Arabs of Egypt). In his view, ‘al-Layth’s presentation of these appointments as jurisdictions over different classes of people (over “Arab” versus “non-Arab/people of the land”) is likely to be another archaic usage reflecting late Umayyad-era termino­logy and conceptions’.182 This archaic termino­logy is also found in Ibn ʿAsākir’s second translation narrative about Syria, in which Sarjūn is described as responsible for all the dīwāns of the Arabs and the non-Arabs (‫وسرجون‬ ‫ — )على جماعة دواوين العرب والعجم‬that is, the military registers (those of the Arabs, in accordance with Sharon’s argument of tribally organized payments) and the fiscal registers (those of the ʿajam/non-Arabs).183 Though al-Balādhūrī does not list the administrators of the dīwāns in Iraq, he reports on the authority of al-Qaḥdhamī that the latter’s grandfather Qaḥdhām was placed in charge of the dīwān jund al-ʿarab by the governor Yūsuf b. ʿUmar.184 It is well known that non-Arabs also served in early Islamic armies, but it would not have been reflected in the name given to the military registers that appears to be linked to the dominant element within the said armies.185 Al-Layth’s presentation suggests that the formal division between ʿajam/kharāj and ʿarab/jund in the duties of governors began only in 96/714 in Egypt. This is in accordance with what can be reconstructed on the basis of Egyptian papyri. I have argued elsewhere that the position that sources written in the Abbasid period call the ṣāḥib al-kharāj was created in Egypt at the beginning of the Marwanid period. This official was in charge of fiscal administration in Fusṭāṭ, and he took over the duties of the dukes of the former Byzantine provinces. The office of the duke and the territorial divisions associated with it were supressed at the same time, the last duke being attested between 704–05 and 711.186 What remains unknown is the exact nature and organization of the registers (the ‘dīwān’ that Sarjūn b. Manṣūr is supposed to have headed since the time of 180 See, for example, P.Cair.Arab. III 148 (Aphrodito, 708–10); Sijpesteijn, ‘Loyal and Knowledgeable Supporters’, pp. 344–45. 181 See Zychowicz-Coghill, The First Arabic Annals, pp. 36, 98 (on Usāma b. Zayd: ‫دخل أسامة بن زيد مصر أميرا‬ ‫)على أهل األرض‬, 102 (when Usāma is appointed in Syria: ‫)خرج أسامة بن زيد إلى الشام فجعل على الدواوين‬, 109 (on Ḥafṣ b. al-Walīd: ‫)وأمر حفص بن الوليد على أهل مصر وفيها نزع القاسم بن عبيد هللا من مصر وجمع لحفص‌عربها‌وعجمها‬. 182 Zychowicz-Coghill, The First Arabic Annals, p. 36. 183 Ibn ʿAsākir, Taʾrīkh madīnat Dimashq, xxii, p. 320. 184 Al-Balādhurī, Futūḥ al-buldān, p. 350; Lynch, Arab Conquests and Early Islamic Historio­graphy, p. 79. According to al-Jahshiyārī, Qaḥdham was appointed to head the fiscal administration of Basra by Yūsuf b. ʿUmar: al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, pp. 64–65. 185 Al-Qāḍī, ‘Non-Muslims in the Muslim Conquest Army in Early Islam’, pp. 83–127. 186 Legendre, ‘Neither Byzantine nor Islamic?’.

121

122

m a r i e l ege n d r e

Muʿāwiya b. Abī Sufyān) before the formalization of the division between ʿarab/ jund and ʿajam/kharāj. This formalization seems to have taken place at a different time and level in Egypt (around 96/714 — meaning after the translation, and at the level of the governors) compared to Syria (under ʿAbd al-Malik, before the translation, and at the level of the dīwān), though in Egypt the division could have happened in the dīwān before it became explicit in the titles of governors.187 It is also possible that after the conquests fiscal information was for the most part not centralized in a single provincial register kept in the capital but rather that fiscal registers remained initially in the hands of pre-Islamic provincial administrators, such as the Egyptian dukes. However, the decentralization of fiscal administration did not equate to a lack of control. In a missive found in Edfu and dating from the 660s–680s, the duke of the Thebaid, as the head of fiscal administration in his province, requests that the pagarchs of the province travel to Fusṭāṭ with the fiscal registers of their districts (pagarchy/kūra) so that they can be checked.188 State control is also evident after the translation in early second/eighth-century Aphrodito, though it involves different agents. In Aphrodito, one of the main duties of the local administrators was to draw up registers for the localities under their jurisdiction with lists of taxable entities, such as people, land, trades, and commercial goods. Assessors were in charge of compiling all this information; they did so under oath and risked a fine of 100 solidi if they falsified the numbers or acted unjustly in their evaluation. The assessment of each district was sent directly to Fusṭāṭ without any involvement on the part of the duke, and there the total sum of yearly dues was decided and communicated to the pagarchs.189 As for military administration, the information provided by Khalīfa b. Khayyāṭ and Ibn ʿAbd Rabbihi shows that non-Muslim administrators such as Sarjūn b. Manṣūr were handling accounting procedures for the payment of military salaries immediately before the translation. This does not mean that they were the deciding authority on the amounts that were paid. This is presented to be at the discretion of the governor, for instance when Athanasius bar Gūmōyē is ordered by ʿAbd al-ʿAzīz b. Marwān to increase the ʿaṭāʾ of a certain Yazīd b. ʿUrwa al-Ḥamalī.190 Athanasius was responsible for keeping proper records of payments. Based on this information, can we assume that Athanasius or Sarjūn would have kept records of military payments in Greek? Morimoto and Hugh Kennedy have argued that military registers were indeed initially written in Greek. Morimoto

187 As Chase F. Robinson puts it for the administrative apparatus of the early Islamic Jazīra: ‘The constituent elements ingredient to later, classical views can be identified relatively early on, but the system emerged only secondarily’. Robinson, Empire and Elites after the Muslim Conquest, p. 33. 188 P.Apoll. 6 (Edfu, 660s–680s). 189 Bell, Greek Papyri in the British Museum, p. xxvii. Eduardo Manzano Moreno refers to the Chronicle of Albelda to show that in al-Andalus, into the second half of the third/ninth century, counts (comites) were accountable for the collection of taxes locally, and the bishop of Malaga, Hostegesis, was expected to update the registers of his district. Manzano Moreno, Conquistadores, emires y califas, p. 78. 190 Al-Kindī, Kitāb al-Wulāt, p. 50; Morimoto, ‘The Dīwāns as Registers of the Arab Stipendaries in Early Islamic Egypt’, p. 364.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

identified P.Lond. IV 1447 (Aphrodito, written before 705) as a fragment of the Greek dīwān. However, Bouderbala has convincingly shown that it is an account of requisitions for the sustenance of officials attached to the service (ḥāshiya) of the governor ʿAbd al-ʿAzīz b. Marwān.191 The very fact that the document was found in Aphrodito shows that it is not a fragment of the central dīwān of Fusṭāṭ, which is our focus here.192 Sijpesteijn has also listed possible fragments of early Arabic dīwāns,193 including P.Mil.Arab. 6, a fragmentary list of names in groups dated paleo­graphically to the first/seventh century,194 and two lists of names with places of residence dated to the first/seventh or early second/eighth century.195 However, lists of names could be drawn up for all sorts of purposes, and it remains conjectural to connect them with certainty to the military dīwān. The debate has been picked up most recently by Mathieu Tillier and Naïm Vanthieghem in their investigation of Arabic debt registers dated between 42–57/662–77, which are concerte evidence of bookkeeping in Arabic before the Marwanid period. While they explore the possibility that the registers were part of the dīwān al-jund, they deem it unlikely in favour of an internal tribal record handled by the ʿirāfa (administrative heads of tribal units).196 Tillier and Vanthieghem also point to a number of documents, published and unpublished, that can be dated to the first sixty years after the conquest of Egypt and that list individuals connected with troops (madad), military units (ʿashīra), and families (ʿiyāl).197 However, the documents are undated and most of them are fragmentary, as such, it is difficult to use them to reconstruct practices of recording military payments in the administration of the Egyptian governor, as opposed to that of individual tribes — and to assert if some of them can be ascribed to the pre-Marwanid period. Ultimately, the translation of registers was part of a wider process of implementation of a new provincial order in Egypt and Syria during which the administration of taxation and that of the army were formalized within two different dīwāns and, in Egypt, in the attributions of governors. This development also promoted new 191 Bouderbala, ‘Ǧund Miṣr’, pp. 174–47, 195–201. P.Lond. IV 1447 is translated in Morimoto, ‘The Dīwāns as Registers of the Arab Stipendaries in Early Islamic Egypt’, pp. 362–64. 192 The same can be said about P.Ness. III 92 (c. 685), a Greek account from Nessana (Palestine), which Kennedy first interpreted as a fragment of a list of payments to the military in kind and in cash, noting that though it can reflect the contents of the central provincial dīwān of Palestine, the document was used locally; Kennedy, ‘The Financing of the Military in the Early Islamic State’, pp. 376–78. Kennedy subsequently reached the conclusion that the document was not part of the dīwān; Kennedy, The Armies of the Caliphs, pp. 66–67. On the correspondence between the administrative documentation found throughout the province and that held in the central dīwān, see below, pp. 124ff. 193 Sijpesteijn, ‘Army Economics’, p. 256. The first attempt at listing documents of this kind is found in Puin, Der Dīwān von ʿUmar ibn al-Ḫaṭṭāb, pp. 120–24. 194 Al-Qāḍī sees in this document a record of the army participating in the conquest of North Africa in 27/647; al-Qāḍī, ‘Death Dates in Umayyad Stipends Registers?’, pp. 70–71. 195 The first is P.Ryl.Arab. p. 31, the second is published in Sijpesteijn, ‘A Seventh/Eighth-Century List of Companions from Fusṭāṭ’. The latter mentions (at pp. 373–74) a mosque that is known to have been built in the Marwanid period (the mosque of Qurūn), which provides a basis for dating. 196 Tillier and Vanthieghem, ‘Recording Debts in Sufyānid Fusṭāṭ’, pp. 149–52. 197 Tillier and Vanthieghem, ‘Recording Debts in Sufyānid Fusṭāṭ’, p. 148.

123

1 24

m a r i e l ege n d r e

types of administrative elites. The available information shows consistently that the distinction between military (ʿarab) and a fiscal administration (ʿajam), as al-Layth b. Saʿd and others describe them, was introduced in the beginning of the Marwanid period. Before that, fiscal administration in Egypt appears to have been decentralized as it had remained in the hands of the dukes. The same could likely be said of the organization of military payments, which were handled by individual tribes or subgroups, a decentralized system before the Marwanid period, though the details of that system entirely escapes us. In Syria, it is said that the registers of the Arabs and of the non-Arabs were both in the hands of Sarjūn b. Manṣūr before the translation, and it is possible that they were all in Greek. The translation, then, was part of a far-reaching creative act that centralized information as much as it translated it. What was transformed was not only the registers but the whole administrative hierarchy, as administrators such as Sarjūn and offices such as those of the dukes were replaced or supressed. The following section dives further into this creative process with a focus on the contents of the registers, the secretaries who handled them, and the milieus in which those secretaries were trained.

Technicalities and Training in Second/Eighth-Century Multilingual Milieus The different Arabic verbs used to refer to the translation of the dīwān testify to the various aspects of the translation process. The earliest Arabic work that refers to the translation, Khalīfa b. Khayyāṭ’s Taʾrīkh, includes the following in a list of officials active at the time of ʿAbd al-Malik: ‫‌‌الخراج والجند ‌سرجون ‌بن ‌منصور الرومي فمات سرجون فولى سليمان بن سعد مولى خشين حي‬ ‫من قضاعة وهو أول من ترجم ديوان الشام بالعربية‬

[In charge of] the kharāj and the jund: Sarjūn b. Manṣūr al-Rūmī. Then Sarjūn died and Sulaymān b. Saʿd, a mawlā of Khushayn, a clan of Quḍāʿa, succeeded him, he was the first to translate the dīwān of al-Shām into Arabic.198 The verb ‫ ترجم‬used by Ibn Khayyāṭ is one of the least commonly used verbs in references to the translation: it appears in only two other sources, the works of Ibn ʿAsākir and al-Ṣafadī, which in fact quote Ibn Khayyāṭ’s text.199 In the second/

198 Khalīfa b. Khayyāṭ, Taʾrīkh, p. 302. In Khalifa Ibn Khayyat’s History on the Umayyad Dynasty, pp. 169–70, Wurtzel and Hoyland offer the following translation: ‘He was the first to use Arabic for the records of the Syrian dīwān’. 199 Ibn ʿAsākir, Taʾrīkh madīnat Dimashq, xxii, p. 320: ‫أخبرنا أبو غالب محمد بن الحسن أنا محمد بن علي السيرافي‬ ‫أنا أحمد بن إسحاق نا أحمد بن عمران نا موسى بن زكريا نا خليفة بن خياط قال في تسمية كتاب عبد الملك الخراج والجند‬ ‫سرجون بن منصور الرومي فمات سرجون فولى يعني عبد الملك سليمان بن سعد مولى خشين حي من قضاعة وهو أول من‬ ‫ترجم ديوان الشام بالعربية‬.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

eighth century, the most common meaning of the verb ‫ ترجم‬was to add a heading or a title to a chapter or a book.200 The most common terms referring to the translation are taken from the root n-q-l. Words sharing this root appear in twenty of the thirty texts, making them the predominant termino­logy. The term naql bears several meanings, both general and technical: it can denote a shift, a transfer, a translation, or a copy.201 The term could thus refer to a change of personnel and a transfer of authority from one official to another. In works such as Ibn al-Nadīm’s Fihrist, al-naql al-qadīm refers to the oldest translations from Greek to Arabic, and the term al-naqala is used for translators.202 As already mentioned in the introduction, this meaning (translation) was selected by al-Qāḍī, focusing on the Iraqi narrative of al-Balādhurī, and by Saliba, who focused on that of Ibn al-Nadīm.203 Across texts, words from the root n-q-l are used as both verbs and nouns: ‫أمر الحجّاج صالحا بنقل الدواوين‬ ‫إلى العربية‬

‫فنقل ديوان الشام إلى العربية سنة‬ ‫إحدى وثمانين‬

Al-Ḥajjāj ordered Ṣāliḥ to [under­take] the translation of the dīwāns into Arabic.204

He translated the dīwān of al-Shām into Arabic in the year 81.205

The second most common verb is ‫حول‬ ّ , which appears in eleven texts. It is mostly used with al-ḥisāb, ‘the accounts’ (ten times across the corpus; on four further occasions, al-ḥisāb is used with the verb naqala). It is a verb that is rarely used in the context of translation of works of literature.206 In that sense, ḥawwala seems

200 Dozy, Supplément aux dictionnaires arabes, i, p. 143. Ibn al-Nadīm lists al-Balādhurī among the trans­ lators from Persian to Arabic but adds that he actually versified or added some verses (‫ )ترجمه بشعر‬to the ʿAhd Ardashīr; Ibn al-Nadīm, Kitāb al-Fihrist, pp. 126, 305. Indeed, according to Jaakko HämeenAnttila, Ibn al-Nadīm is the only one to classify al-Balādhurī as a translator, and the latter’s profile is rather unusual for a translator. Further, al-Balādhurī was active after the heyday of translations from Persian to Arabic. It is very likely that the ʿAhd Ardashīr would have been translated into Persian already before his time. Hämeen-Anttila, Khwadāynāmag, pp. 35, 62 n. 8. 201 In Geniza documents, the term refers to the transcription or copying of documents into other texts or into registers; Rustow, The Lost Archive, p. 323. 202 Ibn al-Nadīm, Kitāb al-Fihrist, p. 305; on naql as transmission or translation, see Hämeen-Anttila, Khwadāynāmag, pp. 99–100. 203 Al-Qāḍī, ‘The Names of Estates in State Registers’, p. 264; Saliba, Islamic Science and the Making of the European Renaissance, pp. 45–47. Sprengling, ‘From Persian to Arabic’, p. 191, prefers the ‘transfer of the official tax books from Persian, Greek, Coptic, and perhaps Soghdian to Arabic’; on p. 209, he describes Ṣalīḥ b. ʿAbd al-Raḥmān as the one who ‘converted the dīwāns from Persian to Arabic’. 204 Al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 38. 205 Al-Māwardī, al-Aḥkām al-sulṭāniyya, p. 301. 206 My thanks to Jaakko Hämeen-Anttila for pointing this out to me. The three most common verbs used with reference to the dīwān are found together in a passage of al-Jāḥiz’̣ s Kitāb al-Ḥayawān, i, p. 53: ‫وحولت آداب الفرس‬ ّ ،‫ وترجمت حكم اليونانيّة‬،‫وقد‌نقلت كتب الهند‬. Al-Jāḥiẓ’s word choice here is guided in part by stylistic considerations, to avoid repeating the same verb three times. The passage is translated in Yücesoy, ‘Translation as Self-Consciousness’, p. 536: ‘The books of India have been transferred

125

1 26

m a r i e l ege n d r e

to be connected to the translation or conversion of numbers, and it can mean to change, to transfer, to convert or to translate:207 ‫فحول ديوان العراق إلى العربية‬ ّ

He converted the dīwān of Iraq into Arabic.208

‫أحول الحساب من الرومية‬ ّ ‫لو شئت‬ ‫إلى العربية لحولته‬

If you wanted that I convert the accounts from Greek into Arabic, I could do it.209

The verb ‫ قلب‬has a similar meaning (‘to change’ or ‘to convert’) and occurs in nine texts, whereas ‫‘( نسخ‬to make a copy’) appears in only three texts, all focusing on Egypt: ‫قلب‌الدواوين من الفارسية إلى العربية‬

‫وأمر عبد هللا بن عبد الملك بالدواوين‬ ‫فنسخت بالعربيّة‬

He converted the dīwāns from Persian to Arabic.210

ʿAbd Allāh b.  ʿAbd al-Malik

ordered that the dīwāns be copied into Arabic.211

Finally, the more general verb ‫‘( جعل‬to make’, ‘to render’) occurs only once: ‫فعزم الحجاج على أن يجعل الديوان بالعربية‬ Al-Ḥajjāj resolved to render the dīwān into Arabic.212 It is worth emphasizing that all the texts are clear in referring to a translation of the dīwān or a conversion of accounts, never to a decree or a reform. The variety of verbs used in the Arabic sources shows that the English ‘to translate’ and ‘translation’ are poor equivalents and simplify the process.213 For lack of a better word, however, they are still used throughout this article, with the hope that what follows will adequately demonstrate how they should be understood in connection with the Marwanid dīwān. The variations of termino­logy in Arabic fully reveal the multidimentional change taking place: a change of personnel, a change of language, and a creative transfer of content — overall, the transformation of a text that includes numbers.

(nuqilat), and the Greek philosophies have been translated (turjimat), and the literature of the Persians has been converted (huwwilat)’. 207 Dozy, Supplément aux dictionnaires arabes, i, p. 338. 208 Al-Ṣūlī, Adab al-kuttāb, p. 192. 209 Ibn al-Nadīm, Kitāb al-Fihrist, p. 303. 210 Ibn al-Faqīh, Kitāb al-Buldān, p. 257. 211 Al-Kindī, Kitāb al-Wulāt, pp. 58–59. 212 Al-Balādhurī, Futūḥ al-buldān, p. 301. 213 Ricci, Islam Translated, pp. 31–32.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

Writing Numbers in the Dīwān

The translation narratives give further details on the question of numbers. In al-Madāʾinī’s Iraqi narrative, Zādhān Farrūkh is said to have boasted to Ṣāliḥ b. ʿAbd al-Raḥmān: ‘He [al-Ḥajjāj] needs me more than he needs you, because he cannot find anyone but me to accomplish his accounting for him’ (‫هو أحوج إل ّي منه إليك ألنه‬ ‫)ال يجد من يكفيه حسابه غيري‬.214 Later in the account, Ṣāliḥ b. ʿAbd al-Raḥmān is asked to pretend that the translation of the dīwān was unachievable. After the translation was decided, a dialogue between Ṣāliḥ b. ʿAbd al-Raḥmān and Zādhān Farrūkh’s son Mardānshāh about the translation of technical terms in the dīwān ensues.215 Mardānshāh asks Ṣalīḥ how fractions could be translated into Arabic, pointing to the challenges of such an undertaking. This is the only part of al-Madāʾinī’s report that is reproduced in al-Rāghib al-Iṣfahānī’s (d. c. 443/1060) Muḥāḍarāt al-udabāʾ wa-muḥāwarāt al-shuʿarāʾ wa-l-bulaghāʾ.216 Theophanes agrees that the numbers posed a problem when the registers were converted to Arabic; the translation was carried out, he says, ‘except for the numerals, because it is impossible in their language to write a unit or a pair or a group of three or a half or a third. For this reason they have Christian notaries until this day’ (χωρὶς τῶν ψήφων, ἐπειδὴ ἀδύ νατον τῇ ἐκείνων γλώσσῃ μονάδα ἢ δυάδα ἢ τριάδα ἢ ὀκτὼ ἥμισυ ἢ τρία γράφεσθαι διὸ καὶ ἕως σήμερόν εἰσι σὺν αὐτοῖς νοτάριοι Χριστιανοί).217 Centuries later, Ibn Khaldūn concurred, saying that in the early Islamic period, ‘accounting remained in the command of mawlās and dhimmīs’ (‫)وبقي أمر الحسبان في الموالي والذميين‬, meaning people of non-Arab origin.218 In the case of Egypt, however, it seems odd that fractions or figures in general needed a translation, since documents written in all the relevant languages — Greek, Coptic, and Arabic — use the same numerical notations, namely, the Greek alphanumerical system. Arabic accounting continued to be done with Greek numerals long after Greek texts had disappeared from the Egyptian documentary record. This means that only the names in the registers needed to be translated, not the numbers. One might wonder whether this was the case everywhere in the empire. Frede Løkkegaard took the dialogue between Ṣāliḥ and Mardānshāh about the difficulties of translating fractions into Arabic as proof that ‘Persian ordinal numbers’ were still used in the registers of Iraq after the translation.219 However, the situation in Khurasan is not clear, as one difference between Egyptian fiscal receipts and those of Khurasan in the early Abbasid period is the absence of figures in the latter. In the Khurasani documents, numerical values are always expressed in

214 Al-Balādhurī, Futūḥ al-buldān, p. 301. 215 See, for instance, Ibn al-Nadīm, Kitāb al-Fihrist, p. 303. Only al-ʿAskarī ascribes this dialogue to Zādhān Farrūkh and not to his son: al-ʿAskarī, Kitāb al-Awāʾil, p. 256. On the issue of fractions, see al-Qāḍī, ‘The Names of Estates in State Registers’, p. 256 n. 3; Sprengling, ‘From Persian to Arabic’, p. 196. 216 Al-Rāghib al-Iṣfahānī, Muḥāḍarāt al-udabāʾ, i, p. 126. 217 Theophanes, Chrono­graphia, p. 376; Hoyland, Theophilus of Edessa’s Chronicle, p. 200. 218 Ibn Khaldūn, Taʾrīkh, i, p. 296; Yarbrough, Friends of the Emir, p. 3. 219 Løkkegaard, Islamic Taxation in the Classic Period, p. 133.

127

1 28

m a r i e l ege n d r e

words, and there is no numerical summary at the end or the top of the documents like that often found in the Egyptian papyri.220 It is very probable, then, that in the process of the naql, the dīwān took a hybrid form, featuring Greek accounting and Arabic names or only headings (tarjama). The evidence for Iraq is less firm given the dearth of documents. Moreover, in the earliest preserved bilingual administrative documents from Egypt, the Greek and Arabic parts are never identical.221 Administrative texts in each language drew from distinctive textual traditions, and, as such, they do not give the same type of information.222 The Greek, Persian and Arabic registers certainly had similar differences. Jaakko Hämeen-Anttila’s description of the transmission of the Khwadāynāmag from Persian into Arabic then applies as well to the translation of the dīwān: One might from a modern point of view speak of adaptations, recreations, or redactions, rather than translations proper. For our purposes, a translation may be defined as any new text in the target language that reproduces, partly or completely, a text in the source language, with or without enlargements and embellishments, abbreviations and changes.223 Overall, although the translation of the dīwān appears as a creative process that certainly took a different shape in the various regions, it does not appear at first as a technical prowess as far as the change from one language to another is concerned: the project entailed translating names but not necessarily numbers. It would not have been an insurmountable problem for non-Muslim secretaries if this had been only about changing a language or even a script. Sarjūn b. Manṣūr, Zādhān Farrūkh, and Athanasius bar Gūmōyē most likely knew Arabic, and they certainly knew enough Arabic to write personal names after spending thirty to fifty years in the Umayyad administration.224 Mastery of the Arabic language was not a skill reserved to the new appointees over the dīwān; the choice of personnel reflected judgements about who could legitimately head the administration of the Arabic dīwān, not about the candidates’ ability to read or write Arabic. The difficulty of drawing up the dīwān lay elsewhere.

220 Khan, Arabic Documents from Early Islamic Khurasan. 221 Papaconstantinou, ‘Administrating the Early Islamic Empire’, pp. 65–66; Richter, ‘Language Choice in the Qurra Dossier’, pp. 198–200; Sijpesteijn, Shaping a Muslim State, pp. 67–69. The same can be said about official seals in Arabic and Greek, which bear different information: Sijpesteijn, ‘Expressing New Rule’, pp. 131–32. 222 On the continuation of Arabian administrative practices as evident in documents preserved throughout the early Islamic empire, see Khan, ‘The Opening Formula and Witness Clauses in Arabic Legal Documents from the Early Islamic Period’, pp. 25–28. 223 Hämeen-Anttila, Khwadāynāmag, pp. 51–58; the quotation is on p. 54. See also Yücesoy, ‘Translation as Self-Consciousness’, pp. 530–31, and Ricci, Islam Translated, pp. 33–34, 41. 224 Zādhān Farrūkh is said to have written letters of appointment for sub-governors on behalf of al-Ḥajjāj, undoubtedly in Arabic: Sprengling, ‘From Persian to Arabic’, p. 189. See also Debié, ‘Christians in the Service of the Caliph’, p. 56, on Athanasius bar Gūmōyē’s likely command of Arabic, an opinion shared by Zychowicz-Coghill, The First Arabic Annals, p. 48.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

Working and Training in the Dīwān

Al-Balādhurī’s final note in the Iraqi narrative on the authority of Sahl b. Abī al-Ṣalt claims that al-Ḥajjāj allocated a specific amount of time for Ṣāliḥ to finish his translation, though he does not say how long.225 According to Ibn Khaldūn, it took Ṣāliḥ one year to finish the translation.226 Al-Nuwayrī is the only source to add that the work of Sulaymān b. Saʿd and Ṣāliḥ b. ʿAbd al-Raḥmān was only the beginning of the translation of the dīwāns.227 Al-Kindī’s Egyptian narrative indicates that ʿAbd Allāh b. ʿAbd al-Malik ordered the production of a copy of the dīwān — an Arabic copy of the Coptic (meaning Greek) original.228 This seems to imply that there would have been a dīwān in Greek and a dīwān in Arabic that could have functioned in parallel for some time. Further information found in Egyptian papyri shows us how Greek registers and the Arabic dīwān were handled together in the long term. As mentioned above, the registers of the various pagarchies of the province were sent to Fusṭāṭ before and after the naql for control and for confirmation of the yearly tax assessments.229 In both known cases, Edfu in the 660s to 680s and Aphrodito in the early second/eighth century, the registers sent to Fusṭāṭ were in Greek.230 In the early second/eighth century, some of the information contained in the Greek registers was surely copied into the new central dīwān; the translation of information from Greek registers to the Arabic dīwān continued. Even after the translation of the central dīwān into Arabic, there was a constant need for a secretarial elite that had ample expertise in the arts of translation required to handle the Greek registers coming in from the kūras but also in Greek accounting, to enable them to maintain the Arabic dīwān.231 Some Islamic sources refer to the son of Sarjūn b. Manṣūr (Manṣūr b. Sarjūn) working in the administration of Syria after the translation.232 In Iraq, we find the son of Zādhān Farrūkh, Mardānshāh, and his great-grandson, Mājusabs b. Bahrām b. Mardānshāh b. Zādhān Farrūkh, working in al-Ahwāz at the time of Marwān b. Muḥammad (r. 127–32/744–50).233 225 Al-Balādhurī, Futūḥ al-buldān, p. 301: ‫ أنبأنا سهل بن أبي‬:‫ قال‬،‫وحدثني عمر بن شبة قال حدثني أبو عاصم النبيل‬ ‫ أجل الحجاج صالح بن عبد الرحمن أجال حتى قلب الديوان‬،‫ قال‬،‫الصلت‬. 226 Ibn Khaldūn, Taʾrīkh, i, p. 303: ‫فأكمله لسنة من يوم ابتدائه‬. 227 Al-Nuwayrī, Nihāyat al-arab, viii, p. 200: ‫هذا ما حكى فى ابتداء نقل الدواوين‬. 228 Al-Kindī, Kitāb al-Wulāt, pp. 58–59. 229 See above, pp. 114–24 and nn. 188–89. 230 The fragments of registers preserved from Aphrodito are dated between c. 705 and 721; Richter, ‘Language Choice in the Qurra Dossier’, p. 196. 231 As al-Qāḍī puts it in ‘Identity Formation of the Bureaucracy of the Early Islamic State’, pp. 143–44: ‘Those hired in the financial department having to know additionally the relevant local language, and to know both languages well enough to be able to move swiftly between them, translating words, concepts, and processes from the local languages into Arabic’. 232 Ibn al-Nadīm, Kitāb al-Fihrist, p. 303; al-Maqrīzī, al-Khiṭaṭ, ed. by Sayyid, i, p. 265; Anthony, ‘Fixing John Damascene’s Bio­graphy’, pp. 618–22. 233 Zādhān’s sons appear in the translation narrative of Ibn al-Nadīm, Kitāb al-Fihrist, p. 303, and Mardānshāh is also mentioned in al-Balādhurī, Futūḥ al-buldān, pp. 300–01; Sprengling, ‘From Persian to Arabic’, pp. 186, 190–91. For Mājusabs, see al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 99. Mājusabs worked for Sulaymān b. Ḥabīb. None of Ṣāliḥ b. ʿAbd al-Raḥmān’s sons is known to have worked in

129

130

m a r i e l ege n d r e

Undoubtably the new leading administrative elites, the mawlās, knew non-Arabic languages as well. It is unlikely that the various dīwāns were fixed texts after the translation. It has already been mentioned that military pay could increase or, supposedly, decrease.234 The composition of the fighting army and the jund changed regularly, and military registers needed updates.235 The same can be said about the fiscal registers, which had to be updated at least once a year. Other evidence indicates that a large body of secretaries was needed to handle this ever-evolving corpus. Tillier and Vanthieghem have remarked that debt registers from Sufyanid Fusṭāṭ were written in several hands, meaning by different scribes.236 The abovementioned P.Lond. IV 1447 (Aphrodito, before 705) refers to Athanasius bar Gūmōyē as responsible for the general affairs of the various provinces and to his son Stephanos, ‘the illustrious chartularius, in charge of requisitioning in Upper Egypt’, on top of no fewer than sixty-four secretaries (νοτάριοι) working for Athanasius and only one Arabic notary, Sulaym b. Simʿān, attached to the governor ʿAbd al-ʿAzīz.237 Even if this document dates from a time right before the translation of the dīwān of Egypt, it points to a significant staff working under the head of the dīwān and gives an idea of the amount of personnel necessary to handle the registers.238 This personnel was certainly delegated the everyday updates of names and numbers in the dīwān. This means that the translation was a collective effort. Al-Qāḍī even posits that Qaḥdham and the other pupils of Ṣāliḥ b. ʿAbd al-Raḥmān mentioned by al-Jahshiyārī continued the translation project after their master’s death.239 the administration, but one son of Sulaymān b. Saʿd, called Thābit, was in charge of the dīwān al-rasāʾil during the short caliphate of Yazīd al-Walīd (r. 126/744): al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 69; Sprengling, ‘From Persian to Arabic’, p. 212. 234 See above, at n. 190. 235 According to Sijpesteijn, ‘Army Economics’, p. 255, the military dīwān of Egypt was updated four times in the Marwanid period. On new recruitments in the Egyptian army in the early Marwanid period, see Bouderbala, ‘Les mawālī à Fusṭāṭ aux deux premiers siècles de l’islam et leur intégration sociale’. On how death dates were recorded in the Marwanid period, see al-Qāḍī, ‘Death Dates in Umayyad Stipends Registers?’. 236 They were also written in the form of rolls rather than of codices: Tillier and Vanthieghem, ‘Recording Debts in Sufyānid Fusṭāṭ’, pp. 150–51. 237 This might simply be the only Arabic notary mentioned in the context of this document; others may have worked for the governor in Fusṭāṭ and Ḥulwān. See below, pp. 136–39, where the eldest son of Athanasius, Peter, is mentioned in the Chronicle of 1234. According to Sebastian Richter, notarios was ‘a title usually borne by professional clerks attached to public offices, who were concerned with administrative writings’; Richter, ‘Language Choice in the Qurra Dossier’, p. 212 (see also pp. 211–14 for notaries in the Aphrodito documents). In L’invention du cadi, pp. 55–58, Mathieu Tillier reconstructs a section of the administration of Fusṭāṭ specialized in judicial questions through mentions of scribes in the letters of the governor Qurra b. Sharīk. One scribe, Muslim b. Labnan is mentioned in all preserved letters stating the instructions of the governor to the pagarchs on legal cases, whereas more than twenty different scribes are attested for fiscal documents (bilingual entagia). 238 In comparison, the Edict XIII of emperor Justinian (r. 527–565) desribes the entire administration of the dux et augustalis at Alexandria in the sixth century, which comprised 600 officials: Palme, ‘Administration, Late Antique Egypt’. 239 Al-Qāḍī, ‘The Names of Estates in State Registers’, p. 264. The other pupils are al-Mughīra b. Abī

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

That Ṣāliḥ is remembered as having trained a large number of secretaries to handle the Arabic dīwān says something about the scope of the translation project. As a true successor to Zādhān Farrūkh, he remained at the head of the dīwān of Iraq for approximately twenty years.240 The importance of the Basran secretarial milieu and the legacy of Ṣāliḥ b. ʿAbd al-Raḥmān are clear in a quotation ascribed to the most famous secretary of the Umayyad period, ʿAbd al-Ḥamīd al-Kātib (d. 132/750), with which most authors transmitting al-Madāʾinī’s report conclude their translation narratives: ‘ʿAbd al-Ḥamīd b. Yaḥyā, the secretary of Marwān b. Muḥammad, used to say: “How excellent was Ṣāliḥ and how great was his gift to the secretaries!”’ ( ‫ هلل در صالح ما أعظم منته على الكتاب‬:‫)فكان عبد الحميد بن يحيى كاتب مروان بن مح ّمد يقول‬.241 Among the Basran secretaries of the first half of the second/eighth century we find another important personality, Dādūye al-Muqaffaʿ, who is said to have been in charge of the fiscal administration of Fars around 120/737–38 and is subsequently encountered as one of the non-Muslim secretaries of the governor Khālid al-Qasrī in Basra.242 Dādūye al-Muqaffaʿ forms a link between the Basran milieu of the translation of the dīwān and the so-called translation movement through his son Rūzbih son of Dādūye, otherwise known as ʿAbd Allāh b. al-Muqaffaʿ (d. 139/756), who, unlike his father, eventually converted to Islam.243 Though Ibn al-Muqaffaʿ is better known for his translations of works of adab, and the details of his bio­graphy are far from clear, he is said by some to have been trained in Basra by his father. He then pursued an administrative career in Nishapur and in Kirman

Qurra, Shayba b. Ayman, al-Mughīra and Saʿīd (two sons of ʿAṭiyya), and Marwān b. Iyās. Al-Qāḍī also writes (pp. 274–75), on the basis of P.Cair.Arab. I 31, that ‘the full completion and implementation of the Arabization of the registers’ can be dated to 110/728. This document is the earliest dated protocol written entirely in Arabic, though it is actually dated to 116–19/734–37. 240 Sprengling, ‘From Persian to Arabic’, p. 205. 241 Al-Balādhurī, Futūḥ al-buldān, p. 301. See also Ibn al-Nadīm, Kitāb al-Fihrist, p. 303; al-Māwardī, al-Aḥkām al-sulṭāniyya, p. 302; al-Nuwayrī, Nihāyat al-arab, viii, p. 200; Ibn Khaldūn, Taʾrīkh, i, p. 303; and al-Maqrīzī, al-Khiṭaṭ, ed. by Sayyid, i, p. 265. ʿAbd al-Ḥamīd al-Kātib’s connection to Basra is less clear, but it is usually assumed that he was of Iraqi origin. On his life and the statement about Ṣāliḥ b. ʿAbd al-Raḥmān, see al-Qāḍī, ‘Identity Formation of the Bureaucracy of the Early Islamic State’, pp. 144–45. According to Sprengling, Ṣāliḥ’s fame amongst later secretaries explains why so many details have been transmitted about his early life and family; Sprengling, ‘From Persian to Arabic’, pp. 192–93. 242 Arjomand, ‘ʿAbd Allah Ibn al-Muqaffaʿ and the ʿAbbasid Revolution’, pp. 13–14; Gundelfinger and Verkinderen, ‘The Governors of al-Shām and Fārs in the Early Islamic Empire’, p. 322. 243 That is in a different way than what George Saliba has argued as he posits that hypothetical administrative manuals were translated as part of the translation movement (see above, at n. 7). On this point, see also the remarks of Petra Sijpesteijn in her contribution to this volume. It may be worth focusing primarily on the mid-second/eighth-century secretaries and their literary activities, as shown here, to find that connection between the translation of administrative documents and the translation movement, and there does not seem to be a need for manuals. Saliba indeed does not focus much on Ibn al-Muqaffaʿ. He says only that the latter was the student of ʿAbd al-Ḥamīd al-Kātib: Saliba, Islamic Science and the Making of the European Renaissance, p. 45; that is following al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 80. However, the few information available on the career of the two administrators are hard to reconcile.

131

13 2

m a r i e l ege n d r e

before returning to Basra.244 In the list of secretaries copied by one of our earliest sources, al-Baghdādī, the Basran translator of the dīwān (Qaḥdham, according to him) and Ibn al-Muqaffaʿ are placed in the same category.245 Other examples of secretaries active in the field of translation are Jabala b. Sālim and, possibly, Sālim Abū al-ʿAlāʾ, both of whom worked in the administration of Hishām b.  ʿAbd al-Malik. This shows that secretaries were involved in the translation of literature not only in Iraq but also in Syria.246 Crucially, Sālim Abū al-ʿAlāʾ was also one of the teachers of ʿAbd al-Ḥamīd al-Kātib.247 In light of this information, we are able to further elaborate on the skills and diverse training of the secretaries employed in the administration in the late Umayyad period. ʿAbd al-Ḥamīd, in his letter to the secretaries (Risāla ilā al-kuttāb), had high expectations for the secretaries of his time, claiming that they should have knowledge of ‘the Qurʾān, religious duties, the Arabic language, penmanship, poetry, the history of pre-Islamic Arabs and of the non-Arabs, and accounting’.248 In Maaike van Berkel’s words, secretaries of the Abbasid world developed over time as ‘men of letters more than of numbers’.249

244 Sourdel, ‘La bio­graphie d’Ibn al-Muqaffa d’après les sources anciennes’, pp. 208–11; Hämeen-Anttila, Khwadāynāmag, pp. 89–99. Several authors posit that Ibn al-Muqaffaʿ was trained by Ṣāliḥ b. ʿAbd al-Raḥmān on the basis of an anecdote reported by al-Balādhurī in the Futūḥ al-buldān. The text reports that Ibn al-Muqaffaʿ wrote documents destined to Ṣāliḥ on scented parchment, something that deeply pleased the latter. The anecdote is transmitted on the authority of al-Madāʾinī through Ibn al-Muqaffaʿ. Al-Balādhurī, Futūḥ al-buldān, pp. 463–64; Sprengling, ‘From Persian to Arabic’, p. 204. According to Francesco Gabrieli, however, the known chrono­logy of the two men’s careers does not agree with this and he rejects the idea that Ibn al-Muqaffaʿ ever worked for Ṣāliḥ; Gabrieli, ‘L’opera di Ibn al-Muqaffa’, p. 207. Said Amir Arjomand argues that the anecdote must have been about Ibn al-Muqaffaʿ’s father, Dādūye, since Ibn al-Muqaffaʿ was born in 101 or 102/720 — that is, after Ṣāliḥ’s dismissal from his post; Arjomand, ‘ʿAbd Allah Ibn al-Muqaffaʿ and the ʿAbbasid Revolution’, p. 13. 245 Sourdel, ‘Le “livre des secrétaires” de ʿAbdallāh al-Baġdādī’, p. 140: .‫أسماء الكتّاب الذين تقدّموا بالبالغة والعلم بالكتابة‬ 246 On Jabala b. Sālim’s translation work from Persian to Arabic, see Ibn al-Nadīm, Kitāb al-Fihrist, p. 305; Hämeen-Anttila, Khwadāynāmag, p. 186. As for Sālim Abū al-ʿAlāʾ, there are disagreements on whether he translated (from Greek or possibly Syriac) or composed some of the letters of the so-called Epistolary Novel of Aristotle and Alexander; see Gutas, ‘On Graeco-Arabic Epistolary “Novels”’ and Rassi, ‘Alchemy in an Age of Disclosure’, pace Maróth, The Correspondence between Aristotle and Alexander the Great, pp. 74–94, and Cottrell, ‘An Early Mirror for Princes and Manual for Secretaries’. My thanks to Salam Rassi for pointing this out to me. 247 Al-Qāḍī, ‘Sālim Abū al-ʿAlāʾ’. 248 Al-Qāḍī, ‘Identity Formation of the Bureaucracy of the Early Islamic State’, p. 147. This is in line with Saliba’s arguments in Islamic Science and the Making of the European Renaissance, pp. 54–58. Bouderbala argues that a good knowledge of tribes and their genealogies was also necessary for the head of the dīwān al-jund in ‘Ǧund Miṣr’, p. 182 n. 46. It is, however, not clear what a ‘good knowledge’ of tribal genealogies would have meant in the early Umayyad period, if some genealogical identities were still taking shape when the dīwān was created. On this topic, see Qaṭṭāṭ, al-ʿArab fī al-jāhiliyya al-akhīra wa-l-islām al-mubakkir; Webb, Imagining the Arabs. On attested tribes and other types of social groups identifiable in pre-Islamic Arabia, see Robin, ‘Tribus et territoires d’Arabie’. 249 Van Berkel, ‘Accountants and Men of Letters: Status and Position of Civil Servants in Early Tenth Century Baghdad’ (unpublished doctoral thesis, University of Amsterdam, 2003), pp. 38–40, quoted in Khalek, ‘Some Notes on the Representation of Non-Muslim Officials in al-Ǧahšiyārī’s Kitāb al-Wuzarāʾ wa-l-Kuttāb’, p. 517.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

The work of numbers was certainly handled by the large teams identified above working for men such as Athanasius bar Gūmōyē and Ṣāliḥ b. ʿAbd al-Raḥmān. According to Hayrettin Yücesoy, translation creates new meanings but also creates power.250 This applies not only to the translation movement, which is the subject of his study, but also to the translation of the dīwān that appears as the defining event allowing for the rise of a class of highly skilled mawlās as the dominant secretarial elite. Though it does not address the translation of the dīwān, ʿAbd al-Ḥamīd’s ahistorical account of the role of the secretaries in his Risāla ilā al-kuttāb and passages from Ibn al-Muqaffaʿ’s Risāla fī al-Ṣaḥāba on the organization of administration are statements of the emerging self-legitimizing discourse of the kuttāb, penned by the heirs of the translators of the Iraqi dīwān.251 In the long term, al-Jahshiyārī’s text may be the most striking attempt to use the translation of the dīwān as a core event that legitimized the Muslim mawlās against non-Muslim secretaries, but the roots of his effort lie almost two centuries earlier in the discourse of Marwanid secretaries.252 The clear role of the large teams of secretaries in general and of translators in particular who were handling the ever-evolving texts that constituted the provincial dīwāns of the Marwanid period is another reason to avoid writing secretarial agency out of the narrative of language change in the Umayyad empire. Further, the central dīwāns were not the only context in which we can identify resourceful secretaries. Non-Muslim secretaries who were experts in the non-Arabic languages of the empire did not become disposable in the early second/eighth century, as some of the translation narratives would like us to believe.253 We have seen that they were retained in the dīwāns of the provincial capitals, and the next section shows that their contribution is also visible in local chanceries within the province of Egypt.

250 Yücesoy, ‘Translation as Self-Consciousness’, p. 533. 251 Al-Qāḍī, ‘Identity Formation of the Bureaucracy of the Early Islamic State’, p. 147. On Ibn al-Muqaffaʿ’s Risāla fī al-ṣaḥāba, see Lucas, ‘Le pouvoir de la terre’, pp. 99–102. According to Khalek (‘Some Notes on the Representation of Non-Muslim Officials in al-Ǧahšiyārī’s Kitāb al-Wuzarāʾ wa-l-Kuttāb’, p. 517), ʿAbd al-Ḥamīd’s career and work ‘served as the prototype for the later ʿAbbāsid representations of kuttāb’. 252 As argued by Luke Yarbrough (Friends of the Emir, p. 22): ‘In writing disapprovingly of non-Muslim officials, Muslim authors pursued resources by deploying competitive practices, thus working to access and accumulate capital within particular fields’. On al-Jahshiyārī’s work, see Khalek, ‘Some Notes on the Representation of Non-Muslim Officials in al-Ǧahšiyārī’s Kitāb al-Wuzarāʾ wa-lKuttāb’, pp. 515–16: ‘The linguistic and technical skills at stake in discourses of competition between non-Muslim and Muslim secretaries illustrate the anxieties felt by ʿAbbāsid-era authors who were struggling with the cultural implications of an asymmetry between Muslim political and non-Muslim knowledge-based power in the early Islamic state’. To reiterate, the first universal history that includes a translation narrative is Miskawayh’s Tajārib al-umam, which is unsurprising since, in the words of Tarif Khalidi, he belonged to the class of ‘the state secretaries and fiscal experts of the old caliphate’. Khalidi, Arabic Historical Thought in the Classical Period, p. 171. 253 That is when Zādhān Farrūkh or Sarjūn b. Manṣūr tell their fellow secretaries to seek employment elsewhere as, for the sake of the narratives, the translation would have made experts in non-Arabic languages redundant. See above, sections 1.1. and 1.2.

133

134

m a r i e l ege n d r e

Contemporary Innovations in Greek and Coptic Secretarial Milieus

Papyro­logical documents from Egypt confirm that there was an increasing use of Arabic in Umayyad administration at the beginning of the second/eighth century that parallels the translation of the dīwān: Arabic administrative letters appear for the first time,254 and papyrus protocols become bilingual (Arabic-Greek) during the caliphate of ʿAbd al-Malik, though the first caliph mentioned in one of these documents is his successor, al-Walīd.255 However, Arabic is not the only language that gains new prominence in administrative communications of that time. Arguably the most interesting change in language use in the early Marwanid state is the introduction of Coptic as an administrative language in Egypt during the governorate of ʿAbd al-ʿAzīz b. Marwān, while ʿAbd al-Malik was caliph.256 The first Coptic administrative documents were issued in the name of the last duke of the Thebaid and Arcadia, Atias son of Juʿayd.257 In the Byzantine period, we never see fiscal documents or administrative correspondence written in Coptic. Only legal documents were written in this language before the Islamic conquest. As Jean-Luc Fournet puts it, in the late Byzantine period Coptic was ‘acquiring an official dimension and becoming an actor in public written culture, to the detriment of the monopoly that Greek had acquired for itself ’.258 This process continued after the Islamic conquest, creating a context in which the use of Greek was no longer promoted by Byzantine regulations although, as will be seen below, it was supported at the local level for a time. Still, possibilities increased for the development of the Egyptian language as an official medium.259 Coptic drew on the Greek administrative 254 This development is most evident in the documents of Aphrodito; see Richter, ‘Language Choice in the Qurra Dossier’, pp. 196–200. On the other side of the empire, Arabic letters from and to the Sogdian ruler Dhēwāshtīch have been preserved, including one addressed to the Umayyad governor of Khurasan, al-Jarrāḥ b. ʿAbd Allāh (in office 718–19); Kratchkovskaya and Kratchkovsky, ‘Drevnejshij arabskij dokument iz Srednej Azii’. See Sijpesteijn’s contribution to this volume for more information on these documents, also showing that documents in Sogdian were issued, at the same time, for the same people or in the same circles. 255 Sijpesteijn, Shaping a Muslim State, pp. 110–11. 256 Cromwell, Recording Village Life, pp. 13–14, 169–70, 184. For a full overview of multilingual administrative practices after the Islamic conquest, see Sijpesteijn’s contribution to this volume. 257 Delattre, ‘Cinq entagia coptes’; Cromwell, ‘The Coptic Texts in the Archive of Flavius Atias’; Delattre, Pintaudi, and Vanthieghem, ‘Un entagion bilingue du gouverneur ʿAbd al-ʿAzīz ibn Marwān’. 258 Fournet, The Rise of Coptic, p. 2. In Sheldon Pollock’s terms, a vernacular language challenged the monopoly of a cosmopolitan one. On his work, see Antoine Borrut’s chapter in this volume. See also Papaconstantinou’s introduction to The Multilingual Experience in Egypt, p. 6: ‘The sixth and seventh centuries comprised perhaps the period of the most balanced societal bilingualism in Egypt, with Greek and Egyptian enjoying similar status in terms of prestige and social use. After the Arab conquest, this balance shifted once again to the clear advantage of Coptic, Greek being gradually restricted, as in the Ptolemaic period, to the broader area of administration’. 259 As Sijpesteijn puts it: ‘The vacuum thus created by a decrease in Greek writings was not, however, filled by Arabic, the language of the new rulers, but rather by Coptic. From the eighth century onwards the use of Coptic, in fact, increased in the administration, being used for purposes and at levels of the administration at which Greek had been used thus far’. Sijpesteijn, ‘Multilingual Archives and Documents in Post-Conquest Egypt’, p. 123. Richter, in ‘Language Choice in the Qurra

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

tradition, using Greek administrative vocabulary and formularies to such an extent that Coptic entagia (tax demand notes) are often edited as bilingual or bi­graphic documents with dates, figures, and invocations all in Greek.260 Coptic and Greek administrative documents related to the administration of Atias son of Juʿayd display another innovation in the late first/seventh and early second/eighth centuries. Jennifer Cromwell has shown that new standardized formularies and a singular ‘Greek chancery style’ were introduced in Greek and Coptic administrative documents of this period.261 All the documents concerned by this change reveal the reach of the Umayyad state through tax collection, requisitions in kind, and forced labour.262 They are first issued for the duke Atias and then for the ʿāmils bearing Arab/Muslim names who progressively replaced the local pagarchs administering the kūras of the Egyptian valley at the turn of the second/eighth century.263 Those documents are attested in Middle Egypt (Antinoopolis, Hermopolis, Aphrodito) and in Djeme (Theban region); they were written by Aristophanes son of Johannes in Djeme and by Theodore in Aphrodito. This development was supported by what Cromwell calls a ‘standardized training’ for administrative secretaries writing Greek and Coptic documents.264 However, several aspects of this shift remain mysterious. Who introduced these innovations in Greek and Coptic? Determining the agency behind developments identifiable in documents is arguably one of the most challenging aspects of their study. Cromwell convincingly argues that Coptic administrative documents were developed to increase the efficiency of administration with a focus on tax collection. Drawing up Coptic administrative documents facilitated communications with the lower levels of the administration and with the tax payers as those documents employed the majority language of the population of the province.265 We need to recognise that those changes could not have been put in place exclusively because of top-down dynamics. Even if they increased the efficiency of administrative procedures, especially the collection of taxes, those changes in the shape and languages of administrative documents needed to mobilize local expertise. The involvement of the administration of the duke Atias appears as crucial, as it is where the documents

Dossier’, p. 216, concurs: ‘Written Coptic knew its widest spread during the first century after the Arab conquest of Egypt. Only then did it become a common means of recording private business and legal events, and a language of private representation in epi­graphy’. 260 De Jong and Delattre, ‘Greek as a Minority Language’, pp. 42–44; Cromwell, Recording Village Life, pp. 101, 235–50; Cromwell, ‘Scribal Networks, Taxation, and the Role of Coptic in Marwanid Egypt’, pp. 357–62. 261 Cromwell, Recording Village Life, pp. 163–78; on the chancery style, see also Bell, ‘Two Official Letters of the Arab Period’, pp. 265–66; Gonis, ‘Reconsidering Some Fiscal Documents from Early Islamic Egypt II’, p. 193. 262 Cromwell, ‘Language Policy and the Administrative Framework of Early Islamic Egypt’, p. 305; Cromwell, ‘Scribal Networks, Taxation, and the Role of Coptic in Marwanid Egypt’, p. 367. 263 Listed in Cromwell, Recording Village Life, pp. 163–64. 264 Cromwell, Recording Village Life, p. 176. 265 Cromwell, ‘Language Policy and the Administrative Framework of Early Islamic Egypt’, pp. 306–08; Cromwell, ‘Scribal Networks, Taxation, and the Role of Coptic in Marwanid Egypt’, pp. 370–72.

135

1 36

m a r i e l ege n d r e

are first attested. This change survived the disappearance of the office of the duke, meaning that local secretaries maintained their position in the administration of the valley despite the increasing involvement of Fusṭāṭ, the dismissal of the dukes, and the progressive replacement of the pagarchs. Cromwell argues that the diffusion of this standardized training could have been because secretaries experts in the non-Arabic languages travelled with Arab officials throughout the province.266 The introduction of ʿāmils in the kūras in place of the pagarchs did not go along with a comprehensive language change. Just as the head of the new Arabic dīwān needed to maintain a multilingual administration, so did the ʿāmils of the governor in the kūras. In this vein, Lajos Berkes and Naïm Vanthieghem have identified four documents containing aides-memoires for toponyms in Greek and Arabic or Greek and Coptic that were certainly used in these local multilingual chanceries.267 Pamela Klasova has shown that the bio­graphies of secretaries, examined in conjunction with contemporary documents, ‘suggest continuity, gradual transitions and multilingualism’ more than reforms or decrees.268 Taking into account the translation of registers into Arabic, on the one hand, and the introduction of new types of Greek and Coptic administrative documents, on the other, allows us to paint a rich picture of the innovations unfolding in the entirety of the state topped by the Umayyads, at least in Egypt. We also have access to different types of secretarial milieus in which the agents of language change operated, including the dynamic Persian and Arabic milieu of Basra, where the translation of the Iraqi dīwān took place, and the context of the Egyptian valley, where the documents of secretaries such as Aristophanes son of Johannes display innovations in Greek and Coptic administrative formularies. These secretaries of diverse backgrounds all contributed to shaping linguistic practices in the early Islamic period. Beyond this, the rise of the mawlās had economic implications both for themselves and for the non-Muslim secretarial elites; these are examined in the next section.

Economic Incentives In the Iraqi narratives on the authority of al-Madāʾinī, Zādhān Farrūkh attempts at several stages either to prevent the translation from happening or to slow down the process by demanding that Ṣāliḥ b.  ʿAbd al-Raḥmān stop the translation or pretend to be ill so that Zādhān would not lose control of the dīwān. What Zādhān

266 Cromwell, ‘Scribal Networks, Taxation, and the Role of Coptic in Marwanid Egypt’, p. 366: ‘The homogeneity must be by design and intentionally disseminated, in order to bring uniformity to the bureaucratic system. Two possibilities seem likely: scribes travelled as part of the retinue of Arab officials and trained local scribes; or local scribes were sent to a regional center to receive training therein. The first option is perhaps the most pragmatic and efficient dissemination method’. 267 Berkes and Vanthieghem, ‘Notes on the Careers of Nāǧid b. Muslim and ʿAbd al-Malik b. Yazīd’, p. 159. 268 Klasova, ‘Empire through Language’, pp. 104–11; the quotation is on p. 104.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

eventually lost was not only a position of prestige in the Umayyad administration: the change also had major economic implications.269 In the tellings in which Zādhān Farrūkh and Sarjūn b. Manṣūr live to see the translation, they are presented as lamenting the loss of a significant source of income for Persian and Greek secretaries.270 In al-Madāʾinī’s narrative, after Zādhān Farrūkh’s death, Ṣāliḥ b. ʿAbd al-Raḥmān is offered a bribe to pretend that it was impossible to translate the dīwān — 100,000 dirhams in most iterations, or 300,000 if we trust Ibn ʿAsākir and al-Dhahabī (d. 748/1348 or 753/1353). The bribe is offered by the remaining Persian secretaries or, according to al-Māwardī and al-Nuwayrī, by Zādhān Farrūkh’s son Mardānshāh.271 Only al-Jahshiyārī leaves out mention of the bribe; Ibn ʿAsākir and al-Dhahabī, meanwhile, take very little else from al-Madāʾinī’s report.272 Reports about Syria on the authority of al-Madāʾinī give further information on how much the translation could have been valued: Sulaymān b. Saʿd requested that he be given the yearly gold taxes (kharāj) of al-Urdunn, which amounted to 180,000 dinars. Al-Balādhurī says that Sulaymān b. Saʿd became governor of al-Urdunn, which was the only way he could explain why Sulaymān would have been given such an amount.273 According to al-Jahshiyārī, Sulaymān was originally in charge of the dīwān al-rasāʾil (correspondence) and after the translation he was placed at the head of the dīwāns of al-Shām.274 In Ibn ʿAsākir’s second narrative, Sulaymān is said to be originally in charge of the dīwān al-Urdunn.275 The amount of 180,000 dinars is very much at odds with information given elsewhere on remuneration for translation and administrative work. Al-Jahshiyārī reports that the secretary in charge of the dīwān al-kharāj received 300 dirhams a month at the time of al-Ḥajjāj (3600 dirhams a year = 144 to 180 dinars), but he does not refer to either Zādhān Farrūkh or Ṣāliḥ b.  ʿAbd al-Raḥmān in this context.276 In his investigation of the translation movement, Dimitri Gutas mentions that translators could be paid 500 dinars monthly to work on medical texts, which would amount to 6000 dinars per year.277 Therefore, the sum given in al-Balādhurī’s Syrian narrative is not merely a salary but a token of the stakes behind the translation of the dīwān. Similarly, Qaḥdham is presented as a wealthy man in al-Jahshiyārī’s text, ‘capable of raising in one day as much as 150,000 dirhams’.278 The economic incentives of heading the dīwān are also made clear in reports about Athanasius bar Gūmōyē. As noted above, the only thing that al-Jahshiyārī has

269 On this point, see also Saliba, Islamic Science and the Making of the European Renaissance, p. 59. 270 See, for instance, Ibn Khaldūn, Taʾrīkh, i, p. 303: ‫صناعة فقد قطعها هللا عنكم‬ ّ ‫اطلبوا العيش في غير هذه ال‬. 271 Al-Māwardī, al-Aḥkām al-sulṭāniyya, p. 301; al-Nuwayrī, Nihāyat al-arab, viii, p. 200. 272 Ibn ʿAsākir, Taʾrīkh madīnat Dimashq, xxii, p. 344; al-Dhahabī, Taʾrīkh al-islām, vii, p. 110. 273 Al-Balādhurī, Futūḥ al-buldān, p. 193. He is followed in this by Ibn Khaldūn, Taʾrīkh, i, p. 303. 274 Al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 40. 275 Ibn ʿAsākir, Taʾrīkh madīnat Dimashq, xxii, p. 320. 276 Al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 61. 277 Gutas, Greek Thought, Arabic Culture, p. 133. 278 Al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, p. 65; Al-Qāḍī, ‘The Names of Estates in State Registers’, p. 265.

137

138

m a r i e l ege n d r e

to say about Athanasius is that he was too rich. At the beginning of the khabar on al-Ḍaḥḥāk b. ʿAbd al-Raḥmān dividing up Athanasius’s wealth on ʿAbd al-Malik’s orders (this wealth consisted of copper objects, jewels, and gemstones), al-Jahshiyārī mentions that ʿAbd al-ʿAzīz also had a palace built for Athanasius in Fusṭāṭ.279 The Chronicle of 1234, reporting on the authority of Dionysius of Tell-Maḥrē, provides further details about Athanasius’ wealth: He was in charge of the tribute in Egypt. Since in addition to the money and honors he received from the king, he and his sons received each year one dinar for each soldier in Egypt, and knowing that there were 30,000 soldiers stationed in Egypt and that he stayed there for 21 years, it is not surprising that he became immensely rich.280 We do not need to trust these numbers to appreciate the message they convey. Being head of the dīwān of an Umayyad province was a highly lucrative position, and losing access to that specific post was a great change of fortune for Christian elites of the time. It was also a position through which wealthy non-Muslim individuals could benefit their wider communities, for example through investments in institutions, as the Chronicle of 1234 shows: He [Athanasius] was zealous for the orthodox faith. He repaired and built churches, gave to the poor and the orphans. He owned 4,000 slaves, villages, houses, gardens, gold, silver, and gemstones. In Edessa he owned 300 shops and 9 hospitals. His elder son Peter was in charge of his possessions in Edessa and the others helped him in the region of Gunada (?). In Egypt he built several churches and monasteries and in the city of Fostat he built two churches. He built in Edessa the beautiful church of the Mother of God. He also built a baptistery with channels of water such as those established by the Bishop Amazonios in the great and old church of Edessa. He adorned it with marble, gold, and silver.281 As Muriel Debié has argued, when Athanasius came back from Egypt to Damascus and ʿAbd al-Malik appropriated part of his property, ‘the challenge was to restrict the visible wealth of Christians to an acceptable level’.282 Restricting their access to the highest positions in the administration and promoting the developing class of mawlās trained as secretaries in their stead was certainly one way to rise to that challenge. One could assume that the head of the dīwān could also be petitioned for financial support or tax exemptions. However, with the translation of the dīwān by the growing class of mawlā secretaries, the privileged access that non-Muslims had

279 Al-Jahshiyārī, al-Wuzarāʾ wa-l-kuttāb, pp. 34–35. 280 The passage is transmitted in the chronicle of Michael the Syrian and the Chronicle of 1234 and is quoted and translated in Debié, ‘Christians in the Service of the Caliph’, p. 55. 281 Debié, ‘Christians in the Service of the Caliph’, p. 55. 282 Debié, ‘Christians in the Service of the Caliph’, p. 63.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

enjoyed to the head of the province’s fiscal administration was reshaped into a very different relationship: the head of the dīwān was now an outsider to their religious community although, in most cases, he remained a local. It has been made clear by several scholars mentioned at the beginning of this article that we should focus on the change of personnel perhaps even more than on the change of language. In all the translation narratives, a non-Muslim secretary is replaced by a mawlā of an Arab tribe — that is, by someone who had converted or at least joined an Arab tribe and was using an Arab/Muslim name.283 According to Elizabeth Urban’s investigation of al-Jahshiyārī’s text, the number of secretaries who were mawlās reached a peak in the early Marwanid period in Syria and Iraq, only to decline by 750.284 As sources composed in the late first/seventh and second/eighth centuries make abundantly clear, assimilation and conversion was one of the issues of the time for Christian authorities.285 The mawlās who replaced the likes of Sarjūn b. Manṣūr, Zādhān Farrūkh, and Athanasius bar Gūmōyē were the sons or grandsons of conquered people or prisoners of war who had assimilated and/or converted. The inability or unwillingness of Christian elites to fit that profile goes some way towards explaining why the change of language of the dīwān is presented in such a negative light in some of the Christian sources examined above. The translation of the registers and the change in the profile of the head of the dīwān did not prevent non-Muslims from working in the Umayyad administration, whether in the provincial centre or anywhere else in the provinces. As mentioned above, one of the grievances against the governor Khālid b. ʿAbd Allāh al-Qasrī in Iraq was his appointment of Zoroastrians in key offices in the provinces, including Dādūye al-Muqaffaʿ.286 Dādūye was tortured and became ‘the cripple’ (al-muqaffaʿ) after being accused of embezzling tax money. His access to wealth eventually proved problematic. His son Ibn al-Muqaffaʿ, as a convert, fitted the profile of the new administrative elite. He is also known to have accumulated great wealth in the course of his administrative career, which undoubtedly allowed him to finance his translation activities.287 There was only one position that was no longer accessible to non-Muslims in the Umayyad empire after the turn of the second/eighth century: that of the head of the dīwān, the highest position in the fiscal and military administration of the provinces of Iraq, Syria, and Egypt. However, the same process did not take place everywhere. The last section delves into the various regions of the Umayyad empire for which no translation narrative is attested and attempts to explain why this is the case.

283 Sijpesteijn, ‘Loyal and Knowledgeable Supporters’. 284 Urban, Conquered Populations in Early Islam, pp. 154–55. However, Urban’s statistics are based on the officials mentioned in al-Jahshiyārī’s text, which focused only on the leading post in the administration. 285 Penn, Envisioning Islam, pp. 142–86. 286 See above, pp. 104–05. 287 Arjomand, ‘ʿAbd Allah Ibn al-Muqaffaʿ and the ʿAbbasid Revolution’, pp. 17–25.

1 39

140

m a r i e l ege n d r e

Mapping the Translation From Ibn Khaldūn’s point of view at the turn of the ninth/fifteenth century, the only translations that were worth mentioning in his account of the dīwān were those of Syria and Iraq.288 He seems to have been particularly attached to ʿAbd al-Malik’s role in the translation, as he is presented as the overarching authority over the process, a depiction that does not agree with earlier translation narratives. Startlingly, Ibn Khaldūn does not refer to the Egyptian translation, even though he spent much of his career in Egypt, nor does he mention anything of note about the early Islamic dīwān of North Africa, his region of origin. After quoting al-Madāʾinī’s translation narratives for Syria and Iraq, he continues with remarks on how the position of the head of the dīwān evolved under the Umayyads of al-Andalus and the Almohads.289 The memory of the translation that he transmits is a caliphal one, and one that is attached to the central regions of the empire. Coming back to the second/eighth century, we find that the regional narratives examined above do not match the map of the whole Umayyad empire. Mapping out the narratives is useful, as it reveals where the translation is said to have taken place but also where it is not. As stated above, there are only four regional narratives: that of Syria in 81/700–01, that of Iraq c. 82/701–02 (dates explicit or deduced from al-Madāʾinī/al-Balādhurī), that of Egypt in 87/705 (al-Kindī), and that of Khurasan, possibly in 124/741–42 (al-Jahshiyārī). For Iraq, though al-Jahshiyārī and al-Ṣūlī clearly refer to two dīwāns (one in Basra and one in Kufa), there is only one narrative. As seen above, the Basran secretarial milieu is the best documented, so it is likely that the translation narrative refers to that city. The length and quality of al-Balādhurī’s informants on Basra also allowed al-Qāḍī to make her compelling demonstration of the translation of the name of estates in the local registers.290 It is notable that there are no translation narratives for the various Iranian regions, the Umayyad North (Armenia and Caucasian Albania), the Jazīra, North Africa, or al-Andalus. The Umayyad North and the Jazīra were integrated into the Umayyad realm in the early Marwanid period, though incursions took place already earlier.291 Based on the copper coinage, both the Jazīra and Armenia appear in the evidence at the time of al-Walīd.292 Only a small part of North Africa had been conquered by the turn of the second/eighth century; the conquest continued in the following decades, reaching the Iberian Peninsula in 711.293 Since al-Jahshiyārī’s Khurasani narrative allow to posit that the translation of local dīwāns could have taken place

288 Ibn Khaldūn, Taʾrīkh, i, p. 303. 289 Ibn Khaldūn, Taʾrīkh, i, pp. 304–05. 290 Al-Balādhurī section on Basra ‘stands out as being the longest and most detailed’ according to al-Qāḍī, ‘The Names of Estates in State Registers’, pp. 257–58. 291 See the relevant sections below for references. 292 Bone, ‘The Administration of Umayyad Syria’, p. 310. 293 On the conquest of North Africa, see Fenwick, Early Islamic North Africa, pp. 36–43; Fenwick, ‘The Umayyads and North Africa’, pp. 294–96. On that of al-Andalus, see Chalmeta, ‘Conquista y sumisión de Hispania’, pp. 19–35.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

as late as the 740s, translation narratives about regions of the Umayyad realm conquered after the initial translation in Syria, Iraq, and Egypt could potentially exist, but this does not seem to be the case. The other region notably absent is Arabia, which otherwise holds centre stage in narratives about the foundation of the dīwān al-jund by ʿUmar b. al-Khaṭṭāb.294 We also have to recognize that the phenomenon studied here — the establishment of administrative registers — is poorly documented in all these regions, and the conclusions I draw here can only be preliminary. Iran

In the case of the various Iranian regions, it is sometimes assumed that since they were dependencies of Iraq, a complete language shift took place there, too, at the turn of the second/eighth century after the translation of the dīwān of Iraq.295 However, the narratives are clear that each translation took place in the provincial capitals, and there is no mention of ramifications on the province’s dependencies. The exclusion of Iran from the process can also be inferred from the organization of al-Balādhurī’s Futūḥ al-buldān, where the account of the translation of the Iraqi dīwān is placed right after the description of the region’s conquest and before the section on the conquest of al-Jibāl, indicating that the translation did not apply to the latter.296 Therefore, there is no reason to believe that anything changed in the languages used in the administration of the Iranian regions around 82/701–02, as the source material is overwhelmingly silent on that matter. This could actually be revealing of the reach of the early caliphate east of Iraq. With regard to the conquest of the area, Milka Levy-Rubin singles out the ‘vassal treaties’ drawn up in various Iranian regions and contrasts them with the ‘surrender treaties’ drawn in Mesopotamia, Syria, and Egypt. She adds that ‘this model acknowledges the continuity of the local leadership under Muslim sovereignty’.297 As for the Umayyad period, Luke Treadwell argues that the copper coinage of Iran, which was produced mainly in Fars and Khuzistan, reflects ‘the relatively tenuous grip which the early Marwānids were able to exert upon the cities of the Iranian plateau’ until the turn of the second Islamic century.298 Similarly, Peter Verkinderen has shown that among the dependencies of Basra, Fars was a contested region since its ‘conquest’ and remained so until the end of the Umayyad period. Local rulers

294 See above, pp. 114–24. 295 This is assuming that a language reform took place under ʿAbd al-Malik. See Spuler, Iran in the Early Islamic Period, pp. 232–33, 352; on p. 232, ‘Until 697 Persian was the language employed for documents and charters in the west and Arabic was only introduced here in connection with the change in all chancelleries (naql al-dīwān) throughout the Empire under the Caliph ʿAbd al-Malik (685–705)’. See also Gyselen, Arab-Sasanian Copper Coinage, p. 101. 296 Al-Balādhurī, Futūḥ al-buldān, pp. 242–302. 297 Levy-Rubin, Non-Muslims in the Early Islamic Empire, pp. 45–50. 298 Treadwell, ‘The Copper Coinage of Umayyad Iran’, p. 373.

141

142

m a r i e l ege n d r e

are even remembered as ‘governors’, which points to a power balance that was very different from the situation in Syria, Iraq, and Egypt.299 As for Isfahan, Ibn Rusta’s al-Aʿlāq al-nafīsa (wr. around 300/912) includes a very short but revealing account on the adoption of Arabic in the city. In a list of awāʾil, he states: ‫واول من كتب بالعربيّة في ديوان اصبهان سعد بن ‌إياس كاتب عاصم بن يونس عامل ابي مسلم‬ ّ ‫صاحب الدولة‬

The first who wrote in Arabic in the dīwān of Isfahan was Saʿd b. Iyās, the secretary of ʿĀṣim b. Yūnus, the appointee of Abū Muslim, ṣāḥib al-dawla.300 Saʿd b. Iyās is otherwise unknown.301 However, his master’s connection to Isfahan is well attested: he is ʿĀṣim b. Yūnus al-ʿIjlī, a mawlā of the Banū ʿIjl, who were known to have estates and trading connections in the city.302 On the basis of the Akhbār al-dawla al-ʿabbāsiyya, Saleh Said Agha presents ʿĀṣim as an early follower of Abū Muslim in the ‘clandestine stage’ of the Abbasid daʿwa.303 Ibn Rusta’s reference to Isfahan can possibly be dated through the mention of Abū Muslim to between 129/747 and Abū Muslim’s death date in 137/755, though ʿĀṣim could have remained in Isfahan after Abū Muslim’s death. Read parallel to al-Jahshiyārī’s translation narrative on Khurasan, Ibn Rusta’s note allows us to connect the memory of the introduction of Arabic in some secretarial practices east of Iraq between the years leading to the Third Fitna and the Abbasid takeover.304 The Umayyad North

As shown by Alison Vacca, the Umayyad North (Armenia and Caucasian Albania) was part of the Iranian oikoumene in the early Islamic period, in line with dynamics established in the Sasanian period.305 Vacca and Michael Bates both argue that an Umayyad governorship connected to the cities of Dabīl/Diun in 299 Gundelfinger and Verkinderen, ‘The Governors of al-Shām and Fārs in the Early Islamic Empire’, pp. 274–75, 276 n. 97, 319, 322, referring to ‘Dādhbeh al-Muqaffaʿ, Farrūkhzād Gushn-Anūshān, and Khālid al-Qasrī’s unnamed ʿāmil of Dārābjird, a dihqān’. 300 Ibn Rusta, Kitāb al-Aʿlāq al-nafīsa, pp. 196–97. 301 All occurrences of this name refer to a different Saʿd b. Iyās, such as Abū ʿAmr al-Shaybānī, a prophetic companion who died at Qādisiyya; see Ibn Saʿd (d. 230/844), Kitāb al-Ṭabaqāt al-kubrā, vi, p. 104. 302 Agha, The Revolution which Toppled the Umayyads, p. 59. 303 Agha, The Revolution which Toppled the Umayyads, pp. 61, 86 n. 78, 328; Akhbār al-dawla al-ʿabbāsiyya, pp. 255, 259. 304 A context in which Persian was used according to al-Ṭabarī, Taʾrīkh al-rusul wa-l-mulūk, vii, p. 441: the phrase ‘Strike them, O youths’ (‫ )دهيد يا‌جوانكثان‬is said to have motivated the armies of Khurasan confronting the forces of the last Umayyad caliph Marwān b. Muḥammad. My thanks to Jaakko Hämeen-Anttila for his assistance with the Persian words in this quotation. See also Arjomand, ‘ʿAbd Allah Ibn al-Muqaffaʿ and the ʿAbbasid Revolution’, pp. 10–11; Lindstedt, ‘The Transmission of al-Madāʾinī’s Historical Material to al-Balādhurī and al-Ṭabarī’, p. 49. 305 Vacca, Non-Muslim Provinces under Early Islam, pp. 11–14, 101–11.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

Armenia and Bardhaʿa/Partaw in Caucasian Albania was established in the early Marwanid period.306 Prior to this period, various Armenian or Albanian cavalries are said to have contributed to the defence of the frontiers, starting in the time of Muʿāwiya b. Abī Sufyān’s governorship of Syria (in office 18–41/639–61).307 Even with the appointment of governors in the region, however, it remains uncertain whether payments to the cavalries were recorded in a military dīwān in the early Marwanid period. Before that time, they were paid for service in lieu of taxation, and according to Ibn Aʿtham al-Kūfī (d. mid-third/ninth century or 314/926), such an arrangement was still expected by the Albanian cavalry in the late second/ eighth century.308 The same uncertainty is found in the various accounts of the fires of Naxčawan: the payment of a stipend from the ‘royal record’ is presented as a promise made by Muḥammad b. Marwān or his appointee to lure members of the Armenian nobility into a church that was then set on fire.309 In Armenian and Arabic narratives, this promise is portrayed as attractive enough for the nobles to enter the church, so it does not seem to have been something that they already had access to.310 According to Vacca and Sergio La Porta, these episodes ‘demonstrate the unease of the Armenian nobility with regard to increased caliphal control over Armenia’ at the beginning of the Marwanid period.311 Several other texts do draw a connection between the Marwanids and the establishment of dīwāns. The early fourth/tenth-century Armenian text of the History of Ałuank‘/Albania transmits the memory of an association between ʿAbd al-Malik and registers that is harnessed to the service of local theo­logical disputes.312 It states that the caliph had established a dīwān to keep track of Dyophysites, reportedly in support of Miaphysite uniformity in the region.313 In this text, the

306 Bates, ‘The Dirham Mint of the Northern Provinces of the Umayyad Caliphate’; Vacca, Non-Muslim Provinces under Early Islam, pp. 69–76. 307 Vacca, Non-Muslim Provinces under Early Islam, pp. 187–88. 308 Ibn Aʿtham, Kitāb al-Futūḥ, viii, p. 257; quoted in Vacca, Non-Muslim Provinces under Early Islam, pp. 202–03. 309 The various Armenian, Syriac, Greek, and Arabic texts referring to that episode have been studied in Vacca, ‘The Fires of Naxčawan’. The reference to the ‘royal record’ is from Łewond, whose account is translated on p. 327. 310 Łewond calls it a ‘pretext’, T‘ovma Arcruni ‘a deceitful trick’, and Yovhannēs Drasxanakertc‘i ‘deceit, fraud, [and] vain hopes’; all of these texts are translated in Vacca, ‘The Fires of Naxčawan’, pp. 327–30. Greek and Syriac texts do not refer to military registers or deceit in any way and present the burning of the Armenian nobles in their church as a straightforwardly barbarous act; ibid., pp. 331–34. Al-Balādhurī and al-Yaʿqūbī do not refer to the dīwān but report that the promise was that the nobles would be treated according to their rank (al-sharaf): al-Yaʿqūbī, Taʾrīkh, ii, pp. 324–25; al-Balādhurī, Futūḥ al-buldān, p. 205. 311 Łewond, History, section V.C. 312 A text possibly attributed to Movsēs Daskhurants‘i or Movsēs Kaghankatuats‘i; see Greenwood, ‘Movses Daskhurants‘i/Movses Kaghankatuats‘i’. 313 ‘All these names were written in the archives [‫ديوان‬, rendered in Armenian as դիւան] of Abdlmēlik‘ Amir Mumin in order that if any of them were found to have become duophysites, they might be destroyed by the sword or imprisonment. Thus was peace achieved in all the churches of Albania’; quoted in Vacca, Non-Muslim Provinces under Early Islam, p. 175. The same text also has ʿAbd al-Malik

143

144

m a r i e l ege n d r e

caliphal reference is meant to convey authoritative validation of the Miaphysite cause.314 As shown by Vacca in this volume, in the second/eighth-century vita of Vahan Gołt‘nec‘i, martyred in 737, the dīwān of Syria is clearly associated with caliphal power.315 In this narrative, Vahan’s role in the dīwān (here meaning the administration) speaks to his skills and his ability to thrive as a member of the secretarial elite of the caliphate, transcending his entry to the Umayyad court as a child and prisoner in the aftermath of the fires of Naxčawan.316 Finally, in Armenia, Łewond’s History (late second/eighth century) makes indirect reference to new fiscal registers being drawn up through a survey (taʿdīl) said to have taken place at the time of the caliph Hishām.317 The Jazīra

Similarly to the Umayyad North, the Jazīra was, according to Chase Robinson, a creation of the Marwanid era.318 Analogous information — partial data about local military registers and the drawing up of new fiscal registers — is available for that region. Accounts of the Kharijite rebellion of Ṣāliḥ b. Musarriḥ and Shabīb b. Yazīd al-Shaybānī (76/695) indirectly document aspects of the military dīwān.319 According to al-Balādhurī, Shabīb revolted either because his name was suppressed from the dīwān, understood here as the register of stipends, or because the Marwanids did not provide stipends to the tribes of Tamīm and Bakr in Syria.320 This latter opinion on the authority of al-Haytham b.  ʿAdī (d. 207/822) agrees with the hypothesis proposed above that the Marwanids provided orderly payments for tribes in their new military dīwān of the Arabs, though they may here have excluded some tribes from those payments.321 In both taking measures against the Chaledonian patriarch of Albania; see ibid., pp. 174–76. դիւան is here not necessarily an Arabic loanword but most probably a Persian one; see above, at n. 151. On the use of Persian sources in the History of Ałuank‘, see Vacca, ‘The Umayyad North’, p. 232. 314 On this phenomenon more broadly in literature referring to the Umayyad North, see Vacca, ‘The Umayyad North’, pp. 232–34. 315 See the contribution of Alison Vacca to this volume. 316 Vacca, ‘The Fires of Naxčawan’, pp. 329–31. 317 See Vacca, Non-Muslim Provinces under Early Islam, pp. 198–200, on fiscal measures in general, and p. 200 on the taʿdīl (p. 200), where Vacca quotes her forthcoming translation of Łewond with Sergio La Porta: ‘Šam, or Hishām b. ʿAbd al-Malik, sent Hert‘, Ḥārith b. ʿAmr b. Ḥaraja l-Ṭāʾī, to the North: ‘to conduct a census of the land of the Armenians in order to intensify the iron yoke of servitude of taxation by means of manifold evils because he was vexed by the goodness of Omar [the caliph ʿUmar b. ʿAbd al-ʿAzīz], as if he had unjustly spent the stores of the treasury that the princes before him had amassed. And so much danger came upon our land until everyone sighed over the tumultuous tribulations, on account of which there was no place to live due to intolerable dangers’. On other regional censuses in the Marwanid period, see al-Qāḍī, ‘Population Census and Land Surveys under the Umayyads’, pp. 364–410. 318 Robinson, Empire and Elites after the Muslim Conquest, p. 44. 319 Robinson, Empire and Elites after the Muslim Conquest, pp. 121–26. 320 Al-Balādhurī, Ansāb al-ashrāf, viii, pp. 17–19. 321 Al-Balādhurī, Ansāb al-ashrāf, viii, p. 19: ‫كان بنو مروان ال يفرضون لبكر وال تميم بالشام‬.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

al-Balādhurī’s and Ibn Aʿtham al-Kūfī’s accounts of the matter, the stipends in question are those of al-Shām, not of al-Jazīra, pointing to the possibility that military pay for the jund al-Jazīra was decided further south. Further, the caliph ʿAbd al-Malik is presented as the authority on the issue.322 This organization can be explained by the history of the formation of the province of al-Jazīra, which is initially presented as an extension of the Syrian jund of Qinnasrīn.323 According to al-Balādhurī’s Futūḥ, ʿAbd al-Malik asked his brother Muḥammad b. Marwān to set up a new jund there relying on local taxation, though al-Balādhurī does not say exactly when this took place.324 Al-Balādhurī places his translation narrative of the dīwān of Syria after the account of the conquest of the Jazīra but before that of Armenia, further confirming that the dīwān of Syria covered the Jazīra but that the Armenian context was different. As for fiscal dīwāns, there are again similarities with the Umayyad North. The Chronicle of Zuqnin refers indirectly to new fiscal registers being drawn up after ʿAbd al-Malik orders a taʿdīl in the Jazīra. It is described as both a register of properties and wealth and a census that served as the basis for the levy of the poll tax, which the text describes as an innovation at the time.325 This taʿdīl was the first of most probably two (not four) such surveys in the Marwanid Jazīra.326 In the Chronicle

322 Al-Balādhurī, Ansāb al-ashrāf, viii, pp. 17–19; Ibn Aʿtham, Kitāb al-Futūḥ, vii, p. 84. 323 Robinson, Empire and Elites after the Muslim Conquest, p. 39. 324 Al-Balādhurī, Futūḥ al-buldān, p. 132: ‫عبْد الملك بن مروان أي أفردها‬ َ ‫وذكروا أن الجزيرة كانت إِلَى قنسرين فجندها‬ .‫ وأن مح ّمد بن مروان كان سأل عبد الملك تجنيدها ففعل‬،‫‌فصار‌جندها يأخذون أطماعهم بها من خراجها‬ 325 The Chronicle of Zuqnīn, trans. by Harrak, pp. 147–48: ‘691–692 The year one thousand and three: ʿAbd-al-Malik made a census (taʿdīl) among the Syrians. He issued a swift decree stating that every person must go to his country, village and paternal house to register his name and that of his father, as well as his vineyards, olive trees, cattle, children and all that he owned. From this time, the poll-tax began to be levied on the male heads and all the calamities began to emerge against the Christian people. Previously, kings used to levy tribute on land, not on men. From this time onward the Sons of Hagar began to reduce the Sons of Aram to Egyptian slavery. But woe unto us! Because we sinned, the slaves ruled over us! This was the first census (taʿdīl) the Arabs had made’. According to Robinson, the term ‘Syrians’ here refers to the ‘Christian inhabitants of the North’; Robinson, Empire and Elites after the Muslim Conquest, p. 45. 326 Robinson, Empire and Elites after the Muslim Conquest, pp. 49–50: ‘The chronicler from Zuqnin records taʿdīls in AG 1003 (691–692) and again in 1020 (708–709); the second of these is said to have been a follow-up to the first. This first taʿdīl is also known to the Edessan and Islamic traditions. In Abū Yūsuf it is credited to al-Ḍaḥḥāk b. ʿAbd al-Raḥmān, a point apparently corroborated by the Chronicle of 846, which, along with the Chronicle of 819, mentions taʿdīls in 1008 (696–697) and 1022 (710–711); the account dated to 1020 (708–709) by the Chronicle of 846, which may mark the beginning of the taʿdīl of 1022, was undertaken by Maslama b. ʿAbd al-Malik (here called the governor of the whole of the Jazira — amīrā d-kullāh gāzartā): “Maslama sent officers throughout northern Mesopotamia in order to measure lands, make a census of vineyards, plantations, livestock and people, and to hang lead seals on everyone’s neck”’. All of these sources might actually be referring to only two taʿdīls, with slightly different dating — the first one in the 690s and the second one around 710. If we follow the dating of the Chronicles of 846 and 819, also confirmed by Abū Yusūf, we find, as in al-Andalus, a fifteen-year gap between the two taʿdīls or population surveys; see below, at n. 343. This is pace al-Qāḍī, ‘Population Census and Land Surveys under the Umayyads’, pp. 365–72, 379–81.

145

146

m a r i e l ege n d r e

of Zuqnin and the Chronicle of 754, these Marwanid fiscal measures are described with similar disdain to that displayed in Łewond’s History in the case of Armenia.327 North Africa

On the other side of the empire, according to Hichem Djaït, the area west of Egypt was divided into ‘Ifriqiya proper (the old Byzantine provinces of Proconsularis and Byzacena), Tripolitania, the Zab (Numsidia), Near Sus and Far Sus’, although he admits that those divisions have been reconstructed on the basis of very confusing source material.328 This was an enormous territory, and we may doubt that it was ever fully integrated into a single province in the few turbulent decades during which the Umayyads attempted to control it.329 The rather meagre information available on Umayyad administration in North Africa is connected to Qayrawān in Ifrīqiya proper. Djaït offers a reconstruction of its dīwān al-jund on the basis of information on Iraq, Syria, and Egypt, as he has nothing to work with on North Africa before the rise of the Aghlabids (184–290/800–909).330 As for fiscal administration the picture is no easier to reconstruct.331 There was no unified pre-Islamic administration over this gigantic region that the conquerors could have taken over, as they did in Egypt, Syria, and Iraq. As Corisande Fenwick has 327 Vacca, Non-Muslim Provinces under Early Islam, pp. 196, 205. 328 The quotation is from Fenwick, Early Islamic North Africa, p. 43, translating Djaït, ‘La Wilāya d’Ifrīqiya au iie/viiie siècle (suite et fin)’, p. 93: ‘Le problème de l’organisation provinciale du Maġrib et de l’Ifrīqiya n’en demeure pas moins extrêmement confus’. 329 Fenwick, Early Islamic North Africa, p. 33: ‘The new Muslim province encompassed — in theory — over five times the amount of territory of Byzantine Africa and twice that of Rome at its height. Both Mediterranean and Saharan, this immense region was inhabited by a range of different peoples, settled and mobile, some of whom had never been ruled by Rome or Byzantium. On scale alone, Byzantine Africa and the far bigger Muslim Africa are simply incommensurable. The caliphate could not rule Africa as the Romans, Vandals or Byzantines had before them: they had to connect regions and peoples that had never in their history been united under one power — and never would be again’. 330 Djaït, ‘La Wilāya d’Ifrīqiya au iie/viiie siècle I’, pp. 113–14. In ‘La Wilāya d’Ifrīqiya au iie/viiie siècle (suite et fin)’, p. 80, Djaït states that ‘toute l’histoire militaire de l’Ifrīqiya des wulāt serait incompréhensible si on ne postulait pas l’existence d’un Dīwān al-Ǧund’; however, this is a circular argument, as the only available sources for the history of this period base their whole narratives on the assumption that there was indeed a dīwān al-jund. See also p. 93 on the applicability of the material from the Aghlabid period: ‘En toute rigueur historique, ce qui est valable du temps des Aġlabites ne l’est pas forcément du temps des Wulāt. Admettons, comme nous l’avons constamment fait au cours de cette étude, que ce qui est aġlabite peut être considéré comme un héritage de l’époque précédente, mais alors il s’agirait d’une identification avec la situation à la fin de la période des Wulāt’. 331 As Djaït himself admits in ‘La Wilāya d’Ifrīqiya au iie/viiie siècle (suite et fin)’, pp. 82, 90, ‘les textes dont nous disposons sont rares, avares d’information, et quand il leur arrive de nous informer, confus et obscurs’. He offers (at pp. 84, 91) the date of ah 110 for the completion of the Arabization of the various dīwāns, following the measures of Ismāʿīl b. Abī Muḥāgir, but he acknowledges that ‘nous ne pouvons que nous baser sur l’intuition pour cela’. The same point is made in Nabia Abbott’s work on the career of ʿUbayd Allāh b. al-Ḥabḥāb (governor of Ifrīqiya 116–23/734–41) in ‘A New Papyrus and a Review of the Administration of ʿUbaid Allāh b. al-Ḥabḥāb’, p. 33: ‘No adequate account exists of the ups and downs of Muslim taxation in North Africa’.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

argued, even in some of the individual regions, such as Far Sūs, there were no known pre-existing taxation systems to takeover. If the Umayyad administration developed taxation registers there from scratch it would be unparalleled and it remains unlikely. The main economic yield of this region appears to have come from the slave trade.332 Most information, again, pertains to Qayrawān. Ibn ʿAbd al-Ḥakam describes the re-foundation of the miṣr around 74/693–94 and reports that the Syrian Ḥassān b. Nuʿmān al-Ghassānī commanded the construction of the mosque, the drawing up of the dīwān, and the establishment of the taxation system.333 The phrase used, ‘he instituted the registers’ (‫دون الدواوين‬ ّ ), is a famous wording echoing what the Caliph ʿUmar b. al-Khaṭṭāb supposedly had done, meaning the report is not specific to Qayrawān.334 It bears a structure common to accounts about the establishment of a miṣr and the accompanying building of monuments and establishment of administrative structures. It also simplifies the process of setting up a dīwān, which, as shown above, was a highly complex affair. It is clear that Ḥassān b. Nuʿmān did not build the mosque himself, nor did he compile the registers himself. The details and the agents involved in these events are beyond the horizon of Ibn ʿAbd al-Ḥakam’s text. The only information that resonates with the other texts discussed above on Egypt and Syria is that taxation is said to have been imposed on ‘the non-Muslims of Ifrīqiya’ (‫)عجم ‌إفريقية‬. In light of the significant Berber element in the North African and Andalusi armies of the Umayyads, one can only wonder whether this term ʿajam was used in North Africa, as it was in Egypt and Syria, to refer to the group that the fiscal administrators focused on.335 The striking difference between accounts concerning North Africa and narratives on Iraq, Syria, and Egypt is the lack of information on the former’s pre-Islamic administration and on administrators of local origin.336 Their inclusion in governance in North Africa must have been as essential as it was elsewhere. The Ifriqīyan context was also that of a new and unknown linguistic sphere, that of Latin; that context gave birth to innovations such as the translation of the shahāda into Latin on North African coins, a phenomenon known only from one earlier coin minted in Sistān with a Middle Persian translation.337 Most historians who have worked on Umayyad Ifrīqiya have noted the ‘delay’ of nearly forty years in the implementation

332 Fenwick, Early Islamic North Africa, p. 36. 333 Ibn ʿAbd al-Ḥakam, Futūḥ Miṣr, p. 201 : ‫ودون الدواوين ووضع الخراج على‌عجم‌إفريقية‬ ّ ‫وبنى مسجد جماعتها‬ ‫وعلى من أقام معهم على النصرانيّة من البربر وعا ّمتهم من البرانس ّإل قليال من البتر‬. On the initial foundation of Qayrawān in 47/667–68 by ʿUqba b. Nāfiʿ, see Fenwick, Early Islamic North Africa, p. 37. Ibn ʿAbd al-Ḥakam is the earliest Islamic source describing in some detail the conquest of North Africa. After him, similar accounts were compiled much later by Ibn ʿIdhārī (d. after 710/1310) and Ibn Khaldūn; both are used by Djaït. 334 Noth and Conrad, The Early Arabic Historical Tradition, p. 35. 335 Rouighi, Inventing the Berbers, pp. 23–33. On Berber troops in al-Andalus, see Manzano Moreno, ‘Conquest and Settlement’, pp. 317–21. 336 Djaït, in ‘La Wilāya d’Ifrīqiya au iie/viiie siècle (suite et fin)’, pp. 88–89, tries to draw conclusions from the names of two possible secretaries. 337 Djaït, ‘La Wilāya d’Ifrīqiya au iie/viiie siècle (suite et fin)’, p. 88; Jonson, ‘A Numismatic History of the Early Islamic Precious Metal Coinage of North Africa and the Iberian Peninsula’, pp. 35–36,

147

1 48

m a r i e l ege n d r e

of ʿAbd al-Malik’s coinage reform in the region, something that also applies to al-Andalus.338 There is no reason to believe that other aspects of administration and governance, such as the translation or compilation of central fiscal or military registers in Arabic, would have been implemented from the very beginning.339 Al-Andalus

Similarly to Ibn Khaldūn, Ibn ʿAbd Rabbihi does not have anything to say about the implementation of the dīwān in his region of origin, in his case al-Andalus. The conquering armies settled throughout the Iberian Peninsula, not in amṣār; settlers collected revenues directly by owning land and did not rely on ʿaṭāʾ payments. This suggests there was no need initially for a military dīwān.340 According to reports transmitted by Ibn al-Khaṭīb (d. 776/1375) on the authority of al-Rāzī (d. 344/955), the implementation of the Syrian jund that came to the region in the aftermath of the so-called Berber revolt of 122–25/740–43 prompted the establishment of a dīwān for military payments.341 However, Eduardo Manzano Moreno points out that it is unclear to which period of Andalusi history this account refers, as the Syrian jund remained the backbone of the Umayyad emirate’s military (138–316/756–928).342 As for fiscal administration in al-Andalus, the Chronicle of 754 reports three censuses, about fifteen years apart, which were undoubtedly focused on the assessment of the poll tax, as in the Jazīra.343 Manzano Moreno interprets the description of the first census, which was started by the governor al-Samḥ b. Mālik al-Khawlānī (in office 100–02/718–21) and redacted propio stilo, as evidence that it was written

92–103. These coins date from 89–90/708–09. The earlier Middle Persian shahāda is dated to 72/691–92 according to Mochiri, ‘A Pahlavi Forerunner of the Umayyad Reformed Coinage’. 338 Djaït, ‘La Wilāya d’Ifrīqiya au iie/viiie siècle (suite et fin)’, p. 79; Fenina, ‘L’arabisation du monnayage d’Ifrīqiya’, pp. 148–62; Fenwick, ‘The Umayyads and North Africa’, p. 299; Jonson, ‘A Numismatic History of the Early Islamic Precious Metal Coinage of North Africa and the Iberian Peninsula’, pp. 407–16. 339 The seals attested in al-Andalus are not an exception, since they relate only to the conquest and collection of booty, not to everyday administration; they were produced for internal use among the conquerors. See Sijpesteijn, ‘Expressing New Rule’, p. 136. 340 Kennedy, Muslim Spain and Portugal, p. 16. However, copper coinage offers ample evidence of organized payments to the army; see Manzano Moreno, ‘Conquest and Settlement’, pp. 324–27. 341 Al-Rāzī’s account is given in Ibn al-Khatīb, al-Iḥāṭa fī akhbār Gharnāṭa, i, pp. 104–05. Kennedy, in Muslim Spain and Portugal, p. 50, writes: ‘The Syrian jundis, whose names were recorded in a diwan, were divided into two groups, one which went on campaign and the other which stayed at home. After three months the two groups would change places. Those on campaign were paid a rizq of 5 dinars at the end of the campaign while their chiefs received 200. The baladis had no diwan and only their leaders were paid 100 dinars. The Syrians were exempt from the ʿushr (tithe) and only had to pay a share of the revenues they collected from the non-Muslims; the baladis, by contrast, were obliged to pay this tithe’. 342 Manzano Moreno, ‘El asentamiento y la organización de los yund-s sirios en al-Andalus’, pp. 348–49. 343 For the second census of 737, the Chronicle explicitly refers to descriptio populi. See Continuatio Isidoriana hispana, pp. 262–63; Manzano Moreno, Conquistadores, emires y califas, p. 78.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

in Arabic.344 Most interestingly, he argues that the regularity of the census was not accidental but rather mirrored the fifteen-year Byzantine fiscal cycle. He contends that the Umayyad authorities of al-Andalus imported this logic of updating fiscal data regularly, a step that is arguably also visible in the Jazīra.345 The account of the third census explicitly mentions registers (publico codice) that needed to be updated as the census was made.346 For most of the regions mentioned in this section, there is evidence for the drawing up or updating of some fiscal registers in the Marwanid period, but the evidence for local military registers is less convincing. The most robust evidence is for population censuses or surveys completed in the early second/eighth century, as they are mentioned both in Andalusi Latin chronicles and in literature concerning the Jazīra and the Umayyad North.347 But across these regions, there is little or no information on how such registers were compiled, who was in charge of them, where they were kept, and — most importantly for our purposes — in which language they were written. Local secretarial practices must have influenced the compilation of registers in the regions examined in this section as much as they did in Syria, Iraq, and Egypt. There is no reason to believe that the translation of registers in the central Umayyad provinces of Syria, Egypt, and Iraq implied uniformity of administration in the regions conquered after the translation was started. I do not mean that Arabic was absent from these regions. There was evident public use of Arabic on coins, seals or inscriptions, however they bear no connection to the drawing up of registers for fiscal or military administration in Arabic which is the focus here. One reason the compilation of such registers might

344 Conquerors and Chroniclers of Early Medi­eval Spain, trans. by Wolf, p. 112: ‘In the western regions the Arabs achieved many military victories under their leader, as-Samh. Having held power in Spain for a little less than three years, he undertook, on his own initiative, a census of Hispania Ulterior and Hispania Citerior. He divided by lot among his allies the booty, arms, and whatever else in the way of plunder the Arab people in Spain had not yet divided, and added a portion of all the moveable and immoveable goods to the fisc’. See also Manzano Moreno, Conquistadores, emires y califas, p. 73 n. 47 (p. 514). Propio stilo is translated as ‘on his own initiative’ in Conquerors and Chroniclers of Early Medi­eval Spain, trans. by Wolf, p. 112 and as ‘con su propria pluma’ in Continuatio Isidoriana hispana, pp. 242–43. 345 Manzano Moreno, Conquistadores, emires y califas, p. 74. 346 Continuatio Isidoriana hispana, pp. 280–81; Conquerors and Chroniclers of Early Medi­eval Spain, trans. by Wolf, p. 125: ‘Yusuf ordered a census of the remaining people to be taken and, although it amounted to embezzlement, he solicitously ordered the scribes to erase from the public records the names of those Christian taxpayers who had been killed by the sword during the great slaughter’. As noted by al-Qāḍī, ‘Population Census and Land Surveys under the Umayyads’, p. 348, censuses were not necessarily at the origin of new registers as old ones could be updated. 347 As al-Qāḍī puts it in ‘Population Census and Land Surveys under the Umayyads’, pp. 364–65, the early Marwanid period (65–105/685–723) ‘is the most crucial period of Umayyad history for matters related to population census, land surveys, and beyond. The fact that it is the best documented, in both the literary Islamic and non-Islamic sources as well as in the papyri, could not be merely an accident of history’. She also concludes (p. 411) that surveys were mostly planned by governors in that period and that they were not a centralized enterprise.

149

150

m a r i e l ege n d r e

be untraceable is that the process was progressive and not associated with a one off event of translation of registers. Another reason registers might have functioned differently in the more far-flung regions is that the nature of Umayyad authority appears to have taken a different shape in those regions. Umayyad representatives did not necessarily manage to extract taxes everywhere, as shown for North Africa. The creation of the Marwanid province of the Jazīra merged territories that had, before the Islamic conquest, been on opposite sides of the Roman-Sasanian frontier. Similarly, Caucasian Albania and Armenia did not have a unified fiscal system prior to the conquest, nor did they immediately acquire one after it. This situation posed a challenge the Umayyads had not faced in Syria, Iraq, or Egypt, which were all unified fiscal spaces prior to the conquest. In Armenia, the Umayyads also had to deal with local princes such as Ašot Bagratuni (r. 726 or 732 to 748) and with cavalry-based armies that had to be incorporated into the caliphal enterprise through payments as in other frontier contexts.348 The power balance was thus very different from that in Syria, Iraq, and Egypt. The Umayyads never had to keep up local fighting contingents active after the conquest of those regions, nor to appoint or recognize local lords as they did in Armenia and in Fars. From the point of view of Umayyad history, these regions are actually the most revealing of the extent and variety of the ruling strategies the caliphate had to champion to establish itself in different contexts, at different times, and within different local dynamics.349 We can make one final observation about the absence of translation narratives for all these regions. As early Islamic literature is often deemed to reproduce topoi across regional narratives of the early empire, especially with reference to the conquest, the capture of cities, surrender agreements, and the establishment of the ruling elite, one might be tempted to think that the implementation of an administration in the language of the conquerors would be among such transferable templates.350 The fact that it is not the case actually gives credence to the idea that translation narratives were not successful topoi, that is, as posited above, apart from one possibility: al-Jahshiyārī’s report on Khurasan.

Conclusion This article has shown that in the first decades of the Marwanid period, the dīwāns of Iraq, Syria and Egypt were translated. The vocabulary used to describe this change in the various sources is very clear: there is no reference to a reform or to a decree, only to the translation of select registers. The situations in Syria and Iraq are the

348 On payments for cavalries in Armenia, Albania, Jurjan, and Tukharistan, see Legendre, ‘Aspects of Umayyad Administration’, pp. 139–40. 349 A point beautifully made by Fenwick (Early Islamic North Africa, esp. p. 293) and Vacca (Non-Muslim Provinces under Early Islam). 350 Noth and Conrad, The Early Arabic Historical Tradition.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

best documented. The translation in each of these regions was the result of a change of personnel at the head of the dīwān, the main dynamic behind shifting linguistic practices being the appointment of new elites. The likes of Zādhān Farrūkh, Sarjūn b. Manṣūr, and Athanasius bar Gūmōyē, who headed the provincial dīwāns before the translation, were certainly bilingual. However, mastery of Arabic did not amount to promotion of Arabic, and it was only with their replacement that the latter was achieved. With the translation, three non-Muslim high officials who had formed the backbone of the administration in the central provinces of the Umayyad realm were replaced, either because they died (Zādhān Farrūkh) or because they were dismissed (Athanasius), while the fate of Sarjūn b. Manṣūr remains uncertain. Through this change, non-Muslim elites lost access to a very lucrative position while mawlās took their place. However, the personnel change at the top did not curtail the employment of non-Muslims in the administration beyond the position of the head of the dīwān. The sons of Zādhān Farrūkh and Sarjūn b. Manṣūr, and the great-grandson of the former, continued to work in the Umayyad administration amongst other non-Muslim secretaries, but in other positions. The investigation of the various translation narratives provides a typical case study in early Islamic historio­graphy. The earliest authors (Khalīfa b. Khayyāṭ, al-Baghdādī) worked with lists of officials, while later authors offer fully formed narratives. Later sources appear better informed than earlier ones and the chrono­logy of events is challenging to reconstruct.351 The most likely chrono­logy for the translations is the following: 81/700–01 for Syria, around 82/701–02 for Iraq, 87/705 for Egypt, and, possibly, 124/741–42 for Khurasan. Reconstructed dialogues and anecdotes of administrators urinating in inkwells set aside, a few stable features in these reports inspire confidence that they indeed reflect aspects of the Umayyad context: the focus on the translation of administrative documents in provincial centres, the role of secretaries, the absence of an overarching authority across provinces, and the prominence of regional dynamics. The core information in the earliest translation narratives fits well with the image reconstructed from contemporary documents, which show the continuity of multilingual practices in the state. The dynamics of change are specific to each of the provinces: this applies to the chrono­logy of the translations as well as to the shape of the documentary record, as seen in the example of numerical notations. Various authorities were involved in the translation of the dīwān, depending on the region — the caliph in Syria, governors in Egypt and Iraq — but in all contexts, secretaries are the key figures and sometimes the initiators of the translation. The milieu of Basran secretaries appears as the best documented in the long term through Ṣāliḥ b.  ʿAbd al-Raḥmān and his student Qaḥdham but also Dādūye al-Muqaffaʿ and his son Ibn al-Muqaffaʿ. The careers of these administrators in addition to Athanasius bar Gūmōyē, Wardān the mawlā of ʿAmr b. al-ʿĀṣ, Jabala b. Sālim, Sālim Abū al-ʿAlāʾ, and ʿAbd al-Ḥamīd al-Kātib

351 The same point is made by Andreas Görke for the prophetic sīra in, for instance, ‘Prospects and Limits in the Study of the Historical Muhammad’ and ‘The Historical Tradition about al-Ḥudaybiya’, pp. 40–48.

15 1

152

m a r i e l ege n d r e

and their interconnections should be the focus of further research. Altogether, this study reveals that the training provided to the secretaries involved in the translation of the dīwān allowed some of them to contribute to or even set off the translation movement. Multilingual secretaries benefiting from highly lucrative positions in the administration could invest their skills and money in projects outside of the administrative realm. In her study of ʿAbd al-Ḥamīd al-Kātib’s letter to the secretaries, al-Qāḍī comments that ‘the secretaries cannot take credit for their work, and the world will always be under the illusion that kings manage their affairs, whereas it is the secretaries who accomplish that behind the scenes on their behalf ’.352 It is hoped that the present research has revealed the many ways in which the secretaries contributed to shaping state practices and discourses. Other secretarial contexts produced innovations in this period, as visible in documents in Greek and Coptic from Egypt. Although the central registers were now written in Arabic, the local ones in the kūras remained in Greek until the end of the period under study, if not beyond it. In these local multilingual chancelleries, Coptic subsequently became an administrative language essential to the functioning of the state until the end of the Umayyad period. New formularies and a new chancery style were developed for Greek and Coptic administrative documents in that same milieu. The papyro­logical documentation is the only corpus that reveals the Marwanid period as a key moment of innovation in the use of pre-Islamic administrative languages. The complex dynamics at play enabled this multilingual state to thrive. What we see in the Umayyad period, with the translation of the dīwān, on the one hand, and the introduction of Coptic as an administrative language, on the other, goes well beyond the idea that ‘Arabization’ aimed at the replacement of pre-Islamic languages was the main dynamic in the written culture of the state. The reason for the lack of uniformity in the evolution of linguistic practices across the Umayyad empire is that in each region or at each echelon of the administrative hierarchy different literate elites were in charge of drawing up documents. Each literate group had its own linguistic background and each introduced innovations of its own in response to opportunities created by the relevant authorities (duke, caliph, governor) in the period in which the Umayyads held the caliphate. Moreover, the use of Coptic in the administration did not stop with the Umayyads, and another necessary step forward is further study of the role of Coptic (and non-Arabic languages in general) in the Abbasid administrative apparatus.353 Much remains to be written about the dīwān. Attempting to ascertain what the term meant in the early Marwanid period, we find that a careful record of early administrative history was preserved for Syria and Egypt. The Egyptian material, 352 Al-Qāḍī, ‘Identity Formation of the Bureaucracy of the Early Islamic State’, p. 150. 353 This is one of the topics examined within the ERC project for which I am the principal investigator: ‘The Finances of the Caliphate: Abbasid Fiscal Practice in Islamic Late Antiquity (Caliphal Finances)’, grant agreement ID: 950414. On this topic, see also Berkes and Younes, ‘A Trilingual Scribe from Abbasid Egypt?’; Berkes and Vanthieghem, ‘A Late Coptic Tax Receipt from the Egyptian National Library’.

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

attributed to al-Layth b. Saʿd among other sources, allows us to discern the Umayyad termino­logy, which refers to the administration of the ʿarab and the ʿajam more than it does to a dīwān al-jund and a dīwān al-kharāj. This distinction appears to have been formalized right before the translation of the dīwān of Syria and after that of Egypt. In other parts of the empire, there are no identifiable translation narratives for the various Iranian regions, the Jazīra, the Umayyad North, North Africa, or al-Andalus. The languages used in the administrative registers of those provinces are mostly not acknowledged in the preserved sources. However, narratives on the compilation or update of registers through censuses or surveys in the early Marwanid period are available for most of these regions. Ultimately, the centre versus periphery narrative can be rejected in studies of the Umayyad state. The idea of a slow diffusion of Arabic from centre to periphery does not do justice to the multiple dynamics at play. Seeing linguistic practices as dictated by reforms is also a misunderstanding of the nature of Umayyad rule. The Umayyad caliphs did not have the will to impose uniformity of administrative languages from al-Andalus to Central Asia. By contrast and as Petra Sijpesteijn’s contribution to this volume makes clear, numerous administrative measures implemented by caliphs across provinces were successful. Regions conquered in the second/eighth century presented challenges different from those encountered in Egypt, Syria, and Iraq. In most of the newly conquered regions, there was no unified fiscal system for the Umayyads to take over, which explains why Umayyad fiscal administration and the corresponding bookkeeping likely took a distinctive shape. The Umayyads also had to deal with militarily competent local elites in some of these regions, which may explain why their military administration likewise took different forms. In light of Antoine Borrut’s contribution to this volume, we can conclude that the phenomena described here are revealing of the formation of the Arabic cosmopolis. Looking at the contribution of translations of administrative texts and multilingual secretaries in the long term allows us to track how it was eventually successful. Benoit Grévin has also invited us to reconsider the possibilities of multilingual practices in his comparative study of the medi­eval Middle East and the Latin West on the basis of language use. He describes Latin and Arabic as ‘langues référentielles’, each drawing prestige from religion and promoted by the ruling elite as part of its religiously inspired authority. These languages were also used as tools in the context of social and religious tensions. There were times when certain languages were promoted, but there were no convincingly established instances of the intentional suppression of others. As Grévin argues, linguistic homogeneity was not the aim, though pressures are visible in the authority attached to the referential languages. This did not represent in the longue durée a barrier to multilingual opportunities, which are consistently visible across the medi­eval world.354 * * *

354 Grévin, Le parchemin des cieux, pp. 34–35.

15 3

154

m a r i e l ege n d r e

I would like to thank Petra Sijpesteijn for always challenging my views on this topic, during the conference in Knoxville, on the plane, in the taxi and during many virtual and in-person meetings over more than a decade. Our exchanges are very much reflected in the following pages, and I am forever grateful for her guidance. While writing this article I have benefited greatly from discussions with other scholars, particularly Hannah-Lena Hagemann, Jaakko Hämeen-Anttila, Pamela Klasova, Peter Verkinderen, Edward Zychowicz-Coghill, and everyone involved in the early Islam reading group at the University of Edinburgh and in The Balzan Seminar on the Formation, Maintenance, and Failure of States in the Muslim World before 1800. I am indebted to Salam Rassi for his generous assistance with the Syriac sources. My deepest thanks to Antoine Borrut, Michael Cook, Alison Vacca, and Luke Yarbrough for their comments and suggestions on previous versions of this article. The abbreviations for editions of papyri follow the Checklist of Editions of Greek, Latin, Demotic, and Coptic Papyri, Ostraca and Tablets (http://scriptorium. lib.duke.edu/papyrus/texts/clist.html) and The (Cumulative) Arabic Papyro­logy Biblio­graphy of Editions and Research (https://www.naher-osten.lmu.de/apb).

Biblio­graphy Primary Sources Abū Yūsuf, Kitāb al-Kharāj, ed. by Muḥammad Ibrāhīm al-Banā (Cairo: Dār al-Salām, 2016) Agapius of Manbij, Kitab al-unvan (Kitāb al-ʿunwān), histoire universelle, ed. and trans. by Alexandre Vasiliev, Patro­logia Orientalis, 7 (1911), 457–591 Akhbār al-dawla al-ʿabbāsiyya wa-fīhi akhbār al-ʿabbās wa-waladihi, ed. by ʿAbd al-ʿAzīz Dūrī and ʿAbd al-Jabbār al-Muṭṭalib (Beirut: Dār al-Ṭalīʿa, 1971) Alexandrinische Patriarchengeschichte von S. Marcus bis Michael I, 61–767 [ce], ed. by Christian Friedrich Seybold (Hamburg: L. Gräfe, 1912) The Armenian History Attributed to Sebeos, trans. by Robert W. Thomson, James HowardJohnston, and Tim Greenwood (Liverpool: Liverpool University Press, 1999) al-ʿAskarī, Kitāb al-Awāʾil, ed. by Walīd Qaṣṣāb and Muhammad al-Misrī (Damascus: Dār al-Bashāʾir, 1975) al-Balādhurī, Futūḥ al-buldān, ed. by Michael Jan de Goeje (Leiden: Brill, 1968) —— , Ansāb al-ashrāf = Kitāb Jumal min Ansāb al-ashrāf, ed. by Suhayl Zakkār and Riyāḍ Ziriklī, 13 vols (Beirut: Dār al-Fikr, 1996) Barhebraeus, Gregorii Barhebræi Chronicon Syriacum, ed. by Paulus Bedjan (Paris: Maisonneuve, 1890) —— , The Chrono­graphy of Gregory Abû’l-Faraj, Commonly Known as Bar Hebraeus, trans. by Ernest Wallis Budge, 2 vols (London: Oxford University Press, 1932) Chronicle of 1234 = Anonymi auctoris chronicon ad Annum Christi 1234 pertinens, ed. by JeanBaptiste Chabot, 2 vols (Leuven: L. Durbecq, 1952) The Chronicle of Zuqnīn, Parts III and IV: a.d. 488–775, trans. by Amir Harrak (Toronto: Pontifical Institute of Medi­eval Studies, 1999)

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

Conquerors and Chroniclers of Early Medi­eval Spain, trans. by Kenneth Baxter Wolf (Liverpool: Liverpool University Press, 1999) Continuatio isidoriana hispana: Crónica mozárabe de 754; Estudio, edición crítica y traducción, ed. and trans. by José Eduardo López Pereira (Madrid: Caja España de Inversiones and Archivo Historico Diocesano de Léon, 2009) al-Dhahabī, Taʾrīkh al-islām wa-wafayāt al-mashāhīr wa-l-aʿlām, ed. by ʿUmar ʿAbd al-Salām Tadmurī, 52 vols (Beirut: Dār al-Kitāb al-ʿArabī, 1990–1999) History of the Patriarchs of the Coptic Church of Alexandria: Agathon to Michael I, ed. and trans. by Basil Thomas Alfred Evetts, Patro­logia Orientalis, 5 (1910), 1–215 Ibn ʿAbd al-Ḥakam, Futūḥ Misr wa-akhbāruhā, ed. by Charles Torrey (New Haven, CT: Yale University Press, 1922) Ibn ʿAbd Rabbihi, al-ʿIqd al-farīd, ed. by ʿAbd al-Majīd al-Raḥīnī, 8 vols (Beirut: Dār alKutub al-ʿIlmiyya, 1997) Ibn ʿAsākir, Taʾrīkh madīnat Dimashq, ed. by ʿAlī Shīrī, 80 vols (Beirut: Dār al-Fikr, 1995–1997) Ibn Aʿtham al-Kūfī, Kitāb al-Futūḥ, ed. by Muḥammad ʿAbd al-Muʿīd Khān, 8 vols (Hyderabad: Dāʾirat al-Maʿārif al-ʿUthmāniyya, 1968–1975) Ibn al-Athīr, al-Kāmil fī al-taʾrīkh, ed. by Carolus Johannes Tornberg, 13 vols (Beirut: Dār Sādir, 1965–1967) Ibn Durayd, Jamharat al-lugha, ed. by Ramzī Baʿalbakī, 3 vols (Beirut: Dār al-ʿIlm li–lMalāyīn, 1987) Ibn al-Faqīh, Kitāb al-Buldān, ed. by Yūsuf al-Hādī (Beirut: ʿĀlam al-Kutub, 1996) Ibn Khaldūn, Taʾrīkh, ed. by Khalīl Shaḥāda, 8 vols (Beirut: Dār al-Fikr, 1988) Ibn al-Khatīb, al-Iḥāṭa fī akhbār Gharnāṭa, ed. by Muḥammad ʿAbd Allāh ʿInān, 4 vols (Cairo: n.p., 1973–1977) Ibn al-Nadīm, Kitāb al-Fihrist, ed. by Riḍā Tajaddud (Tehran: Intishārāt-i Asāṭīr, 1971) Ibn Rusta, Kitāb al-Aʿlāq al-nafīsa, ed. by Michael Jan de Goeje (Leiden: Brill, 1892) Ibn Saʿd, Kitāb al-Ṭabaqāt al-kubrā, ed. by Iḥsān ʿAbbās, 8 vols (Beirut: Dār Ṣādir, 1968) Ibn al-Ṭiqṭaqā, al-Fakhrī fī al-ādāb al-sulṭāniyya wa-l-duwal al-islāmiyya, ed. by ʿAbd alQādir Muḥammad Māyū (Beirut: Dār al-Qalam al-ʿArabī, 1997) Ibn Yūnus, Taʾrīkh Ibn Yūnus al-Miṣrī, ed. by ʿAbd al-Fattāḥ Fatḥī ʿAbd al-Fattāḥ, 2 vols (Beirut: Dār al-Kutub al-ʿIlmiyya, 2000) al-Jāḥiẓ, Kitāb al-Ḥayawān, 5 vols (Beirut: Dār al-Kutub al-ʿIlmiyya, 2002) al-Jahshiyārī, Kitāb al-Wuzarāʾ wa-l-kuttāb, ed. by Musṭafā al-Saqqā, Ibrāhīm al-Abyārī, and ʿAbd al-Ḥafīz Ṣhalabī (Cairo: Muṣṭafā al-Bābī al-Ḥalabī, 1938) Khalīfa b. Khayyāṭ, Taʾrīkh, ed. by Akram Ḍiyāʾ al-ʿUmarī (Riyadh: Dār Ṭayba, 1985) —— , Khalifa Ibn Khayyat’s History on the Umayyad Dynasty (660–750), trans. by Carl Wurtzel and Robert G. Hoyland, Translated Texts for Historians, 63 (Liverpool: Liverpool University Press, 2015) al-Kindī, Kitāb al-Wulāt wa-kitāb al-Quḍāt, ed. by Rhuvon Guest (Leiden: Brill, 1912) Łewond, History, ed. and trans. by Sergio La Porta and Alison Vacca (Chicago: The Oriental Institute at the University of Chicago Press, forthcoming) al-Maqrīzī, al-Khiṭaṭ = al-Mawāʿiẓ wa-l-iʿtibār fī dhikr al-khiṭaṭ wa-l-āthār (Bulaq: n.p., 1853) —— , Kitāb al-Muqaffā al-kabīr, ed. by Muḥammad al-Yaʿlāwī, 8 vols (Beirut: Dār al-Gharb al-Islāmī, 1991)

15 5

156

m a r i e l ege n d r e

—— , al-Khiṭaṭ = al-Mawāʿiẓ wa-l-iʿtibār fī dhikr al-khiṭaṭ wa-l-āthār, ed. by Ayman Fuʾād Sayyid, 4 vols (London: Al-Furqān Islāmic Heritage Foundation, 2013) —— , Book of Exhortations and Useful Lessons in Dealing with Topo­graphy and Historical Remains (al-Khiṭaṭ), Part I, trans. by Karl Stowasser, ed. by Frederic Bauden and Clopper Almon (Leiden: Brill, 2022) al-Māwardī, al-Aḥkām al-sulṭāniyya wa-l-wilāyāt al-dīniyya, ed. by Aḥmad Jād (Cairo: Dār al-Ḥadīth, 2006) Michael the Syrian, Chronique de Michel le Syrien, ed. and trans. by Jean-Baptiste Chabot, 4 vols (Paris: Académie des Inscriptions et Belles-Lettres, 1899–1910) Miskawayh, Tajārib al-umam, ed. by Abū al-Qāsim Imāmī, 8 vols (Tehran: Dār-i Surūsh, 1987–2001) al-Mubarrad, al-Kāmil fī al-lugha wa-l-adab wa-l-naḥw wa-l-taṣrīf, ed. by Aḥmad Muḥammad Shākir, 3 vols (Cairo: Muṣṭafā al-Bābī al-Ḥalabī, 1936–1937) al-Muqaddasī, Aḥsan al-taqāsīm fī maʿrifat al-aqālīm, ed. by Michael Jan de Goeje (Leiden: Brill, 1906) al-Nuwayrī, Nihāyat al-arab fī funūn al-adab, ed. by Saʿīd ʿAbd al-Fattāḥ ʿĀshūr, 33 vols (Cairo: al-Muʾassasa al-Miṣriyya al-ʿĀmma, 1964–1998) al-Qalqashandī, Subḥ al-aʿshā fī ṣināʿat al-inshā, 14 vols (Cairo: Dār al-Kutub al-Miṣriyya, 1913–1922) —— , Maʾāthir al-ināfa fī maʿālim al-khilāfa, ed. by ʿAbd al-Sattār Aḥmad Farrāj, 3 vols (Kuwait: Wizārat al-Irshād wa-l-Inbāʿ, 1964) al-Rāghib al-Iṣfahānī, Muḥāḍarāt al-udabāʾ wa-muḥāwarāt al-shuʿarāʾ wa-l-bulaghāʾ, ed. Riyāḍ ʿAbd al-Ḥamīd Murād, 5 vols (Beirut: Dār Ṣādir, 2006) Sibṭ b. al-Jawzī, Mirʾāt al-zamān fī taʾrīkh al-aʿyān, ed. by Muḥammad Barakāt, 23 vols (Damascus: Dār al-Risāla al-ʿĀlamiyya, 2013) al-Ṣafadī, Kitāb al-Wāfī bi-l-wafayāt, ed. by Muḥammad al-Arnāʾūṭ and Turkī Muṣṭafā, 19 vols (Beirut: Dār Iḥyāʾ al-Turāth, 2000) al-Ṣūlī, Adab al-kuttāb, ed. by Muḥammad Bahjat al-Atharī and Maḥmūd Shukrī al-Alūsī (Cairo al-Maṭbaʿa al-Salafiyya, 1923) al-Suyūṭī, Taʾrīkh al-khulafāʾ, ed. by Ibrāhīm Ṣāliḥ (Beirut: Dār Sādir, 1996) al-Ṭabarī, Taʾrīkh al-rusul wa-l-mulūk, ed. by Muḥammad Abū al-Faḍl Ibrāhīm, 12 vols (Cairo: Dār al-Maʿārif, 1970) Theophanes the Confessor, Theophanis Chrono­graphia, ed. by Carl De Boor (Leipzig: B. G. Teubner, 1883) Theophanes the Confessor, The Chronicle of Theophanes Confessor: Byzantine and Near Eastern History a.d. 284–813, trans. by Cyril A. Mango and Roger Scott (Oxford: Clarendon Press, 1997) al-Yaʿqūbī, Taʾrīkh, ed. by Martijn Theodoor Houtsma, 2 vols (Leiden: Brill, 1883) Yāqūt al-Ḥamawī, Muʿjam al-ubabāʾ, ed. by David S. Margoliouth, 6 vols (Leiden: Brill, 1907–1931)

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

Secondary Works Abbott, Nabia, ‘A New Papyrus and a Review of the Administration of ʿUbaid Allāh b. alḤabḥāb’, in Arabic and Islamic Studies in Honor of Hamilton A.R. Gibb, ed. by George Makdisi (Leiden: Brill, 1965), pp. 21–35 Agha, Saleh Said, The Revolution which Toppled the Umayyads: Neither Arab nor ʿAbbāsid (Leiden: Brill, 2003) Andersson, Tobias, Early Sunnī Historio­graphy: A Study of the Tārīkh of Khalīfa b. Khayyāṭ (Leiden: Brill, 2018) Anthony, Sean W., ‘Fixing John Damascene’s Bio­graphy: Historical Notes on his Family Background’, Journal of Early Christian Studies, 23.4 (2015), 607–27 Arjomand, Said Amir, ‘ʿAbd Allah Ibn Al-Muqaffaʿ and the ʿAbbasid Revolution’, Iranian Studies, 27.1–4 (1994), 9–36 Bates, Michael, ‘The Dirham Mint of the Northern Provinces of the Umayyad Caliphate’, Armenian Numismatic Journal/Hay Dramagitakan Handes, 15.1 (1989), 89–111 Bates, Michael, and Mehdy Shaddel, ‘Note on a Peculiar Arab-Sasanian Coinage of Ibn al-Ash’ath’, Journal of the Royal Asiatic Society, 32.3 (2022), 1–10 Bell, Harold I., Greek Papyri in the British Museum: Catalogue with Texts, iv: The Aphrodito Papyri, with an Appendix of Coptic Papyri edited by W.E. Crum (London: British Museum, 1910) —— , ‘Two Official Letters of the Arab Period’, Journal of Egyptian Archaeo­logy, 12 (1926), 265–81 Berkes, Lajos, and Naïm Vanthieghem, ‘A Late Coptic Tax Receipt from the Egyptian National Library (P. Cair. Nat. Library inv. 3890)’, Journal of Coptic Studies, 21 (2019), 13–17 —— , ‘Notes on the Careers of Nāǧid b. Muslim and ʿAbd al-Malik b. Yazīd’, Chronique d’Égypte, 95.189 (2020), 154–61 Berkes, Lajos, and Khaled M. Younes, ‘A Trilingual Scribe from Abbasid Egypt? A Note on CPR XXII 171’, Archiv für Papyrusforschung, 58.1 (2012), 97–100 Bone, Harry, ‘The Administration of Umayyad Syria: The Evidence of the Copper Coins’ (unpublished doctoral dissertation, Princeton University, 2000) Borrut, Antoine, Entre mémoire et pouvoir: l’espace syrien sous les derniers omeyyades et les premiers abbassides (v. 72–193/692–809) (Leiden: Brill, 2011) —— , ‘Pouvoir mobile et construction de l’espace dans les premiers siècles de l’islam’, in Le gouvernement en déplacement: pouvoir et mobilité de l’antiquité à nos jours, ed. by Sylvain Destephen, Josiane Barbier, and François Chausson (Rennes: Presses Universitaires de Rennes, 2019), pp. 243–67 —— , ‘ʿAbdallāh b. ʿAbd al-Malik b. Marwān’, in Encyclopaedia of Islam, 3rd edn by Kate Fleet, Gudrun Krämer, Denis Matringe, John Nawas, and Everett Rowson (Leiden: Brill, online edn), http://dx.doi.org/10.1163/1573–3912_ei3_COM_24680 [accessed 23 June 2021] Bouderbala, Sobhi, ‘Ǧund Miṣr: étude de l’administration militaire dans l’Égypte des débuts de l’Islam 21/642–218/833’ (unpublished doctoral thesis, Université Paris 1 Panthéon-Sorbonne, 2008)

15 7

1 58

m a r i e l ege n d r e

—— , ‘Les mawālī à Fusṭāṭ aux deux premiers siècles de l’islam et leur intégration sociale’, in Le processus d’islamisation en Sicile et en Méditerranée centrale, ed. by Fabiola Ardizzone and Annliese Nef (Paris: École Française de Rome, 2014), pp. 46–62 Bouderbala, Sobhi, and Abdelhamid Fenina, ‘Fiscalité et pratiques monétaires dans l’Égypte pré-fatimide à la lumière des documents papyro­logiques et numismatiques’, in Money Rules! The Monetary Economy of Egypt, from Persians until the Beginning of Islam, ed. by Thomas Faucher (Cairo: Institut Français d’Archeo­logie Orientale, 2020), pp. 393–410 Brockopp, Jonathan E., ‘The Formation of Islamic Law: The Egyptian School 750–900’, Annales islamo­logiques, 45 (2011), 123–40 Chalmeta, Pedro, ‘Conquista y sumisión de Hispania’, in Arabización, islamización y resistencias en al-Andalus y el Maghreb, ed. by Bilal Sarr and Ángeles Navarro García (Granada: Universidad de Granada, 2019), pp. 19–35 Cobb, Paul, ‘The Empire in Syria’, in The New Cambridge History of Islam, i: The Formation of the Islamic World, Sixth to Eleventh Centuries, ed. by Chase F. Robinson (Cambridge: Cambridge University Press, 2010), pp. 226–68 Conrad, Gerhard, Abū’l-Ḥusain al-Rāzī (347/958) und seine Schriften: Untersuchungen zur frühen damaszener Geschichtsschreibung (Stuttgart: Deutsche Morgenländische Gesellschaft and F. Steiner, 1991) Conterno, Maria, ‘Theophilos, “the More Likely Candidate”? Towards a Reappraisal of the Question of Theophanes’ “Oriental Source”’, in Studies in Theophanes, ed. by Marek Jankowiak and Federico Montinaro (Paris: Centre de Recherche d’Histoire et Civilisation de Byzance, 2015), pp. 383–400 Cottrell, Emily, ‘An Early Mirror for Princes and Manual for Secretaries: The Epistolary Novel of Aristotle and Alexander’, in Alexander the Great and the East: History, Art, Tradition, ed. by Krzysztof Nawotka and Agnieszka Wojciechowska (Wiesbaden: Harrassowitz, 2016), pp. 303–28 Cromwell, Jennifer, ‘The Coptic Texts in the Archive of Flavius Atias’, Zeitschrift für Papyro­ logie und Epi­graphik, 184 (2013), 280–88 —— , Recording Village Life: A Coptic Scribe in Early Islamic Egypt (Ann Arbor: University of Michigan Press, 2017) —— , ‘Language Policy and the Administrative Framework of Early Islamic Egypt’, in Ancient Egyptian Society, Challenging Assumptions, Exploring Approaches, ed. by Danielle Candelora, Nadia Ben-Marzouk, and Kathlyn M. Cooney (London: Routledge, 2022), pp. 299–312 —— , ‘Scribal Networks, Taxation, and the Role of Coptic in Marwanid Egypt’, in Egypt and the Eastern Mediterranean World, From Constantinople to Baghdad, 500–1000 ce, ed. by Jelle Bruning, Janneke de Jong, and Petra M. Sijpesteijn (Cambridge: Cambridge University Press, forthcoming), pp. 353–78 De Jong, Janneke, and Alain Delattre, ‘Greek as a Minority Language’, in The Late Antique World of Early Islam, ed. by Robert G. Hoyland (Princeton, NJ: Darwin Press, 2015), pp. 37–62 Debié, Muriel, L’écriture de l’histoire en syriaque: transmissions interculturelles et constructions identitaires entre hellénisme et islam (Leuven: Peeters, 2015)

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

—— , ‘Theophanes’ “Oriental Source”: What Can We Learn from Syriac Historio­graphy?’, in Studies in Theophanes, ed. by Marek Jankowiak and Federico Montinaro (Paris: Centre de Recherche d’Histoire et Civilisation de Byzance, 2015), pp. 365–82 —— , ‘Christians in the Service of the Caliph: Through the Looking Glass of Communal Identities’, in Christians and Others in the Umayyad State, ed. by Antoine Borrut and Fred M. Donner (Chicago: Oriental Institute of the University of Chicago, 2016), pp. 53–71 Delattre, Alain, ‘Cinq entagia coptes’, Archiv für Papyrusforschung, 54.1 (2005), 79–86 Delattre, Alain, Rosario Pintaudi, and Naïm Vanthieghem, ‘Un entagion bilingue du gouverneur ʿAbd al-ʿAzīz ibn Marwān trouvé à Antinoé’, Chronique d’Égypte, 88.176 (2013), 363–71 Djaït, Hichem, ‘La Wilāya d’Ifrīqiya au iie/viiie siècle: étude institutionnelle I’, Studia Islamica, 27 (1967), 77–121 —— , ‘La Wilāya d’Ifrīqiya au iie/viii siècle: étude institutionnelle (suite et fin)’, Studia Islamica, 28 (1968), 79–107 Dozy, Reinhart, Supplément aux dictionnaires arabes, 2 vols (Leiden: Brill, 1881) Donner, Fred M., The Articulation of Early Islamic State Structures (Burlington, VT: Ashgate, 2012) Dūrī, ʿAbd al-ʿAzīz, ‘Dīwān’, in Encyclopaedia of Islam, 2nd edn, ed. by Peri Bearman, Thierry Bianquis, Clifford E. Bosworth, Emeri van Donzel, and Wolfhart P. Heinrichs, ii (Leiden: Brill, 1991), pp. 323–27 —— , Early Islamic Institutions: Administration and Taxation from the Caliphate to the Umayyads and ʿAbbasids (London: I. B. Tauris, 2011) Fenina, Abdelhamid, ‘L’arabisation du monnayage d’Ifrīqiya: étapes et signification’, in Civilisations en transition, ii: Sociétés multilingues à travers l’histoire du Proche-Orient, ed. by Jean-Luc Fournet, Jean-Michel Mouton, and Jacques Paviot (Byblos: Centre International des Sciences de l’Homme, 2016), pp. 115–68 Fenwick, Corisande, Early Islamic North Africa: A New Perspective (London: Bloomsbury Academic, 2020) —— , ‘The Umayyads and North Africa: Imperial Rule and Frontier Society’, in The Umayyad World, ed. by Andrew Marsham (London: Routledge, 2021), pp. 293–313 Foss, Clive, Arab Byzantine Coins: An Introduction, with a Catalogue of the Dumbarton Oaks Collection (Washington, DC: Dumbarton Oaks Research Library and Collection, 2008) Fournet, Jean-Luc, The Rise of Coptic: Egyptian versus Greek in Late Antiquity (Princeton, NJ: Princeton University Press, 2020) Gabrieli, Francesco, ‘L’opera di Ibn al-Muqaffa’, Rivista degli studi orientali, 13.3 (1932), 197–247 Greenwood, Tim W., ‘Movses Daskhurants‘i/Movses Kaghankatuats‘i’, in Christian-Muslim Relations: A Biblio­graphical History, i: 600–900, ed. by David Thomas and Barbara Roggema (Leiden: Brill, 2009), pp. 261–67 Grévin, Benoît, Le parchemin des cieux: essai sur le Moyen Âge du langage (Paris: Seuil, 2013) Görke, Andreas, ‘The Historical Tradition about al-Ḥudaybiya: A Study of ʿUrwa b. alZubayr’s Account’, in The Bio­graphy of Muhammad: The Issue of the Sources, ed. by Harald Motzki (Leiden: Brill, 2000), pp. 240–75 —— , ‘Prospects and Limits in the Study of the Historical Muhammad’, in The Transmission and Dynamics of the Textual Sources of Islam: Essays in Honour of Harald Motzki, ed. by

1 59

1 60

m a r i e l ege n d r e

Nicolet Boekhoff-van der Voort, Kees Versteegh, and Joas Wagemakers (Leiden: Brill, 2011), pp. 137–51 Gonis, Nikolaos, ‘Reconsidering Some Fiscal Documents from Early Islamic Egypt II’, Zeitschrift für Papyro­logie und Epi­graphik, 150 (2014), 187–93 Gundelfinger, Simon, and Peter Verkinderen, ‘The Governors of al-Shām and Fārs in the Early Islamic Empire: A Comparative Regional Perspective’, in Transregional and Regional Elites: Connecting the Early Islamic Empire, ed. by Hannah-Lena Hagemann and Stefan Heidemann (Berlin: de Gruyter, 2020), pp. 255–330 Gutas, Dimitri, Greek Thought, Arabic Culture: The Graeco-Arabic Translation Movement in Baghdad and Early ʿAbbasid Society (2nd–4th/8th–10th Centuries) (London: Routledge, 1998) —— , ‘On Graeco-Arabic Epistolary “Novels”’, Middle Eastern Literatures, 12.1 (2009), 59–70 Gyselen, Rika, ‘Diwan et “trésorerie” sassanides: Premières attestations sigillo­graphiques’, Studia Iranica, 32.1 (2003), 123–26 —— , Arab-Sasanian Copper Coinage (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2009) Hämeen-Anttila, Jaakko, Khwadāynāmag: The Middle Persian Book of Kings (Leiden: Brill, 2018) Haug, Robert, The Eastern Frontier: Limits of Empire in Late Antique and Early Medi­eval Central Asia (London: I. B. Tauris, 2019) Hawting, Gerald R., The First Dynasty of Islam: The Umayyad Caliphate ad 661–750 (London: Routledge, 2000) Hilkens, Andy, The Anonymous Syriac Chronicle of 1234 and its Sources (Leuven: Peeters, 2018) Hoyland, Robert G., Seeing Islam as Others Saw It: A Survey and Evaluation of Christian, Jewish and Zoroastrian Writings on Early Islam (Princeton, NJ: Darwin Press, 1997) —— , Theophilus of Edessa’s Chronicle and the Circulation of Historical Knowledge in Late Antiquity and Early Islam (Liverpool: Liverpool University Press, 2011) Ilisch, Lutz, ‘Muhammad Drachms and their Relation to Umayyad Syria and Northern Mesopotamia’, Journal of the Oriental Numismatic Society, 193 (2003), 17–24 Jonson, Trent M. H., ‘A Numismatic History of the Early Islamic Precious Metal Coinage of North Africa and the Iberian Peninsula’ (unpublished doctoral thesis, University of Oxford, 2014) Juynboll, Gautier H. A., Muslim Tradition: Studies in Chrono­logy, Provenance and Authorship of Early Ḥadīth (Cambridge: Cambridge University Press, 1983) Kameya, Manabu, ‘From Qusṭāl to Jahbadh: An Aspect of Transition on the Egyptian Tax-Collecting System’, in New Frontiers of Arabic Papyro­logy: Arabic and Multilingual Texts from Early Islam, ed. by Sobhi Bouderbala, Sylvie Denoix, and Matt Malczycki (Leiden: Brill, 2017), pp. 141–60 Khalidi, Tarif, Arabic Historical Thought in the Classical Period (Cambridge: Cambridge University Press, 1994) Khalek, Nancy, ‘Some Notes on the Representation of Non-Muslim Officials in alǦahšiyārī’s (d. 331/942) Kitāb al-Wuzarāʾ wa-l-kuttāb’, Arabica, 62.4 (2015), 503–20 Khan, Geoffrey, Arabic Documents from Early Islamic Khurasan (London: Nour Foundation, 2007) —— , ‘The Opening Formula and Witness Clauses in Arabic Legal Documents from the Early Islamic Period’, Journal of the American Oriental Society, 139.1 (2019), 23–40

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

Kennedy, Hugh, ‘The Financing of the Military in the Early Islamic State’, in The Byzantine and Early Islamic Near East, iii: States, Resources and Armies, ed. by Averil Cameron (Princeton, NJ: Darwin Press, 1995), pp. 361–78 —— , The Armies of the Caliphs: Military and Society in the Early Islamic State (London: Routledge, 2001) —— , Muslim Spain and Portugal: A Political History of al-Andalus (Hoboken, NJ: Taylor and Francis, 2014) Klasova, Pamela, ‘Empire through Language: Al-Ḥajjāj b. Yūsuf al-Thaqafī and the Power of Oratory in Umayyad Iraq’ (unpublished doctoral dissertation, Georgetown University, 2018) Kratchkovskaya, V. A., and I. Kratchkovsky, ‘Drevnejshij arabskij dokument iz Srednej Azii’ [The earliest Arabic document from Central Asia], in Sogdiiskii Sbornik: Sbornik statei o pamyatnikakh sogdiiskogo iazyka i kultury naydennykh na gore Mug v Tadzhikskoiy SSR (Leningrad: Akademiia Nauk, 1934), pp. 52–90 Legendre, Marie, ‘Neither Byzantine nor Islamic? The Duke of the Thebaid and the Formation of the Umayyad State’, Historical Research, 89.243 (2016), 3–18 —— , ‘Caliphal Estates and State Policy over Landholding: Theory and Practice between Literary and Documentary Evidence from Early Islamic Egypt’, in Authority and Control in the Countryside: From Antiquity to Islam in the Mediterranean and Near East (Sixth–Tenth Century), ed. by Alain Delattre, Marie Legendre, and Petra M. Sijpesteijn (Leiden: Brill, 2019), pp. 392–419 —— , ‘Aspects of Umayyad Administration’, in The Umayyad World, ed. by Andrew Marsham (London: Routledge, 2021), pp. 133–57 Lévi, Israël, ‘Une apocalypse judéo-arabe’, Revue des études juives, 67.134 (1914), 178–82 Levy-Rubin, Milka, Non-Muslims in the Early Islamic Empire (Cambridge: Cambridge University Press, 2011) Lindstedt, Ilkka, ‘The Transmission of al-Madāʾinī’s Historical Material to al-Balādhurī and al-Ṭabarī: A Comparison and Analysis of Two Khabars’, in Travelling through Time: Essays in Honour of Kaj Öhrnberg, ed. by Sylvia Akar, Jaakko Hämeen-Anttila, and Inka Nokso-Koivisto (Helsinki: Societas Orientalis Fennica, 2013), pp. 41–64 Løkkegaard, Frede, Islamic Taxation in the Classic Period: With Special Reference to Circumstances in Iraq (Philadelphia: Porcupine Press, 1978) Longworth, Kyle, ‘Islamic Bureaucrats in Late Antiquity: Administration and Elites during the Umayyad Caliphate (ca. 661–750 c.e.)’ (unpublished doctoral dissertation, University of Chicago, 2022) Lucas, Noëmie, ‘Le pouvoir de la terre: éssai d’histoire de la société bas-iraqienne au iie/ viiie siècle’ (unpublished doctoral thesis, Université Paris 1 Panthéon-Sorbonne, 2020) —— , ‘A State Letter from a Marwanid Caliph to his Governor of Iraq: A Historio­graphical Investigation into Khālid b. ʿAbd Allāh al-Qasrī’s Downfall’, in The Ties that Bind: Mechanisms of Social Dependency in the Early Islamic Empire, ed. by Edmund Hayes and Petra M. Sijpesteijn (Cambridge: Cambridge University Press, forthcoming) Lynch, Ryan, Arab Conquests and Early Islamic Historio­graphy (London: Bloomsbury, 2020) Mabra, Joshua, Princely Authority in the Early Marwānid State: The Life of of ʿAbd al-ʿAzīz Ibn Marwān, Islamic History and Thought, 2 (Piscataway, NJ: Gorgias Press, 2017) Manzano Moreno, Eduardo, ‘El asentamiento y la organización de los yund-s sirios en alAndalus’, Al-Qanṭara, 14.2 (1993), 327–59

161

162

m a r i e l ege n d r e

—— , Conquistadores, emires y califas: los omeyas y la formación de al-Andalus (Barcelona: Crítica, 2014) —— , ‘Conquest and Settlement: What al-Andalus Can Tell Us about the Arab Expansion at the Time of the Umayyad Caliphate’, in The Umayyad World, ed. by Andrew Marsham (London: Routledge, 2021), pp. 314–31 Maróth, Miklós, The Correspondence between Aristotle and Alexander the Great: An Anonymous Greek Novel in Letters in Arabic Translation (Piliscsaba: Avicenna Institute of Middle Eastern Studies, 2006) Mikhail, Maged S. A., From Byzantine to Islamic Egypt. Religion, Identity and Politics after the Arab Conquest (London: I. B. Tauris, 2014) Miles, George C., ‘A Byzantine Weight Validated by al-Walīd’, Numismatic Notes and Mono­ graphs, 87 (1939), 1–11 Mochiri, Malek Iradj, ‘A Pahlavi Forerunner of the Umayyad Reformed Coinage’, Journal of the Royal Asiatic Society of Great Britain and Ireland, 2 (1981), 168–72 Morimoto, Koseï, ‘The Dīwāns as Registers of the Arab Stipendaries in Early Islamic Egypt’, in Itinéraires d’Orient: hommages à Claude Cahen, ed. by Raoul Curiel and Rika Gyselen (Bures-sur-Yvette: Groupe pour l’Étude de la Civilisation du Moyen-Orient, 1994), pp. 353–65; repr. in Fred M. Donner, The Articulation of Early Islamic State Structures (Burlington: Ashgate, 2012), pp. 227–39 Noth, Albrecht, and Lawrence I. Conrad, The Early Arabic Historical Tradition: A SourceCritical Study (Princeton, NJ: Darwin Press, 1994) Ostrogorsky, Georg, ‘Die Chrono­logie des Theophanes im 7. und 8. Jahrhundert’, Byzantinisch-neugriechische Jahrbücher, 7 (1928–1929), 1–56 Palme, Bernhard, ‘Administration, Late Antique Egypt’, The Encyclopedia of Ancient History, ed. by Roger S. Bagnall (Malden, MA: Wiley-Blackwell, online edn), https://doi-org. ezproxy.is.ed.ac.uk/10.1002/9781444338386.wbeah07001 [accessed 25 November 2022] Papaconstantinou, Arietta, ‘Administrating the Early Islamic Empire: Insights from the Papyri’, in Money, Power and Politics in Early Islamic Syria, ed. by John Haldon (Farnham: Ashgate, 2010), pp. 57–74 —— , ed., The Multilingual Experience in Egypt, from the Ptolemies to the Abbasids (Abingdon: Routledge, 2010) —— , ‘Arabic: Language of Empire and Language of Egypt’, in Egypt and Empire: The Formation of Religious Identity after Rome, ed. by Elizabeth R. O’Connell (Leuven: Peeters, 2022), pp. 293–309 Penn, Michael Philip, Envisioning Islam: Syriac Christians and the Early Muslim World (Philadelphia: University of Pennsylvania Press, 2015) Perikhanian, Anahit, The Book of a Thousand Judgements (A Sasanian Law-Book) (Costa Mesa, CA: Mazda Publishers and New York: Bibliotheca Persica, 1997) Perlman, Yaara, ‘The Bodyguard of the Caliphs during the Umayyad and Early Abbasid Periods’, Al-Qanṭara, 36.2 (2015), 317–40 Puin, Gerd-Rüdiger, Der Dīwān von ʿUmar ibn al-Ḫaṭṭāb: ein Beitrag zur frühislamischen Verwaltungsgeschichte (Bonn: Rheinische Friedrich-Wilhelms-Universität, 1970) al-Qāḍī, Wadād, ‘Sālim Abū al-ʿAlāʾ’, in Encyclopaedia of Arabic Literature, ed. by Julie Scott Meisami and Paul Starkey, 2 vols (London: Routledge, 1998), i, pp. 681–82 —— , ‘Population Census and Land Surveys under the Umayyads (41–132/661–750)’, Der Islam, 83 (2008), 341–416

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

—— , ‘The Names of Estates in State Registers before and after the Arabization of the “Dīwāns”’, in Umayyad Legacies, ed. by Antoine Borrut and Paul Cobb (Leiden: Brill, 2010), pp. 255–80 —— , ‘Death Dates in Umayyad Stipends Registers (Dīwān al-ʿAṭāʾ)? The Testimony of the Papyri and the Literary Sources’, in From Bāwīṭ to Marw: Documents from the Medi­eval Muslim World, ed. by Andreas Kaplony, Daniel Potthast, and Cornelia Römer (Leiden: Brill, 2015), pp. 59–82 —— , ‘Identity Formation of the Bureaucracy of the Early Islamic State: ʿAbd al-Ḥamīd’s “Letter to the Secretaries”’, in Mediterranean Identities in the Premodern Era: Entrepôts, Islands, Empires, ed. by John Watkins and Kathryn Reyerson (London: Taylor and Francis, 2016), pp. 141–54 —— , ‘Non-Muslims in the Muslim Conquest Army in Early Islam’, in Christians and Others in the Umayyad State, ed. by Antoine Borrut and Fred M. Donner (Chicago: Oriental Institute at the University of Chicago, 2016), pp. 83–127 Qaṭṭāṭ, Ḥayāt, al-ʿArab fī al-jāhiliyya al-akhīra wa-l-islām al-mubakkir (Tunis: Dār al-Amal, 2006) Rassi, Salam, ‘Alchemy in an Age of Disclosure: The Case of an Arabic Pseudo-Aristotelian Treatise and its Syriac Christian “Translator”’, Asiatische Studien, 75.2 (2021), 545–609 Rebstock, Ulrich, ‘Observations on the Diwan al-Kharaj and the Assessment of Taxes in Umayyad Syria’, in The History of Bilad al-Sham during the Umayyad Period (Fourth International Conference, 24–29 October 1987, Amman): Proceedings of the Third Symposium/Bilad al-Sham History Committee, ed. by Muḥammad ʿAdnān Bakhīt, 2 vols (Amman: Yarmouk University, 1989), ii, pp. 229–46 Ricci, Ronit, Islam Translated: Literature, Conversion, and the Arabic Cosmopolis of South and Southeast Asia (Chicago: University of Chicago Press, 2011) Richter, Tonio Sebastian, ‘Language Choice in the Qurra Dossier’, in The Multilingual Experience in Egypt, from the Ptolemies to the Abbasids, ed. by Arietta Papaconstantinou (Abingdon: Routledge, 2010), pp. 189–220 Robin, Christian Julien, ‘Tribus et territoires d’Arabie, d’après les inscriptions antiques et les généalogies d’époque islamique’, Semitica et classica, 13 (2020), 205–50 Robinson, Chase F., Empire and Elites after the Muslim Conquest: The Transformation of Northern Mesopotamia (Cambridge: Cambridge University Press, 2000) —— , ʿAbd al-Malik (Oxford: Oneworld, 2005) Rouighi, Ramzi, Inventing the Berbers: History and Ideo­logy in the Maghrib (Philadelphia: University of Pennsylvania Press, 2019) Rustow, Marina, The Lost Archive: Traces of a Caliphate in a Cairo Synagogue (Princeton, NJ: Princeton University Press, 2020) Saliba, George, Islamic Science and the Making of the European Renaissance (Cambridge, MA: MIT Press, 2011) Scheiner, Jens, ‘Ibn ʿAsākir’s Virtual Library as Reflected in his Taʾrīkh madīnat Dimashq’, in New Perspectives on Ibn ʿAsākir in Islamic Historio­graphy, ed. by Steven Judd and Jens Scheiner (Leiden: Brill, 2017), pp. 156–257 Sharon, Moshe, ‘The Military Reforms of Abū Muslim, their Background and Consequences’, in Studies in Islamic History and Civilization in Honour of Professor David Ayalon, ed. by Moshe Sharon ( Jerusalem: Cana and Leiden: Brill, 1986), pp. 105–43

163

1 64

m a r i e l ege n d r e

Shoemaker, Stephen J., A Prophet Has Appeared: The Rise of Islam through Christian and Jewish Eyes; A Sourcebook (Oakland: University of California Press, 2021) Sijpesteijn, Petra M., ‘A Seventh/Eighth-Century List of Companions from Fusṭāṭ’, in Sixty-Five Papyro­logical Texts Presented to Klaas A. Worp on the Occasion of his 65th Birthday, ed. by Francisca A. J. Hoogendijk and Brian Paul Muhs (Leiden: Brill, 2008), pp. 369–77 —— , ‘Multilingual Archives and Documents in Post-Conquest Egypt’, in The Multilingual Experience in Egypt, from the Ptolemies to the Abbasids, ed. by Arietta Papaconstantinou (Abingdon: Routledge, 2010), pp. 105–24 —— , ‘Army Economics: An Early Papyrus Letter Related to ʿAṭāʾ Payments’, in Histories of the Middle East: Studies in Middle Eastern Society, Economy and Law in Honor of A. L. Udovitch, ed. by Roxani Eleni Margariti, Adam Sabra, and Petra M. Sijpesteijn (Leiden: Brill, 2011), pp. 245–68 —— , Shaping a Muslim State: The World of a Mid-Eighth-Century Egyptian Official (Oxford: Oxford University Press, 2013) —— , ‘Expressing New Rule: Seals from Early Islamic Egypt and Syria, 600–800 ce’, in Seals: Making and Marking Connections across the Medi­eval World, ed. by Brigitte Miriam Bedos-Rezak, The Medi­eval Globe, 4 (Amsterdam: Amsterdam University Press, 2018), pp. 99–148 —— , ‘Loyal and Knowledgeable Supporters: Integrating Egyptian Élites in Early Islamic Egypt’, in Empires and Communities in the Post-Roman and Islamic World (c. 400–1000 ce), ed. by Walter Pohl (Oxford: Oxford University Press, 2021), pp. 329–59 Sokoloff, Michael, A Syriac Lexicon: A Translation from the Latin, Correction, Expansion, and Update of C. Brockelmann’s Lexicon Syriacum (Piscataway, NJ: Gorgias Press, 2009) Sourdel, Dominique, ‘Le “livre des secrétaires” de ʿAbdallāh al-Baġdādī’, Bulletin d’études orientales, 14 (1952), 115–53 —— , ‘La bio­graphie d’Ibn al-Muqaffa d’après les sources anciennes’, Arabica, 1.3 (1954), 307–23 —— , ‘La valeur littéraire et documentaire du “Livre des vizirs” d’al-Gahšiyari’, Arabica, 2.2 (1955), 193–210 Sprengling, Martin, ‘From Persian to Arabic’, American Journal of Semitic Languages and Literatures, 57.3 (1939), 175–224 Spuler, Bertold, Iran in the Early Islamic Period, ed. by Robert G. Hoyland, trans. by Gwendolin Goldbloom and Berenike Walburg (Leiden: Brill, 2014) Stasolla, Maria Giovanna, ‘How a 10th Century Learned Man Reads History: Al-Jahshiyarī (d. 942) and the Barmakids’, Eurasian Studies, 10.1–2 (2012), 221–34 Tannous, Jack, The Making of the Medi­eval Middle East: Religion, Society, and Simple Believers (Princeton, NJ: Princeton University Press, 2018) Tillier, Mathieu, ‘Les “premiers” cadis de Fusṭāṭ et les dynamiques régionales de l’innovation judiciaire (750–833)’, Annales islamo­logiques, 45 (2011), 213–42 —— , L’invention du cadi: la justice des musulmans, des Juifs et des chrétiens aux premiers siècles de l’Islam (Paris: Publications de la Sorbonne, 2017) Tillier, Mathieu, and Naïm Vanthieghem, ‘Recording Debts in Sufyānid Fusṭāṭ: A Reexamination of the Procedures and Calendar in Use in the First/Seventh Century’, in Geneses: A Comparative Study of the Historio­graphies of the Rise of

The Tr a n s l at i o n o f t h e D īwān an d  t h e M arwani d ‘ Language Re fo rm’

Christianity, Rabbinic Judaism and Islam, ed. by John Tolan (London: Routledge, 2019), pp. 148–88 Torallas Tovar, Sofia, ‘Greek and Egyptian: Did Linguistic Policies Exist in Graeco-Roman Egypt?’, in Egypt and Empire: The Formation of Religious Identity after Rome, ed. by Elizabeth R. O’Connell (Leuven: Peeters, 2022), pp. 279–92 Toral-Niehoff, Isabel, ‘History in Adab Context: “The Book on Caliphal Histories” by Ibn ʿAbd Rabbih (246/860–328/940)’, Journal of Abbasid Studies, 2.1 (2015), 61–85 Treadwell, Luke, ‘The “Orans” Drachms of Bishr ibn Marwān and the Figural Coinage of the Early Marwanid Period’, in Bayt al-Maqdis, ii: Jerusalem and Early Islam, ed. by Jeremy Johns and Julian Raby (Oxford: Oxford University Press, 1999), pp. 223–69 —— , ‘The Copper Coinage of Umayyad Iran’, Numismatic Chronicle, 168 (2008), 331–81 Urban, Elizabeth, Conquered Populations in Early Islam: Non-Arabs, Slaves and the Sons of Slave Mothers (Edinburgh: Edinburgh University Press, 2020) Vacca, Alison, ‘The Fires of Naxčawan: In Search of Intercultural Transmission in Arabic, Armenian, Greek, and Syriac’, Le muséon, 129.3/4 (2016), 323–62 —— , Non-Muslim Provinces under Early Islam: Islamic Rule and Iranian Legitimacy in Armenia and Caucasian Albania (Cambridge: Cambridge University Press, 2017) —— , ‘The Umayyad North (or: How Umayyad was the Umayyad Caliphate?)’, in The Umayyad World, ed. by Andrew Marsham (London: Routledge, 2021), pp. 219–39 Vanthieghem, Naïm, and Lev Weitz, ‘A Companion of Muḥammad in the Oldest Egyptian Bilingual Entagion’, Bulletin of the American Society of Papyro­logists, 58 (2021), 203–14 Webb, Peter, Imagining the Arabs: Arab Identity and the Rise of Islam (Edinburgh: Edinburgh University Press, 2017) Wood, Philip, The Imam of the Christians: The World of Dionysius of Tel-Mahre, c. 750–850 (Princeton, NJ: Princeton University Press, 2021) Yarbrough, Luke B., Friends of the Emir: Non-Muslim State Officials in Premodern Islamic Thought (Cambridge: Cambridge University Press, 2019) Yücesoy, Hayrettin, ‘Translation as Self-Consciousness: Ancient Sciences, Antediluvian Wisdom, and the ʿAbbāsid Translation Movement’, Journal of World History, 20.4 (2009), 523–57 Younes, Khaled M., ‘New Governors Identified in Arabic Papyri’, in Authority and Control in the Countryside: Continuity and Change in the Mediterranean (Sixth–Tenth Century), ed. by Alain Delattre, Marie Legendre, and Petra M. Sijpesteijn (Leiden: Brill, 2019), pp. 13–43 Zychowicz-Coghill, Edward, The First Arabic Annals: Fragments of Umayyad History, Studies in the History and Culture of the Middle East, 41 (Berlin: de Gruyter, 2021)

165

Antoine Borrut

Towards an Arabic Cosmopolis Culture and Power in Early Islam

In a ground-breaking study dedicated to the formation of the ‘Sanskrit cosmopolis’, Sheldon Pollock lamented: The story of how this formation arose — how Sanskrit traveled the vast distance it did and came to be used for literary and political texts, and what such texts meant to the worlds of power in which they were produced — has never been told in the historical detail it merits. Indeed, it is unclear whether the fact that there is a story to tell has been fully recognized.1 By contrast, two main narratives have characterized scholarly discussions on the spread of Arabic in the early Islamic empire. As noted at the onset of this volume, the notion of a deliberate Arabicization programme initiated by the Marwanid caliphs, and ʿAbd al-Malik in particular, has dominated modern scholarship. According to one view (which adopts Dimitri Gutas’s chrono­logy), this Arabicization policy was supplemented by the so-called translation movement in Abbasid times; according to the other (which follows George Saliba’s view favouring Umayyad beginnings), it was concomitant to it.2 As abundantly shown in the preceding pages and throughout this volume, however, neither of these narratives of Arabicization and translation prove entirely successful at explaining the startling rise and success of Arabic. Instead, we can detect ‘cracks in the master discourses’3 that invite us to think about language and power in early Islam. That Arabic was the language of the Qur’an does not alone explain its success beyond the sacred sphere. The fact that it was spoken by the first conquerors is also 1 Pollock, The Language of the Gods in the World of Men, p. 11. I am grateful to Alison Vacca, Marie Legendre, and the members of ‘The Balzan Seminar on the Formation, Maintenance, and Failure of States in the Muslim World before 1800’ for their helpful comments on earlier drafts of this paper. 2 Gutas, Greek Thought, Arabic Culture; Saliba, Islamic Science and the Making of the European Renaissance. 3 Pollock, ‘Cosmopolitan and Vernacular in History’, pp. 591–625, crediting the expression to Dipesh Chakrabarty at p. 623.

Antoine Borrut ([email protected]) is Associate Professor of History at the University of Maryland, College Park. Navigating Language in the Early Islamic World: Multilingualism and Language Change in the First Cen­turies of Islam, ed. by Antoine Borrut, Manuela Ceballos, and Alison M. Vacca, ISMAR 2 (Turnhout: Brepols, 2024) pp. 167–191 10.1484/M.ISMAR-EB.5.134626

FHG

168

a n to in e bo r r u t

not enough to explain why it became so successful throughout an immense empire in which Believers/Muslims and Arabophones represented only a small fraction of the population, at a time when identities were still much in flux.4 Whatever their merits and shortcomings, none of these narratives offers a sufficient explanation to resolve the tension surrounding the triumph of Arabic as a language of power and prestige (or langue référentielle, to borrow Benoît Grévin’s termino­logy)5 throughout the caliphate. We need a better model to make sense of why Arabic prevailed and endured as the new prestige language among non-Arabs and new Muslims alike, from Central Asia and the northern Indian subcontinent to Iberia. In Pollock’s footsteps, Ronit Ricci has made a strong case for the emergence of an ‘Arabic cosmopolis’, understood as ‘a translocal Islamic sphere constituted and defined by language, literature, and religion’, in early modern South and Southeast Asia. Ricci sketches a typo­logy of cosmopolitan formations: language emerges as the key element of the Sanskrit and Persian cosmopoleis, whereas the Arabic and Buddhist ones, she contends, ‘were centered around religion, with language playing a greater or lesser role, depending on the circumstances’.6 In that sense, Arabic cosmopolitanism is fundamentally intertwined with Islamic cosmopolitanism. What remains to be addressed, however, is the possibility of a significantly earlier cosmopolitan formation, an Arabic cosmopolis emerging in the first centuries of Islam in the newly conquered caliphal territories.

Cosmopolitan Formations Cosmopolitanism has been reclaimed as a ‘category for historical analysis’7 in recent scholarship. This has taken place in the context of the booming field of empire studies, with scholars seeking to situate cosmopolitanism as ‘an indispensable

4 In his recent ‘Arabicization, Islamization, and the Colonies of the Conquerors’, van Bladel also points out the lack of satisfactory explanations to make sense of the shift towards Arabic in early Islamic times. He notably scorns the emphasis on identity changes to explain the adoption of a new language and advocates instead for social and demo­graphic factors. Thus, he contends that garrison cities (amṣār) proved particularly central to Arabicization processes since they put speakers of different languages in direct contact. In other words, van Bladel urges us to take settlement and migration patterns into consideration to write a social history of Arabicization at different paces throughout the Caliphate. 5 Grévin, Le parchemin des cieux. 6 Ricci, Islam Translated, pp. 11, 268. Ricci’s discussion of a ‘Buddhist cosmopolis’ is primarily based on Monius, Imagining a Place for Buddhism. 7 Lavan, Payne, and Weisweiler, ‘Cosmopolitan Politics’, p. 2. See also Pollock, Bhabha, Breckenridge, and Chakrabarty, ‘Cosmopolitanisms’ and, more broadly, Breckenridge, Pollock, Bhabha, and Chakrabarty ed., Cosmopolitanism; and the additional references listed above in the introduction, pp. 22–23. On the debates generated by the use of the concept in medi­eval Western contexts, see Ganim, ‘Cosmopolitanism and Medi­evalism’ and, for the Persian-speaking world, Bashir, The Market in Poetry in the Persian World, p. 5. Bashir’s critique (based on Mignolo, Local Histories/Global Designs) focuses on the implied ‘built-in elitism’ of the concept; he argues that ‘positing Persian as the lingua franca of a premodern “cosmopolis”, anchored in a literary medium, can have the effect of rendering

towar ds an arab i c co smo po li s

instrument of imperial rule’8 through which the imperial state cultivates the ‘politics of difference’,9 including the ‘politics of cultural difference’10 directly relevant to this volume’s topic. We are indebted to Pollock for the most significant contribution to our understanding of cosmopolitan formation processes — his magisterial study of the ‘Sanskrit cosmopolis’. In it, Pollock offers a new heuristic framework to make sense of the articulation of language, culture, and power in premodern contexts.11 Such approaches seem particularly promising for clarifying the role of Arabic in the first Islamic empire. In what follows, I would like to test the usefulness of cosmopolitanism as a category to shed fresh light on the early Islamicate polity and to document ‘a great moment of transformation in culture and power’, to use Pollock’s phrase,12 in the first centuries of Islam. This will allow me to postulate the emergence of an ‘Arabic cosmopolis’, and the various stages of the process will highlight the tensions between language management and competing cosmopolitan strategies in the early caliphate. Cosmopolitanism refers to a range of ‘practices and ideals that enabled certain individuals not only to cross cultural boundaries but also to establish an enduring normative framework across them’. As such, cosmopolitanism offers a perfect example of ‘theoretical universalism in practice’. In concrete terms, ‘whether a conqueror, a monk, or a merchant, the cosmopolitan regards an encounter with a politically, linguistically, ethnically, and/or religiously distinct group as an opportunity to incorporate its members into their network of political, religious, or economic communities’.13 Cosmopolitanism thus transcends rather than obliterates differences. As aptly pointed out by Myles Lavan, Richard Payne, and John Weisweiler, ‘the analytical value of cosmopolitanism for the study of ancient empires resides in its nonparticipants uncosmopolitan, and hence, lesser beings in terms of cultural sophistication’. I am indebted to Pamela Klasová for this reference. 8 Lavan, Payne, and Weisweiler, ‘Cosmopolitan Politics’, p. 1. 9 Burbank and Cooper, in Empires in World History, p. 8, contend that ‘the concept of empire presumes that different peoples within the polity will be governed differently’. 10 Lavan, Payne, and Weisweiler, ‘Cosmopolitan Politics’, p. 9. 11 Pollock, The Language of the Gods in the World of Men. On the reception of Pollock’s book, see, in particular, Ricci, Islam Translated, and Gould, ‘How Newness Enters the World’. Pollock’s work has proved particularly influential among East Asianists, who have coined the concept of a ‘Sino­graphic Cosmopolis’; see King, ‘Introduction’, p. 2 and ‘Ditching “Diglossia”’. King’s forthcoming edited volume, The Language of the Sages in the Realm of Vernacular Inscription, suggests that the discussion is ongoing. Historians of the Ancient Near East have also fruitfully engaged with Pollock’s work, see most notably Van De Mieroop, Before and After Babel. For a more critical reception of Pollock’s work among scholars of South and Southeast Asia, see, in particular, Ali, ‘The Early Inscriptions of Indonesia and the Problem of the Sanskrit Cosmopolis’, where Ali points out Pollock’s lack of engagement with theories of state formation in this region. Ali also makes the case (at pp. 281–82) for a ‘socio­logical complement’ to Pollock’s ‘Sanskrit Cosmopolis’, with an emphasis on networks and social circulation. For the purpose of this paper, I leave aside Ollett’s important discussion on ‘language order’ that aims to reevaluate the relationship between Prakrit and Sanskrit in premodern India. In so doing, Ollett supplements and sometimes amends Pollock’s model of a ‘Sanskrit Cosmopolis’. See Ollett, Language of the Snakes. 12 Pollock, The Language of the Gods in the World of Men, p. 1. 13 Lavan, Payne, and Weisweiler, ‘Cosmopolitan Politics’, p. 10.

169

170

a n to in e bo r r u t

focus on the intersubjective encounters of imperial elites with their subordinates in the cultural domains of language and literature, religion, and identity’.14 But if cosmopolitanism has been closely linked to imperial rule,15 not all cosmopolitan formations necessitated empire: ‘monks and merchants could construct cosmopolitan networks as expansive and enduring as their imperial counterparts’.16 One can think here of Buddhists,17 Syriac Christians,18 or Sogdian merchants,19 to name but a few non-imperial cosmopolitans. The best example of the emergence of a cosmopolis independent of an imperial setting has been offered by Pollock. Indeed, ‘unlike the comparable trans-regional expansion of Greek, Latin, Chinese, and Arabic literary cultures, the Sanskrit cosmopolis emerged without empire’.20 The Sanskrit Cosmopolis

In 2006, Pollock published his groundbreaking book The Language of the Gods in the World of Men: Sanskrit, Culture, and Power in Premodern India in which he sought to document two ‘great moments of transformation in culture and power in premodern India’. Pollock located the first one around the start of the Common Era, ‘when Sanskrit, long a sacred language restricted to religious practice, was reinvented as a code for literary and political expression’. The dominance of Sanskrit lasted for about a millennium, until local languages acquired a new literary status, thus challenging and eventually replacing Sanskrit.21 In other words, Pollock describes a thousand years of a Sanskrit cosmopolis followed by a vernacular millennium. Pollock tells us that the expression ‘Sanskrit cosmopolis’ carries three key components: The first is its supraregional dimension (‘cosmo-’), which directs attention toward the expansive nature of the formation. The second is the prominence given to the political dimension (‘-polis’), which was of particular importance in this form of global identification. Last, the qualification provided by ‘Sanskrit’

14 Lavan, Payne, and Weisweiler, ‘Cosmopolitan Politics’, p. 11. 15 Not all empires did facilitate cosmopolitan practices though, see Van De Mieroop, Before and After Babel. 16 Lavan, Payne, and Weisweiler, ‘Cosmopolitan Politics’, p. 11. 17 Monius, Imagining a Place for Buddhism. While Monius’s primary focus is on South India, Buddhist networks were also particularly active in Central Asia, thus bridging the gap with China and, eventually, the Islamic world. See, in particular, Hansen, The Silk Road; Elverskog, Buddhism and Islam on the Silk Road; Akasoy, Burnett, and Yoeli-Tlalim ed., Islam and Tibet. I leave aside the most extreme arguments made by Beckwith, Warriors of the Cloisters, which have generated a fair deal of skepticism. 18 The scholarship on the topic is considerable and impossible to summarize here. Among recent contributions, see, most notably, Briquel Chatonnet and Debié, Le monde syriaque. For the role of East Syrian Christians in Iranian imperial history, see Payne, A State of Mixture. 19 La Vaissière, Histoire des marchands sogdiens; Hansen, The Silk Road, pp. 113–39. 20 Lavan, Payne, and Weisweiler, ‘Cosmopolitan Politics’, pp. 11–12. 21 Pollock, The Language of the Gods in the World of Men, p. 1.

towar ds an arab i c co smo po li s

affirms the role of this particular language in producing the forms of cultural and political expression that underwrote this cosmopolitan order.22 From this perspective, philo­logy ‘was inextricably tied to the practices of power’. An elaborate network of patronage fostered language cultivation by supporting ‘grammarians, lexico­graphers, metricians, and other custodians of purity, and through endowments to schools for the purpose of grammatical studies’. It is in this context that the most influential Sanskrit grammars were commissioned.23 The outcome was that ‘power in India now had a Sanskrit voice’.24 Sanskrit’s supremacy was eventually contested around the turn of the first millennium ce. This shift is referred to as vernacularization, ‘the historical process of choosing to create a written literature, along with its complement, a political discourse, in local languages’.25 Through such a process, vernacular languages end up filling ‘all the available cultural space, their expansion as literary-political media limited only by other cosmopolitan cultural formations’.26 Vernacularization implies major social and cultural transformations27 and raises tantalizing questions in Islamicate contexts.28 I will leave aside the question of the most obvious vernacularization processes — the rise of New Persian29 and of literary Turkish,30 ‘cosmopolitan vernaculars’31 that would eventually succeed as new cosmopolitan formations while adopting the ‘Arabic cosmopolitan script’32 — because they largely fall outside of the chrono­logical boundaries of this paper.

22 Pollock, The Language of the Gods in the World of Men, p. 12. 23 Pollock, The Language of the Gods in the World of Men, p. 15. 24 Pollock, The Language of the Gods in the World of Men, p. 122. 25 Pollock, The Language of the Gods in the World of Men, p. 23. 26 Pollock, The Language of the Gods in the World of Men, p. 21. Pollock’s view of vernaculars taking over cosmopolitan languages in a ‘zero-sum-game’ has been challenged, notably because it underestimates the significance of diglossia and multilingualism. See Orsini, ‘The Multilingual Local’, pp. 355–56. I am indebted to Naveena Naqvi for this last reference. 27 A particularly illuminating example has been offered by Spiegel for the medi­eval West. See her Romancing the Past, where she shows that an anxious aristocracy, confronted by social changes and political challenges, prompted the emergence of vernacular historio­graphy. 28 For a useful discussion, see Yücesoy, ‘Language of Empire’. 29 The key role of the Samanids (204–395/819–1005) is highlighted by Yücesoy in ‘Language of Empire’. See also Orsatti, ‘Persian Language in Arabic Script’ and, more broadly, Zadeh, The Vernacular Qurʾan. See also Khodadad Rezakhani’s paper in this volume. 30 The most substantial discussion is offered by Peacock, Islam, Literature and Society in Mongol Anatolia, pp. 147–87, who notably underlines the limitations of Pollock’s vernacularization model with regard to Turkish, pp. 150–51. See also Schmidt, ‘How to Write Turkish?’. 31 The concept was first introduced by Pollock in ‘The Cosmopolitan Vernacular’. See also his ‘Cosmopolitan and Vernacular in History’ and The Language of the Gods in the World of Men, passim. 32 Ricci, in Islam Translated, p. 170, emphasizes the ‘far-reaching cultural implications’ of the adoption of a new script. It is, indeed, impossible to overestimate the significance of the adoption of the Arabic script by these ‘cosmopolitan vernaculars’ since, as pointed out again by Ricci (at p. 154), it ‘allowed Muslims from diverse cultures to share not just a faith but also the system of writing and the heritage it carried with it’. On the widespread use of the Arabic script across literary traditions, see, most recently, Bondarev, Gori, and Souag ed., Creating Standards.

17 1

172

a n to in e bo r r u t

It is important to note, however, that an earlier stage of vernacularization concerns us more directly. Indeed, vernaculars necessarily precede the formation of cosmopolitan languages. But it is only under specific circumstances that a vernacular language can become cosmopolitan. This was the case with Sanskrit, as documented by Pollock, and with Latin, which he uses as an example of a language at the heart of a counter-cosmopolis throughout his book. Vernacular Arabic and Umayyad Policies

With the examples of Sanskrit and Latin in mind, we have fascinating options to consider with regard to the rise of Arabic and its possible evolution from a vernacular to a cosmopolitan language. Epi­graphy reveals that significant linguistic transformations were unfolding both on the eve of Islam and in early Islamic times. Pollock suggests two possible models. The first is one of a local vernacular such as Latin, which was originally restricted to the Latium and then underwent a massive expansion in the footsteps of Roman soldiers embarking on world conquest.33 The case of Sanskrit, by contrast, offers a radically different model: Pollock convincingly shows that Sanskrit was an already transregional vernacular that turned into a cosmopolitan formation without any imperial support.34 From an imperial perspective, the Latin model appears at first sight to be a good fit for the case of Arabic. Such a scenario would imply that the vernacular was initially confined to a specific area, here the Hejaz, before engendering a cosmopolitan formation in the wake of expansion. However, in our current state of knowledge there is, in fact, some evidence to support a scenario rather close to that of the Sanskrit formation (without ignoring the formidable impetus provided by the early Islamic empire and thus the need to somehow combine both models in the case of Arabic). Use of the Arabic language is epi­graphically documented ‘long before the emergence of the script that we call Arabic, […] in many different dialects’, and Arabic inscriptions are occasionally found in ‘the Arabian and Syrian deserts’.35 (We have such inscriptions written in Safaitic, Hismaic, and possibly in Ancient South Arabian scripts, as well as in Nabataean and Greek scripts).36 Texts

33 Pollock, The Language of the Gods in the World of Men, p. 20. 34 Pollock, The Language of the Gods in the World of Men, pp. 20, 262. 35 Hoyland, ‘The Birth of Arabic in Stone’, p. 53. 36 Hoyland points out (‘The Birth of Arabic in Stone’, p. 53) that ‘on the rare occasions that someone wrote a text in Arabic before the rise of Islam, the script of prestige in the locality concerned was employed’, which explains the variety of scripts used. That being said, the ‘Nabatean Aramaic script’ spread and became the most common script ‘used for writing Arabic in these early days’. The influence of Arabic gradually overshadowed the Aramaic component, and it seems that by the sixth century, the Nabatean script had fully transformed into what we call the Arabic script. See Nehmé, ‘A Glimpse of the Development of the Nabataean Script into Arabic’ and ‘Aramaic or Arabic?’; Iema, Al-Jallad, Macdonald, and Nehmé, ‘Provincia Arabia’. Al-Jallad points out that Nehmé situates this gradual shift from the Nabatean to the Arabic script at the ‘courts of tribal chiefs in Northwest Arabia’, which could suggest that Arabic was, in this context, a langue courtoise as defined by Grévin,

towar ds an arab i c co smo po li s

employing the Arabic script seem to emerge in the late fifth century and become more frequent in the course of the sixth century. The spread of the Arabic script is restricted to two distinct areas: the region of Najrān in the Arabian Peninsula, and various locations in Syria.37 In a recent survey, Robert Hoyland notes that these inscriptions point to the ‘emergence of Arab client kingdoms’ and observes that ‘all of these pre-Islamic Arabic inscriptions have a connection with Christianity’, thus revealing missionary efforts, especially at the urging of ‘Syriac Christian church authorities who were involved with an emergent Arab Christianity’.38 To borrow Grévin’s termino­logy again, this suggests that Arabic was being used as langue véhiculaire by Christian missionaries,39 and Christians and Christian institutions indeed seem to have played a central role in the spread of Arabic in the late antique Near East.40 The sixth century saw the continuing ascent of Arabic’s prestige, best exemplified by the ‘the first ever monumental Arabic inscription’ in 568 in Ḥarrān, in the lava fields of the Syrian Lajā. The trend is reinforced by the revelation of the Qur’an, leading to ‘the flowering of Arabic writing in the seventh century’.41 The bureaucracy, however, continued to function in Greek and Middle Persian: Hoyland laments the dearth of Arabic documents prior to the final decade of the seventh century ce, save for ‘a handful of graffiti, almost all of these from west

37

38 39 40

41

Le parchemin des cieux. See Al-Jallad, ‘The Di­graph ‫ اى‬in the Qur’anic Consonantal Text’, p. 9 n. 43 (citing L. Nehmé, ‘Epi­graphy on the Edges of the Roman Empire’ [non vidi]). Al-Jallad also notes, however, that ‘the establishment of Arabic as the language of an empire in the seventh century produced a homogenizing bottleneck, narrowing much of the variation found in pre-Islamic sources’. He identifies evidence of ‘script contact’ situations that suggest that we need to be more sensitive ‘to the diversity of the Arabic script in the 6th century ce’. See Al-Jallad, ‘The Di­graph ‫ اى‬in the Qur’anic Consonantal Text’, pp. 2 and 8. Hoyland, ‘The Birth of Arabic in Stone’, p. 59. The corpus examined by Hoyland is composed of twelve texts from the Najrān area and three from Syria, more specifically from Jabal Says, Zabad (south-east of Aleppo, dated 512 ce), and Ḥarrān (south of Damascus, dated 568 ce). The latter two inscriptions are either bilingual or multilingual. The date given by Hoyland for the Jabal Says inscription, 528 ce, should now be corrected to 532–33 ce, as the reading has been revised to 427 of the era of the province of Arabia, instead of 423 as previously thought. See Robin, ‘L’Arabie à la veille de l’Islam dans l’ouvrage d’Aziz al-Azmeh’. A significantly larger corpus of inscriptions from the Najrān area is presented and discussed by Robin, Najrān en Arabie et ses martyrs chrétiens. Hoyland, ‘The Birth of Arabic in Stone’, pp. 60, 62. On the possible influence of Christian missionaries on the development of the Arabic script, see Robin, ‘La réforme de l’écriture arabe à l’époque du califat médinois’, p. 329. The story of how Arabic evolved from langue véhiculaire to langue référentielle largely remains to be told. The challenge is not specific to Arabic; see Pollock’s remarks on the adoption of Sanskrit by Buddhists and Jains in The Language of the Gods in the World of Men, pp. 56–59. See the contributions of Aaron Butts and Muriel Debié to this volume and Iema, Al-Jallad, Macdonald, and Nehmé, ‘Provincia Arabia’, pp. 429–30. The absence of Arabic Bibles has often been pointed out as a challenge to this theory, but it may not be one if considered through the lens of cosmopolitanism. Indeed, as shown by Pollock, ‘aesthetic’ choices were central in language selection, and so a langue référentielle was likely to be privileged for sacred texts. See the discussion below about the absence of Greek fragments of the New Testament in the Damascus manu­scripts of the Qubbat al-Khazna, pp. 180–81. Hoyland, ‘The Birth of Arabic in Stone’, p. 64.

17 3

1 74

a n to in e bo r r u t

Figure 3.1. Muʿāwiya’s dam near Ṭāʾif, 58 ah/677–78 ce. Photo­graph courtesy of Dr Saad bin Abdulaziz al-Rāshid. Reproduced with permission.

Arabia, beginning in 24/642’.42 The almost complete absence of monumental inscriptions in Arabic is likewise conspicuous, with the notable exception of the dam inscriptions of Muʿāwiya in the Hejaz (see Fig. 3.1).43 To this meagre list, we can add ‘a tombstone from Egypt dated 31/652, a graffito from Iraq dated 65/685, and a moderate number of Arabic papyri from Egypt and southern Palestine, beginning as early as the 20s/640s [that] tell us that Arabic was already in use across the Middle East’.44 In other words, pending new discoveries, we can provisionally conclude that Arabic seems to have been translocal, even if the emergence of the Arabic script itself should probably be placed in north-western Arabia.45 The rapid success of the early Islamic empire undoubtedly reinforced the spread of Arabic in a fruitful combination of the two scenarios envisioned by Pollock. From an epi­graphic perspective, the situation changes dramatically during the reign of ʿAbd al-Malik. The 70s/690s mark ‘the deployment of a more polished form of Arabic script […], often called the imperial Umayyad style’,46 perhaps primarily as a result of the intense competition for cultural authority between the Zubayrids and the Marwanids. In any case, we witness ‘the transformation of the social life’ of Arabic: the epi­graphic record shows that Arabic starts being employed

42 Hoyland, ‘The Birth of Arabic in Stone’, p. 65. I leave aside the complex discussion on the identification of the calendar(s) in use in early Islam; on this, see, notably, Tillier and Vanthieghem, ‘Recording Debts in Sufyānid Fusṭāṭ’, pp. 148–88. 43 Miles, ‘Early Islamic Inscriptions near Ṭāʾif in the Ḥijāz’; al-Rāshid, ‘Sadd al-Khanaq’. 44 Hoyland, ‘The Birth of Arabic in Stone’, p. 65. 45 Iema, Al-Jallad, Macdonald, and Nehmé, ‘Provincia Arabia’, pp. 395 ff., esp. pp. 398–99, 429. 46 Hoyland, ‘The Birth of Arabic in Stone’, p. 65.

towar ds an arab i c co smo po li s

‘in hitherto unprecedented ways’,47 best exemplified by a burst of Umayyad caliphal inscriptions in Syria.48 With ʿAbd al-Malik’s currency reform, coinage immediately follows suit, though not evenly throughout the caliphate.49 Whereas Chase Robinson was arguably correct to depict the Marwanids as ‘linguistic imperialists’,50 the trend reveals — to paraphrase Pollock in Arabic terms — the emergence of ‘a new cultural-political formation’, an Arabic ‘cosmopolitan formation [that] was on the point of being invented’,51 rather than a straightforward process of Arabicization. For a cosmopolis to emerge fully, however, epi­graphy is not enough. Neither is the rise of documentary texts characteristic of the Umayyad period, primarily papyri from Egypt and, to a lesser extent, from Palestine. Literature is required as well; indeed, Pollock contends that epi­graphy and literature are the two key components of any cosmopolitan formation, ultimately leading to the constitution of a formidable corpus of texts, in our case Arabic texts.52

Textual Foundations and Cultural Difference in the First Arabic Cosmopolis Conversion, Translation, and ‘Prior Texts’

At this juncture, we need to introduce a couple of important ideas developed by Ricci. Inspired by Pollock’s pioneering work, she has explored the notion of an ‘Arabic cosmopolis’ in early modern South and Southeast Asia that largely overlapped with the territories of the former ‘Sanskrit cosmopolis’.53 Ricci put forward the notion of ‘literary networks’ and highlighted the close connection between two phenomena of conversion and translation. Her goal was to shed light on how ‘societies in transition, undergoing a profound change such as Islamization, gradually amass for themselves the textual sources allowing an engagement with, and commitment to, a history only recently adopted’.54

47 Pollock, The Language of the Gods in the World of Men, pp. 39, 73. 48 It is impossible to cite here all the relevant inscriptions. For a convenient discussion of building inscriptions, milestones, and a list of mentions of Umayyad caliphs in monumental inscriptions, see Lindstedt, ‘Arabic Rock Inscriptions up to 750 ce’, pp. 418–19, 421, and 428. Lindstedt also provides a comprehensive biblio­graphy. See also the recent contribution of Garosi, ‘Imperial Arabic’. I am indebted to Marie Legendre for bringing this last piece to my attention. 49 Heidemann, ‘The Evolving Representation of the Early Islamic Empire and its Religion on Coin Imagery’. 50 Robinson, ‘Abd al-Malik, p. 126. 51 Pollock, The Language of the Gods in the World of Men, p. 74. 52 Pollock, The Language of the Gods in the World of Men, p. 19. The point is also central to Ricci’s argument in Islam Translated. 53 As far as I know, Ricci’s Islam Translated is the most significant attempt to use Pollock’s work in the definition of an Arabic cosmopolis. 54 Ricci, Islam Translated, p. 245.

17 5

176

a n to in e bo r r u t

The constitution of such a corpus is particularly problematic in the absence of pre-existing layers of what another South Asianist, the linguist Alton Becker (d. 2011), has termed ‘prior texts’.55 Accordingly, Ricci asks: How does a society, in the face of such a significant change as the conversion to a new religion, address the absence of prior text and memory, which are both so important in creating and maintaining a shared identity? If prior texts are, by definition, old and familiar, the challenge of assembling them to fill a void for a society transformed by conversion would seem daunting. How are texts newly created for this purpose, and how are they established so that they, in turn, come to figure as prior texts?56 The answer lies in the fact that ‘new prior texts’ can be generated through two main strategies, which are usually not mutually exclusive. The first option is simply ‘the reformulation of old texts’, while the second tactic is ‘the creation of new ones, often through translation’.57 In our case, the ‘reformulation of old texts’ is perhaps best exemplified by pre-Islamic poetry and genealogies that became much-needed prior texts, thus prompting a fresh ‘construction of the Jāhiliyya’ that provided ‘cultural authority in the making’, as once noted by Rina Drory.58 Obviously, there is much to be said about the rise of Arabic literature already in Umayyad times, before Abbasid-era efforts.59 The topic has sometimes been limited to Umayyad-era poetry and to the epistolary genre, personified by ʿAbd al-Ḥamīd al-Kātib (d. 132/750)60 and possibly by caliphal letters,61 and much remains to be done to re-evaluate cultural life under the first dynasty of Islam.62 Be that as it may, a formidable Arabic literary corpus emerged in Umayyad and early Abbasid times. Part of it was self-generated in Arabic (e.g., poetry, genealogy, ḥadīth), but a sizeable amount was also the outcome of the second tactic outlined above through the ‘translation movement’, understood here as a process that begun in Umayyad times, following Saliba’s chrono­logy.63 The massive translating efforts undertaken

55 Becker, Beyond Translation, especially pp. 285 ff. 56 Ricci, Islam Translated, pp. 245–46. 57 Ricci, Islam Translated, p. 246. 58 Drory, ‘The Abbasid Construction of the Jāhiliyya’. See also Webb, ‘Al-Jāhiliyya’ and Imagining the Arabs; and, more broadly, Schoeler, The Genesis of Literature in Islam. 59 For a generous (albeit somewhat dated) survey of Arabic literature in Umayyad times, see Beeston, Johnstone, Serjeant, and Smith ed., Arabic Literature to the End of the Umayyad Period. See also the stimulating remarks of Schoeler in The Genesis of Literature in Islam, especially ch. 5. 60 On ʿAbd al-Ḥamīd’s letters, see the edition and detailed introduction of ʿAbbās, ʿAbd al-Ḥamīd b. Yaḥyā al-Kātib. The classic study remains al-Qāḍī, ‘Early Islamic State Letters’; see also her entry ‘ʿAbd al-Ḥamīd al-Kātib’. 61 See Robinson’s remarks in his ʿAbd al-Malik, pp. 106–13. 62 A point already made in the late 1930s by Mackensen, ‘Supplementary Notes to “Arabic Books and Libraries in the Umayyad Period”’. See also Eche, Les bibliothèques arabes publiques et semi-publiques, pp. 11–20; and my own discussion in Borrut, ‘An Islamic Late Antiquity?’. 63 See also the relevant chapters in Beeston, Johnstone, Serjeant, and Smith ed., Arabic Literature to the

towar ds an arab i c co smo po li s

during the first three or four centuries of Islam reveal a decisive step towards the formation of an Arabic cosmopolis. Both options are notably achieved through the process of ‘languaging’,64 in which terms and notions are selected and transformed as part of the construction of a new cultural memory. In the context of translations, languaging takes the form of ‘places, flora, fauna, food, music, dress, and other local phenomena, [through which] the new text [develops] certain ties with a known past or present so that audiences may relate to it’.65 Languaging is not restricted to translations: a similar logic guided Umayyad efforts to ‘Qur’anicize’ the terms used for central notions. The trend was first detected by Fred Donner, who pointed out that ‘during the Umayyad period, institutions and practices central to the operation of the state were renamed using terms from the Qur’an’.66 More broadly, the massive number of translations undertaken in early modern South and Southeast Asia, just like in the first centuries of Islam, reveals how ‘communities went about the formation of a new literary and religious repository that was richly interconnected both with a distant past and with a local present’.67 As far as early Islam is concerned, one of the most significant contributions of the translation movement was to participate to the creation of an expansive corpus of prior texts in a new cosmopolitan language. Those new prior texts played a central role in ‘the task of restructuring world knowledge to better define caliphal and communal identity’.68 Translation can thus be understood ‘as a creative act, participating in the creation of knowledge and, by extension, of power’.69 Foreign works also provided ‘inspiration for introducing new themes and literary models’.70 The role of Wahb b. al-Munabbih (d. 110/728 or 114/732), Ibn al-Muqaffaʿ (d. 139/756), and other such cultural brokers was, therefore, of paramount importance.71 It is impossible to overestimate the significance of Sanskrit, Greek, Middle Persian, Syriac, and other literatures that became enshrined in the Arabic End of the Umayyad Period, in particular Goodman, ‘The Greek Impact on Arabic Literature’, and Bosworth, ‘The Persian Impact on Arabic Literature’. 64 Becker, Beyond Translation, esp. pp. 286–88 and passim. 65 Ricci, Islam Translated, p. 248. 66 Donner, ‘Qur’ânicization of Religio-Political Discourse in the Umayyad Period’, p. 79. Donner adduced several examples, ranging from khalīfa to jihād, ribāṭ, qāḍī, fatḥ, fitna, and others. 67 Ricci, Islam Translated, p. 245. 68 Yücesoy, ‘Language of Empire’, p. 386. Yücesoy correctly sees this as ‘an act of conquest’: ‘Approached from this perspective, translation was an intercultural and inter-imperial political confrontation intended to enable the Abbasids to absorb and remold ancient knowledge and cultural traditions so that they could define, position, and deploy their own ideo­logical, cultural, artistic, and literary taste in describing the world anew’. 69 Gentzler, ‘Translation, Poststructuralism, and Power’, p. 216, cited in Yücesoy, ‘Translation as SelfConsciousness’, p. 533. 70 Ricci, Islam Translated, p. 247. 71 De Prémare, ‘Wahb b. Munabbih, une figure singulière du premier islam’; Khoury, ‘Wahb b. Munabbih’; Kristó-Nagy, La pensée d’Ibn al-Muqaffaʿ. See also Yücesoy, ‘Translation as Self-Consciousness’, p. 525, who aptly notes that ‘ancient scientific knowledge had to undergo a process of legitimization to take root in its new context’.

17 7

1 78

a n to in e bo r r u t

literary tradition along the way.72 As the Arabic cosmopolis was coming into being, cosmopoleis came into contact: Sanskrit, Hellenism, and Arabic provide clear evidence of ‘connected histories’.73 This is where literary networks come into play. In Ricci’s words: Literary networks connected Muslims across boundaries of space and culture, and they helped introduce and sustain a complex web of prior texts and new interpretations that were crucial to the establishment of both local and global Islamic identities. Literary networks were comprised of shared texts, including stories, poems, genealogies, histories, and treatises on a broad range of topics, as well as the readers, listeners, authors, patrons, translators, and scribes who created, translated, supported, and transmitted them.74 In other words, the dissemination of texts was key to the making of the ‘cosmopolis as a cultural political space’.75 Pollock has shown that it was precisely the circulation of courtly epics, histories, poetry, and grammar that facilitated the definition and replication of literary standards and models throughout the Sanskrit cosmopolis.76 Ricci asserts that ‘treatises on law, mysticism, grammar, and ritual’ performed an analogous role across the early modern Arabic cosmopolis ‘along with the corresponding Arabicized literature that produced multiple echoes of the cosmopolitan language’.77 Patronage and court culture played an especially significant role in that direction. More broadly, Ricci is right to think ‘of translation and conversion as mutually constitutive processes’, for ‘through translation, communities gradually created, adopted, and accumulated the cultural resources that made memories of an Islamic past and a lived Islamic present possible’.78 A perfect illustration of this process in the early period is offered by Sarah Savant’s work on post-conquest Iran, where ‘new Muslims’ had to build fresh links with the past to make sense of their deeply transformed present.79 These new links were developed in Arabic, the new 72 Just like Arabic and ‘Arabicized’ terms would eventually nurture other literary traditions; see, for instance, Ricci’s rich discussion in her Islam Translated, especially pp. 15–18. 73 The scholarship on ‘connected histories’ is quite extensive. See, in particular, Subrahmanyan, ‘Con­ nected Histories’; Werner and Zimmermann ed., De la comparaison à l’histoire croisée and ‘Beyond Comparison’; Douki and Minard, ‘Histoire globale, histoires connectées’. The need for ‘integrative’ history (as opposed to ‘comparative history’) has also been stressed by Lavan, Payne, and Weisweiler, ‘Cosmopolitan Politics’, pp. 3–4. For a broader discussion of interconnected history in Eurasian Late Antiquity, see Di Cosmo and Maas ed., Empires and Exchanges in Eurasian Late Antiquity. 74 Ricci, Islam Translated, pp. 1–2. 75 Ricci, Islam Translated, p. 261. 76 Pollock, The Language of the Gods in the World of Men, p. 255. 77 Ricci, Islam Translated, p. 261. 78 Ricci, Islam Translated, p. 271. See also Yücesoy, ‘Translation as Self-Consciousness’, p. 525, who suggested the drawing of parallels ‘between the ideo­logies of conquest in early Islam and the translation movement as an act not only to retrieve ancient wisdom from its historical, linguistic, and cultural repositories but also to reclaim, repossess, and reuse it in a new context’. The close connection between ‘language shift and conversion’ was also recently highlighted by van Bladel, ‘Arabicization, Islamization, and the Colonies of the Conquerors’, p. 112. 79 Savant, The New Muslims of Post-Conquest Iran.

towar ds an arab i c co smo po li s

triumphant cosmopolitan language acting as a ‘unifying and standardizing force across the cosmopolis’.80 Cultural integration was a major tool ‘toward the creation of a normative framework shared by universal rulers and local elites’.81 And so, alongside translation (and informed by it), a normative framework was established, first and foremost in the form of a ‘cosmopolitan grammar’.82 Here the role of Sībawayhi (d. 180/796) cannot be overestimated. His grammar is constitutive of the Arabic cosmopolis and so is the foundational work of many more grammarians.83 Moreover, Sībawayhi and other non-Arab converts offer a poignant testimony to the success of the Arabic cosmopolis: ‘the first bio­graphy of Muhammad was by Ibn Ishaq (d. 767), the grandson of a prisoner-of-war from ʿAyn al-Tamr in Iraq; the earliest surviving Qurʾanic commentary was by Muqatil ibn Sulayman (d. 767), born of Persian parents; and the first Arabic grammar was by Sibawayhi (d. 796), a native of Balkh in modern Afghanistan’.84 The universalist dimension of cosmopolitanism is unmistakable. Subordination, Assimilation, and Manu­script Culture

The formation of an Arabic cosmopolis was obviously a long-term process that notably required assessing ‘the political importance of cosmopolitan practices’.85 Indeed, cosmopolitanism implied managing difference via two main competing cosmopolitan strategies, assimilation and subordination: ‘assimilation works by eliding the cultural difference between universal rulers and local elites, whereas subordination operates by recognising, preserving and organising difference’.86 As far language is concerned, what was at stake was the management of different linguistic traditions in a multilingual empire. The rival strategies of assimilation and subordination tend towards opposite approaches. When assimilation is favoured, ‘the imperial state speaks in a universal rather than local idiom’. Subordination, on the other hand, manages ‘difference through categorization’ and ‘ethno­graphic capital’: ‘Imperial interaction with peripheral communities tends to be cast in local rather than universal idioms, inscribing empire into multiple local traditions rather than attempting to speak

80 Ricci, Islam Translated, p. 182. 81 Lavan, Payne, and Weisweiler, ‘Cosmopolitan Politics’, p. 5. 82 Pollock, The Language of the Gods in the World of Men, p. 189. 83 For an accessible introduction to Sībawayhi, see Carter, Sībawayhi and his entry ‘Sībawayhi’. On the reception and legacy of Sībawayhi’s Kitāb, see Bernards, Changing Traditions and Baalbaki, The Legacy of the Kitāb. On the development of Arabic grammar more broadly, see the relevant chapters in Orfali ed., In the Shadow of Arabic. 84 Hoyland, ‘The Birth of Arabic in Stone’, p. 67. The significant proportion of mawālī among early Islamic-era linguists has been highlighted by Bernards, ‘The Contribution of Mawālī’ and ‘Pioneers of Arabic Linguistic Studies’. 85 Lavan, Payne, and Weisweiler, ‘Cosmopolitan Politics’, p. 2. 86 Lavan, Payne, and Weisweiler, ‘Cosmopolitan Politics’, p. 1.

17 9

1 80

a n to in e bo r r u t

with one voice to the entire subject population’.87 That being said, the tactics of assimilation and subordination do not have to be mutually exclusive, and empires could also rethink their strategy along the way, particularly by moving from subordination to assimilation, as illustrated by the Roman Empire.88 The tensions between the two policies are useful for understanding strategies of domination and the articulation of culture and power in early Islam more broadly. The manu­script tradition offers concrete evidence of the different choices and rhythms of the tactics of assimilation and subordination. There are substantial indications that the early Islamic empire originally opted for subordination rather than assimilation, though its strategies could change over time and space. The example of Damascus’s Qubbat al-Khazna is particularly telling from this perspective.89 The Qubbat al-Khazna has been described as an Islamic geniza90 containing numerous texts in Arabic but also in Greek, Syriac, Armenian, Georgian, Coptic, Hebrew, Samaritan, Aramaic, Latin, and even old French. As Arianna D’Ottone has put it, the manu­scripts of the Qubbat al-Khazna offer a window into ‘a multilingual and multicultural society’.91 The Qubbat al-Khazna documents demonstrate ‘the continuous presence of Greek Christian texts in the Damascus region from the fourth to the tenth century’. It is only at the turn of the third/ninth and fourth/ tenth centuries that, ‘after more than a millennium of intellectual supremacy, Greek ceases to be the prevailing script in the region’.92 This transition can be read as evidence of the triumph of the Arabic cosmopolis overcoming the pre-existing Greek/Hellenistic cosmopolis.93 The process is obviously not linear, and more importantly, the lines of division cannot be reduced to religious ones. Thus, in the Damascus find, ‘it

87 Lavan, Payne, and Weisweiler, ‘Cosmopolitan Politics’, p. 6. 88 Lavan, ‘“Father of the Whole Human Race”’, p. 153. Burbank and Cooper have also highlighted the fact that ‘empires’ durability depended to a large extent on their ability to combine and shift strategies’ and to draw on to various repertoires of power; Burbank and Cooper, Empires in World History, p. 16. 89 On this major but understudied corpus see now D’Ottone Rambach, Hirschler, and Vollandt ed., The Damascus Fragments. The name ‘Qubbat al-Khazna’ is a fairly recent one to refer to what was originally known as the Qubbat al-māl, an Abbasid-era structure erected in the courtyard of the Umayyad mosque of Damascus by the governor al-Faḍl b. Ṣāliḥ b. ʿAlī (d. 172/788). See Aljoumani, ‘Taʾrīkh Qubbat al-māl’. On al-Faḍl, see Cobb, White Banners, pp. 27–28 and 151 n. 26. 90 Since Bruno Violet spoke of the ‘Damascener Moschee-Genisah’ in his pioneering work, ‘Ein zweisprachiges Psalmfragment aus Damaskus’, p. 384. 91 D’Ottone, ‘Manu­scripts as Mirrors of a Multilingual and Multicultural Society’. The reasons behind this multilingual and multiscriptural profile remain largely unclear, and there are formidable questions raised by the ‘life cycles’ of such documents; see D’Ottone Rambach, Hirschler, and Vollandt, The Damascus Fragments, pp. 11 ff. On the broader issue of medi­eval archives in the Islamicate world, see Hirschler, ‘From Archive to Archival Practices’; Paul, ‘Archival Practices in the Muslim World prior to 1500’; and Rustow, The Lost Archive. 92 D’Ottone, ‘Manu­scripts as Mirrors of a Multilingual and Multicultural Society’, pp. 70, 73. 93 A similar chrono­logy has been highlighted for the adoption of Arabic in the Melkite liturgy by Rubin, ‘Arabization versus Islamization in the Palestinian Melkite Community’, p. 153.

towar ds an arab i c co smo po li s

is worth noting that no Greek minuscule fragment of the New Testament was found — a detail that reveals that by that time Christian communities had chosen to use Arabic as their liturgical (and thus Scriptural) language instead of Greek’.94 This point is consistent with Pollock’s observation that ‘during the cosmopolitan epoch, one chose a literary language far more often for aesthetic than for theo­logical reasons’.95 Manu­script culture and the materiality96 of writing tells us something else. The fact that ‘the language shift from Greek to Arabic [occurred fully] only after the 9th–10th centuries testifies to the survival of Greek-speaking communities’,97 but it also suggests that Hellenophones remained in a subordinate position up to that point. Petra Sijpesteijn and Marie Legendre forcefully make the case in this volume that non-Arabic documents were produced in Umayyad chanceries, a clear sign of subordination practices. This is a good example of a deliberate effort to continue local administrative traditions — and even create new ones in the case of Coptic — instead of imposing a single voice on the ‘subject population’ throughout the caliphate.98 The ‘thoughtful bilingualism’ (bilinguisme réfléchi) on the walls of Quṣayr ʿAmra noted by Frédéric Imbert perhaps reflects a similar logic.99 Moreover, ‘the Damascus fragments […] do not exhibit the same language “intimacy” that has been observed for Andalusian Spain’, for instance. ‘Instead, the uses of different alphabets for different languages, and the large number of palimpsests, testify both to the cultural wealth of a multi-language society, and to the progressive linguistic adaptation and eventual change’.100 The manu­scripts of the Qubbat al-Khazna thus provide clear evidence of the maintenance of cultural boundaries in the early Islamic empire through typical subordination strategies. The comparison with al-Andalus is also instructive. Early Islamic rule over Iberia witnessed subordination strategies in an age of conquest, as suggested, for instance, by the minting of Latin coins and, subsequently, of bilingual Arabic/Latin coinage under the governorate of al-Ḥurr b. ʿAbd al-Raḥmān al-Thaqafī (97–100/716–19) before the production of the first strictly Arabic coins in the 720s.101 By contrast, in the fourth/tenth century the caliphs of Cordoba seem to have rather opted for assimilation tactics. As noted by Maribel Fierro, ʿAbd al-Raḥmān III and al-Ḥakam II ‘promoted the creation of an Andalusi identity that revolved around the Islamic 94 D’Ottone, ‘Manu­scripts as Mirrors of a Multilingual and Multicultural Society’, p. 71. 95 Pollock, The Language of the Gods in the World of Men, p. 99. The Damascus Psalm fragment, with its Arabic written in Greek script, might represent a transitional phase, though its date has been much debated; suggestions range from pre-Islamic times (Macdonald) to the tenth century ce (Mavroudi). Al-Jallad has recently favoured a mid-eighth-century date; see his The Damascus Psalm Fragment, pp. 35–56. 96 See the suggestive remarks of Anderson in ‘Sovereignty and the Materiality of Caliphal Encounters’. 97 D’Ottone, ‘Manu­scripts as Mirrors of a Multilingual and Multicultural Society’, p. 78. 98 Lavan, Payne, and Weisweiler, ‘Cosmopolitan Politics’, p. 6. 99 Imbert, ‘Le prince al-Walīd et son bain’, p. 347. 100 D’Ottone, ‘Manu­scripts as Mirrors of a Multilingual and Multicultural Society’, pp. 78–79. On the culture of intimacy in medi­eval Iberia, see Dodds, Menocal, and Krasner Balbale, The Arts of Intimacy. 101 See Kennedy, Muslim Spain and Portugal, p. 21.

181

1 82

a n to in e bo r r u t

religion and Arab culture, while trying to relegate Arabic [sic] ethnicity to a secondary position’.102 In other words, manu­script culture and ‘language intimacy’ can reveal cosmopolitan practices and show how linguistic choices allowed rulers to manage the politics of difference.

Conclusion: Culture and Power in Early Islam If we can identify many parallels between the Arabic cosmopolis and Pollock’s Sanskrit cosmopolis, there are also interactions and differences to consider. The presence of Indian scholars at the caliphal court and the translation of Sanskrit scholarship and literature offer the most obvious examples of cosmopoleis in contact. The trend culminated in early Abbasid times, resulting in what Kevin van Bladel aptly termed an ‘Indophile episode’.103 Such intercultural exchanges have deeper roots in Central Asia, especially in Tukharistan through the agency of the Barmakids, but they reached their peak in Baghdad, where ‘the early translations from Sanskrit into Arabic were made available through living practitioners, Indian sages for hire as consultants and who themselves could recite texts they had memorized according to the methods of Indian studies’.104 Contacts between cosmopoleis had other consequences, too, and the expansion of the Arabic cosmopolis impacted both its Sanskrit and its Latin counterparts. Pollock fittingly notes the contemporaneity of the conquests of Iberia and Sind in 92–93/711, ‘bringing new modalities of literary culture to India while in Europe disrupting older forms of cultural reproduction’. Islamic expansion thus created the preconditions for the vernacularization of both Europe and India, since ‘for the first time, the choice to think and write locally began to make better sense than writing and thinking globally’.105

102 Fierro, ‘Abd al-Rahman III, p. 97. Fierro stresses the significance of Ibn ʿAbd Rabbih’s (d. 328/940) work in this regard; on him, see Toral-Niehoff, ‘Writing for the Caliphate’. On the making of an Andalusi identity more broadly, see Martinez-Gros, L’idéo­logie omeyyade and Identité andalouse. 103 Van Bladel, ‘Eighth-Century Indian Astronomy’, p. 258. For Muslim and Hindu encounters more broadly, see Flood, Objects of Translation. 104 Van Bladel, ‘Eighth-Century Indian Astronomy in the Two Cities of Peace’, p. 289; Van Bladel, ‘The Bactrian Background of the Barmakids’. See also Thomann, ‘From Lyrics by al-Fazārī to Lectures by al-Fārābī’. 105 Pollock, The Language of the Gods in the World of Men, p. 611. The impact of Arabic on Romance vernaculars was already stressed by Menocal in The Arabic Role in Medi­eval Literary History. The insufficient reception of Menocal’s work was pointed out by Akbari, ‘The Persistence of Philo­logy’, p. 5, who noted that Menocal’s strong case for ‘the role of the Arabic language in the emergence of Roman vernacular poetics had largely fallen on deaf ears. This lack of receptivity was not due to any weaknesses in the arguments, for the textual evidence was strong; rather, it was due to an unwillingness to acknowledge the permeability of the cultural veil that separated Christian and Muslim communities on the Iberian peninsula during the Middle Ages’. Akbari further noted that Menocal’s work was slowly being situated ‘in the context of cosmopolitanism’ and advocated for further development of this trend (p. 8).

towar ds an arab i c co smo po li s

As far as differences are concerned, perhaps the most obvious one has to do with the fact that ‘Sanskrit was not diffused by a single, scripture-based religion, a condition which was clearly central to the spread of Arabic’.106 And yet we can conclude with Pollock — albeit in the singular in the case of Islam — that when the language of God entered the world of men, power shifted to an Arabic voice in Arabia and, soon, throughout the Islamicate empire. With the subsequent choices made at the court by successive caliphs and cultural brokers, ‘a new era — a cosmopolitan era — began’.107 Indeed, in both the Sanskrit and the Arabic case, ‘it was the circulation of the texts written in these languages within a particular region that produced the cosmopolis as a cultural political space’.108 And so, as argued by Pamela Klasová, there is abundant evidence to show that the Umayyads were making ‘empire through language’109 and that, as Fred Astren posits in his contribution to this volume, an ‘Islamicate mode’ of writing and publishing emerged as an integral part of an Arabic cosmopolis. Perhaps better understanding the formation of the first Arabic cosmopolis and of cosmopolitan policies can help us properly integrate multilingualism into the formative period of Islam and thus better situate nascent Islam within a late antique framework. This should force us to re-evaluate some of our categories and to challenge the logic of opposition still prevailing between Muslim and non-Muslim sources in a variety of languages. As I have argued elsewhere, the so-called gap in the (Arabic) narrative sources that we see up to the middle of the third/ninth century is, in large part, an optical illusion and a historio­graphical construct, both ancient and modern.110 Sources in languages other than Arabic are, in fact, an integral part of a cosmopolitan society. Through intercultural transmission, translation, and contacts between vernacular and cosmopolitan languages we see the critical importance of the politics of difference and of cultural diversity. And this is precisely why we should be interested in unifying the various approaches to cosmopolitanism outlined above. A unified approach allows us to see how some of these cosmopolitan policies generated cultural, social, and political tensions. The success of an Arabic cosmopolis eventually conflicted with the subordination strategies originally implemented by the caliphs. On the other hand, the opposite tactic of assimilation, favoured, for instance, in caliphal al-Andalus,

106 Ricci, Islam Translated, p. 14; see also her discussion on pp. 14–15 more broadly. Ricci also posited that the distinction between a strictly cosmopolitan age and a vernacular one was less applicable to the early modern Arabic cosmopolis since ‘Arabic and Arabicized literature are part of the same continuum and represent both a more recent cosmopolitan stage in South and Southeast Asia, when compared with Sanskrit, and a force that transformed vernaculars already endowed with literary traditions before Islamization’ (p. 261). 107 Pollock, The Language of the Gods in the World of Men, p. 13. 108 Ricci, Islam Translated, p. 261. See also Pollock, The Language of the Gods in the World of Men, p. 255: ‘It was the very circulation of texts that helped produce the cosmopolis as such’. 109 Klasová, ‘Empire through Language’. 110 Borrut, ‘An Islamic Late Antiquity?’.

183

1 84

a n to in e bo r r u t

clashed with vernacularization processes, and thus both options contributed to the political fragmentation of the once united Islamic empire.111 And yet an Arabic cosmopolis endured even if the empire did not. This is perhaps one of the most significant legacies of the Umayyad era. That cultural history could have something to say about empire, power, and specific trends such as political fragmentation should not surprise us for at least two reasons. First, as cultural historians we are simply reuniting the two components of the cosmopolis: the universal and the political. Second, and more broadly, Max Weber and Clifford Geertz have taught us that especially in premodern contexts, culture was power. And so we can agree with the Castilian grammarian Antonio de Nebrija (d. 1522), who dedicated his Castilian grammar to Queen Isabella in 1492 with the observation that language has always been the companion of power and the ‘attendant of empire’ (compañera del imperio).112

111 In that sense, it might be more than centripetal forces that need to be considered, as posited by Yücesoy: ‘Yet, the Abbasids fell victim to their own successes. The political and confessional centripetal pull of the caliphate eventually gave way to political fragmentation and then to challenges of cultural diversity’. Yücesoy, ‘Language of Empire’, p. 386. 112 Pollock, ‘Cosmopolitan and Vernacular’, p. 616.

towar ds an arab i c co smo po li s

Biblio­graphy Secondary Works ʿAbbās, Iḥsān, ʿAbd al-Ḥamīd b. Yaḥyā al-Kātib wa-mā tabaqqā min rasāʾilihi wa-rasāʾil Sālim Abī al-ʿAlāʾ (Amman: Dār al-Sharq, 1988)

Akasoy, Anna, Charles Burnett, and Ronit Yoeli-Tlalim, ed., Islam and Tibet: Interactions along the Musk Routes (Farnham: Ashgate, 2011) Akbari, Suzanne Conklin, ‘The Persistence of Philo­logy: Language and Connectivity in the Mediterranean’, in A Sea of Languages: Rethinking the Arabic Role in Medi­eval Literary History, ed. by Suzanne Conklin Akbari and Karla Malette (Toronto: University of Toronto Press, 2013), pp. 3–22 Ali, Daud, ‘The Early Inscriptions of Indonesia and the Problem of the Sanskrit Cosmopolis’, in Early Interactions between South and Southeast Asia: Reflections on Cross Cultural Exchange, ed. by Pierre-Yves Manguin, A. Mani, and Geoff Wade (Singapore: Institute of Southeast Asian Studies; New Delhi: Manohar India, 2011), pp. 277–95 Aljoumani, Said, ‘Taʾrīkh Qubbat al-māl aw Qubbat ʿĀʾisha aw al-Qubba al-gharbiyya aw Qubbat al-khazna fī al-Jāmiʿ al-umawī, Dimashq’, in The Damascus Fragments: Towards a History of the Qubbat al-Khazna Corpus of Manu­scripts and Documents, ed. by Arianna D’Ottone Rambach, Konrad Hirschler, and Ronny Vollandt (Beirut: Ergon Verlag, 2020), pp. 53–74 Al-Jallad, Ahmad, The Damascus Psalm Fragment: Middle Arabic and the Legacy of Old Ḥigāzī, with a contribution by Ronny Vollandt (Chicago: Oriental Institute, 2020) —— , ‘The Di­graph ‫ اى‬in the Qur’anic Consonantal Text and the Identification of a New Letter Shape for Final hē in the 7th to 8th Century Arabic Script’, Al-ʿUṣūr al-Wusṭā, 29 (2021), 1–19 Anderson, Glaire D., ‘Sovereignty and the Materiality of Caliphal Encounters’, Publications of the Modern Language Association of America, 130.2 (2015), 393–401 Baalbaki, The Legacy of the Kitāb: Sībawayhi’s Analytical Methods within the Context of the Arabic Grammatical Theory (Leiden: Brill, 2008) Bashir, Shahzad, The Market in Poetry in the Persian World (Cambridge: Cambridge University Press, 2021) Becker, Alton L., Beyond Translation: Essays toward a Modern Philo­logy (Ann Arbor: University of Michigan Press, 1995) Beckwith, Christopher I., Warriors of the Cloisters: The Central Asian Origins of Science in the Medi­eval World (Princeton, NJ: Princeton University Press, 2012) Beeston, A. F. L., T. M. Johnstone, R. B. Serjeant, and G. R. Smith, ed., Arabic Literature to the End of the Umayyad Period (Cambridge: Cambridge University Press, 1983) Bernards, Monique, Changing Traditions: al-Mubarrad’s Refutation of Sībawayh and the Subsequent Reception of the Kitāb (Leiden: Brill, 1997) —— , ‘The Contribution of Mawālī and the Arabic Linguistic Tradition’, in Patronate and Patronage in Early and Classical Islam, ed. by Monique Bernards and John Nawas (Leiden: Brill, 2005), pp. 426–53

185

186

a n to in e bo r r u t

—— , ‘Pioneers of Arabic Linguistic Studies’, in In the Shadow of Arabic: The Centrality of Language to Arabic Culture. Studies Presented to Ramzi Baalbaki on the Occasion of his Sixtieth Birthday, ed. by Bilal Orfali (Leiden: Brill, 2011), pp. 197–220 Bondarev, Dmitry, Alessandro Gori, and Lameen Souag, ed., Creating Standards: Interactions with Arabic Scripts in 12 Manu­script Cultures (Berlin: De Gruyter, 2019) Borrut, Antoine, ‘An Islamic Late Antiquity? Problems and Perspectives’, in Journal of Late Antique, Islamic and Byzantine Studies (forthcoming) Bosworth, Clifford E., ‘The Persian Impact on Arabic Literature’, in Arabic Literature to the End of the Umayyad Period, ed. by A. F. L. Beeston, T. M. Johnstone, R. B. Serjeant, and G. R. Smith (Cambridge: Cambridge University Press, 1983), pp. 483–96 Breckenridge, Carol A., Sheldon Pollock, Homi K. Bhabha, and Dipesh Chakrabarty, ed., Cosmopolitanism (Durham, NC: Duke University Press, 2002) Briquel Chatonnet, Françoise, and Muriel Debié, Le monde syriaque: sur les routes d’un Christianisme ignoré (Paris: Les Belles Lettres, 2017) Burbank, Jane, and Frederick Cooper, Empires in World History: Power and the Politics of Difference (Princeton, NJ: Princeton University Press, 2010) Carter, Michael G., Sībawayhi (London: I. B. Tauris, 2004) —— , ‘Sībawayhi’, in Encyclopaedia of Islam, 2nd edn, ed. by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs (Leiden: Brill, online edn) [accessed 6 January 2022] Cobb, Paul M., White Banners: Contention in ʿAbbasid Syria, 750–880 (Albany: State University of New York Press, 2001) Di Cosmo, Nicola, and Michael Maas, ed., Empires and Exchanges in Eurasian Late Antiquity: Rome, China, Iran, and the Steppe, ca. 250–750 (Cambridge: Cambridge University Press, 2018) Dodds, Jerrilynn D., María Rosa Menocal, and Abigail Krasner Balbale, The Arts of Intimacy: Christians, Jews and Muslims in the Making of the Castilian Culture (New Haven, CT: Yale University Press, 2008) Donner, Fred M., ‘Qur’ânicization of Religio-Political Discourse in the Umayyad Period’, in Écriture de l’histoire et processus de canonisation dans les premier siècles de l’islam, ed. by Antoine Borrut (= Revue des mondes musulmans et de la Méditerranée, 129 (2011)), pp. 79–92 D’Ottone, Arianna, ‘Manu­scripts as Mirrors of a Multilingual and Multicultural Society: The Case of the Damascus Find’, in Negotiating Co-existence: Communities, Cultures and Convivencia in Byzantine Society, ed. by Barbara Crostini and Sergio La Porta (Trier: Wissenschaftlicher Verlag Trier, 2013), pp. 63–88 D’Ottone Rambach, Arianna, Konrad Hirschler, and Ronny Vollandt, ed., The Damascus Fragments: Towards a History of the Qubbat al-Khazna Corpus of Manu­scripts and Documents (Beirut: Ergon Verlag, 2020) Douki, Caroline, and Philippe Minard, ‘Histoire globale, histoires connectées: un changement d’échelle historio­graphique? Introduction’, Revue d’histoire moderne et contemporaine, 54.4 (2007), 7–21 Drory, Rina, ‘The Abbasid Construction of the Jāhiliyya: Cultural Authority in the Making’, Studia Islamica, 83 (1996), 33–49 Eche, Youssef, Les bibliothèques arabes publiques et semi-publiques en Mésopotamie, en Syrie et en Égypte au Moyen Age (Damascus: Institut français de Damas, 1967)

towar ds an arab i c co smo po li s

Elverskog, Johan, Buddhism and Islam on the Silk Road (Philadelphia: University of Pennsylvania Press, 2010) Fierro, Maribel, ‘Abd al-Rahman III: The First Cordoban Caliph (Oxford: Oneworld, 2005) Flood, Finbarr Barry, Objects of Translation: Material Culture and Medi­eval ‘Hindu-Muslim’ Encounter (Princeton, NJ: Princeton University Press, 2009) Ganim, John, ‘Cosmopolitanism and Medi­evalism’, Exemplaria, 22.1 (2010), 5–27 Garosi, Eugenio, ‘Imperial Arabic: Some Notes on Visual Symbolism’, in Christians and Muslims in Early Islamic Egypt, ed. by Lajos Berkes (Durham, NC: The American Society of Papyro­logists, 2022), pp. 13–31 Gentzler, Edwin, ‘Translation, Poststructuralism, and Power’, in Translation and Power, ed. by Maria Tymoczko and Edwin Gentzler (Amherst: University of Massachusetts Press, 2002), pp. 195–218 Goodman, L. E., ‘The Greek Impact on Arabic Literature’, in Arabic Literature to the End of the Umayyad Period, ed. by A. F. L. Beeston, T. M. Johnstone, R. B. Serjeant, and G. R. Smith (Cambridge: Cambridge University Press, 1983), pp. 460–82 Gould, Rebecca, ‘How Newness Enters the World: The Methodo­logy of Sheldon Pollock’, Comparative Studies of South Asia, Africa, and the Middle East, 28.3 (2008), 533–59 Grévin, Benoît, Le parchemin des cieux: essai sur le Moyen Âge du langage (Paris: Le Seuil, 2012) Gutas, Dimitri, Greek Thought, Arabic Culture: The Graeco-Arabic Translation Movement in Baghdad and Early ʿAbbāsid Society (2nd–4th/8th–10th centuries) (London: Routledge, 1998) Hansen, Valerie, The Silk Road: A New History (Oxford: Oxford University Press, 2012) Heidemann, Stefan, ‘The Evolving Representation of the Early Islamic Empire and its Religion on Coin Imagery’, in The Qurʾān in Context: Historical and Literary Investigations into the Qurʾānic Milieu, ed. by Angelika Neuwirth, Nicolai Sinai, and Michael Marx (Leiden: Brill, 2010), pp. 149–95 Hirschler, Konrad, ‘From Archive to Archival Practices: Rethinking the Preservation of Mamluk Administrative Documents’, Journal of the American Oriental Society, 136 (2016), 1–28 Hoyland, Robert G., ‘The Birth of Arabic in Stone’, in By the Pen and What they Write: Writing in Islamic Art and Culture, ed. by Sheila Blair and Jonathan Bloom (New Haven, CT: Yale University Press, 2017), pp. 51–67 Iema, Z. T., A. Al-Jallad, M. C. A. Macdonald, and L. Nehmé, ‘Provincia Arabia: Nabataea, the Emergence of Arabic as a Written Language, and Graeco-Arabica’, in Arabs and Empires Before Islam, ed. by Greg Fisher (Oxford: Oxford University Press, 2015), pp. 373–433 Imbert, Frédéric, ‘Le prince al-Walīd et son bain: itinéraires épi­graphiques à Quṣayr ʿAmra’, Bulletin d’études orientales, 64 (2016), 321–63 Kennedy, Hugh, Muslim Spain and Portugal: A Political History of al-Andalus (London: Longman, 1996) Khoury, R. G., ‘Wahb b. Munabbih’, in Encyclopaedia of Islam, 2nd edn, ed. by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs (Leiden: Brill, online edn) [accessed 11 June 2022] King, Ross, ‘Ditching “Diglossia”: Describing Eco­logies of the Spoken and Inscribed in Pre-modern Korea’, Sungkyun Journal of East Asian Studies, 15.1 (2015), 1–19

187

188

a n to in e bo r r u t

—— , ‘Introduction: Koh Jongsok’s Infected Language’, in Infected Korean Language, Purity vs. Hybridity: From the Sino­graphic Cosmopolis to Japanese Colonialism to Global English, ed. by Koh Jongsok, trans. with a critical introduction by Ross King (Amherst, NY: Cambria Press, 2014), 1–16 —— , ‘Ditching “Diglossia”: Describing Eco­logies of the Spoken and Inscribed in Premodern Korea’, Sungkyun Journal of East Asian Studies, 15.1 (2015), 1–19 —— , The Language of the Sages in the Realm of Vernacular Inscription: Reading Sheldon Pollock from the Sino­graphic Cosmopolis (Leiden: Brill, forthcoming) Klasová, Pamela, ‘Empire through Language: Al-Ḥajjāj b. Yūsuf al-Thaqafī and the Power of Oratory in Umayyad Iraq’ (unpublished doctoral dissertation, Georgetown University, 2018) Kristó-Nagy, István, La pensée d’Ibn al-Muqaffaʿ: un ‘agent double’ dans le monde persan et arabe (Versailles: Éditions de Paris, 2013) La Vaissière, Étienne de, Histoire des marchands sogdiens (Paris: De Boccard, 2002), translated as Soghdian Traders: A History (Leiden: Brill, 2005) Lavan, Myles, ‘“Father of the Whole Human Race”: Ecumenical Language and the Limits of Elite Integration in the Early Roman Empire’, in Cosmopolitanism and Empire: Universal Rulers, Local Elites, and Cultural Integration in the Ancient Near East and Mediterranean, ed. by Myles Lavan, Richard E. Payne, and John Weisweiler (Oxford: Oxford University Press, 2016), pp. 153–68 Lavan, Myles, Richard E. Payne, and John Weisweiler, ‘Cosmopolitan Politics: The Assimilation and Subordination of Elite Cultures’, in Cosmopolitanism and Empire: Universal Rulers, Local Elites, and Cultural Integration in the Ancient Near East and Mediterranean, ed. by Myles Lavan, Richard E. Payne, and John Weisweiler (Oxford: Oxford University Press, 2016), pp. 1–28 —— , ed., Cosmopolitanism and Empire: Universal Rulers, Local Elites, and Cultural Integra­ tion in the Ancient Near East and Mediterranean (Oxford: Oxford University Press, 2016) Lindstedt, Ilkka, ‘Arabic Rock Inscriptions up to 750 ce’, in The Umayyad World, ed. by A. Marsham (London: Routledge, 2021), pp. 411–37 Mackensen, Ruth S., ‘Supplementary Notes to “Arabic Books and Libraries in the Umayyad Period”’, American Journal of Semitic Languages and Literatures, 56.2 (1939), 149–57 Martinez-Gros, Gabriel, L’idéo­logie omeyyade: la construction de la légitimité du Califat de Cordoue (xe–xie siècles) (Madrid: Casa de Velázquez, 1992) —— , Identité andalouse (Paris: Actes Sud, 1997) Menocal, María Rosa, The Arabic Role in Medi­eval Literary History: A Forgotten Heritage (Philadelphia: University of Pennsylvania Press, 1987) Mignolo, Walter D., Local Histories/Global Designs: Coloniality, Subaltern Knowledges, and Border Thinking (Princeton, NJ: Princeton University Press, 2000) Miles, George C., ‘Early Islamic Inscriptions near Ṭāʾif in the Ḥijāz’, Journal of Near Eastern Studies, 7.4 (1948), 236–42 Monius, Anne E., Imagining a Place for Buddhism: Literary Culture and Religious Community in Tamil-Speaking South India (Oxford: Oxford University Press, 2001) Nehmé, Laïla, ‘A Glimpse of the Development of the Nabataean Script into Arabic Based on Old and New Epi­graphic Material’, in The Development of Arabic as a Written Language, ed. by M. C. A. Macdonald (Oxford: Archaeopress, 2010), pp. 47–88

towar ds an arab i c co smo po li s

—— , ‘Aramaic or Arabic? The Nabataeo-Arabic Script and the Language of the Inscriptions Written in This Script’, in Arabic in Context: Celebrating 400 Years of Arabic at Leiden University, ed. by Ahmad Al-Jallad (Leiden: Brill, 2017), pp. 75–98 —— , ‘Epi­graphy on the Edges of the Roman Empire: A Study of the Nabataean Inscriptions and Related Material from the Darb al-Bakrah, Saudi Arabia, 1st–5th Century ad’ (unpublished habilitation thesis, École pratique des hautes études [EPHE], 2013) Ollett, Andrew, Language of the Snakes: Prakrit, Sanskrit, and the Language Order of Premodern India (Oakland: University of California Press, 2017) Orfali, Bilal, In the Shadow of Arabic: The Centrality of Language to Arabic Culture. Studies Presented to Ramzi Baalbaki on the Occasion of his Sixtieth Birthday (Leiden: Brill, 2011) Orsatti, Paola, ‘Persian Language in Arabic Script: The Formation of the Ortho­graphic Standard and the Different Graphic Traditions of Iran in the First Centuries of the Islamic Era’, in Creating Standards: Interactions with Arabic Scripts in 12 Manu­script Cultures, ed. by Dmitry Bondarev, Alessandro Gori, and Lameen Souag (Berlin: De Gruyter, 2019), pp. 39–72 Orsini, Francesca, ‘The Multilingual Local in World Literature’, Comparative Literature, 67.4 (2015), 345–74 Paul, Jürgen, ‘Archival Practices in the Muslim World prior to 1500’, in Manu­scripts and Archives: Comparative Views on Record-Keeping, ed. by Sabine Kienitz, Michael Friedrich, Christian Brockmann, and Alessandro Bausi (Berlin: De Gruyter 2018), pp. 339–60 Payne, Richard E., A State of Mixture: Christians, Zoroastrians, and Iranian Political Culture in Late Antiquity (Oakland: University of California Press, 2015) Peacock, Andrew C. S., Islam, Literature and Society in Mongol Anatolia (Cambridge: Cambridge University Press, 2019) Pollock, Sheldon, ‘The Cosmopolitan Vernacular’, Journal of Asian Studies, 57 (1998), 6–37 —— , ‘Cosmopolitan and Vernacular in History’, Public Culture, 12.3 (2000), 591–625 —— , The Language of the Gods in the World of Men: Sanskrit, Culture, and Power in Premodern India (Berkeley: University of California Press, 2006) Pollock, Sheldon, Homi K. Bhabha, Carol A. Breckenridge, and Dipesh Chakrabarty, ‘Cosmopolitanisms’, Public Culture, 12.3 (2000), 577–89 Prémare, Alfred-Louis de, ‘Wahb b. Munabbih, une figure singulière du premier islam’, Annales: Histoire, Sciences Sociales, 60.3 (2005), 531–49 al-Qāḍī, Wadād, ‘Early Islamic State Letters: The Question of Authenticity’, in The Byzantine and Early Islamic Near East I: Problems in the Literary Source Material, ed. by Averil Cameron and Lawrence I. Conrad (Princeton, NJ: Darwin Press, 1992), pp. 215–75 —— , ‘ʿAbd al-Ḥamīd al-Kātib’, in Encyclopaedia of Islam, 3rd edn, ed. by Kate Fleet, Gudrun Krämer, Denis Matringe, John Nawas, and Everett Rowson (Leiden: Brill, online edn) [accessed 6 January 2022] al-Rāshid, Saʿd b. ʿAbd al-ʿAzīz, ‘Sadd al-Khanaq: An Early Umayyad Dam near Medina, Saudi Arabia’, Proceedings of the Seminar for Arabian Studies, 38 (2008), 265–76 Ricci, Ronit, Islam Translated: Literature, Conversion, and the Arabic Cosmopolis of South and Southeast Asia (Chicago: University of Chicago Press, 2011)

189

1 90

a n to in e bo r r u t

Robin, Christian Julien, ‘La réforme de l’écriture arabe à l’époque du califat médinois’, Mélanges de l’Université Saint-Joseph, 59 (2006), 319–64 —— , ‘L’ Arabie à la veille de l’Islam dans l’ouvrage d’Aziz al-Azmeh, The Emergence of Islam in Late Antiquity’, Topoi, 21 (2017), 316–17 —— , Najrān en Arabie et ses martyrs chrétiens (forthcoming) Robinson, Chase F., ‘Abd al-Malik (Oxford: Oneworld, 2005) Rubin, Milka, ‘Arabization versus Islamization in the Palestinian Melkite Community during the Early Muslim Period’ in Sharing the Sacred: Religious Contacts and Conflicts in the Holy Land, First-Fifteenth Centuries ce, ed. by Arieh Kofsky and Guy G. Stroumsa ( Jerusalem: Yad Izhak ben Zvi, 1998), pp. 149–62 Rustow, Marina, The Lost Archive: Traces of a Caliphate in a Cairo Synagogue (Princeton, NJ: Princeton University Press, 2020) Saliba, George, Islamic Science and the Making of the European Renaissance (Cambridge, MA: MIT Press, 2007) Savant, Sarah Bowen, The New Muslims of Post-Conquest Iran: Tradition, Memory, and Conversion (Cambridge: Cambridge University Press, 2013) Schmidt, Jan, ‘How to Write Turkish? The Vagaries of the Arabo-Persian Script in Ottoman-Turkish Texts’, in Creating Standards: Interactions with Arabic Scripts in 12 Manu­script Cultures, ed. by Dmitry Bondarev, Alessandro Gori, and Lameen Souag (Berlin: De Gruyter, 2019), pp. 131–46 Schoeler, Gregor, Écrire et transmettre dans les débuts de l’islam (Paris: Presses Universitaires de France, 2002), rev. edn with and trans. by Shawkat M. Toorawa as The Genesis of Literature in Islam: From the Aural to the Read (Edinburgh: Edinburgh University Press, 2009) Spiegel, Gabrielle, Romancing the Past: The Rise of Vernacular Prose Historio­graphy in Thirteenth-Century France (Berkeley: University of California Press, 1993) Subrahmanyan, Sanjay, ‘Connected Histories: Notes towards a Reconfiguration of Early Modern Eurasia’, Modern Asian Studies, 31.3 (1997), 735–62 Thomann, Johannes, ‘From Lyrics by al-Fazārī to Lectures by al-Fārābī: Teaching Astronomy in Baghdād (750–1000 c.e.)’, in The Place to Go: Contexts of Learning in Baghdād, 750–1000 c.e., ed. by Jens Scheiner and Damien Janos (Princeton, NJ: Darwin Press, 2014), pp. 503–25 Tillier, Mathieu, and Naïm Vanthieghem, ‘Recording Debts in Sufyānid Fusṭāṭ: A Reexamination of the Procedures and Calendar in Use in the First/Seventh Century’, in Geneses: A Comparative Study of the Historio­graphies of the Rise of Christianity, Rabbinic Judaism and Islam, ed. by John Tolan (London: Routledge, 2019), pp. 148–88 Toral-Niehoff, Isabel, ‘Writing for the Caliphate: The Unique Necklace by Ibn ʿAbd Rabbih’, Al-ʿUṣūr al-Wusṭā, 26 (2018), 80–95 Van Bladel, Kevin, ‘The Bactrian Background of the Barmakids’, in Islam and Tibet: Interactions along the Musk Routes, ed. by Anna Akasoy, Charles Burnett, and Ronit Yoeli-Tlalim (Farnham: Ashgate, 2011), pp. 43–88 —— , ‘Eighth-Century Indian Astronomy in the Two Cities of Peace’, in Islamic Cultures, Islamic Contexts: Essays in Honor of Professor Patricia Crone, ed. by Behnam Sadeghi, Asad Q. Ahmed, Adam Silverstein, and Robert Hoyland (Leiden: Brill, 2014), pp. 257–94

towar ds an arab i c co smo po li s

—— , ‘Arabicization, Islamization, and the Colonies of the Conquerors’, in Late Antique Responses to the Arab Conquests, ed. by Josephine van den Bent, Floris van den Eijnde, and Johan Weststeijn (Leiden: Brill, 2022), pp. 89–119 Van De Mieroop, Marc, Before and After Babel: Writing as Resistance in Ancient Near Eastern Empires (Oxford: Oxford University Press, 2023). Violet, Bruno, ‘Ein zweisprachiges Psalmfragment aus Damaskus’, Orientalistische Litteratur-Zeitung, 4.10–12 (1901), 384–403, 425–41, 475–88 Webb, Peter, ‘Al-Jāhiliyya: Uncertain Times of Uncertain Meanings’, Der Islam, 91.1 (2014), 69–94 —— , Imagining the Arabs: Arab Identity and the Rise of Islam (Edinburgh: Edinburgh University Press, 2016). Werner, Michael, and Bénédicte Zimmermann, ed., De la comparaison à l’histoire croisée (Paris: Le Seuil, 2004) —— , ‘Beyond Comparison: Histoire croisée and the Challenge of Reflexivity’, History and Theory, 45 (2006), 30–50 Yücesoy, Hayrettin, ‘Translation as Self-Consciousness: Ancient Sciences, Antediluvian Wisdom, and the ʿAbbāsid Translation Movement’, Journal of World History, 20.4 (2009), 523–57 —— , ‘Language of Empire: Politics of Arabic and Persian in the Abbasid World’, Publications of the Modern Language Association of America, 130.2 (2015), 384–92 Zadeh, Travis, The Vernacular Qurʾan: Translation and the Rise of Persian Exegesis (Oxford: Institute of Ismaili Studies and Oxford University Press, 2012)

191

Section Two

Multilingualism, Empires, and Local Elites

Muriel Debié

The Arabs and Northern Languages and Scripts before Islam When in 800 bce the regent of the Hittite city of Karkemish, Yariris, boasted about knowing twelve languages and four scripts including the ta-i-ma-ni-ti script, which probably refers to ‘the script used in the Arabian oasis of Тауmāʾ, possibly as a representative of the alphabets of Arabia in general’,1 he could not have imagined that he would thereby leave a long-lasting trace of the wide use of Arabic dialects along the caravan roads between the Arabian Peninsula and ancient Syria. Only direct or indirect written traces can give us access to the use of ancient languages when their oral expression has vanished into the silence of past centuries: verba volant, scripta manent. We are thus dependent on written sources to understand the place of Arabic in Late Antiquity, in the centuries before Islam, when the language that was used on a daily basis for economic, religious, and cultural life first started to be written down. The earliest extant literary text in Arabic that was transmitted in written form is the Qurʾān, although it was allegedly revealed to Muḥammad orally and then preserved through recitation within the community of Believers, exemplifying the complex relationship between the oral and written forms of the language.2 So far, no material traces of written texts (other than inscriptions) dating back to the pre-Islamic period have been found, and only a small — although growing — number of inscriptions attest the existence of written Arabic in any script, if understood in a traditional narrow sense, before Islam.3 However, if we take up the suggestion of considering Arabic in a broader sense, as including Safaitic and Hismaic (Ancient North Arabian), then tens of thousands of Arabic inscriptions can be included in the analysis.4



1 I am grateful to Christian Robin for reading a version of this article. All errors remain mine. Macdonald, ‘Arabia and the Written Word’, p. 5. 2 Schoeler, The Oral and the Written in Early Islam. 3 Robin, ‘Les plus anciens monuments de la langue arabe’. 4 Al-Jallad, ‘The Linguistic Landscape of Pre-Islamic Arabia’, p. 114, and ‘What Is Ancient North Arabian?’.

Prof. Muriel Debié ([email protected]) is director of research at École Pratique des Hautes Études, PSL, in Paris and a specialist in the history of Eastern Christianities and in Syriac studies Navigating Language in the Early Islamic World: Multilingualism and Language Change in the First Cen­turies of Islam, ed. by Antoine Borrut, Manuela Ceballos, and Alison M. Vacca, ISMAR 2 (Turnhout: Brepols, 2024) pp. 195–257 10.1484/M.ISMAR-EB.5.134627

FHG

196

m u r i e l de bi é

Pre-Islamic poetry mentioned in Islamic-era texts attests to the existence of literary production in Arabic even though no contemporary written evidence survives, only later accounts. The impression is that written Arabic started with the conquests, when it developed into an official language and became the sacred language of Islam, gradually replacing various neighbouring languages in a hieroglossic relation with Greek and Syriac.5 The oldest dated document known so far is a papyrus written in ah 21 (642 ce), which thus places the beginning of the use of Arabic as a language of economic and administrative documents shortly after the time of Muḥammad and in the very early stages of the Arab polity.6 Both religious reasons (Arabic as the ‘clear’ language of the Qurʾān) and political ones contributed to the impetus for writing the language, but orality remained the ideal according to the early Islamic (written) tradition. Writing was considered a compromise, a second-best solution.7 This explains the absence of written texts from before the Islamic period in spite of oral production and transmission of literary texts (at least poetry) in that time. That does not mean that Arabic was not written at all in the pre-Islamic period. Like many other languages in the Middle and Near East, it had in fact been written for a long time in other scripts. Epi­graphic texts written in two script families used in the Arabian Peninsula — in Arabian alphabets (South Arabian, Dadanitic, Hismaic, and Safaitic) and in Nabataeo-Arabic — have been preserved from the pre-Islamic period.8 A few texts were also written in Greek script in the SinaiticoPalestinian area.9 Such allo­graphic practices had been customary in the Near East since the use of cuneiform. Multilingualism and allo­graphism were everywhere in the Middle East and the Arabian Peninsula in Late Antiquity, and the Arabs10 were part of this multifarious linguistic world. Multiple scripts were part and parcel of their visual landscape. In addition to the official languages of the empires, Arabs in peninsular Arabia were exposed to the web of North and South Arabian scripts that punctuate boulders in the desert near watering holes and in oases. Writing was widespread among both settled people and nomads, although it served different functions for each. Beautifully carved Sabaean inscriptions in the musnad script conveyed official messages of the Ḥimyarite kingdom (275–571 ce). Writing was



5 Jean-Noël Robert defines ‘hiéroglossie’ as ‘le réseau de phénomènes qui unit entre elles des langues et des cultures qui se reconnaissent pour centre une langue primordiale, c’est-à-dire de premier rang, le plus souvent porteuse à l’origine d’Écritures tenues pour consacrées. Il n’est pas besoin de rappeler que cette appellation n’implique nullement un jugement de valeur, mais délimite un fait historique qui évolue d’ailleurs selon une courbe récurrente. Entre deux langues impliquées dans une relation hiéroglossique, le rapport sera d’abord de soumission de la seconde à la première, puis d’émulation et enfin de remplacement’. See Robert, ‘Hiéroglossie des civilisations’. Greek and Syriac are thus in a hieroglossic relation with Arabic. 6 Sijpesteijn, ‘Arabic Script and Language in the Earliest Papyri’. 7 Schoeler, The Oral and the Written in Early Islam. 8 Robin, ‘Les langues de la péninsule Arabique’. 9 Fiema, Al-Jallad, Macdonald, and Nehmé, ‘Provincia Arabia’; Al-Jallad, ‘Graeco-Arabica I’. 10 I will not consider here the vast question of Arab identity and Arab tribes that has been much discussed in the last decades.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

visible everywhere along the caravan roads even though it was not necessary for many aspects of everyday life and was sometimes a mere pastime during boring hours spent watching cattle in the desert.11 Epi­graphy has been instrumental in revealing the history of Ancient Arabia. However, inscriptions, except for the official South Arabian ones, are frustratingly brief and give only little historical or linguistic information. I want to argue in this paper that texts, which have hitherto been little studied from this vantage point, can also throw interesting light on literacy and the use of several languages among the Arabs. The elite, at least, participated in the literate culture of the empires and local kingdoms, implying mastery of languages and scripts other than Arabic. But indirect evidence of literate practices is seldom considered when talking about literacy among the Arabs. As was usual in Late Antiquity, the civic and religious elites, professional scribes in chancelleries, and local administrators and notaries were almost the only ones who knew how to read or write.12 But among the Arabs, even nomads — at least those who travelled and saw graffiti, inscriptions, and official texts along the roads and in the villages and cities of the Peninsula and the Near East, as well as those who fought as allied forces with the armies of the late antique polities — knew of and partly belonged to literate societies. And unlike shepherds, the Arab kings or phylarchs and their courts, military commanders, and chancelleries were integrated into the literate worlds of their times, as is evidenced by the geopolitical interplay of the late antique powers in the Arabian Peninsula and on the borders of the Roman and Sasanian Empires. The religious elites of Arab Jews and Christians were also part of the global geo-religious map and participated in a network of ecclesiastical relations, of exchanges of correspondence and religious commentaries, and of shared scriptures and liturgies according to their religious affiliations and confessions in various languages, both oral and written. My contribution aims to highlight the indirect evidence that demonstrates that at least the Arab elites, and possibly also more ordinary people, were part of the literate life of empires and churches and engaged in correspondence with the higher levels of their respective administrations and representatives. Why did Arabs adopt the Nabataeo-Arabic or rather Palaeo-Arabic alphabet for writing Arabic when they did not need a written medium, either because they did not write in everyday life or because they used other languages in official circumstances such as diplomacy or, among Christians, for Church matters? And what can we qualify as ‘Arabic’ in the pre-Islamic period before the language was standardized under the influence of the Qurʾānic idiom and administrative use and became a cultural medium in Islamic times?13

11 M. Macdonald has rightly drawn attention to this ludic dimension of scripts in several of his publications. I will cite here only the first: Macdonald, ‘Literacy in an Oral Environment’. 12 On ‘simple believers’, see Tannous, The Making of the Medi­eval Middle East. 13 Al-Jallad, ‘The Earliest Stages of Arabic and Its Linguistic Classification’.

197

198

m u r i e l de bi é

Figure 4.1. Distribution of Palaeo-Arabic inscriptions. White: sites of inscriptions; red: major cities. From Al-Jallad and Sidky, ‘A Paleo-Arabic Inscription on a Route North of Ṭāʾif’’, p. 10. Reproduced with permission.

We now have evidence that a distinctive Palaeo-Arabic script14 was used to write the language much earlier than previously thought, since it is attested in inscriptions recently discovered in the area of Ḥimā in modern Saudi Arabia; the oldest of these inscriptions is dated to 470 ce and another one to 513 ce.15 Three other multilingual inscriptions were already known and demonstrate use of the Arabic script in southern Syria in the sixth century: the first is from Zabad, south-east of Aleppo (512 ce); the second from Jabal Usays on the borders of the Ḥawrān, south-east of Damascus (528 ce); and the third from Ḥarrān, south of Damascus (568 ce).16 A fourth inscription, dated to 548/49 ce and discovered at

14 As Al-Jallad and Sidky suggest, this term, coined by Robin, al-Ghabbān, and al-Saʿīd in ‘Inscriptions antiques de la région de Najrān’, is more fitting than Old Arabic (which refers to inscriptions in the Arabic language and in various scripts) or Pre-Islamic Arabic, since it allows for chrono­logical overlaps between pre-Islamic and Islamic forms and at the same time asserts that it was distinct from the standardized idiom of Islamic times (Al-Jallad and Sidky, ‘A Paleo-Arabic Inscription on a Route North of Ṭāʾif ’, p. 2). 15 Robin, al-Ghabbān, and al-Saʿīd, ‘Inscriptions antiques de la région de Najrān’. 16 Macdonald, ‘Old Arabic’, nos Χ–ΧΠ.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

Dūmat al-Jandal, was added recently.17 New findings, some not yet published, add more inscriptions in Palaeo-Arabic to the list.18 Since these last discoveries come from north-west Arabia (one is from Ṭāʾif, north of Mecca in the latter’s fertile hinterland), they challenge the broad consensus that the Ḥijāz stood apart from the regions of the Peninsula where writing was common. Although we still have overall few examples, their number has grown steadily over the past years. These findings have entirely changed our chrono­logy for the emergence of a distinctive Arabic script some 150 years before Islam as well as our mental map of its spread, since it was widely disseminated geo­graphically in and outside the Arabian Peninsula at that time (see Fig. 4.1). Moreover, the images of crosses that accompany the most ancient inscriptions, as well as the Christian contexts of several of them, point to a Christian milieu for the use of this Arabic alphabet — and possibly its emergence — in the pre-Islamic period, although some of these inscriptions are just monotheistic invocations. This paper seeks to elucidate the religious, social, and historical background of Arabs’ use of oral and written Arabic, on the one hand, and other official languages, on the other, in Late Antiquity. Using the inscriptions recently brought to light, we can try to understand how and why a distinctive Palaeo-Arabic script emerged. Syriac sources produced on both sides of the Roman-Sasanian border, Latin and Greek ones in the Roman Empire, and later Islamic sources19 can help illuminate the use of other languages in official contexts, political and economic (Greek, Nabataean, and Middle Persian) as well as cultural and religious (Greek and Syriac). They may also help explain how the Arabic script was used before Islam. A more precise understanding of the interplay of literacy, languages, and scripts among the Arabs before Islam is of great import for understanding the context in which the Qurʾān emerged, since the prevailing idea is that outside the northern oasis towns and South Arabia, Arabian societies were fundamentally non-literate.

The Emergence of the Arabic Script in Late Antiquity The earliest attestations of the Palaeo-Arabic script bring up more questions than answers: What was the impulse for adopting a new alphabet? Persian, for instance, has been recorded in scripts borrowed from other languages: cuneiform for Ancient Persian, Aramaic for Middle Persian, and Arabic for Modern Persian. What sparked the need for an Arabic script before Islam? And what was the social milieu that disseminated it? Although an Arabic script was endorsed, the small number of inscriptions and the absence of extant texts or documents written on soft materials (papyrus or parchment) cannot but have us wonder why the Arabic script emerged 17 Nehmé, ‘New Dated Inscriptions from a Site near al-Jawf ’. 18 Al-Jallad and Sidky, ‘A Paleo-Arabic Inscription on a Route North of Ṭāʾif ’; see p. 11 for a table of all the Palaeo-Arabic inscriptions known so far, including still unpublished ones. 19 We will not refer here to the somewhat artificial distinction between insider and outsider sources (cf. Hoyland, ‘Insider and Outsider Sources’).

199

200

m u r i e l de bi é

but then failed to achieve widespread use before Islam.20 Are the blanks in the documentation due to the dearth of extensive archaeo­logical investigations in these regions so far or the destruction of documents on soft media? Or is there a reason the Arabic script initially had so little success? We have to be cautious when trying to extrapolate from what we have: the recent findings have deeply changed what we thought we knew on the subject, and new archaeo­logical findings will probably again challenge our understanding in the future. The new evidence we have allows us to challenge definitively the concept of Jāhiliyya constructed by the Islamic tradition, which still weighs heavily on our view of pre-Islamic Arabs, and to reassess the place of written Arabic and other languages among pre-Islamic Arabs. The standard perspective on the Jāhiliyya derives clearly from the prejudices of the Qurʾān and later Islamic writers towards the language and (il)literacy of the Arabs, especially nomads, before Islam. But it is also informed by the view of the ‘classical’ literature (historio­graphic and hagio­graphic texts) on which modern historians still rely, which preserved the bias of Herodotus and, more generally, Greek literature against ‘barbarian’ nomads. On the religious side, Jewish supremacism in the Bible — which was continued by the Christians, who deemed themselves the true chosen people and heirs of Jewish sacred history — also shaped the image of the Arabs as secondary and somehow illegitimate heirs of the divine promise. Whether or not the Arabs themselves included this dimension in their concept of Jāhiliyya, it had a profound influence on how patristic authors, as well as modern and even contemporary researchers, who were themselves predominantly from a Jewish or Christian background or tradition, viewed the Arabs from a history of religions perspective.21 Consequently, the language of the Arabs was not included in the circle of sacred languages before Islam, and afterwards it was considered one only by Muslims. As a sacred language it became part of Muslim identity, which obscured the fact that it was also the vernacular and cultural language of many Jews and Christians. However, recent research on the Arab tribes22 before Islam has changed the way we perceive their integration in late antique geopolitics, economy, and religions.23 The new picture drawn by the dates and dissemination of the ancient Arabic

20 Cf. Wasserstein, ‘Why Did Arabic Succeed where Greek Failed?’; Hoyland, ‘Language and Identity’. 21 Millar, ‘Rome’s “Arab” Allies in Late Antiquity’. 22 Modern scholars have rightly emphasized that one should make a termino­logical distinction between the ruling elites and the tribal groups they led. The most famous of these dynasties were the Ḥujrids of Kinda, the Jafnids of Ghassān, and the Naṣrids of Lakhm (Robin, ‘Les Arabes de Ḥimyar, des “Romains” et des Perses’; Fisher, Between Empires, 3–4). However, this distinction does not derive from the sources (Millar, ‘Rome’s “Arab” Allies in Late Antiquity’). This article therefore uses the more conventional designations Ghassānids and Lakhmids. For a synthetic and clear summary of recent scholarship, see Bevan, Fisher, and Genequand, ‘The Late Antique Church at Tall al-ʿUmayrī East’, pp. 58–61; Hoyland, Arabia and the Arabs; Macdonald, ‘Arabs, Arabias, and Arabic before Late Antiquity’. See also Webb’s review of Fisher’s book for important remarks on the methodo­logical problems raised by the near-absence of Arab Muslim sources in recent discussions of pre-Islamic Arabs — a trend also visible in recent Qurʾānic studies. 23 Bowersock, ‘The Ḥadramawt between Persia and Byzantium’; Bowersock, The Crucible of Islam.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

inscriptions also invites a fresh look at the socio-political and religious reasons for the use of a distinct Arabic script. It entails a cognitive jump to realize that the first attested use of the Arabic alphabet is more or less contemporary with the creation of the Armenian alphabet for writing Armenian or the development of specific letters for Christian Palestinian Aramaic (CPA), and that we need to push the emergence of written Arabic back into a historical context that is altogether different from that of the emergence of Islam. We do not have hagio­graphic texts concerning the emergence of the Arabic alphabet that would be similar to those that credit Mesrop Maštoc‘ in the fifth century ce with devising the Armenian, Georgian, and Caucasian Albanian alphabets and Cyril and Methodius in the ninth century with developing the Glagolithic alphabet for the purpose of translating the Bible and ecclesiastical texts. Until the creation of the Armenian alphabet sometime in the fourth or early fifth century (the oldest extant inscription dating to the 480s), Armenian was also written in Aramaic or Greek letters. According to the Life of Maštoc‘, the first attempts at creating an Armenian script were made in a Syriac milieu before the saint took charge of the project; his divinely inspired endeavour was accepted and yielded the official scripts for writing Armenian, Georgian, and Albanian. Armenian was in fact modelled on the Greek alphabet, although Pahlavi and Syriac were also important languages and scripts in the region. Both religious and political reasons provided incentives for creating a new script that would ensure the unity of Armenia, divided as it was between the Roman and Sasanian Empires, and a religious idiom independent from both Greek and Syriac. The alphabet contributed to creating a strong Armenian identity, and it became — and remains — an object of veneration: people display it in their houses, the tradition of t‘ṙčnagir consists in decorating each letter with birds, and even trees are shaped in the form of the letters, which are also displayed as statues. The hagio­graphic tale of the creation of the Armenian script helps us reflect on why a new script is devised, how it becomes authoritative (and not just a pastime), and how it participates in identity formation in relation to neighbouring languages and scripts. The process of emergence of the Arabic alphabet seems to have more in common with the emergence of the Edessan script from official Aramaic and its development into the Syriac scripts used for writing inscriptions and manu­scripts, or with the evolution of the CPA alphabet from Syriac estrangela. In these instances, we have no narrative of the creation of a new alphabet since both Syriac and Arabic letters evolved from existing Aramaic ‘ancestors’ (Old Aramaic and then Edessan in the case of Syriac and Nabataean in the case of Arabic) and were not designed entirely from scratch as Armenian was. Yet at some point the evolving letters were standardized enough to be recognized as Arabic as opposed to the Nabataean or Nabataeo-Arabic used in the same period. Palaeo-Arabic seems to have been seen as the chosen medium for rendering a gradually standardized Arabic language (which retained various Aramaisms, especially for dating or expressing filiation: e.g., ‘br’, son of). Variations in script and ortho­graphy characterize these early stages, yet the script is recognizable as Arabic and close to the more standard Arabic script of the Islamic period. The fact that it is attested in the fifth and sixth centuries ce

201

202

m u r i e l de bi é

from Yemen to Syria (see Fig. 4.1 above) in non-official inscriptions means that it was taught in circles that chose to use this distinctive alphabet. In the case of Syriac, the scribes of the kingdom of Osrhoene who continued the tradition of Imperial Aramaic chancelleries were no doubt instrumental in delineating the Old Edessene script, although the following step — the emergence of the form of what we call Syriac scripts as evidenced in manu­scripts — is obscure. In the case of the Arabic script, it is unclear how and by whom the letter forms of Palaeo-Arabic were fixed and endorsed. The current theory, which presumes that the Arabic script is simply an evolved form of the Nabataean Aramaic alphabet and a result of constant writing of one language in the script of another, does not explain how and why the form of the letters became authoritative for writing ‘Arabic’. We need to consider two related questions: in what form was the language recorded, and which letters were chosen to write it?

Dialects and Standard Languages: Aramaic(s) and Arabic(s) Few studies have been devoted to Old Arabic or pre-Islamic Arabic as attested in scripts other than Palaeo-Arabic.24 E. Knauf argues that ‘all spoken language can be regarded as dialect, (or socio-, idio-, regio-, or whatever-lect), even the spoken version(s) of the standard language. Dialects precede the emergence of a standard language, survive that emergence, and even develop freshly on the basis of standard once established’.25 The variety of ‘Middle Arabics’ in the Middle Ages and modern dialects testify to the enduring creativity of the language alongside the standard form and to the existence of regional variants fuelled by other local languages.26 Diglossia between dialects, on the one hand, and Standard or Classical Arabic, on the other, has been a characteristic of Arabic(s) until today. In the period we are interested in, the process of languages’ standardization played an important role in the languages’ definition and fixation in written form. The relationship between Arabic and the languages attested in the Ancient North Arabian (ANA) inscriptions is still debated. Even if we agree that the languages of the ANA inscriptions (Taymanitic, Thamudic(s), Hismaic, and Safaitic) are all an ancient form of Arabic, they display considerable linguistic diversity.27 Arabic was spoken in an Aramaic linguistic environment that may help us understand the function of dialects versus standard language, an issue shared by most ancient (and many modern) languages.28 ‘Aramaic’ does not exist in the singular except in the case of Imperial Aramaic, which developed out of Ancient 24 Mascitelli, L’arabo in epoca preislamica; Al-Jallad, ‘The Earliest Stages of Arabic and Its Linguistic Classification’. 25 Knauf, ‘Arabo-Aramaic and ʿArabiyya’, p. 204. 26 On the Arabic koine, see Ferguson, ‘The Arabic Koine’. 27 Al-Jallad, ‘The Earliest Stages of Arabic and Its Linguistic Classification’, p. 320; Macdonald, ‘Ancient North Arabian’; Macdonald, ‘Old Arabic’. 28 Gzella, A Cultural History of Aramaic; Gzella, Aramaic.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

Aramaic and became the standard language of Near Eastern chancelleries. Even in the imperial context, however, Aramaic dialects continued to be spoken, although they were not written. With the end of the Achaemenid empire and of the official use of Standard Aramaic, several of the local dialects, collectively known as Middle Aramaic, developed a written form: Hatrean, Palmyrenian, Edessan, and Nabataean. There was no such thing as an Aramaic lingua franca in Late Antiquity; instead, there were various Aramaic dialects and a standard Christian form (Syriac) as well as another, more marginal one (CPA).29 Since the Hellenistic period, Aramaic dialects and scripts have been designated as ‘Syrian’, whatever their form. According to the Greek historian Hieronymus of Cardia, who lived through the events he narrates, the Nabataeans wrote a letter ‘in Syrian characters’ to Antigonos Monophtalmos (382–301 bce) when he attacked them.30 In Late Antiquity, the word used for Aramaic dialects in the Latin of the pilgrim Egeria in the 380s ce is siriste, and in the Greek of Theodoret of Cyrrhus (393–c. 460) it is ē tōn syrōn phōnē, the language of ‘Syrians’. Theodoret was well aware of the variety of the Middle Aramaic dialects: ‘The inhabitants of the provinces of Osrhoene, Syria, Euphratesia, Palestine, and Phoenicia use the language [phonē] of the Syrians, but all the same, the dialect [dialexis] of each exhibits many differences’.31 Each of the abovementioned Aramaic dialects was written in its own alphabet, modelled on the letter forms of Imperial Aramaic. Middle Persian also borrowed its letters from Aramaic and developed a complex written system with ‘Aramaisms’, words that were written as Aramaic words with the flexion of Middle Persian and that were read as Middle Persian, even though the latter is an Indo-European language, whereas Aramaic is Semitic. Inscriptions attest to the use of Middle Aramaic dialects for everyday purposes and as the administrative language of the local kingdoms before they were integrated into the Roman Empire. These dialects were subordinate to Greek, which had become, since the Hellenistic period, the official language in the whole Middle East. Greek was used by the elites in cities, whereas some space for the vernacular was granted in the countryside and small provincial towns. In Roman times, Greek became more widespread as one of the official languages of the empire, alongside Latin. Literature survives in only three varieties of Middle Aramaic: Jewish and Christian literature in Palestinian Aramaic, Jewish Targum and Talmud in Babylonian Aramaic, and Christian literature in Syriac. Being the languages of Jews, the Palestinian and Babylonian Aramaics gave rise to religious texts that were confined to Jewish communities. Edessan Aramaic (that is, Syriac) started as the official language of the chancery of the Abgarid kingdom and was restricted to Osrhoene. But once it became a religious and cultural language in the fourth century ce, it was no longer just a regional dialect. Contrary to the Jewish

29 Recent studies tend to speak of ‘Aramaic’ as a unified language in Late Antiquity, ignoring the existence of a variety of dialects and of only one written literary form, Syriac (and later on CPA). 30 Macdonald, ‘Languages, Scripts, and the Uses of Writing among the Nabataeans’. 31 Brock, ‘Greek and Syriac in Late Antique Syria’ p. 149.

203

2 04

m u r i e l de bi é

Aramaics, it was used far beyond its regional origin, both spoken and written, and also served as a standard religious and cultural idiom for people who spoke other forms of Aramaic or other languages. It became the standard or classical language of Christians from the Mediterranean to India, China, and Arabia. Manichaeans as well as polytheists also used it, although few traces of this use remain, having been erased by the disappearance of these communities and the dominance and control of Christians. In Palestine, Christians, like Jews (including Jesus) and other religious groups, spoke the local dialect of Aramaic. They used Greek as the official language of the Church and Syriac as the standard Christian Aramaic language. In the fifth century, at roughly the same time as scripts developed for Armenian and Arabic, a distinct script emerged to record the local dialect in Palestine, Transjordan, and South Syria used by Christians — the so-called CPA.32 Christians modelled their own script on the forms of estrangela Syriac letters, the classical and prestigious form of Aramaic at the time, and used it to write inscriptions and texts. These were mainly translations from Greek in addition to a handful of original hagio­graphic compositions for communities that remained faithful to the imperial church and accepted the Council of Chalcedon (451 ce) in contrast to most of the Syriac Christians in the empire, who opposed it (the Miaphysites). These Christians’ need to differentiate themselves from Jews, on the one hand, and from the predominantly Syriac Miaphysites, on the other, while representing their own language, can explain the creation of a distinct Christian Palestinian script. Christian Arabs in this area were also exposed to Greek as the official language of both administration and the Church, but CPA was visible in inscriptions on mosaics and stones and was the language of the local population and hence also, unofficially, of the Church. We do not know what the liturgical languages were in the provinces of Arabia and Palestine inhabited by Arabs, but Greek, Syriac, and CPA are a good guess. In the province of Arabia, a local form of Aramaic — with a specific form in the Ḥawrān — and Nabataean were also spoken.33 Aramaic was used in the Arabian Peninsula since Nabonidus conquered Arabian lands in the sixth century bce.34 Although Arabic was not the language of all the Arabian nomads in the earliest periods, as Aramaic and the various forms of Thamudic attested in the region demonstrate, in Late Antiquity Old Arabic was probably the predominant spoken language of Arabs in Arabia and Nabataean Aramaic the only local written language of prestige. The Ancient North Arabian languages used by the settled populations, Taymanitic and Dadanitic, ceased to be written and seem to have disappeared well before the early second century 32 On CPA, see Hoyland, ‘Language and Identity’, with an inventory of Christian Palestinian Aramaic texts and a map of their distribution on pp. 37–39. The expression ‘Syro-Arabia’ sometimes used for the Middle East is problematic in that it does not distinguish between Syriac and CPA or other spoken Aramaic dialects (see, e.g., El-Badawi, ‘Divine Kingdom in Syriac Matthew and the Qur’an’, pp. 3–5. 33 On languages in the province of Arabia, see Hoyland, ‘Late Roman Arabia’, pp. 129–32; see also Contini, ‘Il Ḥawran preislamico’, pp. 56–60, on Syriac in the Ḥawrān. 34 Stein, ‘The Role of Aramaic on the Arabian Peninsula’.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

ce. People speaking Arabic and Aramaic (Nabataean and other local forms) were in touch with people speaking and writing the Ancient North Arabian dialect of Hismaic in southern Jordan and north-west Arabia. Old Arabic was intricately related to the Nabataean used by Arabs and Jews; some six thousand inscriptions and graffiti and a few papyri found near the Dead Sea were produced in Nabataean.35 It was the language of the realm of Nabataea that dominated the region between the Arabian and Sinai Peninsulas from the third century bce until it was subsumed into the Roman Empire in 106 ce as Provincia Arabia. Although the Nabataeans were a mixed population of various nomadic and sedentary groups, they shared Aramaic as their written language. Their language is attested in the Ḥawrān, through Transjordan and the Negev to the Sinai, Egypt, and north-west Arabia. Inscriptions have been found as far south as al-ʿUlā, although the Nabataeans exerted political control also south of al-ʿUlā at the end of the first century ce.36 Nabataean coexisted with Greek and a local form of Aramaic in the Ḥawrān. It remained in use in the province of Arabia and in the peninsula for signatures, votive and dedicatory inscriptions, and a few epitaphs as well as legal inscriptions carved on tombs until the mid-fourth century ce.37 It functioned as ‘a supraregional language and formal means of communication for law, administration, and international correspondence’, as opposed to the local traditions of ANA that in most cases were limited to private purposes.38 A commemorative graffiti mentioning a king of Ghassān was written in Nabataean in the third or fourth century ce not far from al-ʿUlā. It shows that Nabataean was the written language of the kingdom of Ghassān, whose territory lay in northwest Arabia in this period.39 The famous epitaph at al-Namāra dated 328 ce and dedicated to Imruʾ al-Qays, ‘king of all the Arabs’, is Arabic written in Nabataean script, demonstrating an allo­graphic practice and asserting an Arabic identity. It is highly significant that the script was written by someone who identified as king of the Arabs for the first time. The ‘En ‘Evdat inscription that embeds two lines of an Arabic liturgical formula in a Nabataean text also attests to the use of Nabataean as the written language of populations that would otherwise use Arabic for oral communication, religious practice, and cultural traditions.40 Nabataean remained a common written prestige language after the demise of the Nabataean kingdom. Like Aramaic, ‘Arabic’ in pre-Islamic times should be understood as a collection of dialects. The traditional division ‘between Arabic, as defined by the Qurʾān, poetry, the Arabic grammarians, etc., and Ancient North Arabian, the epi­graphic varieties written in the Ancient North Arabian alphabets’, which was based p­ rimarily on

35 On the language and script, see Macdonald, ‘Languages, Scripts, and the Uses of Writing among the Nabataeans’. 36 Robin, ‘Ghassān en Arabie’, p. 97. 37 Macdonald, ‘Languages, Scripts, and the Uses of Writing among the Nabataeans’. 38 Gzella, Aramaic, p. 303. 39 Robin, ‘Ghassān en Arabie’, pp. 96–97 and 100–02. 40 Shaked, ‘The ʿEn ʿAvdat Inscription Revisited’.

205

206

m u r i e l de bi é

variation between the articles ʾl and h(n), has recently been contested.41 A. Al-Jallad has argued that ‘Old Arabic should be viewed as a continuum of dialects stretching from southern Syria into Jordan, the Negev, Sinai, and the northern Higāz, encompassing the dialects composed in the Safaitic, Hismaic, occasionally in the Nabataean, and finally in the Palaeo-Arabic scripts’.42 The reach of Old Arabic should also include Ḥimyar where a variety of dialects spanned from Sabaean to Arabic.43 Arabic underwent what linguists call a ‘homogenizing bottleneck narrowing the variations’ found in its pre-Islamic history.44 Old Arabic was slowly standardized as a language, and this enabled the fixation of a specific script recognizable as Arabic.45 The consensus about the emergence of the Arabic script today is that it is the result of the evolution of the Nabataean that was used by the Arabs.46 A gradual increase in the cursive character of the Nabataean script and the more extensive use of ligatures is noticeable, perhaps under the influence of writing on soft materials, as cursive forms develop to economize writing with a pen and ink.47 The inscriptions exhibiting a more cursive character are termed ‘Nabataeo-Arabic’ and are an intermediary between ‘Classical Nabataean’, namely, the script employed in monuments at Ḥigrā (modern Madāʾin Ṣāliḥ) and Petra, and Palaeo-Arabic script.48 In addition to the development of cursive forms from the day-to-day or documentary script, the later inscriptions are also characterized by the growing presence of Arabic lexical and grammatical features.49 In addition, cursive Nabataean manifests an important change in that the letters are placed on an imaginary line, not suspended from an upper imaginary line as in Ancient Nabataean. This evolution, too, could be due to day-to-day writing on soft material either in Nabataean or in Syriac. The possible influence of Syriac on the increasingly cursive character of Arabic script cannot be dismissed entirely. Since the Arabic script is attested for the first time in Christian inscriptions, it seems likely that its evolution was influenced by people already trained in writing Syriac in an Arabian Christian milieu.50 Comparison of the first Arabic script and the almost contemporary, beautifully handwritten Syriac manu­scripts of the fifth and sixth centuries ce does not reveal many resemblances. No day-to-day Syriac writing contemporary with the oldest Palaeo-Arabic inscriptions has been preserved, only the examples dated 240–43 ce

41 Al-Jallad, ‘The Linguistic Landscape of Pre-Islamic Arabia’, p. 114. 42 Al-Jallad, ‘What Is Ancient North Arabian?’; Al-Jallad, ‘The Linguistic Landscape of Pre-Islamic Arabia’, p. 114. 43 Robin is preparing a book on South Arabia, that deals with Arabs and the use of Arabic, in Ḥimyar. 44 Knauf, ‘Arabo-Aramaic and ʿArabiyya’; Al-Jallad, ‘The Di­graph ‫ اى‬in the Quranic Consonantal Text’, p. 68. 45 The Old Arabic texts that have been discovered so far are listed in Macdonald, ‘Reflections on the Linguistic Map of Pre-Islamic Arabia’, p. 61; cf. also Robin, ‘Inscriptions’, pp. 545–50, and Macdonald, ‘Old Arabic’. 46 Nehmé, ‘Aramaic or Arabic?’, pp. 88–90. 47 Macdonald, ‘Arabia and the Written Word’, p. 52. 48 Nehmé, ‘Aramaic or Arabic?’. 49 Hoyland, ‘Insider and Outsider Sources’, pp. 269–70. 50 See below for the languages of the churches.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

Figure 4.2. P.Euphrate 18 (P.Euphr. inv. 19) © Image: Adam Bülow-Jacobsen. Reproduced with permission.

207

208

m u r i e l de bi é

Figure 4.3. Codex parisino-petropolitanus, Hand E; 6:  76–91 (MS Paris, BnF, Arabe 328, fol. 25v. Reproduced with the permission of Bibliothèque nationale de France.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

in parchments from Dura Europos and the Middle Euphrates. If we examine these documents, the similarity of the page layout and cursiveness between the Syriac, the Palaeo-Arabic script, and the most ancient Qurʾāns in the so-called Ḥijāzī script is more striking. As F. Déroche has noted, the oldest Qurʾānic manu­scripts are similar to documents, which is understandable since the pre-Islamic tradition of writing Arabic was probably exclusively documentary. Inscriptions and the colophons of the most ancient Syriac manu­scripts display similar cursive forms.51 We should imagine that Syriac epistles that circulated in Arabia were written in a form closer to that of documents than to that of manu­scripts, which involved calli­graphy. We can observe that the cursive Syriac script of documents leans towards the left (Fig. 4.2) whereas Palaeo-Arabic (Fig. 4.3) leans towards the right, perhaps in an attempt by scribes to mark the difference between the two visually. The examples of ortho­graphic diversity in Old Arabic ‘indicate that there was no unified ortho­graphy for Arabic in the pre-Islamic period’.52 Different traditions must have evolved locally wherever the Nabataeo-Arabic script was used. The chancelleries of different oases and tribal rulers could have developed their own particularities of writing, and these coexisted until the emergence of the Umayyad state and the increasing promotion of a unified Arabic script in chancelleries, probably during the reign of ʿAbd al-Malik.53 Yet the forms of most letters and in particular of the lām-alif ligature are instantly recognizable as Arabic in the fifth- and sixth-century inscriptions, which implies some form of standardization already by that time, as well as the transmission of this script in heterogenous, (predominantly) Christian Arabic milieus extending from Ḥimā in South Arabia to Dūmat al-Jandal and southern Syria. Before turning to a tentative interpretation of the social context of this early script, a comprehensive overview of the literacy situation of the Arabs before Islam will highlight the variety of languages and scripts they were aware of and used and challenge the idea that they lived in non-literate societies.

Civil Languages Since Arabs were present in the whole Middle East, they were exposed to several official languages, including Greek, Latin (to a much lesser extent) in the Eastern Roman Empire, Middle Persian in the Sasanian Empire as well as in Arabia, Sabaic, and even some Ethiopic in Ḥimyar. The Christian Arabs who read the Bible and liturgical books during the liturgy were in contact with Greek and Syriac as well

51 For a new history of Syriac scripts, see Briquel Chatonnet, ‘De l’écriture édessénienne à l’estrangelâ et au sertô’, and independently Healey, ‘The Early History of Syriac Script’. 52 Al-Jallad, ‘The Di­graph ‫ اى‬in the Quranic Consonantal Text’, pp. 74–75. 53 On the idea that the Arabic script evolved gradually from Nabataean at the courts of tribal chiefs in north-west Arabia, see Nehmé, ‘Epi­graphy on the Edges of the Roman Empire’. The unification of the script does not mean that there was a top-down imposition of Arabic (on this issue, see the contributions of Sijpestejn and Legendre in this volume). Ortho­graphic unification was also a slow process (Déroche, La transmission écrite du Coran, pp. 167–68).

209

21 0

m u r i e l de bi é

as Christian Palestinian Aramaic and Ethiopic, depending on their locations and confessional affiliations. Ethiopic was confined to Ḥimyar, and CPA was the secondary language of the official Chalcedonian Church in the provinces of Palestine and Arabia, whereas Greek was the official language of the Church in the Roman Empire, and Syriac was the language of the Chalcedonian and Miaphysite Churches and of the Church of the East on both sides of the Roman-Sasanian border and as far as India and China in the east and Arabia in the south. We do not know to what extent Arabic was part of the religious life of Christians. Jews, too, used Aramaic for oral as well as written purposes when in a Sabaic and Arabic environment. To make the matter even more complex, interactions between languages meant that loanwords and phrases passed from one language into another. Dialects still thrived, too, although standardized versions of a language were taught in schools and chancelleries, resulting in diglossic practice. The written form of administrative documents entailed the use of the official(s) languages(s) of the empire but also of standardized formulas and chancellery scripts that were different in documents issued in the capital versus those issued at a local level. Such multi­graphies of the same language, exemplifying local differences, are attested in early Arabic and bilingual papyri of the Islamic period.54 In the pre-Islamic period, the Arabs who were involved in translation and writing in the imperial chancelleries as well as the chancelleries of the Arab kings (phylarchs) were thus exposed not only to various languages but also to the various administrative scripts and formulas in each language, which varied according to the administrative and geo­graphic level. Latin and Greek

Since Latin was the official language of the Roman Empire, it was in use in the Eastern Roman Empire as the second language of the administration after Greek, which had been the dominant regional language since Alexander the Great. The phylarchs of the parembolai in the province of Palaestina thus belonged to a Greek environment that did not entirely obliterate the use of Aramaic. A network of phylarchs and the hierarchy among them can be determined thanks to the Greek titles they were granted by the emperors. Archiphylarchoi, symphylarchoi, and phylarchoi were appointed in the southern provinces of Phoinice, Palaestina, and Arabia. Aspebetos, a defector from the Sasanian Empire who became phylarch of Arabia, and Amorkesos, who was appointed phylarch of Palaestina in the 470s ce, are rather well documented in Greek literary texts.55 In the sixth century, al-Ḥārith b. Jabala/Arethas (r. c. 528–569 or 570), who was appointed chief phylarch in Arabia by Justinian, and his brother Abū Karīb in Palaestina were two major figures in the southern provinces in a Greek and CPA linguistic environment and are doc 54 Sijpesteijn, ‘Arabic Script and Language in the Earliest Papyri’. 55 Aspebetos is mentioned around 420 ce by Cyril of Scythopolis in his Life of the Palestinian hermit Euthymius, and Amorkesos is known from Malchus’s History.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

umented in Greek and Syriac sources.56 Likewise in the sixth century, a phylarch was appointed for each ducate in the Middle East.57 We also witness in the early years of Justinian’s reign a complexification in the phylarchs’ administration of the Arab tribes.58 The courts of Arab rulers, which were increasingly integrated into the empire, did not feel a need to invest in the development of an Arabic script that would be of no use in official interactions. Greek was the language of exchanges between Arab courts and imperial officers, even if mediated through interpreters, and it was very likely used in written as well as oral form. A few commemorative bilingual Greek-Safaitic inscriptions also indicate that at least some members of the nomadic communities of the Syro-Arabian desert had a command of Greek.59 In the Madaba area in Transjordan, the majority of inscriptions about the Arabs are in Greek but display traces of spoken Aramaic.60 Middle Persian

Another language that is seldom considered when examining the languages of the Arabs is Middle Persian, the official variety of Iranian in the Sasanian Empire. The Arabs who served the Sasanian kings in the army and tax collection were also exposed to at least oral and probably also written Middle Persian, not only in the Sasanian Empire but also in the Arabian Peninsula. As Sasanian allies, they participated in military and trade treaties with the Persians, like their counterparts in the Roman Empire did with the Romans. The earliest mention of the Lakhmids in service of the Persians appears in a bilingual Persian-Parthian inscription from Paikuli in north-east Iraq that lists among the vassals of the Sasanian king Narseh (293–302 ce) ‘ʿAmru king of the Lakhmids’ (ʿmrw lhmʾdyn MLK).61 This ʿAmr defended Manichaeism just after the execution of Mani and sent a Manichaean envoy to Seleucia-Ctesiphon with letters for Narseh.62 Whatever the language used, defending Manichaean doctrine and dispatching a mission to the Persian capital meant that members of ʿAmr’s Arab court at least wrote letters and perhaps read works of Manichaean doctrine or had such works read (in Syriac? Parthian? Persian?). The historian Ammianus Marcellinus mentions an Arab chief called Podosaces (probably a transcription of a Middle Persian name) fighting for the Sasanians against the Romans as early as 363 ce.63 We have little information about the Arabs allied with the Persians except from Roman historical sources (in Latin,

56 Millar, ‘A Syriac Codex from near Palmyra’, pp. 23–27. 57 Lewin, ‘Did the Roman Empire Have a Military Strategy?’, pp. 168–69. 58 Lewin, ‘Did the Roman Empire Have a Military Strategy?’, p. 174. 59 Al-Jallad, ‘The Linguistic Landscape of Pre-Islamic Arabia’, p. 121. 60 Knauf, ‘Arabo-Aramaic and ʿArabiyya’, pp. 199–200. 61 Humbach and Skjaervo, The Sassanian Inscription of Paikuli, line 92. 62 Trimingham, Christianity among the Arabs in Pre-Islamic Times, p. 168; Weber, ‘Narseh’. 63 Robin, ‘Ghassān en Arabie’, pp. 102–03.

211

212

m u r i e l de bi é

Greek, and Syriac), which consider them enemies, and from Islamic sources, which are suspicious of Sasanian influence.64 Such sources do not provide detailed information, since they know these Arabs only from the perspective of skirmishes with their own Arab allies and of the wars and peace treaties between the Romans and the Persians. Since the third century ce, the Persians had sporadically dominated the region of ʿUmān and the Yamāma for economic reasons.65 They were also present in the Ḥijāz: Mecca and Medina were part of the south-western region of Ērānšahr according to the Šahrestānīhā ī Ērānšahr, a Middle Persian geo­graphical work.66 In the fifth century, Bahrām V Gōr (the son and successor of Yazdgerd I), whose mother was purportedly the daughter of the Jewish exilarch, was raised in Ḥirta/al-Ḥīra at the court of the Lakhmid/Naṣrid king al-Nuʿmān I b. Imruʾ al-Qays (r. 390–418) and his son al-Mundhir (r. 418–462), who put him on the throne in 420 with the help of military assistance.67 Al-Nuʿmān purportedly built the palace of Khawarnaq to accommodate the Sasanian prince while the latter was being brought up at his court.68 Al-Mundhir was a tutor to the Sasanian prince, which implies that he was well educated in matters of war, hunting, and court culture, as well as, possibly, literature — poetry and court literature, either oral or written. At least some Arabs in Ḥirta/al-Ḥīra would have spoken Middle Persian because of the official contacts with the Sasanid court and administration (and the other way round). Since the middle of the sixth century, the Sasanians exercised indirect control over several parts of Arabia, including the Ḥijāz. Kosrow I Anōširavān (r. 531–79) appointed al-Mundhir III (r. c. 504–54) king of the Arabs living between ʿUmān, Baḥrayn, and Yamāma on one side and al-Ṭāʾif and the rest of the Ḥijāz on the other.69 A marzubān al-Zāra, ‘marzbān of the desert’, a Sasanian governor, had his seat in al-Zāra (where a jail of the Persian governor is mentioned) on the coast of the Persian Gulf. He used to receive Kosrow’s messengers who travelled on post-mules (burūd kisrā), which implies the circulation of letters and documents in Middle Persian and a local chancery.70 He ruled also indirectly Medina and Tihāma, where he was represented by an unidentified official (ʿamīl).71 The Jewish tribes of the Qurayẓa and Naḍīr who were ‘kings’ in Medina exacted tribute from the Aws and the Khazraj on behalf of the Sasanians. They probably used Arabic in their dealings with the locals, but the language of their interactions with the

64 Fisher and Wood, ‘Writing the History of the “Persian Arabs”’. 65 See Azarnouche, ‘Arabes et Iraniens avant et au début de l’islam’, pp. 159–60 on Ardashir’s progress in the region, and p. 160. 66 Kister, ‘Al-Ḥīra’, pp. 145–46. 67 Klíma, ‘Bahrām V’. 68 The castle was the subject of a poem written by the Christian poet of al-Ḥīra, ʿAdī b. Zayd (d. about 600). Würsch, ‘Ḵawarnaq’. 69 Lecker, ‘The Levying of Taxes for the Sasanians in Pre-Islamic Medina’, p. 115. 70 Lecker, ‘The Levying of Taxes for the Sasanians in Pre-Islamic Medina’, p. 122. 71 Kister, ‘Al-Ḥīra’; for a new interpretation of the passage of Ibn Khordādbeh, see Lecker, ‘The Levying of Taxes for the Sasanians in Pre-Islamic Medina’, pp. 114–15.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

ʿamīl (through interpreters) and certainly the language of communication of the ʿamīl with the distant marzbān would have been Middle Persian. Administrative

and fiscal documents were probably exchanged between the central chancery in Seleucia-Ctesiphon, the provincial one in al-Zāra, and the representative in Medina, presumably with different scripts and local practices. Accounting documents were also likely written in Middle Persian by bilingual secretaries in Mecca and al-Zāra at the two levels of the Sasanian local administration. In the sixth century, the Sasanians also exploited natural resources in the Najd (silver and copper) and in Yemen (Raḍrāḍ mine), which implies colonies of servicemen and labourers (of at least a thousand people) and some interaction with the local people.72 The memory of the Middle Persian names of the families settled in the village near the mine was transmitted by al-Hamdānī, suggesting direct links between the Persians and the Arabs since the pre-Islamic period that persisted in Islamic literature.73 It seems that Kawād I (r. 488–96 and 499–531) convinced the Ḥujrid/Kindite al-Ḥārith al-Malik b.  ʿAmr to promote Mazdaism (more precisely Mazdakism) in the Najd and Tihāma — which makes sense if the Sasanians exploited mines in the area.74 In 500–520, a Kindite prince was established king of Ḥirta/al-Ḥīra by the Persian King that shows his involvement in Arab affairs. During the latter half of the sixth century, an Arab of the Khazraj, ʿAmr b. al-Itnaba, was made ‘king of Medina’ (or ‘king of the Ḥijāz’) by another king of Ḥirta/al-Ḥīra, al-Nuʿmān III b. al-Mundhir (r. c. 580–602), who, like his fifth-century namesake and predecessor, was a client of the Sasanians. It seems that at this time Medina was no longer controlled from al-Zāra but directly from Ḥirta/al-Ḥīra and that the Jews were no longer tax collectors but tribute payers.75 Although it became less direct, Sasanian control of Medina and Tihāma continued in the latter half of the sixth century. Under Kosrow II Parwez (r. 590–628) Sasanian troops invaded Yemen in 575 and established a direct dominion there.76 The Sasanians may have still considered Mecca part of their realm:77 the king sent his governor from Yamāma to Mecca, and contacts between the governor of Ḥimyar and that of Mecca, evidently over taxes, are attested; these exchanges took place in either Middle Persian or Arabic. Middle Persian was thus used not only in north-east Arabia but also in the west, in the Ḥijāz. Although recent scholarship has argued that the area was less wealthy and more isolated from the rest of the peninsula and the Roman Empire than was previously thought,78 these examples show that the Sasanians were interested in levying taxes from the region’s inhabitants, and probably not only from poor shepherds. We should thus not downplay the importance of the

72 Azarnouche, ‘Arabes et Iraniens avant et au début de l’islam’, pp. 176–77. 73 Azarnouche, ‘Arabes et Iraniens avant et au début de l’islam’, p. 169. 74 According to al-Yaʿqūbī’s Taʾrīkh; Robin, ‘Les religions pratiquées par les membres de la tribu de Kinda’, p. 238. 75 Lecker, ‘The Levying of Taxes for the Sasanians in Pre-Islamic Medina’, p. 123. 76 Azarnouche, ‘Arabes et Iraniens avant et au début de l’islam’, pp. 166–68. 77 Lecker, ‘The Levying of Taxes for the Sasanians in Pre-Islamic Medina’, p. 114 n. 23; p. 121. 78 Shoemaker, Creating the Qur’an.

213

21 4

m u r i e l de bi é

Ḥijāz in Late Antiquity too much. This evidence also attests to the presence of at least minimal Middle Persian, probably in written form, in the region to meet the administrative needs of the Persian empire, and of Arabic (in Nabataean script?) for the intermediary officers who worked for the Nasrids in Ḥirta/al-Ḥīra. Since conversions to Zoroastrianism took place in Arabia and Persians settled in various parts of the peninsula for economic reasons and founded fire temples for their own use, the Arabs had direct knowledge of the religious practices and doctrines of the Persians as well as of the language used for their rituals.79 Economic, administrative, military, and religious Middle Persian loanwords entered Arabic and attest the linguistic interactions between the two languages.80 Sabaic

In southern Arabia, the language originally came from the north, from Ugarit. Official inscriptions in Sabaic and South Arabian script were carved in oasis settlements such as Mārib and in the Jawf settlements.81 They bear mostly building-related or votive texts.82 The writing in these inscriptions was sophisticated, depending on the type of text and the associated medium, which points to scribes trained in each script and mode of writing: the texts were chiselled on inscriptions on stone with a lithic striker or written with ink and reed pen (qalam) on leather, parchment, or papyrus. Some were even written on fabric, wood, camel bones, or palm sticks.83 Two scripts coexisted for Sabaic: musnad, which was used for official inscriptions, and zabūr, a cursive that served for everyday documents and letters. Some 870 documents inscribed in Sabaic have been studied in the past decades, and they attest to the spread of writing for everyday purposes in the Ḥimyarite kingdom. The documents include 214 letters to business partners or family and friends. These letters were written by professional scribes who wrote their messages in their clients’ names but using their own words, referring to the actual sender in the third person.84 The exchange of official letters is also attested in literary texts, especially in the dossier concerning the Christian martyrs of Najrān in the sixth century. The South Arabian script was used to write Arabic, a practice that continued into early Islamic times.85 Inscriptions show the use of Sabaic in official contexts, civic and religious (in temples for instance) but include Arabic words in more mundane contexts like funerary ones. Arabs were present in Ḥimyar as the official title of the Sabaean kings attest (they are called ‘kings of the Arabs’), but the Sabean elites who dominated since the third or fourth century used Sabaic.

79 80 81 82 83 84 85

Azarnouche, ‘Arabes et Iraniens avant et au début de l’islam’, pp. 168–69 and 176. Azarnouche, ‘Arabes et Iraniens avant et au début de l’islam’, pp. 175–76. Robin, ‘Les langues de la péninsule arabique’, pp. 92–101 on South Arabian languages. Stein, ‘Literacy in Pre-Islamic Arabia’. Stein, ‘Literacy in Pre-Islamic Arabia’, pp. 264–65. Stein, ‘Correspondence in Pre-Islamic Yemen’, p. 344. Robin, ‘Les graffites arabes islamiques écrits en caractères sudarabiques’.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

When the Najranites became allies with the Axumites and Sabaic elites were less powerful, the Arabs reemerged as a major group and Arabic came to the fore and became visible in Arabic script. Ge’ez

In the sixth century, King Abraha (r. 525–at least 553), who became king in Sanaa after taking part in the expedition that reconquered it from the Jewish king Dhū Nuwās, purportedly called on his interpreters when he received the Qurayshite king of Mecca for negotiations over a booty of camels, according to the later Life of Muḥammad.86 An Aksumite general and former slave of the Romans who did business in Adulis, Abraha did not speak Arabic himself, according to the historian Procopius. He participated in the Abyssinian expedition in Najrān and revolted against Esimiphaios (Sumuyafaʿ Ashwaʿ), the Christian Ḥimyarite viceroy appointed by the negus Kaleb, with the support of part of the Abyssinian occupation forces. After building a cathedral in Sanaa, Abraha sent a letter to the negus, in all likelihood in Ethiopic.87 His case illustrates the network of exhanges with local Arab chiefs as well as written correspondence between the Ḥimyarite kingdom and the Aksumites. The presence of Ethiopic is also evidenced by official inscriptions.88 Numerous political, economic, and religious links are attested in inscriptions and texts, yet relatively few detailed studies have been devoted to the connections between Ethiopic Christianity and the Arabs before or after Islam.

Languages of Imperial Diplomacy There was no reason for Arabic to be used in the context of imperial diplomacy. However, the situation of the Arabs in the Roman and Sasanian Empires was discussed during the negotiation of peace treaties in the sixth century, and the Arabs took part in these discussions since they were supposed to agree to the conditions. The treaties themselves do not survive; we have only citations in East Roman chronicles and histories. One version in Greek and one in Middle Persian were likely made for each empire, which would then send a copy to its Arab allies. Translators, notaries, and secretaries took part in the process, and some of them were likely Arabs. In Constantinople, a special department dealt with all matters related to the ‘barbarians’: the Scrinium Barbarorum, or Bureau of Barbarians, attested in the fifth-century Notitia Dignitatum. In addition to dealing with tributes and treaties, the bureau appears to have had a corps of translators for embassies, visits, and negotiations. This demonstrates that Romans learned Arabic and that Arabs were 86 Guillaume, The Life of Muhammad, pp. 34 (text), 25 (translation). 87 Guillaume, The Life of Muhammad, pp. 29 (text), 21 (translation). 88 Robin, ‘Les langues de la péninsule arabique’, pp. 96–97, on Ethiopic written as pseudo-Sabaic.

215

21 6

m u r i e l de bi é

trained in Greek and/or Latin for diplomatic purposes. The Arabs took an active part in Roman diplomacy, especially in the sixth century. As foederati they were in a treaty relationship with the Romans that implied the existence of written documents in Greek. When they visited Constantinople, the phylarchs would use Greek directly or through translators.89 In central Arabia, Kindite/Ḥujrid princes fought the Lakhmids and were in a treaty relation with the Roman Empire. Around 502, Anastasius signed a treaty with Arethas/al-Ḥārith b.  ʿAmr b.  Ḥujr, who was at war with the Romans’ Arab allies. During the brief reign of Sumuyafaʿ Ashwaʿ, Justinian sent two embassies to Ḥimyar in order to obtain help for installing his Arab ally Qays b. Salama b. al-Ḥārith over Maʿadd so that they could together fight the Sasanians.90 During his mission, the ambassador Nonnosos visited Kinda and Aksum before going to Ḥimyar. Greek was the language of these discussions, which involved the Arab clients of the Roman Empire in triangular diplomacy against the Sasanians far beyond the borders of the Roman Empire. Each party would have relied on staff trained in Greek for these negotiations. The Roman Empire’s dealings with the Arabs concerned not only their allies but also the allies of the Persians. The peace treaty of 561 between the Romans and the Sasanians indicates that the Romans paid subsidies to al-Mundhir III, the ally of the Persians, so that he did not attack them and their Arab allies. Peter, the magister officiorum and Roman envoy to Seleucia-Ctesiphon, met with the King of Kings Kosrow I Anōšīravān to discuss the subsidies to the Lakhmids. He insisted that the subsidies were not a regular tribute (which would have implied a situation of inferiority for the Romans). These ‘gifts’ were sent by the imperial post, and al-Mundhir reciprocated by sending symbolic gifts in return.91 Six Roman translators and six Persians are mentioned in the discussions, but no Arabs, probably because their presence was not needed for writing the official Greek and Middle Persian records of the outcome. Unfortunately, we do not know whether the Arabs needed Greek or Middle Persian interpreters in these three-way negotiations. Such exchanges were ritualized and described in manuals of diplomacy. They were similar to the arrangements that existed between the Roman Empire and German and Hunnic tribes. In 563, al-Ḥārith b. Jabala/Arethas, the ally of the Romans, came to Constantinople in order to complain about attacks against him by ʿAmr, the ally of the Sasanians, who was angry about the treaty’s conditions. While in the capital, al-Ḥārith met with Theodosius, the pope of Alexandria and the representative of the Miaphysites in Constantinople, and he wrote himself a letter to Bishop Jacob Baradeus, who ordained a new Miaphysite hierarchy.92 Civil and ecclesiastical affairs were thus 89 Malchus in his Greek history describes Amorkesos under Leo’s reign, and John of Ephesus in his Syriac Ecclesiastical History describes the great impression Arethas made when he came to the Roman court. 90 Bowersock, The Throne of Adulis; Robin, ‘Before Ḥimyar’ and ‘Le royaume Hujride dit royaume de Kinda’. According to Islamic sources, al-Ḥārith occupied al-Ḥīra/Ḥirta for a brief period (525–28). 91 Shahid, Byzantium and the Arabs in the Fifth Century. 92 Documenta ad Origines Monophysitarum, p. 100.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

addressed in the capital with various interlocutors and documents. During the discussions and diplomatic exchanges that took place around 567 over the payment of the Lakhmids, the Arab king ʿAmr and forty Arab leaders accompanied the Persian envoy to the emperor Justin II in Constantinople, in what appears to have been a display of power.93 In these circumstances, too, letters were exchanged to summon or announce the parties and organize the embassy in each realm, and multilingual discussions took place through translators. Some delegates probably also knew some Greek or Middle Persian, and it would not be surprising if a few were more or less bilingual. The Arab leaders on both sides participated in the diplomatic exchanges between the Romans and the Sasanians, travelled to Seleucia-Ctesiphon and Constantinople, were received as an official party in both realms, and were included in the face-to-face negotiations. Such participation required sophisticated diplomatic skills and trained personnel at the Arab courts in the sixth century, before the Arab kings fell out of grace in both empires. In the Sasanian Empire, too, a ‘bureau of Arab affairs’ is mentioned in sources of the Islamic period. Although we do not have much information about it, we can surmise that it, like its Roman counterpart, managed the diplomatic, military, and financial aspects of the empire’s dealings with the Arabs along the borders with the Roman Empire and in the regions of the Arabian Peninsula where the Sasanians maintained more indirect rule. The Islamic tradition (al-Yaʿqūbī and al-Iṣfahānī) reports the intrigues of the ʿIbāḍī aristocrat and poet ʿAdī b. Zayd94 at Ctesiphon in 580 to ensure that al-Nuʿmān IV would be selected over his brothers as king of Ḥirta/al-Ḥīra. His father Zayd b. Ḥammād had already been a kingmaker and secured the election of al-Mundhir IV in around 575. Zayd was a secretary of the Naṣrid king, and since he was on good terms with the Persian dihqān Farrukhmāhān, he secured the coveted position of postmaster, normally reserved for the sons of Persian provincial governors. ʿAdī b. Zayd, who was educated with the son of the Persian dihqān, mastered both Middle Persian (and the notoriously difficult Pahlavi script) and Arabic and received an aristocratic education. He caught the attention of the Persian king Kosrow I Anōšīravān (r. 531–79), who appointed him his secretary and interpreter, and he served with his brothers as members of the postal and intelligence service in the bureau of Arab affairs in the capital city Maḥoze (Seleucia-Ctesiphon).95 He was sent to Constantinople as part of an embassy, which signals his high rank. His family acted as intermediaries between the Sasanian court and the Arabs. He purportedly organized a presentation of al-Nuʿmān and his brothers to king Hormizd IV, who interrogated them in order to decide which to name king of Ḥirta/al-Ḥīra. When Zayd b. Ḥammād, suspected of plotting against al-Nuʿmān, was executed by the latter, his son, ʿAdī, who had Kosrow’s ear, managed to turn the Sasanian

93 Shahid, Byzantium and the Arabs in the Fifth Century, p. 309. 94 Powers, ‘Demonizing Zenobia’, pp. 134–38. 95 Rothstein, Die Dynastie der Lahmiden in al-Hîra, pp. 109–10; Shahid, Byzantium and the Arabs in the Sixth Century, p. 315.

217

21 8

m u r i e l de bi é

ruler against al-Nuʿmān. What is interesting here is that that language was a key factor in the intrigue. The ability of ʿAdī to speak both Persian and Arabic allowed him to manipulate the events; the court sent an interpreter who did not understand Arabic well enough to realize what ʿAdī was doing.96 It was thus not just individuals who were influential at the Sasanian court but an entire Arab family/tribe, too. Such families acted a conduit to the capital for the concerns of the Arabs at Ḥirta/ al-Ḥīra and consequently played an important political role. The intricate ArabIranian diplomacy in which they participated took place in Middle Persian as well as Arabic, through translators and direct knowledge of both languages. Beyond the special role of this particular family, this example suggests that Arabs maintained a more or less continuous presence in the capital (though perhaps not always at such a high level), ensuring their representation in the empire’s Arab affairs and the preservation of links between the Sasanian and Arab courts. The Arab kings gained the recognition of the empire they served and entered, at least to some degree, its cultural and linguistic sphere. Pleased with al-Nuʿmān, Hormizd IV gave him a jewelled crown as a sign of the authority granted to him over the Arabs. Justinian took a similar step when in 529 he awarded al-Ḥārith b. Jabala/Arethas the title of basileus, ‘king’, and bestowed on him a crown. The Arabs received titles from the empires with which they were allied. Various inscriptions in Greek list official titles granted to Arab phylarchs by the Roman emperors.97 They enable historians to follow the development of some of these titles’ carriers.98 Such studies demonstrate the Arab phylarchs’ integration into imperial structures, hierarchy, and idiom. Phylarchs were appointed for each province, and they and their chancelleries were well aware of the written practices of the imperial administration. The same was probably true of the Sasanian Empire, as the examples of Podosaces and the abovementioned Aspebetos, two Arab allies of the Sasanians, suggest. Ammianus Marcellinus refers to Podosaces as malechus Podosacis nomine. Malechus may be the transcription of malik, the Arabic for king, which would imply that Podosaces, a name probably transliterated from Middle Persian, was recognized as such by the Sasanians.99 Aspebetos is the Greek transcription of the Parthian title aspbed (chief of the cavalry).100

96 Talib, ‘Fathers and Husbands’, pp. 243–47, and Powers, ‘Demonizing Zenobia’, pp. 133–38. I am grateful to Alison Vacca for these references. 97 Gatier, ‘Les Jafnides dans l’épi­graphie grecque au vie siècle’. 98 Gatier, ‘Les Jafnides dans l’épi­graphie grecque au vie siècle’. 99 Ammianus Marcellinus, XXII. 2. 4. We cannot exclude the possibility that malechus simply refers to the well-known personal name Malik (Malichus was the traditional name of Nabataean kings); but since just a few words earlier Ammianus mentions the title of a very high-ranking official, malechus might well be used as a title here. It could also parallel Caesar and Kosrow as a metonymical designation for a Nabataean king. 100 The name refers to an aspbed, not a spāhbed as is often stated. This was a very high rank, equivalent to general, that would not fit an Arab phylarch. On these titles, see Gyselen, ‘Spāhbed’, and Chaumont, ‘Aspbed’.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

The emperors acknowledged their clients’ authority with symbolic insignia, integrating them into the imperial idiom of power. In both empires, the power relationship entailed control on the part of the emperors and mutual negotiations and lobbying on the part of the Arabs, undertaken in the idiom of the empire and at its geo­graphical core. The Arab elites were active actors on the geopolitical scene and mastered the game rules of their powerful patrons through intermediaries in the capital who had absorbed the imperial language and codes. Between the fourth and sixth centuries, then, the Arabs were increasingly involved in diplomatic exchanges with the Roman and Sasanian Empires. They constituted important subjects for both empires, and each polity sought to satisfy its own allies and to pacify and buy the neutrality of the enemy’s allies, especially in times of war. The Arab leaders received official recognition through the titles and crowns granted by their respective patrons. They were thus integrated into the linguistic, cultural, and political idioms of the empire they served — as well as those of the rival empire, at least on some occasions. The Arab elites participated in the multilingual and multicultural world of Late Antiquity on a macro level. They also engaged in military activities on the ground and interacted with the international milieus of imperial armies and neighbouring allied troops on a micro level.

Languages of the Armies Most nomads were not aware of the affairs of empires, but those who fought as allies to the great powers of the time (the Sasanians, the Romans, and the Ḥimyarites) were part of their respective armies and shared the tactics, commands, and hierarchies of the regular soldiers recruited across each empire. They participated in battles and sometimes shared encampments with soldiers who spoke all sorts of languages and interacted with the upper hierarchy in the official languages of the empires. At least the Arab troops who fought on the borders were thus involved in the multilingual culture of the late antique armies. In the Roman Empire, Latin probably remained the first language of the army, in which ranks and positions were expressed. Latin military titles were simply transliterated into Greek (and into Syriac). Latin was the language of the epitaphs of soldiers dispatched to the different corners of the empire and of the inscriptions in garrison churches. It was a common idiom shared by the soldiers mentioned in these inscriptions who hailed from all parts of the empire. Like French is today the language of the Foreign Legion, its military hierarchy, and its orders as well as the everyday lingua franca of its soldiers from all over the world, Latin fulfilled a similar role in Late Antiquity. For example, the tribe of the ʿAwidh mentioned in inscriptions in the ḥarra (basalt desert) provided troops to the Romans, and other auxiliarii were recruited among the nomads.101 The tribe’s members thus shared the Latin lingua franca of the Roman army. The Life of Symeon Stylite reports that 101 Sartre, ‘Rome et les Arabes nomades’, pp. 45–46.

219

220

m u r i e l de bi é

al-Nuʿmān I invited the commander of the Roman troops in a neighbourhood of Damascus to a feast.102 Accounts of treason also often feature similar circumstances in which Arab leaders and high Roman or Persian commanders shared meals. Table commensality created cultural and linguistic interactions — possibly including written ones if invitation letters were involved. Among professional soldiers we know of strategoi parembolon nomadon, strategoi of the nomads or strategoi of the nomad auxiliarii, a Roman denomination for professional officers103 of whom at least some were recruited among the Arabs. The auxiliary troops of the Arabs were immersed in the multilingualism of the Roman army, where the languages of the individual soldiers would encounter the official languages, Latin and Greek. Phrygian troops, for instance, are mentioned alongside Arab units on the right flank of a battle line in 531.104 The movements of the different units had to be coordinated in a common language, which was Greek or Latin in this instance. The Romans raised auxiliary units from the nomads east of the Ḥawrān, and these were under the command of leaders from prominent tribal groups.105 Greek and Safaitic inscriptions attest to contacts between Greek speakers and Safaitic-writing soldiers who were part of these auxiliary units.106 One Greek inscription contains the Persian name Aurmidarēs, perhaps referring to an Iranian mercenary, which shows how international the units in the desert could be.107 ‘Ethnarchs’, tribal leaders recognized by Rome, also held the title of strategos and were charged with liaising with the nomads. The same was certainly true for the Arabs who fought with the Sasanians and who received orders in Middle Persian and mixed with soldiers from different parts of the Sasanian Empire, although we lack explicit sources on them. Arab troops were active far from their home regions during the wars and were exposed to other cultures and languages within the imperial armies to which they belonged as well as on the battlefield. Arab troops from Kinda, Madhḥij, and Murād who fought for Ḥimyar were immersed in a Sabaic environment. In Haram, a ‘pseudo-Sabaic’ tainted by Arabic is attested since the second century bce and demonstrates the influence of Arabic.108 A chief of the Arab auxiliaries of the king of Saba is mentioned in the fourth century ce, when Arabs increasingly entered the Ḥimyarite kingdom and settled alongside South Arabian tribes, creating a binational identity — and increased linguistic interactions.109

102 Trimingham, Christianity among the Arabs in Pre-Islamic Times. 103 The Notitia Dignitatum lists the units of ‘Arabs’ as equites Saraceni indigenae and equites Saraceni Thamudeni. 104 Lewin, ‘Did the Roman Empire Have a Military Strategy?’, p. 177. 105 Macdonald, ‘Romans go home?’. 106 Al-Jallad and Bernard, ‘New Safaitic and Greek Inscriptions from the Jordanian Ḥarrah’. 107 Al-Jallad and Bernard, ‘New Safaitic and Greek Inscriptions from the Jordanian Ḥarrah’, p. 75. 108 Robin, ‘La pénétration des arabes nomades au Yémen’, p. 77. 109 Robin, ‘La pénétration des arabes nomades au Yémen’, p. 81.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

Languages of Trade Like soldiers, Arab merchants circulated on commercial roads and did business in the many languages of the region. Negotiations took place in Greek and possibly Aramaic, Armenian, or Ethiopic as well as in Middle Persian and other Iranian languages, especially in Central Asia. However, we do not have direct evidence of the languages spoken and possibly also written in the context of trade. Treaties between the Romans and the Sasanians stipulated in which border cities trade was allowed, a situation that implied written permits. Customs tariffs were levied, which also must have required the production of written accounts. Arabs were involved in these exchanges of goods and in the accompanying administrative controls.

Languages and Scripts of Legal Documents Latin was also the language of Roman law, both civil and ecclesiastical, although it had been losing ground in Late Antiquity since the fourth and fifth centuries. However, D. Feissel has shown that it was used again in the sixth century at the time of Anastasius’s and Justinian’s legal enterprises until the disappearance of the law school of Beirut in the earthquake of 551. Latin was not only the language of private and official documents written and read by professionals; it was also sometimes displayed on public milestones, for example to announce the right to provide asylum that was granted to sacred spaces (churches, baptisteries, martyria) in a continuation of a pre-Christian practice. The spatial delimitation of the landscape determined by the emperor was physically marked by milestones inscribed in Greek or Latin.110 Such a milestone was found in Damascus in the district of the Umayyad Mosque, built on the old basilica of Saint John the Baptist.111 Laws were also displayed in Constantinople and provincial cities for variable periods of time in the form of inscriptions on copper, stone, or sometimes fabric that were displayed in public spaces, including, increasingly, churches.112 Some such displays were quite spectacular, such as the twenty-chapter constitution of Anastasius written on two hundred lines and disseminated in several copies in the province of Arabia.113 This general law dealing with military administration was engraved in several garrisons in Arabia, including that of Bostra. Everyone, even those who could not read the text themselves, could see it publicly displayed and have it read (and translated) to them. Everyday legal documents can also offer insights on the use of languages. Archives of documents on papyrus and leather found in the twentieth century in the area of the Dead Sea (the so-called Babatha documents, first century ce) and

110 111 112 113

A revival of Latin is noticeable under Anastasius and Justinian. Feissel, ‘Introduction’, pp. 531 n. 1 and 565 n. 32. Feissel, ‘Épi­graphie et constitutions impériales’. Feissel, ‘Épi­graphie et constitutions impériales’, p. 22.

221

222

m u r i e l de bi é

on the Middle Euphrates (third century ce) have shown that archival practices dating back to the Ancient Near East were still employed in the first centuries of the Christian era, including the use of simple or double documents, sealing practices, and witness signatures. Jewish, Nabataean, and Syriac legal texts and traditions stem from a much older shared Aramaic tradition, an ‘Aramaic Common Law’.114 The charred papyri written in Greek unearthed in Petra and dating to the sixth century ce have also added to our knowledge of legal practices. These documents reveal the entanglement of languages at various levels: onomastics, the language of the document itself, and the languages of the subscriptions that each party wrote in her/his own language and script (Greek, Aramaic, Hebrew, Nabataean in the Dead Sea archive; Greek, Latin, Syriac, and Nabataean in the Euphrates one).115 Although Greek was the official language of the Roman Empire, the Latin vocabulary of law as well as Semitisms derived from Aramaic (including in Egypt) pervaded the Greek. These documents are thus fascinating sites for observing the complex use of languages in everyday activities such as the sale of goods and slaves, marriage, inheritance, the payment of taxes, and the settlement of disputes. They are a window on multilingualism as well as multi­graphism. Several languages and scripts are often used in a single document and in different documents in the same archive. The same degree of complexity applies to the scripts used in the documents. The ‘celestial letters’ of the Latin documents produced in the imperial chancellery in Constantinople are a well-known example of a distinctive documentary script. In the Near East, in the so-called double documents, two different g­ raphics are displayed for the same language and script. The scriptio exterior is of a chancellery type, formal and well traced, in the inferior part of the folio, whereas the scriptio interior, to which one would refer in case of litigation, conveys an almost identical or abridged text in a smaller and more cursive script traced on the upper part of the folio.116 Copies of contracts were kept in city archives, including in Petra, as suggested by the papyri found there, and in Madāʾin Ṣāliḥ, where an inscription on a tomb indicates the existence of a copy on soft material deposited in the archives.117 Arabs were part of this world of written contracts, both among themselves and with the non-Arab populations of the dominions in which they lived. Arabic poetry provides evidence of written treaties, alliances, safe-conducts, testaments, and records.118 One of the Greek papyri from Petra mentions the phylarch Abū

114 Healey, ‘New Evidence of the Aramaic Legal Tradition’, p. 125. 115 On Nabataean papyri, see Yadin and others, The Documents from the Bar Kokhba Period in the Cave of Letters (final publication of the Jewish ‘cave of letters’); Cotton and Yardeni, Aramaic, Hebrew and Greek Documentary texts from Nahal Ever and Other Sites. 116 For a material description of the documents in the Babatha archive, see Puech, ‘La grotte aux lettres de Wadi Khabra’, pp. 37–38. 117 Healey, ‘New Evidence for the Aramaic Legal Tradition’, p. 122. 118 Al-Azmeh, The Emergence of Islam in Late Antiquity.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

Karīb, who intervened as arbitrator in a dispute concerning the sale of a vineyard between two deacons in Petra in Palaestina III.119 Observing the ‘complete absence of literary texts, chronicles, treatises, poetic pieces, myths, or rituals’, some scholars have jumped to the conclusion that ‘the extent of literacy present in the Arabian Peninsula was effectively limited to “the ability to leave behind spontaneous and brief rock graffiti, which serve the sole purpose of passing the time and fulfill no communicative function”’.120 The Ḥijāz, in particular, is still often presented as a region excluded from the major communication routes, both pagan and home to an illiterate population. As S. Shoemaker has also recently asserted, ‘the social and cultural context for Muhammad’s new religious movement was in fact a nonliterate, tribal society where the use of writing was virtually nonexistent’.121 This might be true for literary texts, but it does not mean that there was no written tradition at all. Admittedly, few people were literate,122 but as noted above, a low literacy rate was characteristic of all late antique societies. Several testimonies, albeit difficult to interpret, point to a tradition of contracts between Arabs that imply literate practices, although not on a literary level. As expected and as elsewhere in the late antique world, these were drafted by professional scribes and not by the individuals involved, but they highlight the recourse to written legal documents. Islamic traditions report that the Qurayshites at the beginning of Islam wrote a document on marriage law with other tribes and hung it up in the Kaʿba.123 The name of the writer of the deed was recorded. Although this testimony has been doubted by modern commentators, it would not be surprising if important contracts were archived for future reference and displayed, as laws were elsewhere in Late Antiquity. The tradition also mentions that the seven pre-Islamic poems (al-Muʿallaqāt) that became classics of Arabic literature were written on fabric (silk or linen)124 and hung on the curtains of the Kaʿba. Whatever the historical reality of these claims,125 they may reflect a memory of important documents hanging on display in the Kaʿba. They also suggest the existence of scribes trained in writing documents on soft materials. The Life of Muḥammad records that Umayya b. Abī Ṣalt, who fought on the side of King Abraha when he went up against Mecca, boasted, ‘My people are Iyād [members of the Iyād tribe] … Iraq’s wide plain is theirs … they read and write’.126 It is likely that Arab kings archived or displayed

119 Millar, ‘A Syriac Codex from near Palmyra and the “Ghassanid” Abokarib’. 120 Shoemaker, Creating the Qur’an, p. 126, citing Robin and Stein, respectively. See also Al-Azmeh, The Emergence of Islam in Late Antiquity, p. 40. 121 Shoemaker, Creating the Qur’an, p. 142. Montgomery, ‘The Empty Ḥijāz’ challenges this idea. 122 Macdonald, ‘Literacy in an Oral Environment’. 123 Ibn Hishām, Sīra, ed. by Wüstenfeld, pp. 230 (text), 159 (translation). 124 On writing materials, see Maraqten, ‘Writing Materials in Pre-Islamic Arabia’. 125 The name of the poem collection may not mean that they literally hung in the Kaʿba, but this interpretation seems to imply that written texts may have been thus suspended and displayed there before Islam. 126 Guillaume, The Life of Muhammad, pp. 32 (text), 23–24 (translation).

223

224

m u r i e l de bi é

official agreements between themselves and other rulers, written in Nabataean and possibly also in Arabic (in Nabataean or Arabic script), for the benefit of their communities, even though no such documents have come down to us. The dearth of surviving evidence is not surprising in view of the situation in Syriac: only a handful of legal parchments were preserved in Dura Europos and the Middle Euphrates, and entirely by chance. They constitute the only evidence that Syriac legal documents were ever written. As already noted, the Sasanians levied tariffs in the Arabian Peninsula and more precisely in the Ḥijāz, especially in the last quarter of the sixth century, which means that documents in Middle Persian were written by Persian scribes or by Arabs trained by the Sasanians, although no such documents have been found so far outside Egypt. Another indirect passage in the Qurʾān (2:282) prescribes that all debts should be recorded by a scribe: O ye who believe! When ye deal with each other, in transactions involving future obligations in a fixed period of time, reduce them to writing. Let a scribe write down faithfully as between the parties; let not the scribe refuse to write: as God has taught him, so let him write. This passage suggests that there were scribes to whom one could turn to have legal documents drawn up, and there is no reason to think that the same was not the case in the Ḥijāz. The following verse (2:283) notes that if a scribe cannot be found, the agreement should be backed by some sort of security deposit.127 Obviously scribes could not be found everywhere, and writing contracts required knowing not just how to write but also how to draft legal formulas.128 Professional scribes who were skilled not only in writing but in writing lawful documents, in the manner of notaries, thus existed at the beginning of Islam and likely continued a pre-Islamic tradition. Muḥammad himself had scribes that are mentioned by the Islamic Tradition. This explains how he could have had the ‘Constitution of Medina’ or the ‘Umma Document’ composed: if a tradition of legal documents already existed,129 it would have been natural for him to draw up a written agreement between the Jewish and non-Jewish tribes of Yathrib, outlining the terms of their union as one community under his leadership.130 This was a major political document that is 127 Shoemaker concludes that this is ‘another sign pointing to the Qur’an’s genesis in a fundamentally nonliterate context, even if there may have been some very limited use of writing in its immediate milieu’ (Creating the Qur’an, p. 126). 128 On debt recording in early Islam, see Tillier and Vanthieghem, ‘Recording Debts in Sufyanid Fustat’. 129 See Khan, ‘Remarks on the Historical Background and Development of Early Arabic Documentary Formulae’ and ‘The Historical Development of Early Arabic Documentary Formulae’. 130 Shoemaker, in Creating the Qur’an, p. 131, reaches the opposite conclusion, although he notes that this example is puzzling: ‘How should we square the production of this document with the fact that by all indications, Muhammad and his followers appear to have lived in a nonliterate society? I have to admit, the answer is not entirely certain, and Macdonald, Stein, and others do not provide us with any suggestions’. He adds: ‘Even if there were a few individuals in Mecca and Medina who were capable of writing basic documents — simple contracts, receipts, lists, and so on — there is a clear and broad

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

better understood if a tradition of contract writing already existed in Arabia, in general, and in the Ḥijāz, in particular. Traces of Arabic can also be found in contracts otherwise written in Nabataean, and these suggest that Arabic was the language of the parties involved and thus came to pervade the traditional Aramaic formulas. Strikingly, documents written in Arabic in the Islamic period in turn preserved the Aramaic tradition and display characteristics that do not come from the Greek precedents that scholars until recently thought supplied the model for administrative documents in the former Roman Empire.131 The standard form of Arabic formulas and vocabulary at an early stage seen in the eastern Islamic empire as well as in textual sources from al-Andalus suggests a common pre-Islamic origin.132 A bilingual Greek-Arabic text contained in a 22/643 papyrus names different scribes for the Greek and Arabic parts and suggests that scribes had already mastered not only the Arabic script but the legal formulas that were in use in pre-Islamic times. Unfortunately we know next to nothing about the education and training of Arab scribes in the pre-Islamic and early Islamic periods.133 However, we can infer from both literary sources and later practices that there were already standards for the activity of writing Arabic documents on soft media in legal contexts involving Arabian chieftains.

Religious Languages Inscriptions in Ancient South and Ancient North Arabian contain votive texts and hymns to Arabian gods and goddesses, some of which are mentioned in the Qurʾān. Ancient Arabian inscriptions often invoke a god to bestow blessings on those who read the inscriptions aloud and to curse those who efface them, in a manner similar to that found in the colophons of manu­scripts in other traditions, especially Syriac. These inscriptions indicate that the Aramaic term qrʾ ‘to write’ had entered Arabic, presumably through Nabataean, in a very early period,134 and they manifest a written habit. The South Arabian inscriptions also attest to the disappearance of polytheism in the fourth century ce in favour of Rḥmnn, the equivalent of al-rahṃān or Allāh in the Qurʾān.135 They thus offer insights into traditional religions in Arabia before Islam in the various local languages. The newly found Palaeo-Arabic inscriptions in north-western Ḥijāz likewise bear

consensus that in Muhammad’s Arabia writing was not used to record cultural and religious texts, which remained exclusively oral’. 131 Khan, ‘Remarks on the Historical Background and Development of Early Arabic Documentary Formulae’ and ‘The Historical Development of Early Arabic Documentary Formulae’. 132 Sijpesteijn, Shaping a Muslim State, ch. 2. 133 Sijpesteijn, Shaping a Muslim State, p. 439 n. 15. The scribes used ‘different writing instruments in the process: the Arabic in these texts is invariably written with a much wider cut reed pen (Gr. calamus; Ar. qalam) than the Greek’. 134 Al-Jallad, ‘The Linguistic Landscape of Pre-Islamic Arabia’, pp. 123–24. 135 Robin, ‘Arabie antique’.

225

2 26

m u r i e l de bi é

monotheistic invocations.136 An inscription found in Jabal Dhabūb, written in an early Arabic dialect in late Sabaic script, could be a liturgical prayer to Allāh, al-Raḥmān, also called rabb al-samāwāt (‘Lord of the heavens’; cf. Q 19:65), that precedes the rise of the Medinan state.137 As C. Robin has shown, epi­graphy has already revealed the emergence of ‘henotheism’, or a kind of Jewish monotheism, in South Arabia, and it is increasingly showing that the Ḥijāz was no exception within the Peninsula. The inscriptions in Thamudic and Palaeo-Arabic also attest to a pre-Islamic artery of movement between Mecca and Ṭāʾif and further to the north. The absence of archaeo­logical surveys and the recent demolitions for the purpose of accommodating pilgrims have not entirely erased the past in Mecca and Medina, and chance findings continue to throw unexpected beams of light, albeit thin ones, on the religious and cultural situation in Muḥammad’s region.138 The project of Al-Jallad at a site west of Ṭāʾif between the town and Mecca will no doubt contribute additional material for thought on the continuity between the pre-Islamic and Islamic periods in terms of religious as well as economic trends linking Mecca with the outside world. Languages of Jews in Arabia

The consensus today about the languages of the Jews is that Hebrew was the sacred language (lashon ha-kodesh) of worship, the Bible, and religious discourse, whereas Aramaic was the vernacular in which oral translations of the Hebrew Scriptures were made for the faithful to understand.139 Greek was the official language of the Hellenistic kingdoms and of the Roman Empire, both a language of prestige and a colloquial idiom, especially in the diaspora. The Jewish Bible was translated into Greek (the Septuagint), literature was written in Greek by Jewish authors, and Jewish Greek inscriptions have been found from Judea to Crete and Rome. The presence of Judeans and Jews in Arabia140 and Palestine meant that written Aramaic (biblical and Jewish Aramaic) and to a lesser extent Hebrew were also languages that the pre-Islamic Arabs, who were increasingly monotheists and sometimes converted to Judaism,141 would have been exposed to or even familiar with. By the end of the fourth century, any reference to polytheism had disappeared from Ḥimyarite inscriptions.142 Around the year 380, during the reign 136 Al-Jallad and Sidky, ‘A Paleo-Arabic Inscription on a Route North of Ṭāʾif ’, p. 9 on the so-called ʿAbd Shams inscription. 137 Al-Jallad and Sidky, ‘A Paleo-Arabic Inscription on a Route North of Ṭāʾif ’, p. 9. 138 Shaddel, ‘The Development of the Meccan Sanctuary in the First Two Islamic Centuries’. I am thankful to M. Shaddel and A. Al-Jallad for generously sharing with me their publications, including forthcoming ones. 139 Poirier, ‘The Linguistic Situation in Jewish Palestine in Late Antiquity’, p. 83 n. 82. 140 On Judaism in pre-Islamic Arabia, see Robin, ‘Les religions pratiquées par les membres de la tribu de Kinda’; Robin ed., Le judaïsme de l’Arabie antique; Hoyland, ‘The Jews of the Hijaz’. 141 A Jewish king gained power in sixth-century Arabia. 142 Robin, ‘Les religions pratiquées par les membres de la tribu de Kinda’, p. 235.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

of Malkīkarib Yuhaʾmin (c. 375–c. 400), the Ḥimyarite kingdom abandoned the ancient polytheism that had already been declining during the reign of Malkīkarib’s father Ṯaʾrān Yuhanʿim (324–c. 375) and adopted a Judaeo-monotheism that was not affiliated with any structured Jewish community.143 Judaism was officially dominant in Ḥimyar from 380 until 530.144 After he was enthroned in 522, the king of Ḥimyar Yūsuf Asʾar Yathʾar (also called Yūsuf Dhū Nuwās) led an anti-Aksumite and anti-Byzantine Jewish rebellion until the Ethiopian Negus reacted and put a new Ḥimyarite Christian king on the throne between 525 and 530. The Ḥimyarite kingdom was Christian between 530 and 570. In northern Ḥijāz, the Jewish members of tribes known from inscriptions (prominently the Kinda) and from the Islamic tradition in the period before and during Muḥammad’s lifetime did not benefit from the same semi-official status. Very little evidence has been found, but around thirty graffiti from Arabia that may be called Judean and in some cases Jewish (because they use Hebrew or contain Hebrew names or Jewish formularies) are known today.145 In the Ḥijāz (in Yathrib, Khaybar, and Taymāʾ, the main centres of Jewish presence, as well as Madāʾin Ṣāliḥ), Jewish inscriptions in Nabataean language and script have been found; this is not surprising, since Jews used the same written — and probably oral — language as their neighbours.146 In Ḥimyar they also used the local language, Sabaic, with Aramaic-Hebrew and Arabic components.147 It is more surprising that three Hebrew inscriptions were found in Yemen, whereas the vast majority of Jewish inscriptions in Palestinian synagogues and in the diaspora are in Greek and, to a lesser extent, Aramaic.148 A list of mishmarot ha-kohanim, the priestly classes serving in the Temple, was found engraved on a pillar in the village of Bayt Ḥadir, fifteen kilometres south-east of Sanaa. It likely came from Tanʿim in the same valley, which is well known for its Jewish community, and can be dated to between the fifth and seventh centuries, possibly the sixth.149 In addition to naming the priestly families, it specifies the place in Galilee in which each settled after its return from Babylonia. Other similar inscriptions were found in Galilean synagogues and are attested in the Piyyutim.150 Besides transmitting the memory of the sacerdotal classes after the end of the Temple, these lists functioned as calendars, since the twenty-four classes were called twice a year in the Temple and are sometimes linked to the zodiacal signs. Since they changed every Sabbath and were regular,

143 See Robin ed., Le judaïsme de l’Arabie antique for further biblio­graphy. 144 On another Jewish king in Ḥimyar shortly after the middle of the fifth century, see Rubin, ‘From the Rabbanat at the Court of Sharaḥbiʾīl Yakkuf ’. 145 Robin, ‘Quel judaïsme en Arabie?’, pp. 74–75. 146 Hoyland, ‘The Jews of the Hijaz’. 147 Robin, ‘Quel judaïsme en Arabie?’, p. 188; Tobi, ‘The Jewish Community in Ḥāṣī, South Yemen’, p. 383. 148 Gorea, ‘Les classes sacerdotales (mišmarôt) de l’inscription juive de Bayt Ḥāḍir (Yémen)’, pp. 81–82. 149 Rubin, ‘From the Rabbanat at the Court of Sharaḥbiʾīl Yakkuf ’, p. 451; Robin, ‘Quel judaïsme en Arabie?’, pp. 110–12. 150 Gorea, ‘Les classes sacerdotales (mišmarôt) de l’inscription juive de Bayt Ḥāḍir (Yémen)’, pp. 306–07, 309.

227

228

m u r i e l de bi é

they may also have been used to count the weeks.151 Some knowledge of Hebrew writing is indicated by this pillar (engraved locally?). The language was probably taught locally at least to students of the Torah by the ‘priests’ from Tiberias who are mentioned at the Ḥimyarite court in the fifth and sixth centuries during the reign of the Jewish Ḥimyarite kings.152 It seems that Judaism in the Ḥijāz was close to the Ancient Temple of Jerusalem and sacerdotal as well as synagogal.153 It was also conservative from a linguistic point of view, since it used Hebrew. From the existence of synagogues, mentioned in the Syriac dossier on the massacre of Christians by the Jewish king of Najrān in the sixth century and unearthed by archaeo­logy (ten inscriptions of mikrāb, synagogues, are attested in the triangle of Ẓafār-Khaywān-Marib154), we can infer that Hebrew was used at least for reading the Torah. A seal inscribed in Hebrew and found in South Arabia depicts an arch for the Torah scrolls, showing the symbolic importance of the scriptures.155 We know from the Islamic tradition that a Jewish school called Bayt al-midrās existed in Yathrib (more precisely in al-Quff), though we do not know whether it was a school for boys or an academy for the study of the Torah.156 According to the tradition, Zayd b. Thābit, companion and secretary of Muḥammad who was part of the group that supposedly wrote down the ʿUthmānic version of the Qurʾān, had learned to read and write in this study house of the Jews. It is not clear what languages were used and taught there, but it is likely that they included at least some Hebrew for the purpose of reading the Torah and maybe some biblical Aramaic for the same reason. Zayd reportedly knew how to write both Hebrew and Arabic.157 Another question concerns the identity of the spoken language of the local Jews, yahūdiyya, mentioned by the Arab-Islamic tradition.158 This was a language used specifically by the Jews, and there is no mention of its having been written. Scholars have interpreted it as some form of Judaeo-Arabic or local Yiddish, but Robin has suggested that it is more likely to have been Aramaic, since Aramaicspeaking populations were present in the oases of the Ḥijāz. I would argue that the form of Aramaic that they spoke was probably not Nabataean, since this was not exclusive to the Jews and would not have been singled out as such. Rather, the presence of priests from Tiberias at the Ḥimyarite court in the fifth and sixth centuries159 points to a form of Judaeo-Aramaic, most probably Palestinian. We also 151 Gorea, ‘Les classes sacerdotales (mišmarôt) de l’inscription juive de Bayt Ḥāḍir (Yémen)’, p. 306. 152 Robin, ‘Quel judaïsme en Arabie?’, p. 117; Gorea, ‘Les classes sacerdotales (mišmarôt) de l’inscription juive de Bayt Ḥāḍir (Yémen)’; Rubin, ‘From the Rabbanat at the court of Sharaḥbiʾīl Yakkuf ’, pp. 448–50. 153 Costa, ‘Les juifs d’Arabie dans la littérature talmudique’, p. 477. The seal of the Jewish king Joseph is mentioned during the siege of Najrān but not described in the Syriac letter reporting the events (Robin, ‘Quel judaïsme en Arabie?’, p. 154). 154 Robin, ‘Quel judaïsme en Arabie?’, pp. 122–26. 155 Robin, ‘Quel judaïsme en Arabie?’, p. 272. 156 Robin, ‘Quel judaïsme en Arabie?’, pp. 79–80. 157 Robin, ‘Quel judaïsme en Arabie?’, p. 80. 158 Robin, ‘Quel judaïsme en Arabie?’, pp. 73–74. 159 Gorea, ‘Les classes sacerdotales (mišmarôt) de l’inscription juive de Bayt Ḥāḍir (Yémen)’; Rubin,

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

cannot exclude the possibility that it was derived from Official Aramaic in the case of older Jewish settlers or that it was Babylonian Aramaic. This language seems to have endured for a long time in secluded communities. It may explain why Galilean rabbis in the third century ce went to Ḥigrā (Madāʾin Ṣāliḥ) to enquire about terms in biblical Aramaic that were obscure to them:160 it was a place where ancient forms of Aramaic were still in use for the study of the Torah and/or as a spoken medium. The languages of the Jews are of interest not only for the Jewish communities, as if they were closed in on themselves, but also for understanding the Jews’ interactions with non-Jewish Arabians. The prominent official position of Jews in Ḥimyar, where Jewish kings — and courts — are attested meant that the languages of the Jews were also official languages. The Jewish kings might have used Sabaic as a public language, but the other languages that were in a hieroglossic relation with it (different forms of Aramaic as well as Hebrew, the languages of the Torah and the Talmud) held a special status. In the Ḥijāz, Jews had a prominent position as ‘headmen’ of West Arabian communities.161 Inscriptions have been found, albeit in small numbers, in Hebrew, Nabataean and Jewish Aramaic, Greek (East Sinai), and Arabic.162 Two types of written texts are mentioned in an account of King Yūsuf ’s dispatch of the priests of Tiberias to the Najrānites who refused his authority: the priests took with them both the Torah and written oaths sealed by the king.163 Written forms of a language — probably Sabaic — were thus used for official documents by the Jewish king, in addition to the inscriptions already mentioned. The passage also draws attention to the fact that the religious scriptures of the Jews were used not only for the liturgy in the synagogue but also as guarantees for the promises made by the king during the negotiations with the rebellious Najrānites. According to the Life of Muḥammad and al-Ṭabarī (d. 923), the origin of Judaism in Yemen is attributed to the outcome of an ordeal in which two rabbis with their books around their necks came out unscathed from a fire.164 The account of the arrival of ʿAntar in Khaybar, described in the romance bearing his name, reports that the inhabitants who came out to greet him lifted the sacred books above their heads while chanting the psalms.165 What these anecdotes showcase is the sacred character of the Jewish scriptures and the use of the books or scrolls as objects of public display outside the synagogue, where they could be seen by non-Jews. Scraps of pre-Islamic Arabic poetry by Arabian Jewish authors are preserved from

‘From the Rabbanat at the court of Sharaḥbiʾīl Yakkuf ’, pp. 448–50. Robin, ‘Quel judaïsme en Arabie?’, pp. 72–73; Costa, ‘Les juifs d’Arabie dans la littérature talmudique’. Hoyland, ‘The Jews of the Hijaz’, p. 96. Hoyland, ‘The Jews of the Hijaz’. Cited by Robin, ‘Quel judaïsme en Arabie?’, p. 154. Prémare, ‘Les textes musulmans dans leur environnement’, pp. 172 ff., cited in Robin, ‘Quel judaïsme en Arabie?’, p. 155. 165 Cited in Robin, ‘Quel judaïsme en Arabie?’, p. 155. 160 161 162 163 164

229

23 0

m u r i e l de bi é

the fifth to early seventh centuries.166 Jews were integrated into Arabian society, as the example of the Jewish poet Samuel b.  ʿĀdiyā shows. He was a resident of the oasis of Taymāʾ, where numerus Jews lived, and a whole corpus of poems is attributed to him. He purportedly helped Imruʾ al-Qays (d. c. 550 ce), a prince of the ruling clan of Kinda, on his journey to Constantinople in pursuit of justice for his murdered father.167 This is another example of the links between Arabs from north-west Arabia and the affairs of the Roman Empire and of the mobility of Arab elites between Arabia and Constantinople. The Isrāʾīliyyāt, Jewish (and Christian) narratives, including the Torah, to which the Qurʾān alludes advocate the transmission of either oral or at least partially written Jewish scriptures and commentaries, whatever the language of transmission. The multiform story of Zayd b. Thābit, who was proficient in writing because he had attended a Jewish school, exemplifies in a hagio­graphic way the interplay of languages between Jewish and non-Jewish Arabs. Languages of the Church(es)

Like Symeon Stylite in Syria and the many other ascetics mentioned in hagio­ graphical texts, Christian holy men in the desert were Christian models, wonder workers, and spiritual guides for the Arabs. The texts do not say much about the language(s) involved in these relationships, but besides Greek, the languages of the holy men included Syriac and Palestinian Aramaic. Arabs who knew these languages or monks in the entourages of the ascetics probably translated for them. Once Christianity was adopted as the official religion of the Roman Empire, conversion to it was no longer only a private matter but became a political one. Conversion missions were sponsored by the emperors. Around the year 354, Constantius II sent Theophilus the Indian on a mission to Arabia, where he converted the Ḥimyarites and built three churches for them. Although his trip had also political and trade-related motives, he introduced Greek Christianity to south-western Arabia in its Arian form. A comparable story from Syriac narrates how during Rabbula’s episcopacy (411–35) in Edessa, Paul of Qenṭos and John of Edessa were abducted by a band of Arabs while on a pilgrimage in the Sinai and sold to Ḥimyarites, whom they converted. This story found its way into the Taʾrīkh al-rusul wa-l-mulūk of al-Ṭabarī.168 Interestingly, it creates a link between the Edessan milieu and the conversion of a tribe of Ḥimyarites. That both Greek and Syriac stories of conversion in Ḥimyar were written seems to demonstrate the presence of clerics of both traditions. The monasteries of Bet Qatraye, on the shores of the Arab/Persian Gulf, were another important focus of Syriac culture in the sixth and seventh centuries. They were places in which mystic literature and practices as well as medicine 166 Hoyland, ‘The Jews of the Hijaz’, pp. 92–93. 167 Hoyland, ‘The Jews of the Hijaz’, p. 92. 168 The History of the Great Deeds of Bishop Paul of Qentos and Priest John of Edessa.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

were taught in monastic schools and manu­scripts were copied.169 The monks of Bet Qatraye were either Persians or Arabs; they shared a common education and participated in Church life in Syriac. The diocese of Bet Mazunaye in Oman had been a dependent of Ḥirta/al-Ḥīra and the Lakhmids since the 430s ce, which gave rise to an ecclesiastical continuum in north-east Arabia. Ḥirta/al-Ḥīra, as we will see in greater detail below, was a major centre of Christian presence. In the Ḥijāz, by contrast, no bishopric, church or permanent presence of Christians has been found so far, but our knowledge is impaired by the general lack of archaeo­logical surveys. Yet the newly found Palaeo-Arabic Christian inscriptions tend to show that Christians at least circulated in the area. The region being contiguous with Transjordan to the north and Ḥimyar to the south, it is implausible that no Christians, including priests and deacons (although not bishops), were part of the local communities. We should imagine that Christians belonging to a variety of churches and confessions lived in the region and belonged to the local tribes. In the absence of significant testimonies, however, it is difficult to fathom the extent of the Christian presence, but there is no reason the Ḥijāz would entirely lack the Christian liturgy, discussions, and books attested in other parts of Arabia, especially given the mobility of the Arab elites, troops, and tribes. The Kindite/Ḥujrid princes who dominated northern Arabia in the fifth century for the Ḥimyarite kings and had close links with the Roman Empire extended from near Mecca in the west to Qatar in the east and had its capital in the Najd.170 Although we do not have much information on the kingdom’s religion,171 Judaism was probably the dominant religion of the aristocratic lineages of the tribe of Kinda. As mentioned above, al-Ḥārith al-Malik b.  ʿAmr may have accepted enforcing Mazdakism in the Najd and Tihāma. However, prominent members of the kingdom were also Christians, as shown by the example of the Ḥujrid princess Hind, who married the Lakhmid king al-Mundhir III and founded a monastery in Ḥirta/ al-Ḥīra. The alliance between al-Ḥārith al-Malik b. ʿAmr and the Roman Empire does not necessarily imply conversion to Christianity, but it does suggest at least a positive attitude towards it. In the Ḥujrid/Kindite kingdom too, several official and religious languages — Greek, Middle Persian, Hebrew/Aramaic, and Syriac — were thus used for discussions and very likely also for written correspondence between leaders. Books of the scriptures and of the liturgy, at least, were known and owned by the priests/rabbis, and the words of the liturgy were familiar to the faithful. As they converted to Christianity and became priests and bishops, Arabs became part of Church hierarchy. As such they were exposed to and, at least in some cases, used the official languages of the Church in the Roman and Sasanian Empires and in the Arabian Peninsula. In the Roman Empire, Greek was the lingua franca of the empire and of the Church, and clerics or laypeople, whatever their 169 Archaeo­logical campaigns and research on Syriac authors and written production in the past decades have greatly improved our knowledge of this unique East Syriac milieu. See, for example, Brock, ‘Syriac Writers from Beth Qatraye’. 170 Robin, ‘Le royaume ḥujride, dit “royaume de Kinda”, entre Ḥimyar et Byzance’, p. 695. 171 Robin, ‘Les religions pratiquées par les membres de la tribu de Kinda’.

231

23 2

m u r i e l de bi é

first language, addressed holders of public office in Greek. Syriac was also one of the local languages of the Church and became the standard Christian language for Aramaic-speakers. It was the primary language of the Miaphysite Church since the fifth century (in both the Roman and the Sasanian Empire). Christian Palestinian was the language of the Chalcedonian Church in Palestine and the province of Arabia. Syriac was also the language of the Church of the East in the Sasanian Empire and beyond in India, Central Asia, and China. The bishops of the Arabs were called ‘bishops of the parembolai’ in Greek and ‘bishops of the ʿammē’ in Syriac (τὰ ἔθνη in Greek, umam in Arabic; ‘nations’). They either learned some Arabic or were themselves Arabs and spoke Arabic, although this is not mentioned by ancient authors. Conversely, missionaries probably learned some Arabic. Theodoret of Cyrrhus, in his apo­logetic work Cure of the Pagan Maladies, urges his audience to learn the language of the barbarians in order to Christianize those who are still polytheists.172 Arab bishops participated in the life of the Church and consequently exchanged numerous letters. In the Church of the East, they were expected to send letters to the newly elected catholicos, to receive from him the canonical constitutions of the Church, and to send and receive letters on dogmatic and practical issues. The bishops were not isolated figures but had staff (priests and deacons) who dealt with relations with the patriarchal seat, other bishops (especially suffragants), priests, and the faithful, as well as the local civil authorities and sometimes even the army. They were themselves educated, and they wrote letters and homilies, religious disputes, and hagio­graphical texts. All the Arab members of the Church were involved in at least its basic written culture: the reading of the scriptures and patristic commentaries and the composition — oral or partially written — of homilies and letters. We can assume that alongside deacons and priests, the sons and daughters of the Alliance/Covenant (women and men who were tonsured for Church service and lived in chastity) were ordained to serve churches in Arab communities as they were elsewhere, including in remote places. They received some form of education in order to carry out their duties in everyday liturgy, baptisms, and visits to the poor and sick. Church councils

Syriac was in a hieroglossic relation with Greek not only because the latter was the official language of the empire but also because it was the language of the scriptures (the Septuagint and the New Testament), apostolic missions, liturgy, Church order and hierarchy, dogmatic discussions and writings, ecclesiastical literature, and much more. Greek as the dominant language was thus a sociolinguistic feature of the Eastern Mediterranean Church(es).173

172 Theodoret, Graecarum affectionum curatio, § 5, 60. 173 On multilingualism and diglossia involving Syriac, see Taylor, ‘Bilingualism and Diglossia in Late Antique Syria and Mesopotamia’. On Greek as the language of the Church, see Millar, A Greek Roman Empire.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

A complex hierarchy of languages applied in Western Church councils. As noted earlier, Latin was essential for anyone performing a public function and was still used in letters exchanged between Eastern and Western bishops and for dogmatic texts.174 However, translations (hermeneia) were needed for these documents and for the interventions of papal legates in the councils, since most Eastern bishops knew only Greek (if even that). Although some letters and documents cited in the acts of the councils were originally written in Latin and, in one case, Syriac, they are cited in the acts in Greek translation.175 Similarly, the signatures of Syriac-speaking council participants were recorded in Greek translation, and the original language was indicated in Greek.176 Priests and deacons served as translators for the bishops who did not understand Greek or could not sign in Greek. Syriac was the only other language that had a special place in the Church, although other languages — Armenian, Georgian, Palestinian Aramaic, Arabic, Coptic, and Ethiopic — were also spoken by Christians. Only one reference to a bishop speaking Egyptian (Coptic) appears in the acts of the councils, whereas there are several mentions of Syriac. The translation into Greek thus obscured the original languages allowed in these official gatherings. The acts of the councils kept the memory of Arabs’ participation in the official occasions of the Church. Since Arab clerics adopted Christian names at ordination, their original Arabic names were not preserved and were erased from public memory. However, the lists of participants and signatories mention ‘bishops of the Saracens’ or ‘bishops of the parembolai’.177 We can assume that at least some of these were Arabs or spoke Arabic and other languages. None of them is mentioned as having signed in a language other than Greek, which implies that they had a good command of Greek in order to follow the debates. The most striking case is perhaps that of Aspebetos, the Arab phylarch and ally of the Sasanians mentioned earlier. During the Roman-Sasanian conflict of 420–22 ce he was entrusted with the task of preventing the flight of Christian civilians to the Romans, but he reportedly took pity on the Christians and allowed them to escape.178 Denounced to the Persian king Yazdgerd, he sought refuge in the Roman Empire.179 The magister militum per Orientem appointed him phylarchus of the Saracens in the province of Arabia. Such an appointment must have entailed discussions and negotiations in Greek, although Aspebetos came from the Sasanian Empire. The magister was supported by an administration that included secretaries and probably interpreters for the frontier zones. However, since Aspebetos subsequently converted to Christianity and was ordained bishop of the parembolai under the name of Peter, he may have known Greek already when he crossed the border or learned it quickly enough not 174 Millar, A Greek Roman Empire, pp. 18–19, 93–95. 175 These collections were then translated into Armenian, Coptic, and Syriac and, in the Islamic period, into Arabic. 176 Millar, A Greek Roman Empire, p. 104. 177 Often translated as ‘encampment’, it is better understood as troops. 178 Cyril of Scythopolis, Life of Euthymius. 179 Millar, A Greek Roman Empire, pp. 73, 106.

233

23 4

m u r i e l de bi é

only to become a bishop but to attend the first council of Ephesus in 421, where his Christian name features in the list of participants (his Arab identity would be impossible to guess if we did not know his story from other sources). There is no mention of his not having signed in Greek or having resorted to an interpreter, although such notes appear for other bishops. The acts of the Church councils, which include letters exchanged between members of the Church hierarchy as well as petitions to the emperor and official answers with decrees, provided an occasion for Arab clerics to participate in the written culture of both state and church in Greek and sometimes Latin and Syriac. The famous letter sent in 569–70 by a large number of Miaphysite archimandrites and other monks from Arabia to a group of bishops in Constantinople was written in Greek and very quickly translated into Syriac. Although the signatories seem to have been more familiar with Syriac than with Greek (117 of the 137 original signatures had not been written in Greek), knowledge of Greek was expected, as it was the official language of the Church and of the bishops in the capital.180 In the Church of the East, the acts of the councils were uniformly in Syriac. However, a letter from Western bishops carried to Yazdgerd by Maruta of Maypherkat, acting as ambassador, was written in Greek; clerics translated it to Middle Persian and read it to the king of kings, who then decided to convene the first council of the Church in 410.181 The bishops of Ḥirta/al-Ḥīra ‘d-Ṭayyayē’ (of the ‘Arabs’) attended almost all Church councils after that.182 Their see, though not a metropolitan one, was important in the Church hierarchy and served as the burial place of a number of sixth-century catholicoi.183 Like their peers, the bishops of Ḥirta/al-Ḥīra were convoked by the king and were allowed to use the imperial post, both procedures taking place in Middle Persian. Nothing in the acts indicates that they did not sign and seal confessions of faith and canons in Syriac. As in the Western churches, deacons and notaries could sign for their bishops,184 which shows that clerical staff from the sees accompanied the bishops. A canon of the councils stated that each metropolitan bishop had to write a copy of the canons in order to rule his diocese and suffragant bishops.185 We thus know that canon law was copied and disseminated, which implies the circulation of canonical books and letters exchanged by the catholicos and the bishops through their respective chanceries. We do not know if the bishops of Ḥirta/al-Ḥīra, who were not metropolitans, also had to make copies of the acts of the councils, but they certainly received canonical letters from the catholicos, who had to keep the whole hierarchy informed. Like other bishops for the Arabs in other dioceses — in

180 Millar, ‘A Syriac Codex from near Palmyra and the “Ghassanid” Abokarib’, p. 20; Hoyland, ‘Late Roman Arabia’, p. 129. 181 Synodicon Orientale, pp. 28–19 (text), 256 (translation). 182 On al-Ḥīra, see Toral-Niehoff, Al-Hira. See also Synodicon Orientale, p. 673, for a list of the bishops of Ḥirta present in the councils. 183 On the burials of catholicoi, see Fiey, ‘Résidences et sépultures des patriarches syriaques-orientaux’. 184 Synodicon Orientale, e.g. pp. 59 (text), 306 (translation). 185 Synodicon Orientale, canon X, pp. 25–26 (text), 266 (translation).

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

the Persian Gulf especially — they thus participated in the written culture of the Church of the East in Syriac and in Middle Persian. In 424, a synod of the Church was convened in Markabta of the Arabs, a city not far from Ḥirta/al-Ḥīra, which indicates the importance of the region in the nascent Church of the East as well as the importance of Arabs within the Church. Scriptures and liturgy

Since Christians spoke Arabic and used the Arabic script, we cannot but wonder why Arab Christians did not use this script in the pre-Islamic period to translate the Bible (Old and New Testaments) or to write Christian literature as they did with Syriac, Armenian, Coptic, and Slavonic. No traces of an Arab Christian literature before Islam have been found so far, although this absence may also be due to the disappearance of books. A Syriac text that is often cited, the Dialogue of Patriarch John Sedra with the Emir, mentions a religious gathering purportedly held in the presence of ʿAmr b. Saʿd; John Sedra lived in the seventh century, and the text was likely composed in the eighth. According to the text, the exchange was witnessed by ‘not only the notables of the Hagarenes, but also the chiefs and leaders of the cities and of the believing and Christ-loving peoples the Tanukaye, the Ṭuʿaye, and the ʿAqulaye’ (the pre-Islamic Christian tribes mentioned in Syriac sources).186 It was on this occasion that the emir commissioned a translation of the Gospel into Arabic. Whether or not this account is historical, it is interesting in that it situates the first translation of the Gospel undertaken by members of Christian Arab tribes in the Islamic period. We can understand better the absence of a pre-Islamic Christian Arabic literature if we consider the status of Arabic in the late antique Church(es) and, more generally, the practical use of languages in ecclesiastical contexts. The hieroglossic relation between a sacred language and a vernacular is a phenomenon that was shared by Jews and Christians in Late Antiquity and that explains differences in the respective status and practice of languages. Arabic was neither a high-status nor a sacred language and seems to have been almost entirely confined to non-literary functions, whereas Greek and Syriac dominated literature and literate political expression. Moreover, the churches did not promote the translation of liturgy except in oral form. The liturgy was a site of interaction between official religious languages and vernacular tongues: conducted in the languages of the Church of the eastern Mediterranean, Greek and Syriac, it was translated or explained in a targumic fashion in the vernacular Aramaics and in Arabic. We have two testimonies of this practice in the case of Greek and Aramaic from the fourth century. Eusebius of Caesarea mentions that the martyr Procopius (d. 303) served as an Aramaic translator in his church in Scythopolis.187 And in the 380s, the pilgrim Egeria describes in a well-known 186 Michael the Syrian, Chronicle, ii, p. 422 (Syriac); iv, p. 432 (French). 187 See Hoyland, ‘Language and Identity’, p. 187; Garitte, ‘La version géorgienne de la passion de S. Procope par Eusèbe’, p. 255.

235

2 36

m u r i e l de bi é

passage of her Latin travelogue how Greek and Aramaic interacted in the context of the liturgy in Jerusalem: ‘In this province there are some people who know both Greek and Aramaic (siriste); but others know only one or the other language. The bishop may know Aramaic, but he never uses it. He always speaks in Greek and has a presbyter beside him who translates the Greek into Aramaic so that everyone can understand what he means’.188 This passage makes it clear that bishops, who occupied the top echelon in the hierarchy of the Church, would use only Greek, the official language of the Church in the Eastern Roman Empire, but that for pragmatic reasons priests (and maybe deacons as well) provided oral translations or commentaries for the faithful so that they could understand what was going on. It is difficult to fathom what, exactly, siriste designated here: the higher form of Syriac or the lower CPA? An intra-Aramaic diglossia and hieroglossic relation is evident, with Syriac as the dominant liturgical and cultural language in regions where CPA was the colloquial Aramaic. We can surmise that Greek was used by the bishops even though it was not understood by the faithful not only because it was the official language of the Church but because it enabled them to control the orthodoxy of the liturgy. Christian Theo­logy and the Qurʾān

The sophistication of the arguments exchanged in the disputation sessions as well as in canonical correspondence indicates that Syriac was not just a language used for the liturgy in Arabia but also a vector of religious texts that were discussed at the court of the Arab kings and among the Arab clergy from Syria to Arabia. This point has implications for analysing possible links between Christian texts and the Qurʾān. The arguments put forward during the Tritheistic dispute, for example, may have had echoes in the Qurʾānic passages about the elusive mushrikūn, ‘associators’. Although these associators are generally interpreted as polytheists, we can wonder whether the label might also designate Christians who associated other gods with God, as implied by the Tritheist position, which was accused of creating a ‘quaternity’ (because of the two natures of Jesus Christ plus the nature of the Father and of the Holy Spirit).189 The Church considered heretical the position of those who saw the three persons of the Trinity as separate substances or natures; the official dogma was that they were consubstantial, of the same nature. The proponents of this heresy were called polytheists in the letters mentioned above that circulated in the Miaphysite Arab milieu.190 But Syriac should not be the only prism of interpretation: at least Greek and maybe also CPA should be equally considered. When it comes to the Gospels, for instance, is the Greek or the Aramaic version closer to Qurʾānic passages? If it is the Aramaic, what form of Aramaic, Syriac or CPA?

188 Egeria’s Travels, p. 146. 189 Debié, ‘Les controverses miaphysites en Arabie et le Coran’. 190 Documenta ad Origines Monophysitarum, pp. 167 (text), 116 (translation).

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

And if it is the Syriac, which version can be identified — the Diatessaron, the Old Syriac, the Peshiṭta, or the Harqlean?191 The echoes of Christian apocryphal texts in the Qurʾān might come from a dual Greek and Syriac tradition. The history of the young sleepers in the cave in Sura 18 alludes to several versions of the story, featuring variously six, seven, or eight youngsters. Might these reflect the several known Syriac versions, accessed orally or in written form, even though the Qurʾānic story’s resemblance with the mimro of Jacob of Serugh on the topic is the most conspicuous? Did a Greek version also circulate?192 The story of Moses and his servant in the same sura is close to a Greek or Hebrew version of the Alexander Romance, while the passage on Dhū’l-Qarnayn has striking similarities with the Syriac Exploit of Alexander.193 In terms of literary genres, the aporetic style of disputation found in Syriac bears a close resemblance to the dialectical technique characteristic of Islamic kalām (rational philosophy). It likely had a Greek origin and eventually passed from Greek into Syriac before finding a home in Arabic kalām texts.194 This is not the place for a thorough analysis of the links between Syriac Christianity and the Qurʾān, but in light of the literate culture that we can discern from indirect textual evidence, we should consider the Arabs of Late Antiquity part of the Christian culture of their time — a culture that was transmitted orally as well as in written form. Whatever the date and circumstances of the canonization of the Qurʾānic text, which probably received its final edition under ʿAbd al-Malik, we cannot reject the possibility that the Islamic tradition, which speaks of suras written on various surfaces that were subsequently ‘edited’, reflects the memory of actual practices of writing already in use in the time of Muḥammad.

Languages of Church and State Diplomacy Civil and ecclesiastical diplomacy were increasingly entangled in Late Antiquity, not only in the Roman Empire but also in the Sasanian one, since bishops were involved in negotiations with the emperors and served as intermediaries between empires and with the Arab kingdoms. In the 460s Terebon, the son of Aspebetos, was imprisoned by the dux of Arabia in Bostra because of his participation in a plot. He was liberated thanks to the bishop of Bostra and returned to the desert of Judaea to his protector, Euthymius. We can see in his history both the role of bishops in civil affairs and the importance of holy men in the lives of Christian phylarchs. During Leo’s reign (457–74), Amorkesos/Imruʾ al-Qays, who was allied with the Persians, wished to defect to the Romans. He sent a bishop named Peter as

191 e.g. El-Badawi, in ‘The Impact of Aramaic (Especially Syriac) on the Qurʾān’, speaks about ‘Aramaic’ without specifying the version he means. 192 Griffith, ‘Christian Lore and the Arabic Qurʾan’. 193 Debié, Alexandre le Grand en syriaque. 194 Tannous, ‘Between Christo­logy and Kalām?’, p. 707. Cf. Cook, ‘The Origins of “Kalām”’.

237

23 8

m u r i e l de bi é

an emissary to Emperor Leo, thus showing his knowledge of the diplomatic ways of the Roman Empire.195 The discussions took place in Greek with the help of translators, the bishop and his assistants being first and foremost among them. Arab phylarchs’ conversion to Christianity had political repercussions and was treated as a state matter at the highest level. Religious policy played an important role in the relationship between the Arabs and the Roman Empire. Queen Mawiyya, who ruled the Tanukhids in southern Syria in the latter half of the fourth century, had asked for an Orthodox bishop for her people, but the Roman emperor Valens, an Arian himself, wanted an Arian to be ordained for this bishopric. A revolt ensued, allegedly rooted in this religious disagreement. Familiar with Roman tactics, Mawiyya’s troops defeated the Romans in Palestine and Phoenicia as far as the Egyptian border. As a condition for ending the war, she insisted that the Arab monk Moses be ordained as bishop for her people. In addition, Mawiyya married her daughter Khasidat to Victor, a Roman commander who adhered to the Nicene Creed, thus reinforcing the alliance with the Roman state and her link to the Nicene party within the Church (although — or because — it was not the imperial party at this time). The negotiations took place in Greek, the Arab ecclesiastical hierarchy became part of the Greek Church, and the matrimonial alliance strengthened connections between Arab allies and the Roman elite. Although these relations subsequently deteriorated when Theodosius I favoured the Goths at the expense of the Tanukhs, the Salihs soon became new allies to the Romans. What this famous episode exemplifies is the double significance — both ecclesiastical and military/political — of the Arabs and their language in the Roman Empire. Mawiyya was celebrated in Arabic poems that were still recited in the fifth century and are the earliest reference to Arabic poetry.196 By contrast, the inscription for a (different) Mawiyya who erected a martyrion to honour Saint Thomas outside the walls of Anasartha was in Greek, the language of the empire and of the Church.197 This contrast highlights that Arabic was the ‘internal’ language of the Arabs, whereas Greek was the language of external affairs. It is worth noting that during the Arian controversy as well as the Miaphysite one, the Arabs often supported the non-imperial position — the Nicene party when the emperor was Arian and the Miaphysite one whatever the emperor’s leanings were. The Christo­logical controversies that rocked the Church in the fifth and sixth centuries had political outcomes, and the attempts to recruit the Arabs to one or the other side of ‘orthodoxy’, either that defined by the official Church and backed by the emperors or that promoted by their opponents, contributed directly to the increase in written documents. Patristic and dogmatic proofs were produced during disputes and letters were exchanged to convince adversaries, secure allies, discuss the minutiae of the theo­logical positions at issue, and convene synods and meetings. The Arabs who were sought after by the contending parties were thus involved in a written network of exchanges. 195 Shahid, Byzantium and the Arabs in the Fifth Century, p. 68. 196 Sozomen, Ecclesiastical History, p. 298. 197 Shahid, Byzantium and the Arabs in the Fourth Century, p. 222.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

This practice of writing is attested in the Life of Simeon of Bet Arsham, the Miaphysite ‘Persian debater’ who disseminated professions of faith in multilingual milieus in Persia as well as among the Arabs. He supposedly wrote the creed of each nation on linen, which, according to his bio­grapher, John of Ephesus, he considered less fragile than parchment when travelling. He had the cloths treated so that they could be written on: In order that the certainty of the writing might remain without suspicion of alteration,198 he made great linen cloths and medicated them, so that they might take writing, which also will, I think, be preserved by the believers in the land of the Persians forever. And on them he would accordingly write the belief of every people in their own language from their archbishops, and above the belief he fixed the seal of the king of that people and of their bishops and of their chief men in lead upon these cloths and thus confirmed it. This is how he acted among all people and all tongues among the believers.199 This passage throws light on the multilingual experience of Christians in Persia and Arabia as well as on the practical aspects of writing. Linen is seldom mentioned as a writing surface, and this example should be compared with the mention of written-on linen hanging in the Kaʿba as a possible attestation of an actual practice in Late Antiquity, although no artefacts have come down to us. We have two mentions of clerics sent on missions by Roman emperors to the Lakhmid king al-Mundhir III at Ḥirta/al-Ḥīra. The figure of Abramios, a priest and a diplomat, who was sent by Justin I (r. 518–27), offers the first example of the role priests played in diplomatic contacts between the Romans and the Arabs. He was considered a specialist on Arab affairs. Son of Euphrasius and father of Nonnosos, he belonged to a family of diplomats. We do not know whether the family’s members knew Arabic, but it would not be surprising if they did.200 The second example dates to the reign of Justinian and the Roman-Persian war in 530, when al-Mundhir asked for a deacon he knew, Sergius, to be sent to him so that he might convey peace terms to Justinian. Sergius travelled back and forth with al-Mundhir’s letters. Both missions involved clerics who acted as ambassadors and featured exchanges of documents written in Greek by both parties.201 In the late Sasanian Empire, the story of the conversion of al-Nuʿmān III by Simeon Jabara, the bishop of Ḥirta/al-Ḥīra, is placed in the context of rivalry between the Miaphysites and the Church of the East.202 The catholicos Sabrishoʿ (596–604), newly elected with the permission of Kosrow II (r. 590–628), came to Ḥirta/al-Ḥīra in order to reconvert the Arab king from the Miaphysitic faith he had 198 This shows the high status of writing for defending orthodox beliefs in Christianity, as opposed to the preference for orality in the Zoroastrian and Muslim traditions. 199 John of Ephesus, Lives of the Eastern Saints, p. 156. 200 Martindale, ‘Abramius II’, p. 3; Shahid, ‘Byzantino-Arabica’, p. 116 n. 4; Robin, ‘Le royaume hujride, dit “royaume de Kinda”, entre Himyar et Byzance’, pp. 671–73. 201 Malalas, Chrono­graphia, pp. 466–67 (translation). 202 On this episode, see Wood, The Chronicle of Seert, pp. 192–94.

239

24 0

m u r i e l de bi é

adopted when he converted to Christianity. According to the sources (The Life of Sabrishoʿ and the Chronicle of Sıırt, both East Syriac texts giving a hagio­graphical account of Sabrishoʿ’s role), Kosrow granted him permission to go to Ḥirta/al-Ḥīra and hence to convert al-Nuʿmān. The latter had given shelter to the previous catholicos, Ishoʿyahb, who had angered Kosrow and had consequently been sent into exile by the emperor; this history suggests a certain amount of independence on the part of the Arab king vis-à-vis the Church hierarchy and the shah. Al-Nuʿmān nevertheless converted to the ‘right’ faith. This intricate account demonstrates that the conversion of an Arab king was acceptable only if he joined the Church of the East, under the control of the Sasanian emperor, and not the Syriac Orthodox Church, which was associated with the Roman Empire enemy. His conversion was an affair that concerned the highest levels of the Sasanian Empire (and possibly of the Roman Empire, too), of the Church of the East, and of the Syriac Miaphysite Church. The discussions pertaining to it likely took place in Middle Persian and possibly Arabic as far as al-Nuʿmān was concerned. Miaphysite Diplomacy

The roles played by three major contemporary figures of the Syriac Miaphysite Church, Jacob of Serugh (452–521), Philoxenus of Mabbug (c. 440–523), and Simeon of Bet Arsham (d. c. 540), highlight their links to Arabia. A small dossier of three letters throws light on the relationships that Severus, the bishop of Antioch, and Philoxenus, the bishop of Mabbug in province of Euphratesia), had with the Arabs of Ḥirta/al-Ḥīra at the beginning of the sixth century. Philoxenus exchanged letters with Abū Yaʿfur b.  ʿAlqama, the stratelates (magister militiae in Latin) of Ḥirta/al-Ḥīra, in 503–04, during the Roman-Persian wars, when Abū Yaʿfur succeeded al-Nuʿmān II (r. 499–503) after the latter was killed while campaigning in northern Mesopotamia. Abū Yaʿfur was Christian and had asked questions about the validity of baptism by heretics and the natures of Christ at a time of fierce controversies between Chalcedonians and Miaphysites and in a context in which the Dyophysitic Church of the East, supported by the Sasanian emperors, was powerful. The dogmatic letter of Philoxenus (we do not have the letter written by the phylarch) attests to the high level of the theo­logical discussions between these figures as well as to the political importance of religious affiliations in times of war. It also points to the triangular ecclesiastical and dogmatic dimensions of the relationships between proponents and adversaries of the Council of Chalcedon in the Roman and Sasanian Empires. A little later, when he was patriarch of Antioch (512–18), Severus also sent two missionary bishops to the phylarch al-Mundhir III in order to gain him to the Miaphysite faith.203 Arguments about the immortality of angels and the 203 See Alpi, ‘La correspondance de Sévère d’Antioche et de Philoxène de Mabbōg avec les chrétiens d’al-Ḥīra’ for the reference to Theodore Lector’s Ecclesiastical History. I am grateful to the author for sharing a pre-publication version of his article.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

incorruptibility of God were apparently exchanged (and presented according to a Chalcedonian bias by Theodoret). Angelo­logy, Christo­logical questions, and divine corruptibility — also important themes in the Qurʾān — were thus discussed at the court even though the phylarch was not himself Christian. We can assume that he was curious about the diverging arguments of the various churches in his dominion, where each denomination built churches and monasteries. Understanding the affiliations of the individuals and groups within his realm was also crucial in the context of wars. After he was deposed in 518, Severus also addressed a letter in Greek (of which only a fragment survives) ‘to the priests and archimandrites Jonathan, Samuel, and John, who became stylites, and to the rest of the church of the city of Anbar204 and of the church of Hirtha d-Nuʿmān’. The letter fragment is a tantalizing testimony to the enduring relationships between the highest echelons of the Syrian ecclesiastical hierarchy and the Arabs of Ḥirta/al-Ḥīra in spite of the Roman-Persian wars.205 The content of the letter is unfortunately lost, but the address suffices to reveal that the Miaphysite patriarch, like Philoxenus, reached beyond the borders of the Roman Empire and sent dogmatic letters on Christo­logical matters to the Arabs in Ḥirta/al-Ḥīra. The letter also shows the appeal of the traditional though extreme asceticism of the stylites among the Arabs. Another dossier shows the influence of the Syrian Miaphysite bishops in South Arabia. Philoxenus ordained the first two Miaphysite bishops of Najrān. Jacob of Serugh mentions positive conversions to Christianity (i.e. to the Miaphysite confession) in Najrān. The Miaphysite bishop Simeon of Bet Arsham was at the court of al-Mundhir at Ḥirta/al-Ḥīra when he heard about the massacre of Christians by the Jewish king Dhū Nuwās in 523 and reported the dramatic events in a famous letter to the archimandrite of the monastery of Gabbula in Syria. Their letters attest to politico-religious conflicts between Christians and Jews in Arabia and to intra-Christian polemics over the Trinity and Christo­logical issues. They also reveal the links between the bishops and monks of Syria and the Arab court of Ḥirta/al-Ḥīra, as well as those between the Syrian clerics and the kingdom of South Arabia. When in 526 a war began between the Roman and Sasanian Empires, al-Mundhir III b. al-Nuʿmān attacked and ravaged Syria. Two high-ranking Roman commanders, Timostratus and John, were captured. Justinian sent an embassy to al-Mundhir to sue for peace. It consisted of the priest Abramios/Abraham and Simeon of Bet Arsham, who was the representative of the Miaphysites in the Sasanian Empire, both mentioned above. They were joined by the bishop Sergius of Resafa, whom Justinian later sent with gifts to al-Mundhir. Here, too, bishops were involved in the negotiations and acted as ambassadors between the Roman Empire and the Arab king in matters such as the ransoming of prisoners of war in which the churches were traditionally involved, raising money for the ransoms. We can assume that the

204 Anbar is Pērōz Šāpūr, a Sasanian city on the Euphrates. 205 Alpi, ‘La correspondance de Sévère d’Antioche et de Philoxène de Mabbōg avec les chrétiens d’al-Ḥīra’.

241

242

m u r i e l de bi é

discussions took place in Greek with the help of interpreters. The episode shows the direct Roman-Lakhmid discussions and the role played by the Miaphysite Simeon of Bet Arsham at the court of al-Mundhir. This ‘conference of Ramla’ was preceded by (and in some historical texts confused with) a religious encounter. In 536, when Justinian enacted laws against the Miaphysites, Ephrem, the patriarch of Antioch, tried to win the Jafnid al-Ḥārith b. Jabala/Arethas to the Chalcedonian cause, just as Severus earlier had tried to attract the Naṣrid phylarch to the Miaphysite side. Arab allies were important for the empire’s security, and the church leaders sought to ensure their religious loyalty. In the 540s al-Ḥārith obtained the ordination of two Miaphysite bishops, Jacob Baradeus (the founder of the ‘Jacobite’ Syriac Orthodox Church) for the patriarchate of Antioch and Theodore for the patriarchate of Jerusalem. These appointments were no doubt discussed in Greek, but a letter in Syriac shows al-Ḥārith’s efforts to secure the election of Paul Ukkama, a Miaphysite, to the seat of Antioch.206 It is interesting to observe that the patriarch of Antioch reached out to the Arab king and the latter, in turn, decided to get involved in the affairs of the imperial church. These negotiations and exchanges involved discussions and letters that show how the Arabs participated in the politics of their time in the language of the empire. Greek was the language used within the civil and ecclesiastical hierarchies of the empire, but Syriac was the dominant language of the Miaphysites. The archimandrites of monasteries in the province of Arabia sent a letter in Syriac to Jacob Baradaeus, disavowing Tritheism.207 It was probably in reaction to this letter that Jacob, in turn, wrote a letter — in Greek? Syriac? — that was brought to Constantinople by al-Ḥārith in 563, when he went to complain about his Arab rivals. He also wrote to Jacob himself, and in his letter, he recounts his encounter with the pope of Alexandria, Theodosius, who resided in Constantinople208 and asks Jacob to find trusted people to deliver letters. Miaphysite controversies thus involved written correspondence and found support from the Jafnid phylarch, who presented them to the Church hierarchy in Constantinople. They also provide a glimpse at the interactions between the phylarch and the Miaphysite Church hierarchy in the province and in Constantinople. Al-Ḥārith’s trip to the capital had both political and religious dimensions and entailed participation in the literate idiom of his time, even though he came from a largely illiterate society. When Miaphysite texts were banned by Justinian,209 they were translated into Syriac, and this use of a non-official language enabled the preservation of the condemned religious texts. Christian Arabs in the Roman Empire were thus increasingly attracted to ecclesiastical milieus that employed Syriac for the liturgy as well as for dogmatic texts and correspondence.

206 Shahid, ‘Byzantino-Arabica’, p. 287. 207 On this letter and the geo­graphic limits of the province, see Hoyland, ‘Late Roman Provincia Arabia, Monophysite Monks and Arab Tribes’, pp. 121–25. 208 This dossier was edited in Documenta ad Origines Monophysitarum. 209 It was forbidden to read, copy, or possess these writings.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

The diplomatic role of the Arab kings was not confined to the political and military realms but also touched on theo­logical matters, since they occasionally hosted sessions of theo­logical disputes between the Miaphysite and East Syrian bishops. For example, before 521 or 523, the death date of the East Syrian catholicos Silas, a dispute took place between him and Miaphysites exiled by Emperor Justin who had taken refuge at the court of al-Mundhir.210 Similarly, the phylarch of Jafna sponsored two meetings between Miaphysite opponents in the monastery of Bet Mar Sergius and in the church of Saint Sergius. Both debates took place in Gabitha, his encampment, on neutral ground. The patriarch of Antioch Peter of Callinicum (in office 581–91) and the patriarch of Alexandria Damianus (in office 578–605) debated the properties, roles, and relations of dimensions of the Trinity (Tritheism).211 The letters exchanged on the issue were written in Greek and were rapidly translated into Syriac. The Arab phylarch played the role of a civic authority in organizing a meeting between the highest authorities of the Syrian and Egyptian churches. In Peter of Callinicum’s report on the meetings, he is described as having been impatient to go back to his troops when the patriarchs could not come to an agreement and as complaining about having to correct the two churchmen even though he was a layman.212 The religious controversies within the Miaphysite Church as well as between the Chalcedonian, Syriac Orthodox, and East Syriac Churches were thus present in the correspondence of Miaphysite bishops and the theo­logical sessions organized under the auspices of the Arab kings. Given the technical nature of these discussions, numerous participants attended from both sides of the debate, and manu­scripts, probably in Greek and Syriac, were presented to support the textual testimonies of each party. Arab priests and bishops certainly attended the meetings and made oral as well as written translations of the arguments presented. The Dossier of Najrān

Several texts in Greek, Syriac, Ethiopic, and Arabic describe the events in Ḥimyar in 523 when the Jewish king Yūsuf/Dhū Nuwās rebelled against the Axumite and Byzantine influence over Ḥimyar and massacred the Christians of Najrān.213 This multilingual dossier is indicative of the various parties involved and of the great interest the events held for various audiences in Byzantium, Aksum, Najrān, and Ḥirta/al-Ḥīra, in Miaphysite monasteries, among bishops in Syria, and among East Syriac Christians. Much has been written on this episode, which is central

210 De Blois, ‘The Date of the Martyrs of Najran’, p. 17; Detoraki, Le martyre de Saint Aréthas, pp. 51–52 and 254 n. 55; Beaucamp, ‘Le rôle de Byzance en mer Rouge sous le règne de Justin’, p. 209. 211 Debié, ‘Les controverses miaphysites en Arabie et le Coran’, p. 140. 212 Michael the Syrian, Chronicle, ii, xxii. 213 Beaucamp, Briquel-Chatonnet, and Robin, ‘La persécution des chrétiens de Najrân et la chrono­logie himyarite’; Ryckmans, ‘Les rapports de dépendance entre les récits hagio­graphiques relatifs à la persécution des Ḥimyarites’.

243

24 4

m u r i e l de bi é

to understanding the multireligious geopolitical context of Arabia before Islam. We will concentrate here on what the dossier tells us about literacy and language interactions between the Arabs, the Ethiopians, and the South Arabians. In one of the Syriac letters written in order to make public the massacre in Najrān, the anonymous author says that he wrote from written accounts in the ‘Najrānite’ language. That could have been a southern form of Arabic or possibly Sabaic. This instance alone shows the entanglement of religions and languages as well as the practice of sending letters that was indispensable to every act of civic or ecclesiastic diplomacy or simply to exchanging news across long distances. Clerics of all origins and churches are mentioned in the Syriac Book of the Ḥimyarites: a Najrānite priest, two priests from Ḥirta/al-Ḥīra, a Roman priest and a Roman deacon, a Persian priest, and an Ethiopian priest.214 They illustrate the rainbow of Christian churches, confessions, and theo­logies present in Ḥimyar at the time of the persecution, which explains why so many texts were written. They also demonstrate the political and economic rivalries between the great empires and local kingdoms. The less-known account of the martyrdom of an earlier local martyr, the Martyrdom of Azqir, composed probably in the eighth century, is also interesting, since it might rely on an ancient local record written in Arabic.215 The Literate Culture of Ḥirta/al-Ḥīra

Ḥirta/al-Ḥīra was a place of learning in Late Antiquity. According to the Chronicle

of Sıırt, a Christian school was founded there by Cyrus of Edessa, one of the disciples of the famous catholicos and scholar Mar Aba (in office 540–52), over the grave of the latter.216 As in other schools of the Church of the East, here, too, Syriac was taught to future clerics. This school is sometimes called a monastery in the sources, maybe because the students resided there, like they did in the school of Nisibis, which was organized as a monastery. We can assume from the scholarly profile of Mar Aba and those of his disciples Cyrus and Sergius that the curriculum of the school was not limited to elementary reading and music for the liturgy but followed the traditional Greek paideia adopted in Syriac schools that included mathematics and philosophy (logic and metaphysics), medicine as in Nisibis, and of course theo­logy and commentaries on the scriptures and liturgy.217 Mar Aba learned Greek in the Roman Empire, and he probably taught it to his disciples. We cannot therefore exclude the possibility that some Greek was taught also in the school of Ḥirta/al-Ḥīra.

214 The Book of the Ḥimyarites, p. cix, 14a. 215 Robin, ‘Nagrān vers l’époque du massacre’, p. 97. 216 Fiey, Assyrie chrétienne III, p. 208; Becker, Fear of God and the Beginning of Wisdom, pp. 162–63. On the East Syriac Chronicle of Sıırt written in Arabic and datable to the eleventh century, see Debié, L’Écriture de l’histoire en syriaque, A7, pp. 634–36. 217 Debié, ‘Livres et monastères en Syrie-Mésopotamie d’après les sources syriaques’.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

The Syro-Arabic Chronicle of Sıırt preserves the name of one teacher in Ḥirta/ al-Ḥīra: Sahiq, who was the father of Sergius, Mar Aba’s student.218 Barṣauma, an interpreter (of the scriptures, a traditional title for a high-level teacher) from Ḥirta/al-Ḥīra, was made the metropolitan of Nisibis by Isoʿyahb II (in office 638–45), and Elijah, the founder of the influential Upper Monastery in Mosul (sixth–seventh century), studied in Ḥirta/al-Ḥīra (probably because he was originally from there). The Christian names of the clerics obscure their original Arabic names except in the case of Sahiq, who may have been a layman. Although it is impossible to confirm continuity, it is interesting to note that Ḥunayn b. Isḥāq al-ʿIbādī (809–73), a prominent Christian scholar of the Abbasid period who was active in translating from Greek into Syriac, from Greek into Arabic, and from Syriac into Arabic, originally came from Ḥirta/al-Ḥīra and may have received his early education in Syriac and Arabic there.219 The school may have existed up to the eleventh century.220 Syriac Christian texts of the Church of the East belonging to the period before the separation of the churches were studied, interpreted, and possibly copied there, in an Arab environment. Since the Miaphysites were also present in Ḥirta/al-Ḥīra, the competition between two churches makes it likely that the texts of each tradition were discussed against the background of the rival church. It is interesting to note that when the antiquarian Hishām b. Muḥammad al-Kalbī (d. 819) recounts the history of the Lakhmid dynasty, he says that he ‘took the accounts of the Arabs and the genealogies of the clan of Naṣr b. Rabīʿa and the lifespans of those who acted as agents for the Persian imperial family and the history of their times’ from the monasteries of Ḥirta/al-Ḥīra.221 He does not mention the language of his sources, but it may have been Arabic. He wrote a now lost book on the churches and monasteries of Ḥirta/al-Ḥīra, which is mentioned in the Fihrist of Ibn al-Nadīm and possibly in the tenth/eleventh-century Syro-Arabic chronicle Mukhtaṣar al-akhbār al-biʾiyya.222 The Chronicle of Sıırt, which preserves the same information on the succession of the Lakhmid kings, also points in the direction of a local record of political events. The Mukhtaṣar al-akhbār al-biʾiyya preserves three notices (nos 84–87) on Ḥirta/al-Ḥīra, King al-Nuʿmān, the bishop Simeon b. Jābir, and the tribe of the ʿIbād. Whether the information in the Syro-Arabic chronicles came from direct sources or from Hishām b. Muḥammad al-Kalbī remains to be determined, since the question has received very little attention so far. These Arabic sources, both Christian and Islamic, suggest the existence of at least a written record of the succession of the Lakhmid kings produced in the Christian milieu of the city, and there may also have been other written historical material centred on the city. 218 Becker, Fear of God and the Beginning of Wisdom, p. 163. 219 See the chapter by Aaron Butts in this volume, as well as Becker, Sources for the History of the School of Nisibis; Debié, ‘Sciences et savants syriaques’; Rosenthal, ‘Ishâq b. Hunayn’s Taʾrîh al-atibbâʾ’; Anawati and Iskandar, ‘Ḥunayn b. Isḥāq al-ʿIbādī, Abū Zayd’. 220 Becker, Fear of God and the Beginning of Wisdom, p. 163. 221 Cited by Hoyland, ‘Insider and Outsider Sources’, p. 272. See also Toral-Niehoff, Al-Hira. 222 On this chronicle, see Debié, L’Écriture de l’histoire en syriaque, A8, pp. 636–38.

245

246

m u r i e l de bi é

The so-called Khuzistan Chronicle223 mentions the transfer of Patriarch Ishoʿyahb’s corpse from the village of Beth Qushay near Ḥirta/al-Ḥīra, where he died, to a monastery newly built by al-Nuʿmān’s sister Hind, who went out of the city to receive the corpse with the priests and the faithful. This information likely came from a local record written and kept in the monastery she built.224 Such a record would follow the tradition of monastic foundations to celebrate their holy founder. In the absence of a saint, the remains of a catholicos were prestigious enough. Frustratingly, the Syriac and Arabic histories do not mention the language in which these records about Ḥirta/al-Ḥīra were written. Syriac seems the obvious answer, especially in the episode about Hind, with a later translation into Arabic that the Arabic Christian and Muslim authors copied. We cannot, however, exclude entirely the possibility that some information, at least, was originally written in Arabic. The succession of the Lakhmid kings was likely compiled locally by their chancellery. Such lists were needed for calendric reasons: dating according to regnal lists was traditional in Late Antiquity, when several dating eras were in use.225 In Ḥirta/al-Ḥīra, where we have no traces of dating according to the era of the province of Arabia, although it was probably in use there too, the regnal years of the local dynasty probably served as a dating system.226 It is striking that in the Roman Empire a Syriac manu­script produced near Tadmor/Palmyra in the 550s or 560s is dated with reference to the days of the Arab ‘King Abū Karīb’.227 In the absence of similar ancient East Syriac manu­scripts, we do not have equivalent evidence about the Lakhmids. It is worth noting, however, that the local authority of the Arab kings was acknowledged for relative dating. It is likely that lists of regnal years were compiled in Ḥirta/al-Ḥīra at least for fiscal, legal, and archival reasons. It has to remain hypothetical that this record was maintained in Arabic (and in Arabic script?), but this type of information — dealing with Arab kings, recorded in a city in which the East Syriac Christians, who had a strong written tradition and a school, were Arabs, and reflecting a time in which Arabic had already acquired a ­graphic form — is a perfect candidate for a record in Arabic. Grist to the mill is added by an Islamic tradition that identifies Ḥirta/al-Ḥīra as the place where the Arabic alphabet first evolved and from where knowledge of it spread across the Najd to Mecca in western Arabia over the course of the sixth century.228 Another outcome of the cohabitation of languages in Ḥirta/al-Ḥīra is that Persian loanwords likely entered Arabic already before the Islamic period. The 223 This East Syriac chronicle may be dated to the 660s. Debié, L’Écriture de l’histoire en syriaque, E12, pp. 611–13. 224 Hoyland, ‘Insider and Outsider Sources’, pp. 272–73. 225 See Debié, L’Écriture de l’histoire en syriaque, pp. 73–75 on chrono­logical lists as a basis for writing history. 226 On the Lakhmid king lists and chrono­logy, see Rothstein, Die Dynastie der Lahmiden in al-Hira, pp. 5–60. 227 BL, MS Add 14559, fol. 107v (Wright, Catalogue of Syriac Manu­scripts in the British Museum, pp. 468–69). The manu­script is also placed in the days of bishops Jacob (Baradeus) and Theodore. Cited by Nöldeke, Die ghassanischen Fürsten aus dem Hause Gafna’s, p. 26. 228 Trimingham, Christianity among the Arabs in Pre-Islamic Times, pp. 156–57, 227.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

poet al-Aʿshā Maymūn (d. after 3/625), born in the vicinity of Ḥirta/al-Ḥīra and educated there, is reported to have been fond of introducing Persian words into his verses.229

Conclusion This chapter has shown that Arabs were part of the literate administrative and religious cultures of late antique empires and kingdoms. The Arab kingdoms participated in the global exchange of documents, and they produced their own legal and perhaps also historical documents. Phylarchs, soldiers, and merchants travelled in and outside the Arabian Peninsula all the way to Constantinople and Seleucia-Ctesiphon, to the imperial garrison cities, to battlefields, and to fairs and were thus exposed to the many languages and scripts used in the everyday life of courts, armies, and markets. Jewish and Christian Arabs were also involved in the multilingual lives of their own religious networks. Highly sophisticated discussions on Christian theo­logical issues took place at the courts of the Arab phylarchs. Monks, bishops, and priests travelled to and participated in Church councils that were convened by the Roman and Sasanian emperors. They not only met co-religionists who spoke all sorts of languages but also used the official languages of the Church in each empire. Textual testimonies thus enable us to map a different relationship to literacy than inscriptions do. Although scanty, they reveal networks of correspondence within and beyond the Arabian Peninsula as well as the existence of a legal tradition of contract writing. Epi­graphy remains central for identifying paths of communication, mapping religious practices, and tracing the slow standardization of the Arabic language and script. Yet texts make visible traces of long-disappeared documents and texts. The names ‘Saracens’ and ‘Tayyaye’ given to the Arabs since the third century suggest that already the Roman Empire recognized an Arab identity. In turn, the inscription that calls Imruʾ al-Qays ‘king of all the Arabs’ points to self-identification by the Arabs at the beginning of the fourth century. When the Arabic script emerged from the evolution of Nabataean forms and possibly Syriac influences, it became a medium of expression for Arabs, especially Christians. The concentration of inscriptions in the Nabataeo-Arabic script in north-west Arabia suggests that the Arabic script developed in the chancelleries of the principalities in this area, from where it was brought to Syria in the late fifth or early sixth century.230 Although South Arabian scripts would have been more appropriate for representing Arabic than the Aramaic alphabet was, since the latter lacks certain letters required to render Arabic properly, the choice of Aramaic as a ­graphic model must have had an ideo­logical as well as a practical reason: legal documents as well as inscriptions were written in Nabataean by Arabs, so scribes were trained in that script. The absence

229 Bosworth, ‘Iran and the Arabs before Islam’, pp. 609–11. 230 Nehmé, ‘Epi­graphy on the Edges of the Roman Empire’, p. 14.

247

2 48

m u r i e l de bi é

of centralization explains the variations in post-classical Nabataean, but it does not explain how the form of Palaeo-Arabic imposed itself beginning in the fifth century and why it was practiced and transmitted especially by Christians. Some form of centralization is necessary for common forms to be adopted in different places. We can surmise that the Ḥimyarite kings’ official support of Judaism since the 380s was an incentive for Arab Christians to express themselves in their own written medium that differed from the South Arabian scripts linked to Judaism and used by Sabean elites. It was a means of asserting their Arab identity in the South Arabian kingdom of Ḥimyar that had extended its domination to central and western Arabia. The same was true of the Christian Arabs under Sasanian dominion in Ḥirta/al-Ḥīra, who would have wanted to express an identity different from that of the Persians in the Church of the East. The Ghassānids are attested in the mid-fourth century in west central Arabia,231 and at least some of them headed north in the fifth and sixth centuries and served the Roman Empire. They were present in the regions in which the most ancient Arabic inscriptions have been found, in Ḥimā, and they were Christians, as were those who made these inscriptions. They are one of the possible candidates for the agents of the diffusion of the Palaeo-Arabic script, although Ghassān did not exist anymore as a tribe in the sixth century. Kindite princes dominated western and central Arabia on behalf of Ḥimyar and, during two decades (500–520), were kings of Ḥirta/al-Ḥīra as vassals of the Sasanians. As such, they represented a major force in the region and potential agents for the spread of the Arabic script. The ʿIbād in Ḥirta/al-Ḥīra are similarly possible agents of the development of Arabic writing. We have not thus far found any Arabic inscriptions by the main phylarchs known from the texts or from inscriptions in other languages. In the Roman Empire, they were immersed in a web of Greek literacy: their titles were Greek, and their official civic and ecclesiastical correspondence was likewise in Greek — or in Syriac for those who were Miaphysite. Arabic was squeezed in alongside Greek or between Greek and Syriac in ecclesiastical contexts. But it was used for internal military organization in the armies of the Arab phylarchs, as shown by the inscription recording the dispatch of a soldier by the phylarch. Variations in the Late Nabataean and Nabataeo-Arabic scripts might be explained by differences in the languages they convey and the milieus that used them. The newly found Christian Arabic inscription of ‘Yazīd the king’ (near Qaṣr Burquʿ in north-eastern Jordan) that may date to the beginning of Islam232 seems to attest continuous use of a pre-Islamic palaeo­graphy (and wawation) in Christian milieus.233 It could thus point to a distinctive tradition of writing that endured after Islam and existed in parallel to other evolved forms of the Arabic script. It is difficult to go further in the identification of the milieu that saw the emergence and diffusion of the Arabic

231 In the fourth century, a tribe in the region of Medina claims Ghassanid origins and Ammianus Marcellinus mentions a Ghassanid king in Mesopotamia. 232 This Yazīd might be the son of Muʿāwiya b. Abī Ṣufyān, the founder of the Umayyad dynasty. 233 Al-Shdaifat and others, ‘An Early Christian Arabic Graffito Mentioning “Yazīd the King”’.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

script, but new findings will no doubt help pull together the clues offered by both texts and inscriptions for a more complete view of the development of the Arabic script and its use in composition before Islam.

Biblio­graphy Primary Sources The Book of the Ḥimyarites: Fragments of a Hitherto Unknown Syriac Work, ed. and trans. by Axel Moberg (Lund: C. W. K. Gleerup, 1924) Documenta ad Origines Monophysitarum, ed. by J.-B. Chabot, CSCO, 2nd ser., 37, Textus (Paris, 1908; repr. Louvain: Durbecq, 1933) Egeria’s Travels, trans. by John Wilkinson (Oxford: Oxbow Books, 2002) The History of the Great Deeds of Bishop Paul of Qentos and Priest John of Edessa, ed. and trans. by H. Arneson, E. Fiano, and C. Luqris Marquis (Piscataway, NJ: Gorgias Press, 2010) Ibn Hishām, Sīra: Das Leben Muhammed’s nach Muhammed Ibn Ishâk, ed. by Ferdinand Wüstenfeld, 4 vols (Göttingen: Dieterich, 1857–1860) John of Ephesus, Lives of the Eastern Saints, ed. and trans. by E. W. Brooks, Patro­logia Orientalis, 17, 1 (Paris: Firmin-Didot, 1923) Kyrillos von Skythopolis, ed. by E. Schwartz, Texte und Untersuchungen zur Geschichte der altchristlichen Literatur, 4. Reihe, 4. Band, 2. Heft = 49. Band, 2. Heft (Leipzig: J. C. Heinrichs, 1939) Malalas, John, Chrono­graphia, ed. by Hans Thurn (Berlin: Walter de Gruyter, 2000) Malalas, John, The Chronicle of John Malalas, trans. by Elizabeth Jeffreys, Michael Jeffreys, and Roger Scott (Melbourne: Australian Association for Byzantine Studies, 1986) Menander, The History of Menander the Guardsman, ed. and trans. by Roger Blockley (Liverpool: Francis Cairns, 1985) Michael the Syrian, Chronicle = Chronique de Michel le Syrien, French trans. (vols 1–3) and Syriac text (vol. 4) by J.-B. Chabot (Paris: Pierre Leroux, 1899–1910) Notitia Dignitatum: Accedunt Notitia Urbis Constantinopolitanae, Laterculi Provinciarum, ed. by Otto Seeck (Berlin: Weidmanns, 1876) Sozomen, Ecclesiastical History = Kirchengeschichte, ed. by J. Bidez and G. C. Hansen (Berlin: Akademie Verlag, 1960) Synodicon Orientale, ou, Recueil de synodes nestoriens, ed. by J.-B. Chabot (Paris: Imprimerie Nationale, 1902) Theodoret, Graecarum affectionum curatio = Thérapeutique des maladies helléniques, ed. by Pierre Canivet (Paris: Le Cerf, 1958)

249

250

m u r i e l de bi é

Secondary Works Al-Azmeh, Aziz, The Emergence of Islam in Late Antiquity: Allah and His People (Cambridge: Cambridge University Press, 2014) Al-Jallad, Ahmad, ‘Graeco-Arabica I: The Southern Levant’, in Arabic in Context: Celebrating 400 Years of Arabic at Leiden, ed. by Ahmad Al-Jallad (Leiden: Brill, 2017), pp. 99–186 —— , ‘The Earliest Stages of Arabic and Its Linguistic Classification’, in The Routledge Handbook of Arabic Linguistics, ed. by Elabbas Benmamoun and Reem Bassiouney (London: Routledge, 2018), pp. 315–31 —— , ‘What Is Ancient North Arabian?’, in Re-engaging Comparative Semitic and Arabic Studies, ed. by Daniel Birnstiel and Naʿama Pat-El (Wiesbaden: Harrassowitz, 2018), pp. 1–43 —— , ‘The Linguistic Landscape of Pre-Islamic Arabia: Context for the Qur’an’, in The Oxford Handbook of Qur’anic Studies, ed. by Mustafa Shah and Muhammad Abdel Haleem (Oxford: Oxford University Press, 2020), pp. 111–27 —— , ‘The Di­graph ‫ اى‬in the Quranic Consonantal Text and the Identification of a New Letter Shape for Final Hē in the 7th to 8th Century Arabic Script’, Al-ʿUṣūr al-Wusṭā, 29 (2021), 67–85 Al-Jallad, Ahmad, and Chams Bernard, ‘New Safaitic and Greek Inscriptions from the Jordanian Ḥarrah Relating to Auxiliary Roman Military Units’, Zeitschrift der Deutschen Morgenländischen Gesellschaft, 171.1 (2021), 69–80 Al-Jallad, Ahmad, and Hythem Sidky, ‘A Paleo-Arabic Inscription on a Route North of Ṭāʾif ’, Arabian Archaeo­logy and Epi­graphy, 33.1 (2021), 202–15, https://onlinelibrary. wiley.com/doi/10.1111/aae.12203 [accessed 10 September 2022] Alpi, Frédéric, ‘La correspondance de Sévère d’Antioche et de Philoxène de Mabbōg avec les chrétiens d’al-Ḥīra (Hirtha de Nuʿmān): implications d’ordre doctrinal, institutionnel et politique’, forthcoming Anawati, G. C., and A. Z. Iskandar, ‘Ḥunayn b. Isḥāq al-ʿIbādī, Abū Zayd’, Dictionary of Scientific Bio­graphy, 15 (1970), 230–49 Azarnouche, Samra, ‘Arabes et Iraniens avant et au début de l’islam: aperçu de quelques zones de contact et de conflit’, in Le Coran des historiens, ed. by M.-A. Amir-Moezzi and G. Dye (Paris: Le Cerf, 2019), pp. 156–82 Beaucamp, Joëlle, ‘Le rôle de Byzance en mer Rouge sous le règne de Justin: mythe ou réalité?’, in Juifs et chrétiens en Arabie aux ve et vie siècles: regards croisés sur les sources, ed. by Joëlle Beaucamp, Françoise Briquel-Chatonnet, and Christian Julien Robin (Paris: ACHCByz, 2010), pp. 197–220 Beaucamp, Joëlle, Françoise Briquel-Chatonnet, and Christian Julien Robin, ‘La persé­ cution des chrétiens de Najrân et la chrono­logie himyarite’, ARAM, 11 (1999–2000), 15–83 Becker, Adam, Fear of God and the Beginning of Wisdom: The School of Nisibis and Christian Scholastic Culture in Late Antique Mesopotamia (Philadelphia: University of Pennsylvania Press, 2006) —— , Sources for the History of the School of Nisibis, Translated Texts for Historians, 50 (Liverpool: Liverpool University Press, 2008)

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

Bevan, George, Greg Fisher, and Denis Genequand, ‘The Late Antique Church at Tall al-ʿUmayrī East: New Evidence for the Jafnid Family and the Cult of St Sergius in Northern Jordan’, Bulletin of the American Schools of Oriental Research, 373 (May 2015), 49–68 Bosworth, C. E., ‘Iran and the Arabs before Islam’, in The Cambridge History of Iran, iii/1, ed. by E. Yarshater (Cambridge: Cambridge University Press, 1983), pp. 593–612 Bowersock, Glen, ‘The Ḥadramawt between Persia and Byzantium’, in La Persia e Bisantium, Atti dei convegni Lincei, 201 (Rome: Accademia Nazionale dei Lincei, 2004), pp. 263–73 —— , The Throne of Adulis: Red Sea Wars on the Eve of Islam (Oxford: Oxford University Press, 2013) —— , The Crucible of Islam (Cambridge, MA: Harvard University Press, 2017) Briquel-Chatonnet, Françoise, ‘De l’écriture édessénienne à l’estrangelâ et au sertô’, Semitica, 50 (2001), 81–90 Brock, Sebastian P., ‘Greek and Syriac in Late Antique Syria’, in Literacy and Power in the Ancient World, ed. by Alan K. Bowman and Greg Woolf (Cambridge: Cambridge University Press, 1994), pp. 149–60 —— , ‘Syriac Writers from Beth Qatraye’, ARAM, 11 (1999–2000), 85–96 Chaumont, Marie-Louise, ‘Aspbed’, in Enyclopedia Iranica (online edn), https:// iranicaonline.org/articles/aspbed-older-asppat-from-oir [accessed 12 September 2022] Contini, Roberto, ‘Il Ḥawran preislamico: ipotesi di storia linguistica’, Felix Ravenna, 133–34 (1987), 25–79 Cook, Michael A., ‘The Origins of “Kalām”’, Bulletin of the School of Oriental and African Studies, 43 (1980), 32–43 Costa, José, ‘Les juifs d’Arabie dans la littérature talmudique’, in Le judaïsme de l’Arabie antique, ed. by Christian Julien Robin (Turnhout: Brepols, 2015), pp. 453–84 Cotton, Hannah, and Ada Yardeni, Aramaic, Hebrew and Greek Documentary Texts from Nahal Hever and Other Sites (Oxford: Clarendon Press, 1997) De Blois, ‘The Date of the Martyrs of Najran’, Arabian Archaeo­logy and Epi­graphy, 1 (1990), 110–28 Debié, Muriel, ‘Livres et monastères en Syrie-Mésopotamie d’après les sources syriaques’, in Le monachisme syriaque, ed. by F. Jullien, Études syriaques, 7 (Paris: Geuthner, 2010), pp. 123–68 —— , ‘Sciences et savants syriaques: une histoire multiculturelle’, in Les sciences en syriaque, ed. by É. Villey, Études syriaques, 11 (Paris: Geuthner, 2014), pp. 9–66 —— , L’Écriture de l’histoire en syriaque: transmissions interculturelles et constructions identitaires entre hellénisme et islam (Leuven: Peeters, 2015) —— , ‘Les controverses miaphysites en Arabie et le Coran’, in Les controverses religieuses en syriaque, ed. by F. Ruani, Études syriaques, 13 (Paris: Geuthner, 2016), pp. 137–56 —— , Alexandre le Grand en syriaque: du roman grec aux apocalypses chrétiennes et au Coran (Paris: Les Belles Lettres, forthcoming in 2023) Déroche, François, La transmission écrite du Coran dans les débuts de l’islam: le Codex Parisino-Petropolitanus (Leiden: Brill, 2009) Detoraki, Marina, Le martyre de Saint Aréthas et de ses compagnons (BHG 166): édition critique, étude et annotation, traduction par Joëlle Beaucamp, appendice sur les versions orientales par André Binggeli (Paris: ACHCByz, 2007)

25 1

252

m u r i e l de bi é

El-Badawi, Emran, ‘Divine Kingdom in Syriac Matthew and the Qur’an’, Journal of Eastern Christian Studies, 61.1–2 (2009), 1–42 —— , ‘The Impact of Aramaic (Especially Syriac) on the Qurʾān’, Religion Compass, 8.7 (2014), 220–28 Feissel, Denis, ‘Épi­graphie et constitutions impériales: aspects de la publication du droit à Byzance’, in Epigrafia medi­evale greca e latina, ideo­logia e funzione: atti del seminario di Erice (12–18 settembre 1991), ed. by G. Cavallo and C. Mango (Spoleto: Fondazione CISAM, 1995), pp. 67–98 —— , ‘Introduction: ll’asile ecclésiastique, du droit impérial aux témoins documentaires’, in ‘De Tyr à Jérusalem: six inscriptions d’Asylie grecques et latines pour des églises de Phénicie et de Palestine (vie s.)’, ed. by Leah di Segni and D. Feissel, Travaux et Mémoires, 24.2 (2020), 549–56 Ferguson, Charles A., ‘The Arabic Koine’, Language, 35.4 (1959), 616–30 Fiema, Z. T., Ahmad Al-Jallad, Michael Macdonald, and Leila Nehmé, ‘Provincia Arabia: Nabataea, the Emergence of Arabic as a Written Language, and Graeco-Arabica’, in Arabs and Empires before Islam, ed. by Greg Fisher (Oxford: Oxford University Press, 2015), pp. 276–373 Fiey, Jean-Maurice, Assyrie chrétienne III: Bet Garmaï, Bet Aramāyé et Maišān nestoriens (Beirut: Deir el-Mashreq, 1968) —— , ‘Résidences et sépultures des patriarches syriaques-orientaux’, Le Muséon, 98 (1985), 149–68 Fisher, Greg, Between Empires: Arabs, Romans, and Sasanians in Late Antiquity (Oxford: Oxford University Press, 2011) Fisher, Greg, and Philip Wood, ‘Writing the History of the “Persian Arabs”: The PreIslamic Perspective on the “Naṣrids” of al-Ḥīrah’, Iranian Studies, 49.2 (2016), 247–90, http://dx.doi.org/10.1080/00210862.2015.1129763 [accessed 12 September 2022] Garitte, Gérard, ‘La version géorgienne de la passion de S. Procope par Eusèbe’, Le Muséon, 65 (1953), 245–66 Gatier, Pierre-Louis, ‘Les Jafnides dans l’épi­graphie grecque au vie siècle’, in Les Jafnides: des rois arabes au service de Byzance (vie siècle de l’ère chrétienne); actes du colloque de Paris, 24–25 novembre 2008, ed. by Denis Genequand and Christian Julien Robin, Orient et Méditerranée, 17 (Paris: De Boccard, 2015), pp. 193–222 Gorea, Maria, ‘Les classes sacerdotales (mišmarôt) de l’inscription juive de Bayt Ḥāḍir (Yémen)’, in Le judaïsme de l’Arabie antique, ed. by Christian Julien Robin (Turnhout: Brepols, 2015), pp. 297–329 Griffith, Sidney H., ‘Christian Lore and the Arabic Qurʾan: The “Companions of the Cave” in Surat al-Kahf and in Syriac Tradition’, in The Qurʾan in Its Historical Context, ed. by G. S. Reynolds (London: Routledge, 2008), pp. 109–37 Guillaume, Alfred, The Life of Muhammad: A Translation of Ishaq’s Sirat Rasul Allah (New York: Oxford University Press, 1955; 11th edn, 1996) Gyselen, Rika, ‘Spāhbed’, in Encyclopedia Iranica (online edn), https://iranicaonline.org/ articles/spahbed [accessed 12 September 2022] Gzella, Holger, A Cultural History of Aramaic: From the Beginnings to the Advent of Islam (Leiden: Brill, 2015) —— , Aramaic: A History of the First World Language (Grand Rapids, MI: Eerdmans, 2021) Healey, John, ‘Sources for the Study of Nabataean Law’, New Arabian Studies, 1 (1993), 203–14.

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

—— , ‘The Early History of Syriac Script: A Reassessment’, Journal of Semitic Studies, 45.1 (2000), 55–68 —— , ‘New Evidence of the Aramaic Legal Tradition from Elephantine to Edessa’, in Studia Semitica: The Journal of Semitic Studies Jubilee Volume, ed. by P. S. Alexander, G. J. Brooke, A. Christmann, J. F. Healey, and P. C. Sadgrove (Oxford: Oxford University Press, 2005), pp. 115–27 Hoyland, Robert, Arabia and the Arabs: From the Bronze Age to the Coming of Islam (London: Routledge, 2001) —— , ‘Language and Identity: The Twin Histories of Arabic and Aramaic (and: Why Did Aramaic Succeed Where Greek Failed?)’, Scripta Classica Israelica, 23 (2004), 183–99 —— , ‘Late Roman Provincia Arabia, Monophysite Monks and Arab Tribes: A Problem of Center and Periphery’, Semitica et Classica, 2 (2009), 117–39 —— , ‘The Jews of the Hijaz in the Qurʾān and in Their Inscriptions’, in New Perspectives on the Qurʾān: The Qurʾān in Its Historical Context, ii, ed. by G. S. Reynolds (London: Routledge, 2011), pp. 91–116 —— , ‘Insider and Outsider Sources: Historio­graphical Reflections on Late Antique Arabia’, in Inside and Out: Interactions between Rome and the Peoples on the Arabian and Egyptian Frontiers in Late Antiquity, ed. by Jitse H. F. Dijkstra and Greg Fisher (Leuven: Peeters, 2014), pp. 267–80 Humbach, Helmut, and Prods O. Skjaervø, The Sassanian Inscription of Paikuli, iii, pt. 1 (Wiesbaden: Reichert, 1983) Kister, M. J., ‘Al-Ḥīra: Some Notes on Its Relations with Arabia’, Arabica, 15 (1968), 143–69 Khan, Geoffrey, ‘Remarks on the Historical Background and Development of Early Arabic Documentary Formulae’, Asiatische Studien, 62.3 (2008), 885–906 —— , ‘The Historical Development of Early Arabic Documentary Formulae’, in Scribes as Agents of Language Change, ed. by Esther-Miriam Wagner and others (Berlin: Walter de Gruyter, 2013), pp. 199–215 —— , ‘The Opening Formula and Witness Clauses in Arabic Legal Documents from the Early Islamic Period’, Journal of the American Oriental Society, 139.1 (2019), 23–40 Klíma, O., ‘Bahrām V’, in Encyclopaedia Iranica (online edn), http://www.iranicaonline. org/articles/bahram-05 [accessed 30 December 2012] Knauf, Ernst A., ‘Arabo-Aramaic and ʿArabiyya: From Ancient Arabic to Early Standard Arabic, 200 ce–600 ce’, in The Qurʾān in Context: Historical and Literary Investigations into the Qurʾānic Milieu, ed. by Angelika Neuwirth, Nicolai Sinai, and Michael Marx, Leiden Texts and Studies on the Qurʾān, 6 (Leiden: Brill, 2010), pp. 197–254 Lecker, Michael, ‘The Levying of Taxes for the Sasanians in Pre-Islamic Medina’, Jerusalem Studies in Arabic and Islam, 27 (2002), 109–26 Lewin, Ariel S., ‘Did the Roman Empire Have a Military Strategy and Were the Jafnids Part of It?’, in Les Jafnides: des rois arabes au service de Byzance (vie siècle de l’ère chrétienne); actes du colloque de Paris, 24–25 novembre 2008, ed. by Denis Genequand and Christian Julien Robin, Orient et Méditerranée, 17 (Paris: De Boccard, 2015), pp. 168–69 Macdonald, Michael C. A., ‘Reflections on the Linguistic Map of Pre-Islamic Arabia’, Arabian Archaeo­logy and Epi­graphy, 11 (2000), 28–79 —— , ‘Languages, Scripts, and the Uses of Writing among the Nabataeans’, in Petra Rediscovered: The Lost City of the Nabataeans, ed. by Glenn Markoe (London: Thames & Hudson, 2003)

25 3

254

m u r i e l de bi é

—— , ‘Ancient North Arabian (Chapter 16)’, in The Cambridge Encyclopedia of the World’s Ancient Languages, ed. by Roger D. Woodard (Cambridge: Cambridge University Press, 2004), pp. 488–533 —— , ‘Literacy in an Oral Environment’, in Writing and Ancient Near Eastern Society: Papers in Honour of Alan R. Millard, ed. by Piotr Bienkowski, Christopher Mee, and Elizabeth Sater (New York: T & T Clark, 2005), pp. 49–118 —— , ‘Old Arabic’, in Encyclopedia of Arabic Language and Linguistics, ed. by Kees Versteegh, iii, Lat–Pu (Leiden: Brill, 2008), pp. 464–74 —— , ‘Arabs, Arabias, and Arabic before Late Antiquity’, Topoi, 16 (2009), 277–332 —— , ‘Arabia and the Written Word’, in The Development of Arabic as a Written Language, ed. by Michael C. A. Macdonald, Supplement to the Proceedings of the Seminar for Arabian Studies, 40 (Oxford: Archaeopress, 2010), pp. 5–28 ——— ‘Romans Go Home? Rome and Other “Outsiders” as Viewed from the SyroArabian Desert’, in Inside and Out: Interactions between Rome and the Peoples on the Arabian and Egyptian Frontiers in Late Antiquity, ed. by Jitse H. F. Dijkstra and Greg Fisher (Leuven: Peeters, 2014), pp. 145–64 Maraqten, M., ‘Writing Materials in Pre-Islamic Arabia’, Journal of Semitic Studies, 48.2 (1998), 287–310 Martindale, J. R., ‘Abramius II’, in The Prosopo­graphy of the Later Roman Empire, ii, ad 395–527 (Cambridge: Cambridge University Press, 1980), p. 3 Mascitelli, D., L’arabo in epoca preislamica: formazione di una lingua (Rome: ‘L’Erma’ di Bretschneider, 2006) Millar, Fergus, A Greek Roman Empire: Power and Belief under Theodosius II (408–450) (Berkeley: University of California Press, 2006) —— , ‘Rome’s “Arab” Allies in Late Antiquity: Conceptions and Representations from within the Frontiers of the Empire’, in Commutatio et Contentio: Studies in the Late Roman, Sasanian and Early Islamic Near East in Memory of Zeev Rubin, ed. by Henning Börm and Josef Wiesehöfer, Reihe Geschichte, 3 (Düsseldorf: Wellem, 2010), pp. 199–226 —— , ‘A Syriac Codex from near Palmyra and the “Ghassanid” Abokarib’, Hugoye: Journal of Syriac Studies, 16.1 (2013), 15–35 Montgomery, James E. ‘The Empty Ḥijāz’, in Arabic Theo­logy, Arabic Philosophy: From the Many to the One; Essays in Celebration of Richard M. Frank, ed. by James E. Montgomery (Leuven: Peeters, 2006), pp. 37–97 Nehmé, Laïla, ‘Epi­graphy on the Edges of the Roman Empire: A Study of the Nabataean Inscriptions and Related Material from the Darb al-Bakrah, Saudi Arabia, 1st–5th Century ad’ (unpublished thesis, mémoire scientifique d’habilitation à diriger des recherches, École Pratique des Hautes Études, 2013) —— , ‘Aramaic or Arabic? The Nabataeo-Arabic Script and the Language of the Inscriptions Written in this Script’, in Arabic in Context: Celebrating 400 Years of Arabic at Leiden University, ed. by Ahmad Al-Jallad (Leiden: Brill, 2017), pp. 75–98 —— , ‘New Dated Inscriptions (Nabataean and Pre-Islamic Arabic) from a Site near alJawf, Ancient Dūmah, Saudi Arabia’, Arabian Epi­graphic Notes, 3 (2017), 121–66 Nöldeke, Theodor, Die ghassanischen Fürsten aus dem Hause Gafna’s (Berlin: Verlag der Königlichen Akademie der Wissenschaften, 1887)

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

Poirier, John, ‘The Linguistic Situation in Jewish Palestine in Late Antiquity’, Journal of Greco-Roman Christianity and Judaism, 14.3 (2007), 55–134 Powers, David, ‘Demonizing Zenobia: The Legend of al-Zabbāʾ in Islamic Sources’, in Histories of the Middle East: Studies in Middle Eastern Society, Economy and Law in Honor of A. L. Udovitch, ed. by Margariti Eleni Roxani, Adam Sabra, and Petra M. Sijpesteijn (Leiden: Brill, 2010), pp. 127–82 Prémare, Alfred-Louis de, ‘Les textes musulmans dans leur environnement’, Arabica, 47 (2000), 391–408 Puech, Émile, ‘La grotte aux lettres de Wadi Khabra’, in Présence arabe dans le Croissant fertile avant l’Hégire: actes de la table ronde internationale (Paris, 13 novembre 1993), ed. by Hélène Lozachmeur (Paris: Recherche sur les Civilisations, 1995), pp. 37–46 Robert, Jean-Noël, ‘Hiéroglossie des civilisations’, Fondation Collège de France website, April 2021, https://www.fondation-cdf.fr/2021/04/06/hieroglossie-des-civilisations/ [accessed 10 October 2022] Robin, Christian Julien, ‘Les graffites arabes islamiques écrits en caractères sudarabiques de Umm Layla’, Semitica, 26 (1976), 188–93 —— , ‘Les langues de la péninsule Arabique’, Revue du monde musulman et de la Méditerranée, 61 (1991), 89–111 —— , ‘La pénétration des Arabes nomades au Yémen’, Revue du monde musulman et de la Méditerranée, 61 (1991), 71–88 —— , ‘Les plus anciens monuments de la langue arabe’, Revue du monde musulman et de la Méditerranée, 61 (1991), 113–25 —— , ‘Le royaume hujride, dit “royaume de Kinda”, entre Himyar et Byzance’, Comptes rendus des séances de l’Académie des Inscriptions et Belles-Lettres, 140.2 (1996), 665–714 —— , ‘Les inscriptions de l’Arabie antique et les études arabes’, Arabica, 48 (2001), 509–77 —— , ‘Les Arabes de Ḥimyar, des “Romains” et des Perses (iiie–vie siècles de l’ère chrétienne)’, Semitica et Classica, 1 (2008), 167–202 —— , ‘Nagrān vers l’époque du massacre: notes sur l’histoire politique, économique et institutionnelle et sur l’introduction du christianisme’, in Juifs et chrétiens en Arabie aux ve et vie siècles: regards croisés sur les sources, ed. by Joëlle Beaucamp, Françoise BriquelChatonnet, and Christian Julien Robin (Paris: AChCByz, 2010), pp. 39–106 —— , ‘Les religions pratiquées par les membres de la tribu de Kinda (Arabie) à la veille de l’Islam’, Judaïsme Ancien – Ancient Judaism, 1 (2013), 203–61 —— , ‘Before Ḥimyar: Epi­graphic Evidence for the Kingdoms of South Arabia’, in Arabs and Empires before Islam, ed. by Greg Fisher (Oxford: Oxford University Press, 2015), pp. 90–126 —— , ‘Ghassān en Arabie’, in Les Jafnides, des rois arabes au service de Byzance (vie siècle de l’ère chrétienne); actes du colloque de Paris, 24–25 novembre 2008, ed. by Denis Genequand and Christian Julien Robin, Orient et Méditerranée, 17 (Paris: De Boccard, 2015), pp. 79–120 —— , ed., Le judaïsme de l’Arabie antique, Collection Judaïsme ancien et origines du christianisme, 3 (Turnhout: Brepols, 2015) —— , ‘Quel judaïsme en Arabie?’, in Le judaïsme de l’Arabie antique, ed. by Christian Julien Robin, Collection Judaïsme ancien et origines du christianisme, 3 (Turnhout: Brepols, 2015), pp. 15–295

25 5

2 56

m u r i e l de bi é

—— , ‘Arabie antique: variations dans la manière de désigner le Dieu unique’, in Hiéroglossie III. Persan, syro-araméen et les relations avec la langue arabe, ed. by J.-N. Robert (Paris: Collège de France, 2022), pp. 139–70 Robin, Christian Julien, Ali I. al-Ghabbān, and Saʿīd F. Al-Saʿīd, ‘Inscriptions antiques de la région de Najrān (Arabie Séoudite Méridionale): nouveaux jalons pour l’histoire de l’écriture, de la langue et du calendrier arabes’, Comptes rendus des séances de l’Académie des Inscriptions et Belles-Lettres, 3 (2014), 1033–28 Rosenthal, Franz, ‘Ishâq b. Hunayn’s Taʾrîh al-atibbâʾ’, Oriens, 7.1 (1954), 55–80 Rothstein, Gustav, Die Dynastie der Lahmiden in al-Hira: ein Versuch zur arabisch-persischen Geschichte zur Zeit der Sasaniden (Berlin: Reuther & Reichard, 1899) Rubin, Zeev, ‘From the Rabbanat at the Court of Sharaḥbiʾīl Yakkuf to the Tiberian Priests at the Court of Yūsuf Asʾar Yathʾar’, in Le judaïsme de l’Arabie antique, ed. by Christian Julien Robin (Turnhout: Brepols, 2015), pp. 437–52 Ryckmans, Jacques, ‘Les rapports de dépendance entre les récits hagio­graphiques relatifs à la persécution des Ḥimyarites’, Le Muséon, 100 (1987), 297–305 Sartre, Maurice, ‘Rome et les Arabes nomades: le dossier épi­graphique de Eeitha’, Les Jafnides: des rois arabes au service de Byzance (vie siècle de l’ère chrétienne), ed. by Denis Genequand and Christian Julien Robin, Orient et Méditerranée, 17 (Paris: De Boccard, 2015), pp. 37–51 al-Shdaifat, Younis, Ahmad Al-Jallad, Zeyad al-Salameen, and Rafe Harahsheh, ‘An Early Christian Arabic Graffito Mentioning “Yazīd the King”’, Arabian Archaeo­logy and Epi­graphy, 28 (2017), 315–24, https://doi.org/10.1111/aae.12105 [accessed 20 January 2023] Schoeler, Gregor, The Oral and the Written in Early Islam, ed. by James E. Montgomery, trans. Uwe Vagelpohl (London: Routledge, 2006) Shaddel, Mehdy, ‘The Development of the Meccan Sanctuary in the First Two Islamic Centuries’, forthcoming Shahid, Irfan, ‘Byzantino-Arabica: The Conference of Ramla, a.d. 524’, Journal of Near Eastern Studies, 23.2 (1964), 115–31 —— , Byzantium and the Arabs in the Fourth Century (Washington, DC: Dumbarton Oaks Research Library and Collection, 1984) —— , Byzantium and the Arabs in the Fifth Century (Washington, DC: Dumbarton Oaks Research Library and Collection, 1989) —— , Byzantium and the Arabs in the Sixth Century (Washington, DC: Dumbarton Oaks Research Library and Collection, 1995) Shaked, Shaul, ‘The ʿEn ʿAvdat Inscription Revisited’, in Le judaïsme de l’Arabie antique, ed. by Christian Julien Robin (Turnhout: Brepols, 2015), pp. 359–72 Shoemaker, Stephen, Creating the Qur’an: A Historical-Critical Study (Berkeley: University of California Press, 2022) Sijpesteijn, Petra M., ‘Arabic Script and Language in the Earliest Papyri: Mirrors of Change’, Jerusalem Studies in Arabic and Islam, 49 (2020), 433–93 —— , Shaping a Muslim State: The World of a Mid-Eighth-Century Egyptian Official, Oxford Studies in Byzantium (Oxford: Oxford University Press, 2013) Stein, Peter, ‘Literacy in Pre-Islamic Arabia’, abbreviated English version of ‘Stein vs. Holz, Musnad vs. Zabūr: Schrift und Schriftlichkeit im vorislamischen Arabien’, Die Welt des Orients, 35 (2005), 118–57

the a r a bs an d n o rt h e r n l an g uag e s and scri pt s b e fo re i slam

—— , ‘The Role of Aramaic on the Arabian Peninsula in the Second Half of the First Millennium bc’, in Languages, Scripts and Their Uses in Ancient North Arabia, ed. by Michael C. A. Macdonald, Supplement to the Proceedings of the Seminar for Arabian Studies, 48 (Oxford: Archaeopress, 2018), pp. 39–49 —— , ‘Correspondence in Pre-Islamic Yemen’, in Ex Oriente Lux: Sbornik statej k 75-letiju Mikhaila Borisoviča Piotrovskogo (Saint Petersburg: Bronze Horseman Books, 2019), pp. 336–46 Talib, Adam, ‘Fathers and Husbands’, in Arabic Humanities, Islamic Thought: Essays in Honor of Everett K. Rowson (Leiden: Brill, 2017), pp. 233–55 Tannous, Jack, ‘Between Christo­logy and Kalām? The Life and Letters of George, Bishop of the Arab Tribes’, in Malphono w-Rabo d-Malphone: Studies in Honor of Sebastian P. Brock, ed. by George Kiraz (Piscataway, NJ: Gorgias Press, 2008), pp. 671–716 —— , The Making of the Medi­eval Middle East: Religion, Society, and Simple Believers (Princeton, NJ: Princeton University Press, 2018) Taylor, David G. T., ‘Bilingualism and Diglossia in Late Antique Syria and Mesopotamia’, in Bilingualism in Ancient Society: Language Contact and the Written Word, ed. by J. N. Adams, Mark Janse, and Simon Swain (Oxford: Oxford University Press, 2002), pp. 298–331 Tillier, Matthieu, and Naïm Vanthieghem, ‘Recording Debts in Sufyanid Fustat: A Reexamination of the Procedures and Calendar in Use in the First/Seventh Century’, in Geneses: A Comparative Study of the Historio­graphies of the Rise of Christianity, Rabbinic Judaism and Islam, ed. by John Tolan (London: Routledge, 2019), pp. 148–88 Tobi, Yosef Y., ‘The Jewish Community in Ḥāṣī, South Yemen, in the Light of Its Makrab Ṣūrī’el and Cemetery’, in Le judaïsme de l’Arabie antique, ed. by Christian Julien Robin (Turnhout: Brepols, 2015), pp. 373–85 Toral-Niehoff, Isabelle, Al-Hira: eine arabische Kulturmetropole im spätantiken Kontext, Islamic History and Civilization, 104 (Leiden: Brill, 2014) Trimingham, John Spencer, Christianity among the Arabs in Pre-Islamic Times (London: Longman; Beirut: Librairie du Liban, 1979) Wasserstein, David, ‘Why Did Arabic Succeed Where Greek Failed? Language Change in the Near East after Muhammad’, Scripta Classica Israelica, 22 (2003), 257–72 Webb, Peter, Review of Arabs and Empires before Islam, ed. by Greg Fisher, Bulletin of the School of Oriental and African Studies, 79.3 (2016), 640–42 Weber, Ursula, ‘Narseh’, in Encyclopaedia Iranica (online edn, 2016), http://www. iranicaonline.org/articles/narseh-sasanian-king Wood, Philip, The Chronicle of Seert: Christian Historical Imagination in Late Antique Iraq (Oxford: Oxford University Press, 2013) Wright, William, Catalogue of Syriac Manu­scripts in the British Museum, Acquired since the Year 1838, ii (London: British Museum, 1871) Würsch, Renate, ‘Ḵawarnaq’, in Encyclopaedia Iranica (2013), xvi, fasc. 2, pp. 143–45 Yadin, Yigael, and others, The Documents from the Bar Kokhba Period in the Cave of Letters: Hebrew, Aramaic and Nabatean-Aramaic Papyri ( Jerusalem: Israel Exploration Society, 2002)

25 7

Robert Haug

Triliteral Coins and Political Authority along a Contentious Frontier Between Arabic, Bactrian, and Pahlavi in Late Antique Khurāsān Coins are a treasure trove of information for scholars of early Islamic history. In the words of Nicholas Sims-Williams, Their design and legends may reveal the religious and political attitudes and propaganda of those who issued them. Equally importantly, the dates which they bear and the names of rulers and mints can, in principle, document the advance — and occasional retreat — of the forces of Islam in a quite precise way.1 As such, coins are some of the most important and in many cases only primary sources available to historians of the early Islamic world, who often must rely on literary sources that were composed two or more centuries after the events they describe. As Joe Cribb has observed, speaking more broadly about the field of numismatics and its contribution to the study of the past, coins’ ‘contemporaneity with the events they participated in makes them tangible witnesses which do not need to rely on the memory or narratives of individuals. They come to us today as survivors of the events which created them’.2 The information that modern historians of the Islamic world (distinguished here from antiquarians, art historians, and philo­logists) seek from coins generally concerns the date when, the place where, and the authority under which a coin was struck — information largely provided by inscriptions. We then use this information to create and refine lists of emperors, kings, governors, and other state officials and check these lists against the information provided by our literary sources. Beyond the practical value this information holds for deciphering complex and frequently opaque political chrono­logies, modern historians often look to it to discern political statements, too (and here we may also note the importance of icono­graphy and

1 Sims-Williams, ‘The Arab-Sasanian and Arab-Hephthalite Coinage’, p. 115. 2 Cribb, ‘Money as Metaphor 2’, p. 495.

Robert Haug ([email protected]) is Associate Professor of Islamic World History in the Department of History at the University of Cincinnati. Navigating Language in the Early Islamic World: Multilingualism and Language Change in the First Cen­turies of Islam, ed. by Antoine Borrut, Manuela Ceballos, and Alison M. Vacca, ISMAR 2 (Turnhout: Brepols, 2024) pp. 259–279 10.1484/M.ISMAR-EB.5.134628

FHG

26 0

ro be rt hau g

images). Having had one’s name engraved on a coin implies that one held power over a mint that possessed sufficient authority to strike coins people would readily accept as an instrument of exchange. In the Islamic world, as coin production developed and standardized under the Umayyad and Abbasid caliphates, we think of this right to be named on coins — called sikka — as a specific privilege of the caliphs.3 With this understanding, we may then use the appearance and/or disappearance of individual names and titles on coins at specific times and in specific places as evidence of the waxing and waning of an individual’s or a dynasty’s political authority — an assumption we sometimes also apply to coins struck before minting practices were standardized or in areas where the reach of the centralized state was limited or ineffective. But as we think about the value of coins to the modern study of the distant past, we may ask whether this information would have been of any interest or concern to someone living in that past and using a coin as an instrument of exchange. For most monetary systems throughout history, and especially monetary systems that relied upon precious metal coinage, the value of a coin, which constitutes the information that is most important and relevant to anyone using that coin for its original and primary purpose, is transmitted through its size, weight, colour, or material, not through its inscriptions. If the most essential information was not transmitted through inscriptions, we must ask: who read the inscriptions on coins? In fact, the written information modern historians fixate upon would not have been transmitted to contemporaries via coins at all. As Michael Bates put it, proclamations with news of political changes and interpretations of such news arrived long before the arrival of coins issued in the capital, while of course no coins reflecting the new situation could be minted locally before the news arrived. By the time people saw the coins, the information on them was old news.4 In this case, for what audience was this information engraved upon coins? These questions are especially interesting when we encounter coins that include inscriptions not only in multiple languages but also in multiple scripts, with each inscription providing unique information. An interesting case is that of the so-called Arab-Hephthalite coins. These coins — produced primarily in the region of Ṭukhāristān (northern Afghanistan) in the late seventh century5 and described in detail below — appear at first glance to be standard Arab-Sasanian coins, modelled on the issues of the Sasanian Empire but with Arabic inscriptions added. These were the standard coinage throughout the lands of the former Iranian empire — primarily Iraq and Iran — until the late 690s. The notable difference found in Arab-Hephthalite coins is the addition of inscriptions in the Bactrian language,



3 Though the rules surrounding this imperial prerogative are not always clear or apparent. Bosworth, Darley-Doran, and Freeman-Greenville, ‘Sikka’. 4 Bates, ‘Who was Named on Abbasid Coins?’, p. 89. 5 Some argue for the early eighth century. The debates over dating the Arab-Hephthalite coins will be discussed later.

triliteral coins and political authority along a contentious frontier

an Iranian language used in Ṭukhāristān that employed a modified Greek script.6 These coins therefore include inscriptions in Pahlavi using a script derived from the Aramaic script, in Arabic using the Arabic script, and in Bactrian using a modified Greek script. In order to read and comprehend all the information transmitted by these coins, one must have some familiarity with all three languages and all three scripts. So who read these coins, and what were the intentions of the authorities who struck them? What do they tell us about the intersection of Arab-Muslim, Iranian, and Hephthalite7 authorities in early Islamic Ṭukhāristān and neighbouring Khurāsān? Such questions speak directly to the challenges facing the early caliphate as it attempted to assert its authority over a complex, multi-ethnic and multi-linguistic domain, and to the role that language, including written language, played in such assertions of power. Furthermore, these coins were struck at a time when Muslim authority was still establishing itself in the east, a process interrupted by the Second Fitna (680–92). A period of instability in a rapidly changing political, cultural, and linguistic environment created openings for competing expressions of political authority to appear on coins struck in Ṭukhāristān in the late seventh century even while changes to coin design occurred within a recognized imperial framework. In order to address the above questions, this paper explores the small series of Arab-Hephthalite coins in the context of Sasanian, Hephthalite, and Arab-Sasanian coinage as well as of the political situation in Ṭukhāristān and Khurāsān at the assumed time of their production.

The Arab-Hephthalite Coins Before placing the Arab-Hephthalite coins in their historical and political context, I will take a moment to describe the coins themselves.8 The Arab-Hephthalite coins are a rare and short-lived series of silver drachms that appear close to contemporary





6 On the Bactrian language and its relationship to New Persian and other Iranian languages, see Khodadad Rezakhani’s essay in this volume. 7 The Hephthalites were a group of Iranian Huns who used an Iranian language, Bactrian, at least in their names, titles, and inscriptions and who came to political dominance in Ṭukhāristān in the mid-fifth century. For approximately a century, a period known as the Imperial Hephthalite era, the Hephthalites ruled Ṭukhāristān and Sogdiana from their capital near Kundūz until they were conquered by the Western Türk Khāqānate in the mid-sixth century, possibly with the support of the Sasanians, who claimed Hephthalite domains south of the Amu Darya. Following the Imperial Hephthalite period, various Hephthalite lords maintained principalities in Ṭukhāristān (especially in the highlands), often as subjects of Turkic overlords. The last of these principalities survived into the eighth century. 8 The most complete description of this series is found in Göbl, Dokumente zur geschichte der Iranischen Hunnen in Baktrien und Indien, Emissions 271–81. Göbl’s types, which he called ‘Emissions’ using the abbreviation Em., are traditionally numbered using the abbreviation NumH following the work of Helmut Humbach in his Baktrische Sprachdenkmäler. In this schema, Göbl’s numbers do not change; the Em. is simply replaced with the NumH. An updated overview of this series can be found in Vondrovec, Coinage of the Iranian Huns and their Successors from Bactria to Gandhara, ii, pp. 531–35.

261

262

ro be rt hau g

Figure 5.1. Arab-Hephthalite coin of Zhulād Gōzgān from Ambīr, British Museum. 69 ah/ 688 ce. Photo courtesy of the British Museum. Reproduced with permission.

Arab-Sasanian issues but are distinguished by the inclusion of inscriptions in Bactrian as well as those in the typical Pahlavi and Arabic.9 They are so close in style to contemporary Arab-Sasanian issues minted in Khurāsān that when John Walker made the first attempts to identify them in 1941, unable to read the Bactrian inscriptions, he assumed them to be the product of the Arab governors of Khurāsān Rabīʿ b. Ziyād al-Ḥārithī (d. 673) and ʿAbd Allāh b. Khāzim al-Sulamī (d. 691–92), even speculating that the Bactrian inscriptions were forms of their names.10 The quality of the engraving on these coins is also comparable to that of contemporary Arab-Sasanian issues, so we should think of them not as ‘barbarous imitations’ like those minted in Ṭabaristān, Khwārazm, Bukhārā, and eastern Sīstān but as high-quality coin issues themselves.11 As Klaus Vondrovec observed, ‘This group is completely Arab-Sasanian in style and manufacture; its true identity is revealed only by its Bactrian legends’.12 It was another quarter century after Walker’s attempt to classify the Arab-Hephthalite coins before Robert Göbl was able to decipher the Bactrian and propose a firmer identification.13 Even then, we are still left with many questions about who struck them, when, and where. Although there are many small differences between issues and even individual coins of the same issue, I would like to take a moment to describe a typical ArabHephthalite coin. A sample coin of Göbl’s type NumH 280 is pictured in Fig. 5.1. The obverse of the Arab-Hephthalite silver drachms features a bust of the Sasanian Shāhānshāh Khusrow II (r. 590, 591–628), also the model for the majority of Arab-Sasanian coins. In front of the bust, in the typical position of the name of a governor on Arab-Sasanian issues from the 670s onwards, we find one of two names written in Bactrian. On Göbl’s NumH 272, the name is κορμοζαδο, or Kurmzād.

9 Göbl’s types NumH 271 and 271A are both copper issues that appear quite different from the silver drachms in style and should be addressed separately. They will not be a part of the current analysis. 10 Walker, A Catalogue of the Muhammadan Coins in the British Museum, i, pp. lxv–lxix, 127–29. 11 Album and Goodwin, Sylloge of Islamic Coins in the Ashmolean, i, p. 41. 12 Vondrovec, Coinage of the Iranian Huns and their Successors from Bactria to Gandhara, ii, p. 531. 13 Göbl, Dokumente zur geschichte der Iranischen Hunnen in Baktrien und Indien,, i, pp. 185–93, 257.

triliteral coins and political authority along a contentious frontier

On NumH 273–81, the name is ζολαδο γωζογανο, or Zhulād [of] Gōzgān.14 Two coins in Walker’s collection, numbers 138 and 150, include the name of Salm b. Ziyād (r. 680–83), the Umayyad governor of Khurāsān, in Arabic instead. Although these two coins are more clearly connected to the Arab governor of Khurāsān, they, too, include other inscriptions in Bactrian. To the left of the bust, one finds the Pahlavi phrase GDH ʾpzwtˊ, ‘the Royal Glory has been increased’. This phrase also appears in the margin on some issues. Around the obverse margin we find the Arabic phrase ‫بسم هللا‬, ‘in the name of God’, or some variant. In most cases, the basmala is the only Arabic included on the coins. The reverse features the traditional motif of a Zoroastrian fire altar flanked by two guardians. To the left of the altar and guardians is a date written in Pahlavi. The calendrical system used on these coins is debated, with some arguing for a Hijri15 dating and others arguing for a post-Yazdgird dating. The mint name is written in Pahlavi to the right of the altar. In some cases, the locations of the date and the mint name are reversed. A second mint marking in Bactrian is found around the reverse margin. The reverse margin also carries the Bactrian inscription γαϱιγο ϸαυο, ‘king of Ghar’ or ‘king of the mountains’, which we presume to be Zhulād’s title. In addition, many specimens are countermarked with a tamgha, or seal. Most Arab-Hephthalite issues originate from the mint of Ambīr (modern Sar-i Pul in northern Afghanistan) in the region of Gūzgān ( Jūzjān in Arabic) in Ṭukhāristān. The mint name given in Bactrian (αμβιρο) and sometimes in Pahlavi as well (ANBYR). Many Ambīr issues also name the mint as HWR’, or Khurāsān in Pahlavi, which prompts speculation about the origin of the dies and about whether they were engraved in Khurāsān — likely at the provincial capital of Marw — and then sent to Gūzgān, where the Bactrian inscriptions were added later. Similarly, NumH 281 names the mint as Ambīr in Bactrian and as Marw in Pahlavi (MRW’). NumH 272, the only type to name Kurmzād, includes Pahlavi mint marks for the unknown mints of APNWRAN or SPNWRAN. The presence of multiple languages on the coins may sometimes help in deciphering such unusual mint marks. For example, Walker’s type 150 includes the otherwise unknown Pahlavi mint mark of PHRZZ, but the Bactrian παϱσο, which could denote Pāryāb or Fāryāb, a city just west of Gūzgān, provides a likely solution to the undecipherable Pahlavi.16

14 Zhulād’s name also appears on the copper coins NumH 271 and 271A, but on the reverse and in place of the fire altar. 15 Throughout this article I will use the term Hijri to refer to the calendar system used by Muslims. The Hijri era begins with the hijra, or emigration, of the Prophet Muhammad and his followers from Mecca to Medina in 622 ce. ‘Hijri’ remains the most commonly used and recognized name for this calendrical system, but recently some debate has emerged on its appropriateness and association with the hijra because of the late appearance of the term in the papyro­logical record. Some have suggested the era be renamed “the jurisdiction of the Believers” in line with the earliest document which gives a name to the era. Donner, ‘Qur’ânicization of Religio-Political Discourse in the Umayyad Period’, pp. 86–88. 16 Sims-Williams, ‘The Arab-Sasanian and Arab-Hephthalite Coinage’, 121.

263

264

ro be rt hau g

Dating the Arab-Hephthalite Coins Let us turn now to the dating of the Arab-Hephthalite coins and attempt to place these coins in their historical context, an essential step towards understanding their use of language. As mentioned above, the dates for all issues were written in Pahlavi, but they do not give any clear indication of which calendrical system they are employing, a problem compounded by the multi-lingual and multi-cultural elements of the coin design. The coin of Kurmzād (NumH 272 in Göbl’s classification) is dated 39. Of Zhulād’s coins, NumH 273 and 274 are dated 66,17 NumH 275–79 are dated 68, and NumH 280 and 281 are dated 69. Zhulād’s copper coins, NumH 271 and 271A, are dated 62. The dates point to rather different historical contexts depending on whether they refer to the Hijri, Yazdgird, or post-Yazdgird eras. Here we will focus primarily on the silver issues of Zhulād. In his initial study, Walker assumed the dates to pertain to the Hijri calendar (remember that he thought these coins were likely produced by the Arab governors of Khurāsān, and Arab-Sasanian coins had used the Hijri calendar, written in Pahlavi, since the 660s), and he dated them roughly to the years 685,18 687, and 688 ce. The coins of Salm b. Ziyād with Bactrian inscriptions, Walker’s types 138 and 150, are dated 68 and 63, respectively, and these dates correspond with the Hijri dates of Arab-Sasanian coins minted in Salm’s name, despite the fact that Salm left Khurāsān in 63 ah.19 The Hijri dating would place these coin issues in the midst of the Second Fitna (680–92). On the other hand, Göbl and Vondrovec have argued for the post-Yazdgird era, meaning that the coins of Zhulād would be dated roughly 717, 719, and 720 ce, respectively.20 One reason for choosing this later date is the presence of a fourth circle surrounding the outer margin on the reverse of the coin. This is a feature that does not appear on Arab-Sasanian coins until 72 ah/691 ce, so if we presume that these coins were struck in direct imitation of Arab-Sasanian coins or from dies that had been engraved by the Arab governors of Khurāsān, we must push the date of these coins forward to at least 691.21 However, this argument is problematic, as the coins would then postdate the coinage reforms of Caliph ʿAbd al-Malik (r. 685–705), which had replaced the Arab-Sasanian and Arab-Byzantine issues with the aniconic

17 There is some debate over the dating of these coins, with Walker and others arguing for a reading of 63. Walker, A Catalogue of the Muhammadan Coins in the British Museum, i, p. 127. 18 Walker would have dated these coins to 682, as he read the date as 63 rather than 66. See previous note. 19 Salm’s tenure in Khurāsān actually ended sometime in the year 63 ah, but his name continued to be found on coins minted in the region until 71 ah because of the chaos caused by the Second Fitna in Khurāsān. 20 Göbl, Dokumente zur geschichte der Iranischen Hunnen in Baktrien und Indien, i, p. 189; Vondrovec, Coinage of the Iranian Huns and their Successors from Bactria to Gandhara, ii, p. 533. 21 Göbl, Dokumente zur geschichte der Iranischen Hunnen in Baktrien und Indien, i, pp. 191; Album and Goodwin, Sylloge of Islamic Coins in the Ashmolean, i, p. 40; Vondrovec, Coinage of the Iranian Huns and their Successors from Bactria to Gandhara, ii, p. 533. It is not clear why the possibility of a Yazdgirdera dating is eliminated on these grounds, as it would situate the issues of Zhulād between 697 and 700 ce, after the appearance of the fourth marginal circle on Arab-Sasanian coins but not so distant from the introduction of ʿAbd al-Malik’s aniconic reformed coinage.

triliteral coins and political authority along a contentious frontier

coinage that would be used by the Umayyad and ʿAbbāsid caliphs; in other words, the Arab-Hephthalite coins would still diverge from the imperial standards. Attempting to contextualize the Arab-Hephthalite coins while arguing for a post-Yazdgird date, Vondrovec points out that these dates correspond with the conclusion of the conquests of Qutayba b. Muslim (governor of Khurāsān 705–15) in Ṭukhāristān and Sogdiana — conquests that went as far as the Ferghana Valley. He implies that this would be a likely context for coins in the Arab-Sasanian style with the added basmala to appear in Ṭukhāristān.22 Vondrovec assumes that Qutayba had successfully conquered Gūzgān and incorporated it into the caliphate and, therefore, that the region was then paying tribute to the Umayyads and striking coins that coincided with the expectations for imperial tribute. Qutayba had successfully campaigned against Gūzgān in 710 in response to the region’s support for a revolt by the Hephthalite lord of Bādhghīs, the Nīzak Ṭarkhān,23 but if we accept the post-Yazdgird dating, these coins would not have appeared in the immediate aftermath of this campaign. Instead, they would have been struck in the five years following Qutayba’s death in 715. Therefore, if we want to give a post-Yazdgird date to the coins of Zhulād, we must link them not to the impact of Qutayba’s conquests but rather to the aftermath of his murder at the hands of his own troops and the activities of his successors, who did little to consolidate or expand upon his campaigns in Ṭukhāristān or Transoxiana. In fact, these coins would not even have appeared during the governorship of Qutayba’s immediate successor, Yazīd b. al-Muhallab (672–720), who during his second tenure in the east (715–17) turned the focus of the army of Khurāsān away from Sogdiana and Ṭukhāristān and instead led campaigns into Gurgān and Ṭabaristān on the southeastern shore of the Caspian Sea. Instead, the dates 717–20 would correspond directly with the caliphate of ʿUmar II (r. 717–20) and the rule in Khurāsān of his governor, Jarrāḥ b. ʿAbd Allāh al-Ḥakamī (r. 717–19). Jarrāḥ’s governorship is perhaps best known for his abusive taxation — continuing to collect the jizya from people even after they had converted to Islam,24 instituting a circumcision test for converts seeking tax relief, and conscripting local populations for campaigns without paying them their stipend (ʿaṭāʾ).25 Several revolts broke out in the east during Jarrāḥ’s governorship, and a group of local lords even travelled to China seeking the support of the Tang Emperor against the Umayyads, as recorded in Chinese sources.26 This group included representatives

22 Vondrovec, Coinage of the Iranian Huns and their Successors from Bactria to Gandhara, ii, p. 533. 23 On the revolt of the Nīzak Ṭarkhān, see Haug, The Eastern Frontier, pp. 119–21. When faced with Qutayba’s army, the king of Gūzgān, simply called al-Jūzjānī in our literary sources, ran for the hills, and the people of the region surrendered. Ṭabarī, Ta’rīkh al-rusul wa-l-mulūk, ii, pp. 1205–06, 1218. 24 Even though the termino­logy used in our literary sources is certainly anachronistic in this case, it is clear that it was expected that conversion would result in a lightened tax burden, regardless of what the taxes were specifically called in early eighth-century Central Asia. 25 Haug, The Eastern Frontier, pp. 125–26. 26 Chavannes, Documents sur les Tou-Kiue (Turcs) occidentaux, pp. 203–05; Bielenstein, Diplomacy and Trade in the Chinese World, pp. 329, 334.

265

26 6

ro be rt hau g

of the Bukhārkhudā Ṭughshāda (d. 739); the Ikhshīd of Samarqand, Ghūrak (r. 710–38); and the kings of Kumād and Chaghāniyān in Upper Ṭukhāristān, the latter acting as representatives of the Turkish Yabghu of Ṭukhāristān. Any alliances made between Jarrāḥ and local lords for which we have evidence were quite conditional. For example, among the Mount Mugh documents we have a letter in Arabic from the Sogdian ruler of Panjikant named Dīwāshtīch (d. 722) to Jarrāḥ in which Dīwāshtīch seeks an alliance against the Ikhshīd Ghūrak.27 This alliance was based on Dīwāshtīch’s claim to Ghūrak’s title and the authority over Sogdiana that came with it. When Ghūrak would later find a modicum of accord with the Umayyads, Dīwāshtīch, still seeking to seize Ghūrak’s throne, rebelled against the Arabs and was subsequently defeated, captured, and crucified. In short, we do not have much reason to think that Zhulād of Gūzgān would have added the basmala to coins minted in Ambīr out of a sense of loyalty or allegiance to Jarrāḥ or the empire he represented. One hypothesis regarding the Arab-Hephthalite coins is that they were struck in order to pay taxes to Arab authorities.28 Although this argument might line up well with the governorship of an aggressive tax collector such as Jarrāḥ al-Ḥakamī, we would have to question why the coins preserved the Arab-Sasanian style, including dates according to the post-Yazdgird era, rather than the aniconic style introduced by the Umayyads decades earlier along with Hijri dates.29 Given their high quality, it is likely that the dies for the Arab-Hephthalite coins were engraved in Khurāsān and then sent to Gūzgān, where the Bactrian inscriptions were possibly added. This could explain the two mint names found on the coins, Ambīr in Bactrian with Khurāsān in Pahlavi being the most common combination. But aniconic coins were minted in Khurāsān during the caliphate of ʿUmar II, most notably at the provincial capital of Marw, which is also named on some Arab-Hephthalite issues, so we may find it strange that these same mints would simultaneously produce Arab-Sasanian style dies to send to Gūzgān, especially if the purpose of the latter was to produce coinage for tribute. Of course, there were places in the east where coinage modelled on Sasanian issues continued long after the region had been incorporated into the caliphate and the imperial coinage had undergone reform. These included a band of territories stretching from Ṭabaristān to Khwārazm, Bukhārā, and, finally, eastern Sīstān.30

27 For this letter, see Frye, ‘Tarxūn-Türxün and Central Asian History’, pp. 107–14; on Dīwāshtīch, see Grenet and La Vaissière, ‘The Last Days of Panjikent’. 28 Bates, ‘Hunnic Coinage’. 29 Hijri dates written in Pahlavi had first appeared on coins minted in Baṣra in the 650s and quickly spread across the eastern part of the caliphate. Album and Goodwin, Sylloge of Islamic Coins in the Ashmolean, i, pp. 8–9. 30 On the coins of Ṭabaristān, see Malek, The Dābūyid Ispahbads and Early ʿAbbāsid Governors of Tabaristān. On the coins of Khwārazm, see Vainberg, Monety drevnego Khorezma and Fedorov, ‘Notes on the Numismatics of Ancient Khwarezm’. On the Bukhārkhudā coins of Bukhārā, see Frye, Notes on the Early Coinage of Transoxiana, pp. 24–31, and Naymark, ‘Some Observations on Bukhar Khuda Coins’, pp. 1–2. On the coins of Sīstān, see Sears, ‘The Sasanian Style Drachms of Sistan’.

triliteral coins and political authority along a contentious frontier

Although these coins were modelled on Sasanian coinage, they were purely local in their design, based not on imperial Sasanian issues but rather on local currencies. For example, the coins of the Bukhārkhudās maintained the image of Shāhānshāh Bahrām V (r. 420–38) into the Abbasid era, even if his bust became increasingly stylized over time, instead of being updated as Sasanian models changed. Their similarity with their models was rough, not the near-identical imitation we find with the Arab-Hephthalite coins. Sometimes these local issues even included the names of Abbasid governors, which implies that they were produced under some form of imperial authority. They were also, in some cases, minted alongside more standardized caliphal issues at the same mints in the same years. Even in these longer-lasting series, the purpose and use of the local issues are not always clear; scholars have suggested that they were used for local circulation, taxation, and even to export silver.31 There is one unique Arab-Hephthalite issue that should be discussed here. First published by John Walker in 1952, this coin names the governor of Khurāsān Yazīd b. al-Muhallab on the obverse along with the inscription ‫( بسم هللا العظيم‬in the name of God, the All-Powerful), both written in Kufic Arabic. The greatest divergence from standard Arab-Sasanian issues is found on the reverse, where the traditional fire altar has been replaced with a standing figure holding a spear and a sword along with the name of the mint, HWR’ or Khurāsān; the date (84) in Pahlavi; the Bactrian inscription ‘Zhu[…] Gōzgān’; and, finally, the phrase ‫ضرب جزية بالجوزجان‬, ‘struck as tribute in Gūzgān’, in Arabic.32 The crown worn by both the bust on the obverse and the standing figure on the reverse does not match the crown of any Sasanian shāhānshāh, meaning that these coins were not direct imitations of any specific Sasanian issue. They may instead mimic the silver coins minted first at Kūfa and then in Syria in the 690s, which portrayed the caliph on both sides (bust on obverse and standing on reverse) wearing a similar conical crown, but these issues had fallen out of production before Yazīd’s governorship.33 This coin is certainly dated in the Hijri calendar, as the date corresponds with Yazīd’s first governorship of Khurāsān immediately before the reign of Qutayba, during which Yazīd led a successful campaign against the Nīzak Ṭarkhān of Bādhghīs, an event Ṭabarī dates to 84 ah/703 ce.34 This coin provides us with an interesting point of comparison with the other Arab-Hephthalite issues, as it is clearly dated in connection with a known event and stems from a context in which tribute was taken from western Ṭukhāristān. The most interesting difference from the standard Arab-Sasanian

31 As Étienne de la Vaissière has shown, the Bukhārkhudā coins were meant entirely for local use, acting more like tokens than actual currency, and did not circulate outside Bukhārā. La Vaissière, Sogdian Traders, pp. 171–74. For a discussion of the role of Bukhārkhudā coins under the Sāmānids (r. 819–999), see Treadwell, ‘The Monetary History of the Bukharkhuda Dirham in Samanid Transoxiana’, pp. 24–39. 32 Walker, ‘Some New Arab-Sasanian Coins’, pp. 108–10. 33 Heidemann, ‘The Evolving Representation of the Early Islamic Empire and its Religion on Coin Imagery’, pp. 176–78. 34 Ṭabarī, Taʾrīkh al-rusul wa-l-mulūk, ii, pp. 1129–32.

267

26 8

ro be rt hau g

issues, perhaps, is the additional Arabic inscriptions that specifically emphasize the place where the coins were minted(Gūzgān) and the reason for their minting (tribute or taxation). Arabic is used here to send a message to the Arab authorities in Khurāsān: ‘We are minting these coins for you as tribute (jizya)’.

The Hephthalites and the Second Fitna If a post-Yazdgird dating does not fit so neatly with events occurring on the ground in Khurāsān and Ṭukhāristān in the years 717–20, how about a Hijri dating? As mentioned above, assuming the dates on these coins are given in the Hijri calendar, the silver issues of Zhulād would have appeared between 685 and 688, which would place these coins firmly in the middle of the Second Fitna. The Second Fitna was incredibly disruptive to Arab rule in Khurāsān, resulting in a spate of intertribal violence as ʿAbd Allāh b. Khāzim of the Banū Sulaym violently seized control of the province for himself; but the Hephthalites of Ṭukhāristān played a role, too, and the Arab-Hephthalite coins may help us make sense of their actions. In taking control of Khurāsān, Ibn Khāzim went to war first with Sulaymān b. Marthad (d. c. 683–84) and the Arab tribe of Bakr b. Wāʾil — who had been given authority over the important frontier garrison city of Marw al-Rūd along the River Murghāb and the frontiers of Ṭukhāristān, including Gūzgān, when the civil war broke out35 — and then against his erstwhile allies, the Banū Tamīm. Elsewhere, I have used competing Arab-Sasanian issues minted in Khurāsān during the Second Fitna, some of which name Ibn Khāzim as governor while others name his predecessor, Salm b. Ziyād (governor of Khurāsān 680–83), to provide a more detailed chrono­logy of the battles between Ibn Khāzim, the Bakr, and the Tamīm.36 The Arab-Hephthalite coins of Zhulād, when their dates are read according to the Hijri calendar, were an important contribution to this project. A single coin issue struck at Herat in the name of Ibn Khāzim’s son Muḥammad and dated to 67 ah/686–87 ce gives us our clearest evidence for establishing when Ibn Khāzim and his supporters took Herat from the Bakr (67 ah or earlier) and when the Tamīm subsequently took Herat from him, killing Muḥammad in the process (67 ah or later). Although these are both important moments in the battles over Khurāsān during the Second Fitna, neither is clearly dated in the chronicles. Our literary sources tell us that while Ibn Khāzim was laying siege to Herat and taking it from the Bakr, the Hephthalites (called Turks) raided deep into Khurāsān, reaching as far as Nīshāpūr.37 Ṭabarī specifies that the fighting between Ibn Khāzim and the Tamīm, which began with the Tamīm’s attack on Herat, lasted for two years,

35 Ṭabarī, Taʾrīkh al-rusul wa-l-mulūk, ii, p. 489. 36 Haug, The Eastern Frontier, pp. 101–05. 37 Yaʿqūbī, Tārīkh al-Yaʿqūbī, ii, p. 252; Balādhūrī, Kitāb Futūḥ al-buldān, p. 414; Ṭabarī, Taʾrīkh al-rusul wa-l-mulūk, ii, p. 493.

triliteral coins and political authority along a contentious frontier

likely from 67 until 69 ah (686–88 ce) according to the numismatic evidence.38 Into this conversation, let us add two coins that Göbl did not include in his study but that do appear in Walker’s catalogue. The first, Walker’s type 138, was minted in the name of Salm b. Ziyād in the year 68. On the reverse, the mint is named as Marw in Pahlavi in the normal location, but along the margin, the name Ambīr is engraved in Bactrian. Walker could not identify the Bactrian inscription and included this coin among the coins of Salm, noting that it was minted five years after he had left Khurāsān.39 The second, Walker’s type 254, is dated to the year 69 and mentions not only both mints, Marw in Pahlavi and Ambīr in Bactrian, but the name of Zhulād Gūzgān, also in Bactrian. If we read these dates according to the Hijri calendar, they correspond to the two years after Ibn Khāzim’s siege of Herat and therefore fit into the timeline of the Hephthalite raid on Nīshāpūr, the Tamīm’s attack on Herat, Ibn Khāzim’s punitive counter-attack against the Tamīm, and the two-year period of fighting between Ibn Khāzim and the Tamīm. This was a period of continued disruption to the Arab authorities in Khurāsān during which the Hephthalites were able to raid and, perhaps, even conquer and hold Marw. The only year during the decade from 63 until 73 ah (682–93 ce) from which we have no known coins minted at Marw in Ibn Khāzim’s name is 68, implying that in this year he had no control over the city or its mint. Furthermore, as the Banū Bakr b. Wāʾil had been tasked with maintaining the frontier between Khurāsān and Ṭukhāristān, their destruction at the hands of Ibn Khāzim and his supporters, which happened no later than 67 ah, would have opened the frontier up further to Hephthalite raids at a moment in which the Arabs of Khurāsān were largely focused on Herat and fighting among themselves. It could even be hypothesized that the Hephthalites were acting in alliance with the Bakr or the Tamīm, creating a distraction to split Ibn Khāzim’s forces, as there was a long history of alliances between the Hephthalites, other Hunnic groups, or the Turks, on the one hand, and certain factions within and along the frontiers of the Sasanian Empire, on the other.40 The inclusion of Salm’s name on these coins mirrors the practice in areas under the control of the Bakr and contrasts with that manifest in coins minted in territories more firmly under Ibn Khāzim’s authority. Let us now insert the Arab-Hephthalite coins that do not name Marw into this timeline. The latest date found on any of Zhulād’s coins, including those naming only Ambīr, is 69, which is also the latest date found on coins naming both Marw and Ambīr. In view of the clearer political circumstances at Marw and the manner in which these coins fit into the chrono­logy of other Arab-Sasanian coins struck at Marw that lack Bactrian inscriptions, we can confidently say that the coins naming 38 A poem attributed to the leader of the Tamīm in Khurāsān, Ḥarīsh b. Hillāl al-Qurayʿī, also emphasizes that the fighting lasted two years. Ṭabarī, Taʾrīkh al-rusul wa-l-mulūk, ii, pp. 595, 598. 39 As noted above, coins continued to be struck in Salm’s name long after he had left Khurāsān, perhaps by an anti-Ibn Khāzim party, up to the year 71 ah. 40 Haug, The Eastern Frontier, pp. 61–64. Ṭabarī reports that the Tamīm initially regrouped in Marw at the outset of the two-year battle with Ibn Khāzim, thus implying that the city was not fully under the grasp of the rebellious governor. Ṭabarī, Taʾrīkh al-rusul wa-l-mulūk, ii, p. 595.

269

270

ro be rt hau g

both Marw and Ambīr are dated according to the Hijri calendar. Because the coins of Zhulād naming only Ambīr include dates that lead up to and match the dates on the Marw/Ambīr issues (66, 68, and 69), it is safe to assume that all the coins that mention Zhulād use the same dating system and did not appear three decades later, long after the reform coinage of ʿAbd al-Malik had spread across the caliphate, as reading the dates as referring to the post-Yazdgird era would require us to believe. We should therefore think of the period 685–88 as a kind of high point in Zhulād’s career when he was able to flex his authority, symbolized by his ability to strike high-quality coins in his own name from Ambīr. We are not sure what happened to him after 688 — the Umayyads quickly regained control of Marw and Khurāsān after the conclusion of the fitna in 692 and then began their conquests of Ṭukhāristān and later Sogdiana in earnest — but his family may have continued to hold power in Gūzgān. A Zhulād Gūzgān b. Kānag, potentially the grandson of the Zhulād of the Arab-Hephthalite coins, appears in a Bactrian contract dated around 705.41 It is possible that this later Zhulād is Jūzjānī, the king Qutayba b. Muslim chased out of Gūzgān in 710. Another Zhulād Jūzjān b. al-Jūzjān appears as a witness in an Arabic document of manumission dated to Shaʿbān 160/May–June 777.42

Why Strike Coins in Three Languages? Now that we have situated the Arab-Hephthalite coins of Zhulād Gūzgān in a specific historical context, we can return to the questions set out at the beginning of this article. Why would the ‘king of the mountain’ strike coins in three languages, and who would be expected to read them? The unspoken underlying question here is why Zhulād would include the basmala on coins struck at Ambīr. The presumption is that the basmala was a symbol of allegiance with or acknowledgement of the authority of the Arab governors of Khurāsān and that the inscriptions and images found on coins represent political or ideo­logical messages. Against such assumptions, I ask: in the context of the Second Fitna when the governorship of Khurāsān was being actively and violently contested, to whom would Zhulād have been paying homage? Following this logic, Vondrovec made yet another argument for the dating of the Arab-Hephthalite coins according to the post-Yazdgird era, asserting that they could not be using the Hijri dates because they ‘were struck by the adversaries of the Arab Governors’.43 In this case, Vondrovec was arguing not only that the lords of Gūzgān were in clear opposition to the Arab governors of Khurāsān but that the adoption of a calendrical system on a coin meant the adoption of a whole political and cultural milieu. Never mind the presence of the basmala, which, one assumes, would have been equally if not more problematic

41 Sims-Williams, ‘Nouveaux documents bactriens du Guzgan’, pp. 1055–56; Sims-Williams, Iranisches Personennamenbuch, ii, fasc. 7, pp. 65–66. 42 Document 30 in Khan, Arabic Documents from Early Islamic Khurasan, pp. 155–57. 43 Vondrovec, ‘Coinage of the Nezak’, p. 185.

triliteral coins and political authority along a contentious frontier

than a Hijri date written in Pahlavi. Such assumptions should be questioned given that the dominant language of the Arab-Sasanian coins, Pahlavi, and the majority of the coins’ visual and epi­graphic programme were borrowed from Sasanian and Zoroastrian precedents. As Stefan Heidemann has recently argued while tracing the evolution of coin types in the early caliphate, coins are necessary for exchange, and this necessity can and often does trump ideo­logical concerns. Further, people respond more conservatively to changes in gold and silver coins than they do to changes in the petty coinage of daily life, tending to be hesitant towards precious metal coins that differ greatly from the coins with which they are familiar. This means that any changes to precious metal coinage must be introduced gradually.44 It also means that these changes may remain long after the initial historical impetus for the change has passed. As Jere Bacharach has explained, since coinage tends to be conservative, many elements of a new ‘coin type’ such as images, inscriptions, script, and layout, are often carried forward in time, even when the reason for the initial changes are no longer relevant, meaningful, or even remembered.45 Consequently, many aspects of coin design, including size and weight, would often endure even after other features had been modified or changed. With this in mind, it is important to note that Hephthalite coins imitated Sasanian issues well before the arrival of the Arabs. The earliest Hephthalite coins were modelled on coins issued by the Sasanian Shāhānshāh Pīrūz (r. 459–85), the Sasanian emperor at the time of the Hephthalites’ rise to power. These issues were certainly inspired by the substantial tribute Pīrūz paid the Hephthalites following his defeat and capture at the hands of his newly arrived eastern rivals.46 By the end of the sixth century, imitations of the issues of Khusrow II became more prominent among the Hephthalites, as Khusrow struck coins at an unprecedented rate and flooded the market. These same Khusrow issues would later serve as a model for the Arab-Sasanian coins. The Sasanian model was especially popular for imitation because of the high purity level of the Sasanian silver drachms, which likewise hit its peak during the reign of Khusrow II.47 In view of the tendency to conservatism when minting coins in silver, in many ways the Arab-Hephthalite issues demonstrate

44 In these instances, Heidemann is speaking directly about the inclusion of Christian symbols on ArabByzantine coins, though he does point to similar trajectories in Arab-Sasanian coinage. Heidemann, ‘The Evolving Representation of the Early Islamic Empire and its Religion on Coin Imagery’, pp. 158–61. A well-known case cited by Heidemann comes from a seventh-century Maronite Syriac chronicle in which it is reported that indigenous Christians rejected coins issued by Caliph Muʿāwiya because they lacked crosses. Palmer, The Seventh Century in the West-Syrian Chronicles, p. 32. 45 Bacharach, ‘Signs of Sovereignty’, p. 3. 46 Pīrūz would later be killed during another engagement with the Hephthalites that led to a Hephthalite occupation of Khurāsān. 47 Skaff, ‘Sasanian and Arab-Sasanian Coins from Turfan’, pp. 68–69.

27 1

272

ro be rt hau g

a longer continuation of minting practice than do the Arab-Sasanian coins.48 The unique aspects of the Arab-Hephthalite coins show that minor changes were made to give even the earliest Hephthalite imitations, those modelled on the coins of Pīrūz, a distinct identity. For example, many coins included the Bactrian inscription ηβ (the letters ēb), which have been interpreted as shorthand for ηβοδαλο, ēbodalo (‘Hephthalite’).49 These coins also feature four large dots at 2, 5, 7, and 10 o’clock around the obverse margin, which are not seen on Sasanian issues.50 This brings us to the question of use and circulation. On the one hand, the imitation of Arab-Sasanian issues implies that the Arab-Hephthalite coins were meant to circulate within the caliphate, either to be paid in taxes/tribute or to serve in trade with the Hephthalites’ western neighbours. This follows the assumption that the basmala on standard Arab-Sasanian coins was intended as a validating mark to ease the minds of Arab-Muslim warriors who were paid in these issues.51 Therefore, we may assume that the basmala was there to help the Arab-Hephthalite coins circulate within the caliphate. On the other hand, the inclusion of not only Bactrian inscriptions but also the tamghas and other countermarks found on many of the coins imply a local circulation.52 As Stephen Album and Tony Goodwin have noted, coins minted in Khurāsān during the Second Fitna — particularly those naming either Salm b. Ziyād or ʿAbd Allāh b. Khāzim — are more likely than contemporary coins struck in the west are to feature countermarks, including both Hephthalite countermarks and tamghas, and they are much less likely to have circulated outside of Greater Khurāsān, which includes Ṭukhāristān. More than 85 per cent of the coins issued by these two governors that are in the collection of the British Museum, for example, include a countermark.53 The implication is that many of the coins of Salm and Ibn Khāzim circulated primarily in the east and rarely found their way to Iraq or points west — which makes sense in the context

48 It is worth noting here that the copper issues of Zhulād (NumH 271 and 271A), both dated to 62 ah/681–82 ce, do not follow Arab-Sasanian models closely and instead appear closer to other Hunnic models, specifically the coinage of Ṣāhi Tigin of the Nīzak Huns who ruled from Kapiśā near modern Kabul. Therefore, the coinage of daily life reflected a more localized icono­graphy. 49 For the interpretation of this inscription, see Humbach, ‘The Peroz Hephthalite Coin’ and Alram, ‘A Rare Hunnish Coin Type’, pp. 150–51. 50 Vondrovec, Coinage of the Iranian Huns and their Successors from Bactria to Gandhara, ii, p. 399; Heidemann, ‘The Hephthalite Drachms Minted in Balkh’, p. 332. These four large dots predate the appearance of the large crescent and stars motif that was initially placed at 3, 6, 9, and 12 o’clock on the obverse of Sasanian coins struck during the second reign of Kawād I (r. 486–96, 498–531). During his interregnum, Kawād fled east to the Hephthalites, perhaps establishing the district of Qubādiyān near Balkh, and we therefore may draw a connection between these Hephthalite dots and the crescents and stars of later Sasanian coins, which demonstrates some icono­graphic influence flowing west, too. Rezakhani, ReOrienting the Sasanians, p. 133 n. 23. 51 Heidemann, ‘The Evolving Representation of the Early Islamic Empire and its Religion on Coin Imagery’, p. 159. 52 Bates, ‘Hunnic Coinage’. As Malek put it, ‘The purpose of countermarks is to confirm the acceptability of a coin so it can circulate in a specific region or among specific groups’; Malek, ArabSasanian Numismatics and History during the Early Islamic Period in Iran and Iraq, i, p. 379. 53 Album and Goodwin, Sylloge of Islamic Coins in the Ashmolean, i, p. 24.

triliteral coins and political authority along a contentious frontier

of the Second Fitna and the disruptions occurring across the caliphate. When it comes to eastern movement, the collapse of the Türk Khāqānate may have played a similar constricting role. The Tang Empire of China defeated the Turks in 659, breaking up the steppe empire that had dominated the space between the Iranian and Chinese worlds for a least a century and that had provided the necessary stability and security to facilitate long-distance trade.54 At the same time, it is also important to note that since the fifth century, Hephthalite coins, including those that imitated Sasanian issues and those that included Hephthalite countermarks, had also travelled far to the east.55 As mentioned above, Sasanian coins were well regarded for the purity of their silver content, and unlike many contemporary coinages that imitated Sasanian motifs, such as the local coinage of Bukhārā, both the Hephthalite and the Arab imitations maintained this high quality and were therefore desirable instruments for long-distance trade.56 Arguments focusing on circulation within the caliphate have the potential to ignore the larger trade patterns that the Hephthalites were a part of as well as the Hephthalites’ important historical role in providing the stability required to move goods and people between different imperial realms. In other words, if we emphasize the functionality of the coins over the ideo­logical message of their inscriptions, we should be thinking about the Arab-Hephthalite coins circulating in at least three distinct economic zones: the Iranian zone increasingly dominated by Arab-Sasanian coins; the Hunnic/Hephthalite zone, from which these coin designs naturally evolved; and the larger ‘Silk Road’ sphere of trans-Eurasian trade, in which Hephthalite and Sasanian coins had long played an important role. In order for these coins to continue to fulfil their economic function, they needed to be accepted across all three zones.57 This brings us to an important point. As coins were struck in imitation of each other and circulated in similar markets, influences went both ways and coin styles were part of a larger dialogue.58 Earlier, I presented and questioned the argument 54 For an exploration of the role of the Turks in securing the Sogdian trade networks across Central Asia, see Payne, ‘The Silk Road and the Iranian Political Economy in Late Antiquity’, p. 229. 55 The coins of Pīrūz, which were heavily imitated by the Hephthalites, were the first Sasanian coins to appear in China in any serious number, most likely as a result of the large tribute paid to the Hephthalites by the shāhānshāh and the Hephthalites’ role in securing the trade routes between eastern Iran and the western frontiers of China. See Thierry, ‘Sur les monnaies sassanides trouvées en Chine’, p. 115; La Vaissière, Sogdian Traders, pp. 111–12. 56 Skaff, ‘Sasanian and Arab–Sasanian Coins from Turfan’, pp. 68–69. Imitations of coinage widely accepted for transregional trade was nothing new; a prominent example is the widespread imitation of the Athenian owl, as the Athenian tetradrachm was popularly known, around the classical Mediterranean. 57 For an overview of these interactions and the role of the Sogdians in Silk Road trade and exchange, see Hansen, The Silk Road, especially Ch. 4. For the role of Sasanian coins in particular and the distinctions between tokens used exclusively for local trade and coins used for long-distance trade within the Sogdian trade networks, see La Vaissière, Sogdian Traders, pp. 171–74. 58 Here we may turn to Hayrettin Yücesoy’s discussion of inter-imperial and intra-imperial factors in language policy: Yücesoy, ‘Language of Empire’. Staying closer to the present topic, we may look at the twinned coinage reforms of the Byzantine Emperor Justinian II (r. 685–95, 705–11) and the Umayyad

27 3

2 74

ro be rt hau g

that the fourth marginal ring found around the reverse of Arab-Hephthalite coins was evidence of a post-Yazdgird dating. This argument emphasized that a Hijri date would put the Arab-Hephthalite coins earlier than the earliest known ArabSasanian coins to include a fourth marginal ring by at least three years and thus mean that it was the Arab-Sasanian coins that were following the Arab-Hephthalite coins. In examining this question, Alexander Nikitin argued in 1984 that this stylistic development originated with the series of so-called Anāhitā coins of Khusrow II. These coins — which feature a bust of the Iranian goddess Anāhitā on the reverse — were circulated among the Alchon Huns and became a model for coins produced in Arachosia, south of the Hindu Kush, in the early 680s. From there, the style apparently spread north of the Hindu Kush, perhaps as the Alchons migrated and intermingled with the Nīzak Huns and other Hephthalites, and influenced the Arab-Hephthalite coins of the late 680s.59 It may be important to note that the fourth marginal ring is present on coins struck in the name of Salm b. Ziyād, who was governor of Khurāsān before the fourth ring appeared on standard Arab-Sasanian issues. These are coins that also include the mint name Ambīr in Bactrian (αμβιρο) on the reverse but without reference to any Hephthalite or Hunnic authority.60 The fourth marginal ring does not appear on coins struck in the name of Salm or his rival ʿAbd Allāh b. Khāzim that list Ambīr as the mint in Pahlavi but not in Bactrian, leaving the possibility that the coins with the Bactrian inscription and fourth marginal ring on the reverse were struck using a Hephthalite reverse die with an Arab-Sasanian obverse die. Similar arguments against early dating for the Arab-Hephthalite coins had been made regarding the lengthy marginal legends found on their reverse until the dating of NumH 271A (62 ah/681 ce) and Walker’s type 138 (68 ah/687 ce) became readily accepted even though they predated the earliest known standard Arab-Sasanian coins using this innovation.61 It is also important to keep in mind that these types of influences were not necessarily linear. The coin of Kurmzād (NumH 272) was mentioned in passing earlier in this article. This coin was dated 39, which, if we assume the date to be a Hijri one, translates to the year 659 ce, meaning that it predated the coins of Zhulād. The major problem with this dating is that Hijri dates were not regularly used on Arab-Sasanian coins until the 40s ah/660s ce,62 and it would be incredibly surprising if this innovation had reached the distant frontiers of the caliphate before it had been widely adopted elsewhere. By contrast, reading the date according to the Yazdgird Era places Kurmzād’s coin in the year 670 ce, right at the end of the switch to Hijri dates on Arab-Sasanian coins. It would make great sense that a caliph ʿAbd al-Malik and the debates over their mutual influence on each other. See Humphreys, ‘The “War of Images” Revisited’. 59 Nikitin, ‘Monnaies d’Arachosie du haut moyen-âge’, pp. 237–38. 60 Malek, Arab-Sasanian Numismatics and History during the Early Islamic Period in Iran and Iraq, i, pp. 401–02. 61 Gyselen, ‘Two Notes on Post-Sasanian Coins’, p. 147. 62 Malek, Arab-Sasanian Numismatics and History during the Early Islamic Period in Iran and Iraq, i, p. 247.

triliteral coins and political authority along a contentious frontier

mint on the frontiers of the caliphate under the rule of a Hephthalite king would be one of the last to adopt an innovation such as this, but this dating would mean that Kurmzād’s issues predated Zhulād’s by fifteen years. The third possibility, a post-Yazdgird dating, would place Kurmzād’s coin in the year 690 ce, right after the issues of Zhulād, thereby demonstrating a continuity of minting practices in Ambīr during the final years of the Second Fitna and before caliphal authority was fully reasserted in the east. However, this would leave us with the questions of why Kurmzād changed the dating convention. Without more sources to contextualize the career of Kurmzād, we cannot be certain. So why would Zhulād mint coins in three languages? By the time the ArabHephthalite coins were struck, Arab-Sasanian coins that included the basmala as a validating expression had been in circulation for approximately forty years, having first appeared during the reign of Caliph ʿUthmān (r. 644–56) — long enough for the phrase to have become an expected component of silver drachms. We should not necessarily see the inclusion of Arabic in this case as a sign of political changes, like many have wanted to do. It could instead be a reflection of the conservative nature of precious metal coinage. Inertia had prevailed over the decades in the move from Sasanian models to Arab-Sasanian models, and Zhulād was following suit. An alternative (but related) possibility is that the growing status of Arabic as the language of empire may have lent it a cosmopolitan imperial prestige that Zhulād had hoped to invoke. Imperial Iran had long had a tradition of cosmopolitan subordination through which the empire facilitated ‘recognition, subordination, and granting of comparative status to the various human communities within and beyond the empire that its ruling elites encountered’.63 Zhulād may have been seeking just such a recognition and status with the Arabs, and an acknowledgement of their symbols of authority may have been a part of that process. If we focus on the growing pre-eminence of Arab-Sasanian coinage within both political and economic spheres, it is truly the inclusion of Bactrian that makes Zhulād’s coins unique. The appearance of countermarks on coins is often seen as a reflection of either a fiscal crisis (there is not enough money being minted and therefore older coins must be revalidated and reused) or a political crisis (the state most likely to issue coinage does not have the ability to do so). The Second Fitna opened a vacuum of political authority in which minting practices broke down not only in Khurāsān but across Iran. This development created openings for competing expressions of political authority to appear on the coins. The result can be seen in Khurāsān, with competing issues naming Salm b. Ziyād and ʿAbd Allāh b. Khāzim, respectively, appearing from the same mints in the same years. Taking into consideration the evidence of Hephthalite raids into Khurāsān at this time, we may consider Zhulād likewise trying to flex his own political authority as a legitimate issuer of coins, perhaps not necessarily laying claim to Khurāsān but instead asserting his status as a ruler capable of striking coins in his own name.

63 Payne, ‘Iranian Cosmopolitanism’, p. 211. On cosmopolitan subordination, see Lavan, Payne, and Weisweiler, ‘Cosmopolitan Politics’.

27 5

2 76

ro be rt hau g

At a time when Arab authority in Khurāsān was experiencing growing pains in the form of civil war, Zhulād may have employed the tradition of imperial cosmopolitanism to gain recognition and status for himself and Gūzgān. But who would have been expected to read the inscriptions on these coins? I assume it was not many people at all (as highlighted in the introduction, news travels slow on coins), but this is where the differing scripts become important. Although the average person in Khurāsān or Ṭukhāristān might not have been able to read Pahlavi, Arabic, and/or Bactrian, they most likely could recognize the different scripts and associate them with different political authorities thanks not only to coins but also to seals, inscriptions, and government documents including tax and sale receipts, wedding contracts, and other economic records.64 By putting his name in Bactrian, Zhulād made it clear that these coins were struck under a Hephthalite authority, and he did so with a level of conviction greater than a simple countermark. By following the (Arab-)Sasanian icono­graphic model and adopting Pahlavi inscriptions, he likewise tied his authority to the legacies of imperial Iran and his coinage to the existing transregional economic networks. Finally, by including the Arabic inscriptions, he integrated his coins with the rising political and economic power of the region. The fall of the Sasanian Empire followed by the sudden dissipation of the caliphate during the Second Fitna in the west and the dispersal of the Türk Khāqānate in the east may have provided Zhulād and the Hephthalites of Gūzgān with the impetus to assert their own authority and identity by issuing the coinage that linked these zones together, even if for only a couple of years.

64 On Bactrian documents and, especially, their seals, see Judith Lerner’s essay in this volume.

triliteral coins and political authority along a contentious frontier

Biblio­graphy Primary Sources Balādhūrī, Aḥmad b. Yaḥyā, Kitāb Futūḥ al-buldān, ed. by M. J. de Goeje (Leiden: Brill, 1866) Ṭabarī, Abū Jaʿfar Muḥammad b. Jarīr, Taʾrīkh al-rusul wa-l-mulūk, ed. by M. J. de Goeje, 3 vols (Leiden: Brill, 1879) Yaʿqūbī, Aḥmad b. Abī Yaʿqūb, Tārīkh al-Yaʿqūbī, 2 vols (Beirut: Dār Ṣadr, 2010) Secondary Works Album, Stephen, and Tony Goodwin, Sylloge of Islamic Coins in the Ashmolean, i: The PreReform Coinage of the Early Islamic Period (Oxford: Ashmolean Museum, 2002) Alram, Michael, ‘Hunnic Coinage’, in Encyclopaedia Iranica, XII/6, pp. 570–73, available online [accessed 26 July 2022] —— , ‘A Rare Hunnish Coin Type’, Silk Road Art and Archaeo­logy, 8 (2002), 149–53 Bacharach, Jere L., ‘Signs of Sovereignty: The Shahāda, Qur’anic Verses, and the Coinage of ʿAbd al-Malik’, Muqarnas, 27 (2010), 1–30 Bates, Michael, ‘Who Was Named on Abbasid Coins? What Did It Mean?’, in Iranian Numismatic Studies: A Volume in Honor of Stephen Album, ed. by Mostafa Faghfoury (Lancaster, PA: Classical Numismatic Group, 2017), pp. 89–99 Bielenstein, Hans, Diplomacy and Trade in the Chinese World, 589–1276 (Leiden: Brill, 2005) Bosworth, C. E., R. E. Darley-Doran, and G. S. P. Freeman-Greenville, ‘Sikka’, in Encyclopaedia of Islam, 2nd edn, ed. by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs (Leiden: Brill, online edn) [accessed 26 July 2022] Chavannes, Édouard, Documents sur les Tou-Kiue (Turcs) occidentaux (Paris: Librairie d’Amérique et d’Orient Adrien Maisonneuve, 1903) Cribb, Joe, ‘Money as Metaphor 2: Money Is Order’, Numismatic Chronicle, 166 (2006), 493–517 Donner, Fred, ‘Qur’ânicization of Religio-Political Discourse in the Umayyad Period’, Revue des mondes musulmans et de la Méditerranée, 129 (2011), 79–92 Fedorov, Michael, ‘Notes on the Numismatics of Ancient Khwarezm’, Revue Numismatique, 161 (2005), 163–75 Frye, Richard, Notes on the Early Coinage of Transoxiana (New York: American Numismatic Society, 1949) —— , ‘Tarxūn-Türxün and Central Asian History’, Harvard Journal of Asiatic Studies, 14 (1951), 105–29 Göbl, Robert, Dokumente zur geschichte der Iranischen Hunnen in Baktrien und Indien, (Wiesbaden: Otto Harrassowitz, 1967) Grenet, Frantz, and Étienne de La Vaissière, ‘The Last Days of Panjikent’, Silk Road Art and Archaeo­logy, 8 (2002), 155–96 Gyselen, Rika, ‘Two Notes on Post-Sasanian Coins’, in Sources pour l’histoire et la géo­graphie du monde iranien, 224–710, ed. by Rika Gyselen (Bures-sur-Yvette: Groupe pour l’Étude de la Civilization du Moyen-Orient, 2009), pp. 143–72

27 7

278

ro be rt hau g

Hansen, Valerie, The Silk Road: A New History (Oxford: Oxford University Press, 2012) Haug, Robert, The Eastern Frontier: The Limits of Empire in Late Antique and Early Medi­eval Central Asia (London: I. B. Tauris, 2019) Heidemann, Stefan, ‘The Evolving Representation of the Early Islamic Empire and its Religion on Coin Imagery’, in The Qur’ān in Context: Historical and Literary Investigations into the Qur’ānic Milieu, ed. by Angelika Neuwirth, Nicolai Sinai, and Michael Marx (Leiden: Brill, 2010), pp. 149–94 —— , ‘The Hephthalite Drachms Minted in Balkh: A Hoard, a Sequence, and a New Reading’, Numismatic Chronicle,175 (2015), 330–44 Humbach, Helmut, Baktrische Sprachdenkmäler (Wiesbaden: Otto Harrassowitz, 1966) —— , ‘The Peroz Hephthalite Coin’, in Buddiĭskie kompleksy v starom Termez, Osnovnye itogi rabot 1978–1989 gg., ed. by Boris J. Staviskiĭ (Moscow: Izdatelʹskai︠a︡ firma ‘Vostochnai︠a︡ lit-ra’ RAN 1996), pp. 209–12 Humphreys, Michael, ‘The “War of Images” Revisited: Justinian II’s Coinage Reform and the Caliphate’, Numismatic Chronicle, 173 (2013), 229–44 Khan, Geoffrey, Arabic Documents from Early Islamic Khurasan (London: Nour Foundation, 2007) La Vaissière, Étienne de, Sogdian Traders: A History, trans. by James Ward (Leiden: Brill, 2005) Lavan, Myles, Richard E. Payne, and John Weisweiler, ‘Cosmopolitan Politics’, in Cosmopolitanism and Empire: Universal Rulers, Local Elites, and Cultural Integration in the Ancient Near East and Mediterranean, ed. by Myles Lavan, Richard E. Payne, and John Weisweiler (Oxford: Oxford University Press, 2016), pp. 1–28 Malek, Hodge Mehdi, The Dābūyid Ispahbads and Early ʿAbbāsid Governors of Tabaristān: History and Numismatics (London: Royal Numismatic Society, 2004) —— , Arab-Sasanian Numismatics and History during the Early Islamic Period in Iran and Iraq, (London: Royal Numismatic Society, 2019) Naymark, Aleksandr, ‘Some Observations on Bukhar Khuda Coins’, Oriental Numismatic Society Newsletter, 160 (1999), 1–2 Nikitin, Alexandre B., ‘Monnaies d’Arachosie du haut moyen-âge’, Studia Iranica, 13 (1984), 233–39 Palmer, Andrew, The Seventh Century in the West-Syrian Chronicles (Liverpool: Liverpool University Press, 1993) Payne, Richard, ‘Iranian Cosmopolitanism: World Religions at the Sasanian Court’, in Cosmopolitanism and Empire: Universal Rulers, Local Elites, and Cultural Integration in the Ancient Near East and Mediterranean, ed. by Myles Lavan, Richard E. Payne, and John Weisweiler (Oxford: Oxford University Press, 2016), pp. 209–30 —— , ‘The Silk Road and the Iranian Political Economy in Late Antiquity: Iran, the Silk Road, and the Problem of Aristocratic Empire’, Bulletin of the School of Oriental and African Studies, 81 (2018), 227–50 Rezakhani, Khodadad, ReOrienting the Sasanians: East Iran in Late Antiquity (Edinburgh: Edinburgh University Press, 2017) Sears, Stuart, ‘The Sasanian Style Drachms of Sistan’, Yarmouk Numismatics, 11 (1999), 18–28 Sims-Williams, Nicholas, ‘Nouveaux documents bactriens du Guzgan’, in Comptes rendus des séances de l’Académie des Inscriptions et Belles-Lettres, 146 (2002), 1047–58

triliteral coins and political authority along a contentious frontier

—— , ‘The Arab-Sasanian and Arab-Hephthalite Coinage: A View from the East’, in Islamisation de l’Asie centrale: processus locaux d’acculturation du viie au xie siècle, ed. by Étienne de La Vaissière (Paris: Association pour l’Avancement des Études Iraniennes, 2008), pp. 115–30 —— , Iranisches Personennamenbuch, ii: Mitteliranische Personennamen, fasc. 7: Bactrian Personal Names (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2010) Skaff, Jonathan Karam, ‘Sasanian and Arab-Sasanian Coins from Turfan: Their Relationship to International Trade and the Local Economy’, Asia Major, 3rd ser., 11 (1998), 67–115 Thierry, François, ‘Sur les monnaies sassanides trouvées en Chine’, in Circulation des monnaies, des marchandises, et des biens, ed. by Rika Gyselen, Res Orientalis, 5 (Leuven: Peeters, 1993), pp. 89–139 Treadwell, Luke, ‘The Monetary History of the Bukharkhuda Dirham (“Black Dirham”) in Samanid Transoxiana (204–395/819–1005)’, Oriental Numismatic Society Journal, 193 (2007), 24–39 Vainberg, B. I., Monety drevnego Khorezma (Moscow: Ĭzd-vo Nauka, 1977) Vondrovec, Klaus, ‘Coinage of the Nezak’, in Coins, Art and Chrono­logy II: The First Millennium c.e. in the Indo-Iranian Borderlands, ed. by Michael Alram, Deborah Klimburg-Salter, Minoru Inaba, and Matthias Pfisterer (Vienna: Österreichische Akademie der Wissenschaften, 2010), pp. 169–90 —— , Coinage of the Iranian Huns and their Successors from Bactria to Gandhara (4th to 8th Century ce), ii (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2014) Walker, John, A Catalogue of the Muhammadan Coins in the British Museum, 2 vols (London: British Museum, 1941) —— , ‘Some New Arab-Sasanian Coins’, Numismatic Chronicle, 12 (1952), 106–10

27 9

Alison M. Vacca

Arabic and the Public Performance of Power in Early Medi­eval Armenia Several chapters in this volume have cast doubt on the famous story that ʿAbd al-Malik ordered the transfer of the dīwāns into Arabic because a Greek scribe urinated in an inkpot. If the Umayyads ran a multilingual administration — and the chapters in the first section of this volume show that they did — why did the story of the inkpot proliferate? Certainly, one reason is the potty humour of it all. However, the story also offers a deceptively comfortable narrative consistency. ʿAbd al-Malik’s coins, which served as visible cues to caliphal control, were no longer minted in Greek or Middle Persian.1 The Umayyads seem to have participated in the display of political authority by patronizing their own translation movement, like the Sasanians had before them and, more famously, the ʿAbbāsids would do after them.2 The patronage of translation projects reflected the preeminent position of Arabic under the Umayyads. It is easy to believe, then, that ʿAbd al-Malik’s reforms would also Arabicize their administrative records because Arabicization fits neatly into the narrative of Umayyad history. The story of the inkpot retains valence because of the connection between the idea of the Caliphate and the Arabic language. Even in areas of the Caliphate where very few people could read or understand Arabic, the language became a symbol of power.3





* I would like to thank Leone Pecorini Goodall, Marie Legendre, and Robert Hoyland for their comments on earlier drafts of this paper; Hasmik Hovhannisyan for helping arrange the images; and Antoine Borrut and Manuela Ceballos for their constant support and friendship throughout the workshop and editing process. 1 There are, of course, exceptions to this, in some famous examples of post-reform coins that were not exclusively in Arabic. The coins of Ifrīqiyya and al-Andalus continued to be minted in Latin and Arabic at least into the reign of Sulaymān b. ʿAbd al-Malik, while the Dābūyids in northern Iran maintained the Sasanian models for their coins. To my knowledge, the last coins to lose the bust of the Sasanian emperors were minted under the Ṭāhirids in the early 820s. 2 See chapter 2 of Saliba, Islamic Science and the Making of the European Renaissance; on the Sasanian and ʿAbbāsid angles, see Gutas, Greek Thought, Arab Culture. 3 See, for example, Wasserstein, ‘Why Did Arabic Succeed Where Greek Failed?’, p. 266, as well as Robert Haug’s chapter in this volume. For a comparable study related to Arabic and the performance of power in Egypt, see Garosi, ‘Imperial Arabic’, especially p. 17: ‘In a time in which Arabic writing was almost exclusively experienced as a manifestation of Islamic rule, the use of Arabic script in official

Alison M. Vacca is Gevork M. Avedissian Associate Professor of Armenian History and Civilization at Columbia University. Navigating Language in the Early Islamic World: Multilingualism and Language Change in the First Cen­turies of Islam, ed. by Antoine Borrut, Manuela Ceballos, and Alison M. Vacca, ISMAR 2 (Turnhout: Brepols, 2024) pp. 281–305 10.1484/M.ISMAR-EB.5.134629

FHG

282

a l i s o n  m . vacc a

This chapter traces the development of cultural associations of Arabic with the ‘public performance of power’ in a caliphal province in which Arabic never gained popular ascendance.4 As in the other regions of the Iranian cultural sphere, in Armenia Arabic never supplanted the local language. In the Marwānid period,5 Arabic appears in Armenian texts as the ‘barbarous’ (աղխատրոյզ) language of foreign rulers, but also on public objects as a visual reminder of empire.6 In this context, the visible presence of Arabic on coins and inscriptions made a statement above and beyond the literal message conveyed in the words themselves. With the rise of independent kingdoms in Armenia in the ninth century ce, Armenians selectively, albeit rarely, adopted Arabic as a vehicle to assert their own claims to power. When medi­eval Armenians employed Arabic, this choice was deliberate. We might expect Armenian princes and kings to employ Arabic to assert claims outside of Armenia, framing power in a way that was comprehensible in the broader context of ʿAbbāsid-era emirates. However, Arabic frequently appears associated with the Armenian performance of power even when the texts were aimed at Armenian audiences. The use of Arabic then becomes engrained into Armenian political practice, recast from an imperial setting into a regionally specific expression of power. Unsurprisingly, the valence of Arabic in Armenia changes dramatically over time. Given the nature and quantity of sources available on the topic, tracking such developments is admittedly a necessarily speculative process.

A Quick Caveat: Uses of Language in a Multilingual Province Humanity rarely conforms to easy maxims, and consequently, asserting that Arabic was a language of power can hardly suggest that it served that purpose exclusively. Arabic appears in many settings, from armies to the marketplace to churches. Close analysis of literary works in Arabic and Armenian suggests that historical reports moved between the two languages. In the eighth century, Łewond reported information that he had culled from Arabic sources, as he himself claims: ‘as we learned from the enemy himself ’ (որպէս ուսաք ի նոյն ինքն ի թշնամեացն).7 Balādhurī, writing in the ninth century, relied on information from an Armenian or Albanian source, supplied to him via Muslims living in Armenia and Albania.8

documents, inscriptions, architectonic complexes, and coinage guaranteed the visual distinctiveness of the imperial authority’. 4 See Geary, Language and Power in the Early Middle Ages, p. 68, on the public performance of power. 5 I have elsewhere referred to the importance of the Marwānid reforms. The use of the term ‘reforms’ has been rightly critiqued, particularly in the work of Marie Legendre (including in this volume). The policies enacted during the reign of ʿAbd al-Malik in Armenia were particularly transformative, but the term ‘reforms’ does not encapsulate regional specificities across the Caliphate. 6 Drasxanakertc‘i, Hayoc‘ patmut‘iwn, p. 98. 7 Łewond, An Armenian Futūḥ Narrative, ed. and trans. by La Porta and Vacca, based on Matenadaran 1902 (Erevan), fol. 114v. 8 Vacca, ‘The Armenian Sources of Balādhurī’s Kitāb Futūḥ al-buldān’.

arabic and the public performance of power in early medi­e val armenia

In addition to the written evidence, modern scholars have traced Arabic-Armenian engagement via orally transmitted stories, such as the ‘City of Bronze’.9 Clearly, people talked with one another, and it makes sense that they would have done so in Arabic. Tracing such conversations through Arabisms is difficult and pertains more to the study of transmission than to the performance of power; as a result, such enquiry falls outside of the purview of this paper. However much Arabic served the elite that ruled Armenia, it also helped form non-imperial networks based on trade and religion.10 People living in and crossing through medi­eval Armenia navigated a multilingual space. Iṣṭakhrī explains concisely that ‘the languages of Azerbaijan, Armenia, and Albania are Persian and Arabic, except that the people of Dabīl [Duin] and its environs speak Armenian and the language in the vicinity of Bardhaʿa [Partaw] is Albanian’ (‫ولسان اذربيجان‬ ّ ‫وارمينية والران الفارسيّة والعربيّة غير‬ ‫ان اهل دبيل وحواليها يتكلّمون باالرمينيّة ونواحى برذعة لسانهم‬ ‫)الرانيّة‬.11 Ibn Ḥawqal builds on his predecessor and similarly attests to the multiple languages current in ʿAbbāsid-era Armenia: ‫فا ّما لسان اهل اذربيجان واكثر اهل ارمينية فالفارسيّة تجمعهم والعربيّة بينهم مستعملة وقلّمن بها‬ )‫م ّمن يتكلّم بالفارسيّة ال يفهم بالعربيّة ويفصح بها من التجّار وارباب الضياع ولطوائف من (فى‬ ‫االطراف من ارمينية وما شاكلها السنة اخر يتكلّمون بها كاالرمنيّة مع اهل دبيل ونشوى ونواحيهما‬ 12.‫ويتكلّم اهل برذعة بالرانيّة‬

As for the language of the people of Azerbaijan and most of the people of Armenia: Persian is a common language and Arabic is used among them. Few of those who speak Persian there would not understand Arabic. The merchants and the lords of the estates speak it well. Groups who live on the frontiers of Armenia and similar [places] have other languages that they speak there, including Armenian among the people of Dabīl [Duin], Nashawā [Naxčawan], and their environs, while the people of Bardhaʿa [Partaw] speak Albanian. The assertion of multilingualism appears later, too, when Nāṣir-i Khusraw reports that ‘in the city of Akhlāṭ [Xlat‘] they speak three languages, Arabic, Persian, and Armenian’ (‫ تازى وپارسى وارمنى‬:‫)ودر اين شهر اخالط به سه زبان سخن گویند‬.13 All of these passages appear in the context of the economic sector: Iṣṭakhrī records the weights and types of currency accepted in Armenia in the same passage, Ibn Ḥawqal refers explicitly to merchants as Arabic-speakers in Armenia, and Nāṣir’s reference to languages appears after a passage on the permissibility of the sale of pork and wine in Armenia. This multilingualism of markets correlates very well with other descriptions of trade in Arabic, Greek, and Armenian sources. Armenia was the

9 Mkrtč‘yan, ‘Hay-arabakan banahyusakan kaperi patmut‘yunic‘’, pp. 54–60. 10 On non-imperial networks and language, see Antoine Borrut’s chapter in this volume. 11 Iṣṭakhrī, Kitāb al-Masālik wa-l-mamālik, pp. 191–92. 12 Ibn Ḥawqal, Kitāb Ṣūrat al-ʿarḍ, pp. 348–49. 13 Nāṣir-i Khusraw, Book of Travels, p. 8.

283

284

a l i s o n  m . vacc a

site of merchants and mujāhids who came from afar, strangers congregating at the nexus of empires.14 Arabic was merely one of many possible vehicles of communication at such entrepôts. The significance of Arabic in this mêlée, then, cannot rest solely on its intelligibility, since many other languages could also have served as a lingua franca. Given the widespread use of Arabic in a variety of settings and among a mix of people, its significance is manifold. This paper does not attempt to outline the significance of language in medi­eval Armenia writ large. Rather, it serves as a study of one particular role of a single language. Caliphal representatives and Armenian royalty publicly employed Arabic in specific settings to assert claims to power.

The Language of Conquest When Muslim armies entered Armenia in the seventh century, they negotiated with the Armenian elite. However, little evidence survives to indicate the prevalence of Arabic in the conquest period. We read Arabic into the conquest by examining word choice and formulae in extant treaties, as the only explicit references to conquest-era translators come from a much later source. We must extrapolate the language of negotiation because the authors of our texts either took the lingua franca for granted or did not think such information was important. Accordingly, we have no reason to think that either Armenians or Arabs invested Arabic with any particular interpretive significance in the negotiations. A number of treaties pertaining to Armenians appear in the pages of Arabic futūḥ narratives. Although these texts contain familiar phrases and provisions that generate useful interpretations about the deployment of literary formulae in ʿAbbāsid-era histories, they do not inform our understanding of the actual circumstances of the conquest of Armenia.15 The most useful record of a treaty for our purposes is in Armenian. The seventh-century historian Pseudo-Sebēos claims that Muʿāwiya — then governor of Syria — promised that no emirs would set foot on Armenian territory. T. Greenwood has rightly drawn attention to the literary character of this treaty, pointing out that it conforms to narrative expectations patterned on Armenian accounts of Sasanian-era treaties.16 While this study establishes the setting of the treaty within Armenian historio­graphy, it also reflects a new imperial context through Arabisms. Pseudo-Sebēos uses the word amiray (ամիրայ) to render the Arabic amīr. M. Jinbashian has argued that the appearance of the Arabic word in Pseudo-Sebēos suggests instead that the record of the treaty is authentic and likely translated from Arabic. According to Pseudo-Sebēos, Muʿāwiya magnanimously decided to offer to tax the Armenians

14 See Vacca, ‘The Politics of Trade in Abbasid Armenia’. 15 On these formulae, see Levy-Rubin, Non-Muslims in the Early Islamic Empire. 16 Greenwood, ‘Negotiating the North’. I would like to thank the author for providing this paper before its publication.

arabic and the public performance of power in early medi­e val armenia

‘however much you want’ (որչափ և դուք կամիջիք),17 which could potentially be a translation of the Arabic ‘according to your capacity’ (‫)على قدر طاقتكم‬, as found in other Arabic accounts of conquest treaties.18 The evidence from Pseudo-Sebēos has thus been read in a number of ways, whether to suggest that a single Arabic word made its way into an Armenian literary construction or that the Armenian source preserved an Arabic treaty from the first moments of Islamic presence in Armenia. Although logically we expect there to have been translators involved in deliberations during the conquest period, we hear next to nothing about them. Only one comparatively late conquest narrative explicitly refers to the translators who facilitated negotiations during the Islamic incursions into Armenia. The colourful history attributed to Abū ʿAbd Allāh Muḥammad b. ʿUmar al-Wāqidī includes an extensive discussion of the conquest of Akhlāṭ/Xlat‘ and Bidlīs/Bałēš. In it, PseudoWāqidī records multiple reports, each with its own line of transmission, about the language that was dominant in the negotiations. Yūqinnā, the lord of Aleppo reportedly responsible for the conquest of Bidlīs/Bałēš, sent a messenger to invite the Armenians either to convert to Islam or to submit to taxation. Pseudo-Wāqidī adds: ‘so the translator informed the [Armenian] king about what Yūqinnā had told him’ (‫)فاعلم الترجمان الملك بما قاله يوقنا‬. The storyteller then inserts a challenge on the authority of another transmitter: ‘Qudāma informed us that they did not have a translator, for Yūqinnā spoke Greek, which was the language of the people’ (‫حدثنا‬ ‫ وانما كان المتكلم يوقنا بالرومية وهو لسان القوم‬،‫)قدامة انه لم يكن بينهم ترجمان‬.19 This attempted correction is immediately quashed, however, as the storyteller puts the challenger Qudāma in his place: ‘The narrator said: someone who is most trustworthy on this [topic] informed me that they had a translator because the king was an Armenian who understood only the language of the Armenians, while Yūqinnā was a Greek who did not understand any other language’ (‫ قال كان‬.‫ حدثني من اثق به‬:‫قال الراوي‬ ‫)الترجمان بينهم الن الملك ارمني ال يفهم اال بلسان االرمن ويوقنا كان روميا ً ال يفهم لسانا ً آخر‬.20 Unlike Pseudo-Sebēos, Pseudo-Wāqidī deals with the question of language explicitly. However, also unlike Pseudo-Sebēos, Pseudo-Wāqidī is reporting centuries-old legends that reflect popular beliefs. Although Wāqidī himself lived in the ninth

17 Sebēos, Patmut‘iwn, p. 164. 18 Jinbashian, ‘The Arabo-Armenian Peace Treaty of ad 652’. The Armenian would be closer to the Arabic if it read կարիցէք instead of կամիջիք. 19 Ter-Łevondyan, ‘Ałǰik Taroni zruyc‘ə XII arabakan ałbyurum’, suggests that this may be a scribal error, mistakenly reading ‫‘( لسان القوم‬the language of the people’) for ‫‘( لسان القوة‬the language of power’). Although this emendation is certainly feasible given the quality of extant editions of Wāqidī’s text, it is unnecessary. Ter-Łevondyan proposes changing the reading because Greek could not realistically have been called the ‘language of the people’ in medi­eval Armenia. However, the transmitter corrects Qudāma specifically to tell him that his claim that Greek was the ‘language of the people’ was wrong. There is no reason to challenge Qudāma if Greek had in fact been the ‘language of the people’. This passage reflects the fact that not every transmitter knew about the language spoken in Armenia, as Qudāma likely made his claim solely on the basis of the perceived ties between Armenia and Byzantium rather than on the basis of actual knowledge of the languages current in Armenia at the time of the conquests. 20 Wāqidī, Futūḥ al-Shām, p. 163.

285

286

a l i s o n  m . vacc a

century, the only extant version of his conquest narrative was heavily redacted in the twelfth or thirteenth century.21 The redaction emphasizes the concept of religious triumphalism, the idea that the Islamic conquest spread not so much administrators and tax collectors as specialists in Islamic law and the Qurʾān. In this context, the redactors wanted Armenia to fall at the hands of Christians who had already converted, namely the Greek lord Yūqinnā and the Armenian princess Ṭāriyūn. An Arab Muslim, ʿIyāḍ b. Ghanm, appears at the end of the story, but the action remains with the Greek and Armenian Muslims. The transmitter wants local Christians to convert to Islam and lead the armies themselves, so it would make no sense for the invading forces to speak Arabic. The references to translators in this text serve the specific goals of the author or redactor, then, and cannot inform us about the negotiation of multiple languages or the use of translators in the conquest period. Because of the nature of our sources, we do not gain from the conquest narratives any sense that Arabic held symbolic value in negotiations or treaties. We are left either reading between the lines of the early sources or encountering the presumptions of later generations who projected their sense of how things should have been back onto the seventh century. It seems likely that any encounters between Armenians and Arabs during the conquests occurred in Arabic. Nevertheless, nothing in our admittedly difficult source base indicates that Arabic held a position of prestige or that Arabic was a form of the presentation of power. This conclusion is far from surprising. When Armenian elites entered into conversations with Muʿāwiya, Arabic was not unknown in Armenia, but it was not yet an imperial language in the North. The seventh-century Islamic conquests of Armenia functioned as temporary incursions to wrest resources from the North, rather than empire-building expansions. Without direct rule by caliphal representatives in Armenia, Arabic did not emerge as a Reichssprache in the wake of the conquests. Arabic only attained new significance as a visual cue to caliphal power in the North when the political situation changed in the Marwānid period.

Arabic as Caliphal Reichssprache In the Marwānid period, Arabic became a vehicle for expressing caliphal rule and a visible reminder of caliphal presence in Armenia. The Umayyad caliph ʿAbd al-Malik instigated significant changes in the patterns of caliphal rule. He dispatched his brother Muḥammad b. Marwān to serve as governor on the ground in light of an Armenian rebellion against caliphal control during the upset of the second fitna.22 The arrival of Muḥammad b. Marwān dramatically changed the status of

21 See Farsani, ‘Text und Kontext des al-Wāqidī zugeschriebenen Futūḥ aš-Šām’; Paret, ‘Die legendäre Futūḥ-Literatur, ein Arabisches Volksepos?’; and Vacca, ‘The Queen of Akhlāṭ’. 22 On Muḥammad’s response to this rebellion, see Ter-Łevondyan, ‘703 t‘vakani apstambut‘yunə Hayastantum xalifayut‘yan dem’; Vacca, ‘The Fires of Naxčawan’.

arabic and the public performance of power in early medi­e val armenia

Armenia. Pseudo-Sebēos’s treaty promised no permanent caliphal presence in Armenia; even if a literary construct, it seems to have reflected Sufyānid norms, as no governors are attested in Armenian. This arrangement clearly became defunct in the Marwānid period as administrators appeared to collect taxes and maintain the caliphal frontiers. More reliable records of the use of Arabic reflect the empire’s more direct claims over the province. Coins

As governor of the North, Muḥammad b. Marwān first issued Sasanian-style transitional dirhams that paired Middle Persian with Arabic inscriptions.23 S. Sears categorizes these into three groups:24 1. the bust of Hormizd IV with the ‘Eastern’ affirmation of faith along the margins: ‫ بسم هللا ال اله اال هللا وحده محمد رسول هللا‬25 2. the bust of Khusraw II with the phrase ‘this is allowable’: ‫جاز هذا‬26 3. the bust of Khusraw II with the name Muḥammad in reference to Muḥammad b. Marwān and the inscription ‘full [value]’: ‫واف‬ Although many of these coins do not sport recognizable mint names, Sears has demonstrated that they likely originated in the North, given the higher quantities of these specific coins in hoards found there. These coins reveal concern with administrative matters (e.g., the name of the governor), control over the markets (e.g., the acceptability of the currency), and religious claims (e.g., the prophethood 23 The dating of these coins is not firmly established. Sears places them under the governorship of Muḥammad because the shahāda was an important part of ʿAbd al-Malik’s branding, comparing them to the coins of ʿAbd al-Malik and Muḥammad’s brother Bishr from 692–93. See also Heidemann, ‘The Evolving Representation of the Early Islamic Empire’, p. 183 n. 97. 24 Sears, ‘Before Caliphal Coins’. 25 On the categories of affirmations on reform coins, see Bacharach, ‘Signs of Sovereignty’. 26 Sears claims that the second group and one type of the third group were likely minted in Dabīl/ Duin. To justify this claim, he cites the fact that Dabīl/Duin was the only city explicitly identified on Umayyad-era coins. However, the Umayyad specimen with an explicit claim to a Dabīl/Duin mint dates from 703–04. Although Łewond once refers to Dabīl/Duin as a base for Muḥammad b. Marwān, he also describes the rebuilding of the city under ʿAbd al-ʿAzīz al-Bāhilī (MS Matenadaran 1902, fol. 28r). The rebuilding of Dabīl/Duin is also attested in Balādhurī, Kitāb Futūḥ al-buldān, p. 205; Ibn Khayyāṭ, Taʾrīkh, p. 184. Dabīl/Duin may have emerged as a more significant center of caliphal administration only after the rebuilding. If so, identifying the mint for these Arab-Sasanian coins as Dabīl/Duin is anachronistic. Without further data, the suggestion thus remains speculative. For possible alternative explanations, first, we should return to the seminal article by M. Bates, which argued for a travelling mint that remained associated with the governor’s residence; Bates, ‘The Dirham Mint of the Northern Provinces’. Second, Ḥarrān/Xaṙan remains another possible location. Despite the fact that, as Sears correctly points out, Ḥarrān/Xaṙan lay quite far from Muḥammad b. Marwān’s activities in Armenia, many sources (including Łewond) identify the city as his main residence. Third, Łewond repeatedly asserts the significance of Nashawā/Naxčawan in the early Marwānid period.

287

288

a l i s o n  m . vacc a

Figure 6.1. Coin of ʿAbd al-Malik minted in Dabīl/Duin. American Numismatic Society NS 1988.132 (public domain).

of Muḥammad).27 The minting of Sasanian-style coinage with both Middle Persian and Arabic inscriptions makes a visible claim to power, circulating in the hands of the population of the North. When Arabic appears side by side with Middle Persian, the people who use these coins can see a shift in administrative norms. The Caliphate does not emerge with its own agenda; rather, it replaces the Sasanian Empire. Whereas the transitional Arab-Sasanian dirhams reflect local minting norms in their weight and style, ʿAbd al-Malik also issued reform dirhams minted in Armenia starting in 697–98 (see Fig. 6.1).28 These reform dirhams followed the patterns common across the Caliphate with the noticeable exception of the frequent addition of the word ‫ فى‬before ‫سنة‬, which is usually absent from coins outside of the North.29 On the obverse, we find the affirmation of faith in the centre: ‘There is no god except God alone who has no associate’ (‫)ال اله اال هللا وحده ال شريك له‬. The margin offers information on the minting of the coin: ‘In the name of God, this dirham was struck in Dabīl [Duin] in the year 86’ (‫)بسم هللا ضرب هذا الدرهم بدبيل سنة ست وثمنين‬. On the reverse are two partial anti-Trinitarian Qurʾānic quotations: 1. in the field: ‘God is One, God is Eternal. He neither begets nor is He begotten. He is without equal’ (‫)هللا احد هللا الصمد لم يلد ولم يولد ولم يكن له كفوا احد‬ 2. on the margins: ‘Muḥammad is the Messenger of God who sent him with guidance and the religion of truth so that he may proclaim it above all religions even though those who associate [others with God] detest it’ (‫محمد رسول هللا‬ ‫)ارسله بالهدى ودين الحق ليظهره على الدين كله ولو كره المشركون‬ The anti-Trinitarian message also appeared on Umayyad coinage in Syria, where the majority of the population remained Christian throughout the early Islamic period. The same message was subsequently used across the entire Caliphate. Accordingly, these verses on the coins minted in Armenia are not aimed specifically at Armenian Christians. However they were intended, though, the coins circulated in Armenia and so became part of the visible apparatus of the Umayyad state in Armenia. 27 Sears, ‘Before Caliphal Coins’, p. 97. 28 Salmān, ‘Aqdam dirham muʿarrab li–l-khalīfat ʿAbd al-Malik b. Marwān’. 29 Bates, ‘The Dirham Mint of the Northern Provinces’, p. 103.

arabic and the public performance of power in early medi­e val armenia

As J. Bacharach has pointed out, the common use of the terms ‘transitional’ and ‘reformed’ skews how scholars read such coins. Forearmed with the knowledge that the ʿAbbāsids would continue to mint aniconic Arabic-Islamic coinage for centuries, numismatists and historians rely on this termino­logy without adequately investigating its implications. These terms presuppose the idea that the Umayyads were on a mission to Islamicize their coins, following a neat trajectory with a clear goal. Instead, Bacharach argues, we should recognize that every change introduced into ‘reformed’ dirhams signals a conscious decision, the result of some specific historical impetus. The coins were political and religious ‘propaganda’ or ‘symbolic rhetoric’.30 For our purposes here, this argument means that removing the Middle Persian from the coins and minting them only in Arabic were deliberate decisions that served a broader purpose in the Umayyad ruling agenda. Arabic emerged as the companion of empire, even if the administration continued to function in multilingual complexity. In a province such as Armenia, these coins were the most tangible symbols of caliphal rule. Deprived of the familiar images of Sasanian and Byzantine kings, they still represented power because only the political elite minted coins. Some Armenians must have understood the content of these messages, discerning the political and religious claims of the Umayyad elite, while others read only the visual vocabulary proclaimed by the coins. For the segment of the population who could not read the Qurʾānic inscriptions, Arabic itself was the symbol of power. It replaced not only the Middle Persian but also the images of kings. Muḥammad b. Marwān’s coins reflected caliphal expectations but embodied local practices familiar in pre-conquest Armenia. By contrast, ʿAbd al-Malik’s reformed coins were spread across the Caliphate and were not specific to Armenia. The use of Arabic in both cases connected Armenia to the political claims of the Umayyads, but Muḥammad b. Marwān’s coins represent a sort of hybridity because they reflect the specificities of regional practice rather than a uniform imperial agenda. Inscriptions

The few Arabic inscriptions from the period of direct caliphal control appear more in line with Muḥammad b. Marwān’s transitional dirhams than with ʿAbd al-Malik’s caliphal coinage. Although Arabic appears in relation to the ruling elite, the content and formulae suggest that this use of Arabic was at least sometimes regionally specific. The inscriptions at Zuart‘noc‘ are perhaps the most intriguing, placed inside one of the most famous medi­eval Armenian churches (see Fig. 6.2).

30 Bacharach, in ‘Signs of Sovereignty’, uses ‘propaganda’, while Heidemann, in ‘The Evolving Representation of the Early Islamic Empire’, p. 170, refers to ‘symbolic rhetoric’.

289

290

a l i s o n  m . vacc a

Figure 6.2. Eighth-century Arabic inscriptions in the church of Zuart‘noc‘, after Xač‘atryan (public domain).

These inscriptions have been transcribed and translated as follows:31 1. ‘and he wrote in the year four | and fifty and one hundred | … in the governorship of Yazīd b. Jarad and Bjr ys (?) | b. Maryam (?)’ (‫وكتب في سنة اربع وخمسين وماية‬ ‫)في والية يزيد بن جرد بجريس بن مريم‬ 2. ‘Ilyās b. ʿĪsā al-Waddākī wrote | May God pardon | him and the entirety of his family and his house | and | … the village | he wrote in the year’ (‫كتب الياس‬ ‫القرية كتب في سنة‬...‫)الوداكي غفر هللا له واهله و‬ 3. ‘In the name of God, the beneficent, the merciful | Muḥammad b. ʿAbd Allāh b. Ḥamad’ (‫رحيم محمد بن عبد هللا بن حمد‬... ‫)بسم هللا الرحمن‬ Two other Arabic inscriptions from the eighth century have been preserved in or on churches. First, the name Muḥammad b. Sālim appears in Č‘ič‘xanavank‘. This inscription does not include a date, but it has been identified on the basis of its 31 Greenwood, ‘A Corpus of Early Medi­eval Armenian Inscriptions’, p. 89; Xač‘atryan, Divan Hayastani arabakan vimagrut‘yan, p. 48. I have changed some small details to match the transcription adopted in this chapter.

arabic and the public performance of power in early medi­e val armenia

Figure 6.3. Eighth-century Arabic inscription on the church of Aruč. Photo by Astghik Babajanyan. Reproduced with permission.

association with the ʿAbbāsid governor Muḥammad b.  Ṣūl.32 The final inscription is more firmly dated to the eighth century, and it is found on the exterior of the church at Aruč (see Fig. 6.3). It reads: ‘Ḥusayn b. Muḥammad b. | ʿĪsā al-Qalsānī (?) wrote | in the year seventy and one hundred’ (‫كتب الحسين بن محمد بن عيسى القلساني‬ ‫)سنة سبعين وماية‬.33 T. Greenwood argues that these inscriptions point to a more involved and active caliphal administration in the eighth century; he sees the appearance of Arabic in public places as a reflection of hands-on ʿAbbāsid policies.34 The reference in the first inscription to the governorship of a certain Yazīd b. Jarad adds a political dimension. Three details signal the local specificity of the religious and political messages conveyed in the inscriptions. First, the people mentioned by name do not appear in narrative records of ʿAbbāsid rule. All Arabic-language histories that contain information about caliphal rule in Armenia were composed outside of the North, usually in centres of the Islamic world. The inscriptions mention people who lived in or visited Armenia, not officials placed there by the caliph. Second, the placement of the inscriptions in churches indicates that their authors had access

32 Nalbandean, Alek‘sanean, and Miriǰanean, Č‘ič‘xanavank‘, pp. 56–57. 33 Greenwood, ‘A Corpus of Early Medi­eval Armenian Inscriptions’, p. 89; Xač‘atryan, Divan Hayastani arabakan vimagrut‘yan, p. 47. I have changed some small details to match the transcription adopted in this chapter. 34 I have argued elsewhere that these more invasive policy changes should be dated earlier. They derive from the Marwānid period, which saw the in-person meddling of Muḥammad b. Marwān. The appearance of Arabic as a symbol of caliphal rule in Armenia cannot originate in the ʿAbbāsid period, given the evidence of the coins above. See Vacca, Non-Muslim Provinces under Early Islam.

291

292

a l i s o n  m . vacc a

to (or control over?) Christian sites of worship. Perhaps centuries ago there were more inscriptions in a variety of settings, and only those in the churches survived to today. However, the very fact that someone would place an inscription inside of a church in the first place indicates that he acknowledged the religious and/or cultural significance of these sites. If so, this deliberate choice demonstrates the authors’ manoeuvring within local concepts of powerful places to make their own claims. Third, Greenwood points out that the content of the inscriptions at Zuart‘noc‘ is unusual for Arabic inscriptions. Both the author’s requested intercession for himself and his family and the synchronism of aligning the governorship of Yazīd b. Jarad with the year 154 follow patterns common in Armenian rather than Arabic inscriptions. ‘The startling conclusion is that these inscriptions represent a fusion of Arabic and Armenian elements. On the one hand, their language is Arabic, they employ hijra dates, and at least two of them invoke Allah; on the other, the synchronism and the scope of the intercession both sit very comfortably within the Armenian tradition’.35 These inscriptions, then, are much like Muḥammad b. Marwān’s transitional dirhams. They employ an imperial language in regionally specific ways. Again, this weaving of imperial and local threads reflects how claims to power work. It does little good for a foreign ruler to demand recognition in ways that his subjects do not comprehend. Therefore, effective claims to power are frequently hybridized, the result of engagement between rulers and ruled, rather than decrees from an imperial centre.36 Literary Sources

In addition to the material evidence from the period of direct caliphal rule, Armenian sources can also help us think about the role of Arabic in Armenia. These sources refer to meetings and correspondence between the Armenian elite and Arabs, whether caliphal governors or the caliph himself. Most of these references do not specify the language of exchange and so do not serve our purposes here, though it would certainly make sense for these meetings and correspondence to have been in Arabic or bilingual. However, two stories stand out as particularly important. First, the eighth-century martyro­logy of Vahan Gołt‘nac‘i attests to Vahan’s mastery of Arabic. As a four-year-old child, Vahan the son of Xosrov was taken to Damascus as a prisoner following a rebellion during the reign of ʿAbd al-Malik. He was renamed Wahhāb (Վահապ) and raised as a Muslim in Damascus. He grew up to become a scribe in the Umayyad dīwān (դիւանադպիր արքունի).37 When he decided to return to Gołt‘n, he reportedly debated this decision with the caliph ʿUmar b. ʿAbd al-ʿAzīz himself. When Vahan subsequently reverted to Christianity, 35 Greenwood, ‘A Corpus of Early Medi­eval Armenian Inscriptions’, p. 77. 36 Several examples from Flood, Objects of Translation illustrate this point quite well, though a more regionally specific example is that in Eastmond, ‘Other Encounters’. 37 ‘Vkayabanut‘iwn srboyn Vahanay Golt‘nac‘woy’, p. 20.

arabic and the public performance of power in early medi­e val armenia

he travelled to the court of the caliph Hishām b. ʿAbd al-Malik in Rusāfa/Ṙucap‘. Several officials of the court, including his former supervisor in the dīwān, tried to waylay him or to convince him to leave quietly without announcing his religious convictions. The vita of Vahan repeatedly refers to his eloquence: he was reportedly endowed ‘with brave and clever speech’ (խօսիւք քաջ և կորովի), and the caliph Hishām ‘was surprised at the eloquence and truthfulness of his words’ (զարմացաւ ընդ գեղեցկագիծ եւ ընդ ճշմարտարան դիրս բանիցն).38 Such details underscore Vahan’s abilities in religious disputation and, accordingly, his resolute yearning for martyrdom when he was clearly capable of talking his way out of a painful death. These passages are replete with hagio­graphic tropes, placing an eloquent Armenian martyr before two Umayyad caliphs. For our purposes, then, the story certainly cannot confirm or disprove knowledge of Arabic in Umayyad Armenia. Rather, the vita attests that in the eyes of a medi­eval Armenian hagio­grapher, there was a direct association between the dīwān, scribal eloquence, and caliphal power. In this particular story, Umayyad-era Armenians deploy this association in order to cast Vahan as the hero who can both integrate into this imagined setting but also rise above it. The second relevant story shares certain elements with Vahan Gołt‘nac‘i’s. It also posits both written and in-person conversations in which the Umayyads acknowledge the religious superiority of the Armenian religious elite. According to Łewond, the Armenian catholicos Sahak travelled to Ḥarrān/Xaṙan to acknowledge caliphal authority in the aftermath of an Armenian rebellion. He died of an illness before being able to argue his case before Muḥammad b. Marwān, so he left a letter for the governor. Muḥammad b. Marwān was reportedly so moved by Sahak’s eloquence and wisdom that he visited with the catholicos’s corpse and promised to fulfil his requests and grant amnesty to the Armenians. He wrote an ‘oath of the written promise’ (բան երդմանն խոստման գրոյն), which convinced the Armenians to trust caliphal governors.39 In a later record of this same event, Yovhannēs Drasxanakertc‘i claims that Sahak wrote to Ogbay (an unidentified ʿUqba) instead of Muḥammad b. Marwān. Ogbay reportedly visited Sahak after his death, and ‘he moved his hand according to their tradition, as if he were alive, and greeted him in his language, salamalēk‘’ (ձեռն նմա շարժէր ըստ օրինին իւրոց իբրև կենդանւոյ և ողջունէր զնա ի լեզու իւր՝ Սալամալէք).40 As in the first story, here the eloquence of an Armenian Christian speaks to representatives of caliphal power. Particularly in Drasxanakertc‘i’s version, Arabic holds symbolic significance in the engagement, as it appears when caliphal representatives acknowledge and express respect for Armenian culture. Both of these stories must be analysed in light of literary conventions, whether hagio­graphic or martyro­logical. They cannot be taken at face value to prove anything about knowledge of Arabic in Umayyad Armenia; in fact, both stories

38 ‘Vkayabanut‘iwn srboyn Vahanay Golt‘nac‘woy’, pp. 21 and 39. 39 Łewond, ed. and trans. by La Porta and Vacca, fol. 24r. 40 Drasxanakertc‘i, Hayoc‘ patmut‘iwn, p. 100; Asołik, Patmut‘iwn tiezerakan, p. 102.

293

2 94

a l i s o n  m . vacc a

concern Armenians who travelled to centres of caliphal power, rather than caliphal representatives exerting their power in Armenia proper. However, the hagio­graphic details in fact work to our advantage here. Our goal in this chapter is not to determine whether Armenians could speak Arabic, but to convey how people associated the Arabic language with power. Language appears in certain contexts in which the Armenian elite negotiated their relationship with caliphal power, either with the caliph himself or with his representative. Presumably, these hagio­graphical texts circulated only among Armenians. The significance of Arabic in these texts thus demonstrates that Armenians expected their elite to participate in the expression of power in Arabic.

Arabic Expressions of Power after the Rise of the Armenian Kingdoms By the end of the period of direct caliphal rule, Armenians had already adopted and adapted Arabic speech patterns pertaining to the ruling elite. We see this, for example, in the use of Arabic kunyas in ruling Armenian families in the ninth century: the son of Ašot Msaker was named Smbat Aplabas (Abū al-ʿAbbās); an Arcruni nobleman was named Apu-Pelč (Abū Balj); Ašot Arcruni was also known as Abū al-ʿAbbās, and the usurper Gagik Arcruni was Abū Marwān. There are many such examples of Armenian princes or scions of the main noble families with names that would have been familiar in Arabic. This trend continued into the tenth century, when Abusahl-Hamazasp ruled the Kingdom of Vaspurakan (953–72). The adoption of kunyas is not the only way in which members of the medi­eval Armenian elite projected their own claims to power via Arabic. Again, we can trace Arabic in Armenia through both material evidence and texts. Inscriptions

The majority of the Arabic inscriptions found in Armenia from the period of independent Armenian kingdoms display a very clear pattern. They are short, informal graffiti or personal inscriptions that list Muslim names and, at most, add one or two words to request either the mercy or forgiveness of God. None of these could be considered political statements. Rather, they retain some fragmentary traces of Muslims living in Armenia at the time. Examples of this type can be found at Jurnā/Gaṙni, Erevan, and Agarak.41 Starting in the eleventh century and continuing through the twelfth and thirteenth centuries, the length, type, and location of Arabic inscriptions in Armenia change dramatically. Inscriptions such as ‘God, in your mercy, forgive ʿAbd al-ʿAzīz al-Kindī’ (‫ )هللا اغفر برحمتك | عبد العزيز الكندى‬give way to longer, flowery dedicatory inscriptions such as this eight-line inscription at Janza/Ganjak: 41 Xač‘atryan, Divan Hayastani arabakan vimagrut‘yan, pp. 50–53, 59.

arabic and the public performance of power in early medi­e val armenia

‫| االجل شافر بن الفضل ادام‬ ‫| امر باتخاذ هذا الباب موالنا االمير السيد‬ ‫| الرحيم‬ ‫بسم هللا الرحمن‬ ‫| عمل ابرهيم بن عثمان‬ ‫| هللا ادام هللا توفيقه‬ ‫| يدى القاضي ابي الفرج محمد بن عبد‬ ‫هللا سلطانه على‬ 42.‫| الحداد سنة خمسة وخمسين واربعمائة‬ ‫بن عبدويه‬

In the name of God, the merciful | the beneficent. Our master the emir, the most glorious lord, Shāfur b. al-Faḍl — may God extend his reign — ordered the construction of this gate under the direction of the judge Abū al-Faraj Muḥammad b.  ʿAbd Allāh — may God bless him. Ibrahīm b.  ʿUthmān b. ʿAbduwayh al-Ḥaddād completed [it] in the year 455 [1063]. Other inscriptions from the latter half of the eleventh century include the dedicatory inscription of the mosque of Ani, which similarly praises the political leader (but with many more honorifics) and commemorates his involvement in erecting a visible Muslim footprint in the city.43 As we move into the twelfth and thirteenth centuries, such Arabic inscriptions multiply across Armenia on towers, fortresses, mosques, tombs, and the vaguer designation ‘buildings’ (‫)بنيان‬. Some of these inscriptions served as ‘stone charters’ with the practical effect of expressing authority over mixed populations, particularly in the urban areas of Armenia.44 Others were expressions of power and performances of princely acts: like Shāvur’s inscription above, these publicly extol the leaders with honorifics and inscribe their authority across the visible pathways in Armenian cities. Whether or not the average Armenian could read such dedicatory inscriptions in Arabic, their visual presence signalled the authority of Muslim rule. These examples show that the significance of Arabic as a language of power was inextricable from a specifically Muslim milieu. When Muslims were not in power, Arabic inscriptions in Armenia were short and personal; later, under the rule of Kurdish, Turkish, and Iranian Muslims, Arabic emerged yet again as an expression of authority. Armenians at times harnessed the potential symbolic value of Arabic in these later centuries,45 but for the most part, Armenian rulers expressed their authority in Armenian.46 One exception from the pre-Seljuk period, however, demonstrates that Armenian princes could harness the symbolic value of Arabic in their own inscriptions. At Hałbat monastery, there is a statue of the donor, Smbat Bagratuni (r. 997–990), presenting a model of the church along with his brother Gurgēn. According to A. Ter-Łevondyan, plaited into the folds of his turban is the inscription ‫شاهنشاه انه الملك‬, ‘the King of Kings of Ani, the king’ (Fig. 6.4).47 With this inscription, Smbat projects himself as a prince in a way that would have been comprehensible across the Iranian world, to Muslims and Christians alike. First, he employs the title shāhanshāh in Persian, rather than the Armenian 42 Xač‘atryan, Divan Hayastani arabakan vimagrut‘yan, p. 80. 43 Xač‘atryan, Divan Hayastani arabakan vimagrut‘yan, pp. 53–54. 44 For an excellent discussion of the ways in which the language of inscriptions (and, in particular, multi­ lingual inscriptions) lends a ‘public presence’ in Ani, see Eastmond, ‘Inscriptions and Authority in Ani’. 45 Ter-Łevondyan, ‘Zak‘ariayi ev Ivanei arabataṙ arjanagrut‘yunə Anberdum’. 46 Rapti, ‘Displaying the Word’. 47 Ter-Łevondyan, ‘Hałbati araberen arjanagrut‘yunə ev Bagratuni t‘agavorneri titłosnerə’.

295

296

a l i s o n  m . vacc a

Figure 6.4. Smbat Bagratuni on the walls of the monastery at Hałbat. Photo by Christina Maranci. Reproduced with permission.

translation. Second, he sports a turban. Although the ʿAbbāsids frequently gifted turbans to contemporary emirs, turbans’ perceived relation to and association with Islam has been questioned. In Armenia, the turban was a symbol of political power.48 Finally, Smbat relies on Arabic to declare himself as ‘the king’. On the one hand, this inscription demonstrates that Armenians were familiar with the ways in which Iranian Muslims presented themselves as rulers and that they could adopt the latter’s trappings for themselves. On the other hand, the inscription is pleated into the folds of Smbat’s turban, barely visible. For all that Smbat’s turban inscription taps into broader Iranian and Islamic idioms of power, it is also far from clear that anyone would have been able to read it from the ground. Its audience thus remains uncertain, but perhaps it was meant to praise Smbat and so was intended for his eyes alone. For our purposes, it provides a rare example of an independent Armenian king employing Arabic to make a claim to power. Material Sources

Smbat’s turban inscription is not the only example of an Armenian prince’s harnessing Arabic for use in an official capacity as a ruler of medi­eval Armenia. The Armenian king Ašot Bagratuni came to power as prince of Armenia only a few months after the death of the caliph Mutawakkil and the subsequent disruptions 48 Jones, ‘ʿAbbasid Suzerainty in the Medi­eval Caucasus’, p. 147.

arabic and the public performance of power in early medi­e val armenia

of the so-called Decade of Anarchy. The caliph Mustaʿīn recognized him as the prince in 862 and he became king in 884.49 Ašot Bagratuni’s seal survives, a tiny oval garnet measuring only 15 mm. On this surface, his name appears skilfully inscribed in two lines of floriated Kūfic Arabic text: ‫اشوط بن سنباط‬.50 The script used for the name Ashūṭ b. Sunbāṭ shows a notable flourish. The letter alif has a ligature at the bottom, where it should be a simple vertical line. The addition of the ligature is consistent with ʿAbbāsid coins of the ninth and tenth centuries, including those minted in Armenia. So, for example, the statement of faith on Mutawakkil’s coin from Dabīl/Duin shows a ligature on the alif of the words ‫ اله‬and ‫اال‬. However, the base of the alif on the seal is forked, lending a sort of floriation. B. Krachkovskaya has identified this type of floriation of the ligature of the alif on tombstones and certain ʿAbbāsid coins from the ninth and tenth centuries. She also points to a similar floriation on the tops of letters on a dirham minted in Armenia in the name of the caliph Muhtadī in 870. In other words, although both of the names on the seal, Ašot and Smbat, were incredibly common in the medi­eval period, the decorative elements used in the seal’s script place it in the reign of Ašot I, the son of Smbat VIII the Confessor.51 The first Bagratuni king, then, demonstrably expressed his claim to power in Arabic. Presumably, his daily administration worked largely in Armenian; however, asserting his name on official documents in Arabic suggests that he participated in a broader political space where Arabic was the primary language in which rulers could stake claims outside of their immediate surroundings. Literary Sources

Ašot’s seal could have been intended for use in correspondence with non-Armenian rulers. However, references to Armenian princes using Arabic appear also in literary texts that presumably circulated mainly among Armenians. Given this intended audience, it seems that Armenians expected their elite to know Arabic. The history attributed to Šapuh Bagratuni relates a story about the Arcruni prince Derēn, probably referring to Grigor-Derenik Arcruni before he came to power (r. 857–868). He was a wasteful spendthrift who squandered the wealth of his house drinking and feasting, employing actors and musicians to please him. Eventually, he ran out of money, and only one of his servants remained. The servant came across a wounded man who was lost, but as he could not communicate with the man he brought him to Derēn, ‘for Derēn had learned the language of the tačiks’ (զի ուսեալ էր Դէրէն զլեզուն տաճկաց).52 Derēn had also learned medicine and so was able to attend to the man’s wounds, caring for him and then playing music

49 Hakobian, ‘La date de l’avènement d’Ašot’. 50 Krachkovskaya, ‘Печать Багратида Ашота С Арабской Надписью’. 51 One of the reasons we might doubt the seal’s identification is its discovery outside of Armenian territory in Alania, which — to my knowledge — has not been explained. 52 Bagratuni, ‘The Anonymous Storyteller’, pp. 204–05; Bagratuni, Patmut‘iwn ananun, p. 117.

297

298

a l i s o n  m . vacc a

for him. The stranger wove his own tale the next day: he was the king of Baghdad who had been unfairly ousted, and he was on a mission to garner the support of his family in Khorāsān so that he could retake the throne. The stranger-king had heard of an illustrious man named Derēn and hoped he could help. Derēn gifted his last possessions, his horse and his robe, to the stranger-king and hired a guide to take him to Khorāsān. After the king had regained Baghdad, Derēn visited the city but did not recognize the king as the stranger he had helped. The king of Baghdad heaped honours and gifts upon him for a full week. He brought Derēn to sit on his throne in his stead. He ordered the courtesans to make obeisance to Derēn and the merchants to offer him costly gifts. Finally, the king addressed the court: Թէ չէր լեալ սայ՝ չէի ապրիլ ընդ ժուկն, այլ դայ ինձ շնորհաւոր եղև և դորայ շնորհօք տիրացայ ես տանս Բաղտատայ և Բաբելացոց ամէնի։ Եւ ասեմ քազ, Դերեն, այս Բաբելացոց աշխարհն է և քաղաքս Բացտատ՝ քեզ լիցի ի ծառայութիւն, և ամենայն գանձք իմ և ձիան և չորեանք իմ և ամենայն ուղտեչամակք իմ քեզ լիցի ի ծառայութիւն։53

Had it not been for him, I would not have survived for a moment. But he was gracious to me, and thanks to him I recovered this house of Baghdad and all of Babylon. I tell you, Derēn, this land of the Babylonians and the city of Baghdad shall be subject to you, and all my treasures, and horses, and my mules, and all my herds of camels shall be subject to you.54 Pseudo-Šapuh’s history is replete with stunning flourishes and entertaining encounters. It is hardly realistic to expect that an impoverished Armenian prince was solely responsible for getting Maʾmūn to his relatives — for indeed, only the fourth fitna makes sense of a ninth-century king of Baghdad fleeing to find familial aid in Khorāsān. Nevertheless, as always, the history is invaluable for the detail it provides; its veracity as a ‘true’ or ‘convoluted’ record of what really happened is irrelevant here. The interesting feature of this passage is that it revolves around the expression of ruling mores. Derēn was a plucky prince who dined and drank beyond his means because his wealth was not what made him royal. He could squander it or bestow it upon worthy supplicants, just as he could regain it many times over. The prince was educated in medicine and music, an entertainer who knew how to uphold the unwritten laws of hospitality. Fluency in Arabic was one of the ways in which Derēn performed his royalty. It allowed the ‘king’ (pseudo-Maʾmūn) to beseech his help. And it later allowed Pseudo-Šapuh to position Derēn at the same level as the king of Baghdad, quite literally placing him on the caliphal throne. In this story, Arabic opened doors and allowed Baghdad to recognize the worth of Armenian royalty. The Arcruni prince was not alone in performing Arabic to undergird claims to power. Another informative example shows Arabic bolstering the claims of the Bagratuni house and similarly bridging caliphal and independent rule. Between 53 Bagratuni, Patmut‘iwn ananun, pp. 125–27. 54 Bagratuni, ‘The Anonymous Storyteller’, p. 207 [slightly changed].

arabic and the public performance of power in early medi­e val armenia

811 and 813, Theodore Abū Qurra wrote an Arabic defence of Chalcedonianism in an Epistle to the Armenians, subsequently delivered to the prince Ašot Bagratuni Msaker in Greek. In 817, Ašot arranged for a disputation between Abū Qurra and Nonnus of Nisibis. Disputing in Arabic before the Bagratuni prince, Nonnus outwitted and bested his opponent, confirming the non-Chalcedonian doctrines of the Armenian Church. Ašot’s son, Bagarat Bagratuni, subsequently commissioned Nonnus to write the Commentary on the Gospel of John in Arabic, but relying on Syriac traditions.55 This case confirms not only medi­eval Armenian participation in multilingual discourse but also the varied registers in which medi­eval Armenians engaged with Arabic literature. However, for our purposes, this anecdote is particularly significant for what it tells us about the relationship between power and language. Although Nonnus’s Commentary provides rich information the social, scholarly, and religious networks of Christians living under Islamic rule,56 it also intersected with Armenian political claims at every step along the way. Nonnus of Nisibis entered the disputation at the request of an Armenian prince, not of the catholicos or the hierarchy of the Armenian Church. He composed his Commentary by translating his sources from Syriac to Arabic at the behest of Bagarat Bagratuni, ‘a man who has acquired a very high repute and was even once named as “head of the princes” by the barbarous nation of the Hagarenes, as if thereby they considered him worthy of greater honour’ (այր յոյժ բարձրագոյնս ստացաւ զանուն, որ և պետ իշխանաց երբեմն ի բարբարոս ազգէն հագարացւոց անուանեալ, որպէս թէ այնու զնա առաւել պատուի արժանաւոր վարկուցեալ).57 Bagarat was imprisoned

in Samarrāʾ following the campaigns of Bughā in the 850s. T‘ovma Arcruni records a conversation between Bagarat and Mutawakkil that ended in Bagarat’s conversion to Islam; this conversation, set at the caliphal court in Iraq, presumably took place in Arabic.58 The translation of the Commentary from Arabic into Armenian was undertaken at the request of Bagarat’s brother, Smbat Bagratuni the Confessor, who was similarly imprisoned in Samarrāʾ but refused to convert. The translator insists that he accepted his task out of friendship and devotion to Smbat, but he also addresses Smbat directly: ‘You imposed your immediate and princely commands’ (իսկ առընթեր և զիշխանական հրամանսդ ի վերայ ածեր) to ensure the translation

55 This overview is based on Thomson’s introduction to the text in English; Nonnus, Commentary on the Gospel of Saint John. My interest in Nonnus’s work is focused on its translation from Arabic to Armenian in the ninth century. For more on the text itself, see Akinean, ‘T‘ēodoros Apikuṙa ew Nana Asori’; Bundy, ‘The Commentary of Nonnus of Nisibis on the Pro­logue of John’; Mariès, ‘Un Commentaire sur l’évangile de Saint Jean’. For Nonnus’s other texts, see Griffith, ‘The Apo­logetic Treatise of Nonnus of Nisibis’. 56 On this, see Ter-Łevondyan, ‘K‘ristonya arabnerə ev hayerə miǰnadarum’; Ter-Łevondyan, ‘Hay ev k‘ristonya arabakan matenagrakan aṙnč‘ut‘yunneri patmut‘yunic‘’. 57 Nonnus, Commentary on the Gospel of Saint John, p. 3; Nonnus, Nanayi Asorwoy vardapeti meknut‘iwn Hovhannu awetaranin, pp. 5–6. 58 Arcruni, History of the House of the Artsrunik‘, pp. 222–23. This is presumably what led Mariès, ‘Un Commentaire sur l’évangile de Saint Jean’, p. 275, to conclude that Bagarat knew Arabic well, though he cites only Č‘amč‘yan.

299

3 00

a l i s o n  m . vacc a

of the Commentary.59 The translator then places the volume in the hands of Smbat’s granddaughter Marem Bagratuni, the princess of Sīsajān/Siunik‘, who ordered that it be translated ‘a second time’ (երկիցս անգամ) before her father became king of Armenia in 884.60 Here, the translator unsurprisingly vaunts not Marem’s political power but her piety and her nurturing role as mother to four children. Such anecdotal evidence cannot speak to how widespread or well known Arabic was in medi­eval Armenia. The translator’s repeated references to his own inability to render the Arabic adequately reflect the norms of self-abasing scribal traditions rather than an actual ineptitude in the language. Instead, this story is interesting for the intersection of language, religion, and princely authority that it portrays. However, the account of the wayward commentary as it moved from Syriac into Arabic and then into Armenian projects a very different view of the Bagratunik‘ compared to the account of Derēn Arcruni’s princely antics. None of the Bagratunik‘ in this story make any claim to personal knowledge of Arabic, though we might surmise that the prince of Armenia would be able to communicate with Arabs and functionaries of the Caliphate. This supposition is particularly true of Bagarat Bagratuni. Bagarat is established in the eyes of the ‘Hagarenes’ as the political leader of Armenia, and he personally converses with the caliph himself. He cements his position in the eyes of the translator through his patronage of Nonnus and, though it, the promulgation of Christian thought. This status exists against the backdrop of his supposed conversion to Islam. Smbat Bagratuni’s command to translate the book into Armenian is likewise a princely act. Just as the Sasanians, the Umayyads, and the ʿAbbāsids displayed their princely power through their patronage of translation projects, so, too, can we see the expression of Bagratuni power through the creation and translation of the commentary. At no point does the translator intimate that the masses might learn from the erudition of Nonnus. His exuberance in extolling his patrons’ support thus obscures any possible interpretation of the translation’s purpose outside of the royal setting. Smbat performs his own devotion to Christianity while also opening a means of communication between Arabic and Armenian. Although access to Arabic appears in both anecdotes as a marker of princely authority, Derēn’s is a personal skill that expresses his participation in a shared political culture while the Bagratuni translation process instead signals the rulers’ patronage of Christian scholarship and performance of Christian deeds. Despite their dissimilarities in focus and genre, the two references to Arabic share the same political setting, designed to showcase the princely behaviours or deeds of the Armenian royalty. The role of Arabic in such stories is not just practical but symbolic. The Bagratuni and Arcruni princes can engage with and project their own political claims outside of Armenia because they have access to Arabic, but

59 Nonnus, Commentary on the Gospel of Saint John, p. 2; Nonnus, Nanayi Asorwoy vardapeti meknut‘iwn Hovhannu awetaranin, p. 4. 60 Nonnus, Commentary on the Gospel of Saint John, p. 5; Nonnus, Nanayi Asorwoy vardapeti meknut‘iwn Hovhannu awetaranin, p. 8.

arabic and the public performance of power in early medi­e val armenia

the references to their speaking Arabic are circulated in texts meant for Armenian audiences. This detail implies that Armenians expected their leaders to be able to function as elites in Arabic, outside of Armenia.

Conclusions on Power and Language The idea that language could serve to signal power was hardly anomalous in medi­eval Armenia, where the Armenian script was vaunted as a revelation from heaven itself and inscriptions asserted Armenian identity across the walls of the churches. Even though Arabic claimed a number of uses — notably, to facilitate communication between commanders in the ʿAbbāsid army, between merchants in the marketplace, and between theo­logians of various Christian sects — the most visible thread that is woven through the history of Arabic in early medi­eval Armenia is political power. R. Thomson argues that the increased use of Arabic after Armenian independence from the Caliphate was a result of the growth of cosmopolitanism in medi­eval Armenian culture. As he claims, ‘somewhat ironically, the Armenians learned more about Islamic culture and took greater interest in Arabic learning after the military control of the Caliphate over the region declined, as Armenia’s prosperity increased during the tenth century, and as Armenians developed a greater love for city life, untypical of earlier traditional society’.61 Here, he alludes to N. Garsoïan’s famous argument that cities were ‘alien’ to traditional Armenian life before the foundation of the Bagratuni capital of Ani in 961.62 Thomson’s line of reasoning suggests that Armenians’ increased interaction with Arabs led to the growth of metropolises that gathered people together, thus jumpstarting interest in Arabic. The examples we have reviewed above do not prove or disprove Thomson’s argument, but they do suggest that we qualify it. First, Arabic sources frequently refer to fluency in Arabic in cities, specifically in reference to the economic sector cited in the caveat at the start of this paper. On the surface, these references support Thomson’s theory. However, the cases of Armenians using Arabic during the period of the Armenian kingdoms do not always seem to have any clear relation to cities (e.g., the inscriptions in Zuart‘noc‘, Aruč, or Hałbat; the commentary on the Gospel of John). Second, Thomson reaches his conclusion because he looks solely for references in Armenian texts to Armenians speaking Arabic. When we expand the source base to include both written and material evidence, we find plenty of evidence of Arabic in Armenia before the rise of the Armenian kingdoms. Arabic was the Reichssprache during the period of direct caliphal control. These examples do not appear in Thomson’s study because they are material (i.e., coins and inscriptions)

61 Thomson, ‘Arabic in Armenia before the Tenth Century’, p. 706. 62 Garsoïan, ‘The Early-Medi­eval Armenian City’. This claim has been deflected by work on the later period, but it is worth reiterating that Garsoïan’s argument rested specifically on the early medi­eval period, and she made no such generalizations about the ancient period or the period of Bagratuni rule.

301

302

a l i s o n  m . vacc a

instead of textual and because they do not demonstrate that Armenians could speak or read Arabic. Many of the examples of the use of Arabic in the period of direct caliphal control instead convey the claims of non-Armenians living in Armenia. Both qualifications to Thomson’s argument demonstrate that we have to acknowledge both the identity of the speaker and that of the intended audience. Armenian adoption of Arabic did not happen in a vacuum. Bringing examples of Armenian and non-Armenian uses of Arabic to the same table demonstrates that the language had a long history in the province already. Under caliphal rule, Arabic was the symbol of imperial power, even if it was deployed in ways that were sometimes specific to Armenia (hence contributing to a hybridity of language).63 With the collapse of direct caliphal rule and the rise of the Armenian kingdoms, Armenian elites employed Arabic because it already had a special standing in the province. Armenian adoption of Arabic was practical, as it allowed the Armenian elite to converse with other emirs in the ʿAbbāsid commonwealth; but it was also symbolic, as even Armenian authors writing for Armenian audiences asserted Armenian knowledge of Arabic as part of princely etiquette. Just as the caliphs had once patronized translation projects, so, too, did the Armenian elite: the famous example of Qusṭā b. Lūqā, whom an Armenian ruler invited to translate texts from Greek into Arabic in Armenia,64 can be added to the case of Nonnus’s translated commentary on the Gospel of John. In this way, Armenian adoption of Arabic reveals the aftermath of empire. The Armenian elite relied primarily on Armenian in their administration and public performance of power, but their sporadic use of Arabic is noteworthy specifically because of the connotations left from the period of direct caliphal control. With the disintegration of caliphal rule following the so-called Decade of Anarchy, we can chart the rise of other languages, such as New Persian, at the courts of the emirates. Armenia offers a counterpoint to this development. Once they were no longer subject to caliphal rule, the Armenian elite recast Arabic to serve their own claims to the language of empire.

63 On the relationship between empire and ‘hybridity of language and discourse’, see Yücesoy, ‘Language of Empire’. 64 On Qusṭā b. Lūqā’s role in the Greek–Arabic translation movement, see Hoyland, ‘Agapius of Manbiǧ, Qusṭā ibn Lūqā and the Graeco-Roman Past’, pp. 25 ff. On Qusṭā b. Lūqā’s relationship to the ruler of Armenia, see Gabrieli, ‘Nota biobibliografica su Qusṭā ibn Lūqā’.

arabic and the public performance of power in early medi­e val armenia

Biblio­graphy Primary Sources Arcruni, T‘ovma, History of the House of the Artsrunik‘: Translation and Commentary, trans. by Robert Thomson (Detroit: Wayne State University Press, 1985) Asołik, Patmut‘iwn tiezerakan (Saint Petersburg: Tparan Skoroxodovi, 1885) Bagratuni, Šapuh, Patmut‘iwn ananun zruc‘agri karcec‘eal Šapuh Bagratuni (Erevan: Haykakan SSH Gitut‘yunneri Akademiayi Hratarakč‘ut‘yun, 1971) —— , ‘The Anonymous Storyteller (Also Known as Šapuh Bagratuni)’, Revue des études arméniennes, 21 (1989), 171–232 Balādhurī, Aḥmad b. Yaḥyā, Kitāb Futūḥ al-buldān (Leiden: Brill, 1866) Drasxanakertc‘i, Yovhannēs, Hayoc‘ patmut‘iwn (Erevan: Erevani Hamalsarani Hratarakč‘ut‘yun, 1996) Ibn Ḥawqal, Muḥammad Abū al-Qāsim, Kitāb Ṣūrat al-ʿarḍ [al-Masālik wa-l-mamālik] (Leiden: Brill, 1939) Ibn Khayyāṭ, Khalīfa, Taʾrīkh (Beirut: Dār al-Kutub al-ʿIlmiyya, 1995) Iṣṭakhrī, Abū Isḥāq Ibrāhīm b. Muḥammad, Kitāb al-Masālik wa-l-mamālik (Leiden: Brill, 1927) Łewond, An Armenian Futūḥ Narrative, ed. and trans. by Sergio La Porta and Alison Vacca (forthcoming) Nāṣir-i Khusrow, Book of Travels [Safarnāma]: A Parallel Persian-English Text, trans. by Wheeler M. Thackston (Costa Mesa, CA: Mazda, 2001) Nonnus of Nisibis, Nanayi Asorwoy vardapeti meknut‘iwn Hovhannu awetaranin (Venice: Tparan S. Łazaru, 1920) —— , Commentary on the Gospel of Saint John, trans. by Robert Thomson (Atlanta: SBL Press, 2014) ‘Ołbk‘ vasn č‘areac‘n ašxarhis Hayoc‘ ew vkayabanut‘iwn srboyn Vahanay Golt‘nac‘woy’, in Sop‘erk‘ Hayakakank‘ (Venice: Tparan Mxit‘areanc‘, 1854) Sebēos, Patmut‘iwn (Erevan: Haykakan SH Gitut‘yunneri Akademiayi Hratarakč‘ut‘yun, 1979) Wāqidī, Abū ʿAbd Allāh Muḥammad b. ʿUmar, Futūḥ al-Shām (Beirut: Dār al-Kutub alʿIlmiyya, 1997) Secondary Works Akinean, Nerses, ‘T‘ēodoros Apikuṙa ew Nana Asori Hayastani mēǰ ew Nanayi Hovhannu meknut‘ean hayerēn t‘argmanut‘iwnə’, Handēs Amsōreay, 36 (1922), 193–205, 357–68, 417–24 Bacharach, Jere, ‘Signs of Sovereignty: The Shahāda, Qurʾanic Verses, and the Coinage of ʿAbd al-Malik’, Muqarnas, 27 (2010), 1–30 Bates, Michael, ‘The Dirham Mint of the Northern Provinces of the Umayyad Caliphate’, Armenian Numismatic Journal, 15 (1989), 89–111 Bundy, David, ‘The Commentary of Nonnus of Nisibis on the Pro­logue of John’, in Actes du premier congrès international d’études arabes chrétiennes, ed. by Khalil Samir (Rome: Pontificium Institutum Studiorum Orientalium, 1982), pp. 123–33

303

3 04

a l i s o n  m . vacc a

Eastmond, Antony, ‘Inscriptions and Authority in Ani’, in Der Doppeladler: Byzanz und die Seldschuken in Anatolien vom späten 11. bis zum 13. Jahrhundert, ed. by Neslihan AustayEffenberger and Falko Daim (Mainz: Byzanz zwischen Orient und Okzident, 2014), pp. 71–84 —— , ‘Other Encounters: Popular Belief and Cultural Convergence in Anatolia and the Caucasus’, in Islam and Christianity in Medi­eval Anatolia, ed. by A. C. S. Peacock, Bruno De Nicola, and Sara Nur Yıldız (Burlington, VT: Ashgate, 2015), pp. 183–213 Farsani, Yoones Dehghani, ‘Text und Kontext des al-Wāqidī zugeschriebenen Futūḥ aš-Šām: Ein Beitrag zur Forschungsdebatte über frühe Futūḥ-Werke’ (unpublished doctoral thesis, Georg-August-Universität Göttingen, 2017) Flood, Finbarr Barry, Objects of Translation: Material Culture and medi­eval Hindu-Muslim Encounter (Princeton, NJ: Princeton University Press, 2009) Gabrieli, G., ‘Nota biobibliografica su Qusṭā ibn Lūqā’, Rendiconti della Reale Accademia dei Lincei (Classe di scienze morali, storiche e filo­logiche), 21 (1912), 341–82 Garosi, Eugenio, ‘Imperial Arabic: Some Notes on Visual Symbolism’, in Christians and Muslims in Early Islamic Egypt, ed. by Lajos Berkes (Durham, NC: American Society of Papyro­logists, 2022), pp. 13–32 Garsoïan, Nina, ‘The Early-Medi­eval Armenian City: An Alien Element?’, in Church and Culture in Early Medi­eval Armenia (London: Ashgate, 1999), pp. 67–83 Geary, Patrick J., Language and Power in the Early Middle Ages (Waltham, MA: Brandeis University Press, 2013) Greenwood, Timothy, ‘A Corpus of Early Medi­eval Armenian Inscriptions’, Dumbarton Oaks Papers, 58 (2004), 27–91 —— , ‘Negotiating the North: Armenian Perspectives on the Conquest era’, in Festschrift for Hugh Kennedy, ed. by Letitzia Osti and Maaike L. M. van Berkel (Leiden: Brill, October 2022), pp. 591–613 Griffith, Sidney, ‘The Apo­logetic Treatise of Nonnus of Nisibis’, Aram, 3.1–2 (1993), 115–38 Gutas, Dimitri, Greek Thought, Arab Culture: The Graeco-Arabic Translation Movement in Baghdad and Early ʿAbbasid Society (2nd–4th/8th–10th C.) (New York: Routledge, 1998) Hakobian, Vazgen, ‘La date de l’avènement d’Ašot, premier roi bagradite’, Revue des études arméniennes, 2 (1965), 273–82 Heidemann, Stefan, ‘The Evolving Representation of the Early Islamic Empire and its Religion on Coin Imagery’, in The Qurʾān in Context: Historical and Literary Investigations into the Qurʾānic Milieu, ed. by Angelika Neuwirth, Michael Marx, and Nicolai Sinai (Leiden: Brill, 2010), pp. 149–96 Hoyland, Robert, ‘Agapius of Manbiǧ, Qusṭā ibn Lūqā and the Graeco-Roman Past: The Beginnings of Christian Arabic and Muslim Historio­graphy’, Quaderni di Studi Arabi, 16 (2021), 7–41 Jinbashian, Manuel, ‘The Arabo-Armenian Peace Treaty of ad 652’, Haykazean Hayagitakan Handēs, 6 (1978), 169–74 Jones, Lynn, ‘ʿAbbasid Suzerainty in the Medi­eval Caucasus: Appropriation and Adaptation of Icono­graphy and Ideo­logy’, Gesta, 42.3 (2004), 143–50 Krachkovskaya, В. А., ‘Печать Багратида Ашота с арабской надписью’, in Краткие Сообщения о докладах и полевых исследованиях института истории материальной культуры XII (Moscow: Академия Наук СССР, 1946), pp. 112–17

arabic and the public performance of power in early medi­e val armenia

Levy-Rubin, Milka, Non-Muslims in the Early Islamic Empire: From Surrender to Coexistence (Cambridge: Cambridge University Press, 2011) Mariès, Louis, ‘Un commentaire sur l’évangile de Saint Jean, rédigé en arabe (circa 840) par Nonnos (Nana) de Nisibe, conservé dans une traduction arménienne (circa 856)’, Revue des études arméniennes, 1 (1921), 273–96 Mkrtč‘yan, Hasmik, ‘Hay-Arabakan banahyusakan kaperi patmut‘yunic‘’, Lraber Hasarakakan Gitut‘yunneri, 3 (1987), 54–60 Nalbandean, Arman, Tigran Alek‘sanyan, and Diana Miriǰanean, Č‘ič‘xanavank‘ (Erevan: Gasprint Hratarakč‘ut‘yun, 2015) Paret, Rudi, ‘Die Legendäre Futūḥ-Literatur, ein arabisches Volksepos?’, in La poesia epica e la sua formazione (Rome: Accademia nazionale dei Lincei, 1970), pp. 735–49 Rapti, Ioanna, ‘Displaying the Word: Words as Visual Signs in the Armenian Architectural Decoration of the Monastery of Noravank‘’, in Viewing Inscriptions in the Late Antique and Medi­eval World, ed. by Antony Eastmond (Cambridge: Cambridge University Press, 2015), pp. 187–204 Saliba, George, Islamic Science and the Making of the European Renaissance (Cambridge, MA: MIT Press, 2007) Salmān, ʿĪsā, ‘Aqdam dirham muʿarrab li–l-khalīfat ʿAbd al-Malik b. Marwān’, Sumer, 27.1/2 (1971), 147–52 Sears, Stuart D., ‘Before Caliphal Coins: Transitional Drahms of the Umayyad North’, American Journal of Numismatics, 15 (2003), 77–110 Ter-Łevondyan, Aram, ‘Zak‘ariayi ev Ivanei arabataṙ arjanagrut‘yunə Anberdum’, Patmabanasirakan Handes, 1 (1971), 185–87 —— , ‘Hay ev k‘ristonya arabakan matenagrakan aṙnč‘ut‘yunneri patmut‘yunic‘’, Ēǰmiacin, 11 (1977), 57–63 —— , ‘K‘ristonya arabnerə ev hayerə miǰnadarum’, Ēǰmiacin, 3 (1977), 34–39 —— , ‘Ałǰik Taroni zruyc‘ə XII arabakan ałbyurum’, Patma-banasirakan Handes, 3 (1978), 265–86 —— , ‘Hałbati araberen arjanagrut‘yunə ev Bagratuni t‘agavorneri titłosnerə’, Lraber Hasarakakan Gitut‘yunneri, 1 (1979), 73–80 —— , ‘703 t‘vakani apstambut‘yunə Hayastantum xalifayut‘yan dem’, Patma-banasirakan Handes, 3 (1982), 33–45 Thomson, Robert, ‘Arabic in Armenia before the Tenth Century’, in Mélanges Jean-Pierre Mahé, ed. by Aram Mardirossian, Agnès Ouzounian, and Constantin Zuckerman (Paris: Association des Amis du Centre d’Histoire et Civilisation de Byzance, 2014), pp. 691–706 Vacca, Alison, ‘The Fires of Naxčawan: In Search of Intercultural Transmission in Arabic, Armenian, Greek, and Syriac’, Le Muséon, 129.2–3 (2016), 323–62 —— , Non-Muslim Provinces under Early Islam: Islamic Rule and Iranian Legitimacy in Armenia and Caucasian Albania (Cambridge: Cambridge University Press, 2017) Wasserstein, David, ‘Why Did Arabic Succeed Where Greek Failed? Language Change in the Near East after Muhammad’, Scripta Classica Israelica, 22 (2003), 257–72 Xač‘atryan, Alexandr, Divan Hayastani arabakan vimagrut‘yan, VIII–XVI dd (Erevan: Haykakan SSH GA Hratarakč‘ut‘yun, 1987) Yücesoy, Hayrettin, ‘Language of Empire: Politics of Arabic and Persian in the Abbasid World’, Publications of the Modern Language Association of America, 130.2 (2015), 384–92

305

Judith A. Lerner

From Bactrian to Arabic Changes in Seals and Sealing Practices as Observed in the Documents from Bactria Introduction Primary to our understanding of the history and culture of Iran and Central Asia (that is, Greater Iran) from the pre-Islamic into the Islamic era are the Bactrian documents and the related Arabic documents. The first is a corpus of more than 150 legal and economic texts and letters ranging in date from the first half of the fourth century ce to the middle of the eighth, written in the cursive Greek script that expressed the Bactrian language. Beginning in the 1990s, the documents were revealed to the world, read and interpreted by Nicholas Sims-Williams.1 Sometime after they became known, another group of documents — thirty-two in all — came to light.2 Written in Arabic and datable to 755–777 ce (138–160 ah), that is, the early Abbasid period, they appear to have belonged to the same private family archive as some of the Bactrian documents: fourteen of them concern a particular Mir b. Bek and members of his family, some of whom are also mentioned in the Bactrian documents, and both sets contain many of the same place names. The Arabic documents have been translated and fully analysed by Geoffrey Khan, who has carefully reconstructed the genealogy of Mir b. Bek’s family through three generations.3

1 The legal and economic documents were first published in full in 2000 with a revised edition in 2012, and the letters and Buddhist texts appeared in 2007; an edition of plates of all the documents appeared in 2012. See Sims-Williams, Bactrian Documents from Northern Afghanistan, hereafter abbreviated as BD. Subsequently, a volume on the dates of the documents was published in 2018: Sims-Williams and de Blois, Studies in the Chrono­logy of the Bactrian Documents from Northern Afghanistan. Not all the documents published by Sims-Williams belong to the Khalili Collection. Several, also via the market, are held in other private collections but are included in the volumes. For a preliminary report on the sealings on or associated with Khalili Bactrian documents, see Lerner, ‘An Introduction to the Sealings on the Bactrian Documents in the Khalili Collection’. 2 Khan, Arabic Documents from Early Islamic Khurasan. 3 The Bek family first appears in Document T (‘Kamird-far son of Bek, the priest’) dated by SimsWilliams in BD III (2012) to June/July 700 ce; see also BD III, p. 16; BD I, pp. 98–103. It is also found

Judith A. Lerner is Research Associate at the Institute for the Study of the Ancient World (NYU) and co-editor of the Inner and Central Asian Art and Archaeo­logy series (Brepols). Dr Lerner focuses on the visual culture of pre-Islamic Iran, Bactria, and Sogdiana. Navigating Language in the Early Islamic World: Multilingualism and Language Change in the First Cen­turies of Islam, ed. by Antoine Borrut, Manuela Ceballos, and Alison M. Vacca, ISMAR 2 (Turnhout: Brepols, 2024) pp. 307–326 10.1484/M.ISMAR-EB.5.134630

FHG

3 08

j u di th a . l e rn e r

Figure 7.1. Map of the Hindu Kush and Northern Afghanistan, showing the places mentioned in the Bactrian documents (Δ). Drawn by François Ory. Sims-Williams and François de Blois 2018.

The Bactrian documents revealed a Middle Iranian language that had been imperfectly known, and they reflected more than four centuries of political, economic, and social history that heretofore had been relatively elusive. The 400-year span represented by the documents and their sealing method underscores the strength of Bactria’s Hellenistic past. Within this span, the sealings themselves illustrate the variety of cultural and ethnic elements that characterized the inhabitants of the region prior to their coming under Arab rule. With Islamization, a change occurs not only in the method used to seal documents but, even more strikingly, in the character of the seals themselves. Thus, these two sets of documents show the transition to Muslim rule in north-central/north-eastern Afghanistan (what is considered a part of early Islamic Khorasan)4 and its effect on various aspects of life, among them language, document sealing practices, and seal design. It is these last two — sealing methods and the seals themselves — that are the subject of this paper.



in U: ‘Bek son of Kamid-far’, and ‘Kamid-far son of Bek’ (dated 712–13); X: ‘Kamird-far and Bab, sons of Bek’ (dated March/April 750); Y: ‘Mir son of Bek’ (dated 771–72). See also letter je: ‘Meyam son of Bek’ (undated). See Khan, Arabic Documents from Early Islamic Khurasan, pp. 22–24, for his reconstruction of the Bek family genealogy from the Arabic documents. 4 Khan defines ‘Khorasan’ as encompassing an area from the south-eastern Caspian littoral to the Pamir and Hindu Kush mountain chains; thus, he includes the northeast of modern Iran and much of modern Afghanistan, with Khorasan’s eastern political boundary rarely extending east of Baghlan on the Kunduz River (Khan, Arabic Documents from Early Islamic Khurasan, p. 13).

fro m b act ri an to arab i c

The Bactrian Documents The evidence of the documents indicates that most were written in the principality of Rob (modern Rui) in the northern Hindukush and some were written in Guzgan, to the north and west (Fig. 7.1). Sims-Williams categorizes the documents as legal documents (contracts and receipts); similar but mostly fragmentary documents; lists and accounts; wooden tally slips; letters; fragments of ‘uncertain character’; and Buddhist texts.5 Only the legal documents (except the few fragmentary ones) and a few of the letters bear dates in an unspecified era, which Sims-Williams and de Blois refer to as the ‘Bactrian era’ and which they calculate to have begun in 223 ce.6 In addition to the documents, the corpus contains a large number of sealings that are no longer associated with the materials they had once sealed; either they were removed when the documents or parcels were opened and then stored separately, presumably for record-keeping,7 or they became detached from the documents they originally sealed.8 A number of these detached sealings suggest origins other than Rob and Guzgan, such as Balkh in Bactria9 and Samarkand in Sogdiana.10 Relevant to this essay is the manner in which the legal documents were sealed. Although two basic formats can be observed, each with a variation,11 most of the 5 Sims-Williams and de Blois, Studies in the Chrono­logy of the Bactrian Documents from Northern Afghanistan, p. 9. 6 The first document, A, is thus dated to 13 October 332 ce, and the last, Y, to June/July 771. For discussion of the dated Bactrian documents (along with related inscriptional material), see Sims-Williams and de Blois, Studies in the Chrono­logy of the Bactrian Documents from Northern Afghanistan, pp. 15–46. 7 As most likely were the sealings from Sasanian Takht-e Suleiman (Huff, ‘Techno­logical Observations on Clay Bullae from Takht-i Suleiman’) and Sogdian Kafir Kala, near Samarkand (Cazzoli and Cereti, ‘Sealings from Kafir Kala’, pp. 134–36; Begmatov and others, ‘New Discoveries from Kafir-kala’). The archival practice of storing sealings from documents was dramatically illustrated by the discovery in a large storage building at Tel Kedesh, Israel, of more than two thousand stamped lumps of clay from the Hellenistic period that had originally sealed rolled-up papyrus documents (Berlin and Herbert, ‘The Story of a Site and a Project’, pp. 26–27). 8 Many loose sealings, particularly those in the Aman ur Rahman Collection, had been fired — more accurately, burned — and the documents they sealed had, of course, been destroyed. See Lerner and Sims-Williams, Seals, Sealings and Tokens from Bactria to Gandhara. Those from the sites mentioned in note 7 had also been either deliberately fired for storage purposes (hence their preservation) or accidentally burned. Some of the loose sealings in the Khalili and other private collections were attached to sealed documents or were wrapped in already opened documents to which they may not have originally belonged. 9 See Lerner and Sims-Williams, Seals, Sealings and Tokens from Bactria to Gandhara, p. 100, AB 1.1 (Hc170), for an impression of the seal belonging to Mihr-meh, the shahrab (governor) of Balkh, of which eight other sealings are known from the site of Jiga-tepe, near Balkh in northern Afghanistan. 10 From Samarkand come the seal impressions of the Kidarite Kushanshah, Uglarg, who ruled in Samarkand in the fifth century ce, bearing the legend ‘the lord Uglarg [?], the king of the Huns, the great Kushan-shah, the afshiyan of Samarkand’. See Lerner and Sims-Williams, Seals, Sealings and Tokens from Bactria to Gandhara, pp. 72–74, AA 2 (Hc008, Hc009, Hc158); Cazzoli and Cereti, ‘Sealings from Kafir Kala’, Figs 9, 30, and 31. 11 For full discussion on the various document and letter formats and their variations, see SimsWilliams, BD III, pp. 9–12. My description of the different formats and sealing protocols is based on his careful and precise wording.

309

310

j u di th a . l e rn e r

a

c

Figure 7.2. Document J (517 ce), purchase contract (Khalili collection; with permission of the Nour Foundation). a. Sealed; b. Unsealed; c. Back of the unsealed lower copy with the names of the seal-owners

b

fro m b act ri an to arab i c

legal documents were written on a single sheet in two copies. The upper copy was then rolled and secured by string(s) or leather thong(s) inserted through holes pierced in the space between the upper and lower copies, while the lower one was simply rolled or folded (Fig. 7.2a). In a variation of this method, the blank space between the two copies was slit from the right edge to the middle and the upper copy was folded once vertically before being rolled and sealed. To seal the document, pieces of clay were applied to each string- or thong-hole to secure the knot so that, were it suspected that the unsealed (lower) copy had been tampered with, the sealed (upper) copy could be opened to show the original text (Fig. 7.2b). Sometimes the names of the seal owners were written vertically downwards on the back of the lower, unsealed copy; when the document is turned 90 degrees, the seal owners’ names can be read, one above the other, with the name of the first seal owner appearing at the top (Fig. 7.2c). The other format (used for both documents and letters, though less frequently for documents) is either a single, open copy, authenticated by up to six sealings, or a double, closed copy, rolled or folded several times.12 As in the first format, documents of this type are secured by a string or thong threaded through the hole(s), a lump of moist clay applied over each knot and then stamped by the seal of the contractor or witness. In this format the document may be sealed by a single sealing or by multiple ones, each attached to a string or thong inserted through a hole near the bottom of the leather sheet. The double document is an ancient means of ensuring that no part of the document’s contents will be altered. Originating in Mesopotamia of the third millennium bce,13 it remains in use wherever clay is the principal medium for writing and has been adapted to new scripts and writing surfaces, such as papyrus and parchment. Thus, throughout the Hellenistic world, contracts and other official documents were executed in this double document format. In Hellenistic Egypt, for example, a contract or other type of document was written twice on a sheet of parchment or papyrus; the upper text (scriptura interior) was cut halfway across and then rolled and sealed, while the lower text (scriptura exterior) was left unsealed so the contents could be consulted (Fig. 7.3).14 This is the same format as that used in some of the Bactrian double documents previously described.15

12 Document R, dated not earlier than 671 ce (Sims-Williams and de Blois, Studies in the Chrono­logy of the Bactrian Documents from Northern Afghanistan, pp. 63–64). 13 The earliest such documents were the hollow clay balls filled with tokens that represented counts of a particular commodity, their surfaces incised and often marked by seal impressions that apparently represented the buyers’, sellers’, or witnesses’ identities. For discussion of clay tablet envelopes, see Walker, ‘Tontafel, Tontafelhülle: A. In Mesopotamia, § 3. Envelopes’, p. 104. 14 See Vandorpe, Breaking the Seal of Secrecy, pp. 3–5 for the variations of this sealing method, and her Figs 1a–g illustrating the different variations of sealing a double contract. The specimens she cites come from Elephantine in southern Egypt and date to the late fourth and early third centuries bce; see also her ‘Seals in and on the Papyri of Greco-Roman and Byzantine Egypt’. 15 Sims-Williams, BD III, pp. 9–12.

311

312

j u di th a . l e rn e r

a

b Figure 7.3. Double document examples from Egypt. a. Contract written twice and cut halfway through; sealed with three seals after the upper rolled part has been folded over. Papyrus 6, Elephantine (late fourth–early third century bce). After Vandorpe 1995, pp. 232–33; b. Six-witness contract sealed with three clay sealings, each bearing three or four seal impressions (285/284 bce). After Vandorpe 1995, Pl. 45:1.

A variation also popular on Greek contracts involved six witnesses (called ‘six-witness contracts’, συγγραφαι εζαμαρτυροι), three or four of whom would impress their seals on one of three clay sealings, while each of their names (written in the genitive: ‘[seal] of Mister X’) was written next to the appropriate clay sealing, as on the Bactrian documents. Thus, the two contracting parties and the six witnesses have all impressed their seals on one of three clay sealings. This method evolved, so that by the end of the third century bce, the upper part of the papyrus sheet was no longer cut and the entire upper text was rolled up and sealed with three clay sealings. By the end of the following century, contractors and witnesses no longer impressed their seals on the clay; instead, a single person pressed his seal on all three clay sealings. A constant feature, however, is that as on the Bactrian documents, the name of each contractor and witness was written next to the sealing that bore that person’s imprint or would have, had he used a seal. Roman conquest of the West virtually ended this contractual practice,16 but in the Hellenized East the double document on parchment survived for close to 16 Vandorpe (‘Seals in and on the Papyri of Greco-Roman and Byzantine Egypt’, p. 239) also cites the

fro m b act ri an to arab i c

Figure 7.4. Avroman Document II. After Minns 1915, p. 22.

a millennium. Three such documents found in a cave near Avroman in Persian Kurdistan date from Parthian times.17 Two are written in Greek and the third is in Aramaic. One of the Greek parchments (‘Parchment II’), a deed for a vineyard, still has the uppermost version of its text tightly rolled up and held by string on the right side (Fig. 7.4). Two oval sealings — much abraded — remain on the document, which, based on the number of holes in the blank space between the two versions, originally bore nine seal impressions. After searching for comparable letter forms, Ellis H. Minns concluded ‘that we have […] a representative, very probably degraded, of an independent branch of Greek cursive, and it is not quite inconceivable that in some ways it is nearer to the ancestor of the vellum minuscule than is the Graeco-Egyptian cursive’.18 The dating formula in the document names the Parthian ‘King of Kings, Arsaces’, his wives, and the year of some era; Minns identifies Arsaces as Phraates IV and thus calculates that it was written in 22–21 bce.19 With this background, we return to Document J, a purchase contract written in November/December 517 (295 of the Bactrian era) in Malr/Madr, south of Rob (modern Rui), one the three regions in modern Afghanistan from which most of the Bactrian documents come (see Fig. 7.1). It has retained its original five sealings, Roman diptycha, which consisted of two wax tablets. The contract was written on the first, over which the second tablet was fitted, bearing the contract again as well as the witnesses’ signatures. 17 Minns, ‘Parchments of the Parthian Period from Avroman in Kurdistan’, and Choksy, ‘Loan and Sales Contracts in Ancient and Early Medi­eval Iran’, pp. 192 and 194. A possible earlier example may be a scrap of parchment from the second century bce (170 bce?), written in Greek and bearing the name ‘King Theos Antimachos’, its eight holes perhaps made by sealings (Rea, Senior, and Hollis, ‘A Tax Receipt from Hellenistic Bactria’, p. 268 and pl. V). 18 Minns, ‘Parchments of the Parthian Period from Avroman in Kurdistan’, p. 27. 19 Minns, ‘Parchments of the Parthian Period from Avroman in Kurdistan’, pp. 36–38.

313

31 4

j u di th a . l e rn e r

affixed to the rolled upper copy (Fig. 7.2a and 7.5a–e). From left to right, we see the seal impressions of the two contractual parties, Wyem son of Burz-mihr and Bag-re-mareg son of Bur, followed by those of three witnesses, Yatak Sibukan, Mir-Shabur Yolaban, and Bag-wanind Mashan.20 What immediately strikes the eye is that each of the contractors’ sealings is a fingernail impression of the contractor’s nail or of a seal carved to resemble one (Fig. 7.5a and 7.5b).21 Also striking is the variety of styles represented by the witnesses’ seals. The third sealing from the left depicts a male head in profile with the flattened skull associated with the Alkhan Huns, known from their coinage, seals, and metalwork from this period (Fig. 7.5c).22 The fourth sealing bears the graceful figure of a flute player, her pose an echo of figures known from the glyptic and other arts of the classical world (Fig. 7.5d).23 The fifth and last fifth sealing (on the far right of the document) shows a putto holding a ribboned ring or diadem — a frequent figure in Sasanian metalwork and seal design that was borrowed from Western art — here executed in a Hellenized Bactrian style (Fig. 7.5e).24 This mix of style and subject matter characterizes Bactria’s visual culture and demonstrates Bactria’s position as a crossroads that absorbed and reflected the political and religious changes of the region, especially the most momentous and longest-lasting, the advent of Islam. This change, though inexorable, was gradual. Bactrian Document Uu, the settlement of a dispute, was written in Lizag in the region of Guzgan, west of Balkh, in 722 ce, about seventy years after an Arab army had first come to Marv and large numbers of Arabs had begun settling in Khorasan (Figs 7.1, 7.6a, and 7.6b).25 By 736 ce (118 ah) the Umayyad governor of Marv had transferred the capital to Balkh; by 748 (130 ah) Marv had come under Abbasid control. Originally, six seals closed the document. Following centuries of convention, the names of the parties involved were written on the back at a right angle to the relevant sealing.26 The single document format is a departure from 20 Other witnesses are noted in the document but are unnamed: ‘the other freemen [and] witnesses who were there among them’, i.e., among the three who placed their seals on the document. SimsWilliams, BD I (2012), p. 54. 21 Both space and relevance preclude discussion here of this sigillo­graphic phenomenon, which characterizes the surviving contractors’ sealings on six documents, ranging in date from 517 to 747 ce. It will be treated in my forthcoming publication of the sealings on the Bactrian documents in the Khalili Collection and elsewhere. For now, I note that impressions of fingernails, made by the edge of an actual fingernail, in place of an individual’s seal are found in the ancient Near East throughout the history of cuneiform clay tablets as well as among the eleven impressions on the bulla sealing Arabic Document 11, dated 762 ce (Khan, Arabic Documents from Early Islamic Khurasan, no. 25, pp. 141–43). 22 Lerner, ‘Observations on the Typo­logy and Style of Seals and Sealings from Bactria and Greater Gandhara’; and Lerner, ‘Glyptic Art from Bactria to Gandhara’, Figs 2, 5, and 11. 23 See, for example, the flutist on a first-century ce Roman gem: Sena Chiesa, Gemme di Luni, pl. IV:26 and pp. 63–64. 24 For additional seals from the region, see Callieri, Seals and Sealings from the North-West of the Indian Subcontinent and Afghanistan, cat. 2.7 and 2.8. 25 Sims-Williams, BD III, pl. 84; Sims-Williams and de Blois, Studies in the Chrono­logy of the Bactrian Documents from Northern Afghanistan, p. 127. 26 Unfortunately, seals A and B are missing. The first seal served for the declarants, two brothers (and

fro m b act ri an to arab i c

a

b

c

d Figure 7.5. Document J sealings (Khalili collection; with permission of the Nour Foundation). a. Seal/imprint of Wyem (contractor); b. Seal/imprint of Bagre-mareg (contractor); c. Seal of Yatak Sibukan (witness); d. Seal of Mir-Shabur Yolaban (witness); e. Seal of Bag-wanind Mashan (witness).

e

315

31 6

j u di th a . l e rn e r

b

a Figure 7.6. Document Uu (722 ce/104 ah), undertaking to keep the peace (with permission of the Schøyen Collection and the Nour Foundation). a. Open document; b. Sealings attached to the lower part of the document (author’s photo­ graphs); c. Impression of a Sasanian seal (after Bibliothèque Nationale de France, 1972.1317.50; public domain).

c

fro m b act ri an to arab i c

convention: the six sealings were pressed onto the cord that secured the folded lower edge of the document, and the document was then rolled downwards to the row of sealings, where it was bound by the extra cord at the lower left. Of the remaining four sealings, two belonged to named witnesses, one represented the ‘citizens of Lizag’, and the sixth was that of an unnamed witness.27

Towards a New Sealing Protocol and New Seal Designs: The Arabic Documents To appreciate the change in sealing practice in Bactria/Islamic Khorasan, a brief description of Sasanian sealing practice is needed. Typically, official documents and goods were sealed with a lump of clay or bulla on which a number of individuals — officials, contractors, or shippers of goods, along with witnesses to the transaction — impressed their seals.28 Another mode, less common than bullae with multiple seal impressions, was the use of a single large official seal, impressed on a lump of clay that covered the thong or cord that secured the document.29 Bullae impressed with multiple intaglios continued to seal documents in post-Sasanian Iran; major evidence for this is the extensive eighth-century ‘Pahlavi archive’ from Tabaristan (Fig. 7.7).30 Returning to the Bactrian documents, we note that Y is the last precisely datable document written in the Bactrian script and language (Fig. 7.8a and 7.8b). Its Bactrian-era date corresponds to 771 ce, which falls between 154 and 155 ah. It is a singular document. As a judicial declaration, it absolves Mir son of Bek of responsibility for any debts or disputes involving his brother Bab, who had decamped from the family home, whereabouts unknown. Its format is that of an open letter with a single sealing conveying a declaration authenticated by the seal, presumably of its sender, the ruler of Kadagstan, which was situated to the north-east of Rob (Fig. 7.1). More interesting is that although it is written in Bactrian, this ruler of Kadagstan and perhaps also the letter’s recipient seem to have been converts to Islam (or to have adopted one of its tropes), because its salutation reads ‘In the name of God’ (πιδο ναμο ιεζιδασο). Along with this change in greeting, Document

27 28

29 30

their progeny) who agreed to quittance of a debt; the second was that of the god Wakhsh (the river Oxus) or, rather, of his priest, who was a witness to the agreement. Interestingly, this sealing was made by a Sasanian intaglio of ordinary quality, probably carved in the late fifth or sixth century ce and now used far from its place of origin nearly two hundred years later (see Fig. 6c for comparison). The literature on Sasanian glyptics and sealing practices is vast. Comprehensive introductions can be found in Gyselen, Sasanian Seals and Sealings in the A. Saeedi Collection (which contains an extensive biblio­graphy up to its year of publication); also, Bivar, Catalogue of the Western Asiatic Stamp Seals II; and Frye ed., Sasanian Remains from Qasr-i Abu Nasr. Lerner and Skjærvø, ‘Some Uses of Clay Bullae in Sasanian Iran’. See Weber, ‘Studies in Some Documents from the Pahlavi Archive’, with an extensive biblio­graphy on this cache of documents, now divided between the University of California at Berkeley and the Freie Universität of Berlin.

317

31 8

j u di th a . l e rn e r

Figure 7.7. Letter from the Tabaristan Archive (Doc. Tab. 28; 8th century ce). © Res Orientales, XXI, 2012, p. 85.

Y reflects an important shift in convention. As a judicial declaration with an open format, it is not truly comparable with the contracts and settlements on the earlier documents, although as an open document, it follows the single-document format of Document Uu (Fig. 7.6a). A significant difference, however, lies in how it has been sealed and the nature of the sealing: stitched to the bottom left of the document by a short thong that has been threaded through it and reinforced by a rectangular leather patch, the sealing is a thick, circular pad of clay impressed with an intaglio displaying a new kind of image and style (Fig. 7.8b).31 On the clay disc — different in size and shape from the relatively small clay lumps on which the seals of the earlier Bactrian documents were impressed — is

31 Document Y may have always been an open letter; its verso is blank and there is no record of its having been rolled. An examination of it turned up no evidence of its having been folded when it came into the Khalili Collection. See Sims-Williams, BD III, pl. 102.

fro m b act ri an to arab i c

319

b Figure 7.8. Document Y (771 ce/154–55 ah), judicial declaration (Khalili collection; with permission of the Nour Foundation). a. Open document; b. Seal of Keratonga Tonga-spara, king of Kadag.

a

the impression of a seated feline, most likely a panther, its head turned back over its shoulder, perched atop a three-lobed mountain. The animal’s stylization matches other ‘post-Sasanian/early Islamic’ depictions of felines in both glyptics and metalwork as taut and attenuated creatures.32 An example of these elongated felines, which typically have frontal heads, is the walking lion on one of the seals impressed on the bulla that sealed Arabic Document 31, written for the emancipation of a slave in 146 ah/763 ce, eight years before Document Y (Fig. 7.9a and b). Despite the apparent rejection of the human face and figure (both popular motifs in Sasanian and Bactrian glyptic art), the Iranian desire for animate decoration persisted into the early centuries of Muslim dominance, and various kinds of quadrupeds, especially felines, as well as birds embellish personal signets and decorate metalwork.33 32 In metalwork, such felines often serve as the handles of ewers produced in eastern Islamic Khorasan in the late seventh to tenth centuries ce. Important examples are a bronze ewer in the Metropolitan Museum of Art (47.100.90) and a bronze ewer attributed to Samarkand (Marshak, ‘Bronzovyj kuvshin iz Samarkanda’, p. 84, Fig. 15). I plan to treat these and similarly stylized felines in more detail in a future discussion of ‘transitional markers’ of post-Sasanian Greater Iran. 33 It is noteworthy that seventh- and eighth-century Arabic Egyptian personal seals, ‘found on many different kinds of official documents issued by the administration, in Greek, Coptic or Arabic’, as well as on personal letters and tax-demand notices, depict a variety of animals and even the human figure

3 20

j u di th a . l e rn e r

Figure 7.9. Document 31 (763 ce/146 ah), emancipation of a slave (Khalili collection; with permission of the Nour Foundation). a. Complete document; b. Seal impression of a feline on the bulla, Arabic document 31 (author’s photo­ graph, with permission of the Nour Foundation).

b

a

As can be seen from the list of legal and administrative documents that mention members of the Bek family (see Table 7.1), somewhat more than a year after Bactrian Document Y was written, a tax quittance (Arabic Document 16) was issued to Bab b. Bek (this is the same brother who, in the interim, seems to have left the family home). As one of several surviving legal documents and tax quittances issued to Bek family members beginning in July 755, it was written five years after Document X. Although both the Bactrian and the Arabic documents must have been part of a large family archive and the survival of the documents we have is fortuitous, some observations can still be made on the basis of the twenty-year gap between Bactrian Documents W (747 ce) and X (750 ce), on the one hand, and Bactrian Document Y (771 ce) and Arabic Document 16 (772 ce), on the other. The earliest tax quittance (although not the earliest Arabic document from the Bek family archive) is Document 1, dated 147 ah/764 ce. It concerns Mir b. Bek al-Bamiyani, most likely the same Mir son of Bek about whom the judicial declaration, Bactrian Document Y, would be written ten years later. The series

(Sijpesteijn, ‘Seals and Papyri from Early Islamic Egypt’, pp. 164–65; Wassiliou and Harrauer, Siegel und Papyri]).

fro m b act ri an to arab i c Table 7.1. Legal and Administrative Documents from Abbasid Bactria/Khorasan Mentioning Members of the Bek Family

BACTRIAN

W: Purchase Contract, 525 = August/September 747: Wahran and Mir, sons of Bek; “large Arab poll-tax and harvest-tax” X: Settlement of a Dispute, 527= March/April 750: Kamird-far and Bab, sons of Bek, settle with brothers, Wahran and Mir sons of Bek; Kamirdfar, Wahran and Mir are to share woman Zeran

ARABIC

29 (Inv. no. 5): Emancipation of Zeran, Ṣafar 138H = July 755: Muslim husband, Sa‘īd (name adopted by one of Mir’s brothers?) manumits Zeran and her four children 25 (Inv. no. 11): Renunciation of claim against Mīr ibnBēk, Muḥarram 145 = April 762 1 (Inv. no. 9): Tax quittance, Rabī II 147H= June 764 CE 2 (Inv. no. 3): Tax quittance, Dhū al-Qa‘da 147H= Jan 765 3 (Inv. no. 18): Tax quittance, Sha‘bān 148H=Sept 765 4 (Inv. no. 30): Tax quittance, Dhū al-Hijja 148 H=Jan 766 5 (Inv. no 13): Tax quittance, Rajab 149H=August 766 6 (Inv. no. 8): Tax quittance, ca. 150H=ca.767 7 (Inv. no. 23): Tax quittance, ca. Ṣafar 150H=March 767 8 (Inv. no. 16): Tax quittance, Jumādā I 150H=June 767 12 (Inv. no. 32): Tax quittance, Rajab 151H=July 768 9 (Inv. no 10): Tax quittance, Rabī‘151H=April 769 13 (Inv. no. 15): Tax quittance, Ṣafar 152H=Feb 769 10 (Inv. no. 28): Tax quittance, Dhū al-Qa‘da152H=Dec 769 14 (Inv. no. 27): Tax quittance, Dhū al-Qa‘da152H=Dec 769 15 (Inv. no. 4): Tax quittance, Ṣafar 154H=Jan 771

BACTRIAN

Y: Judicial Declaration, 549=June/July 771

ARABIC

11 (Inv. no. 6): Tax quittance, Ramadān 154H=August 771 16 (Inv. no. 26): Tax quittance, Dhū al-Hijja 155H=Oct 772

Bactrian source: Sims-Williams BD III (2012); Sims-Williams & de Blois (2018) Arabic source: Khan, Arabic Documents (2007)

continues after Arabic Document 16 (dated to Dhu al-Hijja 155/October 772) to Arabic Document 20, written in Shaʿban 157/June 774, also a tax quittance and the last of the known surviving Arabic documents. What is significant is that each of these tax quittances (all beginning with the formula ‘In the name of God, the Merciful and Compassionate’) employs a single sealing, attached to the bottom of the document and impressed with a single seal — that of the issuing party, the financial administrator under the authority of the local governor.34 34 See Khan, Arabic Documents from Early Islamic Khurasan, p. 19.

321

3 22

j u di th a . l e rn e r

In their format, the tax quittances differ from what we know of writing practices in Bactria, even though we have nothing comparable written by a Bactrian official for a resident within his jurisdiction. A more consequential shift, however, can be seen in the legal transactions in the Bek family archive. Like the Bactrian legal documents, the writing of these agreements, now in Arabic, required the presence of the parties involved as well as witnesses. However, in a significant change in practice and format, the Arabic legal documents were sealed with a single clay bulla on which all parties to the transaction impressed their respective personal seals. This, of course, is the Sasanian practice, adopted by Iran’s Muslim conquerors in the latter part of the seventh century ce.35 Its use in eighth-century Bactria marks a shift from the centuries-old Bactrian format for legal documents to the Sasanian one, utilized by Bactria’s new Muslim overlords36. Another major shift is the almost complete change in imagery on the seals used by the citizens of this now-Islamized region. I have already noted the seated feline on the sealing of Bactrian Document Y and its Islamic counterpart, the walking lion on Arabic Document 31. Lions and other felines had been popular motifs for seal designs in Sasanian Iran and in Bactria,37 and a particular form of feline — the taut, attenuated creature already noted in our discussion of the sealing on Bactrian Document Y — now came to characterize the felines on seals as well as those found in three-dimensional form as the handles of contemporary metal ewers.38 Also new are two astral elements: a five-pointed star (formed in

35 As remarked by Khan (Arabic Documents from Early Islamic Khurasan, p. 65), ‘the custom of witnesses to legal documents leaving their mark in a clay bulla is […] unattested in the [Egyptian] papyri’. Instead, these texts feature auto­graphed witness clauses written at the bottom; these clauses, however, were instituted only at the beginning of the third century ah (Khan, ‘The Pre-Islamic Background of Muslim Legal Formularies’, p. 201). The Sasanian legacy of a chancery system and the use of seals to register documents is succinctly discussed in Porter, Arabic and Persian Seals and Amulets in the British Museum, p. 2. Its continuation into post-Sasanian times in Iran is exemplified by the ‘Tabaristan Archive’, which has been and continues to be studied by Philippe Gignoux and Dieter Weber and by Rika Gyselen (for seal impressions) in the series Res Orientales; see, for example, Gignoux, ‘Une archive post-sassanide du Tabaristān’. The introduction to Islamic Egypt of the practice of sealing the bottom part of tax-related notes and receipts and its echoing of similar practice and seal usage in the ancient Near East have been explored comprehensively by Petra M. Sijpestein (‘Seals and Papyri from Early Islamic Egypt’ and ‘Expressing New Rule’). 36 For a discussion of later administrative developments, as observed in the Bamiyan Papers, see the forthcoming book by Arezou Azad, Warehouse Worlds: Interpreting Documents on Economic Life along the Hindukush, 1155-1221, Islamicate East: New Approaches to Eastern Islamicate Texts and History (Edinburgh: Edinburgh University Press, forthcoming). 37 For Sasanian Iran, among the numerous catalogues of Sasanian glyptics, see the references in n. 28 above; for Bactria, see Lerner and Sims-Williams, Seals, Sealings and Tokens from Bactria to Gandhara, pp. 158–62. 38 See n. 32 above, and Fig. 7.8b. For similar felines, see Bivar, Catalogue of the Western Asiatic Stamp Seals II, pl. 10: DE 1–6, DF 1–5. I plan a fuller study of the pictorial images that appear on late Sasanian/early Islamic seals, specifically the use of animate forms such as lions, the star-and-crescent motif, and the related ‘Seal of Solomon’. For now, see Buhl, ‘Some Islamic Objects in the Danish National Museum’, pl. XI, 2 and 4: two green jasper stamp seals, both incised with the figure of a lion walking to the left in impression (a departure from Sasanian glyptic convention, which shows animals

fro m b act ri an to arab i c

the way one might draw a star without removing pen from paper) with four small crescents filling the corners of a square or rectangular field (i.e., the shape of the intaglio), and a six-rayed star (the sun?) placed as a subsidiary motif to the main design or text and rendered by two overlapping — often intertwined — triangles.39 Contrary to Geoffrey Khan’s explanation of these ‘astral symbols’ as deriving from Sasanian coinage, I suggest that these pentagrams and hexagrams are versions of the ‘seal of Solomon’ (khatam Sulaymān, ‫)ختم سليمان‬, widely used on coins and seals as well as in other media.40 With this, the transition from an archival format of Hellenistic origin that survived in the Hellenized East well into the eighth century ce is complete. A similar claim might be made about the intaglios that served to seal and authenticate legal documents, except for the difference that the personal nature of seals (in contrast to seals used solely for administrative purposes) allowed their owners to showcase receptivity to a range of stylistic and icono­graphic influences, local, traditional, and foreign. The Bactrian and Arabic documents allow us to follow aspects of the development of Islamic rule in the Greater Iranian space over a span of seventeen years ( July 755/Safar 138 to October 772/Dhu al-Hijja 155) until the Islamization of the region was complete. The inclusion of Arabic documents within the Bek family’s archive of documents written in Bactrian offers a view of Islamization in this part of what was already a larger Muslim world.

facing right in impression), the first inscribed in Arabic with ‘Muhammad’ above the lion and ‘Atan’ or ‘Aban’ or ‘Anan’ below; the second includes a scorpion directly below the lion, a six-pointed star above the feline, another below its belly, and three similar stars below the scorpion. The broad frontal faces of the lions are characteristic of the early Islamic period — not only on seals but on objects in other media, such as glass and ceramics. See also Porter, Arabic and Persian Seals and Amulets in the British Museum, p. 85, no. 368, a carnelian cabochon with a reclining lion in full-face, a six-pointed star below, and a Quranic text above. 39 See Khan, Arabic Documents from Early Islamic Khurasan, p. 86, for photo­graphic details of both symbols. 40 The literature on this symbol is vast; I plan to explore in greater depth its use and its putative relation to Sasanian numismatics and sigillo­graphy. For now, I refer readers to Graham, ‘The Seven Seals of Judeo-Islamic Magic’. For a helpful discussion of ‘Solomon’s seal’ in Umayyad coinage, I am grateful to Lutz Ilisch (personal communication in St Petersburg, Russia, September 2015, and email 24 April 2017).

323

3 24

j u di th a . l e rn e r

Biblio­graphy Primary Sources Khan, Geoffrey, Arabic Documents from Early Islamic Khurasan (London: Nour Foundation and Azimuth Editions, 2007) Sims-Williams, Nicolas, Bactrian Documents from Northern Afghanistan, I: Legal and Economic Documents, Studies in the Khalili Collection, III = Corpus Inscriptionum Iranicarum, pt. II, vi: Bactrian (Oxford: Nour Foundation and Oxford University Press, 2000) —— , Bactrian Documents from Northern Afghanistan II: Letters and Buddhist Texts, Studies in the Khalili Collection, III = Corpus Inscriptionum Iranicarum, pt II: Inscriptions of the Seleucid and Parthian Periods and of Eastern Iran and Central Asia, iii: Bactrian (London: Nour Foundation and Azimuth Editions, 2007) —— , Bactrian Documents from Northern Afghanistan, I: Legal and Economic Documents, rev. edn, Studies in the Khalili Collection, III = Corpus Inscriptionum Iranicarum, pt II, vi: Bactrian (London: Nour Foundation and Azimuth Editions, 2012) —— , Bactrian Documents from Northern Afghanistan III: Plates, Studies in the Khalili Collection, 3 = Corpus Inscriptionum Iranicarum, pt II: Inscriptions of the Seleucid and Parthian Periods and of Eastern Iran and Central Asia, vi: Bactrian (London: Nour Foundation and Azimuth Editions, 2012) Secondary Works Begmatov, Alisher, and others, ‘New Discoveries from Kafir-kala: Coins, Sealings, and Wooden Carvings’, Acta Asiatica, 119 (2020), 1–20 Berlin, Andrea, and Sharon Herbert, ‘The Story of a Site and a Project: Excavating Tel Kedesh’, Archaeo­logy, 65.3 (2012), 24–29 Bivar, A. D. H., Catalogue of the Western Asiatic Stamp Seals II: The Sassanian Dynasty (London: British Museum, 1969) Buhl, Marie-Louise, ‘Some Islamic Objects in the Danish National Museum’, Berytus, 11 (1953), 61–65 Callieri, Pierfrancesco, Seals and Sealings from the North-West of the Indian Subcontinent and Afghanistan (4th century bc–11th century ad): Local, Indian, Sasanian, Graeco-Persian, Sogdian, Roman (Naples: Istituto Universitario Orientale di Napoli, 1997) Cazzoli, Sara, and Carlo G. Cereti, ‘Sealings from Kafir Kala: Preliminary Report’, Ancient Civilizations from Scythia to Siberia, 11.1–2 (2005), 133–64 Choksy, Jamsheed K., ‘Loan and Sales Contracts in Ancient and Medi­eval Iran’, IndoIranian Journal, 31 (1988), 191–218 Frye, Richard N., ed., Sasanian Remains from Qasr-i Abu Nasr: Seals, Sealings, and Coins (Cambridge, MA: Harvard University Press, 1973) Gignoux, Philippe, ‘Une archive post-sassanide du Tabaristān (I)’, in Objets et documents inscrits en Pārsīg, ed. by Rika Gyselen, Res Orientales, 21 (Bures-sur-Yvette: Groupe du Moyen-Orient, 2012), pp. 29–96

fro m b act ri an to arab i c

Graham, Lloyd D., ‘The Seven Seals of Judeo-Islamic Magic: Possible Origins of the Symbols’ (ePublication on Academia.edu, 15 April 2012) unpublished, undated manu­ script, 29 pp.), https://www.academia.edu/1509428/The_Seven_Seals_of_Judeo_ Islamic_Magic_Possible_Origins_of_the_Symbols Gyselen, Rika, Sasanian Seals and Sealings in the A. Saeedi Collection (Leuven: Peeters, 2007) Hartner, Willie, and Richard Ettinghausen, ‘The Conquering Lion, the Life Cycle of a Symbol’, Oriens, 17 (1964), 161–71 Huff, Dietrich, ‘Techno­logical Observations on Clay Bullae from Takht-i Suleiman’, Mesopotamia, 22 (1987), 367–90 Khan, Geoffrey, ‘The Pre-Islamic Background of Muslim Legal Formularies’, ARAM, 6 (1994), 193–224 —— , ‘The Opening Formula and Witness Clauses in Arabic Legal Documents from the Early Islamic Period’, Journal of the American Oriental Society, 139.1 (2019), 23–39 Krachkovskaja, V. A., and I. J. Krachkovskij, ‘Drevnej arabskij dokument iz Srednej Azit’, in Sogdijskij sbornik (Leningrad: Nauk, 1934), pp. 52–90 Lerner, Judith A., ‘An Introduction to the Sealings on the Bactrian Documents in the Khalili Collection’, in Ērān ud Anērān: Studies Presented to Boris Il’ič Maršak on the Occasion of his 70th Birthday, ed. by Matteo Compareti, Paola Raffetta, and Gianroberto Scarcia (Venice: Libreria Editrice Cafoscarina, 2006), pp. 371–86 —— , ‘Observations on the Typo­logy and Style of Seals and Sealings from Bactria and Greater Gandhara’, in Coins, Art and Chrono­logy, ii: The First Millennium c.e. in the Indo-Iranian Borderlands, ed. by Michael Alram, Deborah Klimburg-Salter, Minoru Inaba, and Matthias Pfisterer (Vienna: Verlag der Österreichen Akademie der Wissenschaften, 2010), pp. 245–66 —— , ‘Glyptic Art from Bactria to Gandhara: The Aman ur Rahman Collection in Context’, in Judith A. Lerner and Nicholas Sims-Williams, Seals, Sealings and Tokens from Bactria to Gandhara (4th to 8th Century ce) (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2011), pp. 17–54 Lerner, Judith A., and Nicholas Sims-Williams, Seals, Sealings and Tokens from Bactria to Gandhara (4th to 8th Century ce) (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2011) Lerner, Judith A., and P. Oktor Skjærvø, ‘Some Uses of Clay Bullae in Sasanian Iran: Bullae in the Rosen and Museum of Fine Arts Collections’, in Sceaux d’Orient et leur emploi, ed. by Rika Gyselen, Res Orientales, 10 (Bures-sur-Yvette: Groupe pour l’Étude de la Civilisation du Moyen-Orient, 1997), pp. 67–78 Marshak, B. I., ‘Bronzovyj kuvshin iz Samarkanda’, Srednjija Azija i Iran, (1972), 61–90, 180–81 Minns, Ellis H., ‘Parchments of the Parthian Period from Avroman in Kurdistan’, Journal of Hellenic Studies, 25 (1915), pp. 22–65 Porter, Venetia, Arabic and Persian Seals and Amulets in the British Museum (London: British Museum, 2011; 2nd printing with corrections, 2017) Rea, John R., R. C. Senior, and A. S. Hollis, ‘A Tax Receipt from Hellenistic Bactria’, Zeitschrift für Papyro­logie und Epi­graphik, 104 (1994), 261–80 Sena Chiesa, Gemma, Gemme di Luni (Rome: G. Brettschneider, 1978)

325

3 26

j u di th a . l e rn e r

Sijpesteijn, Petra J., ‘Seals and Papyri from Early Islamic Egypt’, in Seals and Sealing Practices in the Near East: Developments in Administration and Magic from Prehistory to the Islamic Period, ed. by Ilona Regulski, Kim Duistermaat, and Peter Verkinderen, Orientalia Lovaniensia Analecta, 219 (Leuven: Peeters, 2012), pp. 165–74 —— , ‘Expressing New Rule: Seals from Early Islamic Egypt and Syria, 600–800 ce’, The Medi­eval Globe, 4.1 (2018), 99–148 Sims-Williams, Nicholas, and François de Blois, with contributions by Harry Falk and Dietrich Weber, Studies in the Chrono­logy of the Bactrian Documents from Northern Afghanistan (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2018) Vandorpe, Katelijin, Breaking the Seal of Secrecy: Sealing-Practices in Greco-Roman and Byzantine Egypt Based on Greek, Demotic and Latin Papyro­logical Evidence (Leiden: Papyro­logisch Instituut, 1995) Vandorpe, Katelijin, ‘Seals in and on the Papyri of Greco-Roman and Byzantine Egypt’, in Archives et sceaux du monde hellénistique/Archivi e sigilli nel mondo ellenistico, ed. by Marie-Françoise Boussac and Antonio Invernizzi (Paris: De Boccard, 1996), pp. 232–91 Walker, C. B. F., ‘Tontafel, Tontafelhülle: A. In Mesopotamia, § 3. Envelopes’, in Reallexikon der Assyrio­logie und Vorderasiatischen Archäo­logie, xiv (Berlin: De Gruyter, 2014–2016), pp. 101–04. Wassiliou, Alexandra-Kyriaki, and Hermann Harrauer, Siegel und Papyri: das Siegelwesen in Ägypten von römischer bis in früharabische Zeit (Vienna: Österreichische Verlagsgesellschaft, 1999) Weber, Dietrich, ‘Studies in Some Documents from the Pahlavi Archive’, Bulletin of the Asia Institute, 26 (2012 [2016]), 61–95

Section Three

Languages in Contact and Shared Spaces

Khodadad Rezakhani

Navigating Persian The Travels and Tribulations of Middle Iranian Languages Languages, as the main means of communication and one of the primary tools for the performance of power throughout history, are integral parts of political action. The position of a language within the structure of power often determines the position of its speakers within society and their power to negotiate this position.1 In the early Islamic world, the quick rise of Arabic as the language of religion, as well as the language of the rising power of the Islamic state, threatened the survival of the many languages within the growing empire and challenged the social position of their speakers. Some languages, such as Aramaic (and its dialects) and Coptic, almost disappeared or were reduced to minority status, their speakers largely adopting Arabic or else moving to the margins of society. In the ‘lands of the Eastern Caliphate’ — the former territories of the Sasanian empire in Iran and Iraq — the parallel rise of New Persian (NP) alongside Arabic presented an interesting challenge.2 While Arabic remained the language of learning and religion in the Persophone world for a long time, up to the Safavid period, Persian slowly replaced it east of the Tigris. It was in fact the rise of Persian, and not the dominance of Arabic, that doomed the survival of languages such as Khwarazmian, Sogdian, Bactrian, and many other local languages. It wasn’t the survival of Persian in defiance of Arabic — the usual way the process is viewed in modern nationalist discourse — that is significant, but its birth and growth alongside Arabic that is of great interest for the study of the rise of Persian and the Persianate world. In what follows, I survey the prevailing theories regarding the development of Persian, present a slightly modified view of the circumstances of its rise, and add a discussion of East Iranian languages (mainly Sogdian and Bactrian) to the debate. Though I aim to be efficient, the limits of space and the extent of the discussion prevent me from achieving any of the above in a comprehensive manner. Instead, I seek to make a case for the rebirth of Early New Persian (ENP) in the East (the

1 In the context of ancient empires, see Ando, Roman Social Imaginaries, pp. 29–52 for the example of the Roman empire. 2 See the contribution by Borrut to the present volume. Khodadad Rezakhani is a Historian of Late Antique Iran and Central Asia, Senior Researcher and Lecturer, Leiden University, the Netherlands. Navigating Language in the Early Islamic World: Multilingualism and Language Change in the First Cen­turies of Islam, ed. by Antoine Borrut, Manuela Ceballos, and Alison M. Vacca, ISMAR 2 (Turnhout: Brepols, 2024) pp. 329–363 10.1484/M.ISMAR-EB.5.134631

FHG

330

k ho da da d r e z akh an i

so-called Greater Khorasan) and thereby to contribute to a debate on the origins of New Persian, a topic that has been the subject of many scholarly debates already and whose linguistic details have been discussed by many erudite scholars of the field, particularly in the Persophone scholarly sphere itself.3

What is Persian? The term Persian and its variants, as a reference to a people and to a language, has a rather conspicuous history, from the self-referencing of Darius the Great in Old Persian and its Greek use as both an ethnonym and the name of the language to the modern appellation of Farsi as a name for the language. Persian is the only member of the Iranian language family4 to have been attested in all three stages of the Iranian languages.5 Old Persian was the language of the Achaemenid royal inscriptions, whereas Middle Persian (MP) was used in Sasanian inscriptions and Pahlavi literature, as well as in Manichaean and other literary production.6 Classical New Persian (CNP) was used as the administrative and literary language of much of Iran and Central Asia from the tenth century onwards. The late stage of the language, a direct continuation of CNP, remains the modern language of the countries of Iran, Afghanistan, and Tajikistan (as well as parts of Uzbekistan). At one point in its history, Persian dominated as an cosmopolitan language over an area stretching from Burma to Bosnia, used by kings, administrators, merchants, poets, and common people.7 It has been, at various times, a language of high literature and power as well as one of cross-community communication.8 Perhaps amongst its most significant aspects is the fact that in its latest incarnation, the so-called New Persian, the language has stayed comprehensible to its native speakers since at least the tenth century ce. Despite many phono­logical and a few morpho­logical variations, the language of the tenth-century Samanid

3 In European languages, the classical study and discussion of the rise of New Persian is Lazard, ‘Pahlavi, pârsi, dari’, with reconsiderations and revisions in Lazard, ‘The Rise of the New Persian Language’ and ‘Pârsi et dari’, as well as many studies by Bo Utas, including ‘A Multiethnic Origin of New Persian?’ and many articles included in Utas, From Old to New Persian. In Persian, the matter has been discussed by most scholars of Classical Persian, perhaps best represented by Sadeghi, Takvin-e zaban-e farsi and Azarnouche, Chalesh-e miyan-e farsi o arabi. The many other works of these scholars, particularly Sadeghi, are referenced throughout this paper. 4 Skjærvø, ‘Iran vi. Iranian Languages and Scripts’. 5 Abolghassemi, Zaban-e farsi o sargozasht-e an o goftarha-ye digar, p. 13. See Windfuhr, ‘Dialecto­logy and Topics’ for a complete survey of Iranian languages, including the three stages of Persian. 6 Macuch, ‘Pahlavi Literature’. 7 Eaton, ‘The Persian Cosmopolis and the Sanskrit Cosmopolis’, building on Pollock, ‘The Cosmopolitan Vernacular’ and subsequent work. I think the case might be slightly different, as the process of Persian cosmopolitanization before the Mongol period does not lead from a hegemonic, literary language like Sanskrit to the rise of a vernacular; rather, it is a top-to-bottom process of formation, as will be discussed further below. For the later, post-Timurid, expansion of New Persian as a cosmopolitan language, see Green ed., The Persianate World. 8 For studies of this Persianate world see Amanat and Ashraf ed., The Persianate World.

navi gat i ng pe rsi an

court, as recorded in great works of literature, is still largely accessible to even an uneducated modern reader.9 This is, of course, partly a by-product of the fact that the language of the classical literature of the tenth to fourteenth centuries ce has been frozen as a high register, refined and preserved as the perfect form of the language, resembling the case of Arabic.10 However, this tenth-century classical language was itself in many senses a new invention, a language assembled from the remnants of a vernacular intensely peppered with neo­logisms, loanwords, and grammatical simplifications. Some have called this a sign that ENP was a creole language. However, considering the circumstances of its earliest development under the Sasanian rule, I prefer not to use the term creole, which can both act as a marker of exceptionalism and indicate a misunderstood social construct.11 ENP was, for all practical purposes, based on an earlier stage of the language known as Middle Persian, although it evidently derives from a dialect parallel to the official Middle Persian of the Sasanian court. The question of how this colloquial form of the language managed to develop into the prestige language of classical poets such as Ferdowsi and Saʿdi is a matter of speculation and much scholarly debate that lies outside the purview of this paper.12

From Middle to (Early) New Persian As the language of the Samanid court and its successors (as opposed to the Buyid court, which never adopted it), New Persian gained an unprecedented position within the Islamic world. By the time of the Mongols, it was the de facto language of power east of the Tigris and to the Jaxartes.13 The Mongols extended this range even further, all the way to China, and under the Timurids and their successors, New Persian reached its zenith as a prestige language of administration and literature.14 This success of the Persian language needs to be contextualized within its origin as the administrative language of the Samanid empire in the tenth century.15 The Samanid adoption of the language, initially as an administrative language in their capital, Bukhara, and then as the chosen language of literary production, indeed provides an anchoring point for the transition from Early New Persian to Classical New Persian.16 9 Abolghassemi, Zaban-e farsi o sargozasht-e an o goftarha-ye digar, pp. 30–34. 10 For many examples of this continued comprehensibility and the dominance of the classical heritage over the later stages of the language, see Thackston, A Millennium of Classical Persian Poetry (with English translations and commentary) as well as the monumental collection of Keshavarz, Hezar saal nasr-e parsi. See Perry, ‘New Persian’, p. 82 for a different view of what he calls its homoglossia. 11 Ansaldo, Matthews, and Lim, Deconstructing Creole, pp. 3–4. 12 For a summary, see Paul, ‘Persian Language i. Early New Persian’. 13 Eaton, ‘The Persian Cosmopolis and the Sanskrit Cosmopolis’, pp. 69–74. 14 For a discussion of different views of this, see Khanbaghi ‘Champions of Persian’, pp. 193–215. 15 Eaton, ‘The Persian Cosmopolis and the Sanskrit Cosmopolis’, p. 71. 16 Eaton, ‘The Persian Cosmopolis and the Sanskrit Cosmopolis’, p. 69, following Perry, ‘New Persian’, p. 72, who argues that ‘since it did not serve as the vehicle for any scripture or liturgy, New Persian

331

332

k ho da da d r e z akh an i

At the same time, we must consider the coexistence of Zoroastrian Middle Persian literature (that is, the ‘Pahlavi literature’) and the earliest stages of New Persian before its adoption by the Samanids.17 The apparent gap between the two languages and the appearance of antiquity attached to the former corpus are rooted in historical and stylistic reasons. Namely, we tend to think of the Zoroastrian MP texts as reflections of a much ‘older’ tradition of writing, preserving a historical language, frozen in time. Additionally, the stylistic maturity of these texts leads us to think of them as reflecting an older stage of the language.18 However, the facts of the redaction and composition of these texts in the ninth and tenth centuries are of central importance for understanding the historical setting of the texts themselves, as well as the almost contemporary rise of New Persian literature. We can perhaps consider the close affinity between the two languages as being reflective of two registers of the same language, instead of representing two stages of language development.19 The prominence of Pahlavi Middle Persian as a written language outside the purview of Zoroastrian liturgical texts and its remarkable stylistic integrity is evident in the Pahlavi archives newly discovered in the regions of Qom and Tabarestan.20 Both archives date to the late seventh and early eighth centuries ce and include similar content, and they can partly cover the gap between the fall of the Sasanians and the more significant appearance of ENP in the ninth century.21 Their documents pertain mostly to official and legal issues, ranging from the purchase of land and labour to the payment of taxes. Legal content, in the form of material relating to court cases, appeals, and subpoenas are also prominent. In one sense, the archived documents, written after the end of Sasanian rule, show the remarkable continuity of Sasanian legal and administrative frameworks even after the fall of the dynasty. Additionally, they also showcase the uniformity of Sasanian administrative and legal language in the form of written Pahlavi Middle Persian and the established legal and administrative vocabulary that distinguished the language.22 The Qom and

17 18

19 20

21 22

posed no ideo­logical threat to Arabic, the language of Iran’s seventh century Muslim victors’. While I do not disagree with Perry’s earlier statement about the language’s lack of connection to any liturgy, I am not sure this should be seen as a prime reason for the spread of ENP, as I discuss further below. See Daryaee, ‘Middle Persian (Pahlavi)’, pp. 103–16, and Sheffield, ‘New Persian’, for a survey of the use of New Persian alongside Zoroastrian Middle Persian for recording Zoroastrian religious texts. Boyce ‘Middle Persian’, pp. 38–47, with some discussion of the stylistic characteristics of ninthcentury texts within a discussion of dating the texts; Kreyenbroek, ‘The Zoroastrian Tradition from an Oralist’s Point of View’. See also Vevaina, ‘“The Ground Well Trodden but the Shah Not Found …”’ for a discussion of style, as well as orality, in Zoroastrian literature. Lazard, ‘The Rise of the New Persian Language’, p. 596. See the biblio­graphy for the many publications of Gignoux and their revisions by Weber, too numerous to list here, mostly published in of the series Res Orientales. Weber’s ‘Pahlavi Legal Documents from Tabarestān on Lease, Loan and Compensation’ and ‘Pahlavi Legal Documents from Tabarestān: Two Claims’ also provide useful biblio­graphies. Weber, ‘New Arguments for Dating the Documents from the “Pahlavi Archive”’. Macuch, ‘The Legal Context of the Tabarestān Court Records’, ‘Pahlavi Legal Documents from Tabarestān on Lease, Loan and Compensation’, and ‘Pahlavi Legal Documents from Tabarestān: Two Claims’.

navi gat i ng pe rsi an

Tabarestan archives demonstrate a notable resemblance to the collection of papyri preserved from Sasanian-controlled Egypt in the seventh century. Consequently, it seems quite clear that the Sasanian administration had managed to establish a developed administrative language throughout its domains. Although we lack similar archives for the north-east, the regions of Abarshahr, Marv, and Poshang, there is no reason to think that the situation was any different in these places. Of course, the Bactrian and Sogdian archives from outside the borders of the Sasanian empire23 perhaps attest to the imperial nature of Pahlavi Middle Persian as the specific tool of the Sasanian state, indeed marking the extent of Sasanian control. The context of these Pahlavi archives of Tabarestan and Qom and their coexistence with the earliest mentions of ENP in Arabic texts24 point to the role of script as a marker of the differences between the two languages. Middle Persian (written in Pahlavi) and New Persian (written in the Arabic script) show close phono­logical affinities and were most likely distinguished by the way they were written.25 Thus, the way a language was written could determine its identity, too. This holds true for New Persian as an independent language, with its adoption of the Arabic script as a marker of its independence from literal Middle Persian written in the Pahlavi script.26 Beyond script and stylistic differences, the two registers of the language were closely related and shared a majority of their basic lexical components, as well as morpho­logical and phono­logical characteristics.27 So, for those familiar with the administrative Middle Persian of the Sasanian state, the spoken New Persian would have been familiar and largely comprehensible. Different registers were most likely already familiar in any case, in view of the presence of a dialect in the north-east influenced by Parthian and preserving many older Middle Persian forms.28 The rising status of Early New Persian as the language of administration and of a new religion in the former Sasanian territories would then have caused little disconnection in the established administration and among the general populace already familiar with administrative Middle Persian. Here it is perhaps worth noting that Sasanian administration was established, at least on a temporary

23 See below for a discussion of these languages and the related archives. 24 Azarnouche, Chalesh-e miyan-e farsi o arabi, pp. 16–18. 25 For an example of this, see the passage from Ibn al-Muqaffaʿ discussed below. It is significant that the passage appears in a section of Ibn al-Nadim’s Fihrist dedicated to scripts and script forms. 26 Perry, ‘New Persian’, pp. 75–81. 27 See Lazard, ‘The Rise of the New Persian Language’, pp. 596–97 for a discussion of the closeness of ENP with what we know as written Middle Persian. 28 See Sadeghi, Takvin-e zaban-e farsi, pp. 7–11, and Tedesco, ‘Dialekto­logie der westiranischen Turfantexte’ for a comparison of the various Middle Iranian languages as reflected in texts found outside Sasanian territories and their initial differentiation as ‘Parthian’ and ‘Middle Persian’. Indeed, early researchers found the two languages so close that they did not notice such ‘dialectal’ differences (my thanks to Dan Sheffield for pointing this out). Durkin-Meisterernst, Grammatik des Westmitteliranischen provides an exhaustive introduction to the grammatical details of Parthian and Middle Persian accompanied with texts, in fact demonstrating the overlapping characteristics of the two languages. I will discuss these in a historical context below.

333

334

k ho da da d r e z akh an i

basis, in the extreme north-east as far east as the city of Balkh itself in the aftermath of the Western Türk destruction of the Hephthalite power in the second half of the sixth century ce.29 The lack of stylistic qualities in Early New Persian was a significant aspect of the language, making it particularly open to absorbing features from the languages with which it came to contact. The exposure of the language to Arabic as the language of power and ideo­logy enriched it with a new vocabulary, allowing it to reflect the needs of the new system.30 This is perhaps the most remarkable feature that allowed a non-Arabic language to flourish under the Islamic system and come to define certain characteristics even in Arabic itself.31 The association of ENP with Islam then made it a suitable vehicle for the eastern penetration of Islam in the late seventh and early eighth centuries. This is perhaps most visible in the presence of a few Arabic expressions in the Judaeo-Persian letter from Dandan Uiliq.32 It shows that the borrowed Arabic lexicon was so central to the language that it could not be entirely separated from it even when writing in a different script. At this stage of the language, of which we have little evidence, ENP appears to have undergone a process of simplification, including its decisive abandonment of Middle Persian grammatical features such as the ergative construction.33 The simplification of certain aspects (such as the disappearance of most oblique pronouns)34 indicates ENP’s status as a transitional idiom and its function as a connective language. At this point, the borrowings from the neighbouring languages also start to define the character of New Persian, setting it apart from Middle Persian and giving it its familiar ENP flavour. As I will discuss below, the simplification, combined with exposure to a panoply of new languages from Sogdian to Khwarazmian, Bactrian, and Turkic, plus a range of less-known local dialects (e.g. Bukharan), launched ENP into a new stage.35 A significant feature of this shift from ENP to CNP might be the retention of multiple cognates of a word, often with the non-Persian word taking precedence over the Persian one; for example, Sog. mul versus Pers. mey ‘wine’, and Sog. malax ‘locust’ versus Pers. mēg (preserved in southern dialects as meygu ‘prawn’).

29 Potts, ‘Sasanian Iran and its Northeastern Frontier’, pp. 297–300; Rezakhani, ReOrienting the Sasanians, pp. 180–82; Rezakhani, ‘Shahnameh as a Historical Source’. This assumes that the earlier period of Sasanian control over the east, in the guise of the Kushano-Sasanian dynasty, had not left any clear linguistic effects (Rezakhani, ReOrienting the Sasanians, pp. 72–86), despite possible periods of direct Sasanian control; see Nikitin, ‘The Sasanian Šahrab of Balkh’. 30 Azarnouche, Chalesh-e miyan-e farsi o arabi, pp. 195–202. 31 Azarnouche, Chalesh-e miyan-e farsi o arabi, pp. 253–64; Azarnouche, ‘La formation du persan sous l’influence de la langue arabe au iv/xe siècle’. 32 Utas, ‘The Jewish-Persian Fragment from Dandan-Uiliq’ and ‘A Multiethnic Origin of New Persian?’. 33 Windfuhr, ‘Dialecto­logy and Topics’, pp. 31–32. 34 This, of course, is mostly a feature of the written and literary form of Middle Persian and of the modern day eastern/Khurasani dialects. At the same time, Central and Western dialects have preserved oblique pronouns in colloquial forms: Windfuhr Persian Grammar, pp. 52–53. 35 Utas, ‘A Multiethnic Origin of New Persian?’, which basically presents the possibility that ENP is the product not of the development of a single language but rather of a merging of several languages.

navi gat i ng pe rsi an

This hyperglossia, however, generally left ENP without a significant written tradition. The language had already taken shape as a spoken tongue in the late Sasanian period, as is evident from various mentions of it.36 But there does not seem to be a written tradition associated with this stage of the language, which is perhaps the reason for its heterodox writing traditions, including early Judaeo-Persian letters, which commit the language to Hebrew letters. This, of course, does not exclude the possibility that epistolary conventions from Pahlavi Middle Persian had permeated into New Persian’s writing style. Examples of compositions that are considered linguistically ENP but are in fact written in Pārsīg/Pahlavi script further strengthen the idea of the coexistence of the two languages.37

The ‘Persian’ Languages from an Early Islamic Viewpoint Before ENP’s adoption by the Samanid court as an administrative and literary language, the earliest traces of the language appear in bits and pieces of poetry, often attributed to various poets originating in East Iran or the Islamic ‘Greater Khurasan’.38 These poems were often commissioned verses, appearing in isolation in sources such as Tarikh-e Sistan, Ibn Khurradadbih’s al-Masalik wa-l-mamalik, or Chahar maqala, giving the impression of a fully developed literary language. The eastern origin of these poems fits well the context given in an excerpt attributed to Ibn al-Muqaffaʿ and quoted by a few, including Ibn al-Nadim.39 The passage reads as follows: ‫ فأما‬.‫وقال عبدهللا بن المقفع لغات الفارسية الفهلویة و الدریة و الفارسیة والخوزیة والسریانیة‬ ‫الفهلویة فمنسوب الی فهله اسم یقع علی خمسة البلدان و هی اصفهان و الری و همدان و ماه نهاوند‬ ‫ و أما الدریه فلغة مدن المدائن و بها کان یتکلم من بباب الملک و هی منسوبة إلی‬.‫و اذربیجان‬ ‫ واما الفارسیة فیتکلم‬.‫حاضرة الباب والغالب علیها من لغة أهل خراسان و المشرق لغة اهل بلخ‬ ‫ و اما الخوزیة فیها کان یتکلم الملوک و‬.‫بها الموابدة و العلماء و أشباههم و هی لغة أهل الفارس‬ ‫ و اما السریانیة فکان یتکلم بها اهل‬.‫االشراف فی الخلوة و مواضع اللعب و اللذة و مع الحاشیة‬ 40... ‫السواد و المکتابة فی نوع من اللغة بالسریانی فارسی‬ ʿAbd Allāh ibn al-Muqaffaʿ said, ‘The languages of the Persians are the

Pahlāwī (sic), the Derī, the Parsī, the Khuzistānī, and the Syriac. The Pahlāwī (al-Fahlawīyah) is related to Pahlav (Fahlah), a region which includes five cities: Iṣbahān, Rayy, Hamadhān, Mah Nahāwand, and Ādharbayjān. The Dērī (al-Durīyah) was the language of the cities of al-Madā’in, spoken at the king’s court. It was derived from the presence at the court (al-bāb), 36 See Azarnouche, Chalesh-e miyan-e farsi o arabi, pp. 16–18 for examples. 37 De Blois, ‘A Persian Poem Lamenting the Arab Conquest’. 38 For a study of these poems, see Lazard, ‘The Rise of the New Persian Language’; Utas, ‘A Multiethnic Origin of New Persian?’, pp. 181–82. 39 Lazard, ‘Pahlavi, pârsi, dari’, pp. 361–62 quotes the passage in its three versions by Ibn al-Nadim, al-Khwarazmi, and Yaqut (quoting Hamza al-Isfahani), writing the Arabic in Latin transcription. I have used a standard edition here (Ibn al-Nadim, Fihrist, ed. by Tajaddod). 40 Ibn al-Nadim, Fihrist, ed. by Tajaddod, p. 15.

335

3 36

k ho da da d r e z akh an i

coming chiefly from the language of the people of Khurāsān and the East, the speech of the people of Balkh. Priests, scholars, and their like speak Parsī (al-Farsīyah), the speech of the people of Fars. The kings and nobles used to speak the Khuzistāni (al-Khūziyah) in privacy, in places of play and amusement, and with their retinues. The people of al-Sawād used to speak Syriac (al-Suriyanīyah), writing in one form of Persian Syriac.41 In addition to the liberties that the English translator of the text has taken (such as translating al-Farsiyya once as ‘the Persians’ and on another occasion as ‘the Parsī’), most other parts of this passage are disputed. Gilbert Lazard’s translation to French yields the same sense: Le dari est la langue des villes de Madāʾin; il était parlé par ceux qui étaient à la cour du roi; (son nom) se rapporte à la présence à la cour; parmi les langues des gens du Khorassan et de l’Orient, c’est celle des gens de Balkh qui y domine.42 The Persian translation of the text, the only one based on the most complete edition of the Arabic text,43 renders the sentence as follows: ‫ و درباریان با آن سخن می گفتند و منسوب بدربار پادشاهی است‬،‫و اما دری زبان شهرنشینان بود‬ 44.‫ زبان مردم بلخ در آن بیشتر است‬،‫و از میان زبانهای اهل خراسان و مشرق‬

The English translator of the text suggests that the language (dialect?) itself is derived from the speech spoken at the court of the (presumably Sasanian) king. The French translation, by contrast, does not propose a derivation but states that Dari was the dialect spoken at the court. The same meaning is rendered by the Persian translator, who uses ‫ به آن سخن می گفتند‬and also interprets the phrase ‫فلغة مدن المدائن‬ as ‘the language of the city-dwellers’. The passage, significantly, conveys the geo­graphically dispersed and typically mixed state of Dari, the most likely candidate for the language we now call ENP — the form of the language before its rise as a classical language in the tenth and eleventh centuries.45 Perhaps more important still is the second part of the sentence. The English translation retains the ambiguity of the Arabic sentence, rendering the phrase ‫ لغة اهل بلخ‬vaguely and suggesting that ‘the language of the people of Khurasan and the East’ is in fact ‘the speech of the people of Balkh’. In the

41 Dodge 1970, p. 24. 42 Lazard, ‘Pahlavi, pârsi, dari’, p. 361. 43 Dodge and Lazard (the latter surprisingly) still rely on Flügel’s 1871 edition of the book. The Persian translation by Tajaddod himself takes advantage of the newer edition by Tajaddod, which is based on newer manu­scripts. I have used this edition and Tajaddod’s excellent translation throughout this paper. 44 Ibn al-Nadim, Fihrist, trans. by Tajaddod, p. 22. 45 Throughout this paper, when the context allows, I occasionally use NP for New Persian, ENP for Early New Persian, CNP for Classical New Persian, and MP for various forms of Middle Persian. See Paul, ‘Persian Language i. Early New Persian’. The primacy of the name ‘Dari’ for the language is mostly assumed in Persian; see Abolghassemi, Zaban-e farsi o sargozasht-e an o goftarha-ye digar, p. 30 and Sadeghi, Takvin-e zaban-e farsi, pp. 20–25.

navi gat i ng pe rsi an

most prominent interpretation of the passage, Lazard has rendered it as ‘parmi les langues des gens du Khorassan et de l’Orient c’est celle des gens de Balkh qui y domine’,46 confirming the centrality of the East in the formation and state of this language by interpreting the ambiguity as meaning that of all the languages of Khurasan and the East, the language that dominates Dari is that of the people of Balkh. The Persian translation does the same thing while moderating the sense of ‫ الغالب‬to ‫ بیشتر‬to suggest, instead of ‘domination’, ‘more of it’ and thus leaving the passage open to further interpretation. The simplest conclusion one can draw from this passage, despite all its problems, is the importance of the languages of the East for understanding what came to be known as New Persian. The prominence of the ‘languages of Khurasan and the East’, particularly that of Balkh, within the setting of New Persian/Dari seems to have been taken for granted, at least from the perspective of an eighth-century translator (Ibn al-Muqaffaʿ) or a tenth-century biblio­grapher (Ibn al-Nadim).47 Such encounters and historical settings can help explain the status of ENP as a language in transition, historically and geo­graphically, interacting with all the other languages it encountered on its journey. Through these interactions, the transitional language formed the means of literary expression and eventually empire-wide communication, eventually becoming the defining characteristic of the Persianate world and a vehicle for expressing power itself. As I will argue further below, I believe that we must consider Ibn al-Nadim’s entire passage in light of tenth-century linguistic realities, instead of seeing it as a genuine reflection of the situation in the eighth century, much less that in the Sasanian period.48 Although its attribution to an original opinion by Ibn al-Muqaffaʿ cannot be fully dismissed,49 certain parts of the passage make sense only in the context of the later development of New Persian and its position vis-à-vis other languages, and they can thus be understood only in a tenth-century context. So let us further consider the position of Dari and other languages mentioned in the passage before proceeding to a discussion of the eastern, ‘Khurasani’, languages with which Dari came to interact. In the geo­graphical sense, it is worth noting the localities to which the various languages are assigned. In Ibn al-Nadim’s passage, fahlawiyya is linked to the area of Fahlav, or Pahlav, and is clearly identified as the area that came to be known as 46 Lazard, ‘Pârsi et dari’, p. 239, supplementing ‘y’ to make sense of the garbled Arabic. 47 Lazard, ‘Pahlavi, pârsi, dari’, pp. 361–62 quotes parallel passages in al-Khawarazmi’s Mafatih al-ʿulum and Yaqut’s Muʿjam al-buldan, all originating with Ibn al-Muqaffa’ and varying to a small degree from Ibn al-Nadim’s quotation. For a discussion of these passages, see also Lazard, ‘Pârsi et dari’. I have taken Ibn al-Nadim’s version here as a representative since the variations make little difference to the historical discussion that is the focus of the current paper. 48 See La Vaissière, Samarcande et Samarra, pp. 132–35 for the development of certain aspects of this linguistic world in Transoxiana. I will elaborate on the matter further below. 49 The assumption has been that Ibn al-Nadim’s quotation of Ibn al-Muqaffaʿ genuinely reflects the situation in the second Hijri century — an assumption that, despite its plausibility, is rather surprising, given that Ibn al-Nadim is obviously providing his own take on the issue. See Sadeghi, Takvin-e zaban-e farsi, pp. 43–44; Markquartt 1901, 89; Lazard, ‘Pahlavi, pârsi, dari’, p. 389.

337

338

k ho da da d r e z akh an i

the Jibal in later Arabic geo­graphies. Dari is the language of the cities of al-Madaʾin, while al-Farsiyya, stricto sensu, belongs to the region of Fars/Persis.50 In addition, al-Farsiya is also described as the learned language of the mobeds, scholars, and others involved with matters of learning, and it is thus given both a devotional and a prestige status.51 A glaring omission is any mention of scribes and administrators, even though the focus of the chapter in which this passage appears is on scripts and scribal habits. Modern scholars take this fārsiyya to refer to the language known from the Zoroastrian religious texts, which were mainly composed or redacted in the ninth and tenth centuries in reaction to and as a result of Islamic dominance.52 Although this language is most commonly known as Pahlavi (after the name of the script), the scholarly convention that has emerged in recent years suggests that it should be called Pārsīg.53 Ibn al-Muqaffaʿ in fact attributes the language to the specific area of Pars/Fars in southern Iran, and thus calling it Pārsīg would more closely connect it to the area of Fars/Persis and the statement of Ibn al-Muqaffaʿ. But the connection of the language to a particular class of society — the mobeds and the learned people — also sets it apart from Dari, which is not given the same status at this stage. Viewed within this context, Ibn al-Muqaffaʿ’s comment that Dari is the name of the language of ‘the cities of al-Madaʾin’ also deserves some attention. In his treatment of the passage, Ali Ashraf Sadeghi makes two statements whose implications, I believe, summarize the modern scholarly attitude towards the entire testimony of Ibn al-Nadim and the assumptions made on its basis. Sadeghi says that ‘Dari is certainly not an invention of the Islamic period. […] It is the language of the Dar, the court of the king, which is Ctesiphon’.54 Paradoxically, in a further passage he then adds that ‘Dari was the language of the majority of the people, and up to now, no document written in Dari has surfaced from the Sasanian period’.55 These statements seem to convey opposite meanings, portraying Dari both as the language of the high court and as the vulgar tongue of the common folk. One might ask what function the language of the court would have served if

50 Note that at the beginning of the passage, Ibn al-Muqaffaʿ/Ibn al-Nadīm places the entire set (including Pahlavi, Dari, Farsiyya, Khuzistani, and Suryaniyya) under the heading of al-fārsiyya (Lazard, ‘Pahlavi, pârsi, dari’, pp. 378–80). This wider termino­logy is to be understood either as referring to the people (‘Persians’) or as a general name for the spoken tongues of the East (‘Iranian’ languages). I am not in favour of either conclusion but shall refrain from a discussion of the matter at present. 51 For Fars in the early Islamic period, see Daryaee ‘The Effects’, pp. 193–204 as well as Sumner and Whitcomb, ‘Islamic Settlement’ pp. 309–24. 52 Lazard, ‘Pârsi et dari’, pp. 240–41, elaborating on his statement in ‘Pahlavi, pârsi, dari’, pp. 384–85. For Zoroastrian books in the tenth century, see Bailey, Zoroastrian Problems in the Ninth-Century Books and Stausberg ‘Canonization’, as well as the detailed studies by Rezania (‘Mazdakism’ and ‘The Dēnakard’) about Pahlavi translations of the Avestan texts, probably a central feature of the work of these mobeds in the early Islamic period. 53 Gyselen ‘Le Tabaristān’, pp. 7–8. 54 Sadeghi, Takvin-e zaban-e farsi, p. 26. 55 Sadeghi, Takvin-e zaban-e farsi, p. 27.

navi gat i ng pe rsi an

it was not ever written as an administrative language. Furthermore, how could a courtly, high-register language be at the same time the unwritten language of the people? These statements, and the way they have been understood thus far, appear contradictory and generally ahistorical. It is perhaps fair to assume that the spoken tongue of the Sasanian capital region, being the language of the powerful ruling elite of Sasanian Asorestan,56 was the basis for the language of those who converted to Islam early on. These Persianspeakers, whose use of Early New Persian is mentioned in various accounts of the conquests, included the military and the small land-holding elite (the so-called dihgān class) of the empire, many of whom became vehicles for the continuation of the language and its traditions.57 Various individuals speaking both Arabic and a version of Persian (ENP/Dari?) mentioned in the early Islamic texts provide examples of this converted elite, which includes such luminaries as Hasan al-Basri.58 We also have evidence of these conversions in the narratives of the early Islamic conquests, for example in accounts of the defection of the Sasanian cavalry (Ar. asāwira) to the Muslim side. Descendants of these Sasanian elite forces were amongst the best-known Quranic exegetes in later generations, and they included Musa b. Sayyar al-Uswari (fl. 800s ce), who preached about the Quran in both Arabic and Persian (Fārsiyya) and was reportedly as erudite in Persian as he was in Arabic.59 The familiarity of the early Islamic administrators with Persian as well as Arabic is another indication that Persian was the language of the upper classes/ dihgāns in the region. The famous anecdote on the switching of the language of the dīwān of Kufa from Persian to Arabic at the end of the seventh century shows the continuity of the administration and the administrative language in the region.60 A point to note is that the emphasis is on the change of language, not of script, showing that the change was not simply from the Pahlavi script to the Arabic one. Here, I would like to consider two historical issues in tandem. First, given the discussion above, it seems rather indisputable that what Ibn al-Nadim understood as Dari referred to what was known at the time as the language of ENP poetry and the administrative language of the Samanids. Furthermore, it appears to me that the attribution of the language to the Sasanian court is in fact a modern conjecture that rests on taking the two parts of Ibn al-Muqaffaʿ’s quotation (‘it was the language of the cities of al-Madaʾin’ and ‘it is attributed to the court of the king’) to conjure a reference to the Sasanian past. It thus assumes that the language of the cities of al-Madaʾin was also that of the ‘court of the king [in al-Madaʾin]’, a connection that is not in fact made in the text. We have no reason to assume that Ibn al-Nadim was referring to the Sasanian court rather than a contemporary court such as that of the Samanids, where ENP was in fact the court language. (It was not the language spoken in the region; this was Bukharan Sogdian, as will be discussed below.) 56 Morony Iraq, pp. 185–89, 199–208. 57 Morony Iraq, pp. 209–11. 58 Morony Iraq, p. 196. 59 Jahiz, Bayan, p. 293. 60 Baladhuri, Futuh, pp. 300–01 (Hitti ‘the Origins’, pp. 465–66) and al-Qadi ‘The Names of Estates’.

339

340

k ho da da d r e z akh an i

Had Ibn al-Nadim seen Dari as strictly connected to al-Madaʾin, the Ctesiphon of the Sasanians, and as different from the best known form of Persian at the time, the language of the Samanid court, he would surely have mentioned the difference. I would thus like to postulate that the entire passage is most likely to reflect the realities of the tenth century, the period of Ibn al-Nadim’s life and career, rather than those of the eighth century, when Ibn al-Muqaffaʿ lived, or of the Sasanian period preceding him. Furthermore, the above statement regarding the status of fārsiyya as the language of clergy and religious scholarship offers additional support for this theory. We have no evidence that there was significant production of Zoroastrian literary and religious texts in the eighth century, whether in the area of Fars or elsewhere.61 Consequently, there is no reason Ibn al-Muqaffaʿ would have connected the language of Fars to the Zoroastrian clergy. By the tenth century, however, Fars and other places were seeing vigorous production of such texts by Zoroastrian mobeds, justifying the attribution of fārsiyya to them. This use of fārsiyya, unlike that mentioned above to denote apparently the entirety of the languages spoken in Eranshahr, appears to be a placeholder for Zoroastrian Middle Persian, the language of literary production by Zoroastrian priests.62 Ibn al-Nadim’s quotation thus seems to offer an impressionistic picture of the situation in his own day, even though the passage is attributed to the quintessential authority on all things pre-Islamic Iranian, namely Ibn al-Muqaffaʿ.

New Persian: The Political Life of a Language The transformation of ENP into CNP is a direct product of the Samanid adoption of ENP as an administrative language and its eventual development into a literary (classical) form. This evolution is most evident in the efflorescence of Persian prose and poetry in the courts of the Samanids, the Ghaznavids, and the Ghurids.63 Consequently, the politics of early Islamic Khurasan, the region where the new language was spoken, are central to the study of this process. The conquest of the east and the north-east of Iran took longer than did the conquest of the rest of the lands of Eranshahr.64 The incursions into Sistan appear to have been early, but

61 Boyce ‘Middle Persian’, pp. 42–45; De Jong, ‘Pāzand and “Retranscribed” Pahlavi’. 62 See Kreyenbroek, ‘The Zoroastrian Tradition from an Oralist’s Point of View’ for an interesting take on methods of preserving the texts and committing them to writing. For a classical study, see Bailey, Zoroastrian Problems in the Ninth-Century Books. See also Cereti, ‘Middle Persian Literature i’. 63 Lazard, ‘The Rise of the New Persian Language’, pp. 606–11. The emphasis of most studies of the development of CNP has been on the poetics, as illustrated by Lazard’s study, of which only two pages are devoted to prose. But as Keshavarz demonstrates in Hezar saal nasr-e parsi, the development of prose is just as important for understanding the development of the classical literary language; see also Meisami, ‘Why Write History in Persian?’. 64 See Haug, The Eastern Frontier for a study of the local politics of the eastern frontier in the period of the conquests.

navi gat i ng pe rsi an

they brought Muslims into conflict with the local rulers (the Rutbils).65 The rest of Sasanian Khwarasan and beyond was much slower to come under the newcomers’ rule. This is perhaps due to the fact that these regions were not under Sasanian control but were more closely tied to the Western Türk power and could afford to put up better resistance.66 The establishment of the Chinese protectorate of the West, and even the possible instalment of a Sasanian puppet state in the region under the descendants of Yazdgerd III, perhaps has to do with the resistance in the region.67 Late Islamic conquest narratives tend to lump all the conquest efforts into a linear narrative, with effective actors appearing only in al-Sawad, directing triumphal campaigns against the East. But even these sources cannot quite hide the difficulty of establishing lasting control, mentioning the need for the re-conquest of each of these regions. The persistence of local powers, from the Rutbils to the local rulers of Ghur, Chaghanian, Sogdiana, and Ustrushana, is further evidence of the struggle. At the same time, the persistence, or perhaps even the new introduction, of the administrative and legal Middle Persian of the Sasanians to the unconquered fringes of the Islamic empire also signals the place of the language as a form of resistance to power. The independent Ispahbeds of Tabarestan, who resisted conquest from their highland refuge in the Alburz range, appear to have re-introduced Sasanian administration in the region and, at the same time, to have considered their territory an extension of the Sasanian administration of Khurasan.68 They did so through the introduction of a remarkably resilient coin series, copied from the Sasanians and preserving certain archaic features.69 The Tabarestan archive also shows the prominence of Middle Persian and its full legal vocabulary in the region.70 Even years after losing its independence and embracing Islam (although preserving regional autonomy), the aristocracy of the region continued to use Middle Persian and the Pahlavi/Pārsīg script as a prestige language and script, as shown by their famous tomb towers.71 In Tabarestan, unlike in the rest of Eranshahr, where ENP was adopted as the language of power, Middle Persian continued to enjoy high status, affirming the claims of the local rulers to Sasanian ancestry.

65 Bosworth, Sistan, pp. 14–17. 66 The Rutbils indeed put up much stiffer resistance in Zabulistan and Kabulistan, essentially keeping the region out of the direct control of the government in Damascus or Baghdad until their defeat at the hands of Yaʿqub b. Layth after 860 ce; Forstner, ‘Yaʿqūb b. al-Laiṯ und der Zunbīl’; Kennedy, The Prophet and the Age of the Caliphates, p. 88; Rezakhani, ‘Pangul and Bunji, Zhulad and Fulad’, p. 221. 67 Compareti, ‘The Last Sasanians in China’; Gyselen, ‘“Umayyad” Zavulistan and Arachosia’. 68 Madelung, ‘The Minor Dynasties of Northern Iran’. On the Tabarestan archive and the Ispahbed of Khurasan, see Weber, ‘Pahlavi Legal Documents from Tabarestān on Lease, Loan and Compensation’, pp. 155–57 (document Tab. 17). 69 Malek, The Dābūyid Ispahbads and Early ʿAbbāsid Governors of Tabaristān. 70 Macuch, ‘The Legal Context of the Tabarestān Court Records’, ‘Pahlavi Legal Documents from Tabarestān on Lease, Loan and Compensation’, and ‘Pahlavi Legal Documents from Tabarestān: Two Claims’. 71 Cereti, ‘Two Late Pahlavi Inscriptions from the Alborz Mountains’.

341

342

k ho da da d r e z akh an i

The favourable attitude of the early Muslim governors towards the language and its adoption in the new system is also notable. Of the early Muslim rulers of the eastern territories, many, starting with ʿAbd Allah b.  ʿAmir and continuing with both Ziyad b. Abihi and his son ʿUbayd Allah, largely adopted Sasanian norms of administration.72 Their efforts in the area of language were significant, as demonstrated by the coins of ʿUbayd Allah, which in fact ‘translate’ his name into Middle Persian, rendering it as ’wbytwl’ y zy’t’n ‘ʿUbayd Allah i Ziyadan’, using the Middle Persian ezafa and the patronymic suffix instead of the Arabic formation. Perhaps this also shows the administrative foresight of the early Muslim rulers in appointing governors from amongst the conquered populations, as both Ziyad’s mother and his wife (that is, the mother of ʿUbayd Allah) evidently hailed from the Sasanian territories and appear to have been speakers of Persian.73 The involvement of Persian, whether the administrative Middle Persian of the Sasanian court or the colloquial of al-Madaʾin, in the nascent Islamic system seems undeniable. On the other hand, the closeness of Middle Persian and ENP must have made it quite hard for those in power to distinguish the new converts from the established aristocracy. The new language, as part of the new system, needed its own character and its own peculiar point of origin.74 This was, in fact, what the conquest of Khurasan and the East might have occasioned. The conquest of the East and the introduction of ENP to the region was a reflection of historical realities rather than of any ideo­logical convictions such as ‘Iranian nationalism’,75 with which it is occasionally associated.76 However, the later adoption of ENP as the main administrative language of Khurasan under the Samanids was probably due to clear policies aimed at the use of the language as a political tool.77 Within its new eastern context, ENP came into contact with other languages that fundamentally influenced it.78 In addition, it was in close contact with Arabic, heavily borrowing from the latter79 and even being influenced by it on the phono­logical level.80

72 Kennedy, The Prophet and the Age of the Caliphates, pp. 74–80. 73 Rubin, ‘“Al-Walad li–l-Firāsh”’, pp. 13–17. 74 Azarnouche, Chalesh-e miyan-e farsi o arabi, pp. 260–66. 75 Zia-Ebrahimi, The Emergence of Iranian Nationalism, pp. 99–125. 76 Mahmoodi-Bakhtiari, ‘Planning the Persian Language in the Samanid Period’, pp. 253–55. 77 Mahmoodi-Bakhtiari, ‘Planning the Persian Language in the Samanid Period’, pp. 255–57. 78 It should be kept in mind that whereas we often concentrate on dialectal varieties of Persian, less attention is paid to varieties of dialects amongst eastern languages, especially Sogdian and Turkic languages. See below on a possible Sogdian variety of Bukhara for an example. 79 Azarnouche, ‘La formation du persan sous l’influence de la langue arabe au iv/xe siècle’. 80 Sadeghi, ‘L’influence de l’arabe sur le système phono­logique du persan’, pp. 145–52.

navi gat i ng pe rsi an

The Languages of the East In the passage by Ibn al-Muqaffaʿ, the dialect of Balkh is depicted as the most influential of the eastern tongues (lughat ahl Khurāsān). A statement by Maqdisi further claims that the language of Balkh is the most elegant of languages.81 Lazard dismisses this statement as ‘légèrement bizarre’, an out-of-place intervention that enters the frame of Iranian languages ‘comme des cheveux sur la soupe’. However, the statement simply confirms the establishment of Dari (the language of al-Madaʾin) in places where the north-eastern dialect was prevalent.82 The ‘hair in the soup’ is perhaps easily removed when we consider the languages of the East more broadly. The Bactrian World

At the time when most of the existing studies on the origins of New Persian were being written, our knowledge of Bactrian was still in its infancy. The totality of our access to the language consisted of a few inscriptions83 and speculations made by scholars hypothetically ascribing a Bactrian origin to particular words.84 Since then, however, we have been fortunate to see the discovery of a significant cache of Bactrian documents and their speedy translation and publication.85 This has put at our disposal a treasury of documents through which to explore Bactrian and its role vis-à-vis New Persian. A first impression of Bactrian is that the language benefited from a highly literate culture, with a well-developed administrative and legal, as well as epistolary, vocabulary.86 Further, across the Bactrian documents at our disposal, we can see the interaction between Bactrian and the Turkic dialects that were entering eastern Iran as a result of the Turkic conquests and the establishment of the Western Türk khaganate in Transoxiana and eastern Iran in the sixth and seventh centuries ce.87 Many aspects of these language interactions that would grant us deeper insights into the development of Bactrian and its literary devices remain to be explored. On the other hand, the adoption and proliferation of Bactrian as an administrative and legal language reveals much about the socio-political setting of the eastern Iranian world in the late antique period. The spread of Bactrian on

81 Gignoux, ‘Les quatre’ p. 241; Sadeghi, Takvin-e zaban-e farsi, p. 43. 82 Lazard, ‘The Rise of the New Persian Language’, p. 241; Sadeghi, Takvin-e zaban-e farsi, pp. 42–44. 83 Henning, ‘The Bactrian Inscription’; Gershevitch, ‘Bactrian Inscriptions and Manu­scripts’; Humbach, ‘Two Inscriptions in Graeco-Bactrian Cursive Script from Afghanistan’. For a classical and still highly influential study of Bactrian based on the material available before the beginning of the current century, see Humbach, Baktrische Sprachdenkmäler. 84 Schwartz, ‘Irano-Tocharica’; Tremblay, ‘Irano-Tocharica et Tocharo-Iranica’; Lurje and Yakubovich, ‘The Myth of Sogdian Lambdacism’. 85 We owe this development to Nicholas Sims-Williams and his publication of the Bactrian economic documents, as well as numerous other studies, some of which are listed in the biblio­graphy below. 86 The full text of Bactrian documents has been published by Sims-Williams, Bactrian Documents from Northern Afghanistan, Vol I and Sims-Williams, Bactrian Documents from Northern Afghanistan, Vol. 2. 87 Sims-Williams, ‘Ancient Afghanistan and its Invaders’.

343

344

k ho da da d r e z akh an i

both sides of the Hindu Kush under Kushan, Hunnic, and Turkic rule is crucial for understanding the language interactions in the region. Analysis of such interactions requires close attention to the little non-textual evidence we have of the use of languages in the eastern Iranian world. A case in point is coins that proliferated under Turkic suzerainty in eastern Iran. Two examples demonstrate the complicated dynamics of language interactions. The first is a coin issued by an authority who has been dubbed the Yabghu of Tokharistan. Issued on the southern side of the Hindu Kush, possibly in the area of Zabulistan (modern Ghazni), the coin bears a Bactrian inscription. In fact, the coin itself is a marvel of multilingual communication, using a Turkic title for a ruler who identifies himself as the ruler of Tokharistan, to the north of the Hindu Kush, while issuing a coin in Zabul, to the south of the mountain range, where previous rulers had used Middle Persian, Bactrian, and Brahmi in their coin inscriptions. A second example is the coins of Tegin and his son Phrom Kesar, rulers of Kabulistan and Zabulistan in the early eighth century ce. Issuing coins carrying inscriptions in Bactrian, Brahmi, and Middle Persian, these authorities identified themselves as rulers of the entire region and claimed lofty titles such as Khūrāsān Shāh, ‘the King of the East/Khurasan’, and Phrōm Kēsār, ‘the Caesar of Rome’.88 I have elsewhere suggested that these titles mark a process of memory creation, giving us a view into the political ideo­logy of these rulers. I believe that the use of these titles, adopted from the Sasanians and the Byzantines, was a way for local rulers to claim inheritance and authority from the vanquished powers of the past. It is probable that the greatly reduced remnants of the (Eastern) Roman Empire were unknown to them, and that for rulers such as Phrom Kesar, the memory of the Byzantines and the Romans, alongside that of the Sasanians, belonged to a past they found obliged to preserve and useful to claim. This was a past to which the rulers of eastern Iran clearly laid claim, seeing themselves as protectors of its memory and authority. This perspective might be reflected in a Middle Persian inscription on the coins of Phrom Kesar, which claims that he has ‘vanquished the Arabs’.89 But what about the use of Sasanian Middle Persian on the coins of the early Islamic rulers of the East? The use of Pahlavi script and Middle Persian has at least one precedent in the coinage of the East, in the issues of the Nezak Shah rulers of Zabulistan and Kabulistan, which date back to the fifth century ce.90 The coins of these authorities, minted in Ghazni and Kabul, bear inscriptions in Pahlavi that identify the rulers as nycky MLK’ ‘Nezak Shah’ and that thus set these issues apart from the coinage of the Iranian Huns to the south of the Hindu Kush, which commonly used Bactrian or Brahmi. The style of the Nezak issues is also quite different from that of other Iranian Hunnic coins, with elaborate headdresses and

88 Vondrovec, Coinage of the Iranian Huns and their Successors from Bactria to Gandhara, pp. 553–55; on Phrom Kesar, see Piras, ‘Fromo Kesaro’. A new inscription bearing this name has now appeared in Jaghori, Afghanistan; see Sims-Williams ‘The Batrian Inscription of Jaghori’. 89 Vondrovec, Coinage of the Iranian Huns and their Successors from Bactria to Gandhara, pp. 447–90. 90 Alram Das Antlitz, pp. 105–11.

navi gat i ng pe rsi an

fine workmanship on the obverse and reverse of the coins.91 A later continuation of this coin series, known as the Alkhan-Nezak Crossover Series, does away with Middle Persian and replaces it with Brahmi, at least in the region of Kabul and probably under the influence of Alkhan rulers who conquered Kabul from Gandhara.92 So we might consider the use of Pahlavi on the coinage of the Nezak Shah series a political statement, perhaps indicating a certain affinity with the Sasanian court following the conquest of the region by Shapur II in the middle of the fourth century and a continuing legacy of Sasanian rule in the region. This rule might have established Pahlavi as the administrative language of the Sistan and Zabulistan region, and it was subsequently inherited by the Nezak Shahs and maintained in the region. Similarly, the use of Pahlavi on the coins of Tegin and Phrom Kesar also shows an aspiration to Sasanian administrative authority and organization. Prior to the advent of Islam, the language of the documents that we possess from the area of eastern Iran is a legally developed and advanced Bactrian. In fact, as far as texts are concerned, we have until recently known very little about the use of Middle Persian as an administrative language in the Sasanian period beyond what could be gleaned from the papyri from early seventh-century Egypt. Most of our evidence of Sasanian administrative structures consists of epi­graphic information in the form of seals and sealings, which draw a picture of Sasanian administration and administrative geo­graphy.93 However, recent discoveries from the late seventh and early eighth centuries, after the demise of the Sasanian imperial system, have brought a wealth of documents to our attention. Geo­graphically focused on the regions of Qom and Mahallat in north-central Iran and in Tabarestan, south-east of the Caspian Sea, these documents offer rich information about the legal termino­logies, administrative divisions, and economic settings of these regions. Although Ibn al-Nadim portrays these regions as part of Fahla (and thus their inhabitants as speakers of Fahlavi),94 the presence of MP documents there shows the penetration of Sasanian administrative language — the origin of Dari — in the regions.95 The connection is also indicated by the fact that the authority issuing and confirming legal documents in Tabarestan identifies himself as the Ispahbed of Khurasan (‘generalissimo of the eastern quarter’),96 reflecting Sasanian administrative structures. We can perhaps conclude from this that the Sasanians indeed conducted their administration in

91 See Vondrovec, ‘Coinage of the Nezak’, pp. 169–90, and Vondrovec, Coinage of the Iranian Huns and their Successors from Bactria to Gandhara, i, pp. 447–60. 92 Alram 2016, 113–15; Vondrovec, Coinage of the Iranian Huns and their Successors from Bactria to Gandhara, pp. 479–88. 93 Gyselen La géo­graphie administrative. 94 On Fahlavi and its problematic status, see Sadeghi, ‘Yek maʿni nashenakhte kalame pahlavi dar zaban-e farsi’; see also Firouzbakhsh, ‘Fahlaviyat-e tarikh-e Qom’. 95 See Macuch, ‘Pahlavi Legal Documents from Tabarestān on Lease, Loan and Compensation’ and ‘Pahlavi Legal Documents from Tabarestān: Two Claims’ for the legal setting. 96 Gyselen, The Four Generals of the Sasanian Empire.

345

346

k ho da da d r e z akh an i

Middle Persian within their territories, including later acquisitions such as Egypt, and that they developed a well-stocked legal and administrative jargon. What, then, can this tell us about the use of Middle Persian on the coinage of the rulers of Zabulistan and Kabulistan in the early eighth century? It appears that the use of Middle Persian and the Pahlavi script, apart from continuing the tradition of the Nezak Shahs and other authorities, also shows a desire to claim a central authority as representatives of the Sasanian court. Much like the use of the titles of Shah and (Roman) Caesar, the use of Middle Persian lays claim to Sasanian administrative authority and control, extending the reach of the government of Ctesiphon to the farthest reaches of the Sasanian realm and even beyond its immediate political reach. Perhaps it is not too far-fetched to also consider the use of ‘Khurasan’ in designating the territory over which Tegin ‘Khurasan Shah’ ruled. Rule over Khurasan was, of course, an important early Islamic claim. As noted above, the highest authority in the Pahlavi documents from Tabarestan claims the Sasanian title of ‘Ispahbed of Khurasan’.97 So, Tegin’s assertion that he is the ‘Shah’ of Khurasan shows an even higher claim to the same authority, incorporating Sasanian administrative as well as royal titulature. The use of Middle Persian on the coinage of the local rulers of Zabulistan is thus one of the earliest pieces of evidence for the presence of Sasanian administrative language, the origin of Ibn al-Nadim’s Dari in the area of ‘Khurasan and the East’. Whether such use can be detected in the area north of the Hindu Kush, particularly in the Balkh region, cannot be determined with certainty at present. The presence of the language in Zabulistan and even Sistan, however, fits well with the earliest evidence we have of New Persian poetry, which comes from Sistan and its vicinity.98 Additionally, the importance of New Persian in the regions previously dominated by Bactrian, as evidenced by Maqdisi’s statement on the purity of the Persian of Balkh,99 is a sign of the importance of the East to this particular form of Persian. In fact, although the statement might sound out of place in the context of Balkh, it is supported by a quotation in the Persian Fazaʾel-e Balkhin which Nadhr b. Shamir states that ‘the language of the people of Balkh is pārsī-ye darī’.100 Sogdian

Sogdiana, and in fact the entire Transoxiana (the land between the Oxus and the Jaxartes), is known as the homeland of a well-developed CNP literary tradition.101 In contrast to the early poets from Sistan and Herat, whose works survive only in scattered quotations in later histories, the Transoxianan literary masters are

97 See Rezakhani, ‘Pangul and Bunji, Zhulad and Fulad’ for a detailed discussion of the use of the title and the understanding of ‘Khurasan’ in this period. 98 Lazard, ‘The Rise of the New Persian Language’; Nezami Aruzi, Chahar maqala. 99 See above and Lazard, ‘Pârsi et dari’, p. 241. 100 Vaez Balkhi, pp. 29–30. 101 Paul, ‘Persian Language i. Early New Persian’.

navi gat i ng pe rsi an

well recorded and widely recognized. Rudaki’s divan, the history of Balʿami, and translations of Tabari’s Tafsir are evidence of a robust literary world that benefited from royal patronage by the Samanids and their successors. Indeed, for the Samanid, Ziyarid, and later Ghaznavid patrons of classical Persian composers, the literary movement in New Persian was as much of a political project as the Abbasid translation movement had been for the Abbasids 150 years earlier. But the tenth-century perfection of classical Persian literature in Transoxiana was based on the earlier introduction of a spoken form of the language to the area. An anecdote in Narshakhi’s History of Bukhara perhaps demonstrates this most clearly: ‫قتیبه بن مسلم مسجد جامع بنا کرد اندر حصار بخارا بسال نود و چهار و آن موضع بتخانه بود مر‬ ‫ چنانکه هر آدینه منادی فرمودی هرکه بنماز‬،‫ فرمود تا هر آدینه در آنجا جمع شدنی‬.‫اهل بخار را‬ ‫ و مردمان بخارا به اول اسالم در نماز قرآن بپارسی خواندندی‬.‫ دو درهم بدهم‬،‫آدینه حاضر شود‬ }‫ مردی بود در پس ایشان که بانگ {زدی‬،‫و عربی نتوانستندی آموختن و چون وفت رکوع شدی‬ 102...‫بکنیتا نکینت و چون سجده خواستندی {کردی} بانک کردی نکونیا نکونی‬

Qutayba b. Muslim built a congregational mosque inside the walls of Bukhara in the year 94 [ah; 712–13 ce], and in that place there was [previously] an idol temple for the people of Bukhara. He said that there must be a gathering there every Friday, and he declared, ‘I will give two dirhams to anyone who shows up to the Friday prayers’. And the people of Bukhara, at the beginning of [their conversion to] Islam, recited the Quranic [verses] in their prayers in Persian and could not learn Arabic, and when it was time for rukūʿ, there was a man behind them who would shout ‘BKNITA NKINT’, and when they wanted to do sujūd, [he] would shout ‘NKUNYA NKUNY’.103 The capitalized phrases at the end are written in the same form in the Persian translation of Narshakhi’s history; they make no sense and are clearly marked as not Persian.104 The passage as a whole suggests that for the people of Bukhara, saying their newly learned prayers in Persian was deemed more manageable. At the same time, however, Persian was not seen primarily as a spoken medium, and the instructions to bow and prostrate thus had to be given in the local tongue, perhaps the Bukharan dialect of Sogdian. The passage further suggests that the Islamic prayer had already been translated into Persian,105 most likely before Qutayba’s conquest, and that the Persian version was in use in other places.106 There is no

102 Narshakhi, Tarikh-e Bokhara, p. 57. 103 Translation by the present author. For a standard translation, see Frye The History of Bukhara, p. 48, who transcribes the final phrases as bknītā nkīnt and nkūnīā nkūnī. I prefer to leave them in the original transliteration, pending a definite study of their meaning. 104 For an interpretation of the meaning of these phrases, see Frye The History of Bukhara, pp. 135–36 n. 184, quoting a personal correspondence with W. B. Henning and a reference to ‘Rozenberg 1921’, but agreeing that they are imperative phrases referring to bowing down and downward. Part of the issue, as Henning has pointed out, is perhaps the normalization of the words by the Persian translator of Narshakhi’s book, and perhaps dialectal differences in the Bukharan dialect and ‘standard’ Sogdian. 105 Zadeh, The Vernacular Qur’an, pp. 302–30. 106 On Qutayba’s conquests and their aftermath, see Guzmán, ‘La expansión árabe-musulmana hacia Ma

347

3 48

k ho da da d r e z akh an i

suggestion here that the prayer was translated specifically for this occasion; instead, it appears that the Bukharans were given permission to use a translation in place of the Arabic prayer and that they made use of it. The prevalence of local languages, in this case Bukharan, in the early Islamic period is also observable from the widespread Sogdian documentary culture in the region.A local king of Sogdiana named Dewāštīč has left us a cache of documents produced around the time of Qutayba’s conquest of Bukhara at Mount Mugh, a few hundred kilometres east of Bukhara.107 Dewashtich was the local ruler of Panjkent in Sogdiana, less than seventy kilometres to the east of Samarqand, Sogdiana’s main city. This cache, which mostly consists of Dewashtich’s correspondence but also contains legal contracts, was found almost by accident during excavations of the Mount Mugh Fortress near Panjkent and was published by Soviet scholars. The archive shows that Dewashtich claimed the kingship of Sogdiana following the demise of the position after the capitulation of Ghurak and his family to the Muslim armies. There is no need to prove the antiquity of Sogdian written tradition given the long history of its use as a legal, commercial, and religious language,108 but the Mount Mugh Archive is nonetheless important as evidence of the continuity of the use of written Sogdian in the early eighth century. The developed state of the archive’s linguistic norms and patterns, correspondence style, and use of legal termino­logy demonstrates the strength of Sogdian in this period. Although certain borrowings from other languages — Turkic, Bactrian, Middle Persian, and even Chinese — are evident, Sogdian as a whole does not appear to have been heavily influenced by philo­logical or socio-linguistic norms from other languages. Apart from Sogdian and before the full-fledged appearance of New Persian, we have evidence of the use of Arabic and Middle Persian in this region at a scribal school (dabīristān) in Marv, which lay halfway between Nishapur and Bukhara and was the capital of Islamic Khurasan. The late Vladimir A. Livshits proposed that the appearance of MP, dated to ‘no later than the mid-eighth century’ in this school’s archive, is related to the weakening state of Sogdian and the scribes’ perceived need to learn (Middle) Persian alongside Arabic and Sogdian in the early Islamic period.109 But this suggestion is historically problematic on several grounds. First, it assumes the antiquity of Sogdian in the Marv region, which can hardly be proven or even reasonably assumed. Marv could, in fact, have been part of the Bactrian language sphere as much as it was that of Sogdian and, naturally, that of Parthian. Second, the suggestion that the use of MP on the ostraca from Marv signals the waning of Sogdian would imply that Pahlavi Middle Persian gained prominence in the Marv oasis only after the Muslim conquest of the region, which is unlikely. In the absence of an independent sub-Sasanian power (such as wara’ al-nahr’, pp. 58–62 for a short but creative summary. Haug, The Eastern Frontier, pp. 111–19. offers an in-depth analysis. 107 For an edition of these documents, see Freiman and others, Sogdiiskie dokumenty s gory Mug, now partially available in English as Livshits, Sogdian Epi­graphy of Central Asia and Semirech’e. 108 Grenet and Sims-Williams, ‘The Historical Context of the Sogdian Ancient Letters’. 109 Livshits, Sogdian Epi­graphy of Central Asia and Semirech’e, pp. 54–55.

navi gat i ng pe rsi an

those in Tabarestan), there is no reason to assume that the Sasanian administrative language was newly introduced to the area in the Islamic period, particularly as Marv was quickly developing into a regional centre of Islamic presence. Third, if we accept that the rise of Middle Persian (‘Pahlavi’) writing in the scribal school of Marv was related to the rise of New Persian in the region, we must then presume that NP was actively written in the Pahlavi script and needed as to be taught in a scribal school. There is simply no evidence for such active writing of New Persian in the Pahlavi script in this period, least of all in the Marv region. I propose, then, that the appearance of Middle Persian on the ostraca, objects that are by nature easy to move, is an isolated event, perhaps having to do with the change in the administrative system in the region, and that it can simply be considered a vestige of the fading Sasanian authority in the area. Instead, Sogdian shows close interactions with New Persian, possibly as a result of the presence of Sasanian soldiers in the region. In this context, we can see the role of Sogdian as the source of many lexical borrowings into New Persian, some of which appear to replace already existing words in Middle Persian.110 As a whole, Sogdian appears to have been of central importance in Sogdiana itself, with various dialects spoken as far west as Bukhara. However, perhaps because of a lack of strong government in Sogdiana, its political influence outside Sogdiana seems to have been limited. At the same time, the reach and influence of the language as a vehicle of Sogdian mercantile activities as far east as Korea is undeniable, and its prevalence as the language of Manichaean and the Christian churches of Turkistan has been established.111 But in its homeland, Sogdian appears to have been influenced by the rising importance of Turkic languages as the result of the advent of the Western Türk and Türgesh empires. In the Sogdian realm, the Turkic-Sogdian interaction was visible in political and social settings after the seventh century ce. The Turkic control of politics in the region is visible in both textual and artistic sources.112 Simultaneously, however, Sogdians influenced the Türk immensely. Beyond commerce, the Sogdians carried their language and script into the Türk chancellery and further influenced the Turkic world by lending their script to the writing of various Turkic languages.113 But Turkic influences on Sogdian are also obvious in the various documents found in the Dun Huang caves.114 They include morpho­logical influences and syntactical changes in written Sogdian, which show that the documents were most probably written by Turkic/Uighur speakers, perhaps the last preservers

110 Henning, ‘Sogdian Loan-Words in New Persian’, as well as Yakubovich, ‘Nugae Sogdicae’ and ‘Nugae Sogdicae II’. 111 Here I mean only the western extent of the reach of Sogdian; I am not concerned with its spread as the lingua franca of the Sogdian trade network eastward to China, for which see La Vaissière, Sogdian Traders. 112 La Vaissière, Sogdian Traders, pp. 200–02. 113 Yoshida, ‘When Did Sogdians Begin to Write Vertically?’. 114 Sims-Williams and Hamilton, Documents turco-sogdiens du ixe–xe siècle de Touen-houang.

349

3 50

k ho da da d r e z akh an i

of Sogdian.115 The origins of the Turks, generally, and the Türk empire that came to dominate Sogdian, specifically, have already been well studied.116 With the conquest of Sogdiana by the Western Türk, however, this Central Asian element penetrated the Sogdian political, economic, and linguistic milieu completely. The result was a Turko-Sogdian fusion that made Sogdians the masters of commerce in the Türk empire and that continued to play a role in China well into the eighth and ninth centuries.117 In this context, traces of early Turkic in New Persian still need to be studied, but they have been the subject of speculation by some scholars.118 Turkic titles such as Khan, Tarkhan, and Iltäbär (NP Rutbil)119 may have been borrowed into NP from either Sogdian or possibly Bactrian — an issue that can be studied better on the basis of the Bactrian economic documents. The rise of ENP in the lands previously occupied by Sogdian and Bactrian120 then points to a deeper level of integration in the region. The Parthian-Middle Persian Continuum

Of all the ‘eastern’ Iranian languages with which Persian has interacted, one stands apart. Although Parthian, the language of the Arsacid king of kings, is geo­graphically focused on the eastern half of today’s Iran, it was nonetheless a West Middle Iranian language (and located west of the East Iranian languages). Unlike Sogdian, Bactrian, and Khwarazmian, Parthian belonged to the same group of Iranian languages as Persian and Median, and thus its interaction with the imperial language of the Sasanians, Middle Persian, was of a different nature than its interactions with the East Iranian languages. This fact offers a good point at which to briefly discuss the position of Parthian in relation to (Middle) Persian and their continued interactions. Quite peculiarly, the flourishing of Parthian paralleled the rise of the Sasanians in the third century ce. Although the basis of our understanding of the Parthian language during the Arsacid period is limited to the Nisa ostraca collection from the early Arsacid period121 and the legal documents of the Avroman letters,122 there 115 See Yoshida, ‘In Search of Traces of Sogdians’, pp. 193–94, and Krippes, ‘Sociolinguistic Notes’. 116 See Sinor, ‘The Establishment and Dissolution of the Türk Empire’, and of course a complete and handy treatment by Peter Golden in An Introduction to the History of the Turkic Peoples, pp. 115–87, including the history of the Uighurs, whose interactions with the Sogdians resulted in interesting linguistic outcomes. 117 For a survey of this, see La Vaissière, Sogdian Traders, pp. 199–225, and further studies, including La Vaissière, Samarcande et Samarra. 118 Utas, ‘A Multiethnic Origin of New Persian?’. 119 Sims-Williams, ‘Ancient Afghanistan and its Invaders’, p. 235. 120 Cf. Maqdisi’s statement about the dialect of Balkh: see Gignoux les quatre, 241; Sadeghi, Takvin-e zaban-e farsi, p. 43. 121 Diakonoff and Livshits, Parthian Economic Documents from Nisa and subsequent publications, including Livshits, ‘New Parthian Documents from South Turkmenistan’ and Livshits and Pilipko, ‘Parthian Ostraca from the Central Building Complex of Old Nisa’. 122 Nyberg, ‘The Pahlavi Documents from Avroman’.

navi gat i ng pe rsi an

appears to have been a great Parthian revival under the Sasanians. This should be mostly credited to the Prophet Mani and his immediate followers, who chose Parthian as the written medium of many of their works, particularly after their flight from Sasanian persecution to Transoxiana.123 However, even the Sasanian rulers Shapur I and his son Narseh chose to represent Parthian as an imperial language on a par with Middle Persian in their respective inscriptions.124 Whereas Parthian, because of its connection with the Arsacid royal house, is generally associated with the area of the north-east, the Sasanian kust of Khwarasan/ Khurasan, some of its most important monuments are found in the west of the empire. Beyond the Avroman letters, which were found in the area of Kurdistan, a great part of our knowledge of Parthian comes from the area commonly known as Media, or the Mah of the early Islamic geo­graphers.125 Evidence from this region includes the Paikuli inscription of Narseh as well as the linguistic influences left by Parthian on languages such as Armenian and Georgian.126 In the early Islamic context, perhaps one of the most interesting items of evidence for the ‘western’ presence of Parthian is the first part of Ibn al-Nadim’s statement quoted above, where he says that ‘the Pahlavi (al-fahlawiyya) is related to Pahlav (Fahla), a region that contains five cities: Isfahan, Rayy, Hamadhan, Mah Nihavand, and Azarbayjan’.127 We know the term ‘Pahlavi’ primarily as the name of the script used in the writing of Zoroastrian Middle Persian texts. However, the name is originally a take on the word Parthaw (Parthian) and, strictly speaking, denotes the language and script of the Parthian language or, more specifically, the Parthian language as it was known in the Arsacid period.128 Ibn al-Nadim’s statement presumably demonstrates the localization of Pahlavi/Parthian in the centre, west, and north-west of present-day Iran, precisely the area of Media/Mah.129 This is also the area where the remnants of various local languages, ranging from the so-called Tati/Taleshi130 continuum to the language of medi­eval poems known as Fahlaviyat, held sway.131 The study of the connection between these languages and Parthian proper is still a growing 123 Boyce ‘Manichaean Literature’ and A Reader in Manichaean Middle Persian and Parthian; Heuser and Klimkeit, Studies in Manichaean Literature and Art. 124 For Shapur’s ŠKZ inscription, see Huyse, Die dreisprachige Inschrift Šābuhrs I. an der Kaʿba-i Zardušt; for Narseh’s Paikuli inscription, see Humbach and Skjærvø, The Sassanian Inscription of Paikuli. 125 Most significantly, the Mah of Kufa (Hamedan) and the Mah of Basra (Nihavand); Ibn Hawqal, Kitab al-Masalik wa-l-mamalik, p. 255. 126 Meyer, ‘Armeno-Iranian Structural Interaction’, as well as Meyer (forthcoming). 127 Other than the above, see Firouzbakhsh, ‘Fahlaviyat-e tarikh-e Qom’, p. 138 n. 72 for all the different meanings of Fahla as a geo­graphical term; see also Sadeghi, Takvin-e zaban-e farsi, pp. 8–9. 128 Cereti, ‘Middle Persian Literature i’. 129 The Muslim geo­graphers often divided this area between Azerbaijan (plus Armenia), on the one hand, and Jibal (mountains), which included the Mahat, on the other (Ibn Hawqal, Kitab al-Masalik wa-l-mamalik, p. 255). 130 See Yarshater, A Grammar of Southern Tati Dialects and other works by Yarshater, most of which are referenced in Windfuhr, ‘Eštehārdī’. 131 Tafazzoli, ‘Fahlaviyat’. The research on Fahlaviyat and its linguistic categorization, including whether Fahlaviyat can be considered part of the literature of the same language, is new and emerging; for a relatively recent example, see Firouzbakhsh, ‘Fahlaviyat-e tarikh-e Qom’, pp. 130–40.

35 1

3 52

k ho da da d r e z akh an i

field, but it is perhaps worthwhile to consider the presence of Parthian in the west, as opposed to its eastern, Khurasani ‘homeland’.132 We should also keep in mind that outside Media/Mah, the other major monuments of the Parthian/Pahlavi language are to be found in Fars/Persis (the inscription of Shapur I) and amongst the writings of Mani, a native of Babylonia who possibly had a Parthian background.133 The interactions between Parthian and Middle Persian may become clearer when we consider the subsequent history of both languages. The early Sasanian kings obviously made a point of using Parthian alongside Middle Persian in order to emphasise some form of continuity with Arsacid power. Whereas the inscription of Shapur I at Kaʿba-i Zardusht (ŠKZ) uses both Parthian and Greek alongside and even prior to MP, the inscription of Narseh at Paikuli omits Greek but preserves Parthian. A comparison of the two texts shows the close structural parallels between Parthian and MP;134 at many points, the difference between lies purely in the scripts used. Both scripts are based on modified forms of Imperial Aramaic, a remainder of Achaemenid administration, and both rely heavily on Arameograms to convey frequent words.135 But even in instances in which words are spelled out, the inscriptions rely on cognates — often quite close ones — for concepts. The two inscriptions offer, in fact, quite an interesting case study, as this is the only instance of a pair of such close parallel texts in the various West Middle Iranian languages. The interaction of MP and Parthian continues in Manichaean texts, although close parallels like the inscriptions are rare. Still, Manichaean Middle Persian texts appear to have preserved a slightly different dialect of MP than did the later Zoroastrian Middle Persian texts. The Manichaean texts, perhaps representing a dialect of MP originating further east than the Zoroastrian MP dialect, may have also had more connections with Parthian, as it preserved certain phono­logical features that were abandoned in Zoroastrian MP but have remained in New Persian.136 These include the voiced velar fricative (Ɣ), a sound that developed into a voiced labio-velar approximate (w) in Zoroastrian MP.137 It is also worth noting that in its ‘homeland’ of Khurasan, Parthian was eventually fully supplanted by ENP — indeed, Khurasan became very much the homeland of Classical New Persian.138 Subsequently, certain Parthian words and even sounds such as the voiced palato-alveolar fricative (ʒ) survived in New Persian and became part of the phono­logy of the language. These Parthian words, such as harās ‘fear’, still exist in the language alongside their original Persian cognates (i.e. tars), contributing to the total lexical reserves of the language. Therefore, the

132 See Sadeghi, Takvin-e zaban-e farsi, pp. 8–11 on the name and geo­graphical reach of Parthian. 133 Sundermann, ‘Mani’. 134 Sadeghi, Takvin-e zaban-e farsi, p. 19. 135 Skjærvø, ‘Iran vi. Iranian Languages and Scripts’. 136 Tedesco, ‘Dialekto­logie der westiranischen Turfantexte’, pp. 184–258. 137 For a more detailed discussion of the phono­logical and morpho­logical system of ENP, see Sadeghi, Takvin-e zaban-e farsi, pp. 131–39. 138 Sadeghi, Takvin-e zaban-e farsi, pp. 10–11.

navi gat i ng pe rsi an

role of Parthian within New Persian needs to be seen as that of a close partner of Middle Persian and a parallel dialectal form, preserving forms that continued into Classical New Persian.

Where Is Persian? The complicated picture I have drawn above prompts an important question: What is Persian, in the end? And where is Persian located in this linguistic hodgepodge? The lexical treasury of Persian, including its rich borrowings from other Iranian languages, does at times drive one to think of it as a hybrid language, a sort of creole made up of the conjunction of many languages. Philo­logists would, of course, argue that the morpho­logical, phono­logical, and syntactic integrity of the language leaves no doubt that New Persian fits well within the continuum of Persian and is a proper successor to Old and Middle Persian. At the same time, the linguistic affinity of Persian with the languages around it at any given stage of its history is a great demonstration of its nature as a language in transition. Early New Persian, as the common spoken tongue of Sasanian Asorestan/Iraq, represented the cosmopolitanism of the Sasanians themselves. In an area where Aramaic dialects and, increasingly, Arabic were the common languages and where Arabic eventually came to dominate, Dari/ENP was transformed from the vernacular of the late Sasanian period to the prestige language of the early Islamic one. In the late Sasanian period, ENP was the language of the slowly emerging dihgān class, petty landowners who often negotiated with the Muslim conquerors when the latter arrived in Asorestan and further east on the plateau. By absorbing features of the languages with which it came into contact, ENP/Dari eventually developed into an in-between tongue that represented no single dialect but preserved many influences. Its arrival and establishment in the East in the early eighth century ce as a cosmopolitan vernacular were facilitated by an already existing form of the language. This was a Partho-Persian language continuum that represented a dialect close to ENP and further interacted with the imported language of Iraq, easing it into the linguistic scene of the north-east. However, ENP was transformed when it came into contact with East Iranian languages such as Sogdian and Bactrian and was adopted as the language of the new Islamic administration. Although early anecdotal evidence shows the parallel existence of local languages such as Bukharan with Dari, it was eventually Dari that won the position of power — absorbing features of the native tongues in the meantime — and developed further. Its success was also partly a function of its role as an urban prestige language representing the centres of the new administration’s power. Here, the interactions of ENP with competing languages, including Sogdian (and its Turkic substratum in Transoxiana), were also part of the challenges it faced. The picture drawn here represents a way of thinking about Early New Persian not as a monolithic language or a simple continuation of ‘Old’ and ‘Middle’ forms of it but as a cosmopolitan tongue of the many. Although its relationship with the structures and spheres of power, both the Sasanian and the early Islamic, is an

35 3

3 54

k ho da da d r e z akh an i

important element of its rise, it was in fact its great ability to serve as the language of a wide variety of people that guaranteed its success. A language rooted in the south-west of Iran, greatly simplified in its grammar and open to the absorption of influences from Aramaic, Arabic, Sogdian, and Turkic, it managed to dominate even outside official avenues of power, to obtain a prime position in various environments, and eventually even morph into a rich literary language.

Biblio­graphy Primary Sources Al-Baladhuri, Abu Hasan Ahmad Yahya. Futuh Al-Buldan, ed. by Michael J. de Goeje (Leiden: Brill, 1866) —— , The Origins of the Islamic State, Being a Translation from the Arabic, Accompanied with Annotations, Geo­graphic and Historic Notes of the Kitâb Fitûh Al-Buldân of Al-Imâm Abu-l Abbâs Ahmad Ibn-Jâbir Al-Balâdhuri, trans. by Philip Khuri Hitti. (New York: Columbia University Press, 1916) Ibn Hawqal, Abu al-Qasim, Kitab al-Masalik wa-l-mamalik, ed. by M. J. de Goeje (Leiden: Brill, 1873) Ibn al-Nadim, al-Fihrist, ed. by Reza Tajaddod, (Tehran: n.p., n.d.) Ibn Nadim, Muhammad b. Ishaq. Al-Fehrest, trans. by Mohammad Reza Tajaddod (Tehran: Donya-ye Ketab, 1394 ah) Jahiz, ʿAmr b. Bihr, Al-Bayan wa-l-Tabyeen, ed. by Hassan Efendi Fakehani, (Cairo: alMatbaʿat al-ʿilmiya: 1311 ah). Narshakhi, Abu Bakr Muhammad b. Jafar, Tarikh-e Bokhara (History of Bukhara), ed. by Mohammad-Taqi Modarres Razavi (Tehran: Tous, 1387 ah) Nezami Aruzi, Chahar maqala, ed. by Mohammad Moein (Tehran: Armaghan, 1331/1952) Vaez Balkhi, Abu Bakr Abdullah, Fazaʾel-e Balkh, ed. by Abdolhayy Habibi (Tehran: Bonyad-e Farhang-e Iran, 1350 ah) Secondary Works Abolghassemi, Mohssen, Zaban-e farsi o sargozasht-e an o goftarha-ye digar, ed. by Mohammad Emami Malekshah (Tehran: Tahouri, 1394 sh) al-Qādị̄, Wadād, ‘The Names of Estates in State Registers before and after the Arabization of the Dīwāns’, in Umayyad Legacies, ed. by Antoine Borrut and Paul M. Cobb, (Leiden: Brill, 2010), pp. 255–80 Alram, Michael, Das Antlitz des Fremden: Die Münzprägung der Hunnen und Westtürken in Zentralasien und Indien. (Vienna: Österreichische Akademie der Wissenschaften, 2016) Amanat, Abbas, and Assef Ashraf, eds, The Persianate World: Rethinking a Shared Sphere (Leiden: Brill, 2018) Ando, Clifford, Roman Social Imaginaries: Language and Thought in the Context of Empire (Toronto: University of Toronto Press, 2015)

navi gat i ng pe rsi an

Andreas, Friedrich Carl, and Walter Bruno Henning, Mitteliranische Manichaica aus Chinesisch-Turkestan, Sitzungsberichte der Königlich Preussischen Akademie der Wissen­schaften zu Berlin, Philosophisch-Historische Klasse, 10 (Berlin: De Gruyter, 1934) Ansaldo, Umberto, Stephen Matthews, and Lisa Lim, Deconstructing Creole (Amsterdam: John Benjamins, 2007) Arjomand, Saïd Amir, ‘Persianate Islam and its Regional Spread’, in Religions, Nations, and Transnationalism in Multiple Modernities, ed. by P. Michel, A. Possami, and B. Turner (New York: Palgrave Macmillan, 2017), pp. 67–84 Asmussen, Jes Peter, Manichaean Literature: Representative Texts Chiefly from Middle Persian and Parthian Writings (Delmar, NY: Scholars Facsimiles and Reprints, 1975) Azarnouche, Azartache, Chalesh-e miyan-e farsi o arabi: sadeha-ye nokhost (Tehran: Nashr-e Ney, 1385 sh) —— , ‘La formation du persan sous l’influence de la langue arabe au iv/xe siècle’, in Iran: questions et connaisances, ed. by Maria Szuppe, Studia Iranica, 26 (Paris: Association pour l’Avancement des Études Iraniennes, 2002), ii, pp. 11–19 Bailey, Harold Walter, Zoroastrian Problems in the Ninth-Century Books (Oxford: Clarendon Press, 1943) Bosworth, Clifford Edmund, Sistan under the Arabs, from the Islamic Conquest to the Rise of the Saffarids (30–250/651–864) (Rome: Istituto Italiano per il Medio ed Estremo Oriente, 1968) —— , ‘The Development of Persian Culture under the Early Ghaznavids’, Iran, 6.1 (1968), 33–44 Bosworth, Clifford Edmund, ‘The Ṭāhirids and Persian Literature’, Iran, 7.1 (1969), 103–06 Boyce, Mary, ‘The Parthian Gōsān and Iranian Minstrel Tradition’, Journal of the Royal Asiatic Society of Great Britain and Ireland, n.s., 89.1–2 (1957), 10–45 —— , ‘Middle Persian Literature’, Handbuch der Orientalistik, 1.2.1 (1968), 31–66 —— , The Letter of Tansar (Rome: Istituto Italiano per il Medio ed Estremo Oriente, 1968) —— , ‘The Manichaean Literature in Middle Iranian’, Handbuch der Orientalistik, 1.4.1 (1968), 67–76 —— , A Reader in Manichaean Middle Persian and Parthian: Texts with Notes, (Leiden: Brill, 1975) Brunner, Christopher J., A Syntax of Western Middle Iranian, Persian Studies Series, 3 (Delmar, NY: Caravan Books, 1977) Cantera, Alberto, Studien zur Pahlavi-Übersetzung des Avesta, vi (Wiesbaden: Otto Harrassowitz, 2004) Cereti, Carlo G., ‘Middle Persian Literature i. Pahlavi Literature’, in Encyclopaedia Iranica (online edn, 2009) —— , ‘Two Late Pahlavi Inscriptions from the Alborz Mountains’, in Translation and Transmission: Collection of Articles, ed. by Jaakko Hämeen-Anttila and Ilkka Lindstedt (Münster: Ugarit-Verlag, 2018), pp. 55–94 Compareti, M., ‘The Last Sasanians in China’, Eurasian Studies, 2.2 (2003), 197–213 Daryaee, Touraj. ‘The Effect of the Arab Muslim Conquest on the Administrative Division of Sasanian Persis/Fars’, Iran, 41.1 (2003), 193–204 Daryaee, Touraj, ‘Middle Persian (Pahlavi)’, in A Companion to Late Antique Literature, ed. by Scott McGill and Edward J. Watts (New York: Wiley, 2018), pp. 103–20

35 5

356

k ho da da d r e z akh an i

De Blois, François, ‘A Persian Poem Lamenting the Arab Conquest’, in Studies in Honour of Clifford Edmund Bosworth, ed. by I. R. Netton (Leiden: Brill, 2000), pp. 82–95 De Jong, Albert, ‘Pāzand and “Retranscribed” Pahlavi: On the Philo­logy and History of Late Zoroastrian Literature’, in Persian Origins: Early Judeao-Persian and the Emergence of New Persian, ed. by Ludwig Paul (Wiesbaden: Otto Harrassowitz, 2003), pp. 67–78 De Menasce, Jean, ‘Zoroastrian Pahlavi Writings’, in The Cambridge History of Iran, iii, pt 2, ed. by Ehsan Yarshater (Cambridge: Cambridge University Press, 1983), pp. 1166–95 Diakonoff, Igor M., and Vladimir A. Livshits, Parthian Economic Documents from Nisa, Corpus Inscriptionum Iranicarum, pt 2: Inscriptions of the Seleucid and Parthian Periods and of Eastern Iran and Central Asia, ii: Parthian, 1 (London: School of Oriental and African Studies, 1976) Durkin-Meisterernst, Desmond, ‘Erfand Mani die manichäische Schrift?’, in Studia Manichaica IV: Internationaler Kongreß zum Manichäismus, Berlin, 14.–18. Juli 1997, ed. by Ronald Emmerick (Berlin: Akademie Verlag 2000), pp. 161–78 —— , Grammatik des Westmitteliranischen: Parthisch und Mittelpersisch (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2014) Eaton, Richard M., ‘The Persian Cosmopolis (900–1900) and the Sanskrit Cosmopolis (400–1400)’, in The Persianate World, ed. by Abbas Amanat and Assef Ashraf (Leiden: Brill, 2018), pp. 63–83 Firouzbakhsh, Pejman, ‘Fahlaviyat-e tarikh-e Qom’, Iran nameh, 2nd ser., 1.1 (2016), 130–40 Forstner, Martin, ‘Yaʿqūb b. al-Laiṯ und der Zunbīl’, Zeitschrift der Deutschen Morgenländischen Gesellschaft, 120.1 (1970), 69–83 Freiman, A. A., Vladimir A. Livshits, M. N. Bogoliubov, and Olga I. Smirnova, Sogdiiskie dokumenty s Gory Mug (Moscow: Izdatel’stvo Vostochnoi Literaturi, 1962) Frye, Richard Nelson, The History of Bukhara: Translated from a Persian Abridgment of the Arabic Original by Narshakhī (Cambridge, MA: Mediaeval Academy of America, 1954) Gerschevitch, Ilya, ‘Bactrian Inscriptions and Manu­scripts’, Indogermanische Forschungen, 72 (1967), 27–57 Gignoux, Philippe, Gignoux, Philippe, Les quatre inscriptions du mage Kirdīr. (Paris: Association pour l’avancement des études iraniennes, 1991) —— , ‘Les comptes de Monsieur Friyag: quelques documents économiques en pehlevi’, in Sources pour l’histoire et la geo­graphie du monde iranien (224–710), ed. by Rika Gyselen, Res Orientales, 18 (Bures-sur-Yvette: Groupe pour l’Étude de la Civilisation du Moyen-Orient, 2009), pp. 115–42 —— , ‘La collection de textes attribuables à Dādēn-Vindād dans l’archive pehlevie de Berkeley’, in Sources for the History of Sasanian and Post-Sasanian Iran, ed. by Rika Gyselen, Res Orientales, 19 (Bures-sur-Yvette: Groupe pour l’Étude de la Civilisation du Moyen-Orient, 2010), pp. 11–134 —— , ‘Une archive post-sassanide du Tabaristān (I)’, in Objets et documents inscrits en pārsīg, ed. by Rika Gyselen, Res Orientales, 21 (Bures-sur-Yvette: Groupe pour l’Étude de la Civilisation du Moyen-Orient, 2012), pp. 29–96 —— , ‘Une archive post-sassanide du Tabaristān (II)’, in Documents, argenterie et monnaies de tradition sassanide, ed. by Rika Gyselen, Res Orientales, 22 (Bures-sur-Yvette: Groupe pour l’Étude de la Civilisation du Moyen-Orient, 2012), pp. 30–71

navi gat i ng pe rsi an

—— , ‘Une archive post-sassanide du Tabarestān (III)’, in Words and Symbols: Sasanian Objects and the Tabarestān Archive, ed. by Rika Gyselen, Res Orientales, 24 (Bures-surYvette: Groupe pour l’Étude de la Civilisation du Moyen-Orient, 2016), pp. 171–84 Golden, Peter B., An Introduction to the History of the Turkic Peoples (Wiesbaden: Otto Harrassowitz, 1992) —— , ‘Turks and Iranians: An Historical Sketch’, in Turkic-Iranian Contact Areas: Historical and Linguistic Aspects, ed. by Lars Johanson and Christiane Bulut (Wiesbaden: Otto Harrassowitz, 2006), pp. 17–38 Green, Nile, ed., The Persianate World: The Frontiers of a Eurasian Lingua Franca (Berkeley: University of California Press, 2019) Grenet, Frantz, and Nicholas Sims-Williams, ‘The Historical Context of the Sogdian Ancient Letters’, in Transition Periods in Iranian History: Actes du symposium de Fribourg-en-Brisgau (22–24 Mai 1985) (Leuven: Peeters, 1987), pp. 101–22 Guzmán, Roberto Marín, ‘La expansión árabe-musulmana hacia Ma wara’ al-nahr y los primeros contactos chinos con el islam: una visión histórica del encuentro de culturas’, in Estudios sobre China desde Latinoamérica: geopolítica, religión e inmigración, ed. by Susan Chen Mok, Jose Bartels Villanueva, and Ricardo Martinez Esquival (San Jose, Costa Rica: Sección de Impresión del SIEDIN, 2013), pp. 44–76 Gyselen, Rika, The Four Generals of the Sasanian Empire: Some Sigillo­graphic Evidence (Rome: Istituto italiano per l’Africa e l’Oriente, 2001) —— , ‘“Umayyad” Zavulistan and Arachosia: Copper Coinage and the Sasanian Monetary Heritage’, in Coins, Art and Chrono­logy II: The First Millennium c.e. in the Indo-Iranian Borderlands, ed. by Michael Alram, Deborah Klimburg-Salter, Minoru Inaba, and Matthias Pfisterer (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2010), pp. 219–41 —— , ‘Le Tabaristān : lieu de mémoire sassanide’, in Objets et documents inscrits en pārsīg, ed. by Rika Gysele. Res Orientales, XXI (Bures-sur-Yvette: Peeters, 2012), pp. 102–22 —— , ‘The Parthian Language in Early Sasanian Times’, in The Parthian and Early Sasanian Empires: Adaptation and Expansion; Proceedings of a Conference Held in Vienna, 14–16 June 2012 (Oxford: Oxbow Books, 2016), pp. 61–68 —— , La géo­graphie administrative de l’Empire sassanide. Les témoignages épi­graphiques en moyen-perse (Paris: Groupe pour l’Étude de la Civilisation du Moyen-Orient, 2019) Haug, Robert, The Eastern Frontier: Limits of Empire in Late Antique and Early Medi­eval Central Asia (London: I. B. Tauris, 2019) Henning, Walter Bruno, ‘Sogdian Loan-Words in New Persian’, Bulletin of the School of Oriental and African Studies, 10.1 (1940), 93–106 —— , ‘Two Manichaean Magical Texts with an Excursus on the Parthian Ending-Ēndēh’, Bulletin of the School of Oriental and African Studies, 12.1. (1947), 39–66 —— , ‘The Bactrian Inscription’, Bulletin of the School of Oriental and African Studies, 23.1 (1960), 47–55 Heuser, Manfred, and Hans-Joachim Klimkeit, Studies in Manichaean Literature and Art, Nag Hammadi and Manichaean Studies, 46 (Leiden: Brill, 1998) Humbach, Helmut, Baktrische Sprachdenkmäler, i (Wiesbaden: Otto Harrassowitz, 1966) —— , ‘Two Inscriptions in Graeco-Bactrian Cursive Script from Afghanistan’, East and West, 18 (1967), 25–26

35 7

3 58

k ho da da d r e z akh an i

—— , ‘The Great Surkh Kotal Inscription’, in Religious Themes and Texts of Pre-Islamic Iran and Central Asia: Studies in Honour of Professor Gherardo Gnoli on the Occasion of His 65th Birthday on 6th December 2002, ed. by Carlo G. Cereti, Mauro Maggi, and Elio Provasi (Wiesbaden: Ludwig Reichert, 2003), pp. 157–66. Humbach, Helmut, and P. O. Skjærvø, The Sassanian Inscription of Paikuli, 3 pts (Wiesbaden: Ludwig Reichert, 1978) Huyse, Philip, Die dreisprachige Inschrift Šābuhrs I. an der Kaʿba-i Zardušt (ŠKZ), i (London: School of Oriental and African Studies, 1999) Kennedy, Hugh, The Prophet and the Age of the Caliphates (New York: Routledge, 2016) Keshavarz, Karim, Hezar saal nasr-e parsi (Tehran: Elmi Farhangi, 1345 sh) Khanbaghi, Aptin. ‘Champions of the Persian Language: The Mongols or the Turks?’, in The Mongols’ Middle East, ed. by Bruno de Nicola & Charles Melville (Leiden: Brill). pp. 193–215 Kreyenbroek, Philip G., ‘The Zoroastrian Tradition from an Oralist’s Point of View’, in KR Cama Oriental Institute, Second International Congress Proceedings (5th to 8th January 1995) (Bombay: KR Cama Oriental Institute, 1996), pp. 221–35 Krippes, Karl. ‘Sociolinguistic Notes on the Turcification of the Sogdians’, Central Asiatic Journal, 35.1/2, 1991, pp. 67–80. Kuwayama, Shoshin, ‘Chinese Records on Bamiyan: Translation and Commentary’, East and West, 55.1/4 (2005), 139–61 La Vaissière, Étienne de, Sogdian Traders: A History (Leiden: Brill, 2005) —— , Samarcande et Samarra: élites d’Asie Centrale dans l’empire abbasside (Leuven: Peeters, 2007) Lazard, Gilbert, ‘Pahlavi, pârsi, dari: les langues de l’Iran d’après Ibn al-Muqaffaʻ’, in Iran and Islam, in Memory of the Late V. Minorsky, ed. by Clifford Edmund Bosworth (Edinburgh: Edinburgh University Press, 1971), pp. 361–91 —— , ‘The Rise of the New Persian Language’, in The Cambridge History of Iran, iv, ed. by Richard Frye (Cambridge: Cambridge University Press, 1975), pp. 595–632 —— , ‘Pârsi et dari: nouvelles remarques’, Bulletin of the Asia Institute, n.s., 4 (1990), 239–42 Livshits, Vladimir A., ‘New Parthian Documents from South Turkmenistan’, Acta Antiqua Academiae Scientiarum Hungaricae, 25.1–4 (1977), 157–85 —— , Sogdian Epi­graphy of Central Asia and Semirech’e, ed. by Nicholas Sims-Williams, trans. by Tom Stableford, Corpus Inscriptionum Iranicarum, pt 2: Inscriptions of the Seleucid and Parthian Periods and of Eastern Iran and Central Asia, iii: Sogdian, 4 (London: School of Oriental and African Studies, 2015) Livshits, V. A., and V. N. Pilipko., ‘Parthian Ostraca from the Central Building Complex of Old Nisa’, Ancient Civilizations from Scythia to Siberia, 10.1–2 (2004), 139–81 Lurje, Pavel, ‘О следах манихеизма в Средней Азии [Traces of Manichaeism in Middle Asia]’, in Согдийцы, их предшественники, современникии наследники [Sogdians, their precursors, contemporaries and heirs] (St Petersburg: State Hermitage Publishers, 2013), pp. 219–51 Lurje, Pavel, and Ilya Yakubovich, ‘The Myth of Sogdian Lambdacism’, in Zur lichten Heimat: Studien zu Manichäismus, Iranistik und Zentralasienkunde im Gedenken an Werner Sundermann, Iranica, 25 (Wiesbaden: Otto Harrassowitz, 2017), pp. 319–41 Macuch, Maria. ‘Pahlavi Literature.’, in The Literature of Pre-Islamic Iran: Companion Volume I to A History of Persian Literature, ed. by Ronald Emmerick and Maria Macuch,

navi gat i ng pe rsi an

A History of Persian Literature, vol. 17 (London; New York: I. B. Tauris, 2009), pp. 116–96. —— , ‘The Legal Context of the Tabarestān Court Records (Tab. 1–8,  10)’, in Words and Symbols: Sasanian Objects and the Tabarestān Archive, ed. by Rika Gyselen, Res Orientales, 24 (Bures-sur-Yvette: Groupe pour l’Étude de la Civilisation du MoyenOrient, 2016), pp. 145–70 —— , ‘Pahlavi Legal Documents from Tabarestān on Lease, Loan and Compensation: The Juristic Context (Tab. 13, 14, 15, 17, 18, 23)’, in Sasanian Coins, Middle-Persian Etymo­logy and the Tabarestān Archive, ed. by Rika Gyselen, Res Orientales, 26 (Bures-sur-Yvette: Groupe pour l’Étude de la Civilisation du Moyen-Orient, 2017), pp. 165–95 —— , ‘Pahlavi Legal Documents from Tabarestān: Two Claims and a Re-evaluation of Crop Yields; The Juristic Context of Tab. 21, 22 and 24’, in Sasanian Persia and the Tabarestān Archive, ed. by Rika Gyselen, Res Orientales, 27 (Leuven: Peeters, 2019), pp. 117–50 Madelung, Wilferd, ‘The Minor Dynasties of Northern Iran’, in The Cambridge History of Iran, iv, ed. by Richard Frye (Cambridge: Cambridge University Press, 1975), pp. 198–249 Mahmoodi-Bakhtiari, Behrooz, ‘Planning the Persian Language in the Samanid Period’, Iran and the Caucasus, 7.1/2 (2003), 251–60 Malek, Hodge M., The Dābūyid Ispahbads and Early ʿAbbāsid Governors of Tabaristān: History and Numismatics, RNS Special Publications, 39 (London: Royal Numismatic Society, 2004) Marquart, Josef. Ērāns̆ahr Nach Der Geo­graphie Des Ps. Moses Xorenacci: Mit HistorischKritischen Kommentar Und Historischen Und Topo­graphischen Excursen. AKGWGötingen, Phil.-Hist. Klasse, III. (Berlin: Weidmannsche Buchhandlung, 1901) Meisami, Julie Scott, ‘Why Write History in Persian? Historical Writing in the Samanid Period’, in Studies in Honour of Clifford Edmund Bosworth, ed. by Carole Hillenbrand (Leiden: Brill, 2000), ii, pp. 348–74 Meyer, Robin, ‘Armeno-Iranian Structural Interaction: The Case of Parthian Wxd, Armenian Ink‘n’, Iran and the Caucasus, 17.4 (2013), 401–25 Meyer, Robin, ‘Alignment Change and Changing Alignments: Armenian Syntax and the First “Death” of Parthian’, Contact, in Variation, and Reconstruction in the Ancient IndoEuropean Languages: Between Linguistics and Philo­logy, ed. by M. Bianconi and others, (Leiden: Brill, forthcoming) Morony, Michael G. Iraq After the Muslim Conquest (Princeton: Princeton University Press, 1984) Nikitin, A. B., ‘The Sasanian Šahrab of Balkh’, Ancient Civilizations from Scythia to Siberia, 1.3 (1995), 365–68 Nyberg, Henrik S., ‘The Pahlavi Documents from Avroman’, Le monde oriental, 17 (1923), 182–230 Paul, Ludwig, ‘Persian Language i. Early New Persian’, in Encyclopaedia Iranica (online edn, 2000) —— , ed., Persian Origins: Early Judaeo-Persian and the Emergence of New Persian; Collected Papers of the Symposium, Göttingen 1999, vi (Wiesbaden: Otto Harrassowitz, 2003)

359

36 0

k ho da da d r e z akh an i

Perry, John R., ‘New Persian: Expansion, Standardization, and Inclusivity’, in Literacy in the Persianate World, ed. by Brian Spooner and William L. Hanaway (Philadelphia: University of Pennsylvania Press, 2012), pp. 70–94 Piras, Andrea, ‘Fromo Kesaro: echi del prestigio di Bisanzio in Asia Centrale’, in Polidoro: studi offerti ad Antonio Carile, ed. by G. Vespignani (Spoleto: Fondazione Centro Italiano di Studi sull’Alto Medioevo, 2013), pp. 671–90 Pollock, Sheldon, ‘The Cosmopolitan Vernacular’, Journal of Asian Studies, 57.1 (1998), 6–37 Potts, Daniel T., ‘Sasanian Iran and its Northeastern Frontier’, in Empires and Exchange in Eurasian Late Antiquity: Rome, China, Iran, and the Steppe ca. 250–750, ed. by Nicola Di Cosmo and Michael Maas (Cambridge: Cambridge University Press, 2018), pp. 287–301 Rezakhani, Khodadad, ‘From Aramaic to Pahlavi: Epi­graphic Observations Based on the Persis Coin Series’, in The Parthian and Early Sasanian Empires: Adaptation and Expansion, ed. by Vesta Sarkhosh Curtis, Elizabeth Pendelton, Michael Alram, and Touraj Daryaee, BIPS Archaeo­logical Mono­graph Series, 5 (Oxford: Oxbow Books, 2016), pp. 69–75 —— , ReOrienting the Sasanians: East Iran in Late Antiquity (Edinburgh: Edinburgh University Press, 2017) —— , ‘Shahnameh as a Historical Source, 1: The Türk and the Hephthalites’, in Holyfest. Org: Festschrift in Honor of Olga (Holly) Davidson on the Occasion of Her Sixty-Sixth Birthday, ed. by Niloofar Fotoohi (Boston: ILEX [online], 2018) —— , ‘Pangul and Bunji, Zhulad and Fulad: A Note on the Genealogy of the Shanasbid Amirs of Ghur’, in Dinars and Dirhams: Festschrift in Honor of Michael L. Bates, ed. by Touraj Daryaee, Judith A. Lerner, and Virginie Rey (Irvine, CA: Jordan Center for Persian Studies and Culture, 2020), pp. 219–30 Rezania, Kianoosh. ‘Mazdakism and the Canonisation of Pahlavi Translations of the Avestan Texts’, The Transmission of the Avesta, ed. by Alberto Cantera, (Wiesbaden: Harrassowitz 2012) pp. 479–94 Rezania, Kianoosh. ‘The Dēnkard Against Its Islamic Discourse’, Der Islam, vol. 94, no. 2, 2017, pp. 336–62 Rubin, Uri, ‘“Al-Walad li–l-Firāsh”: On the Islamic Campaign against “Zinā”’, Studia Islamica, 78 (1998), 5–26 Sadeghi, Ali-Ashraf, Takvin-e zaban-e farsi (Tehran: Daneshgah-e Azad-e Iran, 1357 sh) —— , ‘Yek maʿni nashenakhte kalame pahlavi dar zaban-e farsi’, Vizhe name-ye name-ye farhangestan (Farhangnevisi), 3 (1389 sh), 2–32 —— , ‘L’influence de l’arabe sur le système phono­logique du persan’, La linguistique, 11.2 (1975), 145–52 Schwartz, Martin, ‘Irano-Tocharica’, in Mémorial Jean de Menasce, ed. by Philippe Gignoux and Ahmad Tafazzoli (Louvain: Impression orientaliste 1974), pp. 399–411 Shaked, Shaul, ‘Two Parthian Ostraca from Nippur’, Bulletin of the School of Oriental and African Studies, 57.1 (1994), 208–12 Sheffield, Daniel J., ‘Primary Sources: New Persian’, in The Wiley Blackwell Companion to Zoroastrianism, ed. by Michael Stausberg and Yuhan Sohran-Dinshaw Vevaina (New York: Wiley, 2015), pp. 529–42.

navi gat i ng pe rsi an

Sims-Williams, Nicholas. Bactrian Documents from Northern Afghanistan, Vol I: Legal and Economic Documents (London: The Nour Foundation in association with Azimuth Editions and Oxford University Press, 2001) —— , ‘Ancient Afghanistan and its Invaders: Linguistic Evidence from the Bactrian Documents and Inscriptions’, in Indo-Iranian Languages and Peoples, ed. by Nicholas Sims-Williams (Oxford: Oxford University Press, 2002), pp. 225–42 —— , Bactrian Documents from Northern Afghanistan, Vol. 2: Letters and Buddhist Texts (London: The Nour Foundation in association with Azimuth Editions, 2007) —— , ‘The Bactrian Inscription of Jaghori: A Preliminary Reading’, Bulletin of the Asia Institute, 30, (2020/21), 67-74 Sims-Williams, Nicholas, and James Hamilton, Documents turco-sogdiens du ixe–xe siècle de Touen-houang, Corpus Inscriptionum Iranicarum, pt. 2: Inscriptions of the Seleucid and Parthian Periods and of Eastern Iran and Central Asia, iii: Sogdian, 3 (London: School of Oriental and African Studies, 1990) Sinor, Denis, ‘The Establishment and Dissolution of the Türk Empire’, in The Cambridge History of Early Inner Asia, ed. by Denis Sinor (Cambridge: Cambridge University Press, 1990), i, pp. 285–316 Skjærvø, Prods Oktor, ‘Iran vi. Iranian Languages and Scripts’, in Encyclopaedia Iranica, xiii, fasc. 4, (2006), pp. 344–77 —— , ‘Middle West Iranian’, in The Iranian Languages, ed. by Gernot L. Windfuhr (Abingdon: Routledge, 2009), pp. 196–278 Stausberg, Michael. ‘The Invention of a Canon: The Case of Zoroastrianism’, in Canonization and Decanonization, ed. by Arie van der Kooij and Karel van der Toorn. (Leiden: Brill, 1998), pp. 257–77 Sumner, William M., and Donald Whitcomb. ‘Islamic Settlement and Chrono­logy in Fars: An Archaeo­logical Perspective’, Iranica Antiqua, 34 (1999), 309–24 Sundermann, Werner, ‘Mani’, in Encyclopaedia Iranica (online edn, 2009) Tafazzoli, Ahmad, ‘Fahlaviyat’, in Encyclopaedia Iranica (1999), ix, pp. 158a–62 Tedesco, Paul, ‘Dialekto­logie der westiranischen Turfantexte’, Le monde oriental, 15 (1921), 184–258 Tetley, Gillies E., The Ghaznavid and Seljuk Turks: Poetry as a Source for Iranian History (London: Routledge, 2008). Thackston, Wheeler McIntosh, A Millennium of Classical Persian Poetry: A Guide to the Reading and Understanding of Persian Poetry from the Tenth to the Twentieth Century (Boston: Ibex, 1994) Tremblay, Xavier, ‘Irano-Tocharica et Tocharo-Iranica’, Bulletin of the School of Oriental and African Studies, 68.3 (2005), 421–49 Utas, Bo, ‘The Jewish-Persian Fragment from Dandan-Uiliq’, Orientalia Suecana, 17 (1969), 123–36 —— , ‘A Multiethnic Origin of New Persian?’, in Turkic-Iranian Contact Areas: Historical and Linguistic Aspects, ed. by Lars Johanson and Christiane Bulut. Wiesbaden: Otto Harrassowitz, 2006), pp. 241–51 —— , From Old to New Persian: Collected Essays (Wiesbaden: Ludwig Reichert, 2013) Vevaina, Yuhan Sohrab-Dinshaw, ‘“The Ground Well Trodden but the Shah Not Found …”: Orality and Textuality in the “Book of Kings” and the Zoroastrian

361

362

k ho da da d r e z akh an i

Mythoepic Tradition’, in Orality and Textuality in the Iranian World, ed. by Julia Rubanovich (Leiden: Brill, 2015), pp. 169–90 Vondrovec, Klaus, ‘Coinage of the Nezak’, in Coins, Art and Chrono­logy II: The First Millennium c.e. in the Indo-Iranian Borderlands, ed. by Michael Alram, Deborah E. Klimburg-Salter, Minoru Inaba, and Matthias Pfisterer, (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2010), pp. 169–90 —— , Coinage of the Iranian Huns and their Successors from Bactria to Gandhara (4th to 8th Century ce) (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2014) Weber, Dieter, Berliner Papyri, Pergamente und Leinenfragmente in Mittelpersischer Sprache, Corpus Inscriptionum Iranicarum, pt. 3: Pahlavi Inscriptions, iv–v: Ostraca and Papyri, Texts 2 (London: School of Oriental and African Studies, 2003) —— , ‘Notes on Some Parthian Names from Nisa’, Parthica, 5 (2003), 127–29 —— , ‘Pahlavi Morpho­logy’, in Morpho­logies of Asia and Africa, ed. by Alan S. Kaye (Winona Lake, IN: Eisenbrauns, 2007), i, pp. 941–74 —— , ‘New Arguments for Dating the Documents from the “Pahlavi Archive”’, Bulletin of the Asia Institute, 22 (2008), 215–22 —— , ‘Villages and Estates in the Documents from the Pahlavi Archive: The Geo­graphical Background’, Bulletin of the Asia Institute, 24 (2010), 37–65 —— , ‘An Interesting Pahlavi Letter from Early Islamic Times (Berk. 245)’, Zeitschrift der Deutschen Morgenländischen Gesellschaft, 161.1 (2011), 91–98 —— , ‘Testing Food and Garment for the “Ōstāndār”: Two Unpublished Documents from the “Pahlavi Archive” in Berkeley, CA’, Bulletin of the Asia Institute, 25 (2011), 31–37 —— , ‘Studies in Some Documents from the “Pahlavi Archive”’, Bulletin of the Asia Institute, 26 (2012), 61–95 —— , ‘Arabic Activities Reflected in the Documents of the Pahlavi Archive’, in Documents, argenterie et monnaies de tradition sassanide, ed. by Rika Gyselen, Res Orientales, 22 (Bures-sur-Yvette: Groupe pour l’Étude de la Civilisation du Moyen-Orient, 2014), pp. 179–89 —— , ‘Court Records of Lawsuits in Tabarestān in the Year 86/7 pye (737 ce): A Philo­ logical Examination’, in Words and Symbols: Sasanian Objects and the Tabarestān Archive, ed. by Rika Gyselen, Res Orientales, 24 (Bures-sur-Yvette: Groupe pour l’Étude de la Civilisation du Moyen-Orient, 2016), pp. 121–44 —— , ‘The Legal Context of the Tabarestān Court Records (Tab. 1–8, 10)’ in Words and Symbols: Sasanian Objects and the Tabarestān Archive, ed. by Rika Gyselen, Res Orientales, 24 (Bures-sur-Yvette: Groupe pour l’Étude de la Civilisation du MoyenOrient, 2016), pp. 145–70 —— , ‘Two Documents from Tabarestān Reconsidered (Tab. 12 and 26)’, in Words and Symbols: Sasanian Objects and the Tabarestān Archive, ed. by Rika Gyselen, Res Orientales, 24 (Bures-sur-Yvette: Groupe pour l’Étude de la Civilisation du MoyenOrient, 2016), pp. 185–92 —— , ‘Pahlavi Legal Documents from Tabarestān on Lease, Loan and Compensation: A Philo­logical Study (Tab. 13, 14, 15, 17, 18 and 23)’, in Sasanian Coins, Middle-Persian Etymo­logy and the Tabarestān Archive, ed. by Rika Gyselen, Res Orientales, 26 (Buressur-Yvette: Groupe pour l’Étude de la Civilisation du Moyen-Orient, 2017), pp. 131–64 —— , ‘Pahlavi Legal Documents from Tabarestān: Two Claims and a Re-evaluation of Crop Yields; A Philo­logical Study of Tab. 21, 22 and 24’, in Sasanian Persia and the

navi gat i ng pe rsi an

Tabarestān Archive, ed. by Rika Gyselen, Res Orientales, 27 (Leuven: Peeters, 2019), pp. 91–116 Windfuhr, Gernot, Persian Grammar: History and State of its Study (The Hague: Mouton, 1979) —— , ‘Eštehārdī’, in Encyclopaedia Iranica (online edn, 1998) —— , ‘Dialecto­logy and Topics’, in The Iranian Languages., ed. by Gernot L. Windfuhr (New York: Routledge, 2013), pp. 5–42 Van Bladel, Kevin. ‘The Language of the Xūz and the Fate of Elamite’, Journal of the Royal Asiatic Society, 31.3, (2021), 447–62 —— , ‘Zoroaster’s Many Languages’, Arabic Humanities, Islamic Thought: Essays in Honor of Everett K. Rowson, ed. by Joseph E. Lowry and Shawkat M. Toorawa, (Leiden: Brill, 2017) pp. 190–210 Yakubovich, Ilya, ‘Mugh 1.1. Revisited’, Studia Iranica, 31.2 (2002), 231–53 —— , ‘Nugae Sogdicae’, Bulletin of the School of Oriental and African Studies, 65.3 (2002), 544–49 —— , ‘Nugae Sogdicae II’, in Turfan Revisited: The First Century of Research into the Arts and Cultures of the Silk Road, ed. by Desmond Durkin-Meisterernst, Simone-Christiane Raschmann, Jens Wilken, Marianne Yildiz, and Peter Zieme. Berlin: Dietrich Reimer, 2004), pp. 393–400 Yarshater, Ehsan, A Grammar of Southern Tati Dialects, Median Dialect Studies, 1 (The Hague: Mouton, 1969) Yoshida, Yutaka, ‘In Search of Traces of Sogdians: “Phoenicians of the Silk Road”’, Berichte und Abhandlungen, Berlin-Brandenburgische Akademie der Wissenschaften, 9 (2002), 185–200 —— , ‘When Did Sogdians Begin to Write Vertically?’, Tokyo University Linguistic Papers, 33 (2013), 375–94 Yücesoy, Hayrettin, ‘Language of Empire: Politics of Arabic and Persian in the Abbasid World’, Publications of the Modern Language Association of America, 130.2 (2015), 384–92 Zadeh, Travis, The Vernacular Qur’an: Translation and the Rise of Persian Exegesis (Oxford: Oxford University Press, 2012) Zadeh, Travis, ‘Translation, Geo­graphy, and the Divine Word: Mediating Frontiers in Pre-modern Islam’ (unpublished doctoral dissertation, Harvard University, 2007) Zia-Ebrahimi, Reza, The Emergence of Iranian Nationalism: Race and the Politics of Dislocation (New York: Columbia University Press, 2016)

363

Aaron Michael Butts

Once Again, the Twin Histories of Arabic and Aramaic (with a Focus on Syriac) Introduction In 2003, D. Wasserstein published an article entitled ‘Why did Arabic Succeed Where Greek Failed? Language Change in the Near East after Muhammad’, in which he investigated why Arabic penetrated Near Eastern society in such a short period of time much more deeply than Greek had been able to do over the course of more than a millennium.1 A year later, R. Hoyland published a rejoinder on ‘the twin histories of Arabic and Aramaic’ in which he reframed Wasserstein’s question by adding Aramaic into the discussion: Hoyland rightly pointed out that Aramaic survived long after the rise of Islam alongside Arabic, and that in fact Arabic existed alongside Aramaic before the rise of Islam.2 Thus, for Hoyland, Aramaic is an essential part of the story about language use after Islam. In this paper, I continue Hoyland’s line of enquiry by exploring Arabic before Islam and Aramaic after Islam. However, I narrow the scope of the enquiry by focusing on language use among Syriac Christians.3 The scope of Hoyland’s paper is enormous, covering the three-millennium history of Aramaic as well as the equally long history of Arabic, even if the first part of Arabic’s history is mostly, though not entirely,





1 Wasserstein, ‘Why Did Arabic Succeed Where Greek Failed?’. 2 Hoyland, ‘Language and Identity’. Mention should also be made of the important contribution of Papaconstantinou (‘Why Did Coptic Fail Where Aramaic Succeeded?’), who, though dealing primarily with the equally interesting situation of Coptic, offers a number of pointed insights on Aramaic as well as on language use after the conquests more broadly. 3 I use ‘Syriac’ here, following our practice in the Gorgias Encyclopedic Dictionary of the Syriac Heritage (hereafter GEDSH), to refer to the cultural tradition that developed historically in the Middle East among Christians who spoke primarily (for this adverbial qualifier, see Butts, Syriac in its GrecoRoman Context) the Syriac language but that grew to encompass a number of communities from different backgrounds, cultures, and languages (GEDSH, p. ix), even if this usage implies a coherence and consistency that never actually obtained, as argued in Gross, ‘A Long Overdue Farewell’, p. 222 n. 2 (though the entire essay speaks to the issue). For Syriac Christians, see, in addition to GEDSH, Briquel-Chatonnet and Debié, Le monde syriaque and King, The Syriac World.

Aaron Michael Butts (PhD University of Chicago, 2013) is Professor at Universität Hamburg. Navigating Language in the Early Islamic World: Multilingualism and Language Change in the First Cen­turies of Islam, ed. by Antoine Borrut, Manuela Ceballos, and Alison M. Vacca, ISMAR 2 (Turnhout: Brepols, 2024) pp. 365–401 10.1484/M.ISMAR-EB.5.134632

FHG

36 6

a a ro n m icha e l b u t t s

lost to us.4 Hoyland’s interjection of Aramaic, especially approached through the longue durée, was an important corrective to Wasserstein’s discussion of language use after the rise of Islam. At the same time, however, by narrowing the scope to Syriac Christians, I hope to provide a more textured account of the twin histories of Arabic and Aramaic. The first two sections of the paper approach the topic at hand from a macro perspective, though not as broad as Hoyland’s: in section 1, I begin by discussing the use of Arabic among Syriac Christians before Islam; in section 2, I then turn to the use of Syriac among Syriac Christians after Islam. In sections 3 and 4, I approach these same topics from a micro perspective by focusing on one particular location: the Naṣrid capital al-Ḥīra (Syriac Ḥirtā), located on the Euphrates not far from the modern city of Najaf in south central Iraq. This provides an opportunity to consider, on the one hand, the use of Arabic before Islam in a particular location among a specific population (especially among the ʿIbād) as well as, on the other hand, the use of Syriac in Islamic times. The latter discussion revolves around the famous translator Ḥunayn b. Isḥāq (808–873).

Arabic Before Islam It is still too rarely acknowledged that Arabic existed prior to the rise of Islam. This is at least partly a definitional problem. Traditionally, Arabic has been narrowly defined to correspond more or less to the Classical Arabic language that is first attested in the Qurʾān.5 If other, earlier forms of the language were considered, they were generally called ‘Old Arabic’ or the like, but even then, Old Arabic was basically defined in terms of the isoglosses of the later Classical Arabic, especially the definite article al-.6 More recently, A. Al-Jallad has argued for a more expansive definition of Arabic that includes both the traditional category of (Old) Arabic and a number of other varieties, such as Safaitic, Hismaic, and Nabataean Arabic.7





4 I discuss this issue in the first section below. Since the publication of Hoyland’s paper, there is now a history of Aramaic in Gzella, Cultural History of Aramaic, which is not without its problems, including in the socio-linguistic issues that concern us here. Unfortunately, no such work exists for Arabic, though one is being prepared by A. Al-Jallad under the tentative title The Word, the Blade, and the Pen: Three Thousand Years of Arabic (under contract with Princeton University Press). 5 In fact, the Qurʾān as transmitted to us attests to at least two forms of Arabic, one in the consonantal text and another in the vocalization tradition(s). The classic statement is Nöldeke and others, Geschichte des Qorāns, iii, pp. 1–115 (an English translation by Behn is now available as The History of the Qurʾān). An important series of articles was later published by Diem: ‘Untersuchungen zur frühen Geschichte der arabischen Ortho­graphie’, I–IV. More recently, Van Putten has published several insight­ful articles on the consonantal text: ‘The Development of the Triphthongs in Quranic and Classical Arabic’; ‘The Feminine Ending -at as a Diptote in the Qurʾānic Consonantal Text’; ‘Case in the Qurʾānic Consonantal Text’; ‘Hamzah in the Quranic Consonantal Text’; ‘Inferring the Phonetics of Quranic Arabic from the Quranic Consonantal Text’. See also his mono­graphic treatment Quranic Arabic. 6 See, for example, Macdonald, ‘Old Arabic (Epi­graphic)’ and ‘Reflections on the Linguistic Map of Pre-Islamic Arabia’. 7 See especially Al-Jallad, ‘The Earliest Stages of Arabic and its Linguistic Classification’ and ‘What is Ancient North Arabian?’.

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

With this broader definition, Arabic is abundantly attested in the pre-Islamic period, especially in the Safaitic and Hismaic inscriptions, which, when combined, number more than fifty thousand.8 Nevertheless, even if one wishes to maintain a narrower definition, the traditional (Old) Arabic still existed prior to Islam; it is just that we have little direct evidence for it, since it was only exceptionally written in the pre-Islamic period.9 There is, however, some attestation. Already in cuneiform inscriptions from the first millennium bce, we find a number of mentions of Arabs as well as the occasional Arabic word, especially personal names.10 The first clear attestation of an Arabic word comes from an inscription of the neo-Assyrian ruler Shalmaneser III (853 bce), where one ‘Gindibu the Arab’ (mGi-in-di-bu-u kurAr-ba-a-a) is mentioned along with a number of other rulers including Adad-Idri and Ahab the Israelite (I Kings 16–20) — the first mention of an Israelite king outside of the Bible.11 The name Gindibu can be compared with Arabic names such as Jundub, Jundab, and Jindab from the classical period.12 Outside of personal names, Arabic words are extremely rare, if not non-existent, in cuneiform sources.13 But this should not distract from the point that there was an Arabic language (better: Arabic languages) already in the early first millennium bce. It is not only in cuneiform sources that we find traces of the Arabic language. To give another example: writing in the fifth century bce, Herodotus mentions that the ‘Arabs’ worship a goddess of the sky named Ἀλιλάτ (Histories, i. 131; iii. 8), which is likely the Arabic word al-ʾilat ‘the goddess’.14 If we move slightly later in time, Nabataean Aramaic texts contain numerous Arabic words, likely reflecting the fact that the authors of these inscriptions spoke Arabic but wrote Aramaic since it was the lingua franca of the area.15 These are only three of the many examples of the Arabic language in indirect transmission (Nebenüberlieferung) prior to Islam. It is not until the Common Era — or perhaps a little earlier — that we first have continuous texts written in a form of Arabic that is not too different from that

8 For Safaitic, see Al-Jallad, An Outline of the Grammar of Safaitic Inscriptions. Al-Jallad is currently preparing a grammar of the Hismaic inscriptions. 9 For a reliable, even if now slightly outdated overview, see MacDonald, ‘Old Arabic (Epi­graphic)’. 10 In general, see Retsö, The Arabs in Antiquity, pp. 119–211, though note the comments in Robin, ‘Les arabes vus de Ḥimyar’. See also the earlier study of Ephʿal, The Ancient Arabs. 11 The inscription is edited in Grayson, Assyrian Rulers of the Early First Millennium bc II (858–745 bc), p. 23. For discussion, see Ephʿal, The Ancient Arabs, pp. 21, 75 and especially Retsö, The Arabs in Antiquity, pp. 124–28. 12 See Lane, An Arabic-English Lexicon, p. 388. 13 For one of the better potential examples, see Livingstone, ‘An Early Attestation of the Arabic Definite Article’, but with the comments in Sima, Tiere, Pflanzen, Steine und Metalle in den altsüdarabischen Inschriften, pp. 126–27; Militarev and Kogan, Semitic Etymo­logical Dictionary, ii, pp. 212–13 (no. 161); Hämeen-Anttila, ‘The Camels of Tiglath-Pileser III and the Arabic Definite Article’. 14 This interpretation is not, however, universal; see Hämeen-Anttila and Rollinger, ‘Herodot und die arabische Göttin “Alilat”’. 15 See Butts, ‘North Arabian Features in the Nabataean Aramaic Inscriptions from Madāʾin Ṣāliḥ’, with the additional literature cited there.

367

36 8

a a ro n m icha e l b u t t s

Figure 9.1. Namāra inscription © Musée du Louvre, dist. RMN-GP/ Pierre and Maurice Chuzeville. Reproduced with permission.

which would eventually develop into Classical Arabic.16 I mention only one here: the famous Namāra inscription (see Fig. 9.1).17 This inscription was discovered in 1901 in the desert of southern Syria near Namāra and is now housed in the Louvre (Antiquités Orientales 4083). The inscription contains five lines of text describing the deeds, as an abbreviated res gestae, of one Maraʾ (traditionally: Imruʾ) al-Qays, son of ʿAmr, self-styled king of all the Arabs. The text is written in the Nabataean Aramaic script, but the language is not too distant from what we now know as Classical Arabic. Consider, for instance, the following sentence that occurs almost at the end of the inscription: flm yblǵ mlk mblǵh ‘No king has reached his rank’ (line 4). This might as well be Classical Arabic!18 At the same time, however, some features in this inscription, especially the pronoun ty ‘this’ and the preposition ʿkdy ‘after’, depart from Classical Arabic.19 The Namāra inscription is dated to the year 223, presumably according to the era of the Roman Province of Arabia, which corresponds to 328 ce. Thus, we have here with the Namāra inscription a case of Arabic language written in Nabataean script before the rise of Islam. If we move into the Syriac milieu, which is the subject of this article, evidence for the Arabic language becomes sparser. Most of the evidence is related to the broader topic of the Arabs in Syriac literature.20 We need, of course, to be extremely 16 For inventories, none of which is the last word on the matter, see Macdonald, ‘Reflections on the Linguistic Map of Pre-Islamic Arabia’, pp. 48–57, 61; Macdonald, ‘Old Arabic (Epi­graphic)’, pp. 467–72 (slightly reduced from the previous inventory); Al-Jallad, ‘On the Genetic Background of the Rbbl bn Hfʿm Grave Inscription at Qaryat al-Fāw’, pp. 445–46. For broader studies, see, among many others, Robin, ‘Les langues de la péninsule Arabique’; Hoyland, ‘The Birth of Arabic in Stone’; Nehmé, ‘Aramaic or Arabic?’. 17 The definitive edition is Calvet and Robin, Arabie heureuse, Arabie déserte, pp. 265–69, where many references to the secondary literature can be found; of these, see especially Beeston, ‘Nemara and Faw’, pp. 2–6 and Bellamy, ‘A New Reading of the Namārah Inscription’. 18 Note especially the negation of the past tense with lam and the short prefix-conjugation: this is one of the isoglosses that Al-Jallad (‘What is Ancient North Arabian?’, p. 17) uses to argue for the unity of the traditional (Old) Arabic and Safaitic (and Hismaic). 19 On the latter, see Al-Jallad, ‘The Particle ʿkdy in the Namārah Inscription’. 20 The classic study is Segal, ‘Arabs in Syriac Literature before the Rise of Islam’.

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

careful here: the category of Arabs, whether in antiquity or after, is a slippery one.21 In addition, Arabs and the Arabic language have never been coterminous; rather, Arabic has been used by non-Arabs, and there have been Arabs who did not speak Arabic. With these caveats in mind, it is worth looking briefly at Arabs and Syriac Christians. Arabs were present from the very beginning of Syriac Christianity. Edessa, the cradle of Syriac Christianity, was ruled by the so-called Abgarid dynasty.22 According to the much later Chronicle of Zuqnin (late eighth century), the Abgarid dynasty began in the second century bce around the time of the disintegration of the Seleucid Empire.23 It lasted up until the middle of the third century, when Edessa became a Roman colonia.24 A majority of the kings over this period bear patently Arabic names, including ʿAbdu, Wāʾil, and Sahru, as well as the often reccurring Abgar and Maʿnu.25 It is within the time frame of the Abgarid dynasty that Christians first appear in Edessa and the surrounding area of Osrhoene. The earliest Christian in Edessa whom we know by name is Bardaiṣan (154–222), who, Julius Africanus (Kestoi i. 20) informs us, was active in the court of Abgar VIII (177–212).26 Thus, from the very beginning of Christianity in Edessa, there was an Arab presence, as shown by the names of the Abgarid kings. Arabs continue to surface in Syriac literature throughout Late Antiquity. To give just one of many potential examples, the Vita of Rabbula of Edessa (d. 435/436) relays a brief story about the protagonist’s encounter with ‘a band of Arabs’ ̈ ‫)ܓܝܣܐ‬.27 Rabbula withdraws to the desert, in the model of Antony, and (‫ܕܛܝܝܐ‬ the story continues:

21 Several relevant books have recently appeared, including Fisher, Arabs and Empires before Islam; Webb, Imagining the Arabs; and Mackintosh-Smith, Arabs, the first of which is the most reliable. 22 See Wardle, ‘Abgarids of Edessa’, as well as, with more detail, Millar, The Roman Near East, pp. 457–81 and Ross, Roman Edessa. 23 Edited in Incerti auctoris Chronicon Pseudo-Dionysianum vulgo dictum, ed. by Chabot, i, pp. 50–52; English translation in The Chronicle of Zuqnīn, trans. by Harrak, pp. 98–101. 24 From the Old Syriac documents, it can be surmised that the transition to Roman colonia was not linear: P. Dura 28 shows that the Abgarid dynasty must have come to an end in 212/213, when the city became a Roman colonia. On the basis of P. Euph. 19, which states that 28 December 240 fell in the second year of king Abgar, it can be established that the Abgarid dynasty was restored in 239 (or late 238) under Abgar X, son of Maʿnu. Finally, it seems that Edessa must have reverted to a colonia by 242 on the basis of P. Euph. 20, which gives 1 September 242 as year 30 of the colonia. For an overview of the Old Syriac documents, including publication information, see Butts, ‘Old Syriac’. John Healey and I are currently in the process of re-editing the Old Syriac documents. For the broader political situation, see Ross, Roman Edessa. 25 It is interesting to note that several of these names are written with a final waw, paralleling a phenomenon found with (Arabic) personal names in Nabatean inscriptions. See the recent discussion in Al-Jallad, ‘One wāw to Rule Them All’. 26 For Bardaiṣan, see Brock, ‘Bardaiṣan’, with additional biblio­graphy. For the witness of Julius Africanus, see the recent re-edition with English translation in Iulius Africanus: Cesti, pp. 100–03. 27 The term ṭayyāyē, which etymo­logically looks like a gentilic noun for ‘people belonging to the Arab tribe of Ṭayy’, is used in Syriac as a generic designation for ‘Arabs’.

369

3 70

a a ro n m icha e l b u t t s

ܿ ̈ ‫ܓܕܫ ܘܗܐ ܓܝܣܐ‬ ‫ܕܛܝܝܐ‬ ܼ ‫ܘܟܕ ܒܨܠܘܬܐ ܩܐܡ ܕܥܢܝܢܗ ܡܢ ܐܠܗܐ ܐܠ ܢܦܣܘܩ ܗܘܐ܆‬ ܿ ‫ ܼܗܢܘܢ ܕܝܢ ܚܙܘ ܠܡܝܬܐ‬.‫ܕܡܛܐ ܡܢ ܟܕܘ ܥܕܢܐ ܕܟܘܠܠܗ‬ .‫ ܘܚܕܝ‬.‫ܕܐܬܐ‬ ‫ܐܬܚܙܝ ܠܗ‬ ܼ ‫ܕܣܒܪ‬ ܼ .‫ ܘܒܠܚܘܕ ܠܚܡܗ ܫܩܠܘ ܘܬܟܣܝܬܗ ܘܫܢܝܘ‬.‫ܚܝܐ ܒܢܩܥܐ ܣܦܝܩܐ܆ ܘܒܣܪܘ ܥܠܘܗܝ ܘܫܒܩܘܗܝ‬ ܿ ‫ܕܐܬܐ‬ ‫ ܕܦܓܥ ܒܗܘܢ ܐܢܫ‬.‫ܬܡܗ ܕܝܢ ܒܗܕܐ‬ ܼ .‫ܼܗܘ ܕܝܢ ܐܦ ܥܠ ܗܕܐ ܠܡܪܗ ܡܘܕܐ ܗܘܐ‬ ܿ .‫ܠܣܢܝܩܘܬܗ ܘܐܠ ܐܟܝܘܗܝ‬ ‫ܗܘܐ ܕܢܝܬܐ ܠܗ ܛܝܒܘܬܐ ܠܚܡܐ‬ As he stood in prayer, so as not to break off his conversation with God, he suddenly saw a band of Arabs approaching. He was glad, for they thought that the time for his coronation had already arrived. They, however, took him to be a dying man living in a barren hole, and they despised him. They left him alone, took only his food and his garments, and departed. He also praised his Lord for the following: He was amazed by the fact that a man who was coming as a kindness to bring him bread for his need met them, and they did not harm him.28 This of course does not tell us much, but at the very least, we have a hagio­graphic author writing in probably the mid-fifth century who depicts a group of marauding Arabs attacking a Christian ascetic. It is hard to imagine that such a story would have much currency with the fifth-century audience of the text if Arabs were unknown in Edessa at this time. Many of the ‘Arabs’ mentioned in Syriac literature are not Christian, but this is not exclusively the case. Consider, for instance, a case in the History of the ‘Slave of Christ’, a Syriac hagio­graphic text which narrates a harrowing tale, set on the mountain of Sinjar, of a young Jewish child, Asher, who after converting to Christianity and taking the name ʿAḇdā da-Mšiḥā (‘Slave of Christ’) is martyred by his father, Levi, in a scene reminiscent of Abraham’s offering of Isaac in Genesis 22.29 After the boy is killed, his body is hastily covered with dirt by the Christian children who converted him. Then — and this is important — a caravan of merchants passes on the road, travelling from east back home to the west. As they travel on this road, they see rays shining from the makeshift burial site. They approach, see the body of the child, recognize him as a martyr, and take the body home with them, placing it in a shrine. Interestingly, these merchants are described as follows in the text: ‘They were Christians from the Arab peoples who are in the West’ (§ 17).30 When they eventually discover the name of the martyr, they place a plaque above his sanctuary that is said to read, ‘This is the place of the coronation of the martyr of Christ ʿAbd al-Mašīḥ’ (§ 21).31 Outside of the opening and closing headings, which can obviously be secondary, this is the only place in the Syriac text that the

28 Edited by Overbeck in S. Ephraemi Syri Rabulae episcopi Edesseni Balaei aliorumque Opera selecta, p. 169. English translation in Doran, Stewards of the Poor, p. 74. A more recent Syriac edition with English translation is available in The Rabbula Corpus, ed. by R. R. Phenix, Jr., and C. B. Horn, pp. 24–25. 29 Critically edited with an English translation in The History of the ‘Slave of Christ’, ed. by Butts and Gross. 30 The History of the ‘Slave of Christ’, ed. by Butts and Gross, pp. 136–39. 31 The History of the ‘Slave of Christ’, ed. by Butts and Gross, pp. 148–49.

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac) Figure 9.2. Ḥimà-Sud PalAr 1. Image courtesy of La mission archéo­logique franco-séoudienne de Najrān (MAFSN), directed by C. Robin, ʿA. I. al-Ghabbān, and S. F. al-Saʿīd. Reproduced with permission.

name of the martyr is given in Arabic, even if it is slightly Syriacized with šin instead of the Arabic sīn. So, here, we have Arab Christians using Arabic language. Or, at the very least, that is what the text is presenting. This case becomes more complicated when we consider the date of the text: the History of the ‘Slave of Christ’ is set in the year 390 (§ 1).32 It was, however, likely written at a later date. A terminus ante quem can be established at around 850, since one of the recensions of the Armenian translation is dated to 873 (= 322 of the Armenian era).33 How much earlier the text could have been written is unclear. In our recent edition, Gross and I propose that the text was probably written after 650, but it should be stressed that our arguments are not conclusive, even if we think it likely.34 Regardless, even if the text was written in the Islamic period, the author has no problem in writing about the imagined presence of Christian Arabs who speak Arabic as early as the fourth century. This could, of course, be gross anachronism. But given the other evidence of Arabs across Syria and Mesopotamia throughout Late Antiquity, I think that this story could reflect the presence of Christian Arabs who speak Arabic in this place and time, even if this particular instance has been imagined by a later author. Fortunately, there is some recently-discovered evidence that may more directly illustrate the use of Arabic in the milieu of Syriac Christians — and perhaps even by Syriac Christians — prior to Islam, even though it takes us beyond the homeland of

32 The History of the ‘Slave of Christ’, ed. by Butts and Gross, pp. 88–89. 33 See already Garitte, ‘La passion géorgienne de saint ʿAbd al-Masīh’, p. 188 with n. 6. 34 For the discussion, see The History of the ‘Slave of Christ’, ed. by Butts and Gross, pp. 34–36.

37 1

3 72

a a ro n m icha e l b u t t s Figure 9.3. Ḥimà-Sud PalAr 8. Image courtesy of La mission archéo­ logique franco-séoudienne de Najrān (MAFSN), directed by C. Robin, ʿA. I. al-Ghabbān, and S. F. al-Saʿīd. Reproduced with permission.

Syriac Christianity. Relatively recently, sixteen short inscriptions from the environs of Ḥimā in the southern part of the Arabian Peninsula were published.35 As the editors note, the inscriptions form an assemblage likely written around the same time by the same group of people.36 This is important, since one of the inscriptions (Ḥimà-Sud PalAr 1; see Fig. 9.2) is dated to the month of Burak in the year 364, presumably according to the era of the Roman Province of Arabia (more on this shortly), which corresponds to February–March 470 ce.37 The inscriptions are all marked with a cross and so present as Christian. Twelve of the inscriptions are written in an early form of the Arabic script, and the remaining four are in the Old South Arabian monumental script (musnad). The inscriptions consist mostly of personal names, including several ultimately from the Hebrew Bible, such as Isaac (ʾsḥq), Moses (mwsy), and Elijah (ʾlyʾ).38 The few words that are not personal names paint an intriguing picture of the language of the inscriptions. The definite article al- is employed several times, including in the word ʾl-ʾlh ‘God’ (Ḥimà-Sud PalAr 8; see Fig. 9.3). This strongly suggests that the inscriptions are in the Arabic language. Nevertheless, a few words cannot be Arabic and instead are Aramaic: yrḥ ‘month’, št ‘year’, and br ‘son’. The last word occurs exclusively in personal names and so is not diagnostic. The former two present more complications, but I propose that they are to be understood as Aramaic ideograms in an otherwise Arabic text.39 35 Robin, al-Ghabbān, and al-Saʿīd, ‘Inscriptions antiques de la région de Najrān’. My references to these inscriptions follow the sigla established in this edition. 36 Robin, al-Ghabbān, and al-Saʿīd, ‘Inscriptions antiques de la région de Najrān’, p. 1039. 37 Robin, al-Ghabbān, and al-Saʿīd, ‘Inscriptions antiques de la région de Najrān’, pp. 1091–92. 38 Though these names could also be understood as Jewish, the ubiquitous presence of the cross prompts their interpretation as Christian. 39 For a similar argument involving a slightly earlier group of inscriptions, see Macdonald, ‘Ancient Arabia and the Written Word’, p. 20, and Nehmé, ‘Aramaic or Arabic?’, pp. 92–93.

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

Thus, these inscriptions are, I suggest, written in the Arabic language but influenced by a broader Aramaic ortho­graphic/scribal tradition. If we bear in mind that these inscriptions are Christian, then Aramaic here may well mean more specifically Syriac: After all, Syriac is a dialect of Aramaic. In the context of the present article, we might speculate that these inscriptions were written in Arabic by Christians who belonged to the broader Syriac Christian tradition. Ḥimā, the location in which these inscriptions were found, is only about 100 km north-north-east of Najrān. Najrān was home to (Syriac) Christian communities up to the time of ʿUmar (r. 634–44), when the Christian population was allegedly resettled to Iraq.40 This Christian population became well known throughout the broader Byzantine commonwealth because of its persecution by Joseph, the ruler of Ḥimyar.41 Important for the discussion here is that the Christian communities in Najrān seem to have belonged to the Syriac tradition. In his Letter on the Ḥimyarite Martyrs, which is the earliest hagio­graphic account of the Najrān persecution, the Syriac Miaphysite leader Simeon of Beth Arsham presents the Christians of Najrān as co-religionists.42 Similarly, the eminent Syriac Miaphysite poet Jacob of Serugh wrote a letter to the Christians in Ḥimyar to console them in the face of (presumably an earlier) persecution.43 This and other evidence suggest that the Christians of Najrān were part of the Syriac tradition, and it raises — but ultimately does not answer — the question of whether the Syriac Christians in Najrān were in some way connected with the Christians writing the Arabic inscriptions in Ḥimā, who I suggested earlier on other grounds also belonged to the Syriac tradition. Finally, I should mention that several Christian inscriptions in Arabic similar to those from Ḥimā have been discovered further north in the Arabian Peninsula and in Jordan. Most relevant is a Christian inscription dated to 548/549 ce from Dūmat al-Jandal in the al-Jawf region (DaJ144Par1).44 The script of this inscription is closely similar to that of the Ḥimā inscriptions as is the language, which is almost certainly Arabic (see ʾl-ʾlh ‘God’) but which includes the Aramaic yrḥ ‘month’ just like the Ḥimā inscriptions do.45 Mention should also be made of a recently published 40 In general, see Van Rompay, ‘Nagran’, with additional references. I say ‘communities’ in the plural because there seems to have been both Dyophysites and Miaphysites in Najrān. See Robin, ‘Ḥimyar, Aksūm, and Arabia Deserta in Late Antiquity’, p. 148. 41 A conventional historical narrative of the events is available in Nebes, ‘The Martyrs of Najrān and End of the Ḥimyar’, with a more detailed account in Hatke, ‘Africans in Arabia Felix’. See also the interesting perspective in Bowersock, The Throne of Adulis. For more detail, especially on the various sources, see Beaucamp, Briquel-Chatonnet, and Robin, Juifs et chrétiens en Arabie aux ve et vie siècles. For Joseph in particular, see Robin, ‘Joseph, dernier roi de Ḥimayr’. 42 The letter is edited by Guidi in ‘La lettera di Simeone vescovo di Bêth-Arśâm sopra i martiri omeriti’; an English translation is available in Jeffery, ‘Christianity in South Arabia’. See also the important study by Taylor, ‘A Stylistic Comparison of the Syriac Ḥimyarite Martyr Texts Attributed to Simeon of Beth Arsham’. 43 Edited as Iacobi Sarugensis Epistulae quotquot supersunt, ed. by Olinder, pp. 87–102. For discussion, see Forness, Preaching Christo­logy in the Roman Near East, pp. 115–31. 44 Nehmé, ‘New Dated Inscriptions from a Site near al-Jawf, Ancient Dūmah, Saudi Arabia’. 45 As already noted by Nehmé in the editio princeps (‘New Dated Inscriptions from a Site near al-Jawf, ancient Dūmah, Saudi Arabia’, pp. 128–29).

37 3

3 74

a a ro n m icha e l b u t t s

Christian inscription in Arabic discovered near Qaṣr Burquʿ in northeastern Jordan that mentions ‘Yazīd the king’, whom the editors tentatively identify as Yazīd I (r. 680–83); if the identification is correct, this inscription brings us into the Islamic period.46 Given these similar inscriptions from further north, it is interesting to note that the one dated inscription among the Ḥimā cache (Ḥimà-Sud PalAr 1) seems to be dated according to the era of the Roman Province of Arabia, which might suggest that the people responsible for the inscription also came from further north.47 Regardless, setting aside the Yazīd inscription given its probable later date, the inscriptions from Ḥimā and the inscription of Dūma provide evidence for the use of Arabic in the pre-Islamic period among Christians. Given what we know about the presence of Syriac Christians in the Arabian Peninsula at this time, these Arabic-using Christians were almost certainly in contact with Christians belonging to the Syriac tradition, and they may have even belonged to this very tradition themselves. These inscriptions thus document yet another intersection between the Arabic language and Syriac Christians in the pre-Islamic period.

Syriac after Islam Having discussed in the previous section the use of Arabic among Syriac Christians before Islam, I now want to turn to the use of Syriac among Syriac Christians after Islam. But first, it should be acknowledged that Arabic played an increasingly prominent role among Syriac Christians after the rise of Islam. The earliest surviving Arabic literature written by Syriac Christians stems from the eighth century, and such literature becomes more common in the ninth century.48 One of the earliest known authors of this literature is Theodore Abū Qurra.49 Abū Qurra, who seems

46 Al-Shdaifat, Al-Jallad, al-Salameen, and Harahsheh, ‘An Early Christian Arabic Graffito Mentioning “Yazīd the King”’. 47 As suggested to me by Ahmad Al-Jallad (personal communication). 48 In general, see Griffith, ‘From Aramaic to Arabic’, esp. pp. 24–30. A few words need to be said at this point about the controversial topic of Christian literature in Arabic, especially translations of the Bible, prior to Islam. In a series of articles in the early 1930s, Baumstark argued that the Bible was translated into Arabic by Christians before Islam (see especially his ‘Das Problem eines vorislamischen christlich-kirchlichen Schrifttums in arabischer Sprache’). Although some scholars, such as Graf (Geschichte der christlichen arabischen Literatur, i, pp. 27–52), dissented early on, others, including perhaps most prominently Shahîd (Byzantium and the Arabs in the Fourth Century, pp. 435–43; Byzantium and the Arabs in the Fifth Century, pp. 422–29, 449–50; Byzantium and the Arabs in the Sixth Century, ii, part 2, p. 295), followed Baumstark in arguing for the existence of a preIslamic Arabic Bible. The entire question has recently been reviewed by Griffith (The Bible in Arabic, pp. 41–53, which builds upon his earlier ‘The Gospel in Arabic’), who judiciously concludes that ‘no conclusive documentary or clear textual evidence of a pre-Islamic, written Bible in Arabic translation has yet come to light’ (The Bible in Arabic, pp. 42–43). 49 In general, see Butts, ‘Theodoros Abū Qurrah’; Graf, Geschichte der christlichen arabischen Literatur, ii, pp. 7–26; Griffith, ‘Reflections on the Bio­graphy of Theodore Abū Qurrah’; Griffith, Theodore Abū Qurrah; Lamoreaux, ‘The Bio­graphy of Theodore Abū Qurrah Revisited’; Lamoreaux, ‘Theodore Abū Qurra’; Samir, Abū Qurrah.

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

to have been born in Edessa in 750, eventually became the Melkite bishop of Ḥarrān, dying sometime after 829. An impressive corpus of texts by Abū Qurra in both Arabic and Greek (as well as Georgian translations) survives.50 In addition, according to his own testimony, Abū Qurra also wrote in Syriac, though none of his Syriac writings survive.51 From just a slightly later period, there are surviving Arabic texts written by other Syriac Christians, such as the Syriac Orthodox author Ḥabīb b. Khidma Abū Rāʾiṭa (d. c. 851) and the Church of the East author ʿAmmār al-Baṣrī (fl. c. 850).52 This triumvirate establishes the use of Arabic among Syriac Christians by the first half of the ninth century.53 The new movement of Syriac Christians writing in Arabic notwithstanding, Syriac continued to be used. In fact, the late eighth through ninth centuries witness an impressive literary output in Syriac across a variety of genres.54 To give just a few examples: the Syriac Orthodox author Mushe bar Kipho (d. 903) produced a vast oeuvre, with his surviving writings covering biblical exegesis, theo­logy, and liturgy, not to mention lost works on history, heresio­logy, and philosophy.55 Another Syriac Orthodox author writing at this time is Antony of Tagrit.56 He wrote a number of smaller treatises addressing topics such as divine providence, the sacrament of chrism, and grace as well as encomia and liturgical prayers.57 However, he is best known for his magnum opus, entitled On the Knowledge of Rhetoric, which is the only treatise

50 See the references in n. 49 above. 51 Edited as Mayāmir Thāwudūrus Abī Qurra, pp. 60–61; English translation in Lamoreaux, Theodore Abū Qurrah, p. 119. 52 For the former, see Griffith, ‘Habīb ibn Ḫidmah Abū Rāʾiṭah’; Keating, Defending the ‘People of Truth’ in the Early Islamic Period; Keating, ‘Abū Rāʾiṭa l-Takrītī’. For the latter, see Hayek, ʿAmmār al-Baṣrī; Hayek, ‘ʿAmmār al-Baṣrī’; Beaumont, ‘ʿAmmār al-Baṣrī’. 53 It is worth noting that this occurred slightly later in Egypt: Severus b. al-Muqaffaʿ (d. after 987) was among the first Christians in Egypt to choose to write in Arabic. In general, see Davis, Coptic Christo­logy in Practice, pp. 201–36; Graf, Geschichte der christlichen arabischen Literatur, ii, pp. 300–18; Swanson, ‘Sāwīrus ibn al-Muqaffaʿ’. Another interesting comparison is the adoption of Arabic by Jews: Saʿadya Gaon (882–942) has traditionally — though not universally (see, e.g., Drory, Models and Contacts, pp. 126–232) — been celebrated as the ‘founder’ of Judaeo-Arabic literature, at least among the geonim and their followers. Recent research by Blau and Hopkins has, however, shown that Arabic translations of the Bible and of certain geonic works predate Saʿadya by a century or so (Early Judaeo-Arabic in Phonetic Spelling). For a model of the spread of Arabic among Jews, see BenShammai, ‘Observations on the Beginnings of Judeo-Arabic Civilization’. 54 There were, of course, also Syriac texts written from the time of the conquests through the early eighth century, including works by ʿEnanishoʿ (seventh century), Ishoʿyahb III (d. 659), Severos Sebokht (d. 666/667), Athanasius II of Balad (d. 687), Isaac of Nineveh (late seventh century), Simeon d-Ṭaybutheh (late seventh century), John bar Penkaye (late seventh century), Ḥenanishoʿ I (d. 699/700), Jacob of Edessa (d. 708), George, bishop of the Arabs (d. 724), and Aba II of Kashkar (d. 751), but I focus here on the late eighth though ninth centuries, since it is only from this time that Syriac Christians are also first writing in Arabic, as discussed in the previous para­graph. 55 In general, see Reller, Mose bar Kepha und seine Paulinenauslegung, with its extensive introduction. 56 See Watt, ‘Anṭun of Tagrit’, with further references, to which can be added Nicosia, ‘Reassessing Antony of Tagrit’. 57 These are preserved in two manu­scripts, London, British Library, MSS Add. 14726 and Add. 17208, but all remain unedited.

37 5

3 76

a a ro n m icha e l b u t t s

on rhetoric to survive in Syriac.58 Within the Church of the East I will point out Catholicos Timothy I (727/728–823).59 Timothy had interests in church legislation and canon law, authoring a treatise on the topic as well as possibly playing a role in the assembly of the so-called Synodicon Orientale, the most important juridical text for the Church of the East.60 In addition, Timothy wrote more than two hundred letters, almost sixty of which survive.61 These deal with a range of topics, from biblical studies to a dialogue with Caliph al-Mahdī (r. 775–785).62 I could continue with the historical writing by Theodosios of Edessa (late eighth–early ninth centuries) and Dionysios of Tel Maḥre (d. 845); the monastic writings by Thomas of Marga (ninth century) and Ishoʿdnaḥ (ninth century); and the exegetical writings of the monk Severos (ninth century), Emmanuel bar Shahhare (d. 980), Theodore bar Koni (fl. 792), Ishoʿ bar Nun (d. 828), and Ishoʿdad of Merv (fl. 850), and this does not even cover all of the major authors, much less minor ones such as Benjamin of Edessa (first half of the ninth century), Dawid bar Pawlos (end of eighth century, perhaps into the ninth), and Loʿozar bar Sobhto (early ninth century), to name only a few. My point is clear: the late eighth through ninth centuries witnessed an impressive literary output in Syriac across a variety of genres. Around the turn of the millennium, however, we start to see a transition, as Syriac Christians increasingly choose to write in Arabic rather than in Syriac. There are many ways to illustrate this shift, but I want to look at just one: the literary activity of Ibn al-Ṭayyib (d. 1043).63 Ibn al-Ṭayyib’s oeuvre includes more than forty items, all in Arabic. Among his works is The Law of Christianity (Fiqh al-Naṣrāniyya).64 This text is divided into two parts: the first is an Arabic abridgement of the various Syriac legal collections of the Church of the East, and the second part consists of an Arabic abridgement of Gabriel of Baṣra’s collection of Syriac legal texts, largely overlapping with the first part.65 Thus, The Law of Christianity is an Arabic reworking

58 On the Knowledge of Rhetoric consists of five books, of which only the fifth book has been edited critically: The Fifth Book of the Rhetoric of Antony of Tagrit, ed. by Watt. The entire text, albeit not critically edited, is available in The Book of the Rhetoric of Antony Rhitor of Tagrit, ed. by Sewan d-Bet Qermez. For the first book, see also ‘Antony of Tagrit’s Rhetoric Book One’, ed. by Eskenasy. 59 In general, see Berti, Vita e studi di Timoteo I. 60 The Synodicon Orientale has been edited (with a French translation) by Chabot. A German translation is available as Das Buch der Synhados oder Synodicon Orientale, trans. by Braun. 61 Timothy’s letters are in the process of being (re-)edited by M. Heimgartner: Timotheos I, Ostsyrischer Patriarch, Disputation mit dem Kalifen Al-Mahdi; Die Briefe 42–58 des ostsyrischen Patriarchen Timotheos I; Die Briefe 30–39 des ostsyrischen Patriarchen Timotheos I; Die Briefe 40 und 41 des ostsyrischen Patriarchen Timotheos I. 62 For Timothy as a biblical scholar, see Ter Haar Romeny, ‘Biblical Studies in the Church of the East’. For the dialogue with al-Mahdī, see the overview and biblio­graphy in Heimgartner, ‘Letter 59 …’ as well as his Timotheos I, Ostsyrischer Patriarch, Disputation mit dem Kalifen Al-Mahdi. 63 For this author and his writings, see Butts, ‘Ibn al-Ṭayyib’; Faultless, ‘Ibn al-Ṭayyib’; Graf, Geschichte der christlichen arabischen Literatur, i, pp. 152–55; ii, pp. 160–77. 64 Edited with a German translation by Hoenerbach and Spies as Ibn at-Taiyib: Fiqh an-Nasrânîya, ‘Das Recht der Christenheit’. 65 See Kaufhold, ‘Sources of Canon Law in the Eastern Churches’, pp. 310–11 and, with more detail, Kaufhold, Die Rechtssammlung des Gabriel von Basra.

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

of the existing Syriac juridical literature of the Church of the East. Ibn al-Ṭayyib’s monumental The Paradise of Christianity (Firdaws al-Naṣrāniyya) is similar in nature. This is a commentary on the entire Bible in two parts. One part presents a running commentary on most of the Bible, and this is an Arabic abridgement of Ishoʿdad of Merv’s commentary in Syriac.66 The other part of The Paradise of Christianity is a series of questions and answers on the entire Bible, which is again an Arabic abridgement but this time of Theodoros bar Koni’s Scholion.67 Here, again, we see the systematic reworking of the earlier Syriac heritage as expressed in the Syriac language for an Arabic-reading audience. Thus, Ibn al-Ṭayyib chose not only to write in Arabic instead of Syriac but also — and this is important — to transfer Syriac writings into Arabic. I am convinced that this move is representative of a broader trend starting around the turn of the millennium in which Arabic increasingly came to displace Syriac in the writings of Syriac Christians. Nevertheless, despite this transition from Syriac to Arabic, Syriac continued — and, for that matter, continues — to be used among Syriac Christians. Written Syriac, for instance, experienced what has been termed a renaissance in the thirteenth century.68 This renaissance culminated with the polymath Bar Hebraeus (d. 1286), who wrote more than forty works on a wide range of topics, including exegesis, theo­logy, philosophy, history, grammar, and science, mostly in Syriac but also in Arabic.69 A similar renaissance of Syriac can also be observed at this time in the Church of the East. Consider, for instance, ʿAbdishoʿ bar Brikha (d. 1318), who wrote important works on theo­logy, canon law, liturgy, and theo­logical poetry, as well as a catalogue of Syriac literature, all in Syriac.70 And although Syriac literature that post-dates the thirteenth century has traditionally incited little interest in Syriac scholarship, it does exist.71 Examples include the poetic compositions in Syriac by two fifteenth-century authors, Isḥaq Shbadnaya of the Church of the East and Dawid Puniqoyo of the Syriac Orthodox Church.72 For that matter, classical Syriac

66 Only the running commentary on Genesis has been edited, as Commentaire sur la Genèse, ed. by Sanders. For Ishoʿdad of Merv as the source of this part, see Sanders, Inleiding op het Genesis­ kommentaar van de Nestoriaan Ibn at-Taiyib; Ibn al-Ṭayyib, Commentaire sur la Genèse, pp. ii–iii (‘la source principale’); Cowley, Ethiopian Biblical Interpretation, p. 66; Féghali, ‘Ibn aṭ-Ṭayyib et son commentaire sur la Genèse’; Faultless, ‘Ibn al-Ṭayyib’, pp. 669, 681; Butts, ‘Embellished with Gold’, pp. 140–45. 67 This part remains entirely unedited apart from a small excerpt from Genesis in Butts, ‘In Search of Sources for Ibn al-Ṭayyib’s The Paradise of Christianity’, where I argue, with further references, that Bar Koni is the immediate source of the material. 68 In general, see Teule, ‘Renaissance, Syriac’ as well as Teule and Tauwinkl (with Ter Haar Romeny and van Ginkel), The Syriac Renaissance. 69 See the extremely useful Takahashi, Barhebraeus. 70 See Childers, ‘ʿAbdishoʿ bar Brikha’ as well as Rassi, Christian Thought in the Medi­eval Islamicate World. 71 See the detailed survey in Macuch, Geschichte der spät- und neusyrischen Literatur. 72 For the former, see Carlson, ‘A Light from “The Dark Centuries”’ as well as Carlson, Christianity in Fifteenth-Century Iraq. For the latter, see Butts, ‘The Afflictions of Exile’, with additional references.

37 7

3 78

a a ro n m icha e l b u t t s

continues to function today as a liturgical and literary language for Syriac Christians both in the Middle East and in the worldwide diaspora.73 To summarize, the writing of Syriac did not end with the rise of Islam; far from it! Rather, Syriac continued to flourish at least up to the turn of the millennium. At this time, we start to see Arabic displacing Syriac more and more, but Syriac never became moribund, and it even witnessed periods of increased use, as during the Syriac Renaissance of the thirteenth century. In light of the long history of Syriac — almost two millennia — it is worth reflecting on an additional complicating factor: the socio-linguistic context. There is, I am convinced, a key socio-linguistic difference between the period of Syriac writing that I have been discussing here, which post-dates the rise of Islam (we might call this Post-Classical Syriac), and the earlier periods. In contrast to earlier Syriac, Post-Classical Syriac was never a primary spoken language and perhaps not a native language either.74 After the rise of Islam, part of the Syriac Christian community undoubtedly continued to speak a variety of Aramaic as its native language. There must, however, have been an ever-growing distance between the written and spoken forms of Aramaic. Consider, for instance, Syriac Christians who spoke Neo-Aramaic in the nineteenth and early twentieth centuries, whether Ṭuroyo or related dialects in the Ṭur ʿAbdin region of southeast Turkey or one of the North-Eastern Neo-Aramaic dialects (= NENA) in south-eastern Turkey, northern Iraq, and north-western and western Iran.75 Though the details will have varied by community, in each case (Post-)Classical Syriac functioned as a literary language, a liturgical language, and even, at times, a formal spoken language, and a Neo-Aramaic dialect served as the (primary) spoken language. Thus, we have a diglossic situation, depending on how one exactly defines diglossia.76 This much is uncontroversial. What is important for the present discussion, however, is the question of how far back this diglossic

73 Brock, ‘Some Observations on the Use of Classical Syriac in the Late Twentieth Century’; Kiraz, ‘Kthobonoyo Syriac’; Knudsen and Wardini, Neo­logisms in Modern Literary Syriac. 74 I introduced the term ‘Post-Classical Syriac’ in Butts, ‘The Classical Syriac Language’, p. 231, where a complementary discussion can be found. 75 For a helpful overview of Neo-Aramaic in the context of Syriac Christianity, see Khan, ‘The NeoAramaic Dialects and their Historical Background’. For Neo-Aramaic more broadly, see the overviews in Khan, ‘The Neo-Aramaic Dialects of Eastern Anatolia and Northwestern Iran’, ‘The Neo-Aramaic Dialects of Northern Iraq’, and ‘The Neo-Aramaic Dialects of Western Iran’. 76 In a now classic article, Ferguson defined diglossia as ‘a relatively stable language situation in which, in addition to the primary dialects of the language (which may include a standard or regional standards), there is a very divergent, highly codified (often grammatically more complex) superposed variety, the vehicle of a large and respected body of written literature, either of an earlier period or in another speech community, which is learned largely by formal education and is used for most written and formal spoken purposes but is not used by any section of the community for ordinary conversation’ (‘Diglossia’, p. 336). The concept of diglossia has since been expanded by various scholars, perhaps most prominently by Fishman (especially in his ‘Bilingualism with and without Diglossia’), to the point that diglossia hardly differs from (societal) bilingualism more broadly (in line with its etymo­logy: Greek for ‘two languages’) or even from sociolect variation involving a single language. For histories of research, see Martin-Jones, ‘Diglossia’; Sebba, ‘Societal Bilingualism’, pp. 449–53; Jaspers, ‘Diglossia and Beyond’ — all with numerous additional references. In the case

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

situation extends.77 Personally, I believe — and this is controversial — that diglossia obtains for much, if not all, of the Syriac written after the rise of Islam. There is a good deal of evidence that I could elicit for this claim, but I will mention here only a small portion of it which involves Neo-Aramaic. Though Neo-Aramaic is generally considered a modern, spoken language, it is also attested in written form: the earliest examples among Christians go back to the sixteenth century. These consist of a body of religious poetry written in a NENA koine based on the dialect of Alqosh (and possibly also of Telkepe).78 These texts display a fully developed Neo-Aramaic, the incipient form(s) of which must go back several centuries, given the amount of time necessary for the witnessed changes, such as the restructuring of the verbal system, to take place. What is more, the NENA dialects do not derive directly from Syriac but rather find their ancestors in different dialects of Aramaic. This all points to a diglossic situation for native Aramaic speakers, in which their spoken varieties of Aramaic diverged more and more from Syriac, beginning perhaps as early as the turn of the millennium or even earlier.79 After the rise of Islam, Aramaic would not, of course, have been the primary spoken language of all Syriac Christians. Rather, Arabic would have increasingly served as the primary spoken language and often as the native language. Thus, some of the authors that I have mentioned in this section, such as Bar Hebraeus, also wrote in Arabic. For many others, we can imagine that Arabic served as their primary spoken language, if not their native language. With this segment of the population, we are again faced with a socio-linguistic situation in which one language, Syriac, functioned as a literary language, a liturgical language, and perhaps even, at times, a formal spoken language, while another language, this time Arabic, served as the (primary) spoken language. The situation is further complicated by the fact that Arabic itself was diglossic at this time, as I will discuss at the end of this paper. For now, however, it seems clear enough that these multilingual, diglossic situations add further texture to our understanding of the continued use of Syriac after Islam.

presently under discussion involving Aramaic, we are probably not too far from Ferguson’s original definition. 77 The same problem exists for Arabic, as I will discuss below. 78 See the texts edited, with English translations, in Mengozzi, Israel of Alqosh and Joseph of Telkepe and Religious Poetry in Vernacular Syriac from Northern Iraq. 79 It is interesting to note that Bar Hebraeus, though writing a bit later than the time I am discussing here, distinguishes various dialects of Aramaic: ‘Aramaic/Syriac (al-suryāniyya): God spoke to Adam in it, and it is divided into three languages/dialects (lughāt). The purest is al-ārāmiyya, which is the language of the people of Edessa, Ḥarrān, and outer Syria. After it there is al-falasṭīniyya, which is the language of the people of Damascus, the mountains of Lebanon, and the rest of inner Syria. The worst of these is al-kaldāniyya, [i.e.?] al-nabaṭiyya, which is the language of the people of the mountains of Assyria and southern Iraq’ (Taʾrīkh mukhtaṣar al-duwal, ed. by Ṣālḥānī, p. 18). It is tempting to see these at least partly as the precursors to Central Neo-Aramaic (Ṭuroyo, etc.), Western Neo-Aramaic (Maʿlula, etc.), and NENA, respectively. Regardless, my point here is that Bar Hebraeus, writing in the thirteenth century, distinguished between the Aramaic spoken in the plain of Nineveh and areas south (which can basically only be the predecessor to NENA) from the Aramaic spoken in Edessa.

37 9

3 80

a a ro n m icha e l b u t t s

Language Use at al-Ḥīra In the previous two sections, I explored the use of Arabic before Islam and that of Syriac after Islam on the macro level. I now adopt a narrower focus, looking first at language use in al-Ḥīra as an example of Arabic before Islam in a particular location among a specific Christian population, the ʿIbād, and finally at the language use of Ḥunayn b. Isḥāq as an example of the employment of Syriac in Islamic times. The city of al-Ḥīra (Syriac Ḥirtā) was located on the Euphrates not far from the modern city of Najaf in south central Iraq.80 It was the capital of the Naṣrid dynasty (c. 300 until 602 ce),81 and Christians had a long presence there.82 The earliest known bishop is one Hoshaʿ, who is listed in the acts of the so-called Synod of Seleucia-Ctesiphon (410).83 The city of al-Ḥīra would ultimately be associated with East Syriac Dyophysite Christianity, which would develop into the Church of the East.84 A number of East Syriac catholicoi resided there, and several were also buried there, including important figures such as Dadishoʿ I (in office 421–456), Aba I (in office 540–52), and Ishoʿyahb I (in office 582–95). In later Islamic times, the Christian population of al-Ḥīra was known as the ʿIbād, in the sense of ‘slave’ (ʿabd) of Christ — compare the name of ʿAbdā da-Mšiḥā mentioned above.85 What is important for our purposes here is that al-Ḥīra is widely thought to have been primarily Arabic-speaking, even in pre-Islamic times.86 I do not have space here to rehearse all of the evidence, but I want to look briefly at a single passage. In his History of the Prophets and Kings, al-Ṭabarī relates the following narrative about Khālid b. Walīd’s encounter with the inhabitants of al-Ḥīra: 80 In general, see Fisher and Wood, ‘Writing the History of the “Naṣrid” Dynasty at al-Ḥīra’; Kister, ‘Al-Ḥīra’; Rothstein, Die Dynastie der Laḫmiden in al-Ḥîra; Toral-Niehoff, ‘The ʿIbād of al-Ḥīra’; ToralNiehoff, Al-Ḥîra; Toral-Niehoff, ‘Late Antique Iran and the Arabs’; Wood, ‘Al-Ḥīra and its Histories’; and a number of contributions to Fisher ed., Arabs and Empires before Islam, including Genequand, ‘The Archaeo­logical Evidence for the Jafnids and the Naṣrids’, pp. 207–12; Munt, with others, ‘Arabic and Persian Sources for Pre-Islamic Arabia’, pp. 454–67, 488–90; and Wood, with Greatrex, ‘The Naṣrids and Christianity in al-Ḥīra’, pp. 172–213, 257–63. 81 In earlier scholarship, ‘Lakhmid’ was often used for ‘Naṣrid’. See Fisher, ‘Editor’s Introduction’, in Arabs and Empires before Islam, pp. 6–8 as well as Fisher, ‘Kingdoms or Dynasties?’. 82 In general, see Hunter, ‘The Christian Matrix of al-Hira’; Toral-Niehoff, ‘The ʿIbād of al-Ḥīra’; ToralNiehoff, Al-Ḥîra, pp. 88–105, 151–211; Wood, ‘Al-Ḥīra and its Histories’, pp. 793–97. 83 Synodicon Orientale ou recueil de synodes nestoriens, ed. by Chabot, pp. 36, 275; German translation in Das Buch der Synhados oder Synodicon Orientale, trans. by Braun, p. 35. 84 Note that there was at various points in time a Miaphysite presence in al-Ḥīra as well (Wood, ‘Al-Ḥīra and its Histories’, p. 794 with n. 65). In addition, I should point out that Toral-Niehoff has criticized the use of the term ‘Dyophysite’, charging that it ‘is also problematic, since it implies a belief in two natures in Christ, which the “Dyophysitists” would deny’ (Toral-Niehoff, ‘The ʿIbād of al-Ḥīra’, p. 15 n. 82). She maintains this position also in her book, published a couple of years later: ‘Diophysiten ist sachlich falsch und ebenfalls polemisch’ (Al-Ḥîra, p. xvi). This is simply untrue. It is not factually wrong to say that this group of Christians confesses two natures in Christ: they do. For a general discussion, see Brock, ‘The “Nestorian” Church’. 85 For understanding ʿIbād as ‘slaves (of Christ)’, see already Nöldeke, Geschichte der Perser und Araber zur Zeit der Sasaniden, p. 24 n. 4; Rothstein, Die Dynastie der Laḫmiden in al-Ḥîra, p. 21. 86 For language use in al-Ḥīra, see Toral-Niehoff, Al-Ḥîra, pp. 113–24, especially pp. 114–20 on Arabic.

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

‫فخال خالد بأهل كل قصر منهم دون االخرين وبدأ باصحاب عدي وقال ويحكم ما انتم اعرب فما‬ ‫تنقمون من العرب او عجم فما تنقمون من االنصاف والعدل فقال له عدي بل عرب عاربة وأخرى‬ ‫متعربة فقال لو كنتم كما تقولون لم تحادونا وتكرهوا امرنا فقال له عدي ليدلك على ما نقول انه ليس‬ ‫لنا لسان اال بالعربية فقال صدقتَ فقال اختاروا واحدة من ثلث أن تدخلوا في ديننا فلكم ما لنا وعليكم‬ ‫ما علينا ﺇن نهضتم وهاجرتم وﺇن اقمتم في دياركم او الجزية او المنابذة والمناجزة فقد وهللا اتيتكم‬ ‫بقوم هم على الموت احرص منكم على الحياة فقال بل نعطيك الجزية فقال خالد تبا لكم ويحكم ﺇن‬ ‫الكفر فالة مضلة فاحمق العرب من سلكها‬

Khālid was alone with the people of each fortress without the others. He began with the companions of ʿAdī, saying, ‘Woe to you! You are not Arabs (aʿrub). Why do you take revenge on the Arabs (ʿarab) or non-Arabs (ʿajam)? Why do you take revenge on fairness and justice?’ ʿAdī said, ‘Indeed, we are pure Arabs (ʿarab ʿāriba), and others are arab(ic)ized (mutaʿarriba) [Arabs].’ [Khālid] said, ‘Had you been as you say, you would not have opposed us and despised the command to us.’ ʿAdī said to him, ‘The fact that we have no language apart from Arabic proves what we say to you.’ Khālid said, ‘You are right.’ Then he said, ‘Choose one of the following three: Enter into our religion, and you will have what we have, and there will be on you what is on us, whether you arise and migrate or whether you stay in your homes. Or there is jizya, or resistance and fighting. By God, I have brought a people who desire death more than you [desire] life!’ [ʿAdī] said, ‘Indeed, we will give you the jizya.’ Khālid said, ‘May you perish! Woe to you! Disbelief is a desert that leads astray, for the one who travels through it is the most foolish of the Arabs (ʿarab).’87 This passage is obviously very rich, and I cannot do full justice to it here. But note that the inhabitants of al-Ḥīra, in the voice of ʿAdī, identify as Arabs (ʿarab) as opposed to non-Arabs (ʿajam).88 As Arabs, they then are further divided into two groups: pure Arabs (ʿarab ʿāriba) and Arabs who have been arab(ic)ized (mutaʿarriba).89 What is particularly interesting is that Khālid initially rejects this claim, and the inhabitants of al-Ḥīra respond by saying that their identity as Arabs is proven by the fact that they speak Arabic.90 Active in the latter half of the ninth century up through the first quarter of the tenth century, al-Ṭabarī is, of course, writing long after the events being narrated here allegedly took place (in 633), and in addition he has a clear pro-Arab agenda. Nevertheless, the depiction

87 Al-Ṭabarī, Annales quos scripsit Abu Djafar Mohammed ibn Djarir at-Tabari, ed. by de Goeje, iv, p. 2041. An English translation is available in al-Ṭabarī, The History of al-Ṭabarī, trans. by Blankinship, xi, p. 31. 88 Traditionally ʿajam is understood to refer to Persians/Iranians. More recently, Borrut has suggested that terms such as ʿarab and ʿajam are not always to be read as ethnic but may at least in some cases track along spatial and/or geo­graphic lines (Entre mémoire et pouvoir, pp. 330–51). 89 Note that in a related account transmitted by al-Masʿūdī the answer is ‘We are Nabateanized Arabs and Arab(ic)ized Nabateans’ (ʿarabun istanbaṭnā wa-nabaṭun istaʿrabnā) (edited with French translation in Les prairies d’or, ed. and trans. by Barbier de Meynard and Pavet de Courteille, i, 218, revised in Murūj al-dhahab wa-maʿādin al-jawhar, ed. by Pellat, § 231). 90 For the construction of ‘Arabs’, see the biblio­graphy in n. 21 above.

381

382

a a ro n m icha e l b u t t s

of the inhabitants of al-Ḥīra as speakers of Arabic does fit what we know about al-Ḥīra from other sources.91 Al-Ḥīra thus provides a more specific example of the use of Arabic among (Syriac) Christians before Islam.

Ḥunayn b. Isḥāq One of the most prominent figures to come from al-Ḥīra is Ḥunayn b. Isḥāq, who is best known for the central role that he played in the so-called Graeco-Arabic translation movement of the early Abbasid period.92 Ḥunayn, whose full name was Abū Zayd Ḥunayn b. Isḥāq b. Sulaymān b. Ayyūb al-ʿIbādī, was born in 808 near al-Ḥīra, where his father was a pharmacist.93 He studied medicine in Baghdad under the famous physician Yūḥannā b. Māsawayh (d. 857), who stemmed from one of the prominent medical families of Jundishapur. In Baghdad, Ḥunayn became well known as one of the foremost translators of Greek texts into Syriac and Arabic. In addition to his translation activity, Ḥunayn was also the personal physician of Caliph al-Mutawakkil (r. 847–61). He died in Baghdad in 873 (or, less likely, in 877).94 I am particularly interested here in reflecting on Ḥunayn’s language use. One often finds statements such as the following in the secondary literature: ‘Ḥunayn grew up in a bilingual environment, fluent in both Syriac and Arabic’.95 In fact, I myself have written in a similar vein: ‘He likely grew up bilingual in Arabic and Syriac, and he acquired an excellent knowledge of Greek as well as Persian in the course of his education’.96 Such statements, however, do little more than echo the Arabic bio­graphical sources.97 In his Kitāb al-Fihrist, for instance, Ibn al-Nadīm (fl. 987) describes Ḥunayn as skilled (faṣīḥan) in Greek, Syriac, and Arabic.98 In his ʿUyūn al-anbāʾ fī ṭabaqāt al-aṭibbāʾ, Ibn Abī Uṣaybiʿa (d. 1270) adds Persian to the list: ‘Ḥunayn was the most knowledgeable (aʿlam) of the people of his time in Greek, Syriac, and Persian […] while he also persisted in mastering Arabic, occupying himself with it until he became one of those distinguished in it’.99 What I seek at

91 See references in n. 86 above. 92 For the Graeco-Arabic translation movement, see Gutas, Greek Thought, Arabic Culture. 93 For Ḥunayn in general, see Anawati and Iskandar, ‘Ḥunayn ibn Isḥāq al-ʿIbādī, Abū Zayd’; Bergsträsser, Ḥunain Ibn Ishaḳ und seine Schule; Butts, ‘Ḥunayn b. Isḥāq’; Graf, Geschichte der christlichen arabischen Literatur, ii, pp. 122–29; Gutas, ‘Scholars as Transmitters of Philosophical Thought’, pp. 680–704; Ḥabbi, Ḥunayn b. Isḥāq; Ḥunayn b. Isḥāq, Ḥunayn ibn Isḥāq on his Galen Translations, ed. by Lamoreaux, pp. xii–xviii; Meyerhof, ‘New Light on Ḥunain ibn Isḥāq and his Period’; Monferrer Sala and Roggema, ‘Ḥunayn ibn Isḥāq’; Strohmaier, ‘Ḥunayn ibn Isḥāq und die Bilder’; Strohmaier, ‘Ḥunayn ibn Isḥāq’; Ullmann, Die Medizin im Islam, pp. 115–19, 205–07. 94 For the date of his death, see Graf, Geschichte der christlichen arabischen Literatur, ii, p. 123 n. 2. 95 Ḥunayn b. Isḥāq, Ḥunayn ibn Isḥāq on his Galen Translations, ed. by Lamoreaux, p. xii. 96 Butts, ‘Ḥunayn b. Isḥāq’, p. 205. 97 For an interesting study of the reception of Ḥunayn in these medi­eval sources, see Olsson, ‘The Reputation of Ḥunayn ibn Isḥāq in Contemporaneous and Later Sources’. 98 Ibn al-Nadīm, Kitâb al-Fihrist, ed. by Flügel, Müller, and Roediger, p. 294, line 18. 99 Ibn Abī Uṣaybiʿa, Ibn Abi Useibia, p. 186, lines 21–24.

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

least to initiate here is a more nuanced discussion of Ḥunayn’s language use that moves beyond the medi­eval bio­graphical sources. At the outset, we should bear in mind the broader linguistic landscape of al-Ḥīra, as discussed above: there is a long history of the use of Arabic in al-Ḥīra even prior to the rise of Islam. Thus, there is a strong possibility that Ḥunayn’s native language would have been Arabic, and Syriac would have been a learned language.100 Given this, I am particularly interested in exploring the role that Syriac plays for Ḥunayn. Ḥunayn authored a number of works, most of which have to do with medicine. In his ʿUyūn al-anbāʾ fī ṭabaqāt al-aṭibbā, for instance, Ibn Abī Uṣaybiʿa attributes more than one hundred works to Ḥunayn.101 Many of these are preserved only in Arabic, with no Syriac versions surviving. But arguably the most important of Ḥunayn’s medical treatises was his Book of Questions on Medicine (Kitāb masāʾil fī al-ṭibb), which survives in both Arabic and Syriac versions.102 In addition to medicine, Ḥunayn is also said to have authored works on grammar and lexico­graphy as well as on theo­logy. In general, titles are known (or works survive) in Syriac and in Arabic.103 So, in his own writings, Ḥunayn used either Syriac or Arabic, and in some cases a work exists in both. This multilingual picture of Ḥunayn’s oeuvre is corroborated by his translation activity. Ḥunayn and his ‘school’, including his son Isḥāq b.  Ḥunayn and his nephew Ḥubaysh b. al-Ḥasan, as well as ʿĪsā b. Yaḥyā and others, translated well over a hundred Greek medical works, including writings by Hippocrates and

100 If so, then Ḥunayn would be an exception to Hoyland’s generalization that ‘men competent in all three languages [that is, Aramaic, Arabic, and Greek; AMB] were far more likely to be native Aramaic-speakers than Greek- or Arabic-speakers’ (‘Language and Identity’, pp. 195–96). As a generalization, this is likely true: Syriac speakers were, in general, more likely to learn Arabic than Arabic speakers, especially Muslim ones, were to learn Syriac (for a comparable situation earlier in the history of Syriac, involving Greek, see Butts, Syriac in its Greco-Roman Context, pp. 30–40). But this generalization probably does not apply to Ḥunayn and others from al-Ḥīra. See similarly Toral-Niehoff, ‘The ʿIbād of al-Ḥīrah’, p. 18 (‘The ʿIbād of al-Ḥīrah used Syriac as their church language, as did most of the Christians in Sasanian Iran, while their colloquial language was Arabic’); Toral-Niehoff, Al-Ḥîra, p. 121 (‘Das Syrische war somit die gemeinsame Kirchensprache für die aramäischsprachige christliche Landbevölkerung und die arabischen Christen in al-Ḥīrah’ [emphasis mine; AMB]); Gutas, ‘Scholars as Transmitters of Philosophical Thought’, p. 683 (‘It is clear that he [Ḥunayn; AMB] must have been trilingual, speaking Arabic at home and studying Syriac and Greek at school’); Strohmaier, ‘Ḥunayn ibn Isḥāq’, p. 164 (‘Arabic was his mother tongue, and at school he had to learn the psalms in Syriac’). 101 Ibn Abī Uṣāybiʿa, Ibn Abi Useibia, pp. 184–200. 102 Arabic edited by Abū Rayyān, ʿArab, and Mūsā in al-Masāʾil fī al-ṭibb li-l-mutaʿallimīn li-Ḥunayn b. Isḥāq; English translation in Questions on Medicine for Scholars by Ḥunayn ibn Isḥāq, trans. by Ghalioungui. For the Syriac version, see Degen, ‘The Oldest Known Syriac Manu­script of Ḥunayn b. Isḥāq’. The Arabic version was translated into Latin under the title Isagoge Johannitii, in which form it remained authoritative for medi­eval Europe (see Newton, ‘Constantine the African and Monte Cassio’). 103 For inventories of Arabic works, see Brockelmann, Geschichte der arabischen Litteratur, i, pp. 224–27; suppl. i, pp. 366–39; Sezgin, Geschichte des arabischen Schrifttums, iii, pp. 247–56; Gutas, ‘Scholars as Transmitters of Philosophical Thought’, pp. 685–704. For Ḥunayn’s Syriac works, see Macomber, ‘The Literary Activity of Hunain b. Ishaq in Syriac’.

383

3 84

a a ro n m icha e l b u t t s

Dioscorides as well as almost the entire corpus of Galen. In addition to medicine, they translated Greek philosophical and scientific texts, including Plato, Aristotle, Proclus, and Porphyry. A number of fascinating insights into the method of Ḥunayn and his ‘school’ can be found in Ḥunayn’s Letter to ʿAlī b. Yaḥyā on Galen’s Books which Have Been Translated … (Risāla ilā ʿAlī ibn Yaḥyā fī dhikr mā turjima min kutub Jālīnūs …).104 In the Letter, Ḥunayn gives us fascinating insights into his translation activity, including that he travelled throughout Mesopotamia, Syria, Palestine, and Egypt in order to procure Greek manu­scripts (§ 126; see also § 5, § 85, and passim); that he often collated multiple (Greek) manu­scripts to obtain a sound textual basis for his translations (§ 5, § 22, and passim); that he frequently consulted and/or revised earlier (Syriac) translations of Greek works (§ 22 and passim); and that he preferred a sensus de sensu, or reader-oriented, approach to translation, though sometimes his patrons preferred a more literal, verbum e verbo, or source-oriented, approach (§ 9, § 59.3).105 The Letter may also give us some insights into Ḥunayn’s language use. According to the Letter, Ḥunayn (and his ‘school’) translated the Greek texts of Galen into Syriac, Arabic, or both. The target language seems to have been, at least primarily, a function of the request of the particular patron, and so it is not directly indicative of Ḥunayn’s own personal preferences. There are, however, some potentially interesting details lurking beneath the surface.

104 This work has a complicated transmission and publication history. Ḥunayn originally wrote it in Syriac, then translated it into Arabic in 856 and later revised it in 864. Only the later revision survives, itself in two recensions. The earlier recension (termed B in the scholarly literature) survives in Istanbul, MS Ayasofya 3590, with some additional information (lists of titles and translators) in Istanbul, MS Ayasofya 3593. The later recension (termed A) survives in Istanbul, MS Ayasofya 3631 as well as in indirect transmission in later biblio­graphers, such as Ibn al-Nadīm’s Kitāb al-Fihrist and Ibn Abī Uṣaybiʿa’s ʿUyūn al-anbāʾ fī ṭabaqāt al-aṭibbā. In 1925, Bergsträsser published an edition of recension A on the basis of the only known manu­script at the time, MS Ayasofya 3631, though also taking account of the indirect witnesses (Ḥunain Ibn Isḥāq über die syrischen und arabischen GalenÜbersetzungen). Only later did Bergsträsser learn of MS Ayasofya 3590 and its B recension, which prompted him to publish a supplemental volume that included collations (in transcription) and other integral material (Neue Materialien zu Ḥunain ibn Isḥāq’s Galen-Biblio­graphie). The witness of MS Ayasofya 3593 was discovered and published only fairly recently by Käs (‘Eine neue Handschrift von Ḥunain ibn Isḥāq’s Galenbiblio­graphie’). Even more recently, Lamoreaux has published a new edition and translation of the text (Ḥunayn ibn Isḥāq on his Galen Translations), but as pointed out by Gutas (‘A New “Edition” of Ḥunayn’s Risāla’), this publication suffers from numerous problems, most crucially in the establishment of the Arabic text itself. A new comprehensive, critical edition remains a desideratum. In the following, I refer to the text by Lamoreaux’s section numbers only because one can easily cross-reference from his edition to Bergsträsser’s, whereas the reverse is much more difficult. 105 For the translation method of Ḥunayn and his colleagues, see the classic article of Brock, ‘The Syriac Background to Ḥunayn’s Translation Techniques’ as well as the more recent studies of Arnzen, ‘Proclus on Plato’s Timaeus 89e3–90c7’; Cooper, ‘Ḥunayn ibn Isḥāq’s Galen Translations and GrecoArabic Philo­logy’; Overwien, ‘The Art of the Translator’; Pormann, ‘The Development of Translation Techniques from Greek into Syriac into Arabic’; Vagelpohl, ‘In the Translator’s Workshop’.

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

In some cases, when an Arabic translation was needed, it seems that the Greek text was first translated into Syriac, and that was then translated into Arabic.106 For instance, Ḥunayn says the following about the translation history of Galen’s ‘Unknown Motions’ (§ 49): ‫ ثم اني من بعد ترجمتها‬.‫ فاما انا فلم اكن تفرغت لنقله على انه قد كان عندي نسخة‬.‫وقد نقلها ايوب‬ ‫الى السريانية ثم الى العربية البي جعفر‬

Ayyūb translated it. As for me, I did not have the opportunity to translate it, though I had a copy. I then later translated it into Syriac and then into Arabic for Abū Jaʿfar. (§ 49.3–5) There are various possible ways to understand the last sentence, but the most straightforward is that Ḥunayn first translated the text into Syriac and only then into Arabic (see the same syntax in § 101.8). Note that Abū Jaʿfar is mentioned frequently in the Letter, but he never receives a Syriac copy. In the vast majority of cases in the Letter, Ḥunayn names the patron of each translation, though there are exceptions. Thus, it seems that in this particular case Ḥunayn opted for Syriac to serve as a bridge between the Greek and the Arabic even though there was no patron for the Syriac text.107 Various possible explanations for this approach have been suggested in the secondary literature, but the most likely is that there was a long-established method for rendering Greek into Syriac, developed over centuries, and so it was in some sense easier to translate the Greek into Syriac (rather than directly into Arabic); once there was a Syriac text, this could then more easily be rendered into another Semitic language such as Arabic.108 If this argument is accepted, one might think of Ḥunayn as ‘more fluent’ in Syriac than in Arabic when translating Greek texts. In more technical, linguistic terms, Ḥunayn would have been linguistically dominant in Syriac (and not in Arabic) in this particular socio-linguistic situation.109 Compare this with the classic example discussed by the linguist U. Weinreich: ‘A child learning both languages in its familial and play

106 It should be noted that this trajectory of Greek > Syriac > Arabic can be observed not only in Ḥunayn’s Letter but also, inter alia, in the Fihrist of Ibn al-Nadīm, the Taʾrīkh al-ḥukamāʾ of Ibn al-Qifṭī (d. 1248), and the ʿUyūn al-anbāʾ fī ṭabaqāt al-aṭibbā of Ibn Abī Uṣaybiʿa, where the trajectory is found with translations from Ḥunayn and his ‘school’ as well as from others. I should, however, also point out that it is possible that the Arabic was translated not from the Syriac but from the Greek. In addition, the Greek could presumably have been consulted when translating from Syriac to Arabic. 107 This is not to imply that Syriac always served as a bridge (see, e.g., § 129, § 137). 108 This is often maintained in the scholarly literature, especially in Syriac studies; see, for example, Brock, ‘Changing Fashions in Syriac Translation Technique’. For an insightful study of the broader historical background for Syriac’s role as an intermediary in the so-called Graeco-Arabic translation movement, see Takahashi, ‘Syriac as the Intermediary in Scientific Graeco-Arabica’. For translations from Greek to Syriac, see King, The Syriac Versions of the Writings of Cyril of Alexandria. For a possible precursor, though it is far from certain, see Knauf, ‘Arabo-Aramaic and ʿArabiyya’, pp. 201–02. 109 I adopt the technical term linguistic dominance from the contact linguist F. Van Coetsem (Loan Phono­logy and the Two Transfer Types in Language Contact, pp. 13–17; ‘Outlining a Model of the Transmission Phenomenon in Language Contact’, pp. 70–72; A General and Unified Theory of the Transmission Process in Language Contact, pp. 32, 42, 49, 58–62, 66–67).

385

3 86

a a ro n m icha e l b u t t s

environment […] may be equipped to deal with everyday things in both tongues; but if it studies certain subjects in a unilingual school, it will have difficulty in discussing these “learned” topics in the other language’.110 Might we imagine a similar situation, mutatis mutandis, for Ḥunayn? This line of questioning takes on an additional layer of interest when we note that in many cases mentioned in the Letter to ʿAlī b. Yaḥyā it is not Ḥunayn who does the second translation from Syriac to Arabic but one of his students. Consider, for instance, the case of Galen’s ‘Composition of Drugs’ (§ 84): ‫ ونقلته انا في خالفة امير المؤمنين المتوكل‬.‫وقد كان نقل هذا الكتاب الى السريانية سرجس الرأسي‬ .‫ ونقله من نقلي حبيش بن الحسن البي جعفر محمد بن موسى‬.‫ليحنى بن ماسويه المتطبب‬

Sergis al-Raʾsī translated this book into Syriac. I myself translated it during the caliphate of the Commander of the Faithful, al-Mutawakkil, for Yūḥannā b. Masāwayh, the physician. Ḥubaysh translated it from my translation for Abū Jaʿfar Muḥammad bin Mūsā. (§ 84.6–8) This is one of many cases in which Ḥunayn first renders a text into Syriac and then one of his students translates his Syriac into Arabic (for another interesting example, see § 18).111 Such a process could be motivated by any of a number of factors: – Perhaps Ḥunayn was better than Ḥubaysh at translating from Greek to Syriac? Or perhaps Ḥubaysh was better than Ḥunayn at translating from Syriac to Arabic? These possibilities are not mutually exclusive. – Perhaps Ḥunayn was better at Greek than Ḥubaysh was? Or perhaps he was better at Syriac? – Perhaps a generational factor was also at play: Are we to read into cases such as this a generational change in language use from Ḥunayn to Ḥubaysh? – Or perhaps it was deemed a better use of resources to have Ḥunayn, the master, only oversee what might have been a relatively easier task of translating from Syriac to Arabic. That is, Ḥunayn could have been equally competent in all of these languages (Greek, Syriac, and Arabic) and tasks (translation from Greek to Syriac and translation from Syriac to Arabic). But even if so, the fact that Ḥunayn first rendered the text into Syriac and then Ḥubaysh translated Ḥunayn’s Syriac into Arabic would still presumably tell us something about the language skills of Ḥubaysh.112 These questions take on additional complications when set within the broader language landscape of al-Ḥīra. As discussed above, there is a strong possibility that

110 Weinreich, Languages in Contact, p. 81. 111 This in fact happens often: § 10.4; § 27, § 28, § 29, § 30, § 31, § 34, § 43, § 45, § 47, § 48, § 54, § 58, § 63, § 65, § 77. The opposite process may possibly be found in § 4. 112 Note, however, that Ḥubaysh does seem to have been proficient in Syriac, since he translated from Arabic to Syriac (§ 38, § 40).

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

Ḥunayn’s native language would have been Arabic, and Syriac would have been

his learned language. How does this likelihood alter our view of the situation? It is difficult to escape the conclusion that Ḥunayn was stronger in Greek and Syriac, at least relative to Ḥubaysh — at the very least, he was more active on the Greek and Syriac side.113 It is also at least possible that Ḥubaysh had an advantage over Ḥunayn in Arabic. A final variable that I want to bring up here is the question of what exactly we mean by Arabic in this context. I have already introduced the concept of diglossia above. Diglossia gained widespread usage in the linguistic literature thanks to a 1959 article by C. A. Ferguson, in which he described language use in Arabic-speaking countries in the modern period as follows: alongside the spoken dialects of Arabic there is another, superposed variety that is used for literature and writing more broadly.114 What is interesting for the present paper is that Arabic has likely been diglossic since at least Late Antiquity.115 No longer can we maintain that the modern Arabic dialects are daughters of Classical Arabic.116 Rather, the modern dialects are continuations of ancient Arabic dialects that must have been in use alongside the literary form of the language from at least the beginning of the Islamic period.117 One clear indication of this is Middle Arabic.118 Middle Arabic refers to varieties of written Arabic in which features of Arabic dialects are found. To be clear, Middle Arabic is not a written form of the spoken language, but rather a written form of the classical language in which dialectic features also appear. Middle Arabic is most often found among Christians and Jews, since these communities were, at least in certain contexts, not as tied to the norms of Classical

113 In his Kitāb al-Fihrist, Ibn al-Nadīm already noted this: ‘If we return to the catalogue of the books of Galen that Ḥunayn made for ʿAlī b. Yaḥyā, we learn that most of what Ḥunayn translated was into Syriac, though sometimes he corrected and examined the Arabic of the translations of others’ (Kitâb al-Fihrist, ed. by Flügel, Müller, and Roediger, p. 289, lines 16–18). 114 See n. 76 above. 115 As will become immediately clear, I mean this in a very different sense from that intended by Owens; see, for instance, his A Linguistic History of Arabic. My view is closely aligned with that of Huehnergard, ‘Arabic in its Semitic Context’, pp. 12–14. 116 See, e.g., Huehnergard, ‘Arabic in its Semitic Context’, p. 13; Al-Jallad, ‘What Is Ancient North Arabian?’, p. 8 n. 23. 117 For one such ancient Arabic dialect, see Al-Jallad’s proposed Ancient Levantine Arabic (‘Ancient Levantine Arabic’). Note that my model here differs in crucial ways from that of Knauf (‘AraboAramaic and ʿArabiyya’, esp. p. 204 with Fig. 4). Knauf envisions a necessary ‘bottleneck’ of standardization, whereby dialect diversity before standardization is lost in the standardization process and then new dialect diversity develops after standardization. This bottleneck is, however, theoretically unnecessary and ultimately unlikely: why must standardization lead to the loss of all prior dialect diversity? More to the point with Arabic, Knauf ’s model is demonstrably false: modern Arabic dialects preserve features of Proto-Semitic that are not found in Classical Arabic. Such ProtoSemitic retentions prove that the modern Arabic dialects do not derive from standardized Classical Arabic, as Knauf would seem to have it, but rather go back directly to Proto-Arabic, which in turn ultimately derives from Proto-Semitic. 118 The biblio­graphy on Middle Arabic is immense; for an introduction, see Blau, A Handbook of Early Middle Arabic as well as his collected studies in Studies in Middle Arabic and its Judaeo-Arabic Variety.

387

3 88

a a ro n m icha e l b u t t s

Arabic as their Muslim contemporaries were.119 Nevertheless, Middle Arabic is also attested among Muslims, especially when writing in less formal genres, such as on papyri.120 It is through this lens of the history of Arabic that we need to consider Ḥunayn. Even if we are to assume that Ḥunayn’s native language was Arabic, as I think we are, his Arabic would not have been Classical Arabic but a different form of the language. The Classical Arabic that Ḥunayn wrote he would have learned not strictly as a native language. In addition, it should be noted that Ḥunayn also wrote in Middle Arabic, including the very Letter discussed here. Given all this, there is less difference between Ḥunayn’s writing in Syriac and his writing in Arabic than one might first imagine: in neither case is Ḥunayn writing sensu stricto in his native language; rather, he is writing in a learned language in both cases. Thus, the example of Ḥunayn adds further texture to the complicated intersections of Syriac and Arabic — in their multiple varieties! — in the Islamic period.

Conclusion Much remains unclear about language use among Syriac Christians after the rise of Islam — or before the rise of Islam for that matter. This is mostly due to the nature of the surviving evidence: in contrast to the situation in Egypt, where the papyro­logical record provides precious data for tracing the development of Greek, Coptic, and later Arabic, there is very limited documentary evidence for language use in Syria and Mesopotamia, the homeland of Syriac Christianity.121 Almost all the texts that we have are literary, written by male elites and subsequently transmitted in manu­scripts, sometimes over generations, during which time the language of the texts could be further manipulated.122 Working within these limitations, I have

119 For Christian Middle Arabic, see Blau, A Grammar of Christian Arabic; for Jewish Middle Arabic, see Blau, A Grammar of Mediaeval Judaeo-Arabic and Emergence and Linguistic Background of Judaeo-Arabic. 120 See Hopkins, Studies in the Grammar of Early Arabic. 121 For the situation in Egypt, see the insightful analysis in Papaconstantinou, ‘Why Did Coptic Fail Where Aramaic Succeeded?’. Geo­graphically between Egypt, on the one hand, and Syria and Mesopotamia, on the other, we have, of course, the Nessana papyri (Excavations at Nessana, i, ed. by Colt; ii, ed. by Casson and Hettich; iii, ed. by Kraemer; note also the ongoing project spearheaded by the Institute for the Study of the Ancient World [ISAW] to digitize the documents and make them widely available online) as well as the earlier Petra papyri (Frösén, Arjava, and Lehtinen, with others, The Petra Papyri). The documentary evidence from Syria and Mesopotamia is, however, unfortunately much more limited. For a checklist of papyro­logical texts, now somewhat dated, see Cotton, Cockle, and Millar, ‘The Papyro­logy of the Roman Near East’. An overview is available in Gascou, ‘The Papyro­logy of the Near East’. The best evidence for Syria and Mesopotamia comes from the more than 150 documents, primarily in Greek or Latin with a few also in Iranian and Aramaic (including but not limited to Syriac), discovered at Dura-Europos, an important military outpost on the Euphrates until its destruction in 256 ce. The texts are edited in Welles, Fink, and Gilliam, The Excavations at Dura-Europus. For language use at Dura-Europos, see Gascou, ‘The Diversity of Languages in Dura-Europos’, pp. 74–96. 122 The following comment by Papaconstantinou (‘Why Did Coptic Fail Where Aramaic Succeeded?’, p. 66) is worth quoting in full: ‘The above discussions show how difficult it is to approach language

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

aimed in this paper to provide additional texture to our understanding of the twin histories of Arabic and Aramaic by narrowing the focus specifically to the Syriac dialect of Aramaic. In particular, I have interrogated the use of Arabic by (Syriac) Christians before the rise of Islam (sections 1 and 3) and the use of Syriac by Syriac Christians after the rise of Islam (sections 2 and 4). In addition, I have advocated for a more nuanced approach to language use that is informed by the history of the languages as well as by (socio-)linguistics. Not enough attention has been paid, in my view, to the fact that Syriac as we know it is a highly standardized literary language that may be quite distinct from the everyday spoken language of most Syriac-speakers in Late Antiquity, not to mention in the Islamic period.123 Thus, even when we have Syriac writings from the Islamic period by authors who likely spoke some form of Aramaic as their first language, there was probably a significant gap between their spoken Aramaic and their written Aramaic (i.e., Syriac). In other words, we have a diglossic situation. The same is true for Arabic, which throughout its long history has often been diglossic.124 This is especially evident for Christians, since we have examples of their Middle Arabic writings. Consequently, when discussing Aramaic and Arabic in the Islamic period, we are never talking about only two languages; rather, we are dealing with multiple varieties of each.125 It is only by taking this complexity into consideration that we can arrive at a more textured account of the twin histories of Arabic and Aramaic. * * *

shift when one needs to rely on texts written mostly, if not exclusively, in institutional contexts. These can be analysed linguistically for evidence of language contact, they can be read directly for straightforward pieces of information, and indirectly, in the hope of assessing the authorial or institutional intentions underlying them; but they cannot inform us about the majority of the population and they cannot capture orality — not even for the high-status individuals who produced them’. 123 Again, I quote Papaconstantinou: ‘How biased an image of the linguistic map of seventeenth-century Europe we would have today if the overwhelming majority of surviving sources were the Latin texts produced by the members of the Republic of Letters!’ (‘Why Did Coptic Fail Where Aramaic Succeeded?’, p. 65). For Syriac as a standard(ized) language, see the influential study of Van Rompay, ‘Some Preliminary Remarks on the Origins of Classical Syriac as a Standard Language’, as well as the more recent remarks in Taylor, ‘Bilingualism and Diglossia in Late Antique Syria and Mesopotamia’, p. 325. For some of my (preliminary) thoughts on written versus spoken varieties of Syriac in Late Antiquity, see Butts, ‘The Classical Syriac Language’, pp. 225–31. 124 Again, to be clear, not in the sense of Owens’s A Linguistic History of Arabic. 125 To return to the linguistic literature one last time, this might be called ‘double-nested diglossia’, a term introduced to refer to the situation in the Indian village of Khalapur where Hindi and the local dialect functioned as high and low languages, respectively, but each itself also consisted of high and low varieties (for the term, see Fasold, The Sociolinguistics of Society, pp. 46–48; for the language situation, see Gumperz, ‘Linguistic and Social Interaction in Two Communities’, pp. 137–53). Mutatis mutandis, with some Syriac Christians in the Islamic period, it is conceivable that Arabic functioned as a high language and Aramaic as a low language, and that in addition both Arabic and Aramaic also consisted of high and low varieties.

389

390

a a ro n m icha e l b u t t s

This publication was supported by the Andrew W. Mellon Foundation Fellowship for Assistant Professors at the Institute for Advanced Study (Princeton, NJ). An earlier version was first presented at the symposium ‘Navigating Language in the Early Islamic World’, hosted by the MARCO Institute for Medi­eval and Renaissance Studies at the University of Tennessee, Knoxville (6–7 April 2018), and subsequently in the Near Eastern Seminar at the Institute for Advanced Study (2 October 2019). I am grateful to both audiences for the insightful discussions that followed the presentations of this paper. I would also like to thank the following people for helping in various ways: Kevin van Bladel, Antoine Borrut, Manuela Ceballos, Simcha Gross, Dimitri Gutas, Amir Harrak, Ahmad Al-Jallad, George Kiraz, Marielle Pic, Khodadad Rezakhani, Christian Robin, Sabine Schmidtke, Alison Vacca, Lucas Van Rompay, Lev Weitz, and Adam Zeidan.

Biblio­graphy Primary Sources Abū Qurra, Theodore, Mayāmir Thāwudūrus Abī Qurra usquf Ḥarrān, ed. by C. Bacha (n.p., 1904) —— , Theodore Abū Qurrah, trans. by J. C. Lamoreaux, Library of the Christian East, 1 (Provo, UT: Brigham Young University Press, 2005) Antony of Tagrit, The Fifth Book of the Rhetoric of Antony of Tagrit, ed. by J. W. Watt, Corpus Scriptorum Christianorum Orientalium, 480–81 (Leuven: Peeters, 1986) —— , ‘Antony of Tagrit’s Rhetoric Book One’, ed. by P. E. Eskenasy (unpublished doctoral dissertation, Harvard University, 1991) —— , The Book of the Rhetoric of Antony Rhitor of Tagrit, ed. by E. Sewan d-Bet Qermez (Stockholm: Författeres Bokmaskin, 2000) Bar Hebraeus, Taʾrīkh mukhtaṣar al-duwal, ed. by A. Ṣālḥānī (Beirut: Catholic Press, 1890) Calvet, Y., and C. Robin, Arabie heureuse, Arabie déserte: les antiquités arabiques du Musée du Louvre (Paris: Editions de la Réunion des musées nationaux, 1997), pp. 265–69 [Namāra inscription] Chronicle of Zuqnin = Incerti auctoris Chronicon Pseudo-Dionysianum vulgo dictum, ed. by J.-B. Chabot, Corpus Scriptorum Christianorum Orientalium, 91 (Paris: Typo­grapheo Reipublicae, 1927) The Chronicle of Zuqnīn: Parts I and II, From the Creation to the Year 506/7 ad, trans. by A. Harrak (Piscataway, NJ: Gorgias Press, 2017) Excavations at Nessana, i, ed. by H. D. Colt (Princeton, NJ: Princeton University Press, 1950); ii: Literary Papyri, ed. by L. Casson and E. L. Hettich (Princeton, NJ: Princeton University Press, 1950); iii: Non-literary Papyri, ed. by C. J.Kraemer, Jr. (Princeton, NJ: Princeton University Press, 1958) [Nessana papyri] Frösén, J., A. Arjava, and M. Lehtinen, with contributions by others, The Petra Papyri (Amman: American Center of Oriental Research, 2002–) [Petra papyri]

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

Grayson, A. K., Assyrian Rulers of the Early First Millennium bc II (858–745 bc), The Royal Inscriptions of Mesopotamia: Assyrian Periods, 3 (Toronto: University of Toronto Press, 1996) [Shalmaneser III inscription] The History of the ‘Slave of Christ’: From Jewish Child to Christian Martyr, ed. by A. M. Butts and S. Gross, Persian Martyr Acts in Syriac: Text and Translation, 6 (Piscataway, NJ: Gorgias Press, 2016) Ḥunayn b. Isḥāq, Ḥunain Ibn Isḥāq über die syrischen und arabischen Galen-Übersetzungen, ed. by G. Bergsträsser (Leipzig: F. A. Brockhaus, 1925) —— , Neue Materialien zu Ḥunain ibn Isḥāq’s Galen-Biblio­graphie, ed. by G. Bergsträsser (Leipzig: Deutsche Morgenländische Gesellschaft, 1932) —— , al-Masāʾil fī al-ṭibb li-l-mutaʿallimīn li-Ḥunayn b. Isḥāq, ed. by M. ʿA. Abū Rayyān, M. M. ʿArab, and J. M. Mūsā (Cairo: Dār al-Jāmiʿāt al-Miṣriyya, 1978) —— , Questions on Medicine for Scholars by Ḥunayn ibn Isḥāq, trans. by P. P. Ghalioungui (Cairo: al-Ahram Center for Scientific Translations, 1980) —— , Ḥunayn ibn Isḥāq on his Galen Translations, ed. by J. C. Lamoreaux (Provo, UT: Brigham Young University Press, 2016) Ibn Abī Uṣaybiʿa, Ibn Abi Useibia [ʿUyūn al-anbāʾ fī ṭabaqāt al-aṭibbāʾ], ed. by A. Müller (Königsberg: self-published, 1884) Ibn al-Nadīm, Kitâb al-Fihrist, ed. by G. Flügel, A. Müller, and J. Roediger (Leipzig: F. C. W. Vogel, 1871–1872) Ibn al-Qifṭī, Ibn al-Qifṭī’s Taʾrīḫ al-ḥukamāʾ, ed. by J. Lippert (Leipzig: Dieterich’sche Verlagsbuchhandlung, 1903) Ibn al-Ṭayyib, Ibn at-Taiyib: Fiqh an-Nasrânîya, ‘Das Recht der Christenheit’, ed. by W. Hoenerbach and O. Spies, Corpus Scriptorum Christianorum Orientalium, 161–62, 167–68 (Leuven: L. Durbecq, 1956–1957) —— , Commentaire sur la Genèse, ed. by J. C. J. Sanders, Corpus Scriptorum Christianorum Orientalium, 274–75 (Leuven: Peeters, 1967) Jacob of Serugh, Iacobi Sarugensis Epistulae quotquot supersunt, ed. by G. Olinder, Corpus Scriptorum Christianorum Orientalium, 110 (Leuven: Peeters, 1937) Julius Africanus, Iulius Africanus: Cesti; The Extant Fragments, ed. by M. Wallraff, C. Scardino, L. Mecella, C. Guignard, and W. Adler, Die griechischen christlichen Schriftsteller der ersten Jahrhunderte, n.s., 18 (Berlin: De Gruyter, 2012) al-Masʿūdī, Les prairies d’or, ed. and trans. by C. Barbier de Meynard and A. Pavet de Courteille (Paris: Imprimerie Impériale, 1861–1877) —— , Murūj al-dhahab wa-maʿādin al-jawhar, rev. ed. by C. Pellat (Beirut: al-Jāmiʿa alLubnāniyya, 1965–1979) Nehmé, L., ‘New Dated Inscriptions (Nabataean and Pre-Islamic Arabic) from a Site near al-Jawf, Ancient Dūmah, Saudi Arabia’, Arabian Epi­graphic Notes, 3 (2017), 121–64 [Dūmat al-Jandal inscription] Rabbula, S. Ephraemi Syri Rabulae episcopi Edesseni Balaei aliorumque Opera selecta, ed. by J. J. Overbeck (Oxford: Clarendon Press, 1865) —— , The Rabbula Corpus, ed. by R. R. Phenix, Jr., and C. B. Horn (Atlanta: SBL Press, 2017) Robin, C., ʿA. I. al-Ghabbān, and S. F. al-Saʿīd, ‘Inscriptions antiques de la région de Najrān (arabie séoudite méridionale): nouveaux jalons pour l’histoire de l’écriture, de la langue et du calendrier arabes’, Comptes rendus des séances de l’Académie des Inscriptions et Belles-Lettres, 158 (2014), 1033–28 [Ḥimā inscriptions]

391

392

a a ro n m icha e l b u t t s

al-Shdaifat, Y., A. Al-Jallad, Z. al-Salameen, and R. Harahsheh, ‘An Early Christian Arabic Graffito Mentioning “Yazīd the King”’, Arabian Archaeo­logy and Epi­graphy, 28 (2017), 315–24 [Qaṣr Burquʿ inscription] Simeon of Beth Arsham, ‘La lettera di Simeone vescovo di Bêth-Arśâm sopra i martiri omeriti’, ed. by I. Guidi, Atti della Reale Accademia Nazionale dei Lincei, Serie Terza: Memorie della Classe di Scienze morali, storiche e filo­logiche, 7 (1881), 471–515 Synodicon Orientale = Das Buch der Synhados oder Synodicon Orientale, trans. by O. Braun (Stuttgart: Rothsche Verlagshandlung, 1900) Synodicon Orientale ou recueil de synodes nestoriens, ed. by J. B. Chabot (Paris: Imprimerie Nationale, 1902) al-Ṭabarī, Annales quos scripsit Abu Djafar Mohammed ibn Djarir at-Tabari, ed. by M. J. de Goeje (Leiden: Brill, 1879–1901) —— , The History of al-Ṭabarī [Taʾrīkh al-rusul wa-l-mulūk], xi: The Challenge to the Empires, trans. by K. Y. Blankinship (Albany: State University of New York Press, 1993) Timothy, Timotheos I, Ostsyrischer Patriarch, Disputation mit dem Kalifen al-Mahdi, ed. by M. Heimgartner, Corpus Scriptorum Christianorum Orientalium, 631–32 (Leuven: Peeters, 2011) —— , Die Briefe 42–58 des ostsyrischen Patriarchen Timotheos I, ed. by M. Heimgartner, Corpus Scriptorum Christianorum Orientalium, 644–45 (Leuven: Peeters, 2012) —— , Die Briefe 30–39 des ostsyrischen Patriarchen Timotheos I, ed. by M. Heimgartner, Corpus Scriptorum Christianorum Orientalium, 661–62 (Leuven: Peeters, 2016) —— , Die Briefe 40 und 41 des ostsyrischen Patriarchen Timotheos I, ed. by M. Heimgartner, Corpus Scriptorum Christianorum Orientalium, 673–74 (Leuven: Peeters, 2019) Welles, C. B., R. O. Fink, and J. F. Gilliam, The Excavations at Dura-Europus: Final Report V, Part 1. The Parchments and Papyri (New Haven, CT: Yale University Press, 1959) [DuraEuropos papyri] Secondary Works Al-Jallad, A., ‘Ancient Levantine Arabic: A Reconstruction Based on the Earliest Sources and the Modern Dialects’ (unpublished doctoral dissertation, Harvard University, 2012) —— , ‘On the Genetic Background of the Rbbl bn Hfʿm Grave Inscription at Qaryat alFāw’, Bulletin of the School of Oriental and African Studies, 77 (2014), 445–65 —— , An Outline of the Grammar of Safaitic Inscriptions (Leiden: Brill, 2015) —— , ‘What Is Ancient North Arabian?’, in Re-engaging Comparative Semitic and Arabic Studies, ed. by D. Birnstiel and N. Pat-El (Wiesbaden: Harrassowitz, 2018), pp. 1–45 —— , ‘The Earliest Stages of Arabic and its Linguistic Classification’, in Routledge Handbook of Arabic Linguistics, ed. by E. Benmaoun and R. Bassiouney (London: Routledge, 2018), pp. 315–31 —— , ‘The Particle ʿkdy in the Namārah Inscription and a New Interpretation of Line 4’, Journal of Semitic Studies, 65 (2021), 363–78 —— , ‘One Wāw to Rule Them All: The Origins and Fate of Wawation in Arabic’, in Scripts and Scriptures, ed. by F. Donner and R. Hasselbach, Late Antique and Medi­eval Islamic Near East 3 (Chicago: Oriental Institute, 2022), pp. 87–103

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

Anawati, G. C., and A. Z. Iskandar, ‘Ḥunayn ibn Isḥāq al-ʿIbādī, Abū Zayd’, in Dictionary of Scientific Bio­graphy, xv: Supplement I, ed. by C. C. Gillispie (New York: Scribner, 1980), pp. 230–49 Arnzen, R., ‘Proclus on Plato’s Timaeus 89e3–90c7’, Arabic Sciences and Philosophy, 23 (2013), 1–45 Baumstark, A., ‘Das Problem eines vor-islamischen christlich-kirchlichen Schrifttums in arabischer Sprache’, Islamica, 4 (1929–1931), 562–75 Beaucamp, J., F. Briquel-Chatonnet, and C. J. Robin, ed., Juifs et chrétiens en Arabie aux ve et vie siècles (Paris: Association des Amis du Centre d’Histoire et Civilisation de Byzance, 2010) Beaumont, M., ‘ʿAmmār al-Baṣrī’, in Christian-Muslim Relations: A Biblio­graphic History, i: 600–900, ed. by D. Thomas and B. Roggema, with J. P. Monferrer Sala, J. Pahlitzsch, M. Swanson, H. Teule, and J. Tolan, History of Christian-Muslim Relations, 11 (Leiden: Brill, 2009), pp. 604–10 Beeston, A. F. L., ‘Nemara and Faw’, Bulletin of the School of Oriental and African Studies, 42 (1979), 1–6 Bellamy, J. A., ‘A New Reading of the Namārah Inscription’, Journal of the American Oriental Society, 105 (1985), 31–48 Ben-Shammai, Haggai, ‘Observations on the Beginnings of Judeo-Arabic Civilization’, in Beyond Religious Borders: Interaction and Intellectual Exchange in the Medi­eval Islamic World, ed. by D. M. Freidenreich and M. Goldstein (Philadelphia: University of Pennsylvania Press, 2012), pp. 13–29 Bergsträsser, G., Ḥunain Ibn Ishaḳ und seine Schule (Leiden: Brill, 1913) Berti, V., Vita e studi di Timoteo I, Patriarca cristiano di Baghdad: Ricerche sull’epistolario e sulle fonti contigue, Cahiers de Studia Iranica, 41; Chrétiens en terre d’Iran, 3 (Paris: Association pour l’Avancement des Études Iraniennes, 2009) Blau, J., A Grammar of Mediaeval Judaeo-Arabic [in Hebrew] ( Jerusalem: Magnes Press, 1961) —— , Emergence and Linguistic Background of Judaeo-Arabic (Oxford: Oxford University Press, 1965) —— , A Grammar of Christian Arabic, Corpus Scriptorum Christianorum Orientalium, 267, 276, 279 (Leuven: Peeters, 1966) —— , Studies in Middle Arabic and its Judaeo-Arabic Variety ( Jerusalem: Magnes Press, 1988) —— , A Handbook of Early Middle Arabic ( Jerusalem: Max Schloessinger Memorial Foundation, 2002) Blau, J., and S. Hopkins, Early Judaeo-Arabic in Phonetic Spelling: Texts from the End of the First Millennium [in Hebrew] ( Jerusalem: Ben-Zvi, 2017) Borrut, A., Entre mémoire et pouvoir: l’espace syrien sous les derniers Omeyyades et les premiers Abbassides (v. 72–193/692–8090) (Leiden: Brill, 2011) Bowersock, G. W., The Throne of Adulis (Oxford: Oxford University Press, 2013) Briquel-Chatonnet, F., and M. Debié, ed., Le monde syriaque: sur les routes d’un christianisme ignoré (Paris: Les Belles Lettres, 2017) Brock, S. P., ‘Some Observations on the Use of Classical Syriac in the Late Twentieth Century’, Journal of Semitic Studies, 34 (1989), 363–75 —— , ‘The Syriac Background to Ḥunayn’s Translation Techniques’, ARAM, 3 (1991), 139–62 —— , ‘The “Nestorian” Church: A Lamentable Misnomer’, Bulletin of the John Rylands University Library of Manchester, 78 (1996), pp. 23–35

393

394

a a ro n m icha e l b u t t s

—— , ‘Changing Fashions in Syriac Translation Technique: The Background to Syriac Translations under the Abbasids’, Journal of the Canadian Society for Syriac Studies, 4 (2004), 3–14 —— , ‘Bardaiṣan’, in GEDSH, pp. 56–57 Brockelmann, C., Geschichte der arabischen Literatur, suppl. i (Leiden: Brill, 1937) —— , Geschichte der arabischen Literatur, 2nd edn, i (Leiden: Brill, 1943) Butts, A. M., ‘The Afflictions of Exile: A Syriac Memrā by David Puniqāyā’, Le Muséon, 122 (2009), 53–80 —— , ‘Ḥunayn b. Isḥāq’, in GEDSH, pp. 205–06 —— , ‘Ibn al-Ṭayyib’, in GEDSH, pp. 206–07 —— , ‘Theodoros Abū Qurrah’, in GEDSH, pp. 403–05 —— , ‘Embellished with Gold: The Ethiopic Reception of Syriac Biblical Exegesis’, Oriens Christianus, 97 (2013/2014), 137–59 —— , ‘In Search of Sources for Ibn al-Ṭayyib’s The Paradise of Christianity: Theodore Bar Koni’s Scholion’, Journal of the Canadian Society for Syriac Studies, 14 (2014), 3–29 —— , Language Change in the Wake of Empire: Syriac in its Greco-Roman Context, Linguistic Studies in Ancient West Semitic, 11 (Winona Lake: Eisenbrauns, 2016) —— , ‘North Arabian Features in the Nabataean Aramaic Inscriptions from Madāʾin Ṣāliḥ: A Contact-Linguistic Analysis’, in Near Eastern and Arabian Essays: Studies in Honour of John F. Healey, ed. by R. Smith, Journal of Semitic Studies Supplement Series (Oxford: Oxford University Press, 2018), pp. 45–65 —— , ‘The Classical Syriac Language’, in The Syriac World, ed. by D. King (London: Routledge, 2019), pp. 222–42 —— , ‘Old Syriac’, in Brill Encyclopedia of Early Christianity, ed. by D. G. Hunter, P. J. J. van Geest, and B. J. L. Peerbolte (Leiden: Brill, forthcoming) Carlson, T. A., ‘A Light from “The Dark Centuries”: Isḥaq Shbadnaya’s Life and Works’, Hugoye, 14 (2011), 191–213 —— , Christianity in Fifteenth-Century Iraq (Cambridge: Cambridge University Press, 2018) Childers, J., ‘ʿAbdishoʿ bar Brikha’, in GEDSH, pp. 3–4 Cooper, G. M., ‘Ḥunayn ibn Isḥāq’s Galen Translations and Greco-Arabic Philo­logy: Some Observations from the Crises (De crisibus) and the Critical Days (De diebus decretoriis)’, Oriens, 44 (2016), 1–43 Cotton, H. M., W. E. H. Cockle, and F. G B. Millar, ‘The Papyro­logy of the Roman Near East: A Survey’, Journal of Roman Studies, 85 (1995), 214–35 Cowley, R. W., Ethiopian Biblical Interpretation: A Study in Exegetical Tradition and Hermeneutics, University of Cambridge Oriental Publications, 38 (Cambridge: Cambridge University Press, 1988) Davis, S. J., Coptic Christo­logy in Practice: Incarnation and Divine Participation in Late Antique and Medi­eval Egypt (Oxford: Oxford University Press, 2008) Degen, R., ‘The Oldest Known Syriac Manu­script of Ḥunayn b. Isḥāq’, in Symposium Syriacum 1976, Orientalia Christiana Analecta, 205 (Rome: Pontificium Institutum Orientalium Studiorum, 1978), pp. 63–71 Diem, W., ‘Untersuchungen zur frühen Geschichte der arabischen Ortho­graphie: I. Die Schreibung der Vokale’, Orientalia, 48 (1979), 207–57 —— , ‘Untersuchungen zur frühen Geschichte der arabischen Ortho­graphie: II. Die Schreibung der Konsonanten’, Orientalia, 49 (1980), 67–106

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

—— , ‘Untersuchungen zur frühen Geschichte der arabischen Ortho­graphie: III. Endungen und Endschreibungen’, Orientalia, 50 (1981), 332–83 —— , ‘Untersuchungen zur frühen Geschichte der arabischen Ortho­graphie: IV. Die Schreibung der zusammenhängenden Rede; Zusammenfassung’, Orientalia, 52 (1983), 357–404 Doran, R., Stewards of the Poor: The Man of God, Rabbula, and Hiba in Fifth-Century Edessa, Cistercian Studies Series, 208 (Kalamazoo, MI: Cistercian Publications, 2006) Drory, R., Models and Contacts: Arabic Literature and its Impact on Medi­eval Jewish Culture (Leiden: Brill, 2000) Ephʿal, I., The Ancient Arabs: Nomads on the Borders of the Fertile Crescent 9th–5th Centuries b.c. ( Jerusalem: Magnes Press; Leiden: Brill, 1982) Fasold, R., The Sociolinguistics of Society (Oxford: Basil Blackwell, 1984) Faultless, J., ‘Ibn al-Ṭayyib’, in Christian-Muslim Relations: A Biblio­graphic History, ii: 900–1050, ed. by D. Thomas and A. Mallett, with J. P. Monferrer Sala, J. Pahlitzsch, M. Swanson, H. Teule, and J. Tolan, History of Christian-Muslim Relations, 14 (Leiden: Brill, 2010), pp. 667–97 Féghali, P., ‘Ibn aṭ-Ṭayyib et son commentaire sur la Genèse’, Parole de l’Orient, 16 (1990–1991), 149–62 Ferguson, C. A., ‘Diglossia’, Word, 15 (1959), 325–40 Fisher, G., ‘Kingdoms or Dynasties? Arabs, History, and Identity before Islam’, Journal of Late Antiquity, 4 (2011), 245–67 —— , ed., Arabs and Empires before Islam (Oxford: Oxford University Press, 2015) Fisher, G., and P. Wood, ‘Writing the History of the “Naṣrid” Dynasty at al-Ḥīra: The PreIslamic Perspective’, Iranian Studies, 49 (2016), 247–90 Fishman, J., ‘Bilingualism with and without Diglossia; Diglossia with and without Bilingualism’, Journal of Social Issues, 23 (1967), 29–38 Forness, P. M., Preaching Christo­logy in the Roman Near East (Oxford: Oxford University Press, 2018) Garitte, G. ‘La passion géorgienne de saint ʿAbd al-Masīh’, Le Muséon, 79 (1966), 187–237 Gascou, J., ‘The Papyro­logy of the Near East’, in The Oxford Handbook of Papyro­logy, ed. by R. S. Bagnall (Oxford: Oxford University Press, 2011), pp. 473–94 —— , ‘The Diversity of Languages in Dura-Europos’, in Edges of Empires: Pagans, Jews, and Christians at Roman Dura-Europos, ed. by J. Y. Chi and S. Heath (Princeton, NJ: Princeton University Press, 2012), pp. 74–96 GEDSH = Gorgias Encyclopedic Dictionary of the Syriac Heritage, ed. by S. P. Brock, A. M. Butts, G. A. Kiraz, and L. Van Rompay (Piscataway, NJ: Gorgias Press, 2011) Genequand, D., ‘The Archaeo­logical Evidence for the Jafnids and the Naṣrids’, in Arabs and Empires before Islam, ed. by G. Fisher (Oxford: Oxford University Press, 2015), pp. 172–213 Graf, G., Geschichte der christlichen arabischen Literatur, Studi e Testi, 118, 133, 146, 147, 172 (Vatican: Biblioteca Apostolica Vaticana, 1944–1952) Griffith, S. H., ‘Habīb ibn Ḫidmah Abū Rāʾiṭah, A Christian Mutakallim of the First Abbasid Century’, Oriens Christianus, 64 (1980), 161–201 —— , ‘The Gospel in Arabic: An Inquiry into its Appearance in the First Abbasid Century’, Oriens Christianus, 69 (1985), 126–67

395

396

a a ro n m icha e l b u t t s

—— , ‘Reflections on the Bio­graphy of Theodore Abū Qurrah’, Parole de l’Orient, 18 (1993), 143–70 —— , Theodore Abū Qurrah: The Intellectual Profile of an Arab Christian Writer of the First Abbasid Century (Tel Aviv: Tel Aviv University, 1995) —— , ‘From Aramaic to Arabic: The Languages of the Monasteries of Palestine in the Byzantine and Early Islamic Periods’, Dumbarton Oaks Papers, 51 (1997), 11–31 —— , The Bible in Arabic: The Scriptures of the People of the Book in the Language of Islam (Princeton, NJ: Princeton University Press, 2013) Gross, S., ‘A Long Overdue Farewell: The Purported Jewish Origins of Syriac Christianity’, in Jews and Syriac Christians: Intersections across the First Millennium, ed. by A. M. Butts and S. Gross, Texts and Studies in Ancient Judaism, 180 (Tübingen: Mohr Siebeck, 2020), pp. 121–44 Gumperz, J. J., ‘Linguistic and Social Interaction in Two Communities’, American Anthropo­logist, 66 (1964), 137–53 Gutas, D., Greek Thought, Arabic Culture: The Graeco-Arabic Translation Movement in Baghdad and Early ʿAbbāsid Society (2nd–4th/8th–10th Centuries) (London: Routledge, 1998) —— , ‘Scholars as Transmitters of Philosophical Thought’, in Philosophy in the Islamic World, i: 8th–10th Centuries, ed. by U. Rudolph, R. Hansberger, and P. Adamson, trans. by R. Hansberger (Leiden: Brill, 2017), pp. 680–704, 766–68 —— , ‘A New “Edition” of Ḥunayn’s Risāla’, Arabic Sciences and Philosophy, 28 (2018), 279–84 Gzella, H., Cultural History of Aramaic: From the Beginnings to the Advent of Islam (Leiden: Brill, 2015) Ḥabbi, Y., Ḥunayn b. Isḥāq (Baghdad: Majmaʿ al-Lugha al-Suryāniyya, 1974) Hämeen-Anttila, J., ‘The Camels of Tiglath-Pileser III and the Arabic Definite Article’, in Of God(s), Trees, Kings, and Scholars: Neo-Assyrian and Related Studies in Honour of Simo Parpola, ed. by M. Luukko, S. Svärd, and R. Mattila, Studia Orientalia, 106 (Helsinki: Finish Oriental Society, 2009), pp. 99–101 Hämeen-Anttila, J., and R. Rollinger, ‘Herodot und die arabische Göttin “Alilat”’, Journal of Ancient Near Eastern Religions, 1 (2002), 84–99 Hatke, G., ‘Africans in Arabia Felix: Aksumite Relations with Himyar in the Sixth Century c.e.’ (unpublished doctoral dissertation, Princeton University, 2011) Hayek, M., ‘ʿAmmār al-Baṣrī: la première Somme de théo­logie chrétienne en langue arabe, ou deux apo­logies du christianisme’, Islamochristiana, 2 (1976), 69–133 —— , ʿAmmār al-Baṣrī: apo­logie et controverses (Beirut: Dar el-Machreq, 1977) Heimgartner, M., ‘Letter 59 …’, in Christian-Muslim Relations: A Biblio­graphic History, i: 600–900, ed. by D. Thomas and B. Roggema, with J. P. Monferrer Sala, J. Pahlitzsch, M. Swanson, H. Teule, and J. Tolan, History of Christian-Muslim Relations, 11 (Leiden: Brill, 2009), pp. 522–26 Hopkins, S., Studies in the Grammar of Early Arabic (Oxford: Oxford University Press, 1984) Hoyland, R., ‘Language and Identity: The Twin Histories of Arabic and Aramaic (and: Why Did Aramaic Succeed Where Greek Failed)’, Scripta Classica Israelica, 23 (2004), 183–99 —— , ‘The Birth of Arabic in Stone’, in By the Pen and What They Write: Writing in Islamic Art and Culture; 6th Biennial Hamad Bin Khalifa Symposium on Islamic Art, November 2015, ed. by S. Blair and J. Bloom (New Haven, CT: Yale University Press, 2017), pp. 49–67

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

Huehnergard, J., ‘Arabic in its Semitic Context’, in Arabic in Context: Celebrating 400 Years of Arabic at Leiden University, ed. by A. Al-Jallad, Studies in Semitic Languages and Linguistics, 89 (Leiden: Brill, 2017), pp. 3–34 Hunter, E. C. D., ‘The Christian Matrix of al-Hira’, in Les controverses des chrétiens dans l’Iran sassanide, ed. by C. Jullien (Paris: Association pour l’Avancement des Études Iraniennes, 2008), pp. 41–56 Jaspers, J., ‘Diglossia and Beyond’, in The Oxford Handbook of Language and Society, ed. by O. García, N. Flores, and M. Spotti (Oxford: Oxford University Press, 2016), pp. 179–96 Jeffery, A., ‘Christianity in South Arabia’, Muslim World, 36 (1946), 204–06 Käs, F., ‘Eine neue Handschrift von Ḥunain ibn Isḥāq’s Galenbiblio­graphie’, Zeitschrift für Geschichte der arabisch-islamischen Wissenschaften, 19 (2010–2011), 135–93 Kaufhold, H., Die Rechtssammlung des Gabriel von Basra und ihr Verhältnis zu den anderen juristischen Sammelwerken der Nestorianer (Berlin: J. Schweitzer, 1976) —— , ‘Sources of Canon Law in the Eastern Churches’, in The History of Byzantine and Eastern Canon Law to 1500, ed. by W. Hartmann and K. Pennington, History of Medieval Canon Law (Washington, DC: Catholic University of America Press, 2012), pp. 215–342 Keating, S. T., Defending the ‘People of Truth’ in the Early Islamic Period: The Christian Apo­ logies of Abū Rāʾiṭah, History of Christian-Muslim Relations, 4 (Leiden: Brill, 2006) —— , ‘Abū Rāʾiṭa l-Takrītī’, in Christian-Muslim Relations: A Biblio­graphic History, i: 600–900, ed. by D. Thomas and B. Roggema, with J. P. Monferrer Sala, J. Pahlitzsch, M. Swanson, H. Teule, and J. Tolan, History of Christian-Muslim Relations, 11 (Leiden: Brill, 2009), pp. 567–81 Khan, G., ‘The Neo-Aramaic Dialects of Eastern Anatolia and Northwestern Iran’, in The Languages and Linguistics of Western Asia: An Areal Perspective, ed. by G. Haig and G. Khan, The World of Linguistics, 6 (Berlin: De Gruyter, 2018), pp. 190–236 —— , ‘The Neo-Aramaic Dialects of Northern Iraq’, in The Languages and Linguistics of Western Asia: An Areal Perspective, ed. by G. Haig and G. Khan, The World of Linguistics, 6 (Berlin: De Gruyter, 2018), pp. 305–53 —— , ‘The Neo-Aramaic Dialects of Western Iran’, in The Languages and Linguistics of Western Asia: An Areal Perspective, ed. by G. Haig and G. Khan, The World of Linguistics, 6 (Berlin: De Gruyter, 2018), pp. 481–582 —— , ‘The Neo-Aramaic Dialects and their Historical Background’, in The Syriac World, ed. by D. King (London: Routledge, 2019), pp. 266–89 King, D., The Syriac Versions of the Writings of Cyril of Alexandria: A Study in Translation Tech­nique, Corpus Scriptorum Christianorum Orientalium, 626 (Leuven: Peeters, 2008) —— , ed., The Syriac World (London: Routledge, 2019) Kiraz, G. A., ‘Kthobonoyo Syriac: Some Observations and Remarks’, Hugoye, 10 (2007), 129–42 Kister, M. J., ‘Al-Ḥīra: Some Notes on its Relation with Arabia’, Arabica, 11 (1968), 143–69 Knauf, E. A., ‘Arabo-Aramaic and ʿArabiyya: From Ancient Arabic to Early Standard Arabic, 200 ce–600 ce’, in The Qurʾān in Context: Historical and Literary Investigations into the Qurʾānic Milieu, ed. by A. Neuwirth, N. Sinai, and M. Marx (Leiden: Brill, 2010), pp. 197–254

397

398

a a ro n m icha e l b u t t s

Knudsen, E. K., and E. Wardini, Neo­logisms in Modern Literary Syriac, Perspectives on Linguistics and Ancient Languages, 10 (Piscataway, NJ: Gorgias Press, 2018) Lamoreaux, J. C., ‘The Bio­graphy of Theodore Abū Qurrah Revisited’, Dumbarton Oaks Papers, 56 (2002), 25–40 —— , ‘Theodore Abū Qurra’, in Christian-Muslim Relations: A Biblio­graphic History, i: 600–900, ed. by D. Thomas and B. Roggema, with J. P. Monferrer Sala, J. Pahlitzsch, M. Swanson, H. Teule, and J. Tolan, History of Christian-Muslim Relations, 11 (Leiden: Brill, 2009), pp. 439–91 Lane, E. W., An Arabic-English Lexicon (London: Williams and Norgate, 1863–1893) Livingstone, A., ‘An Early Attestation of the Arabic Definite Article’, Journal of Semitic Studies, 42 (1997), 259–61 Macdonald, M. C. A., ‘Reflections on the Linguistic Map of Pre-Islamic Arabia’, Arabian Archaeo­logy and Epi­graphy, 11 (2000), 28–79 —— , ‘Old Arabic (Epi­graphic)’, in Encyclopedia of Arabic Language and Linguistics, ed. by K. Versteegh (Leiden: Brill, 2005–2009), iii, pp. 464–77 —— , ‘Ancient Arabia and the Written Word’, in The Development of Arabic as a Written Language, ed. by M. C. A. Macdonald, Supplement to the Proceedings of the Seminar for Arabian Studies, 40 (Oxford: Archaeopress, 2010), pp. 5–28 Mackintosh-Smith, T., Arabs: A 3,000-Year History of Peoples, Tribes and Empires (New Haven, CT: Yale University Press, 2019) Macomber, W. F., ‘The Literary Activity of Hunain b. Ishaq in Syriac’, in Mihrajān Afrām wa-Ḥunayn (Baghdad: n.p., 1974), pp. 545–70 Macuch, R., Geschichte der spät- und neusyrischen Literatur (Berlin: De Gruyter, 1976) Martin-Jones, M., ‘Diglossia’, in International Encyclopedia of Linguistics, ed. by W. J. Frawley, 2nd edn (Oxford: Oxford University Press, 2003), ii, pp. 435–38 Mengozzi, A., Israel of Alqosh and Joseph of Telkepe: A Story in a Truthful Language; Religious Poems in Vernacular Syriac (North Iraq, 17th century), Corpus Scriptorum Christianorum Orientalium, 589–90 (Leuven: Peeters, 2002) —— , ed., Religious Poetry in Vernacular Syriac from Northern Iraq (17th–20th centuries): An Antho­logy, Corpus Scriptorum Christianorum Orientalium, 627–28 (Leuven: Peeters, 2011) Meyerhof, M., ‘New Light on Ḥunain ibn Isḥāq and his Period’, Isis, 8 (1926), 685–724 Militarev, A., and L. Kogan, Semitic Etymo­logical Dictionary, ii: Animal Names, Alter Orient und Altes Testament, 278/2 (Münster: Ugarit-Verlag, 2005) Millar, F., The Roman Near East, 31 b.c.–a.d. 337 (Cambridge, MA: Harvard University Press, 1993) Monferrer Sala, J. P., and B. Roggema, ‘Ḥunayn ibn Isḥāq’, in Christian-Muslim Relations: A Biblio­graphic History, i: 600–900, ed. by D. Thomas and B. Roggema, with J. P. Monferrer Sala, J. Pahlitzsch, M. Swanson, H. Teule, and J. Tolan, History of ChristianMuslim Relations, 11 (Leiden: Brill, 2009), pp. 768–79 Munt, H., with others, ‘Arabic and Persian Sources for Pre-Islamic Arabia’, in Arabs and Empires before Islam, ed. by G. Fisher (Oxford: Oxford University Press, 2015), pp. 434–500 Nebes, N., ‘The Martyrs of Najrān and End of the Ḥimyar: On the Political History of South Arabia in the Early Sixth Century’, in The Qurʾān in Context: Historical and Literary Investigations into the Qurʾānic Milieu, ed. by A. Neuwirth, N. Sinai, and M. Marx (Leiden: Brill, 2010), pp. 27–59

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

Nehmé, L., ‘Aramaic or Arabic? The Nabateo-Arabic Script and the Language of the Inscriptions Written in this Script’, in Arabic in Context: Celebrating 400 Years of Arabic at Leiden University, ed. by A. Al-Jallad, Studies in Semitic Languages and Linguistics, 89 (Leiden, Brill, 2017), pp. 75–98 Newton, F., ‘Constantine the African and Monte Cassio: New Elements and the Text of the Isagoge’, in Constantine the African and ʿAlī ibn al-ʿAbbās al-Maǧūsī: The ‘Pantegni’ and Related Texts, ed. by C. Burnett and D. Jacquart, Studies in Ancient Medicine, 10 (Leiden: Brill, 1994), pp. 16–47 Nicosia, M., ‘Reassessing Antony of Tagrit: When Did He Actually Live?’, Oriens Christianus, 104 (2021), 68–89 Nöldeke, T., Geschichte der Perser und Araber zur Zeit der Sasaniden (Leiden: Brill, 1879) —— , Geschichte des Qorāns, 2nd, enlarged edn, ed. by F. Schwally, G. Bergsträsser, and O. Pretzl (Leipzig: T. Weicher, 1909–1938) —— , The History of the Qurʾān, trans. by W. H. Behn (Leiden: Brill, 2013) Olsson, J. T., ‘The Reputation of Ḥunayn ibn Isḥāq in Contemporaneous and Later Sources’, Journal of Abbasid Studies, 3 (2016), 29–55 Overwien, O., ‘The Art of the Translator, or: How Did Ḥunayn ibn ʾIsḥāq and his School Translate’, in Epidemics in Context: Greek Commentaries on Hippocrates in the Arabic Tradition, ed. by P. E. Pormann (Berlin: De Gruyter, 2012), pp. 151–69 Owens, J., A Linguistic History of Arabic (Oxford: Oxford University Press, 2006) Papaconstantinou, A., ‘Why Did Coptic Fail Where Aramaic Succeeded? Linguistic Developments in Egypt and the Near East After the Arab Conquest’, in Multilingualism in the Graeco-Roman Worlds, ed. by A. Mullen and P. James (Cambridge: Cambridge University Press, 2012), pp. 58–76 Pormann, P. E., ‘The Development of Translation Techniques from Greek into Syriac into Arabic: The Case of Galen’s On the Faculties and Powers of Simple Drugs’, in Medi­eval Arabic Thought: Essays in Honour of Fritz Zimmermann, ed. by R. Hansberger, M. Afifi al-Akiti, and C. Charles Burnett (London: Warburg Institute; Turin: Nino Aragno Editore, 2012), pp. 134–63 Rassi, S., Christian Thought in the Medi­eval Islamicate World: ʿAbdīshōʿ of Nisibis and the Apo­ logetic Tradition (Oxford: Oxford University Press, 2022) Reller, J., Mose bar Kepha und seine Paulinenauslegung (Wiesbaden: Harrassowitz, 1994) Retsö, J., The Arabs in Antiquity: Their History from the Assyrians to the Umayyads (London: Routledge, 2003) Robin, C. J., ‘Les langues de la péninsule Arabique’, Revue du monde musulman et de la Méditerranée, 61 (1991), 89–111 —— , ‘Les arabes vus de Ḥimyar’, Topoi, 14 (2006), 121–73 —— , ‘Joseph, dernier roi de Ḥimyar (de 522 à 525, ou une des années suivantes)’, Jerusalem Studies in Arabic and Islam, 34 (2008), 1–124 —— , ‘Ḥimyar, Aksūm, and Arabia Deserta in Late Antiquity’, in Arabs and Empires before Islam, ed. by G. Fisher (Oxford: Oxford University Press, 2015), pp. 127–71 Ross, S. K., Roman Edessa: Politics and Culture on the Eastern Fringes of the Roman Empire, 114–242 c.e. (London: Routledge, 2001) Rothstein, G., Die Dynastie der Laḫmiden in al-Ḥîra: Ein Versuch zur arabisch-persischen Geschichte zur Zeit der Sasaniden (Berlin: Reuther & Reichard, 1899)

399

4 00

a a ro n m icha e l b u t t s

Samir, K., Abū Qurrah: Vida, bibliografía y obras, trans. by J. P. Monferrer Sala (Córdoba: Universidad de Córdoba, 2005) Sanders, J. C. J., Inleiding op het Genesiskommentaar van de Nestoriaan Ibn at-Taiyib (Leiden: Brill, 1963) Sebba, M., ‘Societal Bilingualism’, in The SAGE Handbook of Sociolinguistics, ed. by R. Wodak, B. Johnstone, and P. Kerswill (London: SAGE, 2011), pp. 445–59 Segal, J. B., ‘Arabs in Syriac Literature before the Rise of Islam’, Jerusalem Studies in Arabic and Islam, 4 (1984), 89–123 Sezgin, F., Geschichte des arabischen Schrifttums (Leiden: Brill, 1967–) Shahîd, I., Byzantium and the Arabs in the Fourth Century (Washington, DC: Dumbarton Oaks Research Library and Collection, 1984) —— , Byzantium and the Arabs in the Fifth Century (Washington, DC: Dumbarton Oaks Research Library and Collection, 1989) —— , Byzantium and the Arabs in the Sixth Century (Washington, DC: Dumbarton Oaks Research Library and Collection, 1995–2009) Sima, A., Tiere, Pflanzen, Steine und Metalle in den altsüdarabischen Inschriften: eine lexikalische und realienkundliche Untersuchung (Wiesbaden: Harrassowitz, 2000) Strohmaier, G., ‘Ḥunayn ibn Isḥāq und die Bilder’, Klio, 43–45 (1965), 525–33 —— , ‘Ḥunayn ibn Isḥāq: An Arab Scholar Translating into Syriac’, ARAM, 3 (1991), 163–70 Swanson, M. W., ‘Sāwīrus ibn al-Muqaffaʿ’, in Christian-Muslim Relations: A Biblio­ graphic History, ii: 900–1050, ed. by D. Thomas and A. Mallett, with J. P. Monferrer Sala, J. Pahlitzsch, M. Swanson, H. Teule, and J. Tolan, History of Christian-Muslim Relations, 14 (Leiden: Brill, 2010), pp. 491–509 Takahashi, H., Barhebraeus: A Bio-Biblio­graphy (Piscataway, NJ: Gorgias Press, 2005) —— , ‘Syriac as the Intermediary in Scientific Graeco-Arabica: Some Historical and Philo­ logical Observations’, Intellectual History of the Islamicate World, 3 (2015), 66–97 Taylor, D. G. K., ‘Bilingualism and Diglossia in Late Antique Syria and Mesopotamia’, in Bilingualism in Ancient Society: Language Contact and the Written Text, ed. by J. N. Adams, M. Janse, and S. Swain (Oxford: Oxford University Press, 2002), pp. 298–331 —— , ‘A Stylistic Comparison of the Syriac Ḥimyarite Martyr Texts Attributed to Simeon of Beth Arsham’, in Juifs et chrétiens en Arabie aux ve et vie siècles, ed. by J. Beaucamp, F. Briquel-Chatonnet, and C. J. Robin (Paris: Association des Amis du Centre d’Histoire et Civilisation de Byzance, 2010), pp. 143–76 Ter Haar Romeny, B., ‘Biblical Studies in the Church of the East: The Case of Catholicos Timothy I’, in Historica, Biblica, Theo­logica et Philosophica: Papers Presented at the Thirteenth International Conference on Patristic Studies held in Oxford 1999, ed. by M. F. Wiles, E. Yarnold, and P. M. Parvis (Leuven: Peeters, 2001), pp. 503–10 Teule, H. G. B., ‘Renaissance, Syriac’, in GEDSH, pp. 350–51 Teule, H. G. B., and C. F. Tauwinkl, with R. B. ter Haar Romeny and J. van Ginkel, ed., The Syriac Renaissance, Eastern Christian Studies, 9 (Leuven: Peeters, 2010) Toral-Niehoff, I., ‘The ʿIbād of al-Ḥīra: An Arab Christian Community in Late Antique Iraq’, in The Qurʾān in Context: Historical and Literary Investigations into the Qurʾānic Milieu, ed. by A. Neuwirth, N. Sinai, and M. Marx (Leiden: Brill, 2010), pp. 323–48 —— , Al-Ḥîra: eine arabische Kulturmetropole im spätantiken Kontext (Leiden: Brill, 2013) —— , ‘Late Antique Iran and the Arabs: The Case of al-Hira’, Journal of Persianate Studies, 6 (2013), 115–26

the T w i n H is to r i e s o f Ar a b i c a n d Ar a m a i c (w i t h a Fo cu s o n Sy ri ac)

Ullmann, M., Die Medizin im Islam (Leiden: Brill, 1970) Vagelpohl, U., ‘In the Translator’s Workshop’, Arabic Sciences and Philosophy, 21 (2011), 249–88 Van Coetsem, F., Loan Phono­logy and the Two Transfer Types in Language Contact, Publications in Language Sciences, 27 (Dordrecht: Foris Publications, 1988) —— , ‘Outlining a Model of the Transmission Phenomenon in Language Contact’, Leuvense Bijdragen, 84 (1995), 63–85 —— , A General and Unified Theory of the Transmission Process in Language Contact (Heidelberg: Winter, 2000) Van Putten, M., ‘The Development of the Triphthongs in Quranic and Classical Arabic’, Arabian Epi­graphic Notes, 3 (2017), 47–74 —— , ‘The Feminine Ending -at as a Diptote in the Qurʾānic Consonantal Text and its Implications for Proto-Arabic and Proto-Semitic’, Arabica, 64 (2017), 695–705 —— , ‘Case in the Qurʾānic Consonantal Text’, Wiener Zeitschrift für die Kunde des Morgenlandes, 108 (2018), 143–79 —— , ‘Hamzah in the Quranic Consonantal Text’, Orientalia, 87 (2018), 93–120 —— , ‘Inferring the Phonetics of Quranic Arabic from the Quranic Consonantal Text’, International Journal of Arabic Linguistics, 5 (2019), 1–19 —— , Quranic Arabic: From its Hijazi Origins to its Classical Reading Traditions, Studies in Semitic Languages and Linguistics, 106 (Leiden: Brill, 2022) Van Rompay, L., ‘Some Preliminary Remarks on the Origins of Classical Syriac as a Standard Language: The Syriac Version of Eusebius of Caesarea’s Ecclesiastical History’, in Semitic and Cushitic Studies, ed. by G. Goldenberg and S. Raz (Wiesbaden: Harrassowitz, 1994), pp. 70–89 —— , ‘Nagran’, in GEDSH, pp. 302–03 Wardle, T. S., ‘Abgarids of Edessa’, in GEDSH, pp. 5–7 Wasserstein, D. J., ‘Why Did Arabic Succeed Where Greek Failed? Language Change in the Near East after Muḥammad’, Scripta Classica Israelica, 22 (2003), 257–72 Watt, J. W., ‘Anṭun of Tagrit’, in GEDSH, p. 23 Webb, P., Imagining the Arabs: Arab Identity and the Rise of Islam (Edinburgh: Edinburgh University Press, 2016) Weinreich, U., Languages in Contact: Findings and Problems (New York: Linguistic Circle of New York, 1953) Wood, P., ‘Al-Ḥīra and its Histories’, Journal of the American Oriental Society, 136 (2016), 785–99 Wood, P., with G. Greatrex, ‘The Naṣrids and Christianity in al-Ḥīra’, in Arabs and Empires before Islam, ed. by G. Fisher (Oxford: Oxford University Press, 2015), pp. 172–213, 257–63

401

Arianna D’Ottone

Sharing the Written Space Contact and Interaction between Arabic and Other Cultures and Scripts le difficile est […] de savoir ce qu’il faut savoir voir1

Multi­graphism and Allo­graphy A feature linked to Late Antique book culture, and attested in the Middle East area in particular, is the complex phenomenon of absolute multi-­graphism, which can be exemplified by di­graphic — and bilingual — book production in the GraecoRoman period and beyond.2 This is a point of particular interest in understanding how Arabic script developed, and it is attested in some early manu­scripts.



1 Marichal, Samaran and Dufour ‘Paléo­graphie latine’, p. 416. 2 ‘The dozens of bilingual manu­scripts containing glossaries and scholastic texts, which were certainly produced in the East in the Late Antiquity, clarify where some of the great late-antique codices of the Bible in Greek and Latin were written: from the East came both the Codex Bezae and the [Codex] Claromontanus of St Paul’s Epistles. […] It is in the Orient that we have clear proof of absolute multi­graphism, firstly in the Greek-Coptic codices and also in the tradition that was established for studying sacred languages that goes back to Origen’ (‘Le molte decine di manoscritti bilingui contenenti glossari e testi scolastici, sicuramente prodotti in area orientale durante la tarda antichità, chiariscono dove sono stati scritti alcuni grandi codici tardoantichi contenenti le Sacre Scritture in edizioni grecolatine: dall’Oriente vengono appunto il Codex Bezae ed il Claromontanus delle Epistole paoline. […] È in Oriente che abbiamo testimonianza sicura di multigrafismo assoluto, già nei codici greco-copti, ma anche nella presenza di una tradizione di indagini sulle lingue sacre che rimonta almeno ad Origene’) Radiciotti ‘Il problema del digrafismo’, p. 20. On this last point, it is useful to recall Ronny Vollandt’s recent comments on the translations of the Hebrew Bible into Arabic: ‘The ultraliteral approach, as pedagogical tool in scholastic environments, goes back to antiquity. Aquila’s famous retranslation of the Torah into Greek, often described as a mirror translation, is said to have had a didactic end. Outside the Bible, an illustration can be found in a number of bilingual texts of Vergil and others, with the Latin and Greek in parallel columns’, Vollandt ‘Translations as Linguistic Commentaries’, p. 266. It is important to stress that the use of the term ‘di­graphic’ is often used in classical palaeo­graphy to describe in a context of ‘relative multi­graphism’ the work of a scribe writing in two different styles of script within a single g­ raphic system. On the manus duplex, see, for example, Bianconi, Paleografia.

Arianna D’Ottone Rambach is Professor of Arabic Language and Literature at Sapienza University of Rome, where she teaches Arabic and Arabic Palaeo­graphy, Codico­logy and Numismatics. Navigating Language in the Early Islamic World: Multilingualism and Language Change in the First Cen­turies of Islam, ed. by Antoine Borrut, Manuela Ceballos, and Alison M. Vacca, ISMAR 2 (Turnhout: Brepols, 2024) pp. 403–423 10.1484/M.ISMAR-EB.5.134633

FHG

4 04

a r i a n n a d’ot to n e Figure 10.1. BBAW/ GCS, Akz.-Nr. 481/136, photo­graphed by Bruno Violet in 1901. Third/ninth century.

I have previously had occasion to discuss various specific characteristics of this phenomenon, with particular regard to the Latin script ­graphic skills of the copyist of a bilingual and di­graphic3 Latin-Arabic fragment. The text in question was from St Paul’s Epistle to the Galatians (MS Vat.lat. 12900, fol. 2) that had been produced in the Western part of the Islamic lands between the end of the ninth and the beginning of the tenth centuries.4



3 In that paper, devoted to Oriental manu­scripts produced in a context of ‘absolute multi­graphism’ in which different languages and writing systems were used simultaneously, I adhered to Paolo Radiciotti’s distinction between what is ‘bilingual’ and ‘di­graphic’, adopting a palaeo­graphic rather than linguistic research perspective; Radiciotti ‘Scrivere e leggere il greco’, pp. 185–86. 4 See D’Ottone, ‘Al-khaṭṭ al-maghribī’, p. 153. See also Marcos Marín, ‘Vat.Lat. 12900’, p. 984 and Monferrer-Sala ‘The Fragmentary Ninth/Tenth Century’, p. 152. Their linguistic and philo­logical observations, regarding the attribution of the fragment to the end of the ninth/beginning of the tenth century, support what I had suggested on palaeo­graphic grounds.

s h ari ng t he w ri t t e n space Figure 10.2 BBAW/ GCS, Akz.-Nr. 481/137, photo­graphed by Bruno Violet in 1901. Third/ninth century.

All photos in this chapter © Berlin-Brandenburg Academy of Sciences and Humanities. Reproduced with permission.

The aim of the present study is to examine a Christian-Arabic fragment of Psalms 118:57–125:6 — according to the numbering in the Septuagint5 — found amongst the manu­script material found in the Qubbat al-khazna in Damascus.6 A palaeo­graphical approach will be adopted. Fragments BBAW-GCS, Akz.-Nr. 481/136–37 (Fig. 10.1–2) can be dated and localized, Hjälm suggests, to tenth-century Palestine, and they bear an Arabic translation based on a Greek version of the Psalms. Commenting on the palaeo­graphic features of these fragments, Hjälm attributes them to ‘a transitional stage between New Style scripts and the more simple, round naskh scripts […] which to a large degree superseded both angular and New Style

5 See Hjälm, 2020. On the Arabic translations of the Psalms, see Sasson, ‘Psalms’. 6 On the history of the Qubbat al-khazna find, see D’Ottone Rambach, Hirschler, and Vollandt, eds, The Damascus Fragments.

405

4 06

a r i a n n a d’ot to n e

Figure 10.3. BBAW/GCS, Akz.-Nr. 481/136, detail of line 11, left-hand folio — photo­graphed by Bruno Violet in 1901.

scripts in tenth-century Palestine’.7 In particular, Hjälm stresses ‘the oblique top stroke on final kāf [is] a feature that in general seems to be common from the tenth century onwards’ (Fig. 10.3).8 However, at least in the fragments BBAW-GCS, Akz.-Nr. 481/136–37, the oblique strokes on the top of the final kāf-s seem to be a later addition to the original shape of the letter and not the result of the actual will of the copyist.9 This observation is suggested by the following details: – strokes do not appear systematically in each occurrence of the letter — they are missing in some instances, e.g. in anā sharīk (Fig. 10.4) – strokes are not connected to the body of the letter-s by a consistent movement of the copyist’s hand – the thickness of the strokes does not match that of the script — as if they were realised in a second moment by a different hand, using a different pen. Moreover, in this di­graphic and bilingual fragment in Arabic and Greek script10 it is also possible to note a third component on the written page. Whilst the anonymous

7 Hjälm compares the script to the MS Sinai MS Ar. NF. Parch. 27; see Hjälm, ‘From Palestine to Damascus to Berlin’, p. 256. 8 However at least in the fragments BBAW-GCS, Akz.-Nr. 481/136–37 the oblique stroke of the final kāf seems to be a later addition. 9 If the stroke is an addition made soon after, or much later after, the copy of the text it is not possible to say since we do not have the original manu­script to check, at least, the colour of the ink — if comparable or not to other inks employed for its copy. In any case it would be difficult to determine the closeness in time — or not — of those additional strokes. Possibly the top stroke might have been added in the tenth century — when its use became common — in order to modernize a script that was considered not matching the standard look of the manu­script of the time. A tenth-century scribe, for instance, added g­ raphic elements — such as spirits and accents as well as diacritics — to the famous manu­script B of the Bible — written in the fourth century —, see Versace ‘I marginalia’, pp. 43–50. As it would be misleading considering the codex a tenth century copy on the basis of this additional elements, it seems misleading to date the fragments BBAW-GCS, Akz.-Nr. 481/136–37 to the tenth century. Yet, the additional elements can be a terminus ante quem and the fragments can be considered to be produced earlier than the tenth century. 10 On the Greek script, with its example of maiuscola ogivalis/sloping ogival (or pointed) majuscule of which two variants are known (inclinata and recta), see Cavallo, ‘Funzione e strutture’ (where a Palestinian variant of the maiuscola ogivalis inclinata is identified), Crisci, ‘La maiuscola ogivale diritta’, Harlfinger, ‘Beispiele der maiuscola’. On the presence of graeca in Oriental (Syriac, Arabic and Sogdian manu­scripts), see D’Aiuto, ‘Graeca’; D’Aiuto, ‘Un antico inno’ and D’Aiuto & Bucca, ‘Some

s h ari ng t he w ri t t e n space

407

Figure 10.4. BBAW/GCS, Akz.-Nr. 481/136, detail of line 9, left-hand folio — photo­graphed by Bruno Violet in 1901.

copyist who added single lines in Greek was in all likelihood a different copyist from the one writing in Arabic,11 the writing style of the latter — also anonymous — provides us with a few clues12 as to what his ­graphic education might have been. Indeed, some details in the Arabic text suggest a scribe who, while able to write Hebrew, could also write in Arabic.13 As has been pointed out by Bassal, ‘the translations of the Christian Arabic Bible exists [sic] either in fragments from the 9th century or in complete compositions or complete units of the Bible from the 10th century. They were written using various sources: Syriac-Aramaic, Greek and Latin, Coptic and Hebrew. Actually, it is impossible to talk about one single source of Arabic sources’.14

11

12 13

14

Greek hymno­graphic’. On the Sogdian fragments with Greeks titles, see also Sims-Williams ‘A New Fragment’. It is not surprising to read in Binggeli, ‘Les trois David’, p. 88, in relation to the bilingual Greek-Arabic Book of Psalms — MSS Sinai gr. 35 + Milano, Biblioteca Ambrosiana, L 120 sup. (fols 125/32 and 139/34) — that the Greek script — again with a maiuscola ogivalis/sloping ogival (or pointed) majuscule — bears no similarity to that of Anthony David. Indeed, nothing leads us to conclude that it could have been by his hand (‘L’écriture grecque est une majuscule ogivale incline très excercée, de la seconde moitié du 9e siècle. Elle ne présente toutefois aucun point de contact avec l’écriture d’Antoine David, et rien ne laisse supposer qu’elle puisse être de sa main’). However, this still pertinent observation seems to contradict the aim of that particular study, which, according to the author, was to ‘re-evaluate, in a comparative way, the manu­script production of the three copyists — one of whom is Anthony David — in the multilingual context of the 9th- and 10th-century Palestine, where the presence of Greek scribes deeply influenced and characterized the habits of Arabic scribes’ (‘Le present article se propose ainsi de réévaluer de manière comparative les productions manuscrites de ces trois copistes et de les restituer dans le context plurilingue de la Palestine des 9e–10e siècles, où la presence des copistes grecs a fortement influence et imprégné la pratique des copistes arabes’); Binggeli, ‘Les trois David’, p. 79. In Italian, ‘spie’; see Ginzburg, ‘Spie’, p. 91. In the same way it is possible to detect the original language that lies behind a translated text through the study of the translation technique adopted (Übersetzungsgrammatik). We might mention, for example, the Arabic translation made “directly from a Greek original” of John of Damascus’ Exposition of the Orthodox Faith, studied by Treiger, ‘Syro-Arabic Translations’. In certain cases, it therefore seems possible to detect the ­graphic substratum that produces specific ­graphic forms. See Bassal, ‘An Early Copy of a Christian Arabic Pentateuch’, p. 13.

4 08

a r i a n n a d’ot to n e

Many interesting essays have been published on the influence, either conscious or unconscious, of (above all Hebrew) scribes in the copying and transmission of texts.15 More recently, fascinating observations have been made regarding the techniques adopted in early translations of the Bible from Hebrew into JudaeoArabic.16 Both philo­logical and translation studies have shown that it is possible to detect connections between languages, uncovering the means by which a text in a given language (e.g. Arabic) is connected to an underlying text originally written in a different language (e.g. Greek, Syriac or Hebrew).17 In turn, palaeo­graphy can be helpful in detecting such connections by analysing the script employed. A palaeo­graphic examination of some unintentional elements that emerge from the ‘writing transparency’ of fragments BBAW-GCS, Akz.-Nr. 481/136–37 contributes in determining the cultural identity of its copyist.18 In particular, I consider the shape of some final kāf-s ‫ ك‬in a ‘square’ form to be of particular significance. These have a small serif on top that recalls the Hebrew letter ‫ כ‬kaf or beth ‫ב‬, two letters that have a similar morpho­logy. A further ­graphic element in the Arabic text that seems worthy of consideration are the two diacritical points of the final yā’ placed vertically, a little above the line of writing: two vertical dots in the same position can also be found in Hebrew (Figs 10.4 and 10.5). The fact that the morpho­logy of a letter is affected by the shape of a similar letter in another alphabet is a relatively well attested phenomenon in various linguistic contexts.19 Moreover, in some of the manu­scripts copied at St Catherine

15 Malachi Beit-Arié, for example, makes a distinction between scribes and copyists and their respective mistakes. For scribes, the act of writing and reproducing someone else’s texts was a profession, and they were concerned with neither major nor minor changes and additions in a given text, but with mistakes that might be introduced through any lack of attention when mechanically reproducing the text. For copyists, the act of writing was carried out for their own interests, and any interventions — emendations, modifications and, when needed, reconstructions of the textual sequence based on personal knowledge — went beyond mere correctness to produce a kind of critical edition of the text in question; see Beit-Arié, ‘Transmissions des textes’, p. 182. 16 On the difference between the translation of the Bible by Saadia Gaon and early non-Saadianic translations, see Vollandt, ‘Translations as Linguistic Commentaries’, in particular p. 265. 17 On translation techniques adopted in both Judaeo-Arabic and Syriac versions of the Bible, see Butbul, ‘Translation in Contact’. 18 ‘La comparaison des écritures des copistes qui ont transcrit un même livre (qu’il n’est pas difficile de distinguer les uns des autres à cause des différences dans les formes des lettres), nous livre aussi les éléments individuels à l’intérieur de la stéréotypie traditionnelle. Elle montre que les éléments individuels sont des éléments ­graphiques associés aux lettres de l’écriture. L’écriture elle-même peut être très uniforme, mais c’est par ces éléments qu’on distingue le copiste; par des tendances et des habitudes qui lui sont propres dans le choix étendu des genres en usage dans la tradition g­ raphique d’une même zone culturelle’ Beit-Arié, ‘Stéréoptypies et individualités’, p. 204. The earliest dated biblical codices produced in the Middle East, evidence a square mode in the execution of Hebrew writing ‘rich in decorative elements that increase the number of strokes required for executing the letters’ Beit-Arié, Unveiled Faces, pp. 70–71. This square mode remained in use until the early eleventh century. 19 I recall here the case of the letter l in Latin inscriptions on ostraka from Libya. The particular, bent shape of the horizontal line in the letter l — in these inscriptions has its origins in the same letter in the neo-Punic inscription from the same area; see Radiciotti, ‘Scrittura Latina per ostraka neopunici’,

s h ari ng t he w ri t t e n space 1. wa-‘alà kalāmika

2. fī yadika

3. khurūjaka

409

4. ‘alà yadika

Figure 10.5. Details from Berlin-Brandenburg Academy of Sciences and Humanities, BBAW/GCS, Akz.-Nr. 481/136–37: Square kāf-s and diacritical dots — photo­graphed by Bruno Violet in 1901.

on Mount Sinai, it has also been possible to detect ‘clumsy copies of early Western scripts, such as Visgothic, Beneventan, Luxeil and Caroline minuscule; and these may represent attempts by Sinai scribes to copy the scripts of imported Western manu­scripts or the hands of foreign scribes they had met’.20 In addition, at least a tenth-century palimpsest codex with a Karaite translation of the Old Testament has been identified at St Catherine (Sinai Gr. 930)21. Both voluntary imitations and involuntary traces, such as introducing unusual letter shape(s) associated with a different g­ raphic system into a scribal practice, attest to close links between scripts, texts and — in some cases — languages.22 In this case both the shape of the final square kāf and the notation of the dia­critics (their position and orientation),23 suggest the influence of Oriental manu­ scripts produced in Hebrew script,24 above all, though not exclusively, in Palestine.

20 21 22

23

24

p. 118. In this context, another example of ­graphic influence between different alphabets is the general influence that a certain ­graphic system may exercise over another: there is the particular allure of a different script, even though it does not affect the morpho­logy of the letters. This occurrence is illustrated, for example, by the roundish appearance of the Arabic legends found in the first postreform Islamic coins, the dies of which were produced by engravers who were used to writing Pahlavi, a script characterized by curvilinear letter-forms; see Heidemann, ‘Calli­graphy on Islamic Coins’, p. 163. On the influence of Arabic script on Hebrew script, see Olszowy-Schlanger ‘Crossing Palaeo­ graphical Borders’. On the possible influence of Latin script on a Christian Arabic writer producing a 9th-century Latin manu­script in Visigothic script, see Monferrer-Sala and Cecini, ‘Once Again on Arabic “alkaufeit”’, p. 210. Brown, ‘The Bridge in the Desert’, p. 91. On the connection between the monasteries of St Catherine’s Sinai and Montecassino, and the possible g­ raphic encounter between Arabic and Latin script, see D’Ottone Rambach, ‘From Monte Cassino abbey’. Treiger, ‘A Newly Discovered Karaite Arabic Translation’. On the influence of Latin script on Greek hands, and on the linguistic influence of the dialect of the Salento area that can be found in Greek texts in various manu­scripts and manu­script notes produced in the monastery of Grottaferrata between the 11th and 12th centuries, see Lucà, ‘Interferenze linguistiche greco-latine’, pp. 310–11. Discussing ninth- and tenth-century Arabic Christian manu­script production, Griffith notes that ‘some scribes of the period went so far as to present their Arabic texts construed with the customary punctuation employed by writers of Syriac’ Griffith, Arabic Christianity in the Monasteries, p. 129. Nothing prevents us from applying exactly the same logic to Arabic texts that were penned by scribes with a ­graphic education in Hebrew. Among the features of the Oriental Biblical square script, it is important to stress that ‘vertical and horizontal lines are made by the same width of the stroke, without shading […]. Vertical lines are made by erect strokes ending in a curve’ Engel, ‘Script, History and Development’, p. 487.

410

a r i a n n a d’ot to n e

The elements that determine the particular shape of the square kāf attested in the fragments BBAW-GCS, Akz.-Nr. 481/136–37 include an additional trait — the top stroke — and a decorative element, the serif.25 These elements give a distinctive form to the letters, which differentiates them from previous or contemporary morpho­logies of the same letter.26 In fragments BBAW-GCS, Akz.-Nr. 481/136–37, the influence of the Hebrew script on the Arabic script, in what are already di­graphic and bilingual texts, adds a third, further script — and language — to the page. It is the ­graphic education of the copyist that can elucidate the peculiar morpho­logy of these ‘square kāf-s’, and the connection between the two g­ raphic systems used in writing Hebrew and Arabic, is not necessarily surprising.27 In a study devoted to Stephen of Ramlah, Sidney Griffith did not ignore this palaeo­graphic aspect when discussing the importance of ninth- and tenth-century Christian Arabic manu­scripts. He commented: ‘the persons who wrote these texts, while they were Arabic-speakers, had obviously been educated in Greek and in Syriac, a circumstance that affected their writing in Arabic’.28 The name of this scribe — as recorded in the colophon of the codex London, BL 4950, fol. 197v — is ‘Stephen son of Ḥakam, known as al-Ramlī’ and Griffith unhesitatingly defines him as a native of al-Ramlah.29 Considering the importance of onomastic evidence in the past, the name Stephen can reasonably be linked to a Greek Christian context.30

25 ‘In conclusion, in enquiring writing it is necessary at first to know — and to define scientifically — what to look at, what to dwell on […]. In the synchronic analysis, carried out on a single manu­ script or on a group of similar manu­scripts, the structure, production and style of the script are the elements that should be considered. Among the structural elements of the script we find the shape of the letters […] Finally, aspects of the style are accessory elements of the g­ raphic rendering, such as individual quirks (of an age or a writer), decorations and even further para- and peri-­graphical elements. These characteristics, even if not always together not altogether, are essential if, for example, one wants […] to follow the development of a writing rule or style, to distinguish contemporary hands, or even to demonstrate the identity of the same writer in different pieces of evidence’ Bianconi, ‘Palaeo­graphy: Introduction’, pp. 268–69. 26 Other morpho­logies include the ‘open kāf’ attested in papyri and in ninth-century Arabic-Islamic (non-Qur’anic) manu­scripts, such as the codex Leiden, University Library, Or. 298, dated 253/866 containing the text Gharīb al-ḥadīth by Abū ‘Ubayd (d. 223/837), and the ‘kāf with a tail’, that is, with a long slant trait on top of a ‘hair-pin’ letter-form body, as in the manu­script London, British Library, Add. 26116, a ninth-century fragment of a vellum codex — 15 folios — containing the Book of Job. 27 Hebrew can be considered one of the ‘indigenous languages’ of Palestine in the Islamic period, along with Aramaic and Arabic; see Griffith, ‘From Aramaic to Arabic’, p. 19, note 46. 28 See Griffith ‘Stephen of Ramlah’, p. 38 (= Arab Christianity, VII). 29 ‘Stephen’s own sobriquet, ar-Ramlī, must mean that he was a native of the town of ar-Ramlah’ Griffith, ‘Stephen of Ramlah’, p. 40 (= Arab Christianity, VII). 30 Regarding Christian onomastic evidence in early Byzantine Egypt, Roger Bagnall writes that ‘people in antiquity took more seriously than we do the religious connotations of the names they gave to their children’ and distinguishes Stephanos as one of the names taken from the New Testament that was used by Christians from outside the city walls (chora), Bagnall, ‘Religious Conversion’, p. 108, p. 110. Stephanos remained among the Christian names — many of which were also taken from the Old Testament — in use in Egypt even after the Islamic conquest; see Papaconstantinou, ‘What Remains Behind’. If in seventh- and eighth-century Egypt Biblical names from the Old Testament,

s h ari ng t he w ri t t e n space

If one turns to ninth-century Christian Arabic manu­scripts and their palaeo­ graphic features, it has been remarked that ‘whether intentionally or not, the monks of the Holy Land employed a recognizable distinctive style of writing in the ninth and tenth centuries’.31 In 2011 Alain George examined the script of a group of six Arabic Christian manu­scripts in particular,32 among which he noticed — apart from some variations — a number of specific elements and significant letters. George rightly noted, for example, the isolated form of the alif, resembling an inverted S with a central trait more or less straight or curved, and a left-side ending sometimes appearing like a thick roundish curl and at other times more like a hook.33 This particular morpho­logy of the alif, George observed, is an unusual feature if compared with other coeval Christian Arabic manu­scripts. Another characteristic shape noted by George is that of the final or isolated kāf, the vertical trait of which, he says, recalls the letter alif, while the connection with the horizontal base line consists of a very short, sometimes almost non-existent, oblique trait.34 If we consider that the morpho­logy of the letter alif is so particular and that the final/ isolated kāf resembles it, we can say that both the alif and the final/isolated kāf are the most characteristics letters in this group of manu­scripts. A potential g­ raphic parallel to both these letters is provided by the shape of the vertical trait of the lamed in early tenth-century Hebrew Bibles.35 As far as the size of the script is concerned, the letters alif and kāf have, in Arabic, a medium-high height: this detail seems consistent with the size of the Hebrew lamed-s, a letter

as Papaconstantinou writes, ‘had the advantage of being totally neutral, even with respect to Islam’, Papaconstantinou, ‘What Remains Behind’, p. 460, it would be interesting to uncover the onomastic evidence and any related bio­graphical elements for other regions of the medi­eval Islamic world. However, for the monastic context in particular, one has to distinguish between names given at birth and those adopted on entering monastic life. Changing name when entering a monastery was not, however, always the rule: the Arab Michael the Synkellos ( Jerusalem 760-Costantinople 846), for example, ‘became a monk in the monastery of St Saba, and in 797 or 798 he was ordinated priest by the patriarch of Jerusalem’, but he kept his own name; see Kolia-Dermitzaki, ‘Michael the Synkellos’. 31 See Griffith, ‘The Monks of Palestine’, p. 7. 32 See George, ‘Le palimpseste Lewis-Mingana’. Among this group there is also a dated codex by a certain Isḥaq — the Old Testament name might not be coincidental — who wrote a monastic antho­ logy in 255 ah/ 868 ce. 33 ‘La g­ raphie du texte arabe chrétien de ces six manuscrits est quasiment identique. Nous sommes loin, ici, des codifications formelles des corans de la même époque. Des variations importantes peuvent être constatée dans les tracé de chaque lettre, mais il se dégage néanmoins certains traits d’ensemble: — l’alif indépendant tend vers une forme de ‘S’ inversé; la partie centrale de la haste est plus ou moins rectiligne ou incurvée. La terminaison supérieure oscille entre une épaisse boucle arrondie et un crochet redescendant vers la gauche; elle se limite parfois à un léger épaississement du trait. Cette variabilité de terminaisons, également observée dans les autres lettres montantes, est inhabituelle parmi les manuscrits arabe chrétiens de l’époque’ George, ‘Le palimpseste Lewis-Mingana’, p. 407. 34 ‘le kāf final ou isolé aune haste incurvée rappelant un alif. La liaison avec le retour horizontal inférieur se fait par un trait oblique court, parfois même quasi-inexistant. Le retour horizontal est un simple trait droit’ George, ‘Le palimpseste Lewis-Mingana’, p. 408. 35 In contrast, ninth-century Syriac manu­scripts do not provide any meaningful comparanda for these features; see for example the MS in serṭā script, London, British Library, Add. 14263 (dated 823 ce), illustrated in Schmidt, ‘Syriac Palaeo­graphy’, p. 317, Fig. 2.10.4.

411

41 2

a r i a n n a d’ot to n e

that, in comparison with the rest of the Hebrew script, is usually double, sometime triple (or more) in size due to its long vertical extension. Moreover, all these peculiar features seem to have close parallels in a tenth-century Karaite manu­script such as the codex London, British Library, Or. 2540.36 This specific morpho­logy of ‘hooked/curled’ left-sided top of vertical letters such as alif, lām and kāf is also typical of the hand of Anthony of Baghdād,37 one of the most distinctive ninth-century Christian Arabic manu­script hands.38 Unfortunately, the bio­graphical information at our disposal is not what one might wish. However, in one of the colophons of the manu­script he copied,39 Anthony wrote: ‘The poor sinner, Anthony David the son of Sulayman of Baghdad’.40 Before entering a monastery, Father Anthony/Anbā Anṭūna was Dā’ūd b. Sulaymān al-Baghdādī.41 Considering his name, we might hypothesize that Father Anthony 36 Available online at: https://www.bl.uk/manu­scripts/FullDisplay.aspx?ref Or_2540. 37 ‘C’est une écriture livresque abbaside, plus particulièrement de type palestino-sinaïtique, régulière mais souple, avec une grande régularité dans le tracé des lettres. On la reconnaît au premier abord par quelques tracés caractéristiques, notamment un crochet vers la gauche en haut de la haste verticale des alif, lām, et kāf final; quant à l’alif non lié, il est doté en plus d’un petit trait horizontal vers la droite, qui lui donne un aspect sinueux, presque dansant de S inversé’ Binggeli, ‘Les trois David’, p. 90. 38 Alain George remarks that Anthony of Baghdad used hooks more frequently than curls on the top of the alif and other vertical letters (‘Antoine de Bagdad traçait aussi le plus souvent des crochets plutôt que des boucles en tête de haste pour l’alif et autres lettres hautes’), and also noted that curled terminations recall Syriac script (‘Notons, tout d’abord, les terminaisons arrondies des lettres, qui rappellent le syriaque’) George, ‘Le palimpseste Lewis-Mingana’, p. 409, note 69 and p. 410. The fact that the Arabic hooked lām has some elements in common with the cognate lamed therefore does not seem to be coincidental; see supra. Indeed, the oblique top strokes on the final kāf that Hjälm notes in the fragments from the Qubbat al-khazna could be a possible (later?) addition to a specific morpho­ logy: a hooked kāf. Unfortunately, these fragments are no longer available and are only documented in pictures. It is thus impossible to carry out an autoptic examination of the ink, for example, from which one might determine if the strokes are coeval with the copy of the text or not. If they are additions, then an earlier date — the ninth century — could be put forward for the two fragments. 39 With regard to the colophons in Christian Arabic manu­scripts from ninth- and tenth-century Palestine as a source for discovering the origins of the monks who penned them, Griffith writes that ‘the sur­viving manu­scripts and their few surviving colophons also give the opportunity to gain some under­standing of the origins of the writers themselves, and some sense of the scope of their influence among the Christians living in the caliphate’ Griffith, ‘Greek into Arabic’, p. 123 (= Arab Christianity, VIII). 40 See Griffith, ‘Anthony David of Baghdad’, p. 8 (= Arab Christianity, XI). 41 Comparing the name given in a colophon of another text written by the same monk, Griffith writes: ‘Given the differences, one suspects that Anthony was the copyist’s monastic name, while David was his given name’ Griffith, ‘Anthony David of Baghdad’, p. 10. The patronymic — son of Sulaymān — is now acquired evidence; see (albeit with uncertain transcription) Binggeli, ‘Les trois David’, p. 80 and especially note 47, p. 90. As for the Christian name of Dā’ūd b. Sulaymān — Anbā Anṭūna — one cannot help but note its Coptic flavour, both in the use of the term Anbā/Apa — that ‘is interchangeable with the Arabic abbā, which occurs in Semitic languages, including Syriac, Aramaic, and even Hebrew’ — and in the choice of Anṭūna, perhaps linked to Saint Antony of Egypt, the ‘Father of the monks’. Antony was a third–fourth century hermit who had a considerable influence on the history of monasticism, and whose bio­graphy — written soon after his death by Saint Athanasius, Patriarch of Alexandria — became the cause of sudden conversions and was decisive, it appears, in the conversion of Saint Augustine; see Guillamont, ‘Antony of Egypt, Saint’; Atiya, ‘Apa’. As far as onomastic change goes, it is useful to recall a parallel case of a Christian, the Norman Johannes of

s h ari ng t he w ri t t e n space

came from a Jewish family and, possibly, not from Palestine but from Iraq, which was — at the time — the centre of the Abbasid caliphate.42 It is also useful to recall that the education of Jewish children in the medi­eval Islamic world would have involved writing in Arabic, and that higher education also included the study of the Bible and its Arabic and Aramaic translations.43 As regards manu­script production, the earliest dated extant bibles in Hebrew script date back to the early tenth century. Moreover, from the fourth/tenth century on, Karaites were also producing biblical manu­scripts both in Hebrew and in Arabic script.44 It therefore seems safe to say that Hebrew and Arabic writing systems were both in use in the ninth and tenth centuries in Islamic lands — Egypt, Iraq and Syria/Palestine as well as al-Andalus45 — at least within Jewish communities and, in particular, amongst the learned.46 Despite the fact that there is little evidence of a Judeo-Christian connection through Arabic in Old Testament translations,47 the influence of Saadyah’s Gaaon translation of the Bible was, in fact, quite significant ‘not only to Rabbanite and Karaite Jews and Egyptian and North African Christians, but also to Palestinian Christian communities as well as scholars in Iraq, where Saadyah lived’.48 It is also useful to recall that in the second half of the ninth century Christianity might have been an ‘attractive alternative for intellectual Jews’ in Medi­eval Iraq,49 that is, in the region where Dā’ūd b. Sulaymān and/or his family came from.

42 43 44 45 46 47

48 49

Oppido, who converted to Judaism in 1102 ‘and from that point on was known as Obadiah’, Franklin, Untidy History, p. 63. On the use of onomastic evidence as a means of investigating the process of conversion to Islam based on a case study of a converted Jew, see Wasserstein, ‘Lāwī b. Ismā‘īl b. Rabī‘ b. Sulaymān’. This generally corresponds to the description of Anthony David’s style as an Abbasid book script; see supra. See Frenkel, ‘Education: 1. Medi­eval Period’. This detail is perhaps of particular use in explaining the distinctive shape of the square-kāf(-s) in the Damascus fragments presented here. See Khan, ‘On the Question of Script’. In particular, on the process of acculturation of Jewish scribes to the Arabic language and script, see Beit-Arié, ‘External and Internal Frontiers’, pp. 404. In this respect it is interesting to recall the ninth century Oriental glosses of the Bible of Cava, written in Arabic and in Judaeo-Arabic by the same hand. This suggests a dual skill in writing the same language using different alphabets; see D’Ottone Rambach, ‘Lucifer and Arabic Palaeo­graphy’, p. 130. See Polliak, The Karaite Tradition of Arabic Bible Translation, pp. 7–8; Steiner, A Biblical Translation in the Making; Zewi, The Samaritan Version of Saadiyah Gaon’s. For example, MS Sinai Ar. 2 dated 328 ah/939–40 ce is translated from Hebrew with a Greek influence and its translation could be possibly connected to a convert, on MS Sinai Ar. 2, see Vollandt, Arabic Versions of the Pentateuch, pp. 142–51. On the Biblical translations see also Monferrer-Sala, ‘Jewish and Christian Bible Encounters in Arabic’, p. 98. The fragments BBAW-GCS, Akz.-Nr. 481/136–37 also bear an Arabic version that suggests a Greek original, but this does not run counter to the idea that the scribe, having a Jewish background and Hebrew writing skills, might have copied a Greek-based Arabic translation of the Bible in a monastic context. He copied what he was asked to copy from what he found available in the monastery library. See Monferrer-Sala, ‘Jewish and Christian Bible Encounters in Arabic’, p. 98. Almbladh, ‘Christianity and Judaism’, p. 26.

413

41 4

a r i a n n a d’ot to n e

Figure 10.6. BnF, MS gr. 497, fol. 1v, Arabic note in the upper margin. Photo source gallica.bnf.fr/ Bibliothèque nationale de France. (public domain).

s h ari ng t he w ri t t e n space

While there seems to be no connection between Jews and Christians for translation practices,50 such a link might be found instead in the palaeo­graphic evidence that possibly suggests the presence of former Jews in Palestinian monasteries. An examination of the g­ raphic characteristics that scribes unintentionally left in a written text suggests a complex cultural and linguistic context, both within and beyond the walls — and the languages — of monasteries.

Mobility in the Margins Manu­script notes can be an extremely insightful source of information in tracing the whereabouts of the owners and readers of the manu­scripts on which they are recorded, and they are often ‘a rich source for a number of fields, from the history of ideas, to social, economic and urban history, historical topo­graphy, and bio­graphical studies’.51 Moreover, manu­script notes in Arabic are abundant not only in Arabic manu­ scripts,52 but also in manu­scripts written in other Oriental and Classical languages.53 The Greek manu­script Paris, Bnf, gr. 497 contains the Homilies of Basil of Caesarea as well as the epitaph of Gregory of Nazianzus for Basil, and it was copied in North Africa in 966–67 — possibly at al-Mahdiyya — by a Byzantine dignitary, Niketas Abalante.54 As Leo the Deacon (born c.  950 ce) records, Niketas was admiral of the fleet of the emperor Nikephoros II Phokas. After his defeat at the Battle of the Straits (965 ce), he was imprisoned and taken to ‘Africa’ where, during his captivity, he copied the manu­script found today at the BnF.55 In fact, codex BnF gr. 497 provides us with a number of interesting marginal notes.56 One of these notes, in Arabic, has so far escaped the close scrutiny of Arabist scholars, and it attests the mobility of Arabic speaking Christian monks. At the top edge of fol. 1v, the note in question, partially trimmed and penned in brown ink, reads as follows (Fig. 10.6):

50 51 52 53

See Monferrer-Sala, ‘Jewish and Christian Bible Encounters in Arabic’, p. 99. See Görke & Hirschler ed., Manu­script Notes. See D’Ottone, ‘The Pearl and the Ruby’. For the Oriental glosses — in Arabic and Judaeo-Arabic — in a Visigothic manu­script dating back to the ninth century, known as Bible of Cava dei Tirreni or Danila’s Bible, see D’Ottone Rambach 2018. 54 See Mandalà, ‘Written Culture(s) in Islamic Sicily’, p. 136. 55 In 1676, the manu­script was acquired by Jean Baptiste Colbert, minister of the finances of Louis XIV, and ended up in France. 56 I wish to thank Annangela Germano, who in 2015 asked me to read the contents of the marginal note on fol. 1v. The codex, already known to Montfaucon and Michele Amari, has recently been mentioned by G. Mandalà in an overview devoted to the written culture(s) in Sicily between the ninth and the tenth centuries. The colophon of the manu­script is illustrated in Mandalà 2018 pl. 6, but no mention is made of the marginalia.

415

41 6

a r i a n n a d’ot to n e

‫وقف على هذا الميامن المبارك الحقير في الرهبان اننا يوحنا القس المقدسي [اللذ]ي يقرى بإكباب‬ ‫ من اوالد الحمامي بالقدس الشريف رحم من قرى هلل السماء‬/

The fortunate, the blessed and poor monk dwelled here: we are John the monk from Jerusalem (al-Muqaddasī) — from the line of al-Ḥammāmī in the noble Jerusalem — who reads with commitment. Lord on high be lenient with the reader. Unfortunately, the note lacks a date, but the fluent naskh script — with diacritical points, the emplacement of which confirms the Oriental background of the monk John,57 and a number of abusive ligatures due to his fluent Arabic hand — suggests that it was written well after the copying of the text, possibly between the end of the sixth/twelfth and the beginning of the seventh/thirteenth centuries. Since it is known that Nikephoros donated the manu­script to the church of St George (Ἅγιος Γεώργιος ὁ Μεγαλομάρτυς) in Morphou, Cyprus after Niketas was released in 970–971, it is possible that it was there that the monk John had access to the codex. The link between Cyprus — where the manu­script had been since the fifth/ eleventh century — and Jerusalem — where the monk John al-Ḥammāmī came from — seems particularly significant in the light of the historical events that took place at the end of the twelfth century. In 1186 Guy I of Lusignan (1150–1194) — married to the elder sister of Baldwin IV of Jerusalem (1161–1185), one of the four Christian kingdoms established in the Near East during the First Crusade (1096–1099) — was crowned King of Jerusalem. His reign, however, was brief, as the city was conquered just one year later, in 1187, by the Ayyubids under Salāḥ al-Dīn (531/1137–589/1193). Nevertheless, Guy I of Lusignan acquired Cyprus from the Knights Templar in 1192, and after his death his brother Amaury became King of Cyprus and, between 1197 and 1205, King of Jerusalem too.58 The arrival of the Ayyubids in Jerusalem might explain why we find the Christian monk John in Cyprus, under the protection of a Latin and Christian king, especially given the strict ties that already existed between Jerusalem under Frankish rule and the island.59 As regards the history of culture, it is important to note that in the twelfth to thirteenth centuries a Christian Arabic monk from Jerusalem was able to read Greek — or at least claimed to do so in his note. This detail seems consistent with what has been observed regarding the use of Greek by Christian monks in the Latin Kingdom of Jerusalem in Galadza’s study of two liturgical Greek manu­scripts copied in Palestine in the twelfth century.60 Manu­script BnF gr. 497, presents us with an opportunity to study Mediterranean medi­eval history and culture through a little more than three centuries thanks to its Greek colophon, dating back to the mid-tenth century, and a late twelfth-/ 57 In this respect, the two points above the letter qāf — as in waqafa, are particularly worth noting. 58 See Maddox and Sturm-Maddox, ‘Introduction’, p. 10. 59 ‘Greek manu­script production continued in Palestine under Frankish rule and there was a strong cul­ tural exchange between the churches and monasteries of Palestine and Cyprus’ Galadza 2017, p. 426. 60 See Galadza, ‘Greek Liturgy in Crusader Jerusalem’, with biblio­graphy.

s h ari ng t he w ri t t e n space

early thirteenth-century marginal note written in Arabic by a Christian monk from Jerusalem. This Greek codex, through its text and its notes, also in Arabic, mirrors a multilingual and multi­graphic society and reflects the unity of the Mediterranean history. A Mediterranean history that is made by all the outsiders that, crossing the water — as captives, merchants or pilgrims — became an agent of transformation on the different societies ‘introducing something of the culture of one continent into the outer edges, at least, of another’.61

Concluding Remarks Palaeo­graphy can be intended as a philo­logy of written signs, and it undoubtedly contributes to the history of culture in the broader sense: it not only adds to our knowledge of the history of written culture, but to the history of intellectual and religious culture. Palaeo­graphy draws on both historical and philo­logical evidence in order to contextualize ­graphic phenomena that would otherwise pass unnoticed, despite concurring with determined chrono­logical and linguistic facts. The palaeo­ graphic perspective — the various ­graphic layers of a written text, and the complex geo­graphical and historical path followed by travelling manu­scripts, their scribes, owners and readers — offers a unique opportunity to navigate epochs, g­ raphic systems and languages.

61 See Abulafia, ‘Mediterranean History’, p. 228.

417

41 8

a r i a n n a d’ot to n e

Biblio­graphy Manu­scripts Berlin, Berlin-Brandenburg Academy of Sciences and Humanities, BBAW/GCS, Akz.-Nr. 481/136-137 Paris, BnF, MS gr. 497 Secondary Works Abulafia, David, ‘Mediterranean History as Global History’, History and Theory, 50.2 (May 2011), 220−28 Almbladh, Karin, ‘Christianity and Judaism Under Islam in Medi­eval Iraq: Da’ūd alMuqammas and Sa‘d b. Mansūr Ibn Kammūna’, Nordisk Judaistik – Scandinavian Jewish Studies, 26.1–2 (2008), 19−28 Almbladh, Karin, ‘MS Uppsala O Nova 791 – a Rediscovered Manu­script of the Arabic Translation of and Commentary on the Song of Songs by Japheth ben Eli’, Orientalia Suecana, Supplement 61 (2012), 31−37 Atiya, Aziz S., ‘Apa’, in The Coptic Encyclopedia, vol. 1, ed. by Aziz S. Atiya (New York: Macmillan, 1991), pp. 152−53 Bagnall, Roger S., ‘Religious Conversion and Onomastic Change’, Bulletin of the American Society of Papyro­logists, 19 (1982), 105–24 Bassal, Ibrahim, ‘An Early Copy of a Christian Arabic Pentateuch Ms Sinai Arabic 2 and its Affinity to the Peshitta’, in Graeco-Latina et Orientalia: Studia in Honorem Angeli Urbani heptagenarii, ed. by Samir Khalil Samir and Juan Pedro Monferrer-Sala, Syro-Arabica, 2 (Cordoba-Beirut: Oriens Academic-CNERU-CEDRAC, 2013), pp. 13−33 Beit-Arié, Malachi, ‘Specimens of Mediaeval Hebrew Scripts: Vol. 1’, in Oriental and Yemenite Scripts/Ketav mizrahî u-ketav têmanî. Asûppôt ketavîm ivriyyîm miy-yemê habbênayîm, ( Jerusalem/Yerûshalayim: The Israel Academy of Sciences and Humanities/ ha-Aqademya hal-Le’ûmît hay-Yisre’elît le-Maddaîm, 1987) Beit-Arié, Malachi, ‘Stéréoptypies et individualités dans les écritures des copistes hébraïques du Moyen Âge’, in L’écriture, le cerveau, l’œil et la main, ed. by Colette Sirat, Jean Irigoin, and Emmanuel Poulle, Biblio­logia, 10 (Turnhout: Brepols, 1990), pp. 201−18 Beit-Arié, Malachi, ‘Transmissions des textes par scribes et copistes. Interférences inconscientes et critiques’, in Les problèmes posé par l’édition critiques des textes anciens et médiévaux, ed. by Jacqueline Hamesse, Textes, Études, Congrès, 13 (Louvain-LaNeuve: Publications de l’Institut d’Études médiévales 1992), pp. 173−96 Beit-Arié, Malachi, Unveiled Faces of Medi­eval Hebrew Books: The Evolution of Manu­script Production: Progression or Regression? ( Jerusalem: The Hebrew University Magnes Press, 2003) M. Beit-Arié, Malachi, ‘External and Internal Frontiers in Hebrew Manu­script Production’, in Frontiers in the Middle Ages: Proceedings of the Third European Congress of Medi­eval Studies (Jyväskylä, 10–14 June 2003), ed. by Outi Merisalo, Fédération Internationale des Instituts d’Études Médiévales – Textes et Études du Moyen Âge, 35 (Louvain-laNeuve: Brepols, 2006), pp. 399−407

s h ari ng t he w ri t t e n space

Malachi Beit-Arié, Colette Sirat, and Mordecai Glatzer, Codices Hebraicis Litteris Exarati Quo Tempore Scripti Fuerint Exhibentes. Tome I: Jusqu’à 1020, Monumenta Palaeo­ graphica Medii Aevi – Series Hebraica (Turnhout: Brepols, 1997) Binggeli, André, ‘Les trois David, copistes arabes de Palestine aux 9e–10e s.’, in Manu­scripta Graeca et Orientalia: Mélanges monastiques et patristiques en l’honneur de Paula Géhin, ed. by André Binggeli, Anne Boud’hors, and Matthieu Cassin (Louvain: Peeters, 2016), pp. 70−117 Bianconi, Daniele, ‘Paleografia: riflessioni su concetto e ruolo’, in Storia della scrittura e altre storie, ed. by Daniele Bianconi, Bollettino dei Classici, Supplement 29 (Roma: Accademia Nazionale dei Lincei, 2014), pp. 7–29 Bianconi, Daniele, ‘Palaeo­graphy: Introduction’, in Comparative Oriental Manu­script Studies: An Introduction, ed. by Alessandro Bausi and others (Hamburg: COMSt, 2015), pp. 267−70 Brown, Michelle P., ‘The Bridge in the Desert: Towards Establishing an Historical Context from the Newly Discovered Latin Manu­scripts of St Catherine’s Sinai’, in Palaeo­graphy Between East and West: Proceedings of the Seminars on Arabic Palaeo­graphy at Sapienza University of Rome, ed. by Arianna D’Ottone Rambach, Rivista degli Studi Orientali, Supplement ns 90 (Roma: Fabrizio Serra, 2018), pp. 73−98 Butbul, Sagit, ‘Translation in Contact: Early Judaeo-Arabic and Syriac biblical Translations’, in Beyond Religious Borders: Interaction and Intellectual Exchange in the Medi­eval Islamic World, ed. by David M. Freidenreich and Miriam Goldstein (Philadelphia: University of Pennsylavnia, 2012), pp. 57−64 Cavallo, Guglielmo, ‘Funzione e strutture della maiuscola greca tra i secoli VIII-XI’, in La paléo­graphie grecque et byzantine. Paris 21–25 octobre 1974, ed. by Jean Glénisson, Jacques Bompaire, and Jean Irigoin, Colloques internationaux du Centre National de la Recherche Scientifique, 559 (Paris: CNRS, 1977), pp. 95–137 Crisci, Edoardo, ‘La maiuscola ogivale diritta. Origini, tipo­logie, dislocazioni’, Scrittura e Civiltà, 9 (1985), 103–45 D’Aiuto, Francesco, ‘“Graeca” in codici orientali della Biblioteca Vaticana (con i resti di un manoscritto delle commedie di Menandro)’, in Tra Oriente e Occidente. Scritture e libri greci fra le regioni orientali di Bisanzio e l’Italia, ed. by Lidia Perria, Testi e Studi bizantino-neoellenici, 14 (Roma: Dipartimento di filo­logia greca e latina, Sezione bizantino-neoellenica, Università di Roma La Sapienza, 2003), pp. 227−96 D’Aiuto, Francesco, ‘Un antico inno per la Resurrezione (con nuove testimonianze di “scrittura mista” d’area orientale)’, Rivista di studi bizantini e neoellenici, n.s. 45 (2008) [2009], 3–135 D’Aiuto, Francesco, and Donatella Bucca, ‘Some Greek Hymno­graphic Fragments from Damascus: Scripts Types and Texts’, in The Damascus Fragments: Towards a History of the Qubbat al-khazna Corpus of Manu­scripts and Documents, ed. by Arianna D’Ottone Rambach, Konrad Hirschler, and Ronny Vollandt, Beiruter Texte und Studien, 140 (Beirut: Orient-Institut Beirut, 2020), pp. 291–320 D’Ottone, Arianna, ‘Al-khaṭṭ al-maghribī et le fragment bilingue latin-arabe Vat.lat. 12900: Quelques observations’, Les Rencontres du Centre Jacques Berque, 5 (2013), 7–18 D’Ottone, Arianna, ‘The Pearl and the Ruby: Scribal Dicta and Other Metatextual Notes in Yemeni Mediaeval Manu­scripts’, in The Yemeni Manu­script Tradition, ed. by David

419

4 20

a r i a n n a d’ot to n e

Holleberg, Christian Rauch, and Sabine Schmidtke, Islamic Manu­scripts and Books (Leiden: Brill, 2015), pp. 82−110, pl. 4.1−4.3 D’Ottone Rambach, Arianna, ‘Lucifer and Arabic Palaeo­graphy: A Contribution on the Oriental Glosses of the Bible of Cava dei Tirreni’, in Palaeo­graphy Between East and West: Proceedings of the Seminars on Arabic Palaeo­graphy at Sapienza University of Rome, ed. by Arianna D’Ottone Rambach, Rivista degli Studi Orientali, Supplement n.s. 90 (Pisa: Fabrizio Serra, 2018), pp. 123−39 D’Ottone Rambach, Arianna, ‘From Monte Cassino Abbey to St Catherine’s Monastery on Mount Sinai, and Back: The Journey of a Monk and the Encounter of Graphic Cultures’, in Les mobilités monastiques en Orient et en Occident de l’Antiquité tardive au Moyen Âge (iv–xv siècle), ed. by Olivier Delouis, Maria Mossakowska-Gaubert, and Annick Peters-Custot, Collection de l’École française de Rome, 558 (Rome: École française de Rome, 2019), pp. 391−408 D’Ottone Rambach, Arianna, and Dario Internullo, ‘One Script for Two Languages: Latin and Arabic in an Early Allo­graphic Papyrus’, in Palaeo­graphy Between East and West: Proceedings of the Seminars on Arabic Palaeo­graphy at Sapienza University of Rome, ed. by Arianna D’Ottone Rambach, Rivista degli Studi Orientali, Supplement n.s. 90 (Pisa: Fabrizio Serra, 2018), pp. 53−72 D’Ottone Rambach, Arianna, Konrad Hirschler, and Ronny Vollandt, eds., The Damascus Fragments: Towards a History of the Qubbat al-khazna Corpus of Manu­scripts and Documents, Beiruter Texte und Studien, 140 (Beirut: Orient-Institut Beirut, 2020) Engel, Edna, ‘Script, History and Development’, in Encyclopedia of Hebrew Language and Linguistics, vol. 3, ed. by Geoffrey Khan (Leiden: Brill, 2013), pp. 485−502 Fleischer, Heinrich L., ‘Zur Geschichte der arabischen Schrift’, Zeitschrift der Deutschen Morgenlänischen Gesellschaft, 18 (1864), pp. 288–91 Franklin, Arnold E., ‘Untidy History: Reassessing Communal Boundaries in Light of the Cairo Genizah Documents’, in Age of Transition: Byzantine Culture in the Islamic World, ed. by Helen C. Evans (New York: The Metropolitan Museum of Art, 2015) pp. 54–69 Frenkel, Miriam, ‘Education: 1. Medi­eval Period’, in Encyclopedia of Jews in the Islamic World, vol. 2, ed. by Norman A. Stillman (Leiden: Brill, 2010), pp. 116−19 Gacek, Adam, Arabic Manu­scripts – A Vademecum for Readers, Handbook of Oriental Studies, Section 1, The Near and Middle East Ancient Near East, 98 (Leiden: Brill, 2009) Galadza, Daniel, ‘Greek Liturgy in Crusader Jerusalem: Witnesses of Liturgical Life at the Holy Sepulchre and St Sabas Lavra’, Journal of Medi­eval History, 43.4 (2017), 421–37 George, Alain, ‘Le palimpseste Lewis-Mingana de Cambridge, témoin ancien de l’histoire du Coran’, Compte-rendus des séances de l’Académie des Inscriptions et Belles-Lettres, 155.1 (2011), 377–429 Ginzburg, Carlo, ‘Spie. Radici di un paradigma indiziario’, in Crisi della ragione. Nuovi modelli nel rapporto tra sapere e attività umane, ed. by Aldo Gargani (Torino: Einaudi, 1979), pp. 57−106 Görke, Andreas, and Konrad Hirschler, eds., Manu­script Notes as Documentary Sources, Beiruter Texte und Studien, 129 (Beirut: Orient-Institut Beirut, 2011) Griffith, Sidney H., ‘Stephen of Ramlah and the Christian Kerygma in Arabic in 9th Century Palestine’, Journal of Ecclesiastical History, 36 (1985), 23–45 (= Sidney H. Griffith, Arabic Christianity, VII)

s h ari ng t he w ri t t e n space

Griffith, Sidney H., ‘Greek into Arabic: Life and Letters in the Monasteries of Palestine in the Ninth Century: The Example of the Summa Theo­logiae Arabica’, Byzantion, 56 (1986), 117–38 (= Sidney H. Griffith, Arabic Christianity, VII) Griffith, Sidney H., ‘The Monks of Palestine and the Growth of Christian Literature in Arabic’, The Muslim World, 78 (1988), 1–28 Griffith, Sidney H., ‘Anthony David of Baghdad, Scribe and Monk of Mar Sabas: Arabic in the Monasteries of Palestine’, Church History, 58 (1989), 7–19 Griffith, Sidney H., Arabic Christianity in the Monasteries of Ninth-Century Palestine, Variorum Collected Studies Series, 380 (Aldershot: Ashgate, 1992 [reprint: 2003]) Griffith, Sidney H., ‘From Aramaic to Arabic: The Languages of the Monasteries of Palestine in the Byzantine and Early Islamic Periods’, Dumbarton Oaks Papers, 51 (1997), 11–31 Guillamont, Antoine, ‘Antony of Egypt, Saint’, in The Coptic Encyclopedia, vol. 1, ed. by Aziz S. Atiya (New York: Macmillan, 1991), pp. 149−51 Harlfinger, Dieter, ‘Beispiele der maiuscola ogivalis inclinata vom Sinai und aus Damaskus’, in Alethes philia: Studi in onore di Giancarlo Prato, vol. 2, ed. by Marco D’Agostino and Paola Degni (Spoleto: Fondazione Centro italiano di studi sull’Alto Medioevo, 2010), pp. 461–77 Heidemann, Stefan, ‘Calli­graphy on Islamic Coins’, in The Aura of Alif: The Art of Writing in Islam, ed. by Jürgen W. Frembgen (London: Prestel, 2010), pp. 161−71 Hjälm, Miriam L., ‘From Palestine to Damascus to Berlin: Early Christian Arabic Texts from the Qubbat al-khazna in the Violet Collection’, in The Damascus Fragments: Towards a History of the Qubbat al-khazna Corpus of Manu­scripts and Documents, ed. by Arianna D’Ottone Rambach, Konrad Hirschler, and Ronny Vollandt, Beiruter Texte und Studien, 140 (Beirut: Orient-Institut Beirut, 2020), pp. 245–64 Khan, Geoffrey ‘On the Question of Script in Medi­eval Karaite Manu­scripts: New Evidence from the Cairo Genizah’, Bulletin of John Rylands University Library of Manchester, 75 (1993), 133–41 Kolia-Dermitzaki, Athina, ‘Michael the Synkellos’, in Christian-Muslim Relations: A Biblio­ graphical History. Volume 1 (600–900), ed. by David Thomas and Barbara Roggema (Leiden: Brill, 2009), pp. 627−32 Lucà, Santo, ‘Interferenze linguistiche greco-latine a Grottaferrata tra XI e XII secolo’, in Studi paleografici e papiro­logici in ricordo di Paolo Radiciotti, ed. by Mario Capasso and Mario De Nonno, Papyro­logica Lupiensia, Supplement 24 (Lecce: Pensa MultiMedia, 2015), pp. 295−327 and plates 1–4 Maddox, David, and Sara Sturm-Maddox, ‘Introduction’, in Jean d’Arras, Melusine; or the Noble History of Lusignan, ed. and trans. by David Maddox and Sara Sturm-Maddox (University Park: Pennsylvania State University Press, 2012), pp. 1−16 Mandalà, Giuseppe, ‘Written Culture(s) in Islamic Sicily (827–1091)’, in Multilingual and Multi­graphic Documents and Manu­scripts of East and West, ed. by Giuseppe Mandalà and Immaculada Pérez Martín, Perspectives on Linguistics and Ancient Languages, 5 (Piscataway: Gorgia Press, 2018), pp. 123−147 and pl. 1–11 Marcos Marín, Francisco, ‘Vat.Lat. 12900. El fragmento Latino Visigótico de la Epístola a los Gálatas y su versión árabe andalusí’, in Vir Bonus Dicendi Peritus. Homenaje al profesor Miguel Ángel Garrido Gallardo, ed. by Luis Alburquerque-García, José Luis

421

422

a r i a n n a d’ot to n e

García Barrientos, Antonio Garrido Domínguez, and Ana Suárez Miramón (Madrid: Consejo Superior de Investigaciones Científicas, 2019), pp. 980−88 Marichal, Robert, Charles Samaran and Julien Dufour, ‘Paléo­graphie latine et française’, École pratique des hautes études. 4e section, Sciences historiques et philo­logiques. Annuaire 1973–74 (1974), 407–22 Monferrer-Sala, Juan Pedro, ‘Jewish and Christian Bible Encounters in Arabic? On a Fragment of GN 27:31–34 from the Cairo Genizah in Hebrew and Judaeo-Arabic (T-S AS 70.101)’, in Between the Cross and the Crescent: Studies in Honor of Samir Khalil Samir, S.J. on the Occasion of His Eightieth Birthday, ed. by Željko S.J. Paša, Orientalia Christiana Analecta, 304 (Roma: Pontificio Istituto Orientale, 2018), pp. 85−101 Monferrer-Sala, Juan Pedro, ‘The Fragmentary Ninth/Tenth Century Andalusi Arabic Translation of the Epistle to the Galatians Revisited (Vat.lat. 12900, olim Seguntinus 150 bc Sigüenza)’, Intellectual History of the Islamicate World, 7.1 (2019), 125–91 Monferrer-Sala, Juan Pedro and Ulisse Cecini, ‘Once Again on Arabic “alkaufeit” (Alb. Ind. 23,14): Between Polemics and Inculturation’, Mittellateinisches Jahrbuch, 49.2 (2014), 201–10 Olszowy-Schlanger, Judith, ‘Crossing Palaeo­graphical Borders: Bi-Alphabetical Scribes and the Development of Hebrew Script – The Case of the Maghrebi Cursive’, in Studies in Semitic Linguistic and Manu­scripts: A Liber Discipulorum in Honour of Professor Geoffrey Khan, ed. by Nadia Vidro, Ronny Vollandt, Esther-Miriam Wagner, and Judith Olszowy-Schangler, Studia Semitica Upsaliensia, 30 (Uppsala: Acta Universitatis Upsaliensis, 2018), pp. 299−319 Papaconstantinou, Arietta, ‘‘What Remains Behind’: Hellenism and Romanitas in Christian Egypt after the Arab Conquest’, in From Hellenism to Islam: Cultural and Linguistic Change in the Roman Near East, ed. by Hannah M. Cotton, Robert G. Hoyland, Jonathan J. Price, and David J. Wasserstein (Cambridge: Cambridge University Press, 2009), pp. 447−66 Polliack, Meira, The Karaite Tradition of Arabic Bible Translation: A Linguistic and Exegetical Study of Karaite Translations of the Pentateuch from the Tenth and Eleventh Centuries c.e., Études sur le judaïsme médiéval, 17 (Leiden: Brill, 1997) Radiciotti, Paolo, ‘Il problema del digrafismo nei rapporti fra scrittura latina e greca nel medioevo’, Nea Romi. Rivista di ricerche bizantinistiche, 3 (2006), 5–55 Radiciotti, Paolo, ‘Per Knut Kleve. Riflessioni sulla paleografia’, Papyro­logica lupiensia, 17 (2008), 53–60 Radiciotti, Paolo, ‘Scrittura latina per ostraka neopunici. Osservazioni paleografiche per interpretare un fenomeno culturale’, Studi di Egitto­logia e Papiro­logia, 6 (2009), 115–20 Radiciotti, Paolo, ‘Scrivere e leggere il greco fuori dai confini temporali del mondo antico: il Medioevo latino’, in Leggere greco e latino fuori dai confini nel mondo antico: atti del I Congresso nazionale dell’Associazione italiana di cultura classica, Lecce, 10–11 maggio 2008, ed. by Mario Capasso, I quaderni di Atene e Roma, 1 (Lecce: Pensa Multimedia, 2010) 175–92 Radiciotti, Paolo, ‘Digrafismo nei papiri latini’, in Bilinguisme et di­graphisme dans le monde gréco-romain: l’apport des papyrus latins. Actes de la table ronde internationale (Liège, 12–13 mai 2011), ed. by Marie-Hélène Marganne and Bruno Rochette (Liège: Presses Universitaires de Liège, 2013), pp. 57−69

s h ari ng t he w ri t t e n space

Reif, Stefan C., Published Material from the Cambridge Genizah Collections: A Biblio­graphy 1896–1980 (Cambridge: Cambridge University Press, 1988) Sasson, Ilana, ‘Psalms: Arabic Translations’, in The Hebrew Bible, vol. 1c: Writings, ed. by Lange Armin and Emanuel Tov (Leiden: Brill, 2017), pp. 110−14 Schmidt, Andrea Barbara, ‘Syriac Palaeo­graphy’, in Comparative Oriental Manu­script Studies: An Introduction, ed. by Alessandro Bausi and others (Hamburg: COMSt, 2015), pp. 316−20 Sims-Williams, Nicholas, ‘A New Fragment of the Book of Psalms in Sogdian’, in Bibel, Byzanz und Christlicher Orient: Festchrift für Stephen Gerö zum 65. Gebutsatg, ed. by Dmitrij Bumazhnov, Emmanouela Grypeou, Timothy B. Sailors, and Alexander Toepel, Orientalia Lovaniensia Analecta, 187 (Leuven: Peeters, 2011), pp. 461−65 Steiner, Richard C., A Biblical Translation in the Making: The Evolution and Impact of Saadia Gaon’s Tafsīr (Cambridge, MA: Harvard University Center for Jewish Studies, 2010) Treiger, Alexander, ‘Syro-Arabic Translations in Abbasid Palestine: The Case of John of Apamea’s Letter on Stillness (Sinai Ar. 549)’, Parole de l’Orient, 39 (2014), 79–131 Treiger, Alexander, ‘A Newly Discovered Karaite Arabic Translation of Genesis and Exodus (Undertext of the Palimpsest Sinai Gr. 930)’, in Senses of Scriptures, Treasures of Tradition: The Bible in Arabic among Jews, Christian and Muslims, ed. by Miriam L. Hjälm (Leiden: Brill, 2017), pp. 529–46 Versace, Pietro, I marginalia del ‘Codex Vaticanus’, Studi e Testi, 528 (Città del Vaticano: Biblioteca Apostolica Vaticana, 2018) Vollandt, Ronny, Arabic Versions of the Pentateuch: A Comparative Study of Jewish, Christian, and Muslim Sources, Biblia Arabica, 2 (Leiden: Brill, 2015) Vollandt, Ronny, ‘Translations as Linguistic Commentaries? On the Interpretative Dimension of Early Bible Translations in Judaeo-Arabic’, Philo­logical Encounters, 3 (2018), 253–74 Wagner, Esther-Miriam, ‘The Socio-Linguistic of Judaeo-Arabic Mercantile Writing’, in Merchants of Innovations, ed. by Esther-Miriam Wagner, Bettina Beinhoff, and Ben Outhwaite (Boston: de Gruyter, 2017), pp. 68−86 Wasserstein, David J., ‘Lāwī b. Ismā‘īl b. Rabī‘ b. Sulaymān: An Unnoticed Jewish Convert to Islam in Fifth/Eleventh Century Al-Andalus’, Der Islam, 91.2 (2014), 341–59 Zewi, Tamar, The Samaritan Version of Saadiyah Gaon’s Translation of the Pentateuch. Critical Edition and Study of the MS London BL OR 7562 and Related MSS (BA, 3), Biblia Arabica, 3 (Leiden: Brill, 2015)

423

Fred Astren

ʿAbbāsid Book Culture and Ninth-Century

Jewish Sectarianism

In the tenth century, Rav Sherira bar Hanina Gaon (906–1006) published a circular letter written in Aramaic that lays out the generations of rabbis and their chrono­logies. In it he reports: ‫] ומורא רב יצחק גאון והוא שהיה בפרוז שבור עת שכבשה‬...[ ‫] אותו הﬠת‬...[ ‫ואמרי דהוה‬ ‫עלי בן אבוטאלב ויצא מר יצחק מן פרוז שבור לקראתו והקביל פניו בסבר פנים יפות והיה‬ .‫בפרוז שבור היום ההוא צ׳ אלף מישראל וקבלם עלי בן אביטאל׳ בסבר פנים יפות‬

And they say that […] during this period […] Mar Rabbi Yiṣḥak Gaon was gaon in Fīrūz-Shāpūr. It was he who was in Fīrūz-Shāpūr when ʿAlī b. Abī Ṭālib conquered the city. Mar Yiṣḥak of Fīrūz-Shāpūr went out to [meet] him and welcomed him with great friendliness. At that time, there were in Fīrūz-Shāpūr ninety thousand Jews who were received by ʿAlī b. Abī Ṭālib with great friendliness.1 Such conquest narratives need to be approached by historians with great care, since they often construct retrospective historical world views designed to support agendas contemporaneous with the narrative’s composition rather than with the period they purport to describe.2 Whether this event actually occurred or not, the report acts as a signifier for an important geo­graphic and demo­graphic feature of Jewish history — that the heartland of the greatest Jewish population in Late Antiquity and the early Middle Ages was largely coterminous with the heartland of the ʿAbbāsid caliphate.3 Sherira Gaon, the author of the circular letter, was the gaon, or head, of the yeshiva (rabbinical academy) of Pumbeditha, which was once located close to





1 Sherira Gaon, Iggeret, pp. 100–01. The translation is adapted from Rabinowich, The Iggeres of Rav Sherira Gaon, p. 125. If this is a reliable report, it is more likely based on ʿAlī’s acceptance of allegiance from the city at the time of his accession to the caliphate in 656 or while seeking to consolidate his power in Iraq (either before or after ʿUthmān’s assassination), rather than on the city’s conquest by Muslim forces years earlier. 2 See Astren, ‘Re-reading the Muslim Sources’. 3 See Gafni, ‘The Political, Social, and Economic History of Babylonian Jewry’, p. 805, referring to scholarly estimates of a Jewish population of one million. This figure seems high.

Fred Astren is Professor Emeritus of Jewish Studies at San Francisco State University. He works on the interchange of late antique and early medi­eval Jewish history. Navigating Language in the Early Islamic World: Multilingualism and Language Change in the First Cen­turies of Islam, ed. by Antoine Borrut, Manuela Ceballos, and Alison M. Vacca, ISMAR 2 (Turnhout: Brepols, 2024) pp. 425–465 10.1484/M.ISMAR-EB.5.134634

FHG

426

f r e d a s tr e n

modern-day al-Fallūja. Nearby Fīrūz-Shāpūr corresponds to al-Anbār. Whereas the Arabs named the land al-ʿIrāq, Jews used the biblical name Bavel (Babylonia), harking back to the origins of the community in the biblical exile of the Judeans under Nebuchadnezzar to the heartland of the Neo-Babylonian Empire in the early sixth century bce. Sherira’s narrative, and ones like it, operate at multiple levels. For the purposes of introducing this study, the narrative confirms the deep local Jewish past — Jewish indigeneity — in this region and simultaneously identifies Jews as participants in an originary moment of Muslim history.4 Furthermore, its telling in a late tenth-century text locates Jews in Iraq at the time of the telling.5 As a publication, Sherira’s letter is a kind of conceit. On the one hand, it is written in the time-honoured form of a rabbinic responsum, that is, as an answer to a scholarly query, in this case from Qayrawān in North Africa. As a rule, Sherira’s responsa were written in the language of the original question, whether Hebrew, Aramaic, or Arabic; in this case he writes in Aramaic, whose use among Jews in the late tenth century is associated with the rabbinic establishment. On the other hand, the letter bears many features of a tenth-century publication. As a composition, it is not merely a legal responsum but rather an extended treatise on the office of the gaonate and rabbinic tradition. It was intended for a specific educated readership (in this case, of rabbinic Jews), enjoyed wide distribution, and was probably written on paper. Like Sherira’s other writings, its composition and dissemination can be understood as part of the massive literary and book revolution that was in full bloom at the time6 but in which the rabbis were late participants. This chapter is conceived as a preliminary effort to locate Jews in the new book culture of the early ʿAbbāsid era. Between the time of ʿAlī’s reported visit to Fīrūz-Shāpūr and the era of Sherira — the first three centuries of Islam — massive transformations reordered the Near East and other lands under Muslim rule. It is axiomatic to this study that Jews were participants in the interlinked remaking of the Near East and the making of Islamdom in those centuries, and that they could be subjects of and participants in general historical change, even if in the late tenth century such claims had to be narrativized by Sherira. Accordingly, it will be argued in the following pages that the participation of Jews in urbanization, immigration, and economic reorganization created conditions that were ripe for the florescence of sectarianism among Jews (and Muslims), in part through religious responses to new urban situations. These massive historical trends, which reordered Islamic societies in this period (especially in Iraq and Baghdad), set the stage on which the introduction of paper made possible the unprecedented growth and impact of ʿAbbāsid book culture. These trends also shaped the social history that undergirds ninth-century Jewish sectarianism.



4 On such claims, see Astren, ‘The Gibeonite Gambit in the Middle Ages’. 5 At the time of Sherira, the caliphate was under the control of the Iranian Shīʿite Būyids, for whom ʿAlī b. Abī Ṭālib, as the central figure of Shīʿite Islam, was a key signifier. On this narrative, see Gross, ‘When the Jews Greeted Ali’. Compare to Iranian memory of the Muslim conquests in Savant, The New Muslims of Post-Conquest Iran, especially pp. 90–129. 6 On Sherira, see Brody, ‘The Epistle of Sherira Gaon’.

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

Urbanization and Economy: The Setting for a New Book Culture and New Jewish Heterodoxy By the time of Sherira, the centrality of Pumbeditha’s geo­graphical location was meaningless, since the two rabbinical academies of Pumbeditha and Sura (located on the Euphrates near al-Ḥīra) had moved to Baghdad, probably by the end of the ninth century.7 Founded as Madīnat al-Salām (‘City of peace’) by the ʿAbbāsid caliph al-Manṣūr (r. 754–75), Baghdad was located on the left bank of the Tigris near to where the courses of the Euphrates and the Tigris come closest together in central Mesopotamia, not far from the sites of ancient Babylon and the former Sasanian capital of Ctesiphon-Seleucia, known in Arabic as ‘the Cities’ (al-Madāʾin). The Persian imperial capital’s transriverine district of Veh-Ardashir, known to Jews as Mahoza (Māḥōzē in Syriac), was the site of residence of both the catholicos of the Church of the East and the Jewish exilarch, each of whom led significant ethno-religious communities in the Sasanian empire.8 The exilarch’s claimed descent from the biblical King David made Mahoza the ‘geo­graphical center of authority for Babylonian Jews’ in the late Talmudic (late Sasanian) period,9 while the Babylonian Talmud reports on Mahoza’s urban character, wealth, and mercantile connections.10 This centrally located subregion was the hub of Iraq’s agricultural core, known in Arabic as the Sawād. It was also the core of the area of Jewish settlement in Mesopotamia, with the ancient rabbinic centres of Pumbeditha, Sura, and Nehardea located on the Euphrates side of this zone of convergence and Mahoza on the Tigris side.11 To the rabbis, this was the centre of the Area of Pure Lineage, a geohistorical construct that elevated the status and religious standing of the Jews of Babylonia over those of the Land of Israel.12 Baghdad’s eclipse of Ctesiphon-Seleucia is reflected in the exilarch’s likely move to the new city soon after its foundation, and it is paralleled by the transfer of Catholicos Henanishoʿ II to the city in 775. Undoubtedly, both communal leaders relocated to Baghdad on account of their formal connections to the caliphal court.13 Baghdad’s centrality is illustrated by



7 The academy of Pumbeditha moved to Baghdad during the geonate of Hayya ben David, 890–97. See Brody, The Geonim of Babylonia, p. 36. Brody notes that ‘it is difficult to assess the implications of this relocation of the Geonic academies’. This topic is worthy of further study. 8 On the catholicos of the Church of the East in Baghdad, see Allard, ‘Les Chrétiens à Baġdād’. See also Fiey, ‘Topo­graphie chrétienne de Mahozé’, and Wood, The Chronicle of Seert, pp. 221–56. 9 Boyarin, A Traveling Homeland, p. 73. On the exilarchate, see Herman, A Prince without a Kingdom, pp. 144–58, and Gil, ‘The Exilarchate’. 10 On Jewish Mahoza and its environs, see Simpson, ‘The Land behind Ctesiphon’; Oppenheimer, Baby­ lonia Judaica in the Talmudic Period, pp. 179–235; and Eshel, Jewish Settlements in Babylonia, pp. 141–44. 11 On Jewish urbanism in Late Antiquity, see Ahuvia, ‘Jewish Towns and Neighborhoods in Roman Palestine and Persian Babylonia’. 12 See Oppenheimer, ‘Purity of Lineage in Talmudic Babylonia’, and Herman, ‘Babylonia of Pure Lineage’. 13 This is the opinion of Mann in ‘The Responsa of the Babylonian Geonim, Part 1’, pp. 465–66. Sherira reports that some time before 828 the geonim of Pumbeditha (there were two so named at the time) ‘came to Baghdad and it happened that they were at the synagogue of bar Nashla, where there was a

427

4 28

f r e d a s tr e n

the appointment of a Jacobite bishop to the city soon after its founding and by the success of the Church of the East in thwarting the establishment of a Melkite bishopric there.14 All these moves, including the establishment of Christian monasteries in Baghdad, preserved long-term historical continuity in terms of regional relationships between subordinated indigenous religious communities and imperial dominion. The centres of power barely moved. In concert with the Muslim conquests in the seventh century, urban development and urbanization played an important role in caliphal policy and contributed to changes in social structure. From the beginning, as a strategic and economic corollary to the conquests themselves, new cities, called amṣār (sing. miṣr), were founded as administrative centres and military bases and were later imagined as sites for segregating Arab tribesmen from the local environment. Kūfa and Baṣra in Iraq (established in the 630s) and Fusṭāṭ in Egypt (641) were such cities.15 A few decades later, the Umayyad founding of Wāsiṭ in Iraq — as a site whose purpose was to house Syrian troops in Iraq and thereby de-emphasize the military, political, and economic power of its two Iraqi predecessors — reveals how quickly these new cities gained prominence in a landscape shaped by the oldest urban configurations in human history. Fuelled by immigration, these cities became centres of economic opportunity, wealth, power, and prestige, as well as important sites of writing. This ascendancy of Muslim urban configurations not only extended Sasanian urbanization policies but also fits ancient Near Eastern historical patterns that link empire and urbanization.16 In fact, Kūfa and Baṣra were massive in comparison to Mesopotamian cities in the Sasanian period or for a long time afterwards.17 The foundation of Baghdad in 762 overshadowed all earlier Muslim urban establishments and became paradigmatic in both gross and subtle ways for subsequent Islamic urbanization. An imperial establishment like Wāsiṭ, Baghdad was designed as a utopian reflection of the dawla of the ʿAbbāsids — their revolution, dynasty, and revival.18 More than any other city, the colocation of the caliphal administration and the army served as a magnet, as had the amṣār, for immigration, economic development, and social reconfiguration.19 It is clear from rabbinic sources that Jews lived across the region in the early Islamic centuries. For example, two geonim of the middle and late eighth century

14 15 16 17 18 19

great gathering’ (Sherira Gaon, Iggeret, p. 110). Whether the gathering (kallah) was connected to the exilarchate or to the yeshiva cannot be determined. See Wood, The Imam of the Christians, p. 23, which cites Fiey, Chrétiens syriaques sous les Abbassides, pp. 129–30. See Kennedy, ‘How to Found an Islamic City’, and Denoix, ‘Founded Cities of the Arab World from the Seventh to the Eleventh Century’. On Sasanian urbanization, see Daryaee, Sasanian Persia, pp. 39–41. On Ctesiphon as a major urban centre, see Adams, Land behind Baghdad, pp. 70–74. On long-term ancient Near Eastern patterns of urbanization, see Altaweel and Squitieri, Revolutionizing a World, especially pp. 124–59. Except for possibly Ctesiphon-Seleucia. See Kennedy, ‘The Feeding of the Five Hundred Thousand’, p. 177. See Lassner, ‘The ‘Abbasid Dawla’. See Micheau, ‘Baghdad in the Abbasid Era’.

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

are reported by Sherira to hail from Baghdad and its environs.20 However, in Muslim sources there is only scattered evidence of Jewish inhabitation in the region’s urban centres. Nonetheless, the geo­graphy of Jews portrayed in the Babylonian Talmud can be viewed as an underlay to Jewish urban history in the early Islamic centuries, with many sites indicating continuity of settlement.21 Toponyms with the Arabic word Yahūd (‘Jew’) often indicate a Jewish presence at one time or another. In Kūfa there was the Street of the Jews (Shāriʿ al-Yahūd),22 while just outside of Baṣra was the ‘Canal of the Jew’ (Nahr al-Yahūdī).23 Baghdad had several Jewish toponyms, including Qanṭārat al-Yahūd (‘Bridge of the Jews’), which may have spanned the Kharkāyā Canal connecting Baghdad with the market suburb of al-Karkh,24 where one also found the Tomb of Joshua ben Jehozadak, high priest of the biblical Babylonian exile.25 The Jewish quarter of Baghdad was called Qaṭīʿat al-Yahūd or Dār al-Yahūd.26 And there was a Darb al-Yahūd (‘Street of the Jews’), mentioned by Yāqūt as a street that was home to some well-known Muslim scholars.27 Further afield, one gate in the walls of Aleppo was named Bāb al-Yahūd, while an area of Mosul was known as Maḥallat al-Yahūd.28 In fact, many cities had quarters known as Ḥarat al-Yahūd that were, for obvious or less clear reasons, associated with Jews. This kind of toponymy indicates the segmentation that characterized much of the early Islamic city, where quarters could be named not only for their inhabitants but for different occupations or the markets of particular goods (such as the sūq al-warrāqīn, or Market of the Stationers, on which more below). However, this toponymic segmentation should not be equated with segregation. There were no ghettos in the cities of early Islam, and by the advent of the caliphate rabbinic Judaism had adapted itself to the diasporic city.29 For example, the legal concept of the ʿeruv acted as a conceptual Jewish overlay of gentile urban space that

20 These are Natronai Kahana bar Amuna, who held office sometime in the 750s (‘He was from Baghdad’), and Yeshaya ha-Levi bar Abba, who held office in the years 790–96 (‘He was from klwady in the vicinity of Baghdad’). See Sherira Gaon, Iggeret, pp. 103 and 109, respectively. 21 See Eshel, Jewish Settlements in Babylonia during Talmudic Times, and Morony, Iraq after the Muslim Conquest, pp. 306–12. 22 See Gil, Jews in Islamic Countries in the Middle Ages, p. 511. Kūfa (ʿĀqūlā in Aramaic) was also the tradi­ tional site of the tomb of the biblical exiled Judean king Yehoiachin, a signifier of Jewish indigeneity. 23 See al-Ṭabarī, Taʾrīkh, iii, p. 1900, translated in al-Ṭabarī, The History of al-Ṭabarī, xxxvi, p. 175. 24 Le Strange, Baghdad during the Abbasid Caliphate, p. 150, and Jacob Lassner, The Topo­graphy of Baghdad in the Early Middle Ages, p. 101. However, Moshe Gil claims the bridge spanned the Tigris. See Gil, Jews in Islamic Countries in the Middle Ages, p. 493. 25 The tomb is another signifier of Jewish indigeneity. On rabbinic Jewish attitudes towards Babylonia and its ‘biblicization’, see Gafni, Land, Center, and Diaspora. 26 For Qaṭīʿat al-Yahūd, see Le Strange, Baghdad during the Abbasid Caliphate, p. 150. For Dār al-Yahūd, see Yāqūt, Muʿjam al-buldān, iv, p. 1045. 27 See Gil, Jews in Islamic Countries in the Middle Ages, pp. 415, 493. 28 For Aleppo, see al-Maqdisī (al-Muqaddasī), Description of Syria, including Palestine, p. 13, and Le Strange, Palestine under the Moslems, pp. 362, 365–66, which cites Nāṣir-i Khusraw and Yāqūt. For Mosul, see the translation of al-Balādhurī, Futūḥ al-buldān, ii, p. 32. 29 See Stillman, ‘The Jew in the Medi­eval Islamic City’, p. 7, which cites the Babylonian Talmud, Sanhedrin 17b.

429

430

f r e d a s tr e n

permitted a certain amount of leniency in Sabbath observance.30 In general, the quarters of these cities lacked homogeneity, usually densely housing a variety of people who rubbed shoulders with one another every day. The claim to indigeneity embedded in Sherira Gaon’s letter is supported by Jewish toponyms and reflected in day-to-day life. Economic growth drove much of the intensification of urbanization in Baghdad. One factor was caliphal activism in urban development for purposes of generating income and gaining prestige. Capital projects, large and small, were implemented to build markets, funduqs, bridges, ports, and canals, as well as mosques and even towns and cities. In addition, the operation of private palatial establishments of important officials, courtiers, and members of the caliphal family produced employment and economic activity.31 A much more important driver of the economy was the new salaried class, whose existence was a consequence of empire and its administration. The Muslim conquests of the seventh and eighth centuries established preliminary arrangements for the conquered peoples that eventually were rationalized to a great degree by classifying non-Muslims as dhimmīs (protected peoples) who were subject to a tax regime that generated a great deal of income for the empire and a great deal of record-keeping on the part of the bureaucracy. The government agency that distributed the accumulated wealth, known as the dīwān, kept careful records and gave an annual (or sometimes biannual) payment to registered Muslims, usually fighters and their dependents, but later these payments became salaries for the regular army. The recipients of these payments became urbanized very soon after the conquests, creating a very large urban class whose disposable income fed great demand for both subsistence and luxury goods of all kinds.32 Likewise, the bureaucrats and other government officials who kept tax and administration records received salaries, as did soldiers, which in turn prompted demand for goods and services that gave rise to a kind of middle class with its own spending capacity.33 When the regime was stable and the bureaucracy was working smoothly, money was all around. Hugh Kennedy identifies the demand generated by the circulation of so much money as one of four factors that together constituted a ‘virtuous circle’ of economic boom and mammoth urbanization that surpassed the cities of antiquity and whose mass was not attained again in the region until modernity.34 This circle was based on the high potential for agricultural productivity in Iraq, whose necessary hydraulic infrastructure was supported by a legal framework for land tenure that encouraged investment.35 And unlike in other urban hinterlands, produce and other goods 30 See Fonrobert, ‘From Separatism to Urbanism’, and Fonrobert, ‘The Political Symbolism of the Eruv’. 31 See Micheau, ‘Baghdad in the Abbasid Era’, p. 225, who writes that al-Yaʿqūbī describes a ‘willed urbanism’ in Baghdad. See Scheiner and Janos, ‘Baghdād’. 32 Kennedy, ‘Military Pay and the Economy of the Early Islamic State’. 33 See S. D. Goitein’s reflections on the middle classes in medi­eval Islam in ‘The Rise of the Middle Eastern Bourgeoisie in Early Islamic Times’ and ‘The Mentality of the Middle Class in Medi­eval Islam’. 34 Kennedy, ‘The Feeding of the Five Hundred Thousand’. 35 See Watson, Agricultural Innovation, and a review and critique of Watson’s much-discussed ‘agricultural revolution’ in Nol, Settlement and Urbanization, pp. 26–30.

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

could be transported efficiently using the rivers and canals to and from Baghdad and other cities. The historian al-Ṭabarī (839–923) describes the system in a speech purportedly given to the caliph al-Manṣūr before the foundation of the city by the dihqān (overlord) of what was then the village of Baghdad. ‫يا أمير المؤمنين على الصراة تجيئك الميرة في السفن من المغرب في الفرات وتجيئك طرائف مصر‬ ‫ وتجيئك الميرة في السفن من الصين والهند والبصرة وواسط في دجلة وتجيئك الميرة من أرمينية‬۰‫والشأم‬ ۰‫وما اتصل بها في تأمرا حتى تصل إلى الزاب وتجيئك الميرة من الروم وآمد والجزيرة والموصل في دجلة‬

O Commander of the Faithful, you would be on the Ṣarāt Canal, permitting provisions to come to you along the Euphrates in ships from the west, as well as the choice products of Egypt and Syria. Supplies would also come to you along the Tigris in ships from China, India, al-Baṣra, and Wāsiṭ. Finally, stores would reach you from Armenia and those adjacent areas via the Tāmarrā Canal connecting to the Zāb, as well as from Byzantium, Āmid, the Jazīra, and Mosul down the Tigris.36 These water routes and their cities lay in the heart of Jewish Iraq. For example, west of Baghdad, the ancient Ṣarāt Canal (Nahr Ṣarāt) linked to the Nahr ʿĪsā, the northernmost of the major canals that connected the Euphrates to the Tigris. It also connected the northern axis of Sasanian Jewish centres at Fīrūz-Shāpūr (al-Anbār) and Pumbeditha (al-Fallūja) to al-Karkh, Baghdad’s commercial southern suburb, where many Jews, Armenians, and Shīʿites lived.37 It is not unwarranted to suggest that elements of the vibrant Jewish economy of Sasanian Mahoza, where ‘wealthy, cosmopolitan, canny’ Jews had pursued their fortunes, were easily relocated to the new megacity.38 More broadly, Jews and all other local populations benefited from economic development in the late eighth and ninth centuries, participating in a system that worked mostly with efficiency until the tenth century, when new political and economic factors reshaped the Near East. In Jewish history, this period marks an era during which rural populations may have declined. That is, as Jews participated in massive eighth- and ninth-century immigration to cities (especially Baghdad), the rural agrarian settings often assumed or referred to in the Babylonian Talmud (redacted in the sixth or possibly early seventh century) were being abandoned or transformed through conversion to Islam. Around 787, the Iraqi rabbinic leadership (the two geonim and the exilarch) issued a major legal ruling known as a takkanah that permitted the debts of orphans to be imposed on movable property, thereby expanding the purview of a Talmudic law that required such debts be paid only from real estate.39 The oft-cited responsum

36 Al-Ṭabarī, Taʾrīkh, iii, p. 275, translated in al-Ṭabarī, The History of al-Ṭabarī, xxviii, p. 243. 37 See Gil, Jews in Islamic Countries in the Middle Ages, p. 496. 38 The Babylonian Talmud includes a good deal of information on economic matters that exhibit continuities with the subsequent early Islamic period. See Elman, ‘Babylonian Jews at the Intersection of the Iranian Economy and Sasanian Law’, p. 549, and Elman, ‘Shopping in Ctesiphon’. 39 Sherira Gaon, Iggeret, pp. 105, 108. On the dating of the takkanah, see Brody, ‘Were the Geonim Legislators?’, pp. 304–05.

431

432

f r e d a s tr e n

of Moses ben Jacob, Gaon of Sura from 829 to 839, upholds the takkanah and parenthetically makes an historical observation. ‫תקינו רבנן למגבי כתובת אשה ובע״ח אפי׳ ממטלטלי משום דחכא רובה דעלמא לית להון‬ ‫מקרקעי ועבדו רבנן בתראי תקנתא שלא לנעול דלת לפני לווין ולמסמך אשה בכתובתה וגמרה‬ ‫להדא תקנתא מתקנות ראשונות דתקינו חכמים ראשונים בכל דרא ודרא ואסמכינון על כמה‬ ‫ אבל בשאר מקומות דרובא דאינשי אית להון מקרקעי אי לא‬,‫טעמי דלא אפשר לפרשינון‬ .‫מפרשין מטלטלין בחיי ובמותי לא גבן כתובת אשה ובע״ח אלא ממקרקעי‬

The rabbis decreed collection of a woman’s ketubbah and a lender’s [loan] even from movable property, since here most of the world (ruba de-ʿalma) has no land (leit lahon meqarqaʿei). The later rabbis effected the decree (takkanah) in order not to close the door to borrowers and [to allow] a woman to rely on her ketubbah. They learned [about making such enactments] from the earlier decrees that the sages made in every generation. […] However, in other places, where most people (ruba de-inshei) own land, if they do not specify ‘chattel during my lifetime and after my death’, a woman’s ketubbah and a lender’s [loan] are only to be collected from real property.40 In this translation by Phillip I. Lieberman, the stipulation that a creditor may collect movables instead of land is clear. The gaon’s reasoning is more subtle than his plain writing indicates. Moses ben Jacob’s ‘here’, whether Baghdad or somewhere else, is a place where the ownership of real estate is uncommon, whereas ‘in other places […] most people own land’.41 The description of economic conditions in the Babylonian Talmud suggests that urban dwellers often owned land outside of their towns or cities,42 whereas the responsum describes an urban milieu in which few people possessed extramural agricultural lands.43 Early Islamic urbanization increased the disjunction between city and country, which required legal reform and undoubtedly had social consequences as well. 40 Translation from Lieberman, The Fate of the Jews, pp. 51–52. A ketubbah is a Jewish marriage contract, which includes financial provisions for the wife in case of divorce. 41 The text of the responsum is found in Oṣar ha-ge’onim, ed. Lewin, viii, p. 210 n. 531. 42 On agriculture in the Babylonian Talmud, see Newman, The Agricultural Life of the Jews of Babylonia, which provides a compilation of references to agriculture in the Babylonian Talmud but lacks analysis or argument. 43 From this, Graetz argued that Jews in this period were abandoning agriculture and turning to commerce. See Graetz, Geschichte der Juden von den Ältesten Zeiten bis auf die Gegenwart, p. 196. Mann resists this reading, stating, ‘Perhaps the building of Bagdad in 762, which soon became a large commercial centre and attracted many Jews, contributed to a considerable extent to the necessity of this institution’ (Mann, ‘The Responsa of the Babylonian Geonim, Part 3’, p. 311). Lieberman has changed the terms of debate by examining Iraqi rural landholding patterns, patronage, and taxation, from which he argues that the takkanah and the responsum are insufficient evidence for a major occupational shift on the part of Jews and that Jewish agriculture in the form of metayage persisted into the Islamic period (Lieberman, The Fate of the Jews, 44–74). Since by the ninth century land was increasingly dominated by fewer landowners, who had rapidly Islamized, and whose labourers were becoming increasingly dependent on landowner patronage, the ability of rural Jewsh populations to persist would have been quite limited. On free labour and patronage in early Islamic Iraq, see van Bavel, Campopiano, and Dijkman, ‘Factor Markets in Early Islamic Iraq’.

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

In the sources, which are thin, Jews can be located across the economic spectrum in the burgeoning economy. The great Arab-Muslim littérateur al-Jāḥiẓ (d. 869) remarks in his Refutation of the Christians, ‘among them [Christians] are to be found government secretaries, attendants of kings, physicians of nobles, perfumers, and bankers; whereas you will find a Jew only as a dyer, a tanner, a cupper, a butcher, or a tinker’.44 A few decades later, two Jewish jahābidhā — non-governmental tax administrators who might also operate as financiers and bankers — were important figures in ʿAbbāsid finance and politics.45 One of their sons-in-law, Netira (d. 916), led an influential merchant and banking house in Baghdad.46 A century later, al-Maqdisī says that in Syria and Egypt, ‘most jahābidhā, dyers, money-changers, and tanners are Jews’.47 As already mentioned, Jewish involvement in trade is reported in the Babylonian Talmud, and it has been suggested that Jews were active traders in Khurasan, Sogdia, and even Kashmir in the sixth century, although the evidence is slight.48 A narrative found in medi­eval Arabic and Syriac sources of the late seventh and early eighth centuries is more revealing for the period in question. It describes the Muslim conquest of Rhodes in 654, parenthetically adding that the scrap bronze of the destroyed Colossus was sold to a Jewish merchant who used a huge number of camels to transport the scrap. Although the historicity of the narrative is doubtful, the literary trope of a Jewish merchant would not work with readers unless it had some plausibility.49 The most important evidence of Jewish merchant activity comes from the brief mid-ninth-century report by Ibn Khurradādhbih on the Jewish Radhānite merchants, whose trading networks stretched from the land of the Franks to China and from Khazaria across the Near East, including Constantinople and Baghdad.50 Moshe Gil has shown that Radhān was a district in the area of Baghdad that in pre-Islamic times was associated with nearby Ctesiphon-Seleucia — another suggestion of Jewish economic continuity that crosses the Sasanian–Muslim historical divide.51 Slight corroboration comes from the early seventh century, when the Greek historian

44 Al-Jāḥiẓ, al-Radd ʿalā al-Naṣāra, pp. 17–18. It can be argued that al-Jāḥiẓ’s satirical tone and polemical intent undermines the historical reliability of his information, but satire is not effective if it does not mirror the world it comments upon. See Webb, ‘Foreign Books’. A translation of this passage, with additional polemical description, is found in Stillman, The Jews of Arab Lands, pp. 169–70. 45 See Fischel, ‘The Origin of Banking in Mediaeval Islam’. 46 The Netira narrative was first published from an unidentified and now lost manu­script by Alexander Harkavy in ‘Netira und seine Söhne’. 47 Al-Maqdisī, Aḥsan al-taqāsīm, vii, p. 183. However, in Fāṭimid territories of the tenth and eleventh cen­turies, the title jahbadh applies to cashiers — lower-level officials. See Rustow, The Lost Archive, p. 86. 48 La Vaissière, Sogdian Traders, pp. 184–86. On Sogdians and trade, see also Rustow, The Lost Archive, pp. 118–22. 49 Conrad, ‘The Arabs and the Colossus’. 50 Ibn Khurradādhbih, Kitāb al-Mamālik wa-l-masālik, pp. 153–54. On Ibn Khurradādhbih, see Bosworth, ‘Ebn Ḵordāḏbeh, Abu’l-Qāsem Obayd-allāh’. 51 Gil, ‘The Rādhānite Merchants and the Land of Rādhān’. Gil claims that Jewish participation in trade goes back to Antiquity.

433

434

f r e d a s tr e n

Theophylactes Simocatta mentions Jewish traders whom he describes as active in the Red Sea and the Persian empire.52 In sum, there is enough evidence to posit that Jews were fully embedded participants in major social and economic developments of the era. Jews (and their religious institutions) migrated to Baghdad and other new or newly reordered cities, where they undoubtedly experienced the challenges of urban growth and social dislocation, but also actively participated in many registers of the economy. In such settings, individual Jews, as well as Christians and Muslims, lost connections that were formerly based on village, clan, or local religious community. Correspondingly, many aspects of the formal structures of community were left behind or dissolved through emigration or Islamization. Consequently, individuals in new early Islamic urban environments might seek new social connections and modes of belonging, including through religion. In the new urban settings, new types of intra-urban Jewish communal organization could develop, sometimes based on new forms of socio-religious bonds. This environment is reflected in the eighth-century halakhah ( Jewish law and jurisprudence) of Anan ben David, who is often identified as the founder of Karaism but is better understood as the founder of a smaller movement that later was folded into Karaism.53 Anan is also important as an author. Departing from the corporate authorship used in the rabbinic literature of his day, which was connected to rabbinic preferences for orality,54 he wrote under his own name. The halakhah known from the fragments that have survived of his Sefer miṣvot (‘Book of commandments’) suggests a community facing the challenges of social disruption and transformation. Salo W. Baron characterized the community as legally rigorist, segregationist, and marked by ‘metropolitan individualism’.55 In terms of segregation, Anan stipulates that circumcision is to be carried out using scissors, rather than a knife as prescribed by the rabbis, and by someone who has been likewise circumcised.56 A regulation such as this would support a closed community with high social boundaries that mark a radical separation between it and the outside world. Those who would observe such a law would thereby distinguish themselves from other Jews whose methods of circumcision were less demanding, such as rabbinic Jews. Similar high boundaries are directed against non-Jews in many of Anan’s halakhot. For example, in cities with non-Jewish inhabitants such as Kūfa and Baṣra (and later Baghdad), which predominated in the Jewish Babylonia (ʿAbbāsid Iraq) of Anan’s followers, it is prohibited to leave home on the Sabbath except to go to synagogue or to fulfil some other legal requirement. Anan even forbade the use of milk or water heated by a non-Jew. In another legal context, Anan rules that some 52 Theophylact Simocatta, The History of Theophylact Simocatta, bk 5, sect. 7.7–8. 53 On Anan, see Gil, ‘The Origins of the Karaites’. See also Ben-Shammai, ‘Between Ananites and Karaites’. 54 Elman, ‘Orality and the Redaction of the Babylonian Talmud’. 55 Baron, A Social and Religious History of the Jews, v, pp. 210–22. 56 Al-Qirqisānī, Kitāb al-Anwār wa-l-marāqib, i, p. 53.

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

laws of purity are applicable only to those twenty years old or older. Comparison can be drawn to adult baptism in early Christianity, which was the norm as long as Christians understood themselves to be members of a voluntary association.57 And like the Anabaptists of the sixteenth and seventeenth centuries, who specified that baptism was acceptable only for consenting adults, the Ananite position reflects the social realities of a voluntary association whose adult members make religious choices with extended social implications when they join the new community. From a rabbinic standpoint, some of Anan’s halakhah is highly irregular and breaks sharply with rabbinic norms but does so using rabbinic styles of biblical legal exegesis, including allegory, which he sometimes takes to extremes.58 It subverts rabbinic authority while creating new, non-rabbinic avenues for belonging marked by high social boundaries and legal rigour. Moreover, its textual eclecticism may have appealed to Jewish urban readerships. It constituted a new, communal form of Jewish resistance against the new, disorderly urban milieu. Anan’s Judaism offered a more challenging communal life than that of the rabbis, and as a voluntary association with propensities towards strictness and self-segregation it would necessarily have had limited appeal, especially in an urban setting. Among those who would have found such religious commitments appealing would have been those urban inhabitants who, with literacy and education, could find meaning in Anan’s teaching and writing. Those who could read (and write) would have included rabbis and the rabbinically educated, along with merchants, tax administrators, bankers, and members of some other professions. Some might have been Jewish urbanites who rose above the economic level of roustabout labour to join the new middle classes, perhaps in the employ of those who possessed and mobilized wealth, or as entrepreneurs. Like the ʿAbbāsid elite itself, such educated non-governmental elites and emerging urban middle classes generated demand for foodstuffs and luxury goods, and also sought out reading.

The Ninth-Century Explosion of Arabic Literature, its Readerships, and Jews The life of a ninth-century Muslim scholar illustrates some of the historical conditions discussed above and introduces developments in reading and writing set in the context of urbanization and economic growth. The career of Abū Maʿshar al-Balkhī, who died a centenarian in the year 886, provides a socio-historical frame for understanding the ‘intertwined worlds’ of Jewish writing and sectarianism and the larger Islamic intellectual and cultural milieu of the ninth century.59

57 However, as Christians began to think of themselves as a natural community, infant baptism became more and more acceptable, and eventually normative. See Louth, ‘Fiunt, non nascuntur Christiani’. 58 See the brief review of Anan’s exegesis in Frank, ‘Karaite Exegesis’, pp. 111–12, and the biblio­graphy there. 59 See Lazarus-Yafeh, Intertwined Worlds. On al-Balkhī, see Pingree, ‘Abū Maʿshar al-Balkhī, Jaʿfar ibn Muḥammad’, and Yamamoto and Burnett, The Great Introduction to Astro­logy by Abū Maʿšar, pp. 1–6.

435

4 36

f r e d a s tr e n

Abū Maʿshar al-Balkhī began his intellectual pursuits with the study of ḥadīth, the maturing Muslim pietist traditionist project, and from this intellectual ground he publicly opposed Abū Yūsuf Yaʿqūb b. Isḥāq al-Kindī, the famous polymath and first self-identified Muslim philosopher. Al-Kindī is credited with introducing Aristotelian concepts into Islamic philosophy.60 This opposition is exemplary of the culture war of Abū Maʿshar’s younger days, whose two sides represented a struggle between traditionalist and philosophical approaches to knowledge in Islam, and between the rejection or acceptance of foreign influence and knowledge in Islamic religion and society. It is noteworthy that Abū Maʿshar turned, perhaps under the influence of al-Kindī, to astronomy and astro­logy, the fields in which he would excel and for which he would become well known in the West as Albumasar. Two sources report that Abū Maʿshar’s turn followed his arrival in Baghdad while en route to Mecca on the pilgrimage. There he became enthralled by the books in the library known as the khizānat al-ḥikma (‘Treasury of Wisdom’). Consequently, he abandoned the ḥajj and even Islam.61 This narrative titbit is less likely a reliable historical report than it is a pietist polemic against a traditionist who had become an astro­loger. However, as mentioned above, such polemics have no value unless they resonate with contemporaneous cultural truths, one of which would have been that books have the power to reorient one’s intellectual and spiritual commitments. A parallel Jewish example of changing commitments is found in the figure of Dāʾūd b. Marwān al-Raqqī, known as al-Muqammiṣ or al-Muqammaṣ, one of the first Jewish Muʿtazilite philosophers of the ninth century. According to the tenth-century Karaite al-Qirqisānī, al-Muqammaṣ was converted from Judaism to Christianity by a certain Nānā, who is described as ‘the perfect philosopher’.62 Nānā is identified as the Jacobite physician and theo­logian Nonnus of Nisibis, who was alive in 862.63 After studying with Nonnus for many years, al-Muqammaṣ returned to Judaism under unknown circumstances and ‘wrote two books about the Christians in which he criticises them’.64 Among his other works is the philosophical work ʿIshrūn maqāla (‘Twenty Chapters’), which uses Muʿtazilite kalām techniques and is based on themes similar to those found in contemporaneous works that comment on various Aristotelian concepts. He was as current in his time as al-Kindī was, and as Sarah Stroumsa notes, ‘both the form and the content of this treatise reflect

60 On al-Kindī, see Ivry’s introduction to al-Kindī, Al-Kindi’s Metaphysics, pp. 3–51, and Adamson, Al-Kindī. 61 Yāqūt, Muʿjam al-udabāʾ, p. 2014, citing al-Muḥassin b. ʿAlī al-Tanūkhī (939–94). See also Ibn Ṭāwūs, Faraj al-mahmūn fī taʾrīkh ʿulamāʾ al-nujūm, p. 157. Much thanks to Abeer al-Abbasi for assistance with these sources. 62 Al-Qirqisānī, Kitāb al-Anwār wa-l-marāqib, i, pp. 44–45; but see also the index, s.v. ‘Dāʾūd b. Marwān al-Muqammiṣ’. 63 See Teule, ‘Nonnos of Nisibis’. I follow Philip Wood in referring to the Syrian Orthodox Church as Jacobite. See his The Imam of the Christians, pp. 23–25. 64 Al-Qirqisānī, Kitāb al-Anwār wa-l-marāqib, i, pp. 44–45. The English is from Lockwood’s translation of the first section of Kitāb al-Anwār in al-Qirqisānī, Yaʿqūb al-Qirqisānī on Jewish Sects and Christianity, p. 137.

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

Christian teaching’.65 Most importantly, he was the first to write in what for Jews was the new literary and exegetical form of biblical commentary.66 The most common form of his name, the nisba al-Muqammiṣ, has presented difficulties. It has been rendered as ‘jumper’, thereby connoting his serial conversions, although it could also refer to Dāʾūd’s profession, simply meaning ‘robe maker’. (‘Robe’ in Arabic is qamīs). Many scholarly suggestions have been proffered, but Sarah Stroumsa explains that an Arabic version of the Greek Life of Saint Stephen by Leontios renders the Greek of ‘an indigenous Arab son of an indigenous Arab’ as mqmṣ ibn mqmṣ, in reference to a Muslim convert to Christianity.67 The term may be translated as ‘one who wears a robe’, that is, of a Muslim (in contrast to Christian garb), and hence is a confirmed Muslim. In fact, Muslim robes may have been characteristically made of cotton, in contrast to the commonly worn linen in Egypt or wool in many places.68 Dāʾūd may have been perceived by his new Christian co-religionists as a wearer of Muslim clothing or as an Arabized Jew (who spoke Arabic), which to them carried Muslim associations. The connection of the term with conversion in the stories of both al-Muqammaṣ and the unnamed Muslim is intriguing.69 Abū Maʿshar al-Balkhī and Dāʾūd al-Muqammaṣ are only two of many ‘jumpers’ who exemplify movement between religious communities and intellectual schools in the early centuries of Islam.70 In further examples, in Eastern Christian sources there is concern over the return of apostate clergy to the church and the question of re-baptism,71 and Zoroastrian law deals with apostates who seek to return to the community.72 Similarly, a responsum of the eleventh-century Hayya Gaon, head of the Pumbeditha yeshiva, refers to an eighth-century legal responsum of Yehuday Gaon (d. 761) of the Sura yeshiva that addresses the case of Zoroastrians who become Muslims but are not sincere in their conversion.73 In another responsum,

65 Stroumsa’s introduction to al-Muqammiṣ, Dāwūd ibn Marwān al-Muqammiṣ’s Twenty Chapters, p. 34. Stroumsa also indicates parallels with such Muslim writers as al-Mātūridī, ʿAbd al-Jabbār, al-Ashʿarī, al-Shahrastānī, and others (pp. 23–35). 66 According to al-Qirqisānī, al-Muqammaṣ wrote commentaries on Genesis and Ecclesiastes; only a small fragment of the former is extant. See Stroumsa, ‘From the Earliest Known Judaeo-Arabic Commentary on Genesis’. 67 Magaritès tís huios margaritou tôn autokkhthonôn. The matter is reviewed in Stroumsa, ‘From the Earliest Judaeo-Arabic Commentary on Genesis’. 68 See Bulliet, Cotton, Climate, and Camels in Early Islamic Iran, pp. 44–68. 69 Stroumsa, ‘From the Earliest Judaeo-Arabic Commentary on Genesis’, p. 378. On the basis of the evidence, Stroumsa upholds the preferred vocalization ‘al-Muqammaṣ’. For more examples of the porosity and management of intercommunal boundaries, see Tannous, The Making of the Medi­eval Middle East, and Weitz, Between Christ and Caliph. 70 Many cases of conversion and re-conversion are studied in Simonsohn, ‘“Halting between Two Opinions”’ and ‘Conversion, Apostasy, and Penance’. 71 See a query on this matter from John of Litarb to Jacob of Edessa (d. 708) in The Synodicon in the West Syrian Tradition, ed. Vööbus, vol. 367, p. 253 (Syr.) and vol. 368, pp. 231–32 (Eng.), no. 15. Cited in Simonsohn, ‘Conversion, Apostasy, and Penance’, p. 213. 72 See Simonsohn, ‘“Halting between Two Opinions”’, p. 351 n. 24. 73 The halakhic question is whether wine that has been handled by such Zoroastrians-cum-Muslims has

437

43 8

f r e d a s tr e n

Natronay bar Hilay Gaon (second half of the ninth century) deals with a Jew of priestly ancestry who has left Judaism for a time.74 These and other instances of turning (like Abū Maʿshar) or ‘jumping’ are correlative of Islamization, but they are also coterminous with the dislocations of late eighth- and ninth-century urbanization, immigration, and economic boom that facilitated intellectual and spiritual choice in a new age of writing. The ninth-century religious and intellectual peregrinations of Dāʾūd al-Muqammaṣ are reminiscent of an earlier historical era. Josephus — Jewish general in the first-century war against Rome, then turncoat, and later historian — describes how he studied the doctrines of the Pharisees, the Sadducees, and the Essenes and then spent three years with the hermit Bannus before deciding to join the Pharisees.75 In The Flourishing of Jewish Sects in the Maccabean Era Albert Baumgarten highlights the socio-religious journeying of Josephus in order to help explain the sectarian environment of late antique Judea, making instructive comparisons to the volatile history of seventeenth-century England, including the English Revolution.76 Baumgarten cites Oxford historian Christopher Hill, who describes ‘masterless men’ moving from one ideo­logical group to another in a society that was breaking apart and whose hierarchies and links of patronage were unravelling. These ‘floaters’, as Hill refers to them, found both meaning and social location among such groups as Anabaptists, Levellers, Diggers (later True Levellers), Ranters, Fifth Monarchists, Muggletonians, Quakers, and others.77 Factors that Hill and others have identified as crucial for understanding that period of English history include urbanization, internal immigration, changes in land tenure, and changes in the economy, all of which have cognates in first-century Judea and late eighth- and ninth-century Iraq.78 Both of these periods witnessed a proliferation and intensification of Jewish sectarianism and sectarian writing. In the eighth and ninth centuries, cities such as Baghdad that witnessed economic growth and an influx of people were characterized by the emergence of both an increasingly literate middle class and an urban proletariat whose roots in a distant countryside or an older, pre-Islamic version of their city left them unmoored from social and cultural bonds that had previously provided guidance and grounding. Mirroring seventeenth-century England, the social dislocation in Baghdad and other cities provided fertile ground for the emergence of new social and religious

been rendered unfit. See Teshuvot ha-geʾonim mi-kitve-yad she-bi-genizat Cambridge, ed. Assaf, no. 61, and Abraham ben Isaac, Sefer ha-Eshkol, ii, pp. 73–74. Cited in Simonsohn, ‘Conversion, Apostasy, and Penance’, pp. 197–98. 74 After his return to Judaism the priest was no longer permitted to perform rituals usually reserved for those of priestly background. See Otsar ha-ge’onim, ed. Lewin, i, p. 140 (no. 35) n. 1. Cited in Simonsohn, ‘Conversion, Apostasy, and Penance’, p. 211. 75 See Josephus, Life of Josephus, § 9–12, in Josephus, i, pp. 4–7. 76 Baumgarten, The Flourishing of Jewish Sects in the Maccabean Era. 77 Hill, The World Turned Upside Down. 78 See Campopiano, ‘Land Tenure, Land Tax and Social Conflictuality in Iraq’, which links landholding and ‘social conflictuality’; and van Bavel, The Invisible Hand?, 41–78.

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

affiliations among Jews (as it did among Muslims). These affiliations could centre on a new type of religious leader who, like Anan in the eighth century, offered innovative, if not heterodox scriptural interpretations and laws that could function as doctrine for new urban Jewish socio-religious configurations. Several of these will be discussed below. This is an environment that Hill’s seventeenth-century floaters and Baumgarten’s late antique Judea bring into view. In the case of Jews, much of whose religious tradition was long established (and included traditions of reading, writing, and orality), new types of religious leaders naturally would challenge inherited forms of belonging and belief, and of reading and writing. In an environment of Baron’s ‘metropolitan individualism’, these leaders would certainly attract jumper-floaters like Dāʾūd al-Muqammaṣ. Returning to the ‘turn’ of Abū Maʿshar al-Balkhī, what does it mean to say that books in a library could unhinge the belief and commitment of a pious Muslim? The contents of the khizānat al-ḥikma, where Abū Maʿshar read, and of other libraries were the result of an unprecedented explosion of writing, literary production, and (of course) reading in Arabic that occurred in the midst of the already volatile combination of urbanization, migration, and economic boom. The character of the khizānat al-ḥikma, whose books impelled Abū Maʿshar’s purported turn, is debated by scholars. It may have been a huge personal library open to scholars that had been founded by Abū al-Ḥasan ʿAlī b. Yaḥyā b. al-Munajjim (d. 888), astro­loger and boon companion to caliphs.79 Or it may have been the caliphal library or one of its sub-libraries that specialized in astro­logy.80 It and other contemporaneous libraries are to be associated with the historically enigmatic bayt al-ḥikma (‘House of Wisdom’), an institution at the heart of the intellectual achievements under the early ʿAbbāsid caliphs. The bayt al-ḥikma, in turn, is associated with the translation movement that introduced books originally written in Greek, Syriac, Pahlavi, and Hebrew to Arabic readers and with the contentious question of whether the importation of foreign and pre-Islamic knowledge into the Islamic setting is permissible.81 In this context, we find a new emphasis on Aristotle and the development of the philosophy of kalām, mathematics, science, and esotericism — and the opposition of Abū Maʿshar al-Balkhī and al-Kindī, not to mention the subsequent transformation of Abū Maʿshar. Moreover, other subjects and literary genres developed, including the study of grammar, lexico­graphy, courtly writing (adab), and new ways of writing poetry, history, tradition, and law.

79 See Roggema, ‘ʿAlī ibn Yaḥyā ibn al-Munajjim’. On ninth-century libraries, see McKinney, The Case of Rhyme versus Reason, pp. 108–10, and Toorawa, Ibn Abī Ṭāhir Ṭayfūr and Arabic Writerly Culture, p. 138 n. 70. 80 See Janos, ‘Al-Maʾmūn’s Patronage of Astro­logy’. 81 On the bayt al-ḥikma, see Janos, ‘Al-Maʾmūn’s Patronage of Astro­logy’; Gutas, Greek Thought, Arabic Culture; Saliba, Islamic Science and the Making of the European Renaissance, especially chapters 1 and 2; Balty-Guesdon, ‘Le Bayt al-hikma de Baghdad’; Balty-Guesdon, ‘Bayt al-Ḥikmah et politique culturelle du caliphe al-Maʾmūn’; Balty-Guesdon, ‘La Maison de la Sagesse’; and Di Branco, ‘Un’istituzione sasanide?’.

4 39

440

f r e d a s tr e n

In his book on the ninth-century courtly writer Ibn Abī Ṭāhir Ṭayfūr, Shawkat Toorawa describes the emergence of a ‘writerly culture’ in this period, emphasizing not simply writing or books but the writer, whose agency was manifest both in new social settings such as literary salons and the stationer’s or bookseller’s market and in more established settings such as the caliphal court and the mosque.82 Hugh Kennedy observes that the emergence of readerships and book markets in ninth-century Baghdad meant that ‘Abbasid Baghdad was probably the first place in the world where an author could make a living […] by writing books to be sold in the market to a literate public’.83 Furthermore, writers and readers intensified intellectual exchange that could transcend confessional boundaries between Muslims, Christians, Jews, and even polytheist Ḥarrānians, as philosophers, theo­logians, scientists, and linguists, among others, shared sources and used each other’s writings, whether attributed or plagiarized. In conjunction with reading and writing, they participated in symposia (majālis), both as a spectacle in the caliphal court and in private settings, where communal theo­logical positions and all other kinds of intellectual questions were debated and whose informed audiences came from ʿAbbāsid elites and new educated urban readerships.84 In addition, there was the religious teaching circle (ḥalqa), in which Muslim traditionists (muḥaddithūn) and other religious specialists transmitted information using methods that were dependent upon memorization, with texts committed to memory completely or partially.85 All of these venues and their activities constituted forms of social affiliation as much as sites of learning, and from them we have a fairly well-drawn picture of the intellectual milieu, in which it is plain that Muslims, Christians, polytheist Ḥarrānians, even Zoroastrians, and some Buddhists, as well as Jews of different kinds, participated. Such intensification would have been limited without paper, whose production was introduced in Iraq in the late eighth or early ninth century. Under the ʿAbbāsids, the demand for writing media was at first met by the use of papyrus, exported from Egypt and subject to limits on production, and by parchment, which was costly and whose manufacture was inefficient for the production of large quantities of writing material.86 According to a late tenth- or early eleventh-century report that Marina Rustow describes as an ‘etio­logical tale’, this was to change immediately after the ʿAbbāsid revolution, as an incidental consequence of the Battle of Talas in 751, when Muslim forces defeated a Chinese imperial army in Central Asia.87 82 Toorawa, Ibn Abī Ṭāhir Ṭayfūr and Arabic Writerly Culture. 83 Kennedy, ‘Baghdad as a Center of Learning and Book Production’, p. 103. 84 On majālis al-munāẓara, held at the court, and majālis held in lesser settings, see McKinney, The Case of Rhyme versus Reason, pp. 99–106. 85 See Melchert, ‘The Etiquette of Learning in the Early Islamic Study Circle’, and McKinney, The Case of Rhyme versus Reason, pp. 112–13. 86 On the use of papyrus, parchment, and the codex by late antique Jews, see Hezser, Jewish Literacy in Roman Palestine, pp. 126–44. See also Nongbri, God’s Library, pp. 21–46. 87 See al-Thaʿālibī, Laṭāʾif al-maʿārif, p. 218, translated in The Laṭāʾif al-maʿārif of al-Thaʿālibī, p. 140. See Rustow, The Lost Archive, pp. 113–37, whose revision of the proliferation of paper from China via the Sogdians to the Near East is convincing.

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

As reported, two of the Chinese prisoners of war were papermakers, and their knowledge was tapped to initiate the production of paper initially in Samarqand and then more widely.88 However, other sources point to the late eighth century. In a tenth-century report recorded by al-Jahshiyārī, the caliph al-Manṣūr (r. 754–775) is said to have looked at a receptacle full of papyrus scrolls and remarked in reference to the governor of Egypt, ‘We need to write on something for which we do not pay governors’. He then ordered his secretary to find a way to deal with the problem.89 An early eleventh-century source puts the event in the time of Hārūn al-Rashīd.90 A thirteenth-century report states that a paper-making facility was built in Baghdad in 794–95, during the caliphate of Hārūn al-Rashīd (r. 786–809).91 And in the fifteenth century, Ibn Khaldūn linked the introduction of paper in Baghdad to the Barmakids, the powerful family of viziers under Hārūn who were eventually purged.92 Ibn Khaldūn’s assertion has led scholars to think that initial impulses toward the use of paper and associated literary production came from empire, an assumption that is reasonably upheld by al-Jahshiyārī. However, Rustow argues that it is more likely that Ibn Khaldūn’s claim was a mirror of his own intimate experiences in the Ḥafṣid and Marīnid chanceries.93 Whatever the origins of papermaking in ʿAbbāsid Iraq, the eighth- and ninth-century caliphate required the keeping of records for orderly administration, and paper was one of several media that were used.94 In fact, the dīwān system may have required more record-keeping than did administrative systems in other premodern empires.95 At first, old bureaucracies were inherited and indigenous scribal classes were employed by the administration, as reported by al-Jāḥiẓ above. A tenth-century report states that ʿAbd al-Malik (r. 685–705) addressed such makeshift arrangements when he decreed that all governmental records were to be written in Arabic.96 However, this report is best understood as an anti-dhimmī polemic and a response to the continued recruitment of non-Muslim bureaucrats from the descendants of conquered peoples.97 Even as the payment of salaries drove demand for goods and services, governmental record-keeping generated demand for writing and writing media. Moreover, as a scriptural religion, with the Qurʾān at its heart, Islam constituted in part a textual community. Accordingly,

88 Summarized in Bloom, Paper before Print, pp. 42–45. 89 Al-Jahshiyārī, Kitāb al-Wuzarāʾ wa-l-kuttāb, p. 158. Translation from Gruendler, ‘Book Culture before Print’, p. 2 n. 2. 90 Al-Maqrīzī, al-Mawāʿiẓ wa-l-iʿtibār, i, p. 147. 91 Yāqūt, cited in Bloom, Paper before Print, p. 48. Unfortunately, Bloom’s book does not include notes. 92 Ibn Khaldūn, cited in Bloom, Paper before Print, pp. 49–50. 93 Rustow, The Lost Archive, pp. 132–33, which includes the citations from Ibn Khaldūn. 94 Rustow, The Lost Archive, pp. 133–35. 95 See van Berkel, ‘Reconstructing Archival Practices in Abbasid Baghdad’. However, see Kennedy, ‘Baghdad as a Center of Learning’. 96 See Gruendler, The Rise of the Arabic Book, p. 5; Gruendler cites the Arabic sources and treats the report as reliable. 97 See the chapters in this volume by Marie Legendre and Petra Sijpesteijn.

441

442

f r e d a s tr e n

most Muslim reflections on theo­logy, law, ritual, commentary, and all manner of religious thought eventually came to be expressed in writing. Whatever paper’s origins in Iraq, it seems to have been produced in quantity from the outset.98 Papermaking was not a cottage industry like most other manufacturing in Late Antiquity and the early Middle Ages. It was a kind of industrial enterprise (like sugar refining and dyeing) that required investment in scale, in this case on the part of government officials.99 Paper production, while initially meant for the chancery, the dīwān, and tax administration, produced surpluses that found their way into society, primarily at the sūq al-warrāqīn (literally, ‘market of stationers’, which may also be translated as the ‘market of the papermakers’ or the ‘market of the booksellers’). It was evidently a large place: al-Yaʿqūbī (d. after 908) reports that there were more than a hundred booksellers’ shops located on one particular roadway in al-Karkh.100 There was another sūq al-warrāqīn in al-Ruṣāfa across the Tigris.101 Fundamentally, these markets were places of commerce, where warrāqūn (‘stationers’, literally ‘folio makers’) made their living. Challenging as it did the hegemony of traditional cultural repositories of knowledge, the bookseller’s market was a place where anyone who could read might find authors, information, and arguments, and it thus made possible new kinds of autodidactic and group learning that, in turn, could circumvent or subvert mosque, monastery, or rabbinic academy. The perfect example is a report on the aforementioned al-Jāḥiẓ, who would rent out the shop of a warrāq overnight in order to read.102 This report tantalizingly suggests that like al-Jāḥiẓ, the many literate but dislocated urban intellectuals and jumper-floaters in Baghdad could look for knowledge and meaning (or even entertainment) in the written word’s ready availability on paper in the newly adopted and easily navigated form of the codex. Encompassing everything about writing, the work of warrāqūn included papermaking, ink making, copying, bookbinding, and all aspects of publishing, including finding authors to publish and buyers to purchase the finished products. Sub-specializations included collators, proofreaders, copyists who added vowel markers to the Arabic consonantal text, and copyists who produced clean exemplars, or master copies, in large legible writing from the auto­graph manu­ scripts of authors.103 By the third quarter of the ninth century, between authors and warrāqūn, a new legible text layout evolved for writing in Arabic. It included

98 See Bloom, Paper before Print, pp. 47–52. 99 Ninth-century developments in coordinated production, factor markets, labour patterns, and land tenure in the caliphate indicate increases in efficiencies and scale. See van Bavel, The Invisible Hand?, pp. 41–60; Bessard, Caliphs and Merchants. 100 Al-Yaʿqūbī, Kitāb al-Buldān, p. 245; English translation in The Works of Ibn Wāḍiḥ al-Yaʿqūbī, i, p. 78. See also Le Strange, Baghdad during the Abbasid Caliphate, p. 92, describing the booksellers’ market in al-Karkh as even larger in area than in al-Yaʿqūbī’s description. 101 McKinney, The Case of Rhyme versus Reason, p. 108, noting that it is often mentioned by al-Tawḥīdī. 102 Reported by Abū Hiffān (d. 871) in Yāqūt, Muʿjam al-udabāʾ, xvi, p. 75. 103 See Gruendler, ‘Aspects of Craft in the Arabic Book Revolution’, especially pp. 37–40; and now Gruendler, The Rise of the Arabic Book, pp. 103–39, which distinguishes different types of warrāqūn.

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

word spacing, para­graphs, section headings, indentation, and the use of blank spaces. What would today be called page design even developed a specific form for the insertion of scriptural proof verses and poetry into prose text.104 Books themselves were being transformed, as writing turned texts into objects meant to be read visually rather than experienced in the customary and conventional way of Late Antiquity — aurally. In addition to being a site of papermaking and book production, the sūq al-warrāqīn functioned as a major venue for intellectual exchange alongside the majālis of the court, private salons, ḥalqas, and other settings. Many majālis and other discussions are reported at shops located there.105 According to al-Qifṭī, the shop of a warrāq named ʿAbd Allāh b. Muḥammad al-Azdī (fl. 845) was the meeting place of intellectuals and hosted more debates than any of the literary salons.106 According to the Manāqib Baghdād, the large sūq al-warrāqīn in al-Ruṣāfa was the ‘meeting place of poets and scholars’.107 In fact, many warrāqūn were more than tradesmen. Those who are known from the sources were educated intellectuals, and many authored books of significant influence.108 As the form of books was transformed from that of a memorized and recited text to a written object, the warrāq became an increasingly important figure in the ownership and dissemination of knowledge.109 That importance is caricatured in a polemical narrative directed against the ninth-century scholar Ibn al-Rāwandī, who was considered a heretic in many circles. Josef van Ess writes that Qādī ʿAbd al-Jabbār (d. 1025) claimed that a few warrāqūn ‘got together and just for fun wrote down a particularly heretical theory that none of them believed. Ibn al-Rāwandī then took this treatise and published it under the title Kitāb al-Tāj’.110 This conspiracy theory, designed to dismiss Ibn al-Rāwandī’s theo­logical claims, puts on display the nexus between the sūq al-warrāqīn, writing, and sectarianism.111 Beatrice Gruendler observes that working as a warrāq ‘suited non-conformist members of society, such as libertines and heretics, who are amply attested among freelance stationers’, and that the profession provided ‘a niche for dissent and added to the pluralism of voices in early ʿAbbāsid times’.112 It is obvious that such an environment would attract jumper-floaters, even Jews.

104 Gruendler, ‘Book Culture before Print’, p. 18. 105 Much of the evidence is reviewed by McKinney, The Case of Rhyme versus Reason, pp. 85–119. 106 Al-Qifṭī, Inbāh al-ruwāt, ii, p. 134. Cited by Makdisi, The Rise of Humanism, p. 63. 107 Translated and cited in McKinney, The Case of Rhyme versus Reason, p. 108. The Manāqib Baghdād is often attributed to Ibn al-Jawzī. 108 Perhaps none more significant than the tenth-century Ibn al-Nadīm. See Stewart, ‘Abū al-Faraj Muḥammad ibn Isḥāq al-Nadīm’; and Ibn al-Nadīm, The Fihrist of al-Nadīm. 109 On the transformation of Arabic literature, see Schoeler, The Genesis of Literature in Islam, and a collection of Schoeler’s articles, The Oral and the Written in Early Islam. See also Gruendler, The Rise of the Arabic Book, pp. 103–39. 110 Van Ess, Theo­logy and Society in the Second and Third Centuries of the Hijra, iv, p. 357. 111 The biblio­graphy on early ʿAbbāsid sectarianism is vast. See Yücesoy, Messianic Beliefs and Imperial Politics in Medi­eval Islam, and Tucker, Mahdis and Millenarians. 112 Gruendler, ‘Aspects of Craft in the Arabic Book Revolution’, 39–40.

443

444

f r e d a s tr e n

Here, the parallel with seventeenth-century England needs to be revisited, for that was a period of increasing literacy in England, when society was flooded with printed books and especially pamphlets that contributed to religious and political segmentation in English society — a direct consequence of the breakdown of government control of censorship during the Long Parliament in late 1640 and early 1641.113 The connection between the proliferation of the written word and social instability can be noted in the Islamic world in the ninth century, when following destructive and murderous rioting, the government closed down the booksellers. Al-Ṭabarī reports for the hijrī year 279 (892–93): ‫فمن ذلك ما كان من أمر السلطان بالنداء بمدينة السالم أال يقعد على الطريق وال في مسجد الجامع‬ ۰‫قاص وال صاحب نجوم وال زاجر وحلف الوراقون أال يبيعوا كتب الكالم والجدل والفلسفة‬

Among the events taking place, the authorities decreed in Madīnat al-Salām that no popular preachers, astro­logers, or fortune-tellers should sit [and practice their trade] in the streets or in the Friday Mosque. Moreover, the booksellers were sworn not to trade in books of theo­logy, dialectic, or philosophy.114 Urban rioting in ninth- and tenth-century Baghdad was often motivated by popular attitudes for or against caliphal policy, which were most usually framed using religious rhetoric and ideas.115 In this religious context, urban violence could be impelled by anti-Christian animus or Shīʿite populism or by the instigation of Ḥanbalī preachers, who arose in the late ninth and tenth centuries in opposition to Shīʿism and the theo­logical schools of the Muʿtazila and Ashʿariyya.116 In fact, the sūq al-warrāqīn was also closed on another occasion in 897. In anticipation of public disorder, the caliph al-Muʿtaḍid (r. 892–902) ordered a ban on storytellers and study groups, declaring that ‘people who gather for discussion or disputation no longer enjoy legal protection’.117 This was a world in which intellectual life, religious sensibilities, and public order were all intertwined. Readerships and followers of popular preachers — reinforced by urban jumper-floaters seeking belonging and meaning — could constitute security threats and conceptual challenges to the social order and its hegemons.

113 See Achinstein, ‘Texts in Conflict’, and other chapters in the volume in which it was published. Compare to McElligott, Royalism, Print and Censorship. 114 See al-Ṭabarī, Taʾrīkh al-rusul wa-l-mulūk, iii, p. 2131. The English translation above is adapted from al-Ṭabarī, The History of al-Ṭabarī, xxxvii, p. 176. 115 See Amabe, Urban Autonomy in Medi­eval Islam, pp. 21–22. 116 Popular urban violence in Baghdad has been studied more for the tenth and eleventh centuries than the ninth. See Cahen, ‘Mouvements populaires et autonomisme urbain dans l’Asie musulmane du Moyen Age’; Badrī, al-ʿĀmma bi-Baghdād fī al-qarn al-khāmis al-hijrī; Sabari, Mouvements populaires à Baghdad à l’époque ‘abbasside; and Amabe, Urban Autonomy in Medi­eval Islam. 117 See al-Ṭabarī, Tarīkh al-rusul wa-l-mulūk, iii, pp. 2164–65, and The History of al-Ṭabarī, xxxviii, pp. 46–47.

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

Jewish Readerships and Sectarianism Although there is no direct textual evidence of a Jew buying a book or reading in the sūq al-warrāqīn, by the early tenth century the working scholarly life of the aforementioned Karaite al-Qirqisānī included participation in discussions, majālis, and debates with Muslims, Christians, Jews, and probably Sabians. Their subjects were law, philosophy, theo­logy, and a broad range of other topics. In a word, al-Qirqisānī is as good as any exemplar of the classical Islamicate intellectual world of books, ideas, and writerly culture that had developed and matured over the ninth century.118 Accordingly, he demonstrates conversance — in both theo­logical and philosophical registers — with rabbinic, Muslim, Christian, and non-monotheistic late antique literature. Moreover, he explicitly identifies Dāʾūd al-Muqammaṣ as one of his sources. Al-Qirqisānī’s encyclopaedic Kitāb al-anwār wa-l-marāqib (‘Book of lights and watchtowers’), written in 937, is clearly a product of the highly diverse and active intellectual world of its day, both reflecting the author’s tenth-century intellectual milieu and providing a unique witness to the contexts of ʿAbbāsid book culture in which Jews participated in the preceding century. The first section of the book, from which most of the following evidence is drawn, is an extensive heresio­ graphy that describes and refutes a number of Jewish sects (afāriq al-yahūd) in chrono­logical order. As such, it is a history of unbelief and belief and constitutes the most important historical source for the variety of Judaisms that emerged in the early Islamic centuries.119 Al-Qirqisānī wrote at a time when many of these ‘heresies’ had declined or disappeared and when full-blown Karaite Judaism was flourishing, with its important centre in Jerusalem. I use the word ‘heresy’ advisedly, since al-Qirqisānī refers to religious groups by a variety of different Arabic terms, including madhāhib, ilḥād, and khawārij, which are not merely Arabic words but also Muslim vocabulary. He also uses kānū bi-l-umam (‘to be like Gentiles’), a phrase whose evocation of non-Jewish identity echoes the Muslim debate over the fitness of knowledge from outside of Islam. In his far-ranging work, scholars have noted two important binary distinctions that al-Qirqisānī uses to mark himself and his intellectual adversaries. First, he distinguishes himself as a Karaite, in opposition to the rabbis and their form of Judaism. In this binary, the Bible, halakhah, and their interpretation are paramount. Second, as a theo­logian and philosopher he explicitly divides scholars into two groups: those who use rational speculation (and outside knowledge), and those who do not.120 Reading al-Qirqisānī again reveals a third binary distinction that reflects his writerly world but that has yet to be commented on: those who write and those who do not write.

118 See Astren, ‘Jacob al-Qirqisānī’. 119 Chiesa, ‘Ya‘qūb al Qirqisānī come fonte storiografica’. 120 Astren, Karaite Judaism and Historical Understanding, pp. 98–123.

445

446

f r e d a s tr e n

The ninth century of al-Qirqisānī’s heresio­graphy is populated by six heterodox Jewish figures, framed as interpreters of Jewish law and leaders of religious groups. Of these, al-Qirqisānī identifies four as having generated writing of some sort. Each, in some measure, can be contextualized into the new urban environment, with its jumper-floaters, or into the emergent writerly culture. It can safely be presumed that followers of these interpreter-leaders would congregate for prayer and preaching, if not for ḥalqa- or majlis-like meetings. Whether delivered by way of preaching, debate, or reading, new religious teachings would have been grist for the mill of urban Jews seeking meaning and belonging in Baghdad and other Near Eastern cities. Even though the sources do not provide a view inside such socio-religious phenomena, it is likely that those who could read formed readerships at the core of many Jewish groups and ‘sects’ of the period, and that reading and writing for those readerships were shaped by the explosion and availability of writing in Arabic. The new modes of reading and writing that emerged were well suited for explicit opposition to the rabbis, whose tradition was vulnerable to critique using scripturalist and other interpretive strategies.121 The first of these ninth-century interpreter-leaders was Ismāʿīl al-ʿUkbarī, who lived at the time of the caliph al-Muʿtaṣim (r. 833–42).122 It is not known whether Ismāʿīl lived in Baghdad, but nearby ʿUkbarā is recorded to have had a Jewish population in the Middle Ages. Beyond halakhic elements that were common to or similar in other Jewish sects of the era, Ismāʿīl’s approach to Scripture reflects features of contemporaneous book culture. Al-Qirqisānī notes that Ismāʿīl wrote books (fī kutubihi), in which he belittles Anan ben David and which ‘amount to no more than ravings, at which any educated man, if he heard them, would laugh’. He is reported to have categorically rejected a Jewish textual tradition according to which some words in the Hebrew Bible are to be read (qere) differently than the way they are written (ketiv).123 His claim that the Bible should be read as written, including even pronunciation of the customarily taboo four-letter name of God, the Tetragrammaton, constitutes a strained scripturalist reaction to the rabbis, and especially to the bookish Masoretic project, which was active in Iraq and Palestine in the eighth and ninth centuries.124 The Masoretes, described by David Stern as the first ‘professional Jewish readers of the Bible’, helped initiate the transition from aural to visual reading of the Bible:125 around this time, public recitation and remembered verses were being replaced by private, individualized reading from a

121 See Ben-Shammai, ‘Return to the Scriptures in Ancient and Medi­eval Jewish Sectarianism and in Early Islam’. 122 Al-Qirqisānī, Kitāb al-Anwār wa-l-marāqib, pp. 13, 56–57. 123 Such reading traditions act as a corrective to erroneous readings that result from a language with an abjad writing system (in this case Hebrew, but also Arabic, Aramaic, and Syriac) in which letters represent consonants in most instances and vowels are supplied by the reader. 124 See Martín-Contreras, ‘Safeguarding the Lord’s Word’. For a more complete treatment, see MartínContreras and Seijas de los Ríos-Zarzosa, Masora. 125 Stern, ‘The First Jewish Books and the Early History of Jewish Reading’.

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

written medium.126 The Masoretic project, whose purpose was to preserve and fix the consonantal biblical text and provide a system for its proper pronunciation, required the synoptic ability to identify, enumerate, and compare small features of the written text (such as qere and ketiv) that practicably could not be done from memory or by reading from a scroll without great difficulty. Skills such as these, associated with the new, visual mode of reading, inevitably led to the use of the codex, which Malachi Beit-Arié describes as ‘the much more convenient, capacious, durable, easy to store, carry about, open and refer to book form’127 and which at the time was most easily, and therefore most likely, made using paper.128 It should come as no surprise that all extant Masoretic books are in the codex form, and the oldest known Jewish codex dates to the eighth or ninth century.129 Unsurprisingly, the word for codex used by medi­eval Jews was taken from Arabic (miṣḥaf/muṣḥaf). Furthermore, like the warrāqūn, Masoretes later produced texts by employing specialists who each contributed one feature of the elements of the page, such as a scribe, a punctuator, and someone to enter the Masoretic textual apparatus.130 According to al-Qirqisānī, Ismāʿīl would have agreed with one of the central problems addressed by the Masoretes — that the existing text of the Bible could be corrupted — and accordingly he attributed some scriptural readings to copyist error. However, his position seems to have been utilitarian, adopted in justification of his own, sometimes idiosyncratic and naïve biblical interpretations rather than out of a concern for ascertaining a correct biblical text. This recourse to an unsystematically interrogated biblical textuality supported his naïve textual corrections, whereby he substituted words or phrases that he thought made better sense than those written in the text. Such unmethodical interpretations of difficult biblical passages parallel the kind of simple scripturalism seen in a contemporary literary genre based on biblical questions and answers. David Sklare describes fragments of early Judeo-Arabic texts found in the Cairo Geniza that interrogate contradictions and inconsistencies in the biblical text that ‘result from a close, but sometimes simplistic reading of the Bible’.131 Many of these were written in phonetic ortho­graphy, which reproduces spoken Arabic rather than the slightly later Judeo-Arabic that developed to a degree under the influence of literary high Arabic. It cannot be determined whether Ismāʿīl’s works fall into this early stage of the Jewish adoption of Arabic, since none of his writings survive, and neither is he quoted by other writers. However, his association with a mindset similar to

126 On rabbinic aural textuality, see Wollenberg, The Closed Book. For comparison, see Guy Stroumsa, ‘On the Status of Books in Early Christianity’. 127 Beit-Arié, Hebrew Manu­scripts of East and West, p. 11. 128 It should be noted that although Christians adopted the codex form early on, the Syriac Christian ‘Masoretic’ project was contemporaneous with Muslim and Jewish enquiry into these matters. See Butts, ‘The Classical Syriac Language’, especially pp. 235–37. 129 Stern, ‘The First Jewish Books and the Early History of Jewish Reading’, p. 164; but see Rustow, The Lost Archive, p. 386, which dates them to the first half of the ninth century. 130 Stern, The Jewish Bible, 70. 131 Sklare, ‘Ninth-Century Judeo-Arabic Texts of Biblical Questions and Answers’.

447

4 48

f r e d a s tr e n

that of the early question-and-answer form locates him in an intermediate position between the formulaic and thematic biblical exegesis of the rabbis and the newly evolving modes of close sustained reading. Even if Ismāʿīl’s methods were unsophisticated and did not depend on working visually with the biblical text in codex form, his response to Masoretic matters is testimony that the new visual way of reading, along with the affiliated use of the codex, was being contested by Jews. Such contestation mirrors the wide range of opinion in similar debates among Muslims concerning the text of the Qurʾān, its reading, and its transmission.132 Like the Masoretes, Muslims were concerned with ascertaining the correct reading (and therefore meaning) of the Qurʾān, but also with codification of the evolving Arabic high language at the heart of ʿAbbāsid culture, which differed from common speech.133 Even as Muslim authors, readers, and warrāqūn challenged received notions of text in Arabic, the Arabic grammar of Sībawayhi (d. c. 796) addressed the Qurʾānic text, while The Rectification of Speech by Yaʿqūb b. al-Sikkīt (d. c. 857–861) gave the educated and those who aspired to education (such as urban intellectuals and jumper-floaters) a practical guide to reading and speaking Arabic in its new contexts.134 Perhaps contemporaneous with Ismāʿīl al-ʿUkbarī was the second interpreter-leader, Benjamin ben Moses al-Nahawendī, who probably lived in the first half of the ninth century.135 It is reported that Benjamin wrote commentaries on the Torah and the books of Isaiah, Daniel, Ecclesiastes, and the Song of Songs. Except for a manu­script fragment of his commentary on Genesis, all of Benjamin’s commentaries are known only through quotations in later Karaite works.136 However, portions of a book of legal rulings known as Sefer Dinim (‘Book of rulings’), written in rabbinic Hebrew, are extant, and these include both rabbinic and non-rabbinic halakhot and doctrines, often exemplifying rabbinic biblical exegesis.137 Benjamin is claimed by later Karaites as one of their own, and the compilatory form of his work is similar to Anan’s Sefer Mitsvot. Compendia of legal rulings such as these have similarities in content and form to the post-Talmudic codelike works of the rabbis, such as Halahkot Pesukot, Halakhot Keṣuvot, and Halakhot Gedolot, although these rabbinic works lack individual author attribution.138 Their corporate

132 For the eighth century, see Cook, ‘The Opponents of the Writing of Tradition in Early Islam’, and Cook, ‘Anan and Islam’. In the ninth century, Ḥīwī al-Balkhī may echo Muʿtazilite ‘rationalist’ perspectives. See Gil, Jews in Islamic Countries in the Middle Ages, p. 322. 133 On this problem (known as laḥn in Arabic), see Ayoub, ‘Laḥn’. 134 Sībawayhi, al-Kitāb; Ibn al-Sikkīt, Iṣlāḥ al-mantiq. On the intellectual and social register of Ibn al-Sikkīt’s book, see Gruendler, ‘Aspects of Craft in the Arabic Book Revolution’, pp. 51–57. 135 Benjamin has been little studied by modern scholars. See Sklare, ‘Binyamin ben Mosheh Nahawendi’. 136 Nemoy, Karaite Antho­logy, p. 22. The manu­script fragment is Oxford, Bodleian Library, MS Heb.d. 64 on Genesis 2:9–17 and 6:5–6. See also the references in Zawanowska, The Arabic Translation and Commentary of Yefet ben ‘Eli the Karaite, pp. 17–18 and 99–100. 137 Al-Qirqisānī on Benjamin ben Moses al-Nahawendī: Kitāb al-Anwār wa-l-marāqib, i, pp. 13, 55–56. See Ben-Shammai, ‘Karaite Exegetes and their Rabbanite Environment’. 138 Although there are traditions associating individual rabbis with some of these works, their content and structure reflect traditional rabbinic compilation and corporate authorship. The genre of sifre

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

authorship may be a feature of a rabbinic culture that prized orality over writing. Both Anan and Benjamin, intellectually close to the rabbis in many ways, were nonetheless among the first Jews to write under their own name; and both upheld the individual responsibility of Jews to search carefully in the biblical text in order to understand and observe its teachings. These early examples of Jewish authorship parallel notions of authorship that were evolving among ninth-century Muslims. It is significant that early Karaites in the late ninth and tenth centuries continued to champion individual authorship, in reflection of their embrace of the new ways of reading and of individual interpretive responsibility. The third interpreter-leader is Abū ʿImrān al-Tiflīsī (also known as Mūsā al-Zaʿfarānī), who was from Baghdad but moved to Tbilisi in Georgia.139 The tenth-century Karaite Yefet ben Eli states that because of pressure from the rabbinic academies his followers had to migrate beyond the borders of the caliphate.140 The followers of Ismāʿīl al-ʿUkbarī claimed that Abū ʿImrān was a student of their leader. He did not write books on halakhah, but he did in a small way participate in writerly culture. Abū ʿImrān penned responses to Ḥīwī al-Balkhī (lahu jawāban) a century before any rabbinic scholar would do so. Although only a fragment of Ḥīwī’s work has survived, portions of it can be reconstructed, and these were written in the biblical question-and-answer format mentioned above in connection with Ismāʿīl al-ʿUkbarī. As observed by Judah Rosenthal, this question-and-answer form used by Ḥīwī, who was a younger contemporary of Ismāʿīl, shows parallels with the anti-Jewish polemical questions of the Zoroastrian Persian Škand Gumānīg Wīzār.141 Although the biblical question-and-answer genre may be associated with an elementary form of the new mode of reading Arabic, it had been used by Christians for centuries, including in contemporaneous Syriac writings.142 Both contexts suggest use of the codex, which Christians embraced almost from its beginnings, in support of synoptic and global readings of the biblical text. Accordingly, the question-and-answer genre, including the questions of Ḥīwī al-Balkhī, could function as obvious points of departure for heterodox Jewish interpreter-leaders, including two listed by al-Qirqisānī. Responses to Ḥīwī went on to become a popular subject among both rabbinic and non-rabbinic Jewish writers, including Saʿadya Gaon (882–942), the first full-fledged rabbinic exponent of ʿAbbāsid book culture.

139 140 141 142

miṣvot was continued by Karaites for centuries. See Danzig, Introduction to Halakhot Pesuqot with a Supplement to Halakhot Pesuqot, pp. 272–75, and Polliack, The Karaite Tradition of Arabic Bible Translation, pp. 22–36. Al-Qirqisānī on Abū ʿImrān al-Tiflīsī: Kitāb al-Anwār wa-l-marāqib, i, pp. 13–14, 57. Erder, ‘Abū ʿImrān al-Tiflīsī’. See Rosenthal, ‘Ḥiwi al-Balkhi’. For a recent brief overview, see Zawanowska, ‘Ḥīwī al-Balkhī’. For scholarship on Ḥīwī, see Sklare, Samuel ben Ḥofni Gaon and his Cultural World, p. 127 nn. 86, 87. On the Škand, see Thrope, ‘Contradictions and Vile Utterances’. See the chapters in Volgers and Zamagni ed., Erotapokriseis, especially ter Haar Romeny, ‘Questionand-Answer Collections in Syriac Literature’, pp. 145–63.

449

4 50

f r e d a s tr e n

Abū ʿImrān al-Tiflīsī also wrote ‘pages’ on the permissibility of eating meat (wa-lahu ayḍan awrāq fī iṭlāq al-laḥm), in opposition to Jews associated with the Mourners of Zion movement, which forbade the eating of meat and the drinking of wine while the Jerusalem temple lay in ruins.143 Like Ismāʿīl al-ʿUkbarī, Abū ʿImrān wrote for audiences whose positions on common sectarian matters (such as the questions of Ḥīwī, the consumption of meat and wine, and the perennially contentious subject of calendation) point to overlapping readerships. Al-Qirqisānī lists two other interpreter-leaders who were non-writers. He reports that Malik al-Ramlī proclaimed in Jerusalem that cocks used to be sacrificed in the temple, an interpretive move that took advantage of some ambiguous Torah verses. Little is known about Malik, about whom al-Qirqisānī declares, ‘this is the action of an ignorant man’.144 The other is Mīshawayh al-ʿUkbarī,145 whose halakhah broadly lowered social boundaries and would have been suited to the fluid and changeable identities of complicated urban milieus in Baghdad and other Islamic cities. However, Mīshawayh wrote no books. This may indicate that his movement was based on preaching and oral teaching rather than readerships. Furthermore, al-Qirqisānī states that there were no scholars or thinkers among his followers. As reported by al-Qirqisānī and others, Mīshawayh’s messianism and antinomianism, positions that are often paired in eschato­logical and millenarian teachings, recall the eighth-century Jewish activist messianic movements of the Shepherd, SeverusSerenus, Abū ʿĪsā al-Iṣfahānī, and Yudghān.146 Accordingly, Mīshawayh held most law in abeyance until the coming of the messiah, a position that offered a religious and communal identity with fewer requirements than others. Al-Qirqisānī states that he was ‘like one confused of mind’. Significantly, his movement persisted to the time of al-Qirqisānī and into the middle of the twelfth century.147 Sixth, and most consequential, is the final figure in al-Qirqisānī’s heresio­graphy, Daniel ben Moses al-Qūmisī (also known as al-Damāghānī), whose significant literary output was made up of biblical commentary — the new Jewish writing form — and other writings. These most likely included the famous epistle that called for Jews to immigrate to the Holy Land, perhaps the genesis of a fully articulated Karaite Judaism centred on Jerusalem.148 Exploiting the new way of reading and writing a generation before Saʿadya Gaon, the first rabbi to engage with Arabic book culture, al-Qūmisī is properly understood as the founder of Karaite Judaism. Meira Polliack writes that al-Qūmisī’s verse-by-verse commentaries ‘are characterized

143 144 145 146

Walfish, ‘The Mourners of Zion’. Al-Qirqisānī on Malik al-Ramlī: Kitāb al-Anwār wa-l-marāqib, i, pp. 14, 57. Al-Qirqisānī on Mīshawayh: Kitāb al-Anwār wa-l-marāqib, i, pp. 14, 57–58. See Astren, ‘Non-Rabbinic and Non-Karaite Religious Movements in the Medi­eval Islamic World’, pp. 610–18. 147 On Mīshawayh, see Erder, The Karaite Mourners of Zion and the Qumran Scrolls, pp. 167–307. 148 Al-Qirqisānī on Daniel al-Qūmisī: Kitāb al-Anwār wa-l-marāqib, i, pp. 14, 58–59, and throughout (see index). On Daniel al-Qūmisī, see Frank, ‘Karaite Exegesis’, pp. 112–14; Polliack, ‘Major Trends in Karaite Biblical Exegesis’; Gordon, ‘Does Scripture Really Only Have One Meaning?’; and Nemoy, ‘The Pseudo-Qumisian Sermon to the Karaites’.

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

by lexico­graphical and grammatical notes, semantic explications, analysis of the meaning of a word, phrase or verse according to their immediate and wider contexts, an awareness of thematic units, and general remarks on the literary structure and coherence of the biblical text’.149 Eschewing traditional rabbinic ways of reading Scripture, which were thematic, associative, and atomized, Daniel’s revolutionary commentary style reflected the new visual mode of reading. It framed the Hebrew Bible as a text whose specific language and sequential written character demanded consecutive and contextual linguistic analysis, sometimes using Arabic and Persian glosses.150 Al-Qirqisānī refers to him as a ‘learned philo­logist’ (ʿālim fī al-lugha).151 His Commentary on the Minor Prophets, written in Hebrew, is extant, and this work and the no longer extant commentaries of Dāʾūd al-Muqammaṣ are the earliest known verse-by-verse biblical commentary written by a Jew.152 Al-Qirqisānī says of Daniel: ‫وايضا فانه يكتب الى اصحابه الذين قد نسخوا كتبه يعرفهم ما قد ظهر له مما هو خالف لما كان‬ .‫دونه فى كتبه من قبل ويامرهم بتدوين ذلك فى كتبه‬

Indeed, he writes to pupils who have copied his books, informing them of any opinion which he may have formed that may differ from those given previously in his book, and instructs them to insert such an opinion in his book.153 Although al-Qirqisānī uses this description to lampoon Daniel for being prone to changing his mind, in fact al-Qirqisānī describes a method of publication that was serial and updatable and that maintained active engagement with its readership. This method would have been advantageous for yielding an open book, which allows segments of text to be written and rewritten, a form suitable to Daniel’s biblical commentary.154 In fact, ninth-century books could be published serially in fascicles, known as juzʾ.155 Furthermore, Daniel’s method of publication bypassed well-known proprietary practices in which a single pupil or copyist might have held or inherited ownership of another’s physical book and thereby controlled its access, transmission, and dissemination, whether through memorization, copying, or recitation (at a majlis or a ḥalqa). Such ownership would have stood in an intermediate position between oral textuality and the book as object and commodity.156 In contrast, proprietorship of Daniel’s writings was widely distributed ‘to pupils’, some of whom must have been located beyond Daniel’s local environment and 149 Polliack, ‘Major Trends in Karaite Biblical Exegesis’, p. 374. 150 Polliack, ‘Concepts of Scripture among Jews of the Medi­eval Muslim World’. 151 Al-Qirqisānī, Yaʿqūb al-Qirqisānī on Jewish Sects and Christianity, p. 94. 152 Al-Qūmisī, Pitron Shemeim Asar. 153 Al-Qirqisānī, Kitāb al-Anwār wa-l-marāqib, pp. 4–5. 154 See preliminary comments on the open text or open book in Marcus, Sefer Hasidim and the Ashkenazic Book in Medi­eval Europe, pp. 4–6. See also Ta-Shma, ‘The “Open” Book in Medi­eval Hebrew Literature’. 155 Gruendler, ‘Aspects of Craft in the Arabic Book Revolution’, pp. 38, 54. 156 Gruendler, ‘Aspects of Craft in the Arabic Book Revolution’, pp. 40–47.

45 1

452

f r e d a s tr e n

personal community. This method of publication is consistent with a network, not only of Daniel’s pupils, but of dispersed readerships and even communities. In this brief report we have the earliest explicit evidence of a Jewish species of ʿAbbāsid book culture. Its description of copying and updating of books points to nothing less than the publishing of books intended for wide dissemination and framed by a fully developed notion of authorship, including the authorial voice.157 It stands in contrast to practices and ideo­logies of orality still partially espoused by the rabbis (and some Muslims). And it strongly suggests the use of paper and the codex for the searching and pasting that constituted updating and serial publication. Daniel al-Qūmisī’s method of publishing is given credence by a report in al-Ṭabarī’s Tafsīr that describes the manner by which the prophetic pronouncements of Muḥammad came to be arranged into Qurʾānic sūras, or chapters. By the time of al-Ṭabarī (d. 923), a contemporary of al-Qirqisānī, the sūra was already an established part of the structure of the Qurʾānic text. And even though it was widely acknowledged that the prophetic pronouncements were not revealed in the sequence in which they appear in the Qurʾān, al-Ṭabarī credits the distribution of the revelations across the sūras to Muḥammad himself. Al-Ṭabarī writes about the Prophet: ‫فكان اذا نزل عليه الشئ دعا بعض من كان يكتب فيقول ضعوا هؤالء اآليات في السورة التي ذكر‬ .‫فيها كذا وكذا‬

And when something would be revealed to him, he called to some of those who were writing, and he would say, ‘Place these verses in the sūra in which is mentioned such and such’.158 Anachronistically, this report ascribes to Muḥammad a method of ʿAbbāsid-era book publication similar to that of Daniel al-Qūmisī. The Prophet, in the image of a divinely guided warrāq, directs copyist-assistants in the arrangement of an author’s text (God’s Qurʾān). In fact, arranging the contents of books is an artefact of the new, visual mode of reading, and among warrāqūn it constituted an area of expertise and professional specialization known as taṣnīf.159 Like Daniel’s pupils, Muḥammad’s assistants in this narrative pasted and arranged blocks of writing in a manner that would be cumbersome and costly if done using any writing medium other than the paper codex. Al-Qirqisānī states that Daniel al-Qūmisī ‘established a school of thought’ (taḥaqqaqa bi-madhhab).160 His ‘school’ consisted of networked followers who constituted readerships centred on his writing, and likely on communal activities 157 In al-Qūmisī, Pitron Sheneim Asar, p. 41, Daniel rejects an interpretation by writing, ‘Some people said […] but this is not the case’. Elsewhere, he uses ‘which I explained’ and ‘I would almost suggest’. Cited in Frank, ‘Karaite Exegesis’, p. 113 n. 11. 158 Al-Ṭabarī, Tafsīr al-Ṭabarī, pp. 98–99, cited in Flowers, ‘Writing and the Termino­logical Evolution of the Qurʾānic “Sūrah”’. 159 Gruendler, ‘Aspects of Craft in the Arabic Book Revolution’, p. 61. 160 Al-Qirqisānī, Kitāb al-Anwār wa-l-marāqib, i, p. 14, translated in al-Qirqisānī, Yaʿqūb al-Qirqisānī on Jewish Sects and Christianity, p. 104.

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

such as prayer and study. Whether al-Qūmisī was a jumper-floater himself, like his contemporary Dāʾūd al-Muqammaṣ, or simply a leader who may have attracted them, he was active in a milieu marked by such men, whose changing affiliations and communal commitments contributed to the nascent beginnings of Karaite Judaism.

Conclusion: From New Ways of Reading to Tenth-Century Karaite Judaism Drawing a line from Anan ben David to Daniel al-Qūmisī reveals the changing urban and intellectual social settings in which ninth-century Jewish sectarianism evolved. Whereas Anan held that some purity laws were compulsory only for those who were twenty years of age or older, Daniel’s notion of a Jewish community as voluntary association applied that standard to the entire Law.161 Voluntary associations like his came together out of the voluntary movement of urban Jews towards the rabbis or other religious leaders (or even from Christianity and Islam), mirroring the migrations to Baghdad and other reconstituted Muslim cities that shaped Jewish history in the late eighth and ninth centuries. At first, Daniel designated Anan as rosh ha-maskilim (‘chief of the wise’), but later, using a play on words, declared him rosh ha-kesilim (‘chief of the fools’).162 Daniel described Anan’s interpretive methods as taʾwīl, an Arabic term that often means ‘allegorical interpretation’ but in this instance refers to the rabbinic exegesis that Daniel eschewed in favour of reading Scripture using what we would describe today as systematic philo­logical-contextual literary analysis.163 In fact, Daniel is often concerned with the plain meaning of the text, as opposed to the figurative or allegorical interpretations favoured by the rabbis.164 Although he is aware of and borrows from Muslim grammarians and exegetes, who were developing their own systematic philo­logical-contextual approaches to the Qurʾān, he does so strategically to suit his own purposes.165 Beyond philo­logical-contextual approaches to the Hebrew Bible, the structured reasoning inherent in grammar and in seeking out linguistic context led him to apply Muʿtazilite rationalism in his biblical commentaries.166 (He may have learned this from the writings of Dāʾūd al-Muqammaṣ or from al-Muqammaṣ himself). All of this is in accord with the new ways of reading.

161 Al-Qirqisānī, Kitāb al-Anwār wa-l-marāqib, ii, p. 331. 162 Al-Qirqisānī, Kitāb al-Anwār wa-l-marāqib, i, p. 5, translated in al-Qirqisānī, Yaʿqūb al-Qirqisānī on Jewish Sects and Christianity, pp. 94–95. 163 Polliack, The Karaite Tradition of Arabic Bible Translation, p. 28 n. 22 and p. 30 n. 29. 164 The place of Daniel al-Qūmisī in the evolution of medi­eval Jewish understanding of the plain meaning (peshat) of the Hebrew Bible is discussed in Cohen, The Rule of Peshat; see index under ‘Daniel al-Qūmisī’. 165 On selective use of Muslim termino­logy and concepts by a Karaite in the generation after Daniel, see Andruss, Jewish Piety in Jerusalem, 74–129. 166 See Erder, ‘The Karaites and Muʿtazilism’.

45 3

4 54

f r e d a s tr e n

Whereas Anan’s anti-rabbinic stance remained grounded in rabbinized legal thinking and was written in the Aramaic of the rabbis, Daniel’s interpretations, written in Hebrew, were based on innovative methodo­logies of reading and exegesis. This methodo­logical move — away from memorization, recitation, and hearing towards seeing and reading visually from a written page — marks a point on a line drawn from the Masoretes to the Karaites by way of Ismāʿīl al-ʿUkbarī and the biblical question and answer texts.167 In this regard, some implications of Masoretic ‘professional reading’ and ʿAbbāsid book culture were realized in lesser measure by al-Qirqisānī’s pen-wielding sectarians, but Daniel is to be credited with radically transforming Jewish reading and the value of books in those dual contexts. In his wake, the Karaites would take matters even further by turning from writing in Hebrew to writing in Arabic, a preference that paradoxically situated their separatist project closer to Arabic linguistics and ʿAbbāsid intellectual culture.168 In the latter context, al-Qirqisānī notes that Daniel ‘has given reason its due’, but also that he is sometimes ‘dissatisfied with reason, disowning it and criticizing its practitioners’.169 As if reflecting the opposition of al-Kindī and Abū Maʿshar al-Balkhī, al-Qirqisānī opposes Daniel on the latter point, which mirrors the so-called traditionalist-rationalist disagreements among Muslims and the contentious issue of the use of outside knowledge, but which also pits old styles of biblical interpretation against the new ways of reading. The turn to new ways of reading and the attendant focus on the written text also helped displace the activistic and sometimes militant messianism of eighth-century Jewish sectarian leaders. What had been a sectarian focus on charismatic individuals and messianic expectations was displaced and in part redirected to the Hebrew Bible itself. Beyond that, al-Qūmisī harnessed the old messianic activism by calling for the relocation of Jews to Jerusalem and the Land of Israel without claiming that any particular man was the messiah.170 This itself was a messianic move, for the call to immigrate was framed as a requirement that needed to be fulfilled in order for the messiah to come. With that massive act of voluntary association — an extension of Jewish movements in the eighth and ninth centuries — Karaism rejected both the Babylonian Jewish indigeneity that Sherira Gaon narrativized and the diasporic Muslim city, incubator of the new way of reading and of sectarian socio-religious formation. And, by adopting a quasi-ascetic and legally rigorist lifestyle, these Karaites turned their backs on the promise of prosperity inherent in diasporic urban life. All of these orientations together simultaneously repudiated the rabbis

167 A nexus between the Masoretic project and Karaism has been noted. See Dotan, ‘De la Massora à la grammaire’, and Rustow, Heresy and the Politics of Community, 41–52. On Masora and grammar in tenth-century Karaism, See Sasson, ‘Masorah and Grammar’. 168 See Goldstein, ‘Arabic Composition 101’. 169 Al-Qirqisānī, Kitāb al-Anwār wa-l-marāqib, i, pp. 4–5, translated in al-Qirqisānī, Yaʿqūb al-Qirqisānī on Jewish Sects and Christianity, p. 94. 170 Mann, ‘A Tract by an Early Ḳaraite Settler in Jerusalem’, and Nemoy, ‘The Pseudo-Qūmisīan Sermon to the Karaites’.

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

and their diasporic orientation. For early Karaites, authentic indigeneity was to be found in Jerusalem and the Land of Israel. Following the call attributed to al-Qūmisī, and capitalizing on Ṭūlūnid and then Ihkshīdid political detachment of Palestine from Iraq (r. 868–905 and 935–969, respectively), tenth-century Karaites embarked on an effort at the wholesale reconstruction of a non-diasporic, non-rabbinic Judaism based on the transfer of new modes of reading and writing from Muslim Baghdad to a Jewishly imagined Jerusalem. There, over several subsequent generations, Karaite Jewish writerly culture would take full advantage of the new way of reading through the elaboration of grammar, biblical commentary, and legal interpretation.

Biblio­graphy Primary Sources Abraham ben Isaac of Narbonne, Sefer ha-Eshkol by Abraham ben Isaac of Narbonne, ed. by S. Albeck and H. Albeck, 2 vols (I. Feldman: Jerusalem, 1935) Abū Maʿshar al-Balkhī, The Great Introduction to Astro­logy by Abū Maʿšar, ed. by Keiji Yamamoto and Charles Burnett (Leiden: Brill, 2019) al-Balādhurī, Futūḥ al-buldān = The Origins of the Islamic State, ii, trans. by Francis Clark Murgotten (New York: Columbia University, 1924) al-Jāḥiẓ, al-Radd ʿalā al-Naṣāra, ed. by Joshua Finkel, Journal of the American Oriental Society, 47 (1927), 311–34 al-Jahshiyārī, Abū ʿAbd Allāh Muḥammad b. ʿAbdūs, Kitāb al-Wuzarāʾ wa-l-kuttāb, ed. by Hans von Mžik (Leipzig: Bibliothek Arabischer Historiker und Geo­graphen, 1926–1928) Josephus, Josephus, with an English Translation, Loeb Classical Library, 1 (London: William Heinemann, 1926) al-Kindī, Yaʿqūb b. Iṣḥāq, Al-Kindi’s Metaphysics: A Translation of Ya‘qub ibn Ishaq al-Kindi’s Treatise ‘On First Philosophy’, trans. by Alfred L. Ivry (Albany: State University of New York Press, 1974) Ibn Khurradādhbih, Kitāb al-Mamālik wa-l-masālik, ed. by M. J. De Goeje, BGA 6 (Leiden: Brill, 1889) Ibn al-Nadīm, Abū al-Faraj Muḥammad b. Isḥāq, The Fihrist of al-Nadīm: A Tenth-Century Survey of Muslim Culture, ed. and trans. by Bayard Dodge, 2 vols (New York: Columbia University Press, 1970) Ibn al-Sikkīt, Iṣlāḥ al-mantiq, ed. by Aḥmad Muḥammad Shākir and ʿAbd al-Salām Muḥammad Hārūn (Cairo: Dār al-Maʿārif, 1368/1949) Ibn Ṭāwūs, Raḍī al-Dīn, Faraj al-mahmūn fī taʾrīkh ʿulamāʾ al-nujūm (Najaf: Manshūrāt alMaṭbaʿa al-Ḥaydariyya, 1368/[1948]) al-Maqdisī (al-Muqaddasī), Shams al-Dīn Abū ʿAbd Allāh Muḥammad b. Aḥmad, Descrip­ tion of Syria, including Palestine, trans. by Guy Le Strange (London: Adelphi, 1886) —— , Aḥsan al-taqāsīm fī maʿrifat al-aqālīm, ed. by M. J. de Goeje (Leiden: Brill, 1906)

45 5

4 56

f r e d a s tr e n

al-Maqrīzī, Taqī al-Dīn Aḥmad, al-Mawāʿiẓ wa-l-iʿtibār bi-dhikr al-khiṭaṭ wa-l-athār, 2 vols (Bulaq: Cairo, 1853–1854) al-Muqammiṣ, Dāwūd b. Marwān, Dāwūd ibn Marwān al-Muqammiṣ’s Twenty Chapters (‘Ishrūn Maqāla’), ed. by Sarah Stroumsa (Leiden: Brill, 1989) Natronay bar Hilai, The Responsa of Natronay bar Hilai Gaon [in Hebrew], ed. by R. Brody, 2 vols ( Jerusalem: Ofeq Institute, 2012) Otsar ha-ge’onim: Thesaurus of the Geonic Responsa and Commentaries, Following the Order of the Talmudic Tractates, ed. by Benjamin M. Lewin, 13 vols (Haifa: Hebrew University Press Association, 1928–1943) al-Qifṭī, ʿAlī b. Yūsuf, Inbāh al-ruwāt ʿalā anbāh al-nuḥāt, ed. by Muḥammad Abū al-Faḍl Ibrāhīm, 3 vols (Cairo: Egyptian National Library Press, 1369–1371/1950–1955) al-Qirqisānī, Yaʿqūb, Kitāb al-Anwār wa-l-marāqib = Code of Karaite Law, ed. by Leon Nemoy, 5 vols (New York: Alexander Kohut Foundation, 1939–1943) —— , Yaʿqūb al-Qirqisānī on Jewish Sects and Christianity: A Translation of ‘Kitāb al-Anwār’, Book I, with Two Introductory Essays, ed. by Bruno Chiesa and Wilfrid Lockwood (Frankfurt: Peter Lang, 1984) al-Qūmisī, Daniel, Pitron Sheneim Asar (Commentarius in Librum Duodecim Prophetarum), ed. by I. D. Markon with notes by D. Z. Baneth and E. E. Urbach ( Jerusalem: Mekitse Nirdamim, 1957) Sherira Gaon, Iggeret Rav Sherira Gaon, ed. by Benjamin M. Lewin (Haifa: Itzkovsky, 1921) —— , The Iggeres of Rav Sherira Gaon, trans. by Nosson Dovid Rabinowich ( Jerusalem: Moznaim, 1988) Sībawayhi, Abū Bishr ʿAmr b. ʿUthmān, al-Kitāb, ed. by ʿAbd al-Salām Muḥammad Hārūn (Cairo: Maktabat al-Khānjī, 1966–1977) The Synodicon in the West Syrian Tradition, ed. and trans. by Arthur Vööbus, 4 vols (Leuven: Secrétariat du Corpus Scriptorum Christianorum Orientalium, 1975–1976) al-Ṭabarī, Muḥammad b. Jarīr, Taʾrīkh al-rusul wa-l-mulūk, ed. by M. J. de Goeje, 15 vols (Leiden: Brill, 1879–1901) —— , The History of al-Ṭabarī, xxxviii: The Return of the Caliphate to Baghdad, trans. by Franz Rosenthal (Albany: State University of New York Press, 1985) —— , The History of al-Ṭabarī, xxxvii: The ʿAbbāsid Recovery, trans. by Philip M. Fields, annot. by Jacob Lassner (Albany: State University of New York Press, 1987) —— , The History of al-Ṭabarī, xxxvi: The Revolt of the Zanj, trans. and annot. by David Waynes (Albany: State University of New York Press, 1992) —— , The History of al-Ṭabarī, xxviii: ʿAbbāsid Authority Affirmed, trans. and annot. by Jane Dammen McAuliffe (Albany: State University of New York Press, 1995) —— , Tafsīr al-Ṭabarī: Jāmiʿ al-bayān ʿan taʾwīl āy al-Qurʾān (Riyadh: Dār ʿĀlam al-Kutub li–l-Ṭibāʿa wa-l-Nashr wa-l-Tawzīʿ, 2003) Teshuvot ha-geonim mi-kitve-yad she-bi-genizat Cambridge [Gaonic responsa in the manu­ scripts of the Cambridge Geniza], ed. by S. Assaf ( Jerusalem: Ha-Madpis, 1942) Thaʿālibī, Abū Manṣūr ʿAbd al-Malik b. Muḥammad b. Ismāʿīl, Laṭāʾif al-maʿārif, ed. by P. de Jong (Leiden: Brill, 1897) —— , The Laṭāʾif al-maʿārif of Thaʿālibī: The Book of Curious and Entertaining Information, trans. by C. E. Bosworth (Edinburgh: Edinburgh University Press, 1968)

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

Theophylact Simocatta, The History of Theophylact Simocatta: An English Translation with Introduction and Notes, trans. by M. Whitby and M. Whitby (Oxford: Clarendon Press, 1986) al-Yaʿqūbī, Aḥmad b. Abī Yaʿqūb, Kitāb al-Buldān, ed. by M. J. de Goeje (Leiden: Brill, 1892) al-Yaʿqūbī, Aḥmad b. Abī Yaʿqūb, The Works of Ibn Wāḍiḥ al-Yaʿqūbī, ed. by Matthew S. Gordon, Chase F. Robinson, Everett K. Rowson, and Michael Fishbein, 3 vols (Leiden: Brill, 2018) Yāqūt b. ʿAbd Allāh al-Ḥamawī, Muʿjam al-buldān, ed. by F. Wüstenfeld, 5 vols (Leipzig: Brockhaus, 1866–1873) —— , Muʿjam al-udabāʾ, ed. by Iḥsān ʿAbbās (Beirut: Dār al-Gharb al-Islāmī, 1993) Secondary Works Achinstein, Susan, ‘Texts in Conflict: The Press and the Civil War’, in The Cambridge Companion to Writing of the English Revolution, ed. by N. H. Keeble (Cambridge: Cambridge University Press, 2001), pp. 50–68 Adams, Robert McCormick, Land behind Baghdad: A History of Settlement on the Diyala Plains (Chicago: University of Chicago Press, 1975) Adamson, Peter, Al-Kindī (Oxford: Oxford University Press, 2007) Ahuvia, Mika, ‘Jewish Towns and Neighborhoods in Roman Palestine and Persian Babylonia’, in A Companion to Late Ancient Jews and Judaism: Third Century bce to Seventh Century ce, ed. by Naomi Koltun-Fromm and Gwynn Kessler (Hoboken: Wiley Blackwell, 2020), pp. 33–52 Allard, Michel, ‘Les Chrétiens à Baġdād’, Arabica, 9.3 (1962), 375–88 Altaweel, Mark, and Andrea Squitieri, Revolutionizing a World: From Small States to Universalism in the Pre-Islamic Near East (London: UCL Press, 2018) Amabe, Fukuzo, Urban Autonomy in Medi­eval Islam: Damascus, Aleppo, Cordoba, Toledo, Valencia and Tunis (Leiden: Brill, 2016) Andruss, Jessica, Jewish Piety in Islamic Jerusalem: The Lamentations Commentary of Salmon ben Yerūḥīm (Oxford: Oxford University Press, 2023) Astren, Fred, Karaite Judaism and Historical Understanding (Columbia: University of South Carolina Press, 2004) —— , ‘Re-reading the Muslim Sources: Jewish History and the Muslim Conquests’, Jerusalem Studies in Arabic and Islam, 35 (2009), 83–130 —— , ‘Jacob al-Qirqisānī’, in the Encyclopedia of Jews in the Islamic World, ed. by Norman A. Stillman (Leiden: Brill, 2010), iv, pp. 136–40 —— , ‘The Gibeonite Gambit in the Middle Ages: Harrānians, Karaites, and Khaybarī Jews on the Margins of Islamic Society’, Journal of Medi­eval Worlds, 1.2 (2019), 1–23 —— , ‘Non-Rabbinic and Non-Karaite Religious Movements in the Medi­eval Islamic World’, in The Cambridge History of Judaism: The Medi­eval Era, v, Jews and Judaism in the Islamic World, Seventh through Fifteenth Centuries, ed. by Philip I. Lieberman (Cambridge: Cambridge University Press, 2021), pp. 606–33 Ayoub, Georgine, ‘Laḥn’, in Encyclopedia of Arabic Language and Linguistics, ed. by Kees Versteegh, 4 vols (Leiden: Brill, 2006–2009), ii, pp. 628–34

45 7

4 58

f r e d a s tr e n

Badrī, Muḥammad Fahad, al-ʿĀmma bi-Baghdād fī al-qarn al-khāmis al-hijrī (Baghdad: Matbaʿat al-Irshād, 1967) Balty-Guesdon, Marie-Geneviève, ‘Le Bayt al-hikma de Baghdad’, Arabica, 39 (1992), 131–50 —— , ’Bayt al-Ḥikmah et politique culturelle du caliphe al-Maʾmūn’, Medicina nei secoli Arte e Scienza = Journal of History of Medicine, 6 (1994), 275–91 —— , ‘La Maison de la Sagesse: une institution hors de l’histoire?’, in L’Islam medi­eval et terres chrétiennes, ed. by Max Lejbowicz (Villeneuve d’Ascq: Presses universitaires du Septentrion, 2009), pp. 85–98 Baron, Salo W., A Social and Religious History of the Jews, 18 vols (New York: Columbia University Press and Jewish Publication Society of America, 1952–1983) Baumgarten, Albert I., The Flourishing of Jewish Sects in the Maccabean Era (Leiden: Brill, 1997) Bavel, Bas van, The Invisible Hand? How Market Economies Have Emerged and Declined since ad 500 (Oxford: Oxford University Press, 2016) Bavel, Bas van, Michael Campopiano, and Jessica Dijkman, ‘Factor Markets in Early Islamic Iraq, c. 600-1100 AD’, Journal of the Economic and Social History of the Orient 57 (2014), 262–89 Beit-Arié, Malachi, Hebrew Manu­scripts of East and West: Towards a Comparative Codico­ logy, Panizzi Lectures (London: British Library, 1992) Ben-Shammai, Haggai, ‘Karaite Exegetes and their Rabbanite Environment’ [in Hebrew], Proceedings of the Ninth World Congress of Jewish Studies, Panel Sessions: Bible Studies and Ancient Near East ( Jerusalem: World Union of Jewish Studies, 1988), pp. 43–46 —— , ‘Between Ananites and Karaites: Observations on Early Medi­eval Jewish Sectarianism’, in Studies in Muslim-Jewish Relations, i (Chur, Switzerland: Harwood Academic Publishers, 1993), pp. 19–31 —— , ‘Return to the Scriptures in Ancient and Medi­eval Jewish Sectarianism and in Early Islam’, in Les retours aux Écritures: fondamentalismes présents et passés, ed. by E. Patlagean and A. Le Boulluec (Louvain: Peeters, 1993), pp. 319–39 Berkel, Maaike van, ‘Reconstructing Archival Practices in Abbasid Baghdad’, Journal of Abbasid Studies, 1 (2014), 7–22 Bessard, Fanny, Caliphs and Merchants: Cities and Economies of Power in the Near East (700–950) (Oxford: Oxford University Press, 2020) Bloom, Jonathan M., Paper before Print: The History and Impact of Paper in the Islamic World (New Haven, CT: Yale University Press, 2001) Bosworth, C. Edmund, ‘Ebn Ḵordāḏbeh, Abu’l-Qāsem Obayd-allāh’, in Encyclopaedia Iranica (1997), viii, pp. 37–38 Boyarin, Daniel, A Traveling Homeland: The Babylonian Talmud as Diaspora (Philadelphia: University of Pennsylvania Press, 2015) Brody, Robert, ‘Were the Geonim Legislators?’ [in Hebrew], Shenaton ha-mishpaṭ ha-ivri, 11–12 (1984–1986), 279–315 —— , The Geonim of Babylonia (New Haven, CT: Yale University Press, 1998) —— , ‘The Epistle of Sherira Gaon’, in Rabbinic Texts and the History of Late-Roman Palestine, ed. by Martin Goodman and Philip Alexander (Oxford: Oxford University Press, 2011), pp. 253–64

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

Bulliet, Richard W., Cotton, Climate, and Camels in Early Islamic Iran: A Moment in World History (New York: Columbia University Press, 2009) Butts, Aaron Michael, ‘The Classical Syriac Language’, in The Syriac World, ed. by Daniel King (London: Routledge, 2019), pp. 222–42 Cahen, Claude, ‘Mouvements populaires et autonomisme urbain dans l’Asie musulmane du Moyen Age’, Arabica, 5 (1955), 225–50, and 6 (1959), 25–56, 223–65 Campopiano, Michele, ‘Land Tenure, Land Tax and Social Conflictuality in Iraq from the Late Sasanian to the Early Islamic Period (Fifth to Ninth Centuries ce)’, in Authority and Control in the Countryside: From Antiquity to Islam in the Mediterranean and Near East (Sixth–Tenth Centuries), ed. by Alain Delattre, Marie Legendre, and Patra M. Sijpesteijn (Leiden: Brill, 2019), pp. 464–99 Chiesa, Bruno, ‘Ya‘qūb al Qirqisānī come fonte storiografica’, in Yaʿqūb al-Qirqisānī on Jewish Sects and Christianity: A Translation of ‘Kitāb al-Anwār’, Book I, with Two Introductory Essays, ed. by Bruno Chiesa and Wilfrid Lockwood (Frankfurt: Peter Lang, 1984), pp. 15–47 Cohen, Mordechai Z., The Rule of Peshat: Jewish Construction of the Plain Sense of Scripture and their Christian and Muslim Contexts, 900–1270 (Philadelphia: University of Pennsylvania Press, 2020) Conrad, Lawrence I., ‘The Arabs and the Colossus’, Journal of the Royal Asiatic Society, 3rd ser., 6.2 (1996), 165–87 Cook, Michael, ‘Anan and Islam: The Origins of Karaite Scripturalism’, Jerusalem Studies in Arabic and Islam, 9 (1987), 161–82 —— , ‘The Opponents of the Writing of Tradition in Early Islam’, Arabica, 44 (1997), 437–530 Danzig, Neil, Introduction to Halakhot Pesuqot with a Supplement to Halakhot Pesuqot [in Hebrew] (New York: Jewish Theo­logical Seminary of America, 1993) Daryaee, Touraj, Sasanian Persia: The Rise and Fall of an Empire (London: I. B. Tauris, 2009) Denoix, Sylvie, ‘Founded Cities of the Arab World from the Seventh to the Eleventh Century’, in The City in the Islamic World, ed. by Salma K. Jayyusi (Leiden: Brill, 2008), pp. 115–39 Di Branco, Marco, ‘Un’istituzione sasanide? Il Bayt al-ḥikma e il movimento di traduzione’, Studio Graeco-Arabica, 2 (2012), 255–62 Dotan, Aron, ‘De la Massora à la grammaire: les débuts de la pensée grammaticale dans l’Hébreu’, Journal Asiatique, 278 (1990), 13–30 Elman, Yaakov, ‘Orality and the Redaction of the Babylonian Talmud’, Oral Tradition, 14.1 (1999), 52–99 —— , ‘Babylonian Jews at the Intersection of the Iranian Economy and Sasanian Law’, in The Oxford Handbook of Judaism and Economics, ed. by Aaron Levine (Oxford: Oxford University Press, 2010), pp. 545–63 —— , ‘Shopping in Ctesiphon: A Lesson in Sasanian Commercial Practice’, in The Archaeo­ logy and Material Culture of the Babylonian Talmud, ed. by Markham J. Geller (Leiden: Brill, 2015), pp. 225–44 Erder, Yoram, ‘Abū ʿImrān al-Tiflīsī’, in Encyclopedia of Jews in the Islamic World, ed. by Norman A. Stillman (Leiden: Brill, 2010), i, p. 30

4 59

460

f r e d a s tr e n

—— , ‘The Karaites and Muʿtazilism’, in A History of Jewish-Muslim Relations: From the Origins to the Present Day, ed. by A. Meddeb and B. Stora (Princeton, NJ: Princeton University Press, 2013), pp. 778–87 —— , The Karaite Mourners of Zion and the Qumran Scrolls: On the History of an Alternative to Rabbinic Judaism (Turnhout: Brepols, 2017) Eshel, Ben-Zion, Jewish Settlements in Babylonia during Talmudic Times: Talmudic Onomasticon; Including Geo­graphical Locations, Historical Notes, and Indices of PlaceNames [in Hebrew] ( Jerusalem: Magnes Press, 1979) Ess, Josef van, Theo­logy and Society in the Second and Third Centuries of the Hijra: A History of Religious Thought in Early Islam, 5 vols (Leiden: Brill, 2019) Fiey, J. M., ‘Topo­graphie chrétienne de Mahozé’, L’Orient Syrien, 12 (1967), 397–420 —— , Chrétiens syriaques sous les Abbassides surtout à Bagdad (749–1258) (Leuven: Corpus Scriptorum Christianorum Orientalium, 1980) Fischel, Walter, ‘The Origin of Banking in Mediaeval Islam: A Contribution to the Economic History of the Jews of Baghdad in the Tenth Century’, Journal of the Royal Asiatic Society, 2 (1933), 339–52, 569–603 Flowers, Adam, ‘Writing and the Termino­logical Evolution of the Qurʾānic “Sūrah”’, in Scripts and Scripture: Writing and Religion in Arabia, ca. 500–700 ce, ed. by Fred M. Donner and Rebecca Hasselbach-Andee (Chicago: Oriental Institute, 2022), pp. 223–34 Fonrobert, Charlotte E., ‘From Separatism to Urbanism: The Dead Sea Scrolls and the Origins of the Rabbinic ‘Eruv’, Dead Sea Discoveries, 11.1 (2004), 43–71 —— , ‘The Political Symbolism of the Eruv’, Jewish Social Studies, 11.3 (2005), 9–35 Frank, Daniel, ‘Karaite Exegesis’, in Hebrew Bible, Old Testament, i: From the Beginnings to the Middle Ages (Until 1300), part ii: The Middle Ages, ed. by Magne Sæbø (Göttingen: Vandenhoeck & Ruprecht, 2000), pp. 110–28 Gafni, Isaiah, Land, Center, and Diaspora: Jewish Constructs in Late Antiquity (Sheffield: Sheffield Academic Press, 1997) —— , ‘The Political, Social, and Economic History of Babylonian Jewry, 224–638 ce’, in The Cambridge History of Judaism, iv, ed. by Steven T. Katz (Cambridge: Cambridge University Press, 2006), pp. 792–820 Gil, Moshe, ‘The Rādhānite Merchants and the Land of Rādhān’, Journal of the Economic and Social History of the Orient, 17.3 (1974), 299–328 —— , ‘The Exilarchate’, in The Jews of Medi­eval Islam: Community, Society, and Identity, ed. by Daniel Frank (Leiden: Brill, 1995), pp. 33–65 —— , ‘The Origins of the Karaites’, in Karaite Judaism: A Guide to its History and Literary Sources, ed. by Meira Polliack (Leiden: Brill, 2003), pp. 73–118 —— , Jews in Islamic Countries in the Middle Ages (Leiden: Brill, 2004) Goitein, Shelomo Dov, ‘The Rise of the Middle Eastern Bourgeoisie in Early Islamic Times’, Cahiers d’histoire mondiale, 3.3 (1957), 583–603 —— , ‘The Mentality of the Middle Class in Medi­eval Islam’, in Colloque sur la socio­logie musulmane: Actes, 11–14 septembre 1961 (Brussels: Centre pour l’Étude des Problèmes du Monde Musulman Contemporain, 1962), pp. 249–61 Goldstein, Miriam, ‘“Arabic Composition 101” and the Early Development of JudaeoArabic Bible Exegesis’, Journal of Semitic Studies 55.2 (2010), 451–78

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

Gordon, Nehemia, ‘Does Scripture Really Only Have One Meaning? A Study of Daniel al-Qumisi’s Exegetical Approach in “Pitron Shnem ʿAsar”’ [in Hebrew], Tarbiṣ, 76 (2007), 385–414 Graetz, Heinrich, Geschichte der Juden von den ältesten Zeiten bis auf die Gegenwart: Aus den Quellen neu bearbeitet, 4th edn (Leipzig: O. Leiner, 1909) Gross, Simcha, ‘When the Jews Greeted Ali: Sherira Gaon’s Epistle in Light of Arabic and Syriac Historio­graphy’, Jewish Studies Quarterly, 24.2 (2017), 122–44 Gruendler, Beatrice, ‘Book Culture before Print: The Early History of Arabic Media’, Occasional Papers (American University of Beirut, Department of Arabic), 2011 —— , ‘Aspects of Craft in the Arabic Book Revolution’, in Globalization of Knowledge in the Post-Antique Mediterranean, 700–1500, ed. by Sonja Brentjes and Jürgen Renn (London: Routledge, 2016), pp. 31–66 —— , The Rise of the Arabic Book (Cambridge, MA: Harvard University Press, 2020) Gutas, Dimitri, Greek Thought, Arabic Culture: The Graeco-Arabic Translation Movement in Baghdad and Early ‘Abbasid Society (London: Routledge, 1998) Harkavy, Alexander, ‘Netira und seine Söhne, eine angesehene jüdische Familie in Baghdad am Anfang des X. Jahrhunderts’, in Festschrift zum siebzigsten Geburtstage A. Berliner’s, ed. by A. Freimann and M. Hildesheimer (Frankfurt: J. Kaufmann, 1903), pp. 34–43 Herman, Geoffrey, A Prince without a Kingdom: The Exilarch in the Sasanian Era (Heidelberg: Mohr-Siebeck, 2012) —— , ‘Babylonia of Pure Lineage: Notes on Babylonian Jewish Toponymy’, in Sources and Interpretation in Ancient Judaism: Studies for Tal Ilan at Sixty, ed. by Meron M. Piotrkowski, Geoffrey Herman, and Saskia Dönitz (Leiden: Brill, 2018), pp. 191–228 Hezser, Catherine, Jewish Literacy in Roman Palestine (Tübingen: Mohr Siebeck, 2001) Hill, Christopher, The World Turned Upside Down: Radical Ideas during the English Revolution (London: Temple Smith, 1972) Janos, Damien, ‘Al-Maʾmūn’s Patronage of Astro­logy: Some Bio­graphical and Institutional Considerations’, in The Place to Go: Contexts of Learning in Baghdad, 750–1000 c.e., ed. by Jens Scheiner and Damien Janos (Princeton, NJ: Darwin Press, 2014), pp. 389–454 Kennedy, Hugh, ‘Military Pay and the Economy of the Early Islamic State’, Historical Research, 75 (2002), 155–69 —— , ‘How to Found an Islamic City’, in Cities, Texts and Social Networks 400–1500, ed. by C. Goodson, A. E. Lester, and C. Symes (Aldershot: Ashgate, 2010), pp. 45–63 —— , ‘The Feeding of the Five Hundred Thousand: Cities and Agriculture in Early Islamic Mesopotamia’, Iraq, 73 (2011), 177–99 —— , ‘Baghdad as a Center of Learning and Book Production’, in By the Pen and What They Write: Writing in Islamic Arts and Culture, ed. by Sheila Blair and Jonathan Bloom (New Haven, CT: Yale University Press, 2017), pp. 91–103 —— , ‘The Middle East in Islamic Late Antiquity’, in Fiscal Regimes and the Political Economy of Premodern States, ed. by Andrew Monson and Walter Scheidel (Cambridge: Cambridge University Press, 2018), pp. 390–403 Lassner, Jacob, The Topo­graphy of Baghdad in the Early Middle Ages: Texts and Studies (Detroit: Wayne State University Press, 1970)

461

462

f r e d a s tr e n

—— , ‘The ‘Abbasid Dawla: An Essay on the Concept of Revolution in Early Islam’, in Tradition and Innovation in Late Antiquity, ed. by F. M. Clover and R. S. Humphreys (Madison: University of Wisconsin Press, 1989), pp. 247–70 La Vaissière, Étienne de, Sogdian Traders: A History, trans. by James Ward (Leiden: Brill, 2005) Lazarus-Yafeh, Hava, Intertwined Worlds: Medi­eval Islam and Biblical Criticism (Princeton, NJ: Princeton University Press, 1992) Le Strange, Guy, Palestine under the Moslems: A Description of Syria and the Holy Land from a.d. 650 to 1500 (London: Palestine Exploration Fund, 1890) —— , Baghdad during the Abbasid Caliphate from Contemporary Arabic and Persian Sources (Oxford: Clarendon Press, 1900) Lieberman, Phillip I., The Fate of the Jews in the Early Islamic Near East: Tracing the Demo­ graphic Shift from East to West (Cambridge: Cambridge Univeristy Press, 2022) Louth, Andrew, ‘Fiunt, non nascuntur Christiani: Conversion, Community, and Christian Identity in Late Antiquity’, in Being Christian in Late Antiquity: A Festschrift for Gillian Clark, ed. by Carol Harrison, Caroline Humfress, and Isabella Sandwell (Oxford: Oxford University Press, 2014), pp. 109–19 Makdisi, George, The Rise of Humanism in Classical Islam and the Christian West (Edinburgh: Edinburgh University Press, 1990) Mann, Jacob, ‘The Responsa of the Babylonian Geonim as a Source of Jewish History, Part 1’, Jewish Quarterly Review, n.s., 7.4 (1917), 457–90 —— , ‘The Responsa of the Babylonian Geonim as a Source of Jewish History, Part 3’, Jewish Quarterly Review, n.s., 10.2–3 (1919–1920), 309–65 —— , ‘A Tract by an Early Karaite Settler in Jerusalem’, Jewish Quarterly Review, n.s., 12 (1922), 257–98 Marcus, Ivan G., Sefer Hasidim and the Ashkenazic Book in Medi­eval Europe (Philadelphia: University of Pennsylvania Press, 2018) Martín-Contreras, Elvira, ‘Safeguarding the Lord’s Word: The Work of the Masoretes in Palestine and Babylon’, in The Place to Go: Contexts of Learning in Baghdad, 750–1000 c.e., ed. by Jens Scheiner and Damien Janos (Princeton, NJ: Darwin Press, 2014), pp. 533–48 Martín-Contreras, Elvira, and Guadalupe Seijas de los Ríos-Zarzosa, Masora: la transmisión de la tradición de la Biblia Hebrea (Estella, Spain: Editorial Verbo Divino, 2010) McElligott, Jason, Royalism, Print and Censorship in Revolutionary England (Woodbridge: The Boydell Press, 2007) McKinney, Robert C., The Case of Rhyme versus Reason: Ibn al-Rūmī and his Poetics in Context (Leiden: Brill, 2004) Melchert, Christopher, ‘The Etiquette of Learning in the Early Islamic Study Circle’, in Law and Education in Medi­eval Islam: Studies in Memory of George Makdisi, ed. Joseph E. Lowry, Devin J. Stewart, and Shawkat M. Toorawa (Cambridge: E. J. W. Gibb Memorial Trust, 2004), pp. 33–44 Micheau, Françoise, ‘Baghdad in the Abbasid Era: A Cosmopolitan and MultiConfessional City’, in The City in the Islamic World, ed. by Salma K. Jayyusi (Leiden: Brill, 2008), pp. 221–45 Morony, Michael G., Iraq after the Muslim Conquest (Princeton, NJ: Princeton University Press, 1984)

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

Nemoy, Leon, Karaite Antho­logy: Excerpts from the Early Literature (New Haven, CT: Yale University Press, 1952) —— , ‘The Pseudo-Qumisian Sermon to the Karaites’, Proceedings of the American Academy for Jewish Research, 43 (1976), 49–105 Newman, Julius, The Agricultural Life of the Jews of Babylonia, between the Years of 200 c.e. and 500 c.e. (London: Oxford University Press, 1932) Nol, Hagit, Settlement and Urbanization in Early Islamic Palestine (7th–11th Centuries): Texts and Archaeo­logy Contrasted (London: Routledge, 2022) Nongbri, Brent, God’s Library: The Archaeo­logy of the Earliest Christian Manu­scripts (New Haven, CT: Yale University Press, 2018) Oppenheimer, Aharon, Babylonia Judaica in the Talmudic Period (Wiesbaden: Dr Ludwig Reichert, 1983) —— , ‘Purity of Lineage in Talmudic Babylonia’, in Manières de penser dans l’antiquité mediterranéenne et orientale: mélanges offerts à Francis Schmidt par ses élèves, ses collègues et ses amis, ed. by Christophe Batsch and Madalina Vârtejanu-Joubert (Leiden: Brill, 2009), pp. 145–57 Pingree, David, ‘Abū Maʿshar al-Balkhī, Jaʿfar ibn Muḥammad’, in Dictionary of Scientific Bio­graphy, i (New York: Charles Scribner’s Sons, 1970), pp. 32–39 Polliack, Meira, The Karaite Tradition of Arabic Bible Translation: A Linguistic and Exegetical Study of the Karaite Translations of the Pentateuch from the Tenth to the Eleventh Centuries c.e. (Leiden: Brill, 1997) —— , ‘Major Trends in Karaite Biblical Exegesis in the Tenth and Eleventh Centuries’, in Karaite Judaism: A Guide to its History and Literary Sources, ed. by Meira Polliack (Leiden: Brill, 2003), pp. 363–413 —— , ‘Concepts of Scripture among Jews of the Medi­eval Muslim World’, in Jewish Concepts of Scripture: A Comparative Introduction, ed. by Benjamin D. Sommer (New York: New York University Press, 2012), pp. 80–101 Roggema, Barbara, ‘ʿAlī ibn Yaḥyā ibn al-Munajjim’, Christian-Muslim Relations: A Biblio­ graphical History, i: 600–900, ed. by David Thomas and Barbara Roggema (Leiden: Brill, 2009), pp. 762–67 Rosenthal, Judah, ‘Ḥiwi al-Balkhi, a Comparative Study’, Jewish Quarterly Review, n.s., 38.3 (1948), 317–42; 38.4 (1948), 419–30; and 39.1 (1948), 79–94 Rustow, Marina, Heresy and the Politics of Community: The Jews of the Fatimid Caliphate (Ithaca, NY: Cornell University Press, 2008) —— , The Lost Archive: Traces of a Caliphate in a Cairo Synagogue (Princeton, NJ: Princeton University Press, 2020) Sabari, Simha, Mouvements populaires à Baghdad à l’époque ‘abbasside, ixe–xe siècle (Paris: Librairie d’Amérique et d’Orient, Adrien Maisonneuve, 1981) Saliba, George, Islamic Science and the Making of the European Renaissance (Cambridge, MA: MIT Press, 2007) Sasson, Ilana, ‘Masorah and Grammar as Revealed in Tenth Century Karaite Exegesis’, Jewish Studies, an Internet Journal 12 (2013), 1–36 Savant, Sarah Bowen, The New Muslims of Post-Conquest Iran: Tradition, Memory, and Conversion (Cambridge: Cambridge University Press, 2013)

463

464

f r e d a s tr e n

Scheiner, Jens, and Damien Janos, ‘Baghdād: Political Metropolis and Intellectual Center’, in The Place to Go: Contexts of Learning in Baghdad, 750–1000 c.e., ed. by Jens Scheiner and Damien Janos (Princeton, NJ: Darwin Press, 2014), pp. 1–46 Schoeler, Gregor, The Genesis of Literature in Islam: From the Aural to the Read (Edinburgh: Edinburgh University Press, 2002) —— , The Oral and the Written in Early Islam, ed. by James E. Montgomery (London: Routledge, 2006) Simonsohn, Uriel, ‘“Halting between Two Opinions”: Conversion and Apostasy in Early Islam’, Medi­eval Encounters, 19 (2013), 342–70 —— , ‘Conversion, Apostasy, and Penance: The Shifting Identities of Muslim Converts in the Early Islamic Period’, in Conversion in Late Antiquity: Christianity, Islam, and Beyond, ed. by Arietta Papaconstantinou (Farnham: Ashgate, 2015), pp. 197–215 Simpson, St John, ‘The Land behind Ctesiphon: The Archaeo­logy of Babylonia during the Period of the Babylonian Talmud’, in The Archaeo­logy and Material Culture of the Babylonian Talmud, ed. by Markham J. Geller (Leiden: Brill, 2015), pp. 6–38 Sklare, David E., Samuel ben Ḥofni Gaon and his Cultural World: Texts and Studies (Leiden: Brill, 1996) —— , ‘Binyamin ben Mosheh Nahawendi’, in The Oxford Dictionary of the Jewish Religion, ed. by Adele Berlin (Oxford: Oxford University Press, 1997), p. 130 —— , ‘Ninth-Century Judeo-Arabic Texts of Biblical Questions and Answers’, in Senses of Scripture: The Bible in Arabic among Jews, Christians, and Muslims, ed. by Miriam L. Hjälm (Leiden: Brill, 2017), pp. 105–24 Stern, David, ‘The First Jewish Books and the Early History of Jewish Reading’, Jewish Quarterly Review, 98.2 (2008), 163–202 —— , The Jewish Bible: A Material History (Seattle, WA: University of Washington Press, 2017) Stewart, Devin, ‘Abū al-Faraj Muḥammad ibn Isḥāq al-Nadīm’, in Essays in Arabic Literary Bio­graphy, 925–1350, ed. by Terri DeYoung and Mary St Germain (Wiesbaden: Harrassowitz, 2011), pp. 129–42 Stillman, Norman, ‘The Jew in the Medi­eval Islamic City’, in The Jews of Medi­eval Islam, ed. by Daniel Frank (Leiden: Brill, 1995), pp. 3–13 —— , The Jews of Arab Lands (Philadelphia: Jewish Publication Society, 1979) Stroumsa, Guy G., ‘On the Status of Books in Early Christianity’, in Being Christian in Late Antiquity: A Festschrift for Gillian Clark, ed. by Carol Harrison, Caroline Humfress, and Isabella Sandwell (Oxford: Oxford University Press, 2014), pp. 57–73 Stroumsa, Sarah, ‘From the Earliest Known Judaeo-Arabic Commentary on Genesis’, Jerusalem Studies in Arabic and Islam, 27 (2002), 375–95 Ta-Shma, Israel, ‘The “Open” Book in Medi­eval Hebrew Literature: The Problem of Authorized Editions’, Bulletin of the John Rylands Library, 75.3 (1993), 17–24 Tannous, Jack, The Making of the Medi­eval Middle East: Religion, Society, and Simple Believers (Princeton, NJ: Princeton University Press, 2018) ter Haar Romeny, Bas, ‘Question-and-Answer Collections in Syriac Literature’, in Erotapokriseis: Early Christian Question-and-Answer Literature in Context, ed. by Annelie Volgers and Claudio Zamagni (Leuven: Peeters, 2004), pp. 145–63 Teule, G. B., ‘Nonnos of Nisibis’, Christian-Muslim Relations: A Biblio­graphical History, i: 600–900, ed. by David Thomas and Barbara Roggema (Leiden: Brill, 2009), pp. 743–45

ʿ A bbā s id B o o k Cu lt u r e a n d N i n t h - C e n tu ry Je w i sh Sectari ani sm

Thrope, Samuel, ‘Contradictions and Vile Utterances: The Zoroastrian Critique of Judaism in the Škand Gumānīg Wīzār’ (unpublished doctoral dissertation, University of California, Berkeley, 2012) Toorawa, Shawkat, Ibn Abī Ṭāhir Ṭayfūr and Arabic Writerly Culture: A Ninth-Century Bookman in Baghdad (London: RoutledgeCurzon, 2005) Tucker, William F., Mahdis and Millenarians: Shīʿite Extremists in Early Muslim Iraq (Cambridge: Cambridge University Press, 2008) Volgers, Annelie, and Claudio Zamagni, ed., Erotapokriseis: Early Christian Question-andAnswer Literature in Context (Leuven: Peeters, 2004) Walfish, Barry Dov, ‘The Mourners of Zion (’Avelei Siyyon): A Karaite Aliyah Movement of the Early Arab Period’, in Eretz Israel, Israel and the Jewish Diaspora — Mutual Relations, ed. by M. Mor (New York: Lanham, 1991), pp. 42–52 Watson, Andrew, Agricultural Innovation in the Early Islamic World (Cambridge: Cambridge University Press, 1983) Webb, Peter, ‘“Foreign Books” In Arabic Literature: Discourses on Books, Knowledge and Ethnicity In the Writings Of Al-Ǧāḥiẓ’, Journal of Arabic and Islamic Studies 12 (2012), 16–55 Weitz, Lev E., Between Christ and Caliph: Law, Marriage, and Christian Community in Early Islam (Philadelphia: University of Pennsylvania Press, 2018) Wollenberg, Rebecca Scharbach, The Closed Book: How the Rabbis Taught the Jews (Not) to Read the Hebrew Bible (Princeton, NJ: Princeton University Press, 2023) Wood, Philip, The Chronicle of Seert: Christian Historical Imagination in Late Antique Iraq (Oxford: Oxford University Press, 2013) —— , The Imam of the Christians: The World of Dionysius of Tel-Mahre, c. 750–850 (Princeton, NJ: Princeton University Press, 2021) Yücesoy, Hayrettin, Messianic Beliefs and Imperial Politics in Medi­eval Islam: The ʿAbbāsid Caliphate in the Early Ninth Century (Columbia: University of South Carolina Press, 2009) Zawanowska, Marzena, ‘Ḥīwī al-Balkhī’, in Encyclopedia of Jews in the Islamic World, ed. by Norman A. Stillman (Leiden: Brill, 2010), ii, pp. 427–28 —— , The Arabic Translation and Commentary of Yefet ben ʿEli the Karaite on the Abraham Narratives (Genesis 11:10–25:18) (Leiden: Brill, 2012)

465

Index

Page numbers in italics refer to figures and tables. ʿAbbasid Caliphate see also Baghdad; book culture; specific caliphs Arabic literature during: 176 coins from: 260, 265, 267, 289, 297 intercultural exchanges and: 182 multilingualism in: 29, 152, 283 political fragmentation within: 184 n. 111 scribal education during: 132 translation movement and: 17–18, 167, 177 n. 68, 281, 382 ʿAbd Allāh b. ʿAbd al-Malik: 48, 51 n. 30, 54, 91 n. 14, 97, 105–07, 112–13, 129 ʿAbd Allāh b. ʿAmir: 342 ʿAbd Allāh b. Jābir: 67–68 ʿAbd Allāh b. Khāzim: 262, 268, 269, 269 n. 40, 272, 275 ʿAbd Allāh b. Muḥammad al-Azdī: 443 ʿAbd al-ʿAzīz b. Marwān: 51 n. 34, 53 n. 42, 54, 105–07, 111–13, 122, 134, 138 ʿAbd al-Ḥamīd b. Yaḥyā: 120, 131 ʿAbd al-Ḥamīd al-Kātib: 108 n. 105, 131, 132, 151–52, 176 ʿAbdishoʿbar Brikha: 377 ʿAbd al-Malik b. Marwān (Umayyad caliph) administrative policies of: 49, 50 n. 28 Arabicization under: 16–17, 167, 174 Athanasius and: 111–12, 138 coin reforms of: 52, 54, 55, 55–56 nn. 52 & 53, 175, 264, 274 n. 58, 288–89 death of: 112 dīwāns and: 45 nn. 6 & 7, 47, 48 n. 22, 53 n. 41, 75, 90, 102–03, 108–09, 119–20, 122, 140 gubernatorial appointments by: 51–52 n. 34 Jazīra region and: 145 Sarjūn and: 47, 47 n. 14, 75, 100 symbolic policies of: 51 n. 30 Abgarid dynasty: 203, 369, 369 n. 24 Abraha (king of Sanaa): 215, 223 Abū ʿImrān al-Tiflīsī: 449, 450 Abū Maʿshar al-Balkhī: 435–39

468

i n dex

Abū Muslim: 105, 142 Abū Qurra, Theodore: 299, 374–75 Achaemenids: 203, 330, 352 ʿAdī b. Zayd: 94, 217–18 Agapius of Manbij: 99 Agha, Saleh Said: 142 Akbari, Suzanne Conklin: 182 n. 105 Albania see Caucasian Albania Album, Stephen: 272 ʿAlī b. Abī Ṭālib (Rashidun caliph): 425, 425 n. 1, 426, 426 n. 5 allography: 196, 205, 403–13 Ammianus Marcellinus: 211, 218, 248 n. 231 Amorkesos see Imruʾ al-Qays ʿAmr b. al-ʿĀṣ: 61, 106 ANA languages see Ancient North Arabian languages Anan ben David: 434–35, 439, 446, 448, 449, 453, 454 Anastasius (Byzantine Empire): 216, 221, 221 n. 110 Ancient North Arabian (ANA) languages: 195, 202, 204–06, 225 Ancient South Arabian languages: 172, 225 al-Andalus administrative apparatus of: 122 n. 189 coins from: 17, 52 n. 39, 148, 181 dīwāns in: 148–49, 149 n. 344, 149 n. 346, 153 writing systems used in: 413 Anthony, Sean: 101, 111 Anthony of Baghdad: 412–13, 412 n. 38 Antony of Egypt: 412 n. 41 Antony of Tagrit: 375–76 Arab-Hephthalite coins: 260–76 Arabic inscriptions on: 260–63, 267–68, 275, 276 Bactrian inscriptions on: 260–64, 266, 267, 269, 272, 274–76 circulation and use of: 272–76 dating of: 264–68, 264 n. 21, 264 nn. 17 & 18, 270, 274–75 description of: 261–63, 262 Pahlavi inscriptions on: 261–64, 266, 267, 269, 271, 276 Second Fitna and: 261, 264, 264 n. 19, 268–70, 272–73, 275 Arabicization see also translation movement defined: 15, 30 literature review: 15–23 narratives of: 167–68, 168 n. 4 non-Muslims and: 25, 28, 90 progressive nature of: 76, 78 trade and: 30

index

Arabic language see also Arabicization; pre-Islamic Arabic administrative use of: 16–17, 49–56, 57, 59–60, 59–62, 64–75, 94, 134 alphabet development: 199, 201, 246 Aruč inscriptions in: 291, 291–92 Bactrian documents in: 307, 317–23, 319–21 book culture and literature in: 439 on coins: 17, 21, 22, 28, 28, 54–55, 181, 260–63, 267–68, 275, 276, 287–89 cosmopolitanism and: 24, 153, 168–71, 171 n. 32, 175–84, 183 n. 106, 275 dīwāns in: 44–49, 66–67, 66 n. 84, 75–77, 94–97, 99–107, 115, 123, 129, 142, 339 emergence of: 15–16, 18, 24, 172–73 n. 36, 172–74, 199–202, 206, 247–48 Hebrew compared to: 408–13, 409 as indigenous language of Palestine: 410 n. 27 Judaeo-Arabic: 29, 228, 375 n. 53, 408, 413 n. 45, 447 Latin, interplay with: 20–21 as lingua franca: 17, 91 n. 17 Mar Behnam inscriptions in: 21 non-Arab appropriation of: 24 numerals in: 73, 127 performance of power in Armenia: 282–84, 286–302 protocol texts in: 56, 57, 59, 59, 62, 69, 74–75 Qubbat al-Khazna documents in: 25, 180–81, 405–12 in Rasulid Hexaglot: 21 as Reichssprache: 19, 20, 286–94, 301 role in Islamic studies: 13–14 as sacred language of Islam: 27, 196, 200 standardization of: 206, 209, 209 n. 53, 247 in Syriac Christianity: 368–74, 371–72, 376–77 Syriac in relation to: 24, 206–09, 207–08 as vernacular: 172–75 Zuartʿnocʿ inscriptions in: 289–90, 290, 292 Arabization, defined: 15, 30 Aramaic language see also Syriac language Avroman documents in: 313 Babylonian: 203, 229 Coptic compared to: 19 CPA: 201, 203, 204, 210–11, 230, 233, 236 ideograms in Arabic text: 372 Imperial: 202, 203, 352 as indigenous language of Palestine: 410 n. 27 Jews and: 203, 226, 228–29, 426 for legal documents: 222 Nabataean: 172 n. 36, 202–05, 366–68

469

470

i n dex

Neo-Aramaic: 378–79 Qubbat al-Khazna documents in: 180 reduction to minority status: 329 for trade negotiations: 221 Arethas see al-Ḥārith b. Jabala Arian Christianity: 230, 238 Arjomand, Said Amir: 132 n. 244 Armenia Arabic inscriptions in: 289–92, 290–91, 294–96, 296 Arabic language and performance of power in: 282–84, 286–302 coins in: 287–89, 287 n. 23, 287 n. 26, 288 cosmopolitanism in: 301 elites in: 283–84, 286, 289, 292–94, 297, 301, 302 fires of Naxčawan in: 143, 143 n. 310, 144 Islamic conquest narratives: 284–86 multilingualism in: 282–84, 299 Persian language in: 283, 295 during Second Fitna: 286 Umayyad governorship in: 142–44, 286–89, 293 Armenian language alphabet development: 201 Caucasian Albania’s use of: 31 Christian communities and: 28–29, 233 dīwāns, use of term in: 116 Mar Behnam inscriptions in: 21 Parthian influences on: 351 Qubbat al-Khazna documents in: 180 in Rasulid Hexaglot: 21 for trade negotiations: 221 Armillas-Tiseyra, Magalí: 78 Aruč inscriptions: 291, 291–92 al-Aʿshā Maymūn: 247 al-ʿAskarī: 94, 98, 101, 104 Ašot Bagratuni: 150, 296–97, 299 Aspebetos: 210, 210 n. 55, 218, 233–34 assimilation: 69, 78, 139, 179–84 Astren, Fred: 25, 183, 425 Athanasius bar Gūmōyē: 105–07, 106 n. 95, 107 n. 98, 111–13, 122, 128, 130, 137–38, 151 Avroman documents: 313, 313, 350–51 Ayyubids: 416

index

Babylonian Aramaic: 203, 229 Babylonian Jews: 427, 429, 431–33, 431 n. 38, 454 Bacharach, Jere: 271, 289 Bactrian documents: 307–23 in Arabic language: 307, 317–23, 319–21 in Bactrian language: 309–17, 310, 315–16, 321 Bek family in: 307, 307–08 n. 3, 317, 320–23, 321 on dispute settlement: 314, 316, 317 on emancipation of slaves: 319, 320 judicial declarations: 317–19, 319 locations of: 309, 345 purchase contracts: 310, 311, 313–14 seals from: 308–23, 310, 315–16, 319–20 Bactrian language administrative use of: 50, 343 on Arab-Hephthalite coins: 260–64, 266, 267, 269, 272, 274–76 Bactrian documents in: 309–17, 310, 315–16, 321 for coin inscriptions: 344 New Persian and: 334, 343, 346, 350 spread of: 343–44 Bagarat Bagratuni: 299, 300 Baghdad see also book culture Christian populations in: 433, 434 economic growth in: 430–31, 433, 438 establishment of: 427, 428, 432 n. 43 intercultural exchanges in: 182 Jewish populations in: 429–35, 432 n. 43, 446 literacy in: 438, 440, 442 monasteries established in: 428 popular urban violence in: 444, 444 n. 116 rabbinical academies in: 427, 427 n. 7 translation movement in: 17–19 urbanization in: 430–32, 434 al-Baghdādī, ʿAbd Allāh b. ʿAbd al-ʿAzīz: 96, 98, 119 n. 166, 132 Bagnall, Roger: 410 n. 30 Bakr b. Wāʾil: 268, 269 al-Balādhurī Ibn al-Nadīm on: 125 n. 200 on Iraqi dīwān translation: 46 n. 9, 90, 94, 98, 125, 129, 140, 141 on Kharijite rebellion: 144–45 multilingual sources used by: 282 on Syrian dīwān translation: 47, 75, 100, 104, 109, 114, 137

47 1

472

i n dex

Bardaiṣan (Syriac Christian): 369 Bar Hebraeus: 377, 379, 379 n. 79 Baron, Salo W.: 434, 439 Bashir, Shahzad: 168–69 n. 7 Bassal, Ibrahim: 407 Bates, Michael: 142–43, 260, 287 n. 26 Baumgarten, Albert: 438, 439 Baumstark, A.: 374 n. 48 Becker, Alton: 176 Beit-Arié, Malachi: 408 n. 15, 447 Bek family: 307, 307–08 n. 3, 317, 320–23, 321 Benjamin ben Moses al-Nahawendī: 448–49 Berbers: 69 n. 98, 147, 148 Berkes, Lajos: 136 Bible see also Hebrew Bible codices of: 403 n. 2, 408 n. 18, 447–49 Ibn al-Ṭayyib’s commentary on: 377 religious connotations of names from: 410–11 n. 30 translations of: 201, 226, 235, 374 n. 48, 375 n. 53, 403 n. 2, 407, 408, 413 Blau, J.: 375 n. 53 Blois, François de: 309 book culture: 439–55 see also literacy Arabic literature and: 439 bookseller’s markets and: 442–44 intellectual exchange and: 440, 443 Jews and: 440, 445–55 libraries and: 436, 439 multigraphism and: 403 papermaking and: 426, 440–43 translation movement and: 439 Borrut, Antoine: 13, 24, 69, 153, 167 Bouderbala, Sobhi: 107, 118, 123, 132 n. 248 Brody, Robert: 427 n. 7 Buddhists and Buddhism: 168, 168 n. 6, 170 n. 17, 309, 440 Bukhārkhudā coins: 267, 267 n. 31 Burbank, Jane: 180 n. 88 Butts, Aaron Michael: 25, 365 Byzantine Empire see also specific rulers church and state diplomacy in: 237–39 civil languages in: 134, 209–11 coins from: 54, 55, 55–56 n. 53, 264, 271 n. 44, 273–74 n. 58 fiscal cycle used by: 149 religious language in: 236

index

Caucasian Albania: 31, 142–44, 150, 201 Ceballos, Manuela: 13 Chalcedonian Church: 210, 232, 240–43, 299 Christian Palestinian Aramaic (CPA): 201, 203, 204, 210–11, 230, 233, 236 Christians and Christianity see also monasteries; Syriac Christianity as administrative elites: 138, 139 Arabicization of: 25, 28 Arian: 230, 238 in Baghdad: 433, 434 baptism and: 232, 240, 435, 435 n. 57, 437 book culture and: 440 church councils: 232–35 conversion to: 230, 231, 233, 238, 240, 241, 436 Ethiopic: 215 Greek: 180, 230, 410 languages used by: 203–04, 230–37 missionary efforts: 173, 230, 232, 240 persecution by al-Walīd: 99 politico-religious conflicts with Jews: 241 in pre-Islamic period: 197, 200, 210, 230–37 Qurʾānic passages and: 236–37 religious connotations of multilingualism: 26–29 as scribes: 47–48, 48 n. 23, 73 n. 112, 75, 101–02, 127 scriptures and liturgy: 235–36 in shared space with Islam: 21–22 in translation movement: 19 Church of the East: 210, 232–35, 239–40, 244–45, 248, 375–77, 380, 427–28 Classical New Persian (CNP): 330–31, 334, 340, 340 n. 63, 346, 352–53 Cobb, Paul: 91 n. 15 codices biblical: 403 n. 2, 408 n. 18, 447–49 book culture and: 442, 452 Greek dīwāns and: 116 n. 145 marginal notes: 414, 415–17 Masoretic project and: 447 n. 128 Qurʾānic: 28, 448 coins see also Arab-Hephthalite coins ʿAbbasid: 260, 265, 267, 289, 297 from al-Andalus: 17, 52 n. 39, 148, 181 Arabic language on: 17, 21, 22, 28, 28, 54–55, 181, 260–63, 267–68, 275, 276, 287–89 Bactrian language on: 344

47 3

474

i n dex

Byzantine: 54, 55, 55–56 n. 53, 264, 271 n. 44, 273–74 n. 58 countermarks on: 263, 272–73, 272 n. 52, 275, 276 Ghaznavid: 28, 28 Greek language on: 54–55 iconography on: 52, 55 n. 53, 259, 272 n. 48, 276 from Iranian regions: 141, 260, 344–45 Latin language on: 17, 17, 17 n. 12, 28, 54, 147, 181 Middle Persian on: 17, 28, 147, 287–88, 341–42, 344, 346 Mongol: 21, 22 North African: 55, 147–48 Pahlavi on: 261–64, 266, 267, 269, 271, 276, 344–46 performance of power through: 287–89 Qurʾānic phrases on: 54, 55, 288 role in Islamic studies: 259–60 Roman: 54, 55 n. 53 Sanskrit on: 28, 28 Sasanian: 54, 55 n. 53, 260–62, 264–68, 271–76, 287–88 in trade: 273, 273 n. 56 Umayyad reforms: 17, 17, 52–55, 53 n. 40, 55–56 nn. 52 & 53, 68 n. 95, 113, 175, 264–66, 273–74 n. 58, 288–89 communal identity: 31, 177, 450 Conrad, Lawrence: 107, 110 Constantius II (Byzantine Empire): 230 conversion to Christianity: 230, 231, 233, 238, 240, 241, 436 to Islam: 27, 78, 131, 265, 286, 299, 339, 431, 437 to Judaism: 226, 413 n. 41 language change and: 178 n. 78 prior texts and: 176 translation and: 175, 178 to Zoroastrianism: 214 Cooper, Frederick: 180 n. 88 Coptic language administrative use of: 48, 50, 54, 60–62, 63, 69, 70, 72, 91, 134–35 n. 259, 134–36, 152 Aramaic compared to: 19 Christian communities and: 233 dīwāns in: 48, 107, 115, 129 numerals in: 127 Qubbat al-Khazna documents in: 180 reduction to minority status: 329

index

cosmopolitanism: 167–84 analytical value of: 169–70 Arabic: 24, 153, 168–71, 171 n. 32, 175–84, 183 n. 106, 275 Armenian: 301 assimilation and: 69, 78, 139, 179–84 Buddhist: 168, 168 n. 6 formation processes: 20–21, 168–75 Greek: 178, 180 intercultural exchanges and: 182 Latin: 172, 182 literary networks and: 175, 178 manuscript culture and: 180–82 Persian: 168, 168–69 n. 7, 330, 330 n. 7 prior texts and: 175–78 role in Islamic studies: 22–23, 168–69 Sanskrit: 23, 167–72, 175, 178, 182–83 of Sasanian Empire: 353 subordination and: 179–81, 183–84 textual foundations of: 175–82 universalist dimension of: 169, 179 vernacular languages and: 134 n. 258, 171–72, 171 n. 26, 183, 353 Council of Chalcedon: 204, 240 CPA see Christian Palestinian Aramaic Cribb, Joe: 259 Cromwell, Jennifer: 135–36, 136 n. 266 culture see also book culture articulation of: 169, 180 boundaries of: 24, 169, 181 court: 72, 178 literary: 25, 70, 182 manuscript: 180–82 political: 300 power of: 184 rabbinic: 449 scribal: 61 writerly: 440, 445, 446, 449, 455 written: 134, 152, 232, 234, 235, 417 cuneiform: 196, 199, 314 n. 21, 367 Cyrus of Edessa: 244

47 5

476

i n dex

Dadanitic language: 196, 204–05 Dādūye al-Muqaffaʿ: 131, 139, 151 al-Ḍaḥḥāk b. ʿAbd al-Raḥmān: 112, 112 n. 129, 138, 145 n. 326 Dari language: 336–40, 343, 345, 346, 353 Debié, Muriel: 24, 101 n. 69, 138, 195 Derēn (Grigor-Derenik Arcruni): 297–98, 300 Déroche, François: 209 al-Dhahabī: 98, 137 Dhēwāshtīch (Sogdian ruler): 64–65, 68–69, 134 n. 254, 348 dhimmīs (protected peoples): 127, 430, 441 Dhū Nuwās (Ḥimyarite king): 215, 227, 241, 243 diglossia: 171 n. 26, 202, 210, 236, 378–79, 378 n. 76, 387, 389, 389 n. 125 Dionysius of Tell-Maḥrē: 111, 112, 138 diplomacy: 69, 197, 215–19, 237–47 dīwāns (state registers) see also dīwān translations in Arabic: 44–49, 66–67, 66 n. 84, 75–77, 94–97, 99–107, 115, 123, 129, 142, 339 burning of: 109, 109–10 n. 114 in Coptic: 48, 107, 115, 129 in Greek: 45 n. 7, 47, 47 n. 14, 66, 66 n. 84, 75, 100, 107, 115, 122–24, 129 multilingualism of: 66–67, 66 n. 84, 129 papermaking for: 441, 442 in Persian: 46, 49, 66, 66 n. 84, 104, 115 pre-Islamic utilization of: 44, 66 uses of term: 44–45, 114–24, 114 n. 135, 118–19 n. 165, 118 n. 162 wealth distribution through: 430 dīwān translations dating of: 109–14, 151 economic incentives and: 136–39 Egyptian narrative: 44, 48, 75, 91 n. 14, 105–07, 110–13, 129, 150–53 historicity of accounts: 107–09 Iraqi narrative: 45 n. 6, 46, 46 n. 9, 73, 73 n. 112, 90, 94–97, 98, 110, 113, 125, 127, 129, 140, 141, 150–51 Khurasani narrative: 49, 104–05, 110, 113, 140–41, 151 mapping: 140–50 mawlās and: 46, 91, 102 numerals in: 45 n. 6, 48 n. 23, 73, 73 n. 112, 75, 127–28 personnel changes and: 49, 75–77, 90, 104–05, 139, 151 Syrian narrative: 45 n. 6, 47–48, 53 n. 41, 73, 75, 76 n. 121, 90, 99–103, 104, 109–10, 113, 114, 140, 150–53 verbs used in reference to: 124–26, 125–26 n. 206 working and training in: 129–33, 129 n. 231, 132 n. 248 Dīwāshtīch (Sogdian ruler): 266

index

Djaït, Hichem: 146 Donner, Fred: 89, 177 D’Ottone Rambach, Arianna: 25, 180, 403 double document method: 311–13, 312 double-nested diglossia: 389 n. 125 Drasxanakertcʿi, Yovhannēs: 293 Drory, Rina: 176 Dyophysites: 143, 240, 373 n. 40, 380, 380 n. 84 Early New Persian (ENP): 329–37, 339–42, 350, 352–54 Eastern Roman Empire see Byzantine Empire East Syriac Christianity: 243, 246, 380 economic growth: 430–31, 433, 435, 438, 439 economic power: 276, 428 Egeria (pilgrim): 203, 235–36 Egypt administrative documents in: 56–62, 57–60, 63, 67–70, 72, 134–36 administrative policy implementation in: 53–54, 53–54 n. 42 Arabicization of Christian communities in: 28 arithmetical calculations in: 70–71 n. 101 dīwān administrators in: 105, 120–24 dīwān translation narrative: 44, 48, 75, 91 n. 14, 105–07, 110–13, 129, 150–53 double document method in: 311, 312 Mamluk Sultanate: 21 n. 36, 22 multilingualism in: 56–62, 67–72, 134–36, 134 n. 258 pre-Islamic inscriptions in: 174 sealing practices in: 319–20 n. 33, 322 n. 35 writing systems used in: 413 elites administrative: 75, 90–91, 114, 124, 130, 138–39, 152 Armenian: 283–84, 286, 289, 292–94, 297, 301, 302 cosmopolitanism and: 168–69 n. 7, 170, 179 imperial diplomacy and: 219 language change and: 24, 25 literacy of: 152, 197 political: 49, 108 n. 106, 289 in pre-Islamic period: 197, 214–15, 219, 230, 231 relationship with caliphal courts: 51, 52 religious: 197, 293 in Sasanian Empire: 339 secretarial: 129, 133, 136, 144 ENP see Early New Persian

47 7

478

i n dex

Ess, Josef van: 443 Ethiopic language: 209–10, 215, 221, 233, 243 Eusebius of Caesarea: 235 Feissel, Denis: 221 Fenina, Abdelhamid: 68 n. 95 Fenwick, Corisande: 146–47 Ferguson, C. A.: 378 n. 76, 387 Fournet, Jean-Luc: 134 Fourth Fitna: 298 Gabrieli, Francesco: 132 n. 244 Galadza, Daniel: 416 Garsoïn, Nina: 301 Geertz, Clifford: 184 Ge’ez language: 215 Geniza documents: 30, 125 n. 201, 447 George, Alain: 411, 412 n. 38 Georgian language: 28–29, 180, 201, 233, 351 Ghassānids: 205, 248, 248 n. 231 Ghāzān Khān (Mongols): 21, 21 n. 35 Ghaznavids: 28, 28, 340, 347 Ghūrak (Sogdian ruler): 266, 348 Gil, Moshe: 433, 433 n. 51 Gilbert, Claire: 22 Göbl, Robert: 262, 264, 269 Goodwin, Tony: 272 Graetz, Heinrich: 432 n. 43 Greek language Avroman documents in: 313 for church and state diplomacy: 238, 239, 242, 243 civil/administrative use of: 50, 54, 56–62, 57–58, 60, 67–70, 72–75, 91, 91 n. 15, 134–36, 152, 173, 209–11 on coins: 54–55 cosmopolitanism and: 178, 180 dīwāns in: 45 n. 7, 47, 47 n. 14, 66, 66 n. 84, 75, 100, 107, 115, 122–24, 129 for imperial diplomacy: 215–18 for legal documents: 221–23 Middle Aramaic dialects subordinate to: 203 as military language: 219, 220 Muʿāwiya’s use of: 16, 16 Najrān dossier and: 243

index

Nessana documents in: 18, 123 n. 192 numerals in: 73, 75, 127 pre-Islamic inscriptions: 196 protocol texts in: 56, 57, 62, 69, 74 Qubbat al-Khazna documents in: 17, 180–81, 405–7, 406 n. 10 in Rasulid Hexaglot: 21 religious uses of: 204, 210, 226, 230–37 for trade negotiations: 221 translation into Arabic: 17–19 Greenwood, Timothy: 284, 291–92 Grévin, Benoît: 20, 153, 168, 172 n. 36, 173 Griffith, Sidney H.: 374 n. 48, 409 n. 23, 410, 412 n. 39, 412 n. 41 Gross, S.: 371 Gruendler, Beatrice: 443 Gutas, Dimitri: 17, 18, 137, 167, 383 n. 100 Ḥafṣ b. al-Walīd: 121 al-Ḥajjāj b. Yūsuf: 46, 95, 95 n. 43, 97, 109, 116, 127, 129 Hałbat monastery (Armenia): 295–96, 296 Hämeen-Anttila, Jaakko: 125 n. 200, 128 al-Ḥārith b. Jabala: 210, 216, 216 nn. 89 & 90, 218, 242 Hārūn al-Rashīd (ʿAbbasid caliph): 441 Hasan al-Basri: 94 n. 30, 339 Ḥassān b. Nuʿmān al-Ghassānī: 147 Haug, Robert: 24, 259 Hawting, Gerald R.: 118 n. 162 al-Haytham b. ʿAdī: 144 Hayya Gaon: 437 Hebrew Bible: 26, 226, 372, 403 n. 2, 411, 446–47, 451, 453–54 Hebrew language Arabic compared to: 408–13, 409 as indigenous language of Palestine: 410 n. 27 Judaeo-Arabic translations of: 29 for legal documents: 222 in pre-Islamic period: 226–28 Qubbat al-Khazna documents in: 180 as sacred language of Judaism: 226 Heidemann, Stefan: 271, 271 n. 44 Hephthalites see also Arab-Hephthalite coins Sasanian conquest of: 261 n. 7, 334 Second Fitna and: 268–70 trade patterns involving: 273, 273 n. 55

47 9

480

i n dex

Herodotus: 200, 367 Hexaglot (glossary): 21 Hieronymus of Cardia: 203 Hijri calendar: 59, 263, 263 n. 15, 264, 266–71, 266 n. 29, 274 Hill, Christopher: 438, 439 Ḥimyarites: 196, 214–15, 219–20, 226–28, 230–31, 248 Ḥirta/al-Ḥīra language use at: 212, 380–82 literate culture of: 244–47 Miaphysites in: 245, 380 n. 84 monasteries of: 231, 244–46 Hishām b. ʿAbd al-Malik (Umayyad caliph): 48–49, 51 n. 30, 75, 101, 105, 132, 293 Hishām b. Muḥammad al-Kalbī: 245 Hismaic language: 172, 195, 196, 202, 205, 206, 366–67 Ḥīwī al-Balkhī: 449, 450 Hjälm, Miriam L.: 405–6, 412 n. 38 Hopkins, S.: 375 n. 53 Hormizd IV (Sasanian Empire): 217, 218 Hoyland, Robert: 14 n. 2, 19, 20, 91 n. 13, 172 n. 36, 173–74, 365–66, 383 n. 100 Ḥunayn b. Isḥāq al-ʿIbādī: 25, 245, 366, 382–88, 384 n. 104, 387 n. 113 Ibn ʿAbd al-Ḥakam: 48, 91 n. 14, 106–07, 107 n. 98, 147 Ibn ʿAbd Rabbihi: 96–97, 98, 101, 104, 119–20, 122 Ibn ʿAsākir on ʿAbd al-Malik and Sarjūn: 47 n. 14 on dīwān administrators: 121 on Iraqi dīwān translation: 94, 98, 137 on Syrian dīwān translation: 76 n. 121, 90, 101–03, 104, 110, 137 Ibn Aʿtham al-Kūfī: 143–45 Ibn al-Athīr: 93 n. 26, 97, 98, 108 n. 108 Ibn Durayd: 98 Ibn al-Faqīh: 98 Ibn Ḥawqal: 283 Ibn Khaldūn: 98, 100, 104, 127, 129, 140, 441 Ibn al-Khaṭīb: 148 Ibn Khurradādhbih: 29–30, 335, 433 Ibn al-Muqaffaʿ: 131–33, 132 n. 244, 139, 151, 177, 335–40, 343 Ibn al-Nadīm on ʿAbd al-Malik and Sarjūn: 47 n. 14 on al-Balādhurī as translator: 125 n. 200 dīwāns, use of term by: 116 on Ḥunayn b. Isḥāq al-ʿIbādī: 387 n. 113

index

on Iraqi dīwān translation: 45 n. 6, 46 n. 9, 73 n. 112, 94, 98, 125 on Persian language: 335, 337–40, 351 on Syrian dīwān translation: 48, 100–01, 104 Ibn Qutaybah: 30 Ibn al-Rāwandī: 443 Ibn Rusta: 142 Ibn al-Ṭayyib: 376–77 Ibn al-Ṭiqṭaqā: 97, 98 Ibn Yarbūʿal-Fazārī: 48, 48 n. 21, 72 Ibn Yūnus: 107 n. 98, 118 identity Andalusi: 181–82 Arab: 30, 234, 247, 248, 381 Armenian: 31, 201, 301 communal: 31, 177, 450 cultural: 220, 408 formation of: 25, 201 Hephthalite: 272, 276 language and: 19, 30–31, 78 religious: 104, 200 Imbert, Frédéric: 181 Imperial Aramaic: 202, 203, 352 Imruʾ al-Qays: 205, 210, 210 n. 55, 212, 216 n. 89, 230, 237–38, 247, 368 Iran coins from: 141, 260, 344–45 dīwāns in: 142, 153 literary networks in: 178–79 revival of Middle Persian in: 69 n. 98 sealing practices in: 317, 318, 322, 322 n. 5 Iraq see also Baghdad dīwān administrators in: 89, 121, 131, 136–37 dīwān translation narrative: 45 n. 6, 46, 46 n. 9, 73, 73 n. 112, 90, 94–97, 98, 110, 113, 125, 127, 129, 140, 141, 150–51 Masoretic project in: 446–47 multilingualism in: 29 papermaking in: 440–43 pre-Islamic inscriptions in: 174 writing systems used in: 413 Isḥāq b. Ṭulayb: 104 Islam and Muslims see also Qurʾān Arabic as sacred language of: 27, 196, 200 book culture and: 440

481

482

i n dex

conversion to: 27, 78, 131, 265, 286, 299, 339, 431, 437 as dīwān administrators: 49, 75–76, 76 n. 121 integration into Late Antiquity: 14, 183 oral traditions in: 195, 196, 239 n. 198 on Persian language: 335–40 religious connotations of multilingualism: 27–28 in shared space with Christianity: 21–22 Shīʿites: 426 n. 5, 431, 444 Islamization: 51 n. 30, 76–77, 175, 289, 308, 323, 434, 438 Ismāʿīl al-ʿUkbarī: 446–49, 454 Iṣṭakhrī: 283 Jabala b. Sālim: 132, 151 Jacob of Serugh: 240, 241, 373 Jāhiliyya: 176, 200 al-Jāḥiẓ: 125–26 n. 206, 433, 433 n. 44, 441, 442 al-Jahshiyārī on ʿAbd al-Malik and Sarjūn: 47 n. 14 on Athanasius: 112, 137–38 dīwāns, use of term by: 114–16, 118–19 n. 165 on Iraqi dīwān translation: 94–97, 98, 110, 140 on Khurasani dīwān translation: 104–05, 110, 113, 140–41 on non-Muslim administrators: 47 n. 14, 133 on papermaking: 441 on Qaḥdham as dīwān administrator: 121 n. 184 on Syrian dīwān translation: 101, 104, 137 al-Jallad, Ahmad: 172–73 n. 36, 206, 226, 366 al-Jarrāḥ b. ʿAbd Allāh al-Ḥakamī: 64, 265–66 Jazīra region administrative apparatus of: 122 n. 187 dīwāns in: 144–45, 145 nn. 325 & 326, 149, 153 Kharijite rebellion in: 144–45 in Marwānid period: 140, 144–46, 150 Jews and Judaism see also Hebrew Bible Babylonian: 427, 429, 431–33, 431 n. 38, 454 in Baghdad: 429–35, 432 n. 43, 446 book culture and: 440, 445–55 conversion to: 226, 413 n. 41 Karaite: 409, 412–13, 434, 436, 445, 448–50, 449 n. 138, 453–55 languages used by: 203, 226–30, 426 in Late Antiquity: 425, 438, 439 literacy among: 435

index

as merchants: 30, 433–34, 433 n. 51 persecution by al-Walīd: 100 politico-religious conflicts with Christians: 241 in pre-Islamic period: 197, 200, 210, 226–30 religious connotations of multilingualism: 26, 29 in Sasanian Empire: 213, 431 sectarianism among: 426, 435, 438, 453, 454 Jinbashian, Manuel: 284 John of Damascus: 48 n. 17, 407 n. 13 John of Ephesus: 216 n. 89, 239 Josephus: 438 Judaeo-Arabic: 29, 228, 375 n. 53, 408, 413 n. 45, 447 Judaeo-Persian: 334, 335 Judaism see Jews and Judaism Julius Africanus: 369 Justin I (Byzantine Empire): 239, 243 Justin II (Byzantine Empire): 217 Justinian (Byzantine Empire): 130 n. 238, 210–11, 216, 218, 221, 221 n. 110, 239, 241–42 Justinian II (Byzantine Empire): 273–74 n. 58 Kaplony, Andreas: 50 n. 29 Karaite Judaism: 409, 412–13, 434, 436, 445, 448–50, 449 n. 138, 453–55 kātibs see scribes Kawād I (Sasanian Empire): 213, 272 n. 50 Kennedy, Hugh: 122, 123 n. 192, 148 n. 341, 430, 440 Keshavarz, Karim: 340 n. 63 Khalek, Nancy: 46 n. 8, 47 n. 14, 95–96 nn. 43 & 44, 96, 133 nn. 251 & 252 Khālid b. ʿAbd Allāh b. al-Qasrī: 52 n. 34, 105, 113, 131, 139 Khālid b. Walīd: 380–81 Khālid b. Yazīd: 18, 45 n. 6 Khalīfa b. Khayyāṭ: 93, 100 n. 60, 101–02, 104, 109–10, 119–20, 122, 124 Khan, Geoffrey: 307, 322 n. 35, 323 Kharijite rebellion: 144–45 Khurasan administrative language in: 64, 342 Classical New Persian in: 352 coins minted in: 262–64, 266, 268, 272, 275 dīwān translation narrative: 49, 104–05, 110, 113, 140–41, 151 Jewish merchants in: 433 multilingualism in: 64–65, 68–69 numerical value expression in: 127–28

483

484

i n dex

Sasanian Empire and: 346 during Second Fitna: 268–70, 275 Khwarazmian language: 329, 334, 350 al-Kindī: 48, 75, 106–07, 110, 112–13, 129, 436 Kitan language: 21 n. 33 Klasová, Pamela: 136, 183 Knauf, Ernst A.: 202, 387 n. 117 Kosrow I Anōširavān (Sasanian Empire): 212, 216–18 Kosrow II Parwez (Sasanian Empire): 213, 239–40, 262, 271, 274 Krachkovskaya, B. A.: 297 Kufic language: 61, 267, 297 Kultursprache: 19, 20 Kurmzād: 262–64, 274–75 Lakhmids: 211–12, 216–17, 231, 239, 242, 245–46 language see also cosmopolitanism; multilingualism; vernacular languages; specific languages of diplomacy: 215–19 empire and: 43, 43 n. 1, 78–79, 184, 275, 282 identity and: 19, 30–31, 78 of imperial armies: 219–20 for legal documents: 221–25 power and: 69, 69–70 n. 98, 168, 184, 261, 281–84, 295, 299–302, 329–31, 334 religious/sacred: 27, 170, 196, 200, 225–37, 403 n. 2 trade and: 29–30, 221, 283–84 language change see also translation movement conversion and: 178 n. 78 elite participation in: 24, 25 linear narratives of: 13, 25 multidirectional process of: 15, 25 personnel change and: 43, 49, 75–77, 90, 104–05, 139, 151 progressive: 74, 76, 78, 91, 181 religious connotations of: 26–27 scribal education and: 23, 135–36 sociodemographics and: 19–20 langue courtoise: 20, 172 n. 36 langue référentielle: 20, 153, 168, 173 nn. 39 & 40 langue véhiculaire: 20, 173, 173 n. 39 La Porta, Sergio: 143 Late Antiquity allography: 196, 205 Aramaic dialects in: 203

index

book culture in: 403 coins during: 55 n. 53 integration of Islam into: 14, 183 Jewish populations in: 425, 438, 439 literacy during: 197, 223, 244–47 multilingualilsm in: 15, 24, 196, 219, 239 pre-Islamic Arabic in: 195, 199–202, 204 Latin language Arabic, interplay with: 20–21 civil/administrative use of: 209, 210 on coins: 17, 17, 17 n. 12, 28, 54, 147, 181 cosmopolitanism and: 172, 182 for imperial diplomacy: 216 for legal documents: 221, 222 as military language: 219–20 Qubbat al-Khazna documents in: 180 religious uses of: 233, 234 Lavan, Myles: 169–70 al-Layth b. Saʿd: 118, 120–21, 124, 153 Lazard, Gilbert: 336–37, 340 n. 63, 343 Legendre, Marie: 23–24, 46, 53, 89, 181 Leo I (Byzantine Empire): 237–38 Lerner, Judith A.: 24–25, 307 Levy-Rubin, Milka: 90, 141 Łewond: 144, 144 n. 317, 146, 282, 287 n. 26, 293 Lieberman, Phillip I.: 432, 432 n. 43 lingua franca: 17, 91 n. 17, 168–69 n. 7, 219, 284, 367 lingua sacra see sacred languages literacy see also book culture in Baghdad: 438, 440, 442 of elites: 152, 197 of Jewish populations: 435 in Late Antiquity: 197, 223, 244–47 literary networks: 175, 178–79 Livshits, Vladimir A.: 348 Løkkegaard, Frede: 127 Longworth, Kyle: 90–91, 108, 108 n. 106, 114 n. 135, 118–19 n. 165 al-Madāʾinī: 94–97, 98, 100, 104, 110, 113–14, 127, 136–37 Malek, Hodge Mehdi: 272 n. 52 Malik al-Ramlī: 450 Mallette, Karla: 20–21

48 5

486

i n dex

Mamluk Sultanate: 21 n. 36, 22 Manichaeans: 204, 211, 330, 349, 352 Mann, Jacob: 432 n. 43 al-Manṣūr (ʿAbbasid caliph): 427, 431, 441 Manṣūr b. Sarjūn: 101, 111, 129 manuscript culture: 180–82 Manzano Moreno, Eduardo: 122 n. 189, 148–49 al-Maqdisī: 343, 346, 433 al-Maqrīzī: 48, 94, 97, 98, 102, 104, 107, 115–16 Mar Behnam monastery (Iraq): 21–22 Mardānshāh b. Zādhān Farrūkh: 127, 129, 137 Marem Bagratuni: 300 Marsham, Andrew: 48 n. 21 Marwān b. al-Ḥakam (Umayyad caliph): 106, 111, 119, 120 Marwān b. Muḥammad (Umayyad caliph): 52 n. 34, 119, 120, 129, 131, 142 n. 304 Marwānid period see also dīwāns; Second Fitna; specific caliphs administrative policies of: 16–17, 49–56, 53 n. 41, 73, 112, 120 Armenia during: 282, 286–94 Jazīra region during: 140, 144–46, 150 linguistic imperialism during: 17, 50, 175 mawlā scribes during: 139 al-Qāḍī on: 149 n. 347 Umayyad North during: 140, 142–44 Maslama b. ʿAbd al-Malik: 145 n. 326 Maslama b. Mukhallad al-Anṣārī: 106 Masoretes: 446–48, 447 n.128, 454 al-Māwardī: 94, 98, 100, 104, 137 Mawiyya (Tanukh queen): 238 mawlās as administrators: 48, 72, 90, 105, 106, 124, 127, 130 as scribes: 46, 48 n. 21, 72, 90–91, 102, 108, 133, 138–39 Melkites: 101, 111, 180 n. 93, 375, 428 Menocal, María Rosa: 182 n. 105 merchants see trade Mesrop Maštocʿ: 201 Miaphysites on Council of Chalcedon: 204 diplomacy by: 240–43 hierarchy of: 216, 242 at Ḥirta/al-Ḥīra: 245, 380 n. 84 Syriac as language of: 210, 232, 242, 248 in Umayyad North: 143–44

index

Michael the Synkellos: 411 n. 30 Michael the Syrian: 48 n. 23, 99, 117 Middle Aramaic dialects: 203–04 Middle Persian (MP) see also Pahlavi Aramaisms in: 203 for church and state diplomacy: 240 civil/administrative use of: 173, 209, 211–14, 341–42, 345 on coins: 17, 28, 147, 287–88, 341–42, 344, 346 dīwāns, use of term in: 116, 116 n. 151 for imperial diplomacy: 215–18 for legal documents: 224 as military language: 220 Muʿāwiya’s use of: 16 New Persian compared to: 333–35 Parthian language and: 333 n. 28, 350–53 religious uses of: 231, 234–35 revival of: 69 n. 98 for trade negotiations: 221 translation into Arabic: 17 Zoroastrians and: 332, 340, 351, 352 Miles, George: 89 Minns, Ellis H.: 313 Mīshawayh al-ʿUkbarī: 450 Miskawayh: 93, 95–97, 133 n. 252 monasteries education at: 61, 70, 71, 230–31, 244–45 Hałbat: 295–96, 296 of Ḥirta/al-Ḥīra: 231, 244–46 manuscript production at: 25, 60, 60 n. 64 Mar Behnam: 21–22 name changes upon entrance to: 411 n. 30 Mongols: 21–22, 21 n. 36, 22, 331 Morelli, Federico: 59, 70 n. 101 Morimoto, Koseï: 91 n. 14, 115, 122–23 Moses ben Jacob: 432 Mourners of Zion movement: 450 MP see Middle Persian Muʿāwiya b. Abī Sufyān (Umayyad caliph) Athanasius and: 106 coin reforms of: 52, 55, 55 n. 53 dam inscriptions of: 174, 174 dīwāns and: 100, 119, 122

48 7

488

i n dex

as governor of Syria: 143, 284–86 Greek inscription at Hammat Gader: 16 multilingualism of: 16 al-Mubarrad: 98 Muḥammad (prophet): 114 n. 135, 195, 224, 452 Muḥammad b. Marwān: 52 n. 34, 143, 145, 286–89, 293 multigraphism biblical codices and: 403 n. 2 Hebrew-Arabic comparison: 408–13, 409 Late Antique book culture and: 403 in legal and administrative documents: 210, 222 in Qubbat al-Khazna fragments: 404–7, 405–12, 409 multilingualism in ʿAbbasid Caliphate: 29, 152, 283 administrative documents and: 56–65, 57–60, 63, 67–70 in Armenia: 282–84, 299 of dīwāns: 66–67, 66 n. 84, 129 in Egypt: 56–62, 67–72, 134–36, 134 n. 258 ideological motivations for: 62, 66–68 of imperial armies: 220 in Iraq: 29 in Khurasan: 64–65, 68–69 in Late Antiquity: 15, 24, 196, 219, 239 phases of: 65–72 practical functions of: 62, 65–69 of Qubbat al-Khazna: 180 religious connotations of: 26–29 role in Islamic studies: 14, 20–21 in Sasanian Empire: 69 trade and: 29–30, 221, 283–84 in Umayyad Caliphate: 16–18, 23, 29, 56–72, 78–79, 152, 261, 281 al-Mundhir III (Lakhmid king): 212, 216, 231, 239–43 al-Muqammiṣ: 436–39, 437 n. 66, 445, 451, 453 Musa b. Sayyar al-Uswari: 339 Mushe bar Kipho: 375 Muslims see Islam and Muslims al-Muʿtaḍid (ʿAbbasid caliph): 444 al-Mutawakkil (ʿAbbasid caliph): 51 n. 30, 296–97, 299, 382, 386 Muʿtazilites: 94 n. 30, 436, 453

index

Nabataean Aramaic language: 172 n. 36, 202–05, 366–68 Nabataeo-Arabic language: 196, 201, 206, 209, 247, 248, 366 Nadhr b. Shamir: 346 Najrān dossier: 214, 228, 243–44 Namāra inscription: 368, 368 Narseh (Sasanian Empire): 211 Narshakhī: 27–28, 347 Nāṣir-i Khusraw: 29, 283 Naṣr b. Sayyār: 49, 104, 105, 113 Natronay bar Hilay Gaon: 438 Naxčawan, fires of: 143, 143 n. 310, 144 Nebrija, Antonio de: 43, 43 n. 1, 78, 184 NENA (North-Eastern Neo-Aramaic): 378–79 Neo-Aramaic: 378–79 Nessana documents: 18, 123 n. 192 New Persian (NP) Bactrian language and: 334, 343, 346, 350 Classical: 330–31, 334, 340, 340 n. 63, 346, 352–53 Early: 329–37, 339–42, 350, 352–54 lack of liturgical connections: 331–32 n. 16 Middle Persian compared to: 333–35 Parthian language and: 352–53 political life of: 340–42 Sogdian language and: 334, 346–49 Turkic language and: 350 Zoroastrians and: 332 Niketas Abalante: 415, 416 Nikitin, Alexander: 274 non-Muslims see also Christians and Christianity; Jews and Judaism Arabicization and: 25, 28, 90 caliphal diversity in treatment of: 74 n. 114 as dhimmīs: 127, 430, 441 as dīwān administrators: 47 n. 14, 76–77, 76 n. 121, 90, 103, 105, 127, 138–39 Qurʾānic verses on coins and: 55 role in Islamic studies: 13–14, 14 n. 2 Nonnus of Nisibis: 299, 300, 302, 436 Norman Johannes of Oppido: 412–13 n. 41 North Africa see also specific locations coins from: 55, 147–48 dīwāns in: 146–48, 153 Umayyad Caliphate and: 140, 146–47, 146 n. 329, 150 North-Eastern Neo-Aramaic (NENA): 378–79

48 9

490

i n dex

Noth, Albrecht: 107, 110 NP see New Persian al-Nuʿmān I (Lakhmid king): 212, 220 al-Nuʿmān II (Lakhmid king): 240 al-Nuʿmān III (Lakhmid king): 213, 239–40 al-Nuwayrī: 94, 98, 100, 104, 129, 137 Old Arabic see pre-Islamic Arabic Old Persian: 330, 353 Ottoman sign systems: 22 Pahlavi administrative use of: 50, 345 on coins: 261–64, 266, 267, 269, 271, 276, 344–46 imperial nature of: 333 numerals in: 73 Parthian language and: 351–52 Qom archive and: 332–33 Tabaristan archive and: 317, 318, 322 n. 35, 332–33, 341, 345, 346 Palaeo-Arabic see pre-Islamic Arabic Palestine Arabicization of Christian communities in: 28 indigenous languages of: 410 n. 27 Masoretic project in: 446–47 Middle Aramaic dialects used in: 204 pre-Islamic inscriptions in: 174 writing systems used in: 413 Papaconstantinou, Arietta: 19, 20, 91 n. 13, 134 n. 258, 365 n. 2, 389 n. 123 papermaking: 426, 440–43 Parthian language: 211, 333, 333 n. 28, 348, 350–53 Payne, Richard: 169–70 Perry, John R.: 331–32 n. 16 Persian language see also Middle Persian; New Persian administrative use of: 141 n. 295, 283, 331, 335 in Armenian inscriptions: 295 cosmopolitanism and: 168, 168–69 n. 7, 330, 330 n. 7 dīwāns in: 46, 49, 66, 66 n. 84, 104, 115 Islamic viewpoint on: 335–40 Judaeo-Persian: 334, 335 as lingua franca: 168–69 n. 7 numerals in: 127 Old Persian: 330, 353 Qurʾānic verses in: 27

index

in Rasulid Hexaglot: 21 role in Islamic studies: 13 Peter of Callinicum: 243 Pfeifer, Helen: 22 Philoxenus of Mabbug: 240, 241 Pīrūz (Sasanian Empire): 271, 271 n. 46, 272, 273 n. 55 Podosaces: 211, 218 political power and authority: 24, 260–61, 275–76, 281, 296, 300, 301 Polliack, Meira: 450–51 Pollock, Sheldon: 23, 23 n. 43, 134 n. 258, 167–75, 171 n. 26, 173 n. 40, 178, 181–83 polytheism: 77, 104, 204, 225–27, 232, 236, 440 power of administrative elites: 114 articulation of: 169, 180 balance of: 142, 150 of culture: 184 economic: 276, 428 hierarchies of: 64 imperial repertoires of: 180 n. 88 language and: 69, 69–70 n. 98, 168, 184, 261, 281–84, 295, 299–302, 329–31, 334 political: 24, 260–61, 275–76, 281, 296, 300, 301 public performance of: 282–84, 286–302 of Sanskrit cosmopolis: 171 scribal dynamics of: 102 translation movement and: 133, 177, 281, 300 pre-Islamic Arabic diversity of: 15, 202, 204–06, 209, 366 emergence of: 199–202, 247–48 for imperial diplomacy: 218 inscriptions: 173–74, 174, 195, 198, 198–99, 367–68, 368 for legal documents: 223–25 Najrān dossier and: 243 poetry in: 176, 196, 222, 223, 229–30, 238 as religious language: 225–26 in Syriac Christianity: 368–74, 371–72 terminology considerations: 198 n. 14 pre-Islamic period see also pre-Islamic Arabic Christians in: 197, 200, 210, 230–37 church and state diplomacy in: 237–47 civil languages in: 152, 209–15 dīwāns in: 44, 66, 77 elites in: 197, 214–15, 219, 230, 231

49 1

492

i n dex

imperial diplomacy in: 215–19 Jāhiliyya and: 176, 200 Jews in: 197, 200, 210, 226–30 legal document in: 221–25 military languages in: 219–20 religious languages in: 225–37 scientific traditions of: 18 trade during: 221 prior texts: 175–78 Procopius (historian): 215 Procopius (martyr): 235 Puin, Gerd-Rüdiger: 117–18 al-Qāḍī, Wadād dīwāns, use of term by: 116 on Iraqi dīwān translation: 90, 96–97, 125, 140 on Marwānid period: 149 n. 347 on research gaps in Islamic studies: 89 on secretarial agency: 152 on working in dīwāns: 129 n. 231, 130 Qaḥdham: 96–97, 121, 121 n. 184, 130, 132, 137, 151 al-Qaḥdhamī: 94–97, 98, 101, 104, 110, 121 al-Qalqashandī: 94, 96–97, 96 n. 46, 98, 101–02, 104, 107 al-Qifṭī: 443 al-Qirqisānī: 436, 437 n. 66, 445–47, 449–52, 454 Qom archive: 332–33 Qubbat al-Khazna documents: 17, 25, 180–81, 404–7, 405–12, 409 al-Qūmisī, Daniel: 450–55 Qurʾān Christian theology and: 236–37 codices of: 28, 448 coin reforms and: 54, 55, 288 consonantal text of: 366 n. 5 correct readings of: 448 on debt recording: 224 languaging process and: 177 in nonliterate context: 224 n. 127 oral preservation of: 195 scribal education on: 132 structure of text: 452 translations of: 27–28 ʿUthmānic version of: 228

index

Qurayshites: 215, 223 Qurra b. Sharīk: 62, 121 Quṣayr ʿAmra (desert castle): 18, 18 n. 20, 181 Qutayba b. Muslim: 265, 265 n. 23, 267, 270, 347 Rabbula of Edessa: 369–70 al-Rāghib al-Iṣfahānī: 98, 127 Rashīd al-Dīn: 21, 21 n. 35 Rashidun Caliphate see specific caliphs Rasulid Hexaglot (glossary): 21 al-Rāzī: 101–02, 104, 148 reading see book culture Rebstock, Ulrich: 90 n. 5, 116 Reichssprache: 19, 20, 286–94, 301 Reinfandt, Lucian: 50 n. 28 religion see also conversion; specific religions church and state diplomacy: 237–47 coin reforms and: 55 multilingualism and: 26–29 polytheism: 77, 104, 204, 225–27, 232, 236, 440 sacred languages: 27, 170, 196, 200, 226, 235, 403 n. 2 Religionssprache: 19, 20 Rezakhani, Khodadad: 25, 329 Ricci, Ronit: 168, 171 n. 32, 175–76, 178, 183 n. 106 Richardson, Kristina: 22 Robin, Christian Julien: 226, 228 Robinson, Chase: 17, 49–50, 50 n. 29, 122 n. 187, 144, 175 Romance vernaculars: 182 n. 105 Roman Empire assimilation strategy in: 180 church and state diplomacy in: 237–39 coins from: 54, 55 n. 53 diplomacy and: 215–19 legal documents in: 221–22 military language in: 219–20 official languages of: 203, 210 religious language in: 210, 230–32 trade within: 221 Rosenthal, Judah: 449 Rustow, Marina: 440, 441 Rutbils: 341, 341 n. 66

493

494

i n dex

Saʿadya Gaon: 375 n. 53, 413, 449, 450 Sabaic language: 196, 209–10, 214–15, 220, 226–27, 229, 244 sacred languages: 27, 170, 196, 200, 226, 235, 403 n. 2 Sadeghi, Ali Ashraf: 338 al-Ṣafadī: 101, 102, 104 Safaitic language: 172, 195, 196, 202, 206, 211, 220, 366–67 Sahak: 293 Sahl b. Abī al-Ṣalt: 94, 94 n. 30, 129 Saʿīd b. ʿAbd al-ʿAzīz: 64, 65, 68–69, 72 Saliba, George on arithmetical calculations: 70–71 n. 101 on dīwāns: 90, 116, 125 on scribal education: 71–72 on translation movement: 17–18, 90, 90 n. 8, 131 n. 243, 167, 176 Ṣāliḥ b. ʿAbd al-Raḥmān descendants of: 129–30 n. 233 dīwān translations by: 45 n. 6, 46, 95, 97, 108, 116, 129 family background: 72 scribal students of: 46, 94, 97, 108 n. 106, 131, 151 Zādhān Farrūkh and: 46, 46 nn. 8 & 9, 72 n. 109, 95, 95 n. 43, 127, 136 Ṣāliḥ b. Musarriḥ: 144 Sālim Abū al-ʿAlāʾ: 132, 132 n. 246, 151 Salm b. Ziyād: 263–64, 264 n. 19, 268–69, 269 n. 39, 272, 274, 275 Samanids: 25, 330–32, 335, 339–40, 342, 347 al-Samḥ b. Mālik al-Khawlānī: 148, 149 n. 344 Samuel b. ʿĀdiyā: 230 Sanskrit on coins: 28, 28 cosmopolitanism and: 23, 167–72, 175, 178, 182–83 as sacred language: 170 Šapuh Bagratuni: 297–98 Sarjūn b. Manṣūr al-Rūmī ʿAbd al-Malik and: 47, 47 n. 14, 75, 100 Athanasius and: 111, 112 death of: 76 n. 121, 102, 103, 110, 113, 124 descendants of: 72, 101, 111, 129 as dīwān administrator: 119–22, 124, 137, 151 language competency of: 47 n. 15, 128 Syriac representations of: 101, 101 n. 69 Sasanian Empire see also specific rulers administrative structures of: 345–46 church and state diplomacy in: 237, 239–40

index

civil language in: 209, 211–14 coins from: 54, 55 n. 53, 260–62, 264–68, 271–76, 287–88 cosmopolitanism of: 353 diplomacy and: 215–19 dīwāns in: 94, 116 elites in: 339 Hephthalite conquest by: 261 n. 7, 334 Jewish populations in: 213, 431 legal documents in: 224 military language in: 220 Quṣayr ʿAmra and: 18 n. 20 religious language in: 231, 232 seals from: 317, 322, 322 n. 5, 345 urbanization policies: 428 Savant, Sarah: 30 n. 73, 178 Scheerlinck, Eline: 62 scribes agency of: 102, 108, 133, 152 Arabic: 46, 53, 67, 70, 72, 407 n. 11 Christian: 47–48, 48 n. 23, 73 n. 112, 75, 101–02, 127 Coptic: 61, 70–71, 76 copyists vs.: 408 n. 15 education of: 23, 25, 46, 61, 70–72, 94, 97, 131–32, 135–36, 136 n. 266, 152, 247, 348–49 Greek: 47, 48, 59–62, 67, 70–72, 76, 100, 108, 407 n. 11 mawlās as: 46, 48 n. 21, 72, 90–91, 102, 108, 133, 138–39 milieu of: 131, 134–36, 140, 151 Persian: 46, 46 n. 8, 49, 75, 96, 108, 224 in pre-Islamic period: 197, 202, 209, 214, 223–25 Sogdian: 64, 71, 72 Syrian: 47 seals for administrative documents: 61, 62 animal depictions on: 319, 319–20 n. 33, 322, 322–23 n. 38 of Armenian king Ašot Bagratuni: 297 from Bactrian documents: 308–23, 310, 315–16, 319–20 in double document method: 311, 313 preservation and storage of: 309 nn. 7 & 8 for religious documents: 228, 239 from Sasanian Empire: 317, 322, 322 n. 5, 345 for six-witness contracts: 312, 312 Sears, Stuart D.: 287, 287 n. 23, 287 n. 26

49 5

496

i n dex

Sebēos: 284–85, 287 Second Fitna Arab-Hephthalite coins and: 261, 264, 264 n. 19, 268–70, 272–73, 275 Armenian rebellion during: 286 dīwān administration during: 120 Marwānid restoration following: 16 secretaries see scribes sectarianism: 426, 435, 438, 443, 450, 453, 454 Sergius of Resafa: 241, 244 Severus b. al-Muqaffaʿ: 375 n. 53 Shabīb b. Yazīd al-Shaybānī: 144 Shapur I (Sasanian Empire): 351, 352 Shapur II (Sasanian Empire): 345 Sharon, Moshe: 118, 121 Sherira Gaon: 425–30, 427–28 n. 13, 454 Shīʿite Islam: 426 n. 5, 431, 444 Shoemaker, Stephen: 223, 224–25 n. 130, 224 n. 127 shuʿūbiyya movement: 30, 30 n. 73 Sībawayhi: 179, 448 Sibṭ b. al-Jawzī: 101–02, 104 Sijpesteijn, Petra M.: 23, 43, 91, 91 n. 15, 123, 134 n. 259, 153, 181 Simeon of Bet Arsham: 239–42, 373 Sims-Williams, Nicholas: 259, 307, 309 six-witness contracts: 312, 312 Sklare, David: 447 Smbat Bagratuni: 295–96, 296, 299, 300 Sogdian language: 50, 64–65, 68–69, 334, 346–50 Sprengling, Martin: 89, 89 n. 3, 100 n. 64, 110, 116 Spuler, Bertold: 78 n. 130, 141 n. 295 state registers see dīwāns Stephen of Ramlah: 410, 410 n. 29 Stern, David: 446–47 Stroumsa, Sarah: 436–37 subordination: 69, 179–81, 183–84, 275 Sufyanids: 106, 117, 130, 287 Sulaymān (Umayyad caliph): 56 Sulaymān b. Marthad: 268 Sulaymān b. Saʿd al-Khushanī dīwān translations by: 47, 47 n. 14, 53 n. 41, 76 n. 121, 100–03, 104, 124, 129 family background: 72, 100 n. 60 as governor of al-Urdunn: 53 n. 41, 137 al-Ṣūlī: 47 n. 14, 94–97, 98, 101, 104, 115–16, 140

index

al-Suyūṭī: 93, 97, 97 n. 52, 98 Syria administrative policy implementation in: 50–52, 50 n. 29 Arabicization of Christian communities in: 28 dīwān administrators in: 101, 119–24, 137 dīwān translation narrative: 45 n. 6, 47–48, 53 n. 41, 73, 75, 76 n. 121, 90, 99–103, 104, 109–10, 113, 114, 140, 150–53 pre-Islamic inscriptions in: 173, 198, 368, 368 writing systems used in: 413 Syriac Christianity Arabic used in: 368–74, 371–72, 376–77 on Council of Chalcedon: 204 East Syriac: 243, 246, 380 Masoretic project in: 447 n.128 missionary efforts of: 173 Qurʾānic passages and: 236–37 Syriac language used in: 374–79 Syriac language see also Aramaic language Arabic in relation to: 24, 206–09, 207–08 for church and state diplomacy: 242, 243 emergence of: 201, 202 for legal documents: 222, 224 Mar Behnam inscriptions in: 21 Miaphysites and: 210, 232, 242, 248 Najrān dossier and: 243, 244 Qubbat al-Khazna documents in: 180 religious uses of: 25, 29, 203–04, 210, 230–37 in Syriac Christianity: 374–79 translation into Arabic: 377 Syriac Orthodox Church: 101, 111, 240, 242, 243, 375, 377 al-Ṭabarī: 229–30, 267–69, 269 n. 40, 380–81, 431, 444, 452 Tabaristan archive: 317, 318, 322 n. 35, 332–33, 341, 345, 346 Taymanitic language: 202, 204–05 Ter-Łevondyan, Aram: 285 n. 19, 295 Thamudic language: 202, 204, 226 Theodoret of Cyrrhus: 203, 232 Theodosius (pope): 216, 242 Theophanes the Confessor on Christian scribes: 48 n. 23, 73 n. 112, 75 on coin reforms: 55 n. 53 on dīwāns: 93, 117, 127

49 7

498

i n dex

on Sarjūn: 47 n. 14 on al-Walīd: 99 Theophilus of Edessa: 48 n. 23, 99 Theophilus the Indian: 230 Theophylactes Simocatta: 434 Third Fitna: 105, 142 Thomson, Robert: 301–02 Tillier, Mathieu: 123, 130 Timothy I (Church of the East): 376 Toorawa, Shawkat: 440 Toral-Niehoff, I.: 380 n. 84, 383 n. 100 Tʿovma Arcruni: 299 trade Arabicization and: 30 coins used in: 273, 273 n. 56 Jewish involvement in: 30, 433–34, 433 n. 51 multilingualism and: 29–30, 221, 283–84 translation movement see also dīwān translations as act of conquest: 177 n. 68, 178 n. 78 book culture and: 439 Christian communities in: 19 conversion and: 175, 178 Graeco-Arabic: 17, 382 languaging process in: 177 power and: 133, 177, 281, 300 Saliba on: 17–18, 90, 90 n. 8, 131 n. 243, 167, 176 Treadwell, Luke: 55, 55 n. 47, 141 Turkic language: 21, 334, 343, 348–50, 354 Turkish language: 13, 171 ʿUbayd Allāh b. Ḥabḥāb: 74–75 ʿUbayd Allāh b. Ziyād: 342 ʿUmar b. ʿAbd al-ʿAzīz (Umayyad caliph): 27, 101, 112 n. 129, 265, 266, 292 ʿUmar b. al-Khaṭṭāb (Rashidun caliph): 66, 66 n. 84, 116–18, 141, 147 ʿUmar b. Marwān b. al-Ḥakam: 107 n. 98 Umayya b. Abī al-Ṣalt: 223 Umayyad Caliphate see also dīwāns; Marwānid period; specific caliphs administrative policies: 16–17, 49–56 Arabic literature during: 176 coin reforms: 17, 17, 52–55, 53 n. 40, 55–56 nn. 52 & 53, 68 n. 95, 113, 175, 264–66, 273–74 n. 58, 288–89 dynamics of change within: 113–14

index

multilingualism in: 16–18, 23, 29, 56–72, 78–79, 152, 261, 281 personnel changes: 43, 49, 75–77, 90, 104–05, 139, 151 state-building efforts of: 49, 114 n. 135 symbolic policies: 50–51 n. 30 translation movement and: 17–18, 167, 176–77, 281 urbanization policies: 428 Umayyad North see also Armenia; Caucasian Albania dīwāns in: 143–44, 144 n. 317, 149, 153 in Marwānid period: 140, 142–44 Urban, Elizabeth: 139 urbanization: 428, 430–32, 434, 438, 439 Usāma b. Zayd al-Tanūkhī: 112 n. 128 ʿUthmān (Rashidun caliph): 228, 275 Vacca, Alison M.: 13, 24, 142–44, 281 Vahan Gołtʿnacʿi: 144, 292–93 Vaissière, Étienne de la: 267 n. 31 Valens (Byzantine Empire): 238 van Berkel, Maaike: 132 van Bladel, Kevin: 19–20, 168 n. 4, 182 Vanthieghem, Naïm: 117, 123, 130, 136 Verkinderen, Peter: 141 vernacular languages Arabic: 172–75 cosmopolitanism and: 134 n. 258, 171–72, 171 n. 26, 183, 353 religious languages, interaction with: 235 Romance: 182 n. 105 Vollandt, Ronny: 403 n. 2 Vondrovec, Klaus: 262, 264, 265, 270 Wahb b. al-Munabbih: 177 al-Walīd b. ʿAbd al-Malik (Umayyad caliph): 45 n. 7, 48, 48 n. 22, 52, 97, 99–100, 117 Walker, John: 262–64, 264 nn. 17 & 18, 267, 269, 274 Wāqidī, Abū ʿAbd Allāh Muḥammad b. ʿUmar: 285–86 Wardān al-Rūmī, ʿUbayd Allāh: 106–07, 151 Wasserstein, David: 18–20, 365, 366 Weber, Max: 184 Weinreich, U.: 385–86 Weisweiler, John: 169–70 Weitz, Lev: 117 Wolper, Ethel Sara: 21–22

49 9

500

i n dex

Yaʿqūb b. al-Sikkīt: 448 al-Yaʿqūbī: 442 Yarbrough, Luke: 76 n. 121, 90, 103, 105, 108 n. 107 Yarbūʿ al-Fazārī: 105 Yazīd b. ʿAbd al-Malik (Umayyad caliph): 51 n. 30, 119 Yazīd b. Jarad: 291, 292 Yazīd b. Muʿāwiya (Umayyad caliph): 120 Yazīd b. al-Muhallab: 265, 267 Yazīd b. al-Walīd (Umayyad caliph): 119, 120, 130 n. 233 Yefet ben Eli: 449 Yehuday Gaon: 437 Yücesoy, Hayrettin: 133, 177 n. 68, 178 n. 78, 184 n. 111 Yūsuf b. ʿUmar al-Thaqafī: 49, 104, 105, 113, 121, 121 n. 184 Zadeh, Travis: 28 Zādhān Farrūkh death of: 46 n. 11, 95, 96, 109, 110, 113, 151 descendants of: 47 n. 12, 72, 73 n. 112, 127, 129, 137 Ibn al-Nadīm on: 46 n. 9, 73 n. 112 language competency of: 128, 128 n. 224 reproduction of dīwān from memory: 109–10 n. 114 Ṣāliḥ b. ʿAbd al-Raḥmān and: 46, 46 nn. 8 & 9, 72 n. 109, 95, 95 n. 43, 127, 136 Zayd b. Ḥammād: 217 Zayd b. Thābit: 228, 230 Zhulād of Gōzgān: 263–66, 263 n. 14, 264 n. 21, 268–70, 272 n. 48, 275–76 Ziyād b. Abīhi: 74 n. 114, 95, 342 Zoroastrians and Zoroastrianism in administrative positions: 139 on apostates: 437 book culture and: 440 conversion to: 214 Middle Persian and: 332, 340, 351, 352 oral traditions in: 239 n. 198 as scribes: 49, 77 Zuartʿnocʿ inscriptions: 289–90, 290, 292 Zubayrids: 28, 106, 174 Zychowicz-Coghill, Edward: 120, 121