Late Antique Responses to the Arab Conquests 2021046471, 2021046472, 9789004500617, 9789004500648, 9004500618

Late Antique Responses to the Arab Conquests is a showcase of new discoveries in an exciting and rapidly developing fiel

119 35 28MB

English Pages 288 [289] Year 2021

Report DMCA / Copyright


Polecaj historie

Late Antique Responses to the Arab Conquests
 2021046471, 2021046472, 9789004500617, 9789004500648, 9004500618

Table of contents :
Half Title
Series Information
Title Page
Copyright Page
Figures and Tables
Notes on Contributors
Chapter 1 Late Antique Responses to the Arab Conquests: An Introduction
1 Islam and the End of Antiquity: Revisiting the Pirenne Thesis in the Age of Connectivity
2 The Impact of Empire: Tradition and Change
3 A “Late Antique Turn”
4 A Survey of the Contributions to This Volume
4.1 The Qurʾān as a Late Antique Text
4.2 Late Antiquity to Early Islam: A Gradual Transformation
4.3 Adapting to the Arab Conquests
5 A Small Arab World: The Late Antique Arab Conquests from a Network Perspective
Chapter 2 The Qurʾānic Rūm: A Late Antique Perspective
1 Reading the Qurʾān: Some Introductory Remarks
2 Traditional Tafsīr of Q 30:2–5
3 The Qurʾānic al-Rūm: Some Late Antique Perspectives
4 Jerusalem—Near and Far?
5 Conclusions
Chapter 3 Wine and Impurity in the Sura of the Bees: A Structuralist Interpretation of Qurʾān 16:67
1 The Debilitating and Polluting Aspects of Sakar
2 The Drinks of Life
3 The Opposition between Pure and Impure in Wine Verse 16:67
4 Conclusion
Chapter 4 Historical-Critical Research of the Sīra of the Prophet Muhammad: What Do We Stand to Gain?
1 The State of Affairs
2 A Story of the Sīra
3 Conclusion
Chapter 5 Arabicization, Islamization, and the Colonies of the Conquerors
1 Arabicization
2 Islamization
3 Patterns of Arabicization and Islamization
4 Conclusion
Chapter 6 Continuity and Change: Elite Responses to the Founding of the Caliphate
1 Conquests and Conquerors?
2 Continuity of Emigrant Communities?
3 Sources
4 Changes to Elite Communities: Poetic Indications
4.1 Terms of Communal Identity
4.2 Spatial Considerations
4.3 The Hajj
4.4 Islam and Poetry
5 An Arabness Case Study: al-Azd in Umayyad Politics
6 Conclusions
Chapter 7 Muhammad’s World in Egypt
1 Sources and Methods in the Study of Early Islam
2 A New Regime
3 Consumption Patterns and Material Culture
4 Muslim Rule?
5 Patterns of Innovation and Influence
Chapter 8 “May God be Mindful of Yazīd the King”: Further Reflections on the Yazīd Inscription and the Development of Arabic Scripts
1 Two Other Early Christian Inscriptions and the Shape of the r
2 Is Yazīd the King Really Yazīd i?
3 The Writing School of Medina
4 Conclusion
Chapter 9 Of Siblings, Kingdoms, and the Days of the Messiah: Jewish Literary Responses to the New Order in the Land of Israel in the First Muslim Period
1 The Dissimilar Siblings and Their Dissimilar Offspring
2 Reordering the Kingdoms
3 The Kingdom of Ishmael and Its Role in the Jewish Eschaton
4 Conclusion
Chapter 10 New Light on the Dark Ages: A Byzantine Perspective on the Arab Expansion
1 Introduction
2 Empires in Flux
3 First Case Study: Athens
4 Second Case Study: Butrint
5 Conclusion
Index of Names and Subjects
Index of Biblical and Qurʾānic References

Citation preview

Late Antique Responses to the Arab Conquests

Cultural Interactions in the Mediterranean Editor in Chief Floris van den Eijnde, Utrecht University Editorial Board David Abulafia, Cambridge University Diederik Burgersdijk, Radboud University

volume 5

The titles published in this series are listed at​cim

Late Antique Responses to the Arab Conquests Edited by

Josephine van den Bent, Floris van den Eijnde and Johan Weststeijn


Cover illustration: Gold “Standing Caliph” dinar. Courtesy American Numismatic Society. Public domain: http://​​collection/​1970.63.1. Library of Congress Cataloging-in-Publication Data Names: Bent, Josephine van den, 1988- editor, author. | Eijnde, Floris van den, 1975- editor, author. | Weststeijn, Johan (Researcher), editor, author. | Mohammed en het Einde van de Oudheid (Conference) (2015 : Universiteit van Amsterdam) Title: Late antique responses to the Arab conquests / edited by Josephine van den Bent, Floris van den Eijnde, and Johan Weststeijn. Description: Leiden ; Boston : Brill, [2022] | Series: Cultural interactions in the Mediterranean, 2405-4771 ; volume 5 | This book originated in a Dutch symposium on Late Antiquity and early Islam organized by the Zenobia Foundation in 2015: “Mohammed en het einde van de Oudheid”–Preface. | Includes bibliographical references and index. Identifiers: lccn 2021046471 (print) | lccn 2021046472 (ebook) | isbn 9789004500617 (hardback) | isbn 9789004500648 (ebook) Subjects: lcsh: Islamic civilization. | Mediterranean Region–Civilization– Islamic influences. | Mediterranean Region–Colonization–History–To 1500. | National characteristics, Arab. | Qurʼan–Criticism, interpretation, etc. | Islam–History. | Muḥammad, Prophet, -632. Classification: lcc DS36.85.L38 2022 (print) | lcc DS36.85 (ebook) | ddc 909/.09767–dc23/eng/20211012 lc record available at lc ebook record available at

Typeface for the Latin, Greek, and Cyrillic scripts: “Brill.” See and download:​brill-​typeface. issn 2405-​4 771 isbn 978-​9 0-​0 4-​5 0061-​7 (hardback) isbn 978-​9 0-​0 4-​5 0064-​8 (e-​book) Copyright 2022 by Josephine van den Bent, Floris van den Eijnde and Johan Weststeijn. Published by Koninklijke Brill nv, Leiden, The Netherlands. Koninklijke Brill nv incorporates the imprints Brill, Brill Nijhoff, Brill Hotei, Brill Schöningh, Brill Fink, Brill mentis, Vandenhoeck & Ruprecht, Böhlau Verlag and V&R Unipress. Koninklijke Brill nv reserves the right to protect this publication against unauthorized use. Requests for re-​use and/​or translations must be addressed to Koninklijke Brill nv via or This book is printed on acid-​free paper and produced in a sustainable manner.

Contents  Preface vii  List of Figures and Tables ix  Notes on Contributors xi 1  Late Antique Responses to the Arab Conquests: An Introduction 1 Josephine van den Bent, Floris van den Eijnde and Johan Weststeijn 2  The Qurʾānic Rūm: A Late Antique Perspective 32 Clare Wilde 3  Wine and Impurity in the Sura of the Bees: A Structuralist Interpretation of Qurʾān 16:67 56 Johan Weststeijn 4  Historical-​Critical Research of the Sīra of the Prophet Muhammad: What Do We Stand to Gain? 74 Harald Motzki † 5  Arabicization, Islamization, and the Colonies of the Conquerors 89 Kevin van Bladel 6  Continuity and Change: Elite Responses to the Founding of the Caliphate 120 Peter Webb 7  Muhammad’s World in Egypt 171 Petra M. Sijpesteijn 8  “May God be Mindful of Yazīd the King”: Further Reflections on the Yazīd Inscription and the Development of Arabic Scripts 195 Ahmad Al-​Jallad 9  Of Siblings, Kingdoms, and the Days of the Messiah: Jewish Literary Responses to the New Order in the Land of Israel in the First Muslim Period 212 Constanza Cordoni

vi Contents 10  New Light on the Dark Ages: A Byzantine Perspective on the Arab Expansion 245 Joanita Vroom  Index of Names and Subjects 267  Index of Biblical and Qurʾānic References 273

Preface This book originated in a Dutch symposium on Late Antiquity and early Islam organized by the Zenobia Foundation in 2015: “Mohammed en het einde van de Oudheid.” The Zenobia Foundation presents scholarship on East and West in Antiquity to a wider Dutch audience; the editors of the book you are holding have all been connected at some point to this foundation. Several scholars from Dutch universities presented at the 2015 symposium, among them Ahmad Al-​ Jallad, Petra Sijpesteijn, and Joanita Vroom, all at the time working at Leiden University. Kevin van Bladel from Yale University presented his contribution to the current volume as the English-​language keynote lecture. A Dutch-​ language publication resulted from this symposium entitled Mohammed en de Late Oudheid (Hilversum: Verloren, 2018). Harald Motzki (Radboud University Nijmegen) and Clare Wilde (University of Groningen) contributed additional chapters for the publication. Mohammed en de Late Oudheid and the present volume were funded by Stichting Oosters Instituut, Stichting Boekenfonds Elisabeth Grent, Stichting Sormani Fonds, the Zenobia Foundation, and various individual donors. The Dutch book laid the foundation for an academic version in English. Some contributions to the Dutch book were intended as an introduction to various aspects of the field of study to a wider audience, while others presented cutting-​edge scholarship. It is the latter that we have selected for this volume. Together, these contributions cover a broad range of subfields with regards to research matter, source material, and geography. We have asked two more scholars, focusing on yet other aspects, to contribute as well: Peter Webb (Leiden University) and Constanza Cordoni (Utrecht University). During production of this book the sad news reached us that Harald Motzki passed away on 8 February 2019. Trained in Germany, Motzki had worked at Radboud University Nijmegen since 1991, where he held the chair Methodology of Research in Islamic Studies from 2001. Motzki acquired fame for his innovative isnād-​cum-​matn analysis. Commenting on his legacy, his Nijmegen colleague Kees Versteegh wrote that perhaps the most important lesson Motkzi taught us was that “rejecting the [Arabo-​Islamic] evidence about the early Islamic period out of hand is mistaken, unwarranted, and counterproductive.”1

1 Kees Versteegh, “Harald Motzki (1948–​2019).” Hadis ve Siyer Araştırmaları: Hadīth and Sīra Studies 5 (2019): 397–​99.

viii Preface We are grateful to Hisham Hamad for his translation of Harald Motzki’s contribution, and to Anna Ruitenbeek and Merlijn Veldman for additional translation assistance. Paul Kloeg made beautiful tracings of Arabic inscriptions for the contribution by Ahmad Al-​Jallad, and Anouk de Bruin was of great help in compiling the index. Jennifer Palinkas was essential as our proofreader, keeping us from many a mistake and greatly improving the language of this book. Josephine van den Bent, Floris van den Eijnde and Johan Weststeijn

Figures and Tables Figures 1.1  Mediterranean shipwrecks (By RaphaelG, data source: Parker 1992) 4 7.1  Papyrus perf 558 recto, receipt for the delivery of 65 sheep from 643 ce (Österreichische Nationalbibliothek) 173 7.2  Papyrus perf 558 verso 174 7.3  Seal on a letter from governor Qurra b. Sharīk to the pagarch of Ishqūh/​ Aphrodito, Basileios dated Ṣafar 91 ah/​December 709–​January 710 ce (D. 13296: Papyrus Qurrah iii, detail of lead seal (oim E13756) Courtesy of the Oriental Institute of the University of Chicago) 180 7.4  Coins of Hishām ʿAbd al-​Malik with Qurʾānic phrases from al-​Andalus, Marw, and al-​Jazīra (OMJ 307–​F 03, OMJ–​307–​H 03 and OMJ–​3 08–​c 10. Copyright Universität Jena, Orientalisches Münzkabinett) 181 8.1  The Yazīd inscription (Photo: Z. al-​Salameen) 195 8.2  Superscript dot on UJadh 178 (Nehmé 2010, 78) 198 8.3  dāl/​dhāl in the Yazīd inscription (Al-​Jallad et al. 2017) 198 8.4  Dotted dāl in the Codex Parisino-​Petropolitanus 199 8.5  Saʿd bin Dhakwān inscription (Al-​Saʿīd and al-​Biṭār 2018, 127) 199 8.6  Detail of section with the dotted dāls 200 8.7  Syriac d 201 8.8  Syriac r 201 8.9  Detail of ʾbn ḥdydw in perf 558 (Larcher 2010) 201 8.10  The r of the Zuhayr inscription (643/​644 ce) (Ghabban and Hoyland 2018) 202 8.11  The r of the Yazīd inscription (Al-​Jallad et al. 2017) 202 8.12  The ʿAbd al-​Masīḥ amulet (al-​Jumaili 2016) 203 8.13  Kilwa inscription (Hoyland 2018) 205 8.14  Inscription #163 (al-​Theeb 2002) 208 10.1  Map of Byzantine Empire, with Athens and Butrint. Red line: size of the Empire during the reign of Justinian i (527–​65); region in orange: size ca. 1020; region in red: size ca. 1360 (J. Porck) 246 10.2  Miniature with the use of Greek fire, invented during the Byzantine-​Arab wars. Codex Skylitzes Matritensis, 12th century, Biblioteca Nacional de España, Madrid, Vitr. 26-​2, Bild-​Nr. 77, f 24v.b (from: Pászthory 1986, 31) 248 10.3  Photograph of the antique Agora in Athens (ascsa) 249 10.4  Map of Athens, including the diminution of the area of the antique city in Late Antiquity until 1204 ce (after: Bouras 2010, ­figure 1) 250


Figures and Tables

10.5  Map of the Agora with wells and building structures from the “Dark Ages,” including the late antique city wall (red line) (J. Vroom, E. Tzavella, ascsa) 251 10.6  Graph with wells with finds from the “Dark Ages” in the Athenian Agora (J. Vroom) 252 10.7  Cross-section of well H11:1.2 from section Zeta in the Athenian Agora with finds, varying from the late Roman until the mid-​Byzantine periods (J. Vroom, E. Tzavella, P. Doeve) 253 10.8  Graphs and drawings of finds from well H11:1.2 in the Athenian Agora (J. Vroom, F. Kondyli) 254 10.9  Drawing of the western defence wall in Butrint, Albania, with tower 1 (wd 1) and tower 2 (wd 2) depicted (Butrint Foundation) 256 10.10 Photograph of tower 1 (wd 1) in Butrint, Albania (J. Vroom) 257 10.11  Reconstruction drawing of the interior of tower 1 (wd 1) in Butrint, Albania, including photographs of finds (J. Vroom, W. Euvermans, Butrint Foundation) 258

Tables 3.1  Oppositions between impure and pure in Qurʾānic lā taqrabū … ḥatta … proscriptions 61 3.2  Oppositions between pure and impure in Q 16:65–​70 69 3.3  Structure of the passage on the drinks of life in the Sura of the Bees 70 4.1  Transmission chains (asānīd) from the Sīra by Ibn Hishām and al-​ʿUṭāridī 84 8.1  Inscriptions with the formula dkr ʾl-​ʾlh (may God be mindful of …) 196 8.2  The letter r as a slightly curved line in sixth-​and seventh-​century inscriptions 204 8.3  The Jebel Usays inscription, reading and interpretation following Macdonald 207 8.4  The letter alif in sixth-​and seventh-​century inscriptions 208

Notes on Contributors Ahmad Al-​Jallad is a philologist, epigraphist, and historian of language. His work focuses on the languages and writing systems of pre-​Islamic Arabia and the ancient Near East. Josephine van den Bent is lecturer and postdoctoral researcher in the history department of Radboud University Nijmegen, investigating water management in premodern Iraqi cities (c. 700–​1500) as part of the project “Source of Life,” funded by the Dutch Research Council. Her PhD thesis (University of Amsterdam, 2020) analyzed the representation of the Mongols in the Mamluk Sultanate of Egypt and Syria, c. 1250–​1350. Her research interests include ethnic stereotyping, urban organization, and in general the social and cultural history of the premodern Middle East. Kevin van Bladel is a Professor of Near Eastern Languages & Civilizations at Yale University. He is the author of The Arabic Hermes (2009) and From Sasanian Mandaeans to Sabians of the Marshes (2017), as well as many articles on the classical traditions of western Asia and the language communities of the Near East in the first millennium ce. Constanza Cordoni is the Jewish studies research fellow at the Institute for Medieval Research of the Austrian Academy of the Sciences in Vienna and lecturer at the Department of Jewish Studies of the University of Vienna. She was formerly a postdoctoral researcher at Utrecht University. Her research focuses on rabbinic literature of late antiquity and the early Middle Ages, the literature of formative Judaism. She is especially interested in questions of historical hermeneutics and narratology. She is co-​editor of the Vandenhoeck & Ruprecht series Poetics, Exegesis, and Narratology. Floris van den Eijnde is director of Ancient, Medieval, and Renaissance Studies and lecturer and researcher of Ancient History at the department of History and Art History of Utrecht University. He leads the uu research group Sacrality and the Greek Polis, is founder and co-​director of the Dutch research group Cultural Interactions


Notes on Contributors

in the Ancient World (ciaw, OIKOS) and editor-​in-​chief of the academic book series Cultural Interactions in the Mediterranean (cim, Brill, Leiden). For the latter series he has previously co-​edited the volume Empires of the Sea (2019). Harald Motzki (1948–​2019) studied in Bonn, Paris, and Cologne and wrote his Habilitation in Hamburg, where in 1989 and 1990 he was the substitute professor of Islamwissenschaft. From 1991 to 2011 he worked at Radboud University Nijmegen, first as associate professor and from 2001 as professor of the Methodology of Research in Islamic Studies. He has written more than a hundred books and articles. Petra M. Sijpesteijn is professor of Arabic at Leiden University. Currently she is a senior research fellow at the Historisches Kolleg in Munich. She is the principal investigator of an international research project entitled “Embedding Conquest: Naturalising Muslim Rule in the Early Islamic Empire (600–​1000),” funded by the European Research Council. Joanita Vroom is professor of the Archaeology of Medieval and Early Modern Eurasia at the Faculty of Archaeology, Leiden University, specializing in medieval and post-​medieval archaeology in the eastern Mediterranean and the Near East (including the Byzantine, Islamic, Crusader, and Ottoman periods). She takes a particular interest in the social-​economic (production and distribution) and cultural aspects (cuisine and eating habits) of ceramics in these societies and is series editor of the Medieval and Post-​Medieval Mediterranean Archaeology Series (mpmas) at Brepols Publishers (Turnhout). For her publications, see:​JoanitaVroom and https://​​ profile/​Joanita_​Vroom. For further information on her various (NWO) research projects and field projects, see:​en/​stafmembers/​ joanita-​vroom. Peter Webb is a university lecturer in Arabic Literature and Culture at Leiden University. His research investigates the evolution of Arab identity and Muslim narratives of pre-​Islamic history as developed in pre-​modern Middle Eastern writings. Peter is the author of Imagining the Arabs: Arab Identity and the Rise of Islam (Edinburgh, 2016), and the editor and translator of several classical Arabic texts for the nyp Press Library of Arabic Literature and Brill’s


Notes on Contributors


Bibliotheca Maqriziana series. He is currently the principal investigator of a Veni research grant from the Dutch Research Council, “Epic Pasts: Pre-​Islam through Muslim Eyes.” Johan Weststeijn studies the parallels between Greek myth and Arabic accounts of the Basus War. From 1992 to 2019 he was a student and guest researcher at the Department of Arabic and Islamic Studies of the University of Amsterdam. Clare Wilde (ab, Princeton; Licentiate, Pontifical Institute for the Study of Arabic and Islam; PhD, cua) is the author of Approaches to the Qurʾan in Early Christian Arabic Texts (Academica, 2014). She is a specialist in the history of Arabophone Christianity and early Christian-​Muslim relations, on which themes she has written various articles. She was the research associate for the first edition of Brill’s Encyclopaedia of the Qurʾan and has taught at Georgetown University, the University of Auckland, and the University of Groningen.

­c hapter 1

Late Antique Responses to the Arab Conquests: An Introduction Josephine van den Bent, Floris van den Eijnde and Johan Weststeijn The Arab conquests had far-​reaching and long-​lasting consequences. Not only did they usher in a new religion, but they also heralded a new cultural paradigm, as the Islamic caliphate came to encompass a territory stretching from central Asia to Spain. While this paradigm was indeed novel in many ways, it was also very much heir to what came before it. This assessment has not always received its due attention in the academic world. Twentieth-​century scholarship long took the historical Muhammad as its main point of departure, largely following Islamic tradition, even if it replaced a spiritual-​religious with a historical-​scientific approach. This methodology has had serious implications for the way in which early Arab/​Islamic culture has been studied, to some extent disconnected from either its own prehistory or its wider historical context. In the past decades, however, scholarship has changed considerably. Beyond the traditional historiographical approach, with its attention to literary Arabic texts written sometimes centuries after Muhammad and the initial stages of Arab expansion, there has been a distinct movement towards new readings of these texts as well as to alternative fields of research, such as papyrology, numismatics, architecture, and archaeology. This engagement with alternative sources and perspectives has integrated early Islamic culture into the late antique context in which the Arab conquests took place and Islam came into being, as well as into the early medieval world that it helped shape. This book brings together some of the approaches currently employed to study the impact of the Arab conquests on Mediterranean and Near Eastern cultures. Strongly embedding the rise of Islam within the context of Late Antiquity, it presents early Islam as a religion in-​the-​making that was entrenched within the cultural mosaic of its time, with as many points of connection as elements of departure from the antique past. In addition, we highlight the many new disciplines that have been added to the study of early Islam in hopes of contributing to a deeper understanding of the historical context in which the Arab expansion took place and from which Islam emerged as a dominant world religion.

© van den Bent, van den Eijnde and Weststeijn, 2022 | DOI:10.1163/9789004500648_002


van den Bent, van den Eijnde and Weststeijn


Islam and the End of Antiquity: Revisiting the Pirenne Thesis in the Age of Connectivity

For archaeologists and historians, periodization is indispensable. By generalizing eras through the assignment of certain properties, historical themes can be separated from each other or connected into manageable categories. Inevitably, such an abstraction of the past leads to oversimplification and blind spots. One of history’s major fault lines has traditionally been the transition from Antiquity to the Middle Ages. With the rise of Western European nation states in the eighteenth and nineteenth centuries it became received wisdom that “Western civilization” had its roots in Classical Antiquity, and more specifically in the civilizations of Greece and Rome. According to Enlightenment historians such as Edward Gibbon, a period of darkness set in after the fall of the Western Roman Empire on the fourth of September 476, whereafter reason was exchanged for military force, and law based on the written word for law based on the sword.1 The cause for this decay was found within the rise of Christianity on the one hand, and a so-​called “feminization” of the once martial Romans on the other. Henry Pirenne was one of the first and most important voices to argue against such a periodization.2 He argued that “[f]‌or centuries after the political collapse of the Roman Empire in the West, the economic and social life of Western Europe still moved exclusively to the rhythms of the Ancient world. Romania, a robust ‘functional Romanity’ (…), survived intact from the so-​ called ‘Germanic Invasions’ of the fifth century A.D.”3 Egyptian papyrus was still imported into Marseille from Byzantine-​controlled Alexandria, while textiles, papyrus, wine, and spices provided a “hint of the Roman elegance” that still defined high society in the eastern Mediterranean. This assessment ties in with a recent scholarly consensus that has coalesced onto a continuity of Roman political institutions, trade networks, and cultural forms, as evident in the establishment of a “Romanized” Vandal kingdom in northern Africa but most clearly in the “Justinian Renaissance,” which briefly re-​established Roman rule throughout much of the Mediterranean.4 Perhaps most vocally among recent scholars, Averil Cameron has reinforced Pirenne’s notion of a continued Romania beyond the fifth century, with reference to a variety of cultural indicators such as urban models, institutional forms, and cultural “mentality.”5 1 2 3 4 5

Gibbon 1776–​1788. Pirenne 1937. As condensed in Brown 1974, 26. E.g., Dark 2007, Cameron 2001. Cameron 2012.

Late Antique Responses to the Arab Conquests


Pirenne argued that the real break between Antiquity and the Middle Ages should be sought in the rise and military expansion of an Arab nation and the concomitant spread of a new Islamic faith. This new nation and the Islamic “thalassocracy” effectively disconnected Christian Europe from long-​distance maritime trade routes. Pirenne defined a new field: “Late Antiquity,” the period between the perceived decline of the Roman Empire in the fourth century and the initial stages of Arab expansion in the seventh. Through the conquest of large parts of Spain, North Africa, and the Levant, Pirenne argued, the traditional trading networks of the Mediterranean were disrupted, which led to widespread economic decline in Christian Europe. The fall of Carthage in 698 spelled doom for the international port of Marseilles, effectively cutting Gaul off from “Romania” and its Mediterranean horizons. The return to a virtually autarkic economy in parts of Europe proved to be fertile soil for the feudal political structures that heralded the coming of the Middle Ages. Pirenne furthermore observed that the downfall of Europe was paired with an economic expansion in the regions conquered by the Arabs. This so-​called thèse Pirenne is powerfully vocalized in an often-​quoted citation: “Without Islam, the Frankish Empire would probably never have existed, and Charlemagne, without Mahomet, would be inconceivable.”6 While the incorporation of Late Antiquity into the narrative of the ancient world has demonstrably benefited the field, pushing back the break between ancient and medieval to the seventh century has merely shifted the problem. One critique of Pirenne’s heavy emphasis on commerce has been that the disruption of trading routes should have had a similarly disruptive effect on the southern and eastern coasts of the Mediterranean, something it clearly did not.7 The “early medieval depression”8 of the seventh through ninth centuries—​a period Horden and Purcell aptly call the “Pirenne period”9—​was a European phenomenon with apparently little impact on the economy of an early Islamic state invigorated by conquest and oriented on trade with the Near East, rather than with the northern shores of the Mediterranean.10 6 7 8 9 10

This thesis is explained in depth in the first two chapters of the 1925 book Medieval Cities: Their Origins and the Revival of Trade (citation on p. 27; most recent edition Pirenne 2014, 17). An important and early critique of Pirenne’s thesis was written by Daniel C. Dennett, Jr.: “Pirenne and Muhammad” (1948). For a loving as well as sober consideration of Pirenne’s work, see Brown 1974. Cf. Ehrenkreutz 1972, 102–​4. Cf. Horden and Purcell 2000, 153–​72. Horden and Purcell 2000, 32–​34, 153–​72. Kreuts 1976, 87 speaks of the ninth century in particular as “an age of undisputed mastery over the whole Mediterranean” (i.e., for Islam). Cf. also Ehrenkreutz 1972, who attributes the comparative lack of Arab interest in Mediterranean trade to the severely weakened connectivity in that area. Sijpesteijn 2017, 654–​56 presents a brief overview.


van den Bent, van den Eijnde and Weststeijn

­f igure 1.1  Mediterranean shipwrecks by raphaelg, DATA source: parker 1992

It has, more importantly, been pointed out that the economic depression was a Mediterranean phenomenon that far predated the arrival of Islam. Its roots lie, in fact, not in the early medieval, nor even in the late antique period, but already in the High (Roman) Empire. Judging from a survey of Mediterranean shipwrecks, trade reached an all-​time high during the first centuries bce and ce, after which a gradual but persistent decline set in (Figure 1.1).11 A temporary economic boom may be witnessed in the Age of Justinian until the arrival of the bubonic plague in the late sixth and Persian and Arab invaders in the seventh century. Thus, viewed from a longue durée perspective it is arguably the comparatively short-​lived Justinian revival of the sixth century that stands out against the main trend, rather than the slump succeeding it. Additionally, the question remains how much the Arabs were, in fact, attempting to shut out Europe.12 ʿUmar ii (r. 99–101 AH/717–​20 CE), for example, issued a decree in which he professed that it was God’s will that the ports would be accessible to everyone, and that trade would not be hindered by state intervention.13 He was likely referring to trade between Syrian and Byzantine ports.14 Moreover, an edict was ascribed to the first ʿUmar (r. 13–23 AH/634–​44 CE) which 11 12 13 14

Parker 1992. Ehrenkreutz 1972, 97. Gibb 1955, 6. Gibb 1958, 231.

Late Antique Responses to the Arab Conquests


set taxation rates, including for foreign merchants, aimed specifically at traders from the Byzantine empire. They were to pay 10 percent, as this was reportedly the amount paid by Muslim traders in Byzantine territory. Both Byzantine and Muslim merchants were clearly trading across borders.15 And while much has been made of the disruptive effects of naval warfare in the seventh through ninth centuries, hostilities were probably far from constant in either magnitude or frequency.16 Muslim maritime dominance may furthermore hardly have been as complete as it has sometimes been depicted,17 nor will internecine strife of whatever type have been a great hindrance to maritime redistribution. In any case, there is evidence to suggest that Islamic and Christian tradesmen freely intermingled and even cooperated when opportunity arose.18 Horden and Purcell point to two further factors that should caution against blaming Islam for the evident economic depression that took place in the seventh, eighth, and ninth centuries: contemporary literary testimony and the misunderstood effects of piracy.19 One of the quintessential literary genres of the age, hagiography, recounts the lives of saints including some of their wanderings across the Mediterranean. From these accounts, a general picture emerges that shows the ease with which the protagonists enter Muslim-​controlled territories, as well as the short coastal “hops” that led them from harbor to harbor to their destination.20 The hagiography of St. Willibald moreover mentions the presence of an Egyptian ship in an Italian port in the 720s, close to the acme of the economic slump.21 Similarly, Gregory the Decapolite’s successful navigation through waters supposedly known for Muslim piracy is highly suggestive of an active pan-​ Mediterranean maritime infrastructure.22 From the other side of the supposed divide stems a famous passage by Ibn Khurdādhbih (d. third century AH/ninth century CE) on the Jewish traders known as al-​Rādhāniyya (Radhanites), who 15 16 17 18 19 20 21 22

Gibb 1958, 230. Sijpesteijn 2017, 656 attributes trade restrictions to the ninth century. As Horden and Purcell 2000, 155 point out, naval superiority was always “a highly localized phenomenon, without implications for the overall fortune of one side or the other.” E.g., Kreuts 1976, 87, who speaks of the ninth century in particular as “an age of undisputed mastery over the whole Mediterranean.” The Serçe Liman wreck for instance suggests that its crew may have been of mixed Muslim-​Christian stock (van Doorninck 1991). While this wreck is dated to the eleventh century, it shows that socio-​political factors were not an overruling impediment to trade. Horden and Purcell 2000, 157–​66. Note for example the exploits of Arculf, a Burgundian bishop, who found “the commerce of the whole world” in Alexandria (Adamnan, On the Holy Places, 30.16, ed. Meehan 1958, 102–​3; Bede, Ecclesiastical History, 5.15). Cf. Horden and Purcell 2000, 161. Hodoeporicon (The Itinerary of St. Willibald; trans. Talbot 1954, 153–​77). Cf. Horden and Purcell 2000, 161. Malamut 1993, 247–​48, as quoted by Horden and Purcell 2000, 161–​62.


van den Bent, van den Eijnde and Weststeijn

were ostensibly multilingual and maintained a trade network stretching from Francia, through Egypt and Arabia, all the way to China.23 A levy granted to the Abbey of Corbey in 712 by Chilperic ii confirms the picture that commerce from the east could still reach the port of Marseille in the eighth century.24 Finally, Pirenne’s allegation that Muslim piracy was the main impediment to international trade has been shown to be problematic. As David Abulafia has rightly pointed out, a persistence of trade relations is a prerequisite for piracy, with the latter predicated on the former, the predator unable to exist without prey.25 Nor can the scale of piracy ever exceed a certain ratio of commerce, unless the predator is willing to cannibalize his own livelihood and perish: “Pirates might pick off one or two ships but were highly unlikely to capture the whole lot.”26 The piracy argument moreover overlooks one of the main aspects of buccaneering in any age, the fact that pirates often are tradesmen themselves: pirates need markets.27 Like government requisitions, piracy is thus just another form of economic redistribution.28 And lest we forget, piracy was never a one-​sided (i.e., Muslim) affair, as is evident from the surviving anecdote of a Byzantine ship raiding a merchant ship making its way to North Africa.29 The fact remains that neither piracy nor war were characteristics uniquely attributable to either the early Islamic or even to the Pirenne period as a whole; they were and always had been part of Mediterranean life (periodic attempts to quell the phenomenon notwithstanding). Judged by contemporary standards, the piracy argument is thus revealed as an unsuccessful attempt to discredit early Islam by blaming it for the early medieval depression—​pirates after all are easily cast as “the other.” All this is not to deny that the Pirenne period marks a low tide in Mediterranean connectivity, but rather to emphasize the changing nature of that connectivity. Ports that previously had been bustling with international commerce—​ Marseille, Naples, and Carthage—​ were reduced to regional importance or vanished outright.30 But new actors were about to enter the 23 24 25 26 27 28 29 30

Ibn Khurdādhbih 1889, 153. See also Horden and Purcell 2000, 162. Pertz 1872, no. 86, with Horden and Purcell, 163 and the numismatic (Hendy 1988, 68–​69) and archaeological (Citter et al. 1996) evidence. Abulafia 1987, 415. Horden and Purcell 2000, 157. See also Udovitch 1978, 541: “The scourge of piracy, while fearsome indeed for an unfortunate victim, did not result in any paralysis of maritime commerce.” For a treatment of Mediterranean piracy in the second and first millennia bce, see Van den Eijnde 2019; De Souza 1999. For the medieval period, see Heebøll-​Holm 2013; HeebøllHolm 2019. Horden and Purcell 2000, 158: “Piracy is the continuation of cabotage by other means.” Theophanes Continuatus, Chronicle (5.63–​64, ed. Bekker 1838, 304–​5). Drauschke 2011; Sijpesteijn 2017, 652–​53.

Late Antique Responses to the Arab Conquests


arena, savvily adapting to new realities, in particular the rise of Islam as a dynamic new force with strong ties to the Near East and southern Asia.31 In Africa, Carthage fell to the Umayyads in 698 and was destroyed more comprehensively even than it had been at the hands of the Romans in 146 bce; it was never to recover again. Fearing recapture of the province by the Byzantines, the Arabs reconstituted local control in the new administrative center of Tunis on the opposite end of the great laguna. In Italy, Venice and especially Amalfi were among the first to profit. Amalfi established its position through a consistently pro-​Muslim policy, even as Arabs raided the Italian coasts in the ninth and tenth century. The Amalfitans were rewarded with special commercial concessions and trade rights in Arab ports.32 It is interesting, though perhaps not surprising, that the rise of these new European entrepôts coincides exactly with the heyday of the alleged Muslim thalassocracy. Both in the East and in the West, contacts between Christian and Muslim nations were largely mediated by Christian merchants, rather than Muslims venturing into the dār al-​ ḥarb.33 In the East, Cyprus played a crucial role as a redistribution center for both the Byzantines and Arabs, while trading in the West was now confined to the narrow strait between Ifrīqiyya and Sicily, with the Amalfitans eagerly playing the part of middlemen. Unsurprisingly, the latter were known as much as merchants as privateers, operating as proxies in name of their Muslim allies or operating independently depending on the circumstances.34 If we concede that a quantitative decline in connectivity characterized the Pirenne period, it nevertheless remained dynamic and adaptive to new circumstances, with shifting shipping routes marking rapidly changing consumer markets. A major problem in bringing this vibrancy to light is the fact that the Pirenne period has at times been undervalued by archaeologists (but see the contributions by Al-​Jallad and Vroom in this volume), its material culture underappreciated in comparison to what came before and after it. Both field surveys and excavations generally have neglected to focus on this period—​ owing in part, one is reluctant to admit, to its underwhelming decorative traits. What is more, low-​level connectivity of the kind prevalent in this period is especially hard to prove as it leaves little trace in the material record: coinage may be limited when trade is often in kind, shipwrecks are few and far between, non-​durable materials are used, and pottery in particular is often too stylistically “crude” to be employed in meaningful analysis. Speaking of a 31 32 33 34

Cf. Sijpesteijn 2017, 654–​56. Cf. Wickham 1981, 150. Horden and Purcell 2000, 168. Abulafia 2014, 141–​42; Heebøll-​Holm 2013, 25, 83–​90.


van den Bent, van den Eijnde and Weststeijn

collapse in the volume of redistribution may therefore be too rash. The small scale “hops” alluded to in the hagiographic account of the Burgundian bishop Arculf may leave less of a material footprint than the long-​distance voyages of the sixth century, but they maintain connectivity nonetheless.35 Pirenne may thus be rightfully criticized for substituting one migratory movement (that of the fifth-​century Germanic invasions) with another (the Arab conquests in the seventh) as the effective cause of Antiquity’s demise. But this misconception was mostly a consequence of mixing cause and effect. Pirenne was probably right in seeking a significant chronological extension of Romania—​that is, of the Roman cultural traits transmitted by major urban centers of the ancient world—​well into the early medieval period. The coup de grâce, however, was not delivered by marauding Arab invaders, but by a gradual weakening of Mediterranean connectivity that had its roots in the High Roman Empire. What caused this waning of socio-​economic, cultural, and political ties is difficult to assess and is not the subject of this book. It will, however, not have been monocausal and should be understood as having taken place over a very long period of time, a movement that required the full first millennium to unfold. Returning to the Arab conquests, which is the subject of this book, we may therefore conclude that it was not the conquests that caused the depression, but that the depression (or more precisely, the weakening of Mediterranean connectivity) created the right conditions for the Arab expansion to ignite. This book seeks to understand that moment of ignition as part of its late antique milieu or rather within the context of the remnants of traditions that reach back centuries and more. The contributions to this volume all in one way or another seek to explain how early Arab culture dealt with these remnants and how those remnants were in turn transformed by conquest. 2

The Impact of Empire: Tradition and Change

By narrowing the resolution of our historical lens from the first millennium to only its sixth and seventh centuries, our focus shifts from continuity to disruption and from unity to fragmentation. If the longue durée perspective of the previous section has invited us to look for persisting connectivity, we must now focus our attention on the realities of conquest and empire, both lost and found, set in motion by changes in connectivity. During the sixth-​century 35

See note 20.

Late Antique Responses to the Arab Conquests


Justinian Reconquista, Byzantine naval power provided a brief reprieve from the long-​term trend of weakening Mediterranean ties, ensuring that existing maritime connections remained protected and operable. But the amputation of its African and Asiatic limbs drastically curtailed the naval capabilities of the Byzantines. Likewise, their ability to draw upon the resources of the Mediterranean and Near East were severely impaired as well. Joanita Vroom’s contribution to this volume shows that, for the Byzantine world at least, connectivity had to be rerouted to a certain extent to accommodate its new, strongly diminished dimensions. Items that were previously imported from the Levant or Egypt had to be found within the contracted borders, exchanged for more readily available items, or allowed to disappear altogether. That the importance of the Byzantine fleet in maintaining the empire cannot be overemphasized is borne out by the observable fact that it maintained a tight grip on the Dalmatian coast, even at a time when Slavic tribes overran the rest of the Balkans.36 As long as the fleet remained intact, coastal areas could, within limits, stay protected. Even so, there may well have been a macro-​economic aspect to complete this picture of increased “autarky.” With its fiscal position inherently impaired by the contraction, there may have been a protectionist movement, whether deliberately decreed by Constantinople or imposed upon the commonwealth by economic necessity. In any case, it cannot be a coincidence that the Pirenne period, with its dramatically impaired access to material goods, is commonly seen as a turning point in Byzantine art and culture.37 A similar analysis may doubtlessly be employed for the other Christian regions—​Spain, Gaul, and Italy. Thus, the individual contributions to this volume identify both continuities and discontinuities. At a very minimum, however, this mosaic of case studies shows that a deep, gaping abyss between (Late) Antiquity and the Middle Ages—​whether conceived of as the result of German migrations, following Gibbon, or of the rapidly combusting Arab revolution, following Pirenne— cannot be maintained. The abyss most certainly did not exist for the people living in these times. For them, present and past comprised a continuum of laws and customs, of shapes and symbols, and above all, of a deep-​rooted conviction that the world was God’s unique creation. While the rapidly evolving Islamic civilization generated a unique architecture, an idiosyncratic corpus of literature, and a revolutionary new religious order, early Arab culture could 36 37

Ferluga 1987, 42–​47 describes the continued use of amphorae, glass, and ceramic iron utensils of Byzantine origin. Cf. Haldon 1990. The caesura from early to middle Byzantine art, for example, is generally placed ca. 730.


van den Bent, van den Eijnde and Weststeijn

build upon the existing, and transform the ancient into the new. As some of the authors argue in this book, many old habits took a long time to die, as new ones gradually began to emerge beside them. A quick glance at the Umayyad Mosque in Damascus shows that the first dynasts of this latest empire to claim universal authority imagined themselves as the inheritors of an ancient culture with deep roots in the Greco-​Roman past. Built like Byzantine basilicas, with a large nave, aisles, Corinthian capitals, mosaics, and opus sectile, this first “Islamic cathedral” proclaims a message consisting of well-​recontextualized traditions. Resting upon the foundations of an early Christian basilica dedicated to John the Baptist, the current building still contains a shrine in which the “original” relics of the Baptist are kept. How then to explain the gap that has emerged in later perception? This question does not allow an easy answer, but part of an answer must be found in the period after the earth-​shaking conquests of Muhammad and his descendants, in the inward-​focused world of early medieval Europe. Peter Brown has cogently argued that Christian Europe was one of the few places in the world in which people were not taken by the dazzling dynamics of Arab expansion, where people considered Islam as the photonegative of themselves—​not something to admire, or even to respect, but a despicable heathendom.38 The image of Islam as an anti-​Christian counterpoint that was to be utterly annihilated is in a certain way the real break between Antiquity and the Middle Ages. This break, as Edward Said had shown long ago, has never healed.39 When the nascent Western European nation states started to reinvent themselves according to a largely imagined tradition of a communal Greco-​Roman heritage—​the emergence of Antiquity’s Nachleben—​a recurrent theme was the search for an explanation for the colonial supremacy of Europe. Why was the Islamic world unable to oppose this European expansion? The answer was sought and found in the idea that the Islamic world had turned itself away from the foundation of civilization, laid out in Antiquity. With that, blame for the schism between East and West was retroactively, and squarely, placed upon the Arab expansion. Echoes of these ideas, of a schism between East and West and a schism between Antiquity and Islam, are still prevalent in Western popular perception, politics, and at times even in scholarship. In recent decades, however, more and more scholars have come to recognize the importance of viewing the development and events of early Islam—​religious, economic, political,

38 39

Brown 1979. Said 1978

Late Antique Responses to the Arab Conquests


cultural—​in light of the broader contexts of Late Antiquity. This volume joins them in this focus and argues that Antiquity did not suddenly end with the rise of Islam (emphasizing that the Arab conquests did not originate in a vacuum), but rather took place in the context of a culturally integrated late antique Near East. For a proper understanding of the period and its developments, a perspective in which Islam is studied within the context of Late Antiquity, or even as part of a narrative spanning the first millennium, is essential.40 3

A “Late Antique Turn”

For a long time, since the beginning of the nineteenth century, scholars of Islam and Arabic culture and history generally shared the traditional, Islamic idea that the emergence of Islam took place within relative isolation on the Arabian Peninsula during the life of Muhammad. Following the subsequent conquests, the new but already mature religion spread across the Near East, the Mediterranean, and beyond. This view of early Islam as well as the origins of the Qurʾān primarily followed Islamic, Arabic-​language narrative reports about this early period. Historiographically, however, these sources are riddled with challenges. The accounts contain numerous contradictions within and between source texts as well as other indications that they might not offer a reliable representation of events, exhibiting, for instance, anachronisms, a clear interest in numerological symbolism, topoi, and various other issues. Moreover, the stories about the early Islamic period were not put down in writing until the late eighth or ninth century CE, so that they were strongly colored by later concerns. By the time these sources were put into writing, their authors had an Islamic self-​awareness that led them to present the rise of Islam as a rigorous break with the past and suggest that Islam was fully “developed” as such from the very start. In the same vein, they interpreted the Qurʾānic text from their own cultural and political perspectives.41 Yet, as early as the nineteenth century, scholars such as Gustav Weil, Ignaz Goldziher, and Theodor Nöldeke took a critical approach towards the Islamic source material. Similarly, the nineteenth century saw some scholars 40 41

Fowden 2015. For a discussion of various aspects of these source problems, see for instance Donner 1998, 1–​31; Donner 2010, 50–​52; Reynolds 2010, 18–​22. These problems, especially source material being colored by later concerns, are not exclusive to accounts of Muhammad’s life and times but also appear in reports on later events of early Islamic history. See for instance Noth 1968; Robinson 2000.


van den Bent, van den Eijnde and Weststeijn

recognize the importance of looking at the temporal and cultural contexts of early Islamic history. The first steps in the process of placing the Qurʾān and the rise of Islam within a late antique perspective date back to the work of Abraham Geiger in the 1830s. In an essay entitled Was hat Mohammad aus dem Judenthume aufgenommen?, this German rabbi and scholar investigated the relationship between the Qurʾān and rabbinic literature.42 After a few interesting advances in the examination of connections between Judaism and Islam, this line of research was interrupted in the 1930s, as a result of the persecution of Jewish scholars by the Nazis. In the decades that followed, scholars abandoned the historical inquiry of the emergence of the Qurʾān within its late antique context and shifted their focus primarily to the person of Muhammad.43 From the 1970s onward, however, a number of publications cast doubt on several or even many of the elements of the traditional Islamic narrative. Referencing the reliability problems of the Islamic sources, these works tended to dismiss the traditional narrative, in some cases discarding it altogether, focusing on outside sources instead. The authors, who devised entirely new interpretations of early Islamic history, are also known as “Revisionists”—​even though they did not and do not constitute a coherent group with a singular argument. In fact, their arguments and conclusions frequently contradict one another irreconcilably. The term “Revisionist,” therefore, does not convey much more information than the fact that the authors reject the traditional narrative. The theories themselves are rather controversial, with some leading to heated debate within academic circles while others were outright rejected; at times the theories have even attracted the attention of the popular media. In 1974, Günter Lüling published Über den Ur-​Qurʾān, in which he argued that the Qurʾān consists of different layers: Christian strophic hymns at its base, which were later covered by Islamic layers of different dates.44 More followed in 1977, with the publication of John Wansbrough’s Quranic Studies and Patricia Crone and Michael Cook’s Hagarism. Wansbrough, in his infamously obtusely written Quranic Studies, challenged the traditional narrative in a

42 43 44

Geiger 1833. See also Sinai and Neuwirth 2010, 3–​5. For an overview and analysis of research and researchers in this period, see Stewart 2017. See also Sinai and Neuwirth, 2010, 3–​7; Neuwirth 2014, 6–​10. Lüling 1974. See also Lüling 2003, a translated and expanded edition of this work, entitled A Challenge to Islam for Reformation. For a more extensive summary of Lüling’s theory, see Reynolds 2008, 10–​11.

Late Antique Responses to the Arab Conquests


variety of ways, but the best known is his assertion that the Qurʾān was not canonized until the second/​eighth or third/​ninth century in ʿAbbāsid Iraq.45 Hagarism presented an entirely different theory. Crone and Cook, having completely set aside all Islamic sources on account of their unreliability, used contemporary external sources to reconstruct early Islamic history, suggesting that Islam originated from a Jewish messianic movement, the so-​called Hagarenes (muhājirūn).46 More “Revisionist” theories followed after the turn of the millennium. The first, and arguably most famous, was Christoph Luxenberg’s (pseud.) Die syro-​ aramäischen Lesart des Qurʾan (2000). Pointing to the generally accepted fact that Syriac influenced the language of the Qurʾān, he argued that the Qurʾān was written in a mixed language of Aramaic and Arabic and consequently reread Qurʾānic passages from a Syriac, Christian perspective. His argument that the houris of paradise in Q 44:54 are really bunches of grapes especially attracted ample and enthusiastic attention from Western media outlets.47 A few years later, in 2003, Judith Koren edited Crossroads to Islam, based on the work of Yehuda Nevo, who died in 1992. They argued for a planned but secret, centuries-​long Byzantine withdrawal from its eastern provinces, the power vacuum being filled by their Arab vassals whose religion slowly developed from “indeterminate monotheism” into Islam.48 The proposed “Revisionist” theories, the best known of which have been mentioned here, thus vary wildly and are in many cases mutually exclusive. All share, however, the controversy they generated—​both within and without academia.49 What these “Revisionists” do have in common—​besides this controversy and their departure from the traditional narrative—​is that they devote a great deal of attention to the late antique context in which the Qurʾān and what we now know as Islam emerged. They look at the text of the Qurʾān


46 47 48 49

Wansbrough 1977, followed in 1978 by The Sectarian Milieu. For a brief summary of and some insights into his theses, see for instance Reynolds 2008, 11–​13; Stewart 2017, 18–​19. See also and cf. Sinai and Neuwirth 2010, 7–​10. Their respective discussions of the effect of Wansbrough’s theses on the idea of Qurʾānic chronology gives insight into some of the current debates in the field. Quranic Studies was re-​released in 2004 by Prometheus Books (Amherst, NY), with a foreword by his student Andrew Rippin, and includes explanatory notes and English translations of the quotations used. Crone and Cook 1977. Luxenberg 2000. For some responses to his work, see for instance Böwering 2008; Saleh 2010; and Wild 2010. See also Stewart 2017, 20–​23. Nevo and Koren 2003. For a summary of their argument, see Reynolds 2008, 13–​14. For some general criticisms of pre-​1998 “Revisionists,” or what he calls the “skeptical approach,” see Donner 1998, 25–​30.


van den Bent, van den Eijnde and Weststeijn

from that perspective, and their distrust of Islamic sources has led them to focus on non-​Islamic texts as well as material evidence. In that sense, these theories have posed a challenge and have helped breathe new life and a new sense of urgency into the field. New lines of research opened in the last decades of the twentieth century, but it was in the first decade of the twenty-​first century that they truly gained momentum. A number of conferences were organized, at least in part in response to the “Revisionist” challenge—​taking up the proverbial glove, as it were. The 2004 conference “Historische Sondierungen und methodische Reflexionen zur Koranexegese—​Wege zur Rekonstruktion des vorkanonischen Koran” and the 2005 conference “Towards a New Reading of the Qurʾān?” both resulted in edited volumes with contributions that pay ample attention to the late antique context of the Qurʾān.50 This trend has stuck: Islamicists, Arabists, historians, and other scholars have increasingly incorporated this perspective into their research. Their work acknowledges the importance of this context not only for insights into the text of the Qurʾān itself, but also to shed light on the development and spread of this new religion as well as social, cultural, and political trends in the early Islamic period. Given the wide range of this scholarship—​both in subject matter and in field of research—​constraints of space prohibit an elaborate overview and analysis of individual contributions and contributors.51 However, two clear trends can be observed: renewed readings of the Qurʾān and an interest in the use of different source types, such as non-​Islamic written sources, inscriptions, archaeology, and coinage. Like Islamic/​Arabic literature, these sources are not without interpretative challenges of their own, but they do offer new vantage points. The recognition of the importance of connections between late antique cultures and religions in combination with a multidisciplinary approach offers new insights and has made this field of research a fast-​moving and exciting one. The re-​reading of the Qurʾān takes many different shapes. One approach is a philological perspective, in which scholars look at linguistic elements of and

50 51

Neuwirth et al. 2010 and Reynolds 2008, respectively. In what follows, we will only give a few examples in each category. These publications, and/​or other works by their authors, can then be consulted for further references. See also the bibliographies of the individual contributions to this volume. An elaborate analysis of recent developments in the field of Qurʾānic studies can be found in Stewart 2017, 19–​53. Incidentally, among the nine influential trends that he identifies is one that he calls the “Late Antiquarians”: this trend is the least clearly defined of the nine and—​as Stewart himself acknowledges—​rather overlaps with various other groups.

Late Antique Responses to the Arab Conquests


influences on the Qurʾānic text. A particular interest in this respect in Syriac influences closely connects to another way of re-​reading the Qurʾān: looking at its relationship with other texts, primarily religious, circulating in the region.52 Other scholars have focused on internal elements of the text, while keeping its late antique context in mind. By looking carefully at the Qurʾānic text, Fred Donner for instance has attempted to gain insight into the beliefs of the early followers of the nascent movement that would later become Islam, seeing it as a “Believers” movement that originally included Christians and Jews.53 Others have renewed scholarly efforts to look at the Qurʾān’s composition through its literary elements, such as rhyme and structure.54 Another approach, with Michel Cuypers as its main advocate, has focused on the concentric structure of (parts of) suras, which has been (somewhat unfortunately) dubbed “Semitic rhetoric.”55 At the intersection of Qurʾānic text and new source material is the research into early Qurʾān manuscripts (or rather: fragments thereof), whether discovered relatively recently or unearthed from long-​held collections in European libraries. These manuscripts shed light on the history of the text itself—​and will hopefully, in the not too distant future, contribute to a long-​awaited critical edition of the Qurʾānic text.56 A new focus on source types other than the traditional Islamic texts is similarly helpful in better understanding the Qurʾān. These source types include the abovementioned non-​Islamic writings—​in a variety of languages (Syriac, Hebrew, Ethiopic, Persian) and of different religious (Christian, Jewish, Zoroastrian) and political provenance—​but also pre-​ Islamic Arabian texts such as inscriptions.57 The Qurʾān and the earliest days of Islam are thus now understood more and more in their late antique context. Yet, these developments had relevancy and consequences outside the Arabian Peninsula as well, as they were followed by what is commonly known as the “Arab conquests.” Communities living in areas that had previously been under Byzantine or Sasanian control—​ Christians, Jews, Zoroastrians, and others, of various sects and speaking different languages—​now passed under Arab dominion. To trace how these populations and their cultures interacted with the new conquerors and their

52 53 54 55 56 57

A few examples: Reynolds 2010; Griffith 2008; El-​Badawi 2014. See also various contributions to Reynolds 2008 and Neuwirth et al. 2010. Donner 2010. A few examples: Neuwirth 2014; Stewart 1990; idem 2011. See for instance Cuypers 2012. A few examples: Sadeghi and Goudarzi 2012; Déroche 2013; Hilali 2014. A few examples: Al-​Jallad 2015; Al-​Jallad et al. 2017; Nehmé 2017.


van den Bent, van den Eijnde and Weststeijn

nascent religion, scholars have similarly turned to a much wider variety of sources than the Arabic literary texts. The conquests themselves, their events, causes, and consequences are revisited,58 but so too are the late antique contexts of early conquest societies. How did existing communities in conquered territories and the newcomers perceive and influence each other? How did contact between communities affect the way various cultural and religious groups defined and identified themselves? What continued and what changed in those conquered societies, and when and why? These questions are very broad and cover a wide variety of historical processes and social structures, which means that various scholars are each trying to answer (parts of) these questions for (parts of) the conquered territories.59 What these studies show is that although the Arab takeover indeed brought changes to the conquered territories and populations, the hard break envisioned in past scholarship does not hold. On the contrary, early Islamic history was very much part of wider late antique developments, in which the religion of Islam and Arab identity formed gradually. The Qurʾān is thus a product of its late antique surroundings, and the transition from the cultures of Late Antiquity to those of Islam was a protracted and gradual process. Scholars have been able to come to and corroborate these insights to a significant extent by turning to previously little used and/​ or recently discovered source material. The information offered by early Umayyad coins is of a different type than that which can be gathered from Syriac Christian apocalyptic text, the study of a rock inscription, or a receipt on papyrus. Yet they all offer up some evidence, some piece of the puzzle, shedding new light on this historical period, especially when used in conjunction with a critical engagement with the Arabic/​Islamic narrative sources. We do not yet have a new, widely accepted “grand vision” of the rise and earliest history of Islam. But more and more pieces of it have been recovered, reinterpreted, or replaced.

58 59

For a discussion of some questions and debates on the topic of the conquests, see Donner 2008. See also for instance Robinson 2004; Hoyland 2014. A few examples: Donner 1998; idem 2010; Penn 2015 (on the gradual development of Islam and relations with other religious communities); Webb 2016 (on the gradual development of Arab ethnic identity); Hoyland 1997 (on outside visions on Islam); Levy-​Rubin 2011 (on the position of non-​Muslims); Sijpesteijn 2013 (on post-​conquest Egypt); Robinson 2000 (on the post-​conquest Jazīra).

Late Antique Responses to the Arab Conquests



A Survey of the Contributions to This Volume

This book offers a number of new pieces for the puzzle. We have elected to incorporate a diverse selection of research questions, source material, and methodologies, in order to showcase the range of current scholarship and its great strides. As a relatively young discipline, much remains to be discovered and developed. The broad approach of this volume also highlights that this field of study, on the interface of Late Antiquity and early Islam, requires a multifaceted and comprehensive approach. Only then will it succeed in contextualizing this time period and the many important contemporary developments without being hindered by received ideas about the period. At the same time, such an integration of approaches and methodologies also enables scholarship to deal with the many different challenges posed by the various types of source material. The contributors to this volume have chosen different ways to tackle the problems inherent in the Arabic narrative sources. Both Clare Wilde and Johan Weststeijn have elected to ignore the narrative, literary Islamic sources in their investigations of how the Qurʾān’s first audience would have understood certain Qurʾānic passages. Harald Motzki, by contrast, subjects those very texts to a critical internal analysis. Other contributors consider non-​narrative source types, available across the transitional period from Late Antiquity to Islam, allowing them to compare the situation before and after the Arab conquests. The sources considered range from Arabic poetry (Peter Webb) and Egyptian papyri (Petra M. Sijpesteijn) to rock inscriptions (Ahmad Al-​Jallad) and pottery remains (Joanita Vroom). Their findings highlight three elements in the relationship between Late Antiquity and early Islam: 1) the late antique characteristics of the Qurʾān; 2) the gradual transition from late antique to Islamic cultures—​despite the relative speed of the conquests themselves; and 3) the profound impact of the Arab conquests on existing cultures that did not yield wholly (or at all) to the newly emerging Arabic cultural forms. Moreover, the transition from Late Antiquity to Islam was no clean break. Rather, processes of Arabicization and Islamization were incremental, as Kevin van Bladel, Webb, Sijpesteijn, and Al-​Jallad demonstrate in their studies of ethnicity, language, script, and religion. Not only the Arabicization and Islamization of the conquered was piecemeal, but also the culture of the conquerors was similarly subject to gradual change and was very much in motion during this transitional phase. Among these contributions, Van Bladel stands out: rather than advocating the use of new sources, he proposes a new model to explain the processes of Arabicization and Islamization of the conquered.


van den Bent, van den Eijnde and Weststeijn

4.1 The Qurʾān as a Late Antique Text Clare Wilde and Johan Weststeijn squarely read the Qurʾān in the context of its late antique historical environment: a different approach than the traditional, medieval Islamic exegesis. In traditional exegesis of the Qurʾān, religious scholars try to explain difficult Qurʾānic verses by linking them to events from Muhammad’s life, as they are known from the narrative literary sources that constitute the traditional biography of the Prophet. It is possible, however, that these narratives were invented only later or in hindsight, in order to clarify exactly those unclear Qurʾānic verses, by Islamic authors for whom the late antique context of those verses was no longer understandable.60 Clare Wilde considers Sura 30:2–​5, the only passage in the Qurʾān that mentions al-​Rūm, “the Romans,” i.e., the Eastern Christian Romans or Byzantines. These verses mention a defeat and a victory of the Romans “in the nearby land” and are traditionally interpreted as referring to the military battle for Jerusalem, which the Byzantines lost in 614 in the last Persian-​Byzantine War but then reclaimed in 628. According to Wilde, the first audience of the Qurʾān could have interpreted this passage about a defeat and a victory of the Byzantines in the “nearest land” differently than the later medieval Islamic exegetes. Wilde argues that this earliest audience could potentially have understood the defeat and victory of the Byzantines not only as a military conflict but also as a conflict of doctrine. Namely, as the failure and success of the attempts of the Byzantine emperors to reconcile the different groups of Eastern Christians in their heated theological disputes, beginning with the ecumenical Council of Chalcedon in 451. The “nearest land” was most probably not interpreted by the first audience as Jerusalem—​which is seen, according to the Qurʾān, from the viewpoint of Mecca as the location of the “farthest house of prayer”—​but as southern Arabia. That too was a battlefield in the Persian-​Byzantine War, in which Ethiopians, supported by the Byzantines, fought with the Yemenites, supported by Persia. Like Wilde, Johan Weststeijn argues that parts of the Qurʾān may have had a different meaning to their late antique audience than they did to medieval interpreters, who lacked this late antique context. In contrast to Wilde’s historical context, Weststeijn examines the cultural component of the late antique view of the world, applied in a verse about wine. Aside from three dismissive passages about alcohol, the Qurʾān contains one verse (16:67) that, according to the majority of medieval exegetes as well as modern Western handbooks, is positive about wine. Medieval exegetes explained the paradox between this 60

See Motzki, ­chapter 4.

Late Antique Responses to the Arab Conquests


positive verse and the other negative verses by pointing towards the traditional biography of the Prophet: these verses would have been revealed at different moments in Muhammad’s life. During his activities as a prophet, Muhammad’s contemporaries would increasingly have misbehaved under the influence of wine, and thus the Qurʾān became more critical of this beverage over time. However, Weststeijn argues that if we place verse 16:67 in the context of late antique ideas about the oppositions between pure and impure, it becomes clear that this verse presents wine as something impure: in line with the other argumentative passages in the Qurʾān about the beverage, this verse is negative about wine as well. While Wilde and Weststeijn elect to pass over the Prophet’s biography, Harald Motzki endeavors to date those biographical sources through an internal analysis and to test their historical accuracy. Motzki is well known for his new approach to these texts, which he called the isnād-​cum-​matn method, which combines the study of the chain of transmitters (isnād) with the study of the contents (matn) of a report. By comparing different versions of a report about, for example, the Prophet, and their different chains of transmittance, he tries to reconstruct and date the original text of that report. In Sura 52:29–​ 30 and 74:24, the Qurʾān argues against the criticism that its messenger was a soothsayer, a madman, or poet. According to a report in the traditional biography of the Prophet, these verses were revealed after a discussion among Muhammad’s opponents about how they could best slander his good name and, thereby, stop the spread of his message. Has this report been created in hindsight to explain those Qurʾānic passages? Motzki arrives, through his historical-​critical method, at the conclusion that the report is much older than source critics suppose, and that it is most likely that Muhammad’s opponents indeed painted such a negative picture of him. 4.2 Late Antiquity to Early Islam: A Gradual Transformation Other authors focus on the early days of the young Arabic-​Islamic empire, and its interactions with the late antique cultures of Egypt, Syria, and Iran. They not only show how the conquests changed both conquered and conquerors, but also emphasize the gradual pace of Arabicization and Islamization. Criticizing existing explanations for these processes, Kevin van Bladel proposes a new model to explain differences in the rate and speed of Arabicization and Islamization in the different conquered regions. Rather than reasons of ideology and identity, Van Bladel argues that the differences are the result of variations in the settlement patterns of the conquerors. Van Bladel examines the local demographic state before, during, and after the conquests. Both Arabicization and Islamization occurred fastest in those places where the


van den Bent, van den Eijnde and Weststeijn

Arabs spread themselves over the conquered areas so that many points of contact with the local population emerged. This was, for example, the case in Spain, but also in the Levant, where a rapid transition to the Arabic language was additionally facilitated by the earlier emigration of the Greek-​speaking, Christian elite towards the remnants of the Byzantine Empire. By occupying vacated dwellings, the conquerors mixed rapidly with the remaining, and often less educated, Armenian populace. In places where the conquerors isolated themselves from the local population, such as initially in Egypt and Iraq, the process of acculturation was considerably slower. The conversion to Islam usually followed only after Arabicization had already taken place. Practical considerations played a larger role here than identity. Important sources for this research are Arabic texts written by individuals who had not yet converted to Islam. These sources give an important insight into the lived experiences both before and during the processes of Arabicization and Islamization. Peter Webb shows that the conquests not only affected the conquered but also changed the conquerors. He compares Arabic poetry from before and after the conquests. It is noteworthy that from the pre-​Islamic period, much more Arabic poetry than Arabic narrative prose has been preserved that can within reason be considered as authentic. Webb demonstrates that the ethnonym “Arab” is not found in Arabic poetry before the conquests but only in Arabic poetry from the end of first century ah/​seventh century ce onward. Webb draws the conclusion that the conquests caused the different ethnic groups that formerly inhabited the Arabian Peninsula to consider themselves as belonging to a single Arab ethnos. Just as Webb compares Arabic poetry from before and after the conquests, Petra Sijpesteijn compares Egyptian papyri from the same vantage points to study the effect of these conquests. These papyri are often administrative sources that have been well-​preserved by the dry desert climate. In contrast to the later literary sources, these papyri deal with common daily practice, and so have not been rewritten for later political, theological, or literary goals. For example, these papyri show which foods the Arabic conquerors ordered from their Egyptian subjects, and therefore which foods these new Muslims actually consumed. The conquerors ordered, for example, large amounts of syrup from boiled grape must (a by-​product of wine production) for their workers. This syrup is interesting, not only considering the negative opinion towards wine in later Islamic law, but also considering Weststeijn’s argument that Q 16:67 (“from the fruits of the palms and the grapevines you make intoxicant and good provision”) was considered by the first, still late antique audience of the Qurʾān to denounce wine production and consumption.

Late Antique Responses to the Arab Conquests


One conclusion to emerge from these papyri is that the religious identity of the conquerors did not play an important role at first: the distinction between the new Arab rulers and their Egyptian subjects was not so much expressed in religious terms, but rather in political-​economic terms. Additionally, this research highlights particularly the continuity that existed in administrative practice after the coming of the Arabs. Only at the beginning of the eighth century, argues Sijpesteijn, were administrative practices introduced that aimed for a greater Islamization and Arabicization of the Islamic Empire. Ahmad Al-​Jallad works with epigraphic sources, examining in particular a unique Arabic inscription that reads: “May God be mindful of Yazīd-​w the king.” Using linguistic and palaeographic elements he subsequently dates the inscription and identifies “Yazīd the king.” Doing so, he concludes that the inscription dates to the seventh century. This date suggests that the Yazīd mentioned in the inscription is the Umayyad caliph Yazīd i (r. 60–64 AH/ 680–​83 CE). Moreover, Al-​Jallad examines, from the perspective of the Yazīd inscription, the development of the Arabic script, and concludes that possibly different Arabic scriptural traditions coexisted before the Islamic-​Arabic variant gained the upper hand. This contribution is a case study from Al-​Jallad’s broader research on early Arabic inscriptions, which are documentary evidence for the pre-​Islamic and early Islamic periods—​research that is rapidly developing due to a myriad of new archaeological finds. 4.3 Adapting to the Arab Conquests Many existing communities were affected by the Arab conquests. Joanita Vroom examines by contrast what impact the Arab conquests had on the Byzantines themselves. Vroom uses two case studies, Athens in Greece and Butrint in southern Albania, to sketch an image of the trading contacts of and the daily life in Byzantine cities in the period of the Arab conquests and immediately after it. Using archaeological materials, she shows that the rise of the caliphate negatively impacted, from the seventh century onwards, the Byzantine economy. However, this impact was absorbed through the resilience of the Byzantine trading networks by an active search for new possibilities of exchange within the Byzantine state. These networks inevitably limited themselves to contacts within the Aegean Sea and the southern Adriatic but nonetheless show the adaptive possibilities of the empire. Moreover, Vroom’s study shows a rich city life in which many activities still took place, like the construction of houses and the placement of burials, roads, and wells. Finally, the contribution by Constanza Cordoni asks how post-​classical rabbinic and para-​rabbinic literature—​that is, a subset of post-​Islamic Jewish


van den Bent, van den Eijnde and Weststeijn

literature that emerged in the period of transition between Late Antiquity and the Middle Ages—​formulated a Jewish response to the rise of Islam and to Muslim rule in the land of Israel. The question is approached through three distinct themes. The article first deals with the scriptural Ishmael and his afterlife; it is shown that rabbinic literature tries to come to terms with the Muslim notion that Abraham’s first son, Ishmael, not his second son, Isaac, was the more important of the two and the forefather of the Arabs. Rabbinic literature has downplayed to the point of slandering Ishmael, in order to prove God’s favor of the Jews over the Muslims (who considered Ishmael their progenitor). Second, Ishmael’s descendants are treated as the last in a consecution of world powers, a theme that has long antecedents in rabbinic literature that far preceded the Arab conquests. In this light, it is interesting that they are often included in pre-​existing lists of world kingdoms—​with Greece, Rome, Babylon, and Media/​Persia—​but are generally not directly referred to as a kingdom. Rather, their rule is understood as a necessary precursor to a new order, the Messianic era, which is the third theme of Cordoni’s discussion. The Arab empire is thus understood as the final historical convolution that will lead to the ultimate delivery of the Jews. Perhaps unsurprisingly, the texts that deal with the latter theme take a more positive view of the Arab hegemony than the texts that view it in a longer line of previous world kingdoms. Viewed together, however, the rabbinic sources present an intriguing preoccupation with the integration of the actual historical context (Arab dominance) into scriptural tradition, offering a crucial insight into the mindset of a conquered nation dealing with the realities of Arab worldly power. 5

A Small Arab World: The Late Antique Arab Conquests from a Network Perspective

We noted above that cross-​Mediterranean ties had been weakening since the second century, a process that was briefly interrupted during the Justinian Renaissance, but continued unabated from the late sixth century onwards. Viewed from the vantage point of connectivity it is not the Arab conquests that caused the downfall of the late antique world: it was rather the waning of network ties across the Mediterranean that created the right environment for the Arab conquests to occur. In other words, the weakness of Mediterranean connectivity provided an opportunity for new (Arab) networks to rapidly emerge. Of course, once these new networks of connectivity became successful, they did contribute to a further weakening of pre-​existing ties—​see Vroom in this volume—​reaching an all-​time low in the seventh and especially eighth century

Late Antique Responses to the Arab Conquests


during the Arab conquests. This view takes fundamental issue with placing blame on the Arab conquests for the perceived disintegration of Romanitas. Network connectivity, and in particular the confrontation of old and new networks, is an important concern of this book. We are not just interested in the way Arab culture (an emerging Islam included) developed within a world consisting of a myriad of local but interconnected social networks, but also how previous networks reacted to the newly emerging ones.61 Taking our cue from current networking approaches in late antique and early medieval scholarship as outlined above, we argue that late antique local, regional, and supra-​ regional networks provided the quintessential historical context in which new cultural forms were able to flourish. To be sure, we do not suggest that late antique networks always coincided with early Arab networks. In fact, the contributions in this volume rather show how new networks built upon, coincided with, and sometimes replaced existing forms of connectivity in processes that were crucial to the emergence of Arabic culture. This book thus redirects attention from a top-​down, centralized view of the Arab world to the local dynamics of cultural adaptation. Like other network approaches, ours takes issue with earlier metanarratives that view the Mediterranean ecosystem as neatly distinct “cellular entities.”62 This “network” approach itself seems to have been a result of the fragmentation of political power and social coherence within the past decades and has had a profound influence on the views of social scientists and historians.63 Rather than studying the linear histories of “central” places, they have recalibrated their visors to the smaller “hubs” or “nodes” (in network theory jargon) that comprise such supposedly center-​oriented systems.64 Indeed, looking back, many of the entities that were previously believed to be monolithic, centralized systems, no longer seem so easy to understand as such.65 As others have shown, 61

62 63 64 65

We cannot do justice to the vast scholarship of Mediterranean connectivity, let alone seaborne connectivity in general. For Mediterranean connectivity, see for instance Horden and Purcell 2000; Abulafia 2011; Manning 2018; Horden and Purcell 2019. For global maritime connectivity, including the Mediterranean, see for instance Abulafia 2019; Strootman et al. 2020. Cf. Horden and Purcell 2000, 74. Hobsbawm 2013. E.g., Wasserman and Faust 1994; Castells 2010. As Crump 2015 has argued, Soviet policy was often dictated by the political exigencies caused by internal conflict between the members of the Warsaw Pact, rather than by “Moscow.” Similarly, scholarship on the Roman Empire has now largely moved away from “Rome” as the prime agent of change, in favor of local actors (Rome itself included) that operated in a multidirectional and multiscalar spatial framework. Cf. Pitts and Versluys 2015.


van den Bent, van den Eijnde and Weststeijn

cultural phenomena such as the different religions of the Mediterranean that have often been studied separately are understood better when viewed within a larger cultural context.66 Ancient religions in particular have recently been redefined within a multiscalar67 and network-​oriented framework.68 They were not self-​contained, monolithic phenomena, but were highly mobile and dynamic belief systems. As this characterization applies to monotheistic as well as the polytheistic religions of Antiquity,69 we should expect Islamization to have taken place along a similar multiscalar and multidirectional pathway. So how did this take place in practice? As Kevin van Bladel shows in this volume, local circumstances often determined the degree to which old and new networks interacted and thus impacted one another. In Syria, for example, the mass exodus of (presumably elite) Christians to Byzantine-​controlled territory favored a situation where Arab invaders settled among the remaining population, thus speeding up the processes of Arabicization and Islamization. In areas such as Egypt or Iraq, however, Arabs settled away from existing centers of habitation, leading to a more fragmented cultural landscape and a slower pace of adaptation. In this case, settlement patterns, which are highly indicative of local networks, were responsible for the varying pace with which Arabicization and Islamization took place. But can we take the network approach further to explain why Islam was so successful and why the Arabic language was eventually adopted nearly universally? Or why previous cultural and linguistic differences between the original tribes were obscured rather than magnified during the centuries of conquest? In this introduction we can do no more than make some very general suggestions. But however one may be inclined to approach these fundamental questions, it seems clear that any future explanation must in some shape or form take into consideration the implications of network studies. In particular, we suggest that one specific application of network theory, that of Small 66 67 68 69

Orlin 2016. Kindt 2012. Eidinow 2011; Eidinow 2015. Bonnet and Bricault 2016 have argued that where the movement of cult images allowed polytheistic religions to spread, in the case of Judaism and Christianity it was the word itself that caused new beliefs to diffuse. Cf. p. 250: “C’est donc par le texte et dans le texte que les Juifs vont assurer la ‘transhumance’ du culte de Yahvé.” As Paul’s mission in various Mediterranean harbor cities shows, it was, above all, a networked phenomenon. And of necessity, it had to be a glocalized phenomenon as well, with new ideas (the singular divinity of Christ) being wedded to existing local entities (the Unknown God). Cf. Acts 17.16–​34.

Late Antique Responses to the Arab Conquests


Worlds, opens a door to begin explaining these success stories. Small world theory holds that even in very large and complex societies, a rather limited number of intermediaries—​the well-​known “six degrees of separation”—​is needed for individuals that are widely separated by geography or social status to be connected.70 A “Small World” is predicated on the proposition that while people are connected mostly to their immediate neighbors (strong ties), they are also connected to other local networks through seemingly random connections (weak ties) between certain individuals. As Irad Malkin has argued for the period of Greek colonization (eighth to sixth century bce) these “weak ties” are—​rather paradoxically—​extremely important for the evident homogenization of Greek culture in its period of rapid expansion.71 According to Malkin, the “weak ties” (traders, mercenaries, artisans) connecting Greek communities living far and wide apart were responsible for communicating, very rapidly, new ideas and technologies throughout the Mediterranean and Black Seas. The very fact that these local Greek networks were often relatively isolated and surrounded by culturally distinct, if not downright hostile people, meant that they were extremely dependent on communications from overseas for the maintenance of their identity. Did such homogenization take place in the wake of the Arab expansion? Interestingly, precisely such a process is described by Peter Webb in his contribution to this volume, when he observes that since the ethnonym “Arab” is absent in pre-​conquest Arab poetry, it seems to have been the expansion itself that caused Arab cultural identity to emerge. If the example of Greek colonization is any indication, it was not the natural outcome of various individuals “mixing” together to create a blended whole. While a certain awareness of tribal affiliation among Arabs remained, homogenization over vast geographical expanses appears to have been the norm in the first centuries after Muhammad. Could it be that it was precisely these centrifugal forces that caused the widely dispersed Arab conquerors to seek common ground by means of the weak ties that kept their widely dispersed local networks connected? Did the dispersed Arab groups harken back to an invented and self-​reinforcing tradition of homogeneity? Perhaps herein lies a key to understanding the maintenance—​even reinforcement—​of cohesion in the wake of the Arab conquests: the emergence of what we may call a “Small Arab World.”

70 71

See Schnettler 2009 for an overview of some key publications. Malkin 2011.


van den Bent, van den Eijnde and Weststeijn


Abulafia, David. 1987. “Asia, Africa and the trade of Medieval Europe.” In Trade and Industry in the Middle Ages (The Cambridge Economic History of Europe, vol. 2), edited by Edward Miller, Cynthia Postan, and M.M. Postan, 402–​73, 905–​11. Cambridge: Cambridge University Press. Abulafia, David. 2011. The Great Sea: A Human History of the Mediterranean. Oxford: Oxford University Press. Abulafia, David. 2014. “Thalassocracies.” In A Companion to Mediterranean History, edited by Peregrine Horden and Sharon Kinoshita, 139–​53. Chichester: John Wiley & Sons. Abulafia, David. 2019. The Boundless Sea. New York: Oxford University Press. Al-​Jallad, Ahmad. 2015. An Outline of the Grammar of the Safaitic Inscriptions (SSLL 80). Leiden: Brill. Al-​Jallad, Ahmad, Younis Shdaifat, Zeyad al-​Salameen, and Rafe Harahsheh. 2017. “An Early Christian Arabic Graffito Mentioning ‘Yazīd the King’.” Arabian Archaeology and Epigraphy 28:1–​10. Bekker, Immanuel. 1838. Corpus Scriptorum Historiae Byzantinae. Bonn [Berlin]: Weber. Bonnet, Corinne, and Laurent Bricault. 2016. Quand les dieux voyagent: Cultes et mythes en mouvement dans l’espace méditerranéen antique. Geneve: Labor et Fides. Böwering, Gerhard. 2008. “Recent Research on the Construction of the Qurʾān.” In The Qurʾān in Its Historical Context, edited by Gabriel Said Reynolds, 70–​87. London: Routledge. Brown, Peter. 1974. “ ‘Mohammed and Charlemagne’ by Henri Pirenne.” Daedalus 103:25–​33. Brown, Peter. 1979. “Understanding Islam.” New York Review of Books, 22 February. Cameron, Averil. 2001. “Vandal and Byzantine Africa.” In Late Antiquity: Empire and Successors, AD 425–​600 (The Cambridge Ancient History, vol. 14), edited by Averil Cameron, Bryan Ward-​Perkins, and Michael Whitby, 552–​69. Cambridge: Cambridge University Press. Cameron, Averil. 2012. The Mediterranean World in Late Antiquity, 395–​700 AD, 2nd ed. Abingdon: Routledge. Castells, Manuel. 2010. The Rise of the Network Society, 2nd ed. Chichester: Wiley-​Blackwell. Citter, Carlo, Lidia Paroli, Christophe Pellecuer, and Jean-​Michel Péne. 1996. “Commerci nel Mediterraneo occidentale nell’alto medioevo.” In Città altomedievali nel mediterraneo occidentale, edited by Gian Pietro Brogiolo, 121–​42. Mantua: Padus. Crone, Patricia, and Michael Cook. 1977. Hagarism: The Making of the Islamic World. Cambridge: Cambridge University Press. Crump, Laurien. 2015. The Warsaw Pact Reconsidered. London: Routledge.

Late Antique Responses to the Arab Conquests


Cuypers, Michel. 2012. La composition du Coran: Naẓm al-​Qurʾân, rhétorique sémitique. Paris: Gabalda. Dark, Ken R. 2007. “Globalizing Late Antiquity: Models, Metaphors and the Realities of Long-​Distance Trade.” In Incipient Globalization? Long-​Distance Contacts in the Sixth Century, edited by Anthea Harris, 3–​14. Oxford: British Archaeological Reports. Dennett, Jr., Daniel C. 1948. “Pirenne and Muhammad.” Speculum 23:165–​90. De Souza, Philip. 1999. Piracy in the Graeco-​Roman World. Cambridge: Cambridge University Press. Déroche, François. 2013. Qurʾans of the Umayyads: A Preliminary Overview. Leiden: Brill. Donner, Fred M. 1998. Narratives of Islamic Origins: The Beginnings of Islamic Historical Writing. Princeton: The Darwin Press, Inc. Donner, Fred M. 2008. “Introduction.” In The Expansion of the Early Islamic State, edited by Fred M. Donner, xiii–​xlii. Aldershot: Ashgate Variorum. Donner, Fred M. 2010. Muhammad and the Believers: At the Origins of Islam. Cambridge: The Belknap Press of Harvard University Press. Doorninck, Jr. Frederick H. van. 1991. “The Medieval shipwreck at Serçe Limani: an early 11th century Fatimid-​Byzantine commercial voyage.” Graeco-​Arabica 4:45–​52. Drauschke, J. 2011. Zwischen Handel und Geschenk. Rahden: Verlag Marie Leidorf. Ehrenkreutz, Andrew S. 1972. “Another Orientalist’s Remarks concerning the Pirenne Thesis.” Journal of the Economic and Social History of the Orient 15:94–​104. Eidinow, Esther. 2011. “Networks and Narratives: A model for Ancient Greek Religion.” Kernos 24:9–​38. Eidinow, Esther. 2015. “Ancient Greek Religion: ‘Embedded’ … and Embodied.” In Communities and Networks in the Ancient Greek World, edited by Claire Taylor and Kostas Vlassopoulos, 54–​79. Oxford: Oxford University Press. Eijnde, Floris van den. 2019. “The “First Athenian Empire”? Athenian Overseas Interests in the Archaic Period.” In Empires of the Sea (Cultural Interactions in the Mediterranean 4), edited by Rolf Strootman, Floris van den Eijnde, and Roy van Wijk, 52–​80. Leiden: Brill. El-​Badawi, Emran. 2014. The Qurʾān and the Aramaic Gospel Traditions. London: Routledge. Ferluga, J. 1987. “Navigation et commerce dans l’Adriatique aux VIIe et VIIIe siècle.” Byzantinische Forschungen 12: 39–​51. Fowden, Garth. 2015. Before and After Muhammad. Princeton: Princeton University Press. Geiger, Abraham. 1833. Was hat Mohammed aus dem Judenthume aufgenommen. Bonn: Baden Verlag. Gibb, Hamilton Alexander Rosskeen. 1955. “The Fiscal Rescript of ʿUmar II.” Arabica 2:1–​16.


van den Bent, van den Eijnde and Weststeijn

Gibb, Hamilton Alexander Rosskeen. 1958. “Arab-​Byzantine Relations under the Umayyad Caliphate.” Dumbarton Oaks Papers 12:219–​33. Gibbon, Edward. 1776–​88. The History of the Decline and Fall of the Roman Empire (6 vols.). London. Griffith, Sidney. 2008. “Christian Lore and the Arabic Qurʾan: ‘The Companions of the Cave’ in Sūrat al-​Kahf and Syriac Christian Tradition.” In The Qurʾān in Its Historical Context, edited by Gabriel Said Reynolds, 109–​37. London: Routledge. Haldon, John F. 1990. Byzantium in the Seventh Century: The Transformation of a Culture. Cambridge: Cambridge University Press. Heebøll-​Holm, Thomas K. 2013. Ports, Piracy, and Maritime War. Piracy in the English Channel and the Atlantic, c. 1280—​c. 1330. Leiden: Brill. Heebøll-​Holm, Thomas K. 2019. “Medieval Denmark as a Maritime Empire.” In Empires of the Sea (Cultural Interactions in the Mediterranean 4), edited by Rolf Strootman, Floris van den Eijnde, and Roy van Wijk, 194–​218. Leiden: Brill. Hendy, Michael Frank 1988. “From Public to Private: The Western Barbarian Coinages as a Mirror of the Disintegration of Late Roman State Structures.” Viator 19:29–​78. Hilali, Asma. 2014. “Was the Ṣanʿāʾ Qurʾān Palimpsest a Work in Progress?” In The Yemeni Manuscript Tradition, edited by David Hollenberg, Christoph Rauch, and Sabine Schmidtke, 12–​27. Leiden: Brill. Hobsbawm, Eric. 2013. Fractured Times. London: Little, Brown. Horden, Peregrine, and Nicholas Purcell. 2000. The Corrupting Sea: A Study of Mediterranean History. Oxford: Blackwell. Horden, Peregrine, and Nicholas Purcell. 2019. The Boundless Sea: Writing Mediterranean History. Abingdon and New York: Routledge. Hoyland, Robert G. 1997. Seeing Islam as Others Saw It: A Survey and Evaluation of Christian, Jewish, and Zoroastrian Writings on Early Islam. Princeton: Darwin Press. Hoyland, Robert G. 2014. In God’s Path. The Arab Conquests and the Creation of an Islamic Empire. Oxford: Oxford University Press. Ibn Khurdādhbih, Abū l-​Qāsim ʿUbaydallāh ibn ʿAbdallāh. 1889. Kitāb al-​Masālik wa-​l-​ Mamālik. Edited by M.J. de Goeje. Leiden: Brill. Kindt, Julia. 2012. Rethinking Greek Religion. Cambridge: Cambridge University Press. Kreutz, Barbara M. 1976. “Ships, Shipping, and the Implications of Change in the Early Medieval Mediterranean.” Viator 7:79–​110. Kristof, Nicholas D. 2004. “Martyrs, Virgins and Grapes.” New York Times, 4 August, A17. Levy-​Rubin, Milka. 2011. Non-​Muslims in the Early Islamic Empire: From Surrender to Coexistence. Cambridge: Cambridge University Press. Lopez, Robert Sabatino, and Irving Woodworth Raymond. 1955. Medieval Trade in the Mediterranean World: Illustrative Documents Translated with Introductions and Notes. New York: Columbia University Press.

Late Antique Responses to the Arab Conquests


Lüling, Günter. 1974. Über den Ur-​Qurʾān: Ansätze zur Rekonstruktion vorislamischer christlicher Strophenlieder im Qurʾān. Erlangen: Verlagsbuchhandlung H. Lüling. Lüling, Günter. 2003. A Challenge to Islam for Reformation: The Rediscovery and Reliable Reconstruction of a Comprehensive pre-​Islamic Christian Hymnal Hidden in the Koran under Earliest Islamic Reinterpretations. Delhi: Motilal Banarsidass Publishers. Luxenberg, Christoph. 2000. Die syro-​aramäische Lesart des Koran: Ein Beitrag zur Entschlüsselung der Koransprache. Berlin: Das Arabisch Buch. Malamut, Elisabeth. 1993. Sur la route des saints byzantins. Paris: cnrs Éditions. Malkin, Irad. 2011. A Small Greek World: Networks in the Ancient Mediterranean. Oxford: Oxford University Press. Manning, Joseph G. 2018. The Open Sea. Princeton and Oxford: Princeton University Press. Meehan, Denis, ed. 1958. Adamnan, De locis sanctis (On the Holy Places). (Scriptores Latini Hiberniae 3). Dublin: Dublin Institute for Advanced Studies. Nehmé, Laïla. 2017. “New Dated Inscriptions (Nabataean and Pre-​Islamic Arabic) from a Site near Al-​Jawf, Ancient Dūmah, Saudi Arabia.” Arabian Epigraphic Notes 3:121–​64. Neuwirth, Angelika, Nicolai Sinai, and Michael Marx, eds. 2010. The Qurʾān in Context: Historical and Literary Investigations into the Qurʾānic Milieu. Leiden: Brill. Neuwirth, Angelika. 2014. Scripture, Poetry and the Making of a Community: Reading the Qur’an as a Literary Text. Oxford: Oxford University Press. Nevo, Yehuda D., and Judith Koren. 2003. Crossroads to Islam: The Origins of the Arab Religion and the Arab State. Amherst: Prometheus. Noth, Albrecht. 1968. “Iṣfahān-​Nihāwand: Eine quellenkritische Studie zur frühislamischen Historiographie.” Zeitschrift der Deutschen Morgenländischen Gesellschaft 118:274–​96. Ohlig, Karl-​Heinz, and Gerd-​R. Puin, eds. 2005. Die dunklen Anfänge: Neue Forschungen zur Entstehung und frühen Geschichte des Islam. Berlin: Verlag Hans Schiler. Orlin, Eric M. 2016. The Routledge Encyclopedia of Ancient Mediterranean Religions. London: Routledge. Parker, Anthony John. 1992. Ancient Shipwrecks of the Mediterranean and the Roman Provinces. Oxford: British Archaeological Association. Penn, Michael Philip. 2015. Envisioning Islam: Syriac Christians and the Early Muslim World. Philadelphia: University of Pennsylvania Press. Pertz, Georg Heinrich. 1872. Diplomata regum Francorum e stirpe Merowingica: Monumenta Germaniae Historica, Diplomatum Imperii 1. Hannover: Hahn. Pirenne, Henri. 2014 [1925]. Medieval Cities: Their Origins and the Revival of Trade. Princeton: Princeton University Press. Pirenne, Henri. 1937. Mahomet et Charlemagne. Paris, Bruxelles: F. Alcan, Nouvelle société d’éditions.


van den Bent, van den Eijnde and Weststeijn

Pirenne, Henri. 1939. Mohammed and Charlemagne. London: G. Allen & Unwin. Pitts, Michael, and Miguel John Versluys. 2015. Globalisation and the Roman World: World History, Connectivity and Material Culture. New York: Cambridge University Press. Reynolds, Gabriel Said, ed. 2008. The Qurʾān in Its Historical Context. London: Routledge. Reynolds, Gabriel Said. 2008. “Introduction: Qurʾānic Studies and Its Controversies.” In The Qurʾān in Its Historical Context, edited by Gabriel Said Reynolds, 1–​25. London: Routledge. Reynolds, Gabriel Said. 2010. The Qurʾān and Its Biblical Subtext. Abingdon: Routledge. Robinson, Chase F. 2000. Empires and Elites after the Muslim Conquest: The Transformation of Northern Mesopotamia. Cambridge: Cambridge University Press. Robinson, Chase F. 2004. “The Conquest of Khūzistān: A Historiographical Reassessment.” Bulletin of the School of Oriental and African Studies 67:14–​39. Sadeghi, Behnam, and Mohsen Goudarzi. 2012. “Ṣanʿāʾ 1 and the Origins of the Qurʾān.” Der Islam 87:1–​129. Said, Edward W. 1978. Orientalism. New York: Pantheon Books. Saleh, Walid A. 2010. “The Etymological Fallacy and Qurʾanic Studies: Muhammad, Paradise, and Late Antiquity.” In The Qurʾān in Context: Historical and Literary Investigations into the Qurʾānic Milieu, edited by Angelika Neuwirth, Nicolai Sinai, and Michael Marx, 649–​98. Leiden: Brill. Schnettler, Sebastian. 2009. “A Structured Overview of 50 Years of Small-​World Research.” Social Networks 31.3:165–​78. Sinai, Nicolai, and Angelika Neuwirth. 2010. “Introduction.” In The Qurʾān in Context: Historical and Literary Investigations into the Qurʾānic Milieu, edited by Angelika Neuwirth, Nicolai Sinai, and Michael Marx, 1–​24. Leiden: Brill. Sijpesteijn, Petra M. 2013. Shaping a Muslim State: The World of a Mid-​Eighth-​Century Egyptian Official. Oxford: Oxford University Press. Sijpesteijn, Petra M. 2017. “The Rise and Fall of Empires in the Islamic Mediterranean (600–​1600 CE): Political Change, the Economy and Material Culture.” In The Routledge Handbook of Archaeology and Globalization, edited by T. Hodos, 652–​68. London: Routledge. Stewart, Devin J. 1990. “Sajʿ in the Qurʾan: Prosody and Structure.” Journal of Arabic Literature 21:101–​39. Stewart, Devin J. 2011. “The Mysterious Letters and Other Formal Features of the Qurʾan in Light of Greek and Babylonian Oracular Texts.” In New Perspectives on the Qurʾān: The Qurʾān in its Historical Context 2, edited by Gabriel Said Reynolds, 321–​46. London: Routledge. Stewart, Devin J. 2017. “Reflections on the State of the Art in Western Qurʾanic Studies.” In Islam and its Past: Jahiliyya, Late Antiquity, and the Qurʾan, edited by Carol Bakhos and Michael Cook, 4–​68. Oxford: Oxford University Press.

Late Antique Responses to the Arab Conquests


Strootman, Rolf, Floris van den Eijnde, and Roy van Wijk, eds. 2020. Empires of the Sea (Cultural Interactions in the Mediterranean 4). Leiden: Brill. Talbot, Charles Hugh. 1954. The Anglo-​Saxon Missionaries in Germany. London: Sheed & Ward. Udovitch, Abraham L. 1978. “Time, the Sea and Society: Duration of Commercial Voyages on the Southern Shores of the Mediterranean during the High Middle Ages.” ss 25:503–​46. Wansbrough, John. 1977. Quranic Studies: Sources and Methods of Scriptural Interpretation. Oxford: Oxford University Press. Wansbrough, John. 1978. The Sectarian Milieu: Content and Composition of Islamic Salvation History. Oxford: Oxford University Press. Wasserman, Stanley, and Katherine Faust. 1994. Social Network Analysis: Methods and Applications. Cambridge: Cambridge University Press. Webb, Peter. 2016. Imagining the Arabs: Arab Identity and the Rise of Islam. Edinburgh: Edinburgh University Press. Wickham, Chris. 1981. Early Medieval Italy: Central Power and Local Society 400–​1000. London: Macmillan. Wild, Stefan. 2010. “Lost in Philology? The Virgins of Paradise and the Luxenberg Hypothesis.” In The Qurʾān in Context: Historical and Literary Investigations into the Qurʾānic Milieu, edited by Angelika Neuwirth, Nicolai Sinai, and Michael Marx, 625–​ 47. Leiden: Brill.

­c hapter 2

The Qurʾānic Rūm: A Late Antique Perspective Clare Wilde Rome and Arabia were not strangers to one another in Late Antiquity.1 Distant late antique Christians (including Roman emperors)2 knew about events in Arabia, even those in the southern part of the Peninsula, such as the early sixth-​century ce martyrdom of Christians in Najrān.3 Similarly, both the Qurʾan and later Islamic tradition include numerous events before, during, and after the life of the Prophet Muhammad (570–​632 ce) that are connected to the Roman Empire.4 For example, the story of the “People of the Cave” (Q 18:9–​26) is commonly connected to Christian stories of the “Sleepers of Ephesus.” According to Christian tradition, seven youths had escaped persecution during the reign of Decius (r. 249–​51 ce) by hiding in a cave in Ephesus, in present-​day Turkey. God then placed them into a miraculous sleep for two hundred years. Their “awakening” became part of debates about the reality of the bodily resurrection, a disputed topic in the fifth century ce. Although known by different names, and commemorated on different days, various Christian denominations recognize the seven “Sleepers of Ephesus” as saints.5 This association of the Qurʾānic “Companions of the Cave” with the Christian “Sleepers of Ephesus” requires some familiarity with late antique narratives. For example, the Qurʾānic passage does not include identifying names of places or people. Rather, the Qurʾan alludes to details of a narrative as if its audience knows the story (it also disputes some details, such as the number

1 See the comprehensive works of Irfan Shahid on the subject, beginning with Shahid 1984, as well as the more recent studies of Kaegi 1995; Hoyland 2009; Fisher 2011. These sources discuss Arabs as border guards between Rome and Persia, also found in Casey 1996 and Fowden 1999. 2 Van Rompay 2005. 3 See the discussion and bibliography in Shahid 2001–​6. See also Shahid 1971. Shahid 1979 provides an extensive discussion of Najrān and other south Arabian Christian sites. 4 See Hoyland 2012 for a comprehensive presentation of Islam as a late antique religion. 5 For an excellent overview of the Qurʾānic account and late antique (especially Syriac) literature, see Griffith 2007.

© Clare Wilde, 2022 | DOI:10.1163/9789004500648_003

The QurʾĀnic Rūm


of individuals in the cave and how long they slept).6 Elsewhere, however, the Qurʾān does include proper names of people or places. The beginning of Sura 30 is one such example: … The Rūm have been defeated in a nearby land, but after their defeat they will prevail within a few years. To God belongs the command before and after. On that day the believers will rejoice in the victory of God. He gives victory to whom He wills … Q 30:2 contains the only Qurʾānic reference to “al-​Rūm,” the Arabic term for Romans, which, by this time, signified the Christian Eastern Roman (or Byzantine) Empire. This hapax legomenon also lends itself to the name of this entire sura: the thirtieth chapter of the Qurʾān is entitled “Sūrat al-​Rūm” (The Romans). But, as with the “Companions of the Cave,” the Qurʾān does not provide specific details about the identity of the Rūm or the place or time of their defeat (or victory). Later Islamic tradition, aware of the chronological overlap of the Persian-​Roman war of 602–​28 ce and Muhammad’s reception of the Qurʾān (610–​32 ce), would come to understand the opening verses of Q 30 as an allusion to the Persian conquest of Jerusalem in 614 ce and a prediction of the Roman reconquest in 628 ce. In keeping with this interpretation is the understanding that an individual who recites Q 30 is compensated for each verse in the same manner as someone who visits Jerusalem.7 This reading resonates with Christian narratives about what would prove to be the last of the Sasanian-​Byzantine wars. After the Sasanian rise to power in the third century ce, periodic conflicts between Rome and Persia occurred, often at the eastern border of the Roman Empire. Phocas’s murder of Maurice in 602 ce led to a succession crisis in the Eastern Roman Empire and opened the doors for what initially appeared to be a Persian victory: by 620 ce, Palestine, Egypt, and parts of Asia Minor were under Persian control. The course of the war changed in 622 ce (the year of Muhammad’s emigration from Mecca to Medina): the Eastern Roman emperor, Heraclius, had been building his army for a few years, leading to a number of significant victories in the Caucasus. A coup eventually led to the death of the discredited Persian leader Khosrow and peace was negotiated. Although the Romans definitively defeated the Sasanians, both empires had so depleted their resources that, 6 For two excellent overviews of Qurʾānic language, see Gilliot and Larcher 2001–​6; Neuwirth 2001–​6. 7 Firūzābādī, cited by Busse 2001–​6, 6.

34 Wilde within a few years, Arab armies would conquer not only former Sasanian, but also former Byzantine, strongholds: e.g., Damascus in 634 ce, Jerusalem and Ctesiphon in 637 ce, and Egypt in 639–​42 ce. The Persian conquest of Jerusalem in 614 ce held great symbolic importance for Christian chroniclers,8 particularly the capture and transfer to Ctesiphon of relics relating to Christ’s crucifixion (especially the True Cross). As expected, Christian narrative emphasized Heraclius’s triumphal return to Jerusalem—​ with the True Cross. Much of the post-​Qurʾānic Islamic tradition developed in areas that had been part of either the Byzantine or Sasanian empire. In many places, Christians remained a demographic majority for much of the early Islamic centuries, often continuing to hold relatively high administrative posts even under Muslim rulers.9 Therefore, it is not surprising to find that Islamic tradition echoes apocalyptic and other religious significations ascribed by Christian tradition to the Persian conquest and Roman reconquest of Jerusalem in the early seventh century ce (such as the strong connection between the opening verses of Q 30 and Jerusalem). Western scholarship, likewise the heirs of narratives in which Jerusalem figures prominently, would also accept, with few questions, the classical Islamic connection of Q 30 to Jerusalem. However, these interpretations developed at a chronological and (often) geographical remove from Mecca and Medina in 610–​32 ce, the context of the initial revelation in Islamic tradition. As discussed below, other interpretations of “a nearby land” and the nature of the victory and defeat of the Rūm emerge if the opening passage of Q 30 is read in the light of other Qurʾānic passages, as well as within larger late antique history. Moreover, this approach offers a new interpretation of the identity of the “believers who will rejoice in the victory of God.” 1

Reading the Qurʾān: Some Introductory Remarks

Western scholarship and Islamic tradition have commented on some of the difficulties in understanding the Qurʾānic text: the ambiguities of the earliest manuscripts; the allusive nature of the text; its non-​chronological arrangement; and the debated relationship between the Qurʾān and other Islamic literary sources, such as the prophetic biography and the varied and varying glosses of later commentators that contribute to the multivalent reading of many passages. Islamic tradition developed a genre of literature, the so-​called 8 Recent scholarship has questioned the historical accuracy of some aspects of these narratives. See Avni 2010. 9 For a history of Christians under Islamic rule, see the masterful recent work of Griffith 2012.

The QurʾĀnic Rūm


asbāb al-​nuzūl, “occasions of revelation,” to explain the Qurʾānic verses in the context of Muhammad’s life.10 It is difficult, however, to separate traditions that reflect the later needs of the Muslim community from those that may better reflect the situation in Arabia during Muhammad’s own time.11 Later exegetes, removed chronologically (and geographically) from the Qurʾān’s initial audience, would often interpret Qurʾānic passages through the lenses of (later) Islamic tradition and with respect to the needs of their own times. Therefore, the exegetes often did not uncover the “original” sense—​or reading—​of a given passage.12 While later exegetes may preserve echoes of pre-​Islamic, late antique controversies that are reflected in the Qurʾān, the allusive and sometimes polemical nature of Qurʾānic discourse frequently obscures the original late antique narrative or controversy. Jerusalem exemplifies the interplay of the Qurʾān, late antique events, and later narratives. Although never named in the Qurʾān, Jerusalem figures in common interpretations of a number of Qurʾānic eschatological passages, as well as ones relating to the life of this world, in references to the Children of Israel (before Muhammad’s lifetime) and to events associated with Muhammad himself.13 We understand that Jerusalem is referenced in various Qurʾānic passages and/​or in Islamic tradition for reasons ranging from the deliberate appropriation of Judeo-​Christian traditions to a natural assumption of such traditions. Islam is as much an heir to the legacy of the prophets recorded in the Bible as Judaism and Christianity. Another example of Islamic tradition’s appropriation of late antique narratives is the common gloss for those at whom God is angry (Jews) and those who are astray (Christians; Q 1:6–​7). Although now a common interpretation of Q 1:6–​7, it is significant that the Qurʾānic terms for “Jew” (Ar. yahūd) and “Christian” (Ar. naṣārā) do not appear in this sura. Moreover, Islamic tradition is not unanimous in its interpretation of this passage.14 However, although missing from many contemporary (and classical) discussions of this passage, the polemic of Jews as those at whom God is angry has a rich pre-​Islamic, late 10 11

12 13 14

On the difficulties in identifying historic moments in the Qurʾān, see Neuwirth 2003a. Key examples include the differences between Sunni and Shia understandings of the integrity of Qurʾānic text, and its interpretations, as well as the regard each group gives to various key figures in early Islamic history (such as ʿĀʾisha or the first three caliphs). See especially Modarressi 1993. Hoffman 1998 provides excellent examples of how exegetes would interpret verses in light of their contemporary societal conventions, arguing that exegesis sheds more light on the societies in which the exegetes lived, rather than the conventions in the Qurʾānic milieu. See the discussions in Van Ess 1999 and Neuwirth 2003b. See the extensive discussion (in English) of the interpretations of this passage in Nasr 2015.

36 Wilde antique, history.15 Similarly, the Qurʾānic portrayal and critique of Christian beliefs and internecine disputes has deep resonances with late antique controversies. As will be elaborated below, while very critical of two central Christian doctrines (the Incarnation and the Trinity), both the Qurʾān16 and later Islamic tradition17 offer varied evaluations of Christians themselves.18 To Muslims, Christians19 can be seen either as fellow believers—​monotheists and “People of the Book”—​or as unbelieving polytheists.20 Both the Qurʾān and the early Muslim community had mixed judgments of, and experience with, Christians and Christianity. For example, according to the prophetic biography, the Christian leader of Ethiopia (the Negus) gave sanctuary to some of Muhammad’s followers, and a treaty was made between the first Muslims and the Christians of Najrān. In the Qurʾān, Christians are described as the “closest in love to the believers” (Q 5:82), but the believers are also exhorted not to take Christians as friends, as they are allies with Jews (Q 5:51). And, at a doctrinal level, Christian belief in the Trinity and also the divine Sonship of Christ did indeed incur Muslim (and Qurʾānic) criticism, including the charge of polytheism.21 However, because monotheistic communities commonly accused other groups of “polytheism,” such rhetoric must be read carefully, whether found in the Qurʾān or in later exegetical glosses. Despite these challenges of interpretation, the Qurʾān’s engagement with contemporary events, as well as its seeming awareness of its own revelatory claims,22 has long been observed. Using the Qurʾān’s awareness of its milieu as a starting point, the following will read the Qurʾān as a text emerging from—​ and reflecting—​late antique, rather than later Islamic, traditions: a rereading of the opening verses of Q 30 will question whether they might refer to events other than the Persian conquest and Roman reconquest of Jerusalem in the early seventh century ce. Although it is beyond the scope of this paper

15 16 17 18 19 20 21 22

Cf. e.g., Cohen 1983. Peters 1997. McAuliffe 1991. For a comprehensive overview of classical positions, especially in the first Islamic centuries, see Gilliot 2009. For an excellent review of the Qurʾānic discussions of Christians, see Griffith 2011. On the difficulties of distinguishing historical fact from religious polemic, especially in accusations of idolatry, see Hawting 1997. As seen, for example, in the debate between Theodore Abū Qurrah and a number of Muslim notables in the caliphate of al-​Maʾmūn. See the translation of Nasry 2008. Cf. also Thomas 1992. On this concept see, e.g., the seminal work of Madigan 2001, the edited collection of Wild 2006 and the recent monograph of Boisliveau 2014.

The QurʾĀnic Rūm


to explore how scholars date the collection and codification of the Qurʾānic text,23 it should be noted that the following discussion accepts the traditional date of the prophetic mission (610–​32 ce) and the dissemination of the basic Qurʾānic text within a generation after Muhammad’s death. Analyzing the available literary record to reconstruct the events of the late antique and early Islamic periods is not a simple task. Extant manuscripts, Christian or Muslim, often postdate the events they purport to report, often by centuries. Further, the literature often recounted certain events to serve particular theological or other interpretations, rather than to provide a dispassionate description of events.24 Similarly, the “original” sense of Qurʾānic passages is not always evident. In addition to these challenges, the Arabic script used in the early manuscripts is ambiguous. The Arabic alphabet includes twenty-​eight consonants, to which may be added vowels (a-​i-​u). The consonants, however, are written in a limited number of basic forms. These basic forms (hooks, lines, etc.) become distinct consonants by the presence (or absence) of a dot (or dots) above (or below) the letters. For example, a “b” is represented by a “hook” with a single dot below it; while a “t” is represented by a “hook” with two dots above it. The earliest Qurʾānic manuscripts, however, are written only with the basic Arabic script (the rasm)—​that is, the script does not utilize dots to distinguish the various consonants or the various vowel markers. This fact presents scholars with an interpretive conundrum: could the early Qurʾānic manuscripts have been incomprehensible (or allow a nearly infinite range of possible interpretations)? Or, does the very minimal Arabic script used in early Qurʾānic manuscripts indicate that they were used merely as memory aids, indicating that the Qurʾān was, initially, a memorized (and recited) text?25 Because exegetes do not agree upon the reading or meaning of every word in the opening passage of Q 30, the following discussion seeks another way to hear the verse—​one that might have resonated with the first audience, rather than the apologetic (or polemic) agendas of later exegetes. Through a close reading of late antique history in and around the Arabian Peninsula, I revisit the traditional reading of these passages as a reference to Roman/​Persian battles in Syria. Roman involvement occurred in a number of areas in and around the Arabian Peninsula in Late Antiquity. Therefore, I explore whether the

23 24 25

See Cook 2000. For critiques of the dating of post-​Qurʾānic, early Islamic, literature, see the discussions present in, e.g., Wansbrough 1977; Motzki 2001; Berg 2013. As, for example, with Eusebius’s Life of Constantine. See Jones 2001–​6 for further elaboration of these theories.

38 Wilde Roman victory/​defeat of Q 3026 might refer to events other than the early seventh century ce Persian/​Roman battles for Jerusalem, the traditional understanding of this passage.27 2

Traditional Tafsīr of Q 30:2–​5

Q 30:2–​5 is commonly read as (30:2) The Rūm have been defeated (3) in a nearby land, but after their defeat they will prevail (4) within a few years. To God belongs the command before and after. On that day the believers will rejoice (5) in the victory of God. He gives victory to whom He wills … The traditionally accepted reading of the two verbs in 30:2–​3 is a passive voice (ghulibat) followed by the active voice (sa-​yaghlibūna): the Romans have been defeated, but they will be victorious. This phrase is usually understood as a reference to the Persian victories in Syria that followed Heraclius’s ascension to power in 610 ce (after the turmoil of the depositions of both Maurice and Phocas in the first decade of the 600s) and his later victories over the Persians in the 620s ce, culminating in the return of the True Cross to Jerusalem in 630 ce. Although this understanding came to be common, Islamic tradition is far from unanimous in its reading of Q 30:2–​5.28 While the traditional reading is that the Romans were defeated, but will (eventually) be the victors, the events and the dates intended by the Qurʾānic passage as well as the reason for the “rejoicing” of the believers have multiple possible interpretations. In early glosses, the victory of the Romans over the Persians occurs on the same day as the Battle of Badr (in 624 ce), when the (monotheistic) Muslims defeated the (polytheistic) Meccans. Sometimes, however, the Roman victory is understood 26 27


Depending on the vocalization of the Arabic root letters, the sequence of the Roman victory and defeat in the Qurʾānic passage has been understood in a variety of ways. See the discussion in El-​Cheikh 2001–​6. Little is known for certain about the real impact of the Persian incursions into Syria. Although the Persian capture of Jerusalem (and Heraclius’s triumphal return) figures prominently in Christian literature, these accounts are not necessarily historically accurate. The archaeological record—​although imperfect—​remains the preferred source for this period. See e.g., Foss 2003. For a detailed discussion of one hypothesis, see Shahid 1972. See the thorough discussion found in El-​Cheikh 1998, summarized in El-​Cheikh 2001–​6. The details here follow her overview of the classical literature.

The QurʾĀnic Rūm


as coinciding with the Treaty of Ḥudaybiyya (in 628 ce). Early exegetical works, as well as works discussing the “occasions of revelation,” stress the link between the victories of two monotheistic Peoples of the Book, Christian Byzantines and Muslims, over their respective—​and polytheistic—​enemies (Persians and Meccans). In these early readings, the “believers” who rejoice are generally glossed as the early Muslim community, but the reason for their rejoicing is the victory of fellow monotheists (Christian Byzantines) over unbelievers. Later commentators, while agreeing that the passage alludes to a Roman victory over the Persians, would question why the “believers” (Muslims) would rejoice over the victory of the (unbelieving) Rūm. In these readings, the believers rejoice at the victory of one unbelieving group over another because the constant warring between these groups will eventually enable a Muslim victory over, for example, the Romans. The shift in the understanding of al-​Rūm from believers to unbelievers likely reflects polemics stemming from the geopolitical realities of the commentators’ days, rather than any shift in the Muslim theological estimation of Christian belief (or unbelief). Although the Qurʾān recognizes Christians as “People of the Book,” it is also highly critical of Christian doctrines that could easily be construed as shirk (associating something other than God with God). For example, Q 112 asserts that God neither begets nor is begotten. The Qurʾān does not deny that Mary could have given birth to Jesus without a human father but asserts that the likeness of Jesus is Adam: God need only say “Be” and something becomes (Q 3:59). Q 4:171 exhorts (Christians) not to say “Three.” Although Christians had not changed their doctrinal beliefs between the seventh and tenth or eleventh centuries ce, geopolitical realities differed over time. Despite the initial rapid wave of Arab, Muslim conquests in the first century after the Prophet’s death, the Byzantine Empire persisted as a military and geopolitical rival to the caliphate.29 Then, with the initiation of the Crusades in 1095 ce came the presence of foreign Christians in the Levant.30 The increased exegetical emphasis on the Qurʾānic Rūm as unbelievers31 coincides with an

29 30 31

For an example of a polemical critique of Christianity prior to the Crusades and the Mongol destruction of Baghdad, while the Byzantines still maintained their borders with the caliphate, see Reynolds 2004. For one example of the (d)evolution of Muslim-​Christian perceptions in the Crusades, see Ebied and Thomas 2005; Cucarella 2010. The refusal to include the Byzantines as “believers” also reflects a conceptual trend that, although also present in earlier periods, emerges with particular clarity after the Mongol destruction of Baghdad in 1258 ce. See, for example, Ibn Taymiyya and Ibn Kathīr’s use of Isrāʾīliyyāt, discussed in Tottoli 1999, 201–​6. For other aspects of the evolving image of

40 Wilde increased emphasis on the importance of Jerusalem, mirroring the hopes for a final Muslim victory over the Crusaders. In addition to this variety of interpretations, the basic Arabic text of this passage (the rasm, discussed above) is also vocalized in multiple ways.32 Islamic tradition has read Q 30:2–​5 as containing an active verb (ghalabat) followed by one in the passive voice (sa-​yughlabūna): the Romans have been victorious, but will be defeated.33 This reading (as with others) has been variously emphasized, often in response to the particular circumstances of the exegetes. The ghalabat/​sa-​yughlabūna reading, considered the principal variant, appears particularly in response to the Crusades, emphasizing the eventual defeat of the Christians—​at the hands of Muslims. The passage became part of more abstract theological discussions. For example, the precise date and exact number of years intended by the “few years” of Q 30:4 appear in discussions of Muhammad’s ability to predict future events. The passage also figured in the development of the understanding of the inimitability (iʿjāz) of the Qurʾān, one aspect of which is its knowledge of future events. Al-​Bāqillānī (d. 403 ah/​1 013 ce) cites the opening verses of Q 30 as a major example of this aspect. These glosses reflect later geopolitical realities, theological developments, or apologetic agendas. They do not, however, shed much light on what the Qurʾān’s first audience might have understood when these verses were recited. The following section explores possible identifications of “al-​Rūm” and adnā al-​arḍ (“the nearest land”) that might have resonated with inhabitants of Arabia in Late Antiquity. 3

The Qurʾānic al-​Rūm: Some Late Antique Perspectives

In its references to Jesus and Mary, as well as to Christian factions,34 the Qurʾān appears familiar with the Christian Christological controversies debated at

32 33


Christians in Islamic thought, see Safran 2003. See also the translation of Ibn Taymiyya’s refutation of/​response to Christianity in Michel 1984. On the establishment of the consonantal text, see e.g., Sinai 2014. On the variant readings of the Qurʾān, see Leemhuis 2001–​6. See El-​Cheikh 2001–​6 for a comprehensive discussion of various other interpretations, as well as other possible readings, such al-​Qurṭubī (d. 671 ah/​1273 ce), who argues for two active verbs (ascribing two victories to the Byzantines: their victory over the Persians that coincided with the Muslim victory at Badr and a later Byzantine victory). Q 19:37, when read in conjunction with the preceding passages, is commonly interpreted as a reference to (Christian) Christological disputes: “the factions differed among themselves.”

The QurʾĀnic Rūm


the Council of Chalcedon (451 ce). As with other early ecumenical councils, Chalcedon was called by the emperor (Marcian), in response to Christian disputes. Whereas the Council of Nicea (325 ce) set out to define the relationship of God the Son to God the Father, at Chalcedon the primary theological dispute centered on how properly to understand Christ as both human and divine. The heated disputes over these doctrinal matters attracted imperial attention. Needless to say, the definition of the “hypostatic union” arrived at by Chalcedon—​that the divine and human natures of Christ, though retaining their distinct properties, were united (but not commingled) in a single hypostasis in the person of Christ—​did not meet with immediate, or universal, approval (particularly as political, ethnic, and linguistic divisions factored into the theological divides).35 For over a century after Chalcedon, emperors would attempt—​unsuccessfully—​to appease the Christian factions.36 The disputes over Chalcedon’s definition of the hypostatic union (of Christ’s human and divine natures) have received much scholarly attention, including in the field of Qurʾānic studies.37 In a number of passages, the Qurʾān presumes its audience’s familiarity with the Christian theological disputes—​particularly regarding Jesus and Mary—​that the Council of Chalcedon also addressed. For example, among the heretics and heresies the Council set out to correct were those trying to ruin the proclamation of the truth, and through their private heresies they have spawned novel formulas—​some by daring to corrupt the mystery of the Lord’s economy on our behalf, and refusing to apply the word “God-​bearer” to the Virgin.38 Although Qurʾānic discourse defended Mary’s honor (Q 66:12), she is never called “Mother of Jesus,” let alone “Mother of God.” Nevertheless, the Qurʾān proffers estimations of Mary’s virtue: she may suffer the pangs of childbirth (cf. Q 19:23), but she preserves her chastity (cf. Q 19:20 and

35 36

37 38

For a classic articulation of this thesis, see Jones 1959. Examples of imperial compromises that attempted to unify eastern Christians include the so-​called “Three Chapters Controversy” of Justinian (543 ce) and Zeno’s Henoticon (482 ce). The Egyptian church especially suffered, with over a century of religious strife affecting the occupant of the See of Mark and the local churches. Cf. Butcher 1897; Wigram 1923; Hardy 1952; Sellers 1953; Attwater 1961; Wakin 1963; Atiya 1968. For an overview of some of these perspectives, see Reynolds 2014. Quoted passage from the dogmatic definition of Chalcedon at http://​www.papalencyclicals .net/​Councils/​ecum04.htm. Translation on this website taken from Tanner and Alberigo 1990.

42 Wilde 21:91);39 Mary is also purified, chosen above the women of the world, and among the obedient (Q 3:42–​43). Similarly, the Qurʾān consistently calls Jesus “the Messiah” and “Son of Mary,” but never “Son of God.” In fact, Q 5:116–​17 reminds its audience how God will question Jesus ibn Maryam: “Did you say ‘Take me and my mother as gods besides God’?” Scholars debate how these passages should be read. For example, they may reflect a “Nestorian” trend in Qurʾānic Christology, which sought to maintain a clearer distinction between the divine and human natures of Christ than that promoted by Chalcedon. To emphasize this distinction, Nestorius, Patriarch of Constantinople from 428 to 431 ce, insisted that Mary should be termed “Christokos” rather than “Theotokos,” as she carried “Christ” and not “God” in her womb.40 The passage may instead offer an intentional distortion of, and commentary on, Chalcedonian Christology,41 possibly by Monophysites (the other major anti-​ Chalcedonian Christian group, which, in contrast to Nestorians, emphasized the unity or oneness of Christ’s humanity and divinity in a single nature). Chalcedon attempted—​ unsuccessfully—​to mediate these two positions by establishing the concept of “hypostatic union.” Even today, eastern Christian churches feature Nestorians, Monophysites, and Chalcedonians.42 Further, without specifically naming the Council of Chalcedon, classical exegetes of the Qurʾān gloss Qurʾānic allusions to disagreements among, and between, those who had earlier been given the Book (generally understood as pre-​Islamic Jews and Christians [e.g., Q 3:19], referring especially to the various Christian denominations familiar in their own day that trace their origins to Chalcedon, namely the Chalcedonian Melkites, non-​Chalcedonian Nestorians, and Monophysite Jacobites). In addition to points of theological doctrine, Chalcedon also attempted to resolve matters of ecclesiastical order and discipline. As with the theological disputes, the instructions of the Council did not meet with universal, or immediate, acceptance, let alone approval.43 The Qurʾān offers multivalent discussions of monks (Ar. ruhbān) and monasticism (Ar. rahbāniyya). The Qurʾān praises monks (and priests) as “humble” (Q 5:82), but also criticizes Christians 39 40 41 42 43

The Qurʾān denies Jesus’s divine sonship (cf. Q 19:35) but chastises the Jews for their slander of Mary (Q 4:156) and their claim to have killed the Messiah, Son of Mary, messenger of God (Q 4:157). For more on this Christian denomination, see Brock 1996. See especially the discussion and bibliography in Reynolds 2014. See Griffith 2001 and Griffith 2006 for discussion of the various names of these groups. Chalcedonians, for example, may be called Melkites, whereas Monophysites can be termed Jacobites. On the decline of monastic establishments, especially connected with the iconoclasm of Constantine v in the eighth century ce, see Charanis 1971.

The QurʾĀnic Rūm


for their “innovation” of monasticism (Q 57:27). Additionally, it faults Christians for taking monks as lords besides God (Q 9:31–​32) and accuses monks of “devouring wealth” (Q 9:34). These Qurʾānic allusions to monastic misconduct resonate with the canons of Chalcedon that attempt to regulate monastic communities and individual ascetics, as well as the ecclesiastical hierarchy.44 Later commentators would interpret Q 57:27’s criticism of the Christian innovation of rahbāniyya as a blanket condemnation of monasticism. But, considered together with all the Qurʾānic passages alluding to monks or monastic practices, a far more nuanced picture emerges. The passages sometimes praise monks (e.g., at Q 5:82), they list monasteries among the places of worship in which God’s name is frequently mentioned (Q 22:40), and the passages praise various practices associated with monasticism, such as night vigils, fasting, etc. (cf. e.g., Q 33:35). Although ecclesiastical councils attempted to regulate monasticism, local Christian communities did not always respond readily or happily to imperial and episcopal interference in the administration of their internal affairs. Given the varying assessments of monks and monasticism in the Qurʾānic discourse, the Qurʾān likely reflects debates about the institution of monasticism that were occurring during late antique Christianity45 rather than offering a blanket condemnation of the institution of monasticism. The Qurʾānic allusions to monks and monasticism would have resonated particularly with Christians living under the influence of the Roman (Byzantine) Empire, with its contested relationships among monks, emperors, and bishops in Late Antiquity. For later exegetes who were under Arab Muslim rule, however, ecclesiastical power struggles (especially those in which Christian emperors participated, as through the convening of ecclesiastical councils) would have lost their relevance and hence their significance for the understanding of Qurʾānic allusions to monasticism. Given the theological and administrative disputes that Chalcedon recognized and attempted to settle, it is notable that sectarian divisions figure prominently in Q 30 (Sūrat al-​Rūm). Although later exegesis would overwhelmingly understand the opening verses as alluding to a military victory/​defeat of the Byzantines, could the Qurʾān’s first audience have understood this sura differently? About midway through the sura, Q 30:30 (cf. Q 30:43) refers to “al-​dīn 44


See, for example, the third canon of Chalcedon, which decrees that bishops, monks, and clerics should not manage property or administer worldly business of their own accord. Translation at http://​​Councils/​ecum04.htm. Similar regulations existed for the Church of the East at the Council of Seleucia-​Ctesiphon (481 ce). For some of the institutional challenges Christianity faced in Late Antiquity, see Becker 2006; Sterk 2009; Rapp 2013.

44 Wilde al-​qayyim”—​translated as “straight religion” (akin to the “straight path” of Sūrat al-​Fātiḥa). In classical exegesis, this phrase is glossed as “that which has no crookedness,” as for example by the classical exegete and philosopher Fakhr al-​Dīn al-​Rāzī (d. 606 ah/​1210 ce). Previously, one of the earliest preserved exegetes, Muqātil b. Sulaymān (d. 150 ah/​767 ce), glosses this passage as “tawḥīd, oneness, al-​dīn al-​mustaqīm” in a phrase that evokes the opening chapter of the Qurʾān: the “straight path/​religion” (al-​ṣirāt al-​mustaqīm, Q 1:6). The following verses (Q 30:31–​32) exhort the Qurʾān’s audience not to be “of the associators, of those who divide their religion, becoming sects, each party of which rejoicing in its position.” In classical tafsīr, these verses receive various exegetical glosses. For example, Muqātil understands the verses as referring to humanity’s division of their (original, true) religion, Islam, into the religious parties (aḥzāban fī al-​dīn) of Judaism and Christianity and Zoroastrianism, etc. By contrast, al-​Rāzī argues that these verses refer to the true monotheism of Islam, but that the religious division is a warning against congregations within Islam (lam yajtamiʿū ʿalā al-​islām), with each congregation following its own madhhab.46 Although the victory and defeat of “al-​Rūm” found at the beginning of Q 30 are commonly interpreted in later literature as referring to late antique military engagements of the Byzantine Empire, could they also be understood as referring to contemporary religious disputes? Especially when considered within the discussions of “right” religion elsewhere in the sura, Q 30 might have been understood as a reflection of, or commentary on, Christian internecine troubles, given the duration of the inter-​Christian debates after Chalcedon and the local47 and empire-​wide ramifications of the opposition to Chalcedon. Instead of (or in addition to) a Byzantine military victory/​defeat, could the Qurʾān’s first audience have heard an echo of post-​Chalcedonian Christian polemics, namely the Christian categories of “orthodox” (“straight/​correct belief”) and heterodox (which included accusations of sectarianism, dividing the true religion)? In this interpretation, a victory/​defeat of al-​Rūm might refer to the success and failure of Chalcedonian doctrinal pronouncements or ecclesiastical ordinances, rather than a specific military encounter. In this way, the identity of the (Christian) believers would determine whether the defeat or the victory of al-​Rūm comprised the victory of God that caused rejoicing. Even if the Qurʾān in fact utilizes contemporaneous “Christian” terminology and 46 47

Tafāsīr Rāzi and Muqātil, ad loc. at Translations my own. Perhaps most famously in Alexandria, where in 451 ce a riot greeted Proterius (the Chalcedonian replacement of Dioscorus). Six years later, he was murdered in the baptistery (cf. e.g., Torrance 1998, 8).

The QurʾĀnic Rūm


arguments, it does not necessarily advocate one Christian interpretation over the others. When read in their entirety, the Qurʾānic statements on Christians and Christianity clearly indicate that the Qurʾānic “straight” religion is unlikely to be either the Chalcedonian Christianity of the Byzantines or one of the non-​Chalcedonian interpretations. Instead, for the Qurʾān, “orthodoxy” or the “straight religion” is the “fiṭrat Allāh” and the religion of the ḥanīf (commonly glossed as pre-​Islamic Arabian monotheists) can be reinterpreted as that preached by the Qurʾānic messenger, Muhammad. If this reading is correct, it is important to distinguish Qurʾānic echoes of contemporary narratives or polemics from its intended message. Although the Qurʾān is aware of contemporaneous Christian disputes, and may even have been willing to adopt their terminology (e.g., “orthodoxy” or “straight religion”), this adaptability does not necessarily translate into advocacy of any Christian position. In fact, a holistic reading of the Qurʾānic assessments of Christianity strongly argues against advocacy of the beliefs of any contemporaneous Christian communities. 4

Jerusalem—​Near and Far?

The next question concerns the location of the defeat (or victory) of al-​Rūm. As noted above, Q 30:2–​5 is commonly glossed as a reference to the late antique conflict between Persia and Rome, eventually centering on Jerusalem. Especially with the onset of the Crusades, the Romans were increasingly seen as unbelievers, and the variant reading (that the Romans would be defeated) occurs with greater frequency and the role of Jerusalem increases in importance. Although Muslims would develop a body of literature on the “virtues” of Jerusalem, Jerusalem is never specifically named in the Qurʾān: references to “holy land” (Q 5:21) and land that has been “blessed” (by God; Q 7:137) do, however, occur. Because these references appear in historical narratives about the Children of Israel, they have been interpreted as meaning Jerusalem and her environs, despite alternative possibilities (e.g., Syria or the entire world for Q 7:137, or, among others, Jordan and Mt. Sinai for Q 5:21; see also the various interpretations of Q 2:58, 259; 23:50; 24:35–​36).48 Jerusalem also figures in 48

Other Qurʾānic passages, such as Q 2:114; 95:1, have also been interpreted as allusions to Jerusalem, although modern commentators tend to read these passages as references to Mecca and the life of Muhammad, rather than Jerusalem (and the Children of Israel). See Busse 2001–​6 for a comprehensive bibliography and extensive discussion of these passages and their various interpretations.

46 Wilde the interpretation of verses related to Muhammad’s own life and prophetic career: the change of the qibla (direction of prayer at Q 2:142–​50), the “night voyage” of Q 17:1,49 and the location of the eventual Roman victory (or defeat) at the beginning of Q 30. In addition, the Qurʾān also includes a number of passages that have been interpreted as referring to various places in and around Jerusalem within the eschaton (e.g., Q 50:41; 57:13; 70:43; 79:14). As discussed above, the opening verses of Q 30 are commonly understood as referring to the (brief) Persian rule of Jerusalem in the early decades of the seventh century ce. This interpretation, however, may reflect this event’s later importance in Christian and Muslim literature, rather than contemporaneous interpretations of the significance of the Persian occupation of Jerusalem or the magnitude of the destruction inflicted by the occupiers. Despite this military engagement’s prominence in Christian apocalyptic works, the archaeological record gives little indication that the Roman military defeat and Persian occupation physically affected Jerusalem.50 Further, given the imperial and ecclesiastical discord that arose among Christian factions (for example, in the Christological debates surrounding the mid-​fifth-​century ce Council of Chalcedon), Nestorians and Jacobites would not necessarily have regarded the defeat of the Chalcedonian Byzantines (al-​Rūm) at the hands of the Persians as an absolute tragedy (nor a subsequent Roman victory as a cause of rejoicing).51 If, however, we understand the Qurʾān as embedded in local—​Arabian—​ concerns, can we attempt a reading that does not place the center of its late antique concerns in Jerusalem or other areas more familiar to later exegetes (and Western scholars)? This question becomes particularly pertinent when considering the designation of the Roman victory/​defeat as in the “nearest” land. Here, the Qurʾānic allusion in Q 17:1 to the “farthest” mosque (al-​masjid al-​aqṣāʾ) may be relevant. Islamic tradition has generally understood that Q 17:1 intends Jerusalem as the destination of Muhammad’s “night voyage.” Therefore, Jerusalem is the site of the “farthest” mosque (as opposed to the “sacred” mosque in Mecca). If Mecca is indeed the initial reference point for both Q 30 and Q 17, the traditional understanding of Jerusalem as the other location in each of these passages is potentially problematic. How can Jerusalem be located in the “nearest” land at the beginning of Q 30, and yet 49

50 51

Van Ess 1999, 48 gives a brief overview of the exegetical arguments for Jerusalem as the destination of the night voyage. Neuwirth 2003b provides an extensive discussion. See also Van Ess 1999, 50 for discussion of Orientalist arguments that the traditional markers of the subsequent heavenly ascension were, in fact, local Meccan landmarks. See Avni 2010. For some discussion of these positions, see Griffith 1997.

The QurʾĀnic Rūm


include the “farthest” mosque in Q 17:1? For, while lands and mosques are discrete entities, and therefore different comparisons may indeed be logically consistent (the nearest land may in fact house the farthest mosque), could the Qurʾān’s first audience have understood another place as the “nearest” land at the beginning of Q 30?52 Similarly, if considered within the context of late antique Christian controversies—​and alliances—​could the Qurʾān’s first audience have understood this (Meccan) chapter as referring to Byzantine victories/​defeats in south Arabia in the sixth century ce, rather than in Syria/​Palestine during the seventh century ce?53 In the sixth century ce, Constantinople sought to actively expand its influence on its eastern and southern fringes, partially in response to the Persian presence. Ghassānids (Monophysites) served on its eastern frontier as “border guards” against the Persian Empire.54 (The Ghassānids were Arabs who had migrated from southern Arabia and settled within the Roman frontier by the end of the fifth century ce.) Although the Ghassānids were anti-​Chalcedonian, doctrinal differences came second to the advantages Constantinople gained from this alliance. In the middle of the sixth century ce, the Ghassānids could even secure the ordination of two Monophysite bishops—​and with the assistance of the Byzantine empress Theodora. (One of these bishops was Jacob Baradaeus; thus the Syrian Monophysites became known as “Jacobites.”)55 To the south, the Red Sea held a crucial position in trade routes to Arabia and the Far East.56 As with the Byzantine-​Ghassānid alliance, Justinian attempted alliances with (Monophysite) Christians in both southern Arabia and eastern Africa in the first quarter of the sixth century ce.57 These efforts received particular energy when a Jewish (and Judaizing) king, Dhū Nuwās, came to power in southern Arabia. A timeline of the king’s reign can be reconstructed from Greek, Syriac, and later Arabic accounts. Just before or after 520 ce, Dhū Nuwās rose to power in 52 53

54 55 56 57

The obvious companion question is whether the location of the “farthest” mosque of Q 17:1 is, in fact, correctly identified as Jerusalem. For discussion of the identification of this “farthest mosque” in Islamic tradition, see, e.g., Neuwirth 2003b and Hasson 1996. On the interactions between eastern Christians and early Islam, see the recent study of Penn 2015. For a comprehensive overview of various early understandings of Islam, from different communities, see Hoyland 1997. See also Lamoreaux 2000 and the articles in Grypeou, Swanson, and Thomas 2006. For a recent discussion of this relationship, see Hoyland 2009. For further discussion, see also Greatrex and Lieu 2005. See Shahid 1960–​2007 for a concise overview of the Ghassānids and further bibiography. For further discussion, see Cameron 2015, esp. 173. For detailed discussion of these relationships, see Power 2012.

48 Wilde Ḥimyar, a kingdom in southern Arabia. The accounts portray him as ambitious and merciless, especially towards the Christians of Najrān; Christian hagiography preserves the memory of their martyrdom at his orders in 524 ce.58 In response to this atrocity (and also the disruption of trade under his reign, as for example in the reports of Ethiopian merchants murdered in this period), the Ethiopian king of Axum invaded Ḥimyar in 525 ce with the support of a Roman fleet. By the early 530s ce, however, the Ethiopians had been overthrown and by 570 ce the Persians had annexed Ḥimyar. Much like the different traditions surrounding the “Sleepers of Ephesus” (discussed above), the various historical chronicles and hagiographies that discuss this king do not agree on the details of his reign or life. (For example, did he really convert to Judaism? How many Ethiopian invasions occurred and when?)59 The significance of these events can also be understood on multiple levels: Jewish-​Christian disputes in southern Arabia, local Ethiopian-​Yemeni rivalries, and, at the level of “superpower” conflict, as one of the Roman-​ Sassanian military theaters. In Christian hagiography, however, these “martyrs of Najrān” are the first anti-​Chalcedonian martyrs. To echo Lucas Van Rompay, later accounts of the massacre of men, women, and children at Jewish hands, while valiantly defending their orthodox (i.e., anti-​Chalcedonian) faith, had considerable rhetorical effect.60 For anti-​Chalcedonians, the alliance of the Byzantine (and Chalcedonian) Justin i with the non-​Chalcedonian Ethiopian Christian ruler against Dhū Nuwās, in concert with the martyrdoms at Najrān, vindicated their form of Christianity.61 Instead of being called heretics by imperial and ecclesiastical authorities in Constantinople, (Monophysite) Ethiopians received the support of an imperial fleet as they attempted to avenge the murder of their coreligionists, the (Monophysite) martyrs of Najrān. As a result, the Ethiopians (with Roman support) ruled southern Arabia until the 570s ce, when, thanks to Persian support, the region returned to Ḥimyarite control. Although many contemporary Christians may not have heard of the martyrs of Najrān, this event’s significance in Christian historical memory is indicated by its commemoration in the Roman Catholic calendar on 24 October (despite the anti-​Chalcedonian victims). Islamic tradition also preserves their memory: the “People of the Ditch” of Q 85 are commonly identified as these martyrs.62 Q 85 speaks of the “People of the Ditch” and alludes to tortured 58 59 60 61 62

See, for example, Barrett 2013. See also Shahid 1971. For further details, see the discussion in Power 2012. Van Rompay 2005, 254–​57. Van Rompay 2005, 243–​44. For further discussion, see e.g., Cook 2008. For two contemporary views on the significance of these martyrs, see Mulligan 1977 and Worth 2010.

The QurʾĀnic Rūm


“believers” (both male and female), as well as the promise of a heavenly reward for those who “believe and do good deeds” (Q 85:11). The later exegetical tradition—​even as early as Muqātil—​explained these allusions through the account of Dhū Nuwās and the men and women he ordered to be burnt in Najrān. The reason for their deaths, however, would be attributed to their refusal to apostatize—​from Islam.63 Multiple readings of gh-​l-​b (victory/​ defeat), combined with uncertain chronology and geography, allow a flexible reading of Q 30’s opening passage, as seen in the various interpretations of later exegetes. Because Najrān is closer than Jerusalem to Mecca, would the Qurʾān’s first audience have understood the Roman victory/​defeat mentioned in Q 30 as an allusion to the Roman-​Persian wars in the “closer” battlefield of southern Arabia in the sixth century ce? At least two other Qurʾānic passages may refer to events directly, or indirectly, associated with Roman-​Persian struggles for control of southern Arabia: Q 85 and Q 105. In Q 105, the “People of the Elephant” are commonly understood as an allusion to an expedition against Mecca undertaken by an Ethiopian ruler with ties to Constantinople in the mid-​to late sixth century ce, around the year of Muhammad’s birth.64 Although the date of the expedition is contested, it is generally understood as having been unsuccessful. Soon after (in the 570s ce), a Persian expedition successfully expelled the Ethiopians and reinstated Ḥimyarite rule in southern Arabia. With this late antique background in mind, Q 30’s rejoicing of the “believers” in God’s victory resonates with the reaction of Monophysite Christians at the Byzantine support given to the non-​Chalcedonian Christians after the martyrdom of Najrān as well as with their joy at the ultimate reward they believe will be granted to the martyrs themselves in paradise. Further, al-​dīn al-​qayyim (“the straight religion”), interpreted by later Muslim commentators as Islam, finds a place in various discussions among Christians about “orthodoxy”—​ as seen, for example, in the claims by both Chalcedonian and non-​Chalcedonian Christians to the “orthodox” (straight/​correct) belief. 5


Although Islamic tradition maintains a strong connection between the opening verses of Q 30 and Jerusalem, including the belief that the individual who recites Q 30 receives compensation for each verse in the same manner as someone who 63 64

See the Tafsīr of Muqātil, ad Q 85:4 at For further details on the date, especially the epigraphic record, see the recent discussion in Robin 2015.

50 Wilde visits Jerusalem, the Arabic text does not specify the chronological or geographical location of the events in the passage. However, when the Qurʾān is considered an Arabic bookend to Late Antiquity, rather than an Ur-​text for classical Islamic civilization, the continuing impact of the theological and administrative decisions of the Council of Chalcedon on local Christian communities cannot be overlooked. In this context, Qurʾānic references to a “straight” religion may echo inter-​Christian arguments over “orthodoxy” that increased in intensity after Chalcedon, in reaction to the persistence and growth of numerous and disparate communities that continued to adhere to Christological understandings other than those promulgated by the Council. Each Christian faction considered itself the defender of orthodoxy—​the correct, “straight” belief—​similar to the claims made about al-​dīn al-​qayyim in Q 30. For over a century after Chalcedon, emperors would attempt—​unsuccessfully—​to appease the Christian factions. However, when dealing with those on the periphery, or outside, of the empire, doctrinal differences were sometimes overlooked in the interest of military and trade alliances, as seen in Byzantium’s support of Monophysite regimes in Ethiopia and Yemen in the sixth century ce. Q 85 and Q 105 traditionally refer to south Arabian events associated with Christians and, indirectly, Byzantium. As the sixth-​century ce “Martyrs of Najrān” arguably made a stronger impression on the landscape of Late Antiquity—​in Arabia and beyond—​than the Persian capture of Jerusalem, particularly for the anti-​Chalcedonian Christian factions, this paper has argued that the Qurʾān’s first audience may have heard Q 30 as an allusion to events in southern Arabia. However, the Qurʾān’s own polemics and apologetics should not be confused, or interchanged, with those of its later interpreters. Continued scholarly investigation into the processes by which Late Antiquity transitioned to medieval Christendom or classical dār al-​islām is a desideratum. For historians of late antique Arabia, attempts to hear the Qurʾān in the same way as its first audience may prove more useful than the readings of later exegetes, whose interpretations frequently reflect the inter-​communal polemics and other dynamics of the milieu of their composition, rather than the Sitz-​im-​Leben of the Qurʾān itself. And, because much of the Islamicate world is itself heir to the world of Late Antiquity, readings of the Qurʾān that attempt to situate it within this world may complement the interpretations of later exegetes.


EQ = McAuliffe, Jane, ed. 2001–​6. Encyclopaedia of the Qurʾān. Leiden: Brill. Atiya, Aziz S. 1968. A History of Eastern Christianity. London: Methuen. Attwater, Donald. 1961. The Christian Churches of the East. London: Chapman.

The QurʾĀnic Rūm


Avni, Gideon. 2010. “The Persian Conquest of Jerusalem (614 CE): An Archaeological Assessment.” Bulletin of the American Schools of Oriental Research 357:35–​48. Barrett, Richard. 2013. “Sensory Experience and the Women Martyrs of Najran.” Journal of Early Christian Studies 21.1:93–​109. Becker, Adam H. 2006. Fear of God and the Beginning of Wisdom: The School of Nisibis and the Development of Scholastic Culture in Late Antique Mesopotamia. Philadelphia: University of Pennsylvania Press. Berg, Herbert. 2013. The Development of Exegesis in Early Islam: The Authenticity of Muslim Literature from the Formative Period. London: Routledge. Boisliveau, Anne-​Sylvie. 2014. Le Coran par lui-​même: Vocabulaire et argumentation du discours coranique autoréférentiel. Leiden: Brill. Brock, Sebastian P. 1996. “The ‘Nestorian’ Church: A Lamentable Misnomer.” Bulletin of the John Rylands University Library of Manchester 78.3:23–​35. Busse, Heribert. 2001–​6. “Jerusalem.” In eq. Butcher, E. L. 1897. The Story of the Church of Egypt. London: Smith. Cameron, Averil. 2015. The Mediterranean World in Late Antiquity: AD 395–​700. London: Routledge. Casey, P. J. 1996. “Justinian, the Limitanei and Arab-​Byzantine Relations in the 6th c.” Journal of Roman Archaeology 9:214–​22. Charanis, P. 1971. “The Monk as an Element of Byzantine Society.” Dumbarton Oaks Papers 25:61–​84. Cohen, Jeremy. 1983. “The Jews as the Killers of Christ in the Latin Tradition, from Augustine to the Friars.” Traditio 39:1–​27. Cook, Michael. 2000. The Koran: A Very Short Introduction. Oxford: Oxford University Press. Cook, David. 2008. “The Aṣḥāb al-​Ukhdūd: History and Ḥadīth in a Martyrological Sequence.” Jerusalem Studies in Arabic and Islam 34:125–​48. Cucarella, Diego Sarrió. 2010. “Corresponding across Religious Borders: The Letter of Ibn Taymiyya to a Crusader in Cyprus.” Islamochristiana 36:187–​212. Ebied, Rifaat Y., and David Richard Thomas, eds. 2005. Muslim-​Christian Polemic during the Crusades: The Letter from the People of Cyprus and Ibn Abī Ṭālib al-​Dimashqī’s Response. Leiden: Brill. El-​Cheikh, Nadia Maria. 1998. “Sūrat al-​Rūm: A Study of the Exegetical Literature.” Journal of the American Oriental Society 118:356–​64. El-​Cheikh, Nadia Maria. 2001–​6. “Byzantines: Exegetical Explanations.” In eq. Ess, Josef van. 1999. “Vision and Ascension: Sūrat al-​Najm and its Relationship with Muḥammad’s Miʿrāj.” Journal of Qur’anic Studies 1.1:47–​62. Fisher, Greg. 2011. “Kingdoms or Dynasties? Arabs, History, and Identity before Islam.” Journal of Late Antiquity 4.2:245–​67. Foss, C. 2003. “The Persians in the Roman Near East (602–​630 AD).” Journal of the Royal Asiatic Society of Great Britain & Ireland 13.2:149–​70.

52 Wilde Fowden, Elizabeth Key. 1999. The Barbarian Plain: Saint Sergius between Rome and Iran. Berkeley: University of California Press. Gilliot, Claude. 2009. “Christians and Christianity in Islamic Exegesis.” In Christian-​ Muslim Relations: A Bibliographical History. Volume 1 (600–​900), edited by David Thomas et al., 31–​56. Leiden: Brill. Gilliot, Claude, and Pierre Larcher. 2001–​6. “Language and Style of the Qurʾān.” In eq. Greatrex, Geoffrey, and Samuel N. C. Lieu. 2005. The Roman Eastern Frontier and the Persian Wars AD 363–​628. London: Routledge. Griffith, Sidney. 1997. “From Aramaic to Arabic: The Languages of the Monasteries of Palestine in the Byzantine and Early Islamic Periods.” Dumbarton Oaks Papers 51:11–​31. Griffith, Sidney. 2001. “ ‘Melkites,’ ‘Jacobites,’ and the Christological Controversies in Arabic in Third/​Ninth Century Syria.” In Syrian Christians under Islam: The First Thousand Years, edited by David Thomas, 9–​55. Leiden: Brill. Griffith, Sidney. 2006. “The Church of Jerusalem and the ‘Melkites’: The Making of an ‘Arab Orthodox’ Christian Identity in the World of Islam (750–​1050 CE).” In Christians and Christianity in the Holy Land: From the Origins to the Latin Kingdom, edited by Ora Limor and Guy Stroumsa, 175–​204. Turnhout: Brepols. Griffith, Sidney. 2007. “The ‘Companions of the Cave’ in Sūrat al-​Kahf and in Syriac Christian Tradition.” In The Qurʾān in Its Historical Context, edited by Gabriel Said Reynolds, 109–​38. London: Routledge. Griffith, Sidney. 2011. “Al-​Naṣāra in the Qurʾān: A Hermeneutical Reflection.” In New Perspectives on the Qurʾān: The Qurʾān in its Historical Context, 2, edited by Gabriel Said Reynolds, 301–​22. London: Routledge. Griffith, Sidney. 2012. The Church in the Shadow of the Mosque: Christians and Muslims in the World of Islam. Cambridge, MA: Princeton University Press. Grypeou, Emmanouela, Mark Swanson, and David Thomas, eds. 2006. The Encounter of Eastern Christianity with Early Islam. Leiden: Brill. Hardy, Edward Rochy. 1952. Christian Egypt: Church and People. New York: Oxford University Press. Hasson, Izhak. 1996. “The Muslim View of Jerusalem. The Qurʾān and Ḥadīth.” In The History of Jerusalem. The Early Muslim Period, 638–​1099, edited by Joshua Prower and Haggai Ben-​Shamai, 349–​85. New York: New York University Press. Hawting, Gerald. 1997. “Shirk and Idolatry in Monotheist Polemic.” Israel Oriental Studies: Dhimmis and Others: Jews and Christians and the World of Classical Islam 17:107–​26. Hoffman, Valerie J. 1998. “Qurʾānic Interpretation and Modesty Norms for Women.” In The Shaping of an American Islamic Discourse: A Memorial to Fazlur Rahman, edited by Earle H. Waugh and Frederick M. Denny, 89–​122. Atlanta: Scholars Press. Hoyland, Robert G. 1997. Seeing Islam as Others Saw It: A Survey and Evaluation of Christian, Jewish and Zoroastrian Writings on Early Islam. Princeton: Darwin Press.

The QurʾĀnic Rūm


Hoyland, Robert. 2009. “Late Roman Provincia Arabia, Monophysite Monks and Arab Tribes: A Problem of Centre and Periphery.” Semitica et Classica 2:117–​39. Hoyland, Robert. 2012. “Early Islam as a Late Antique Religion.” In The Oxford Handbook of Late Antiquity, edited by Scott Fitzgerald Johnson, 1053–​77. Oxford: Oxford University Press. Jones, Alan. 2001–​6. “Orality and Writing in Arabia” In eq. Jones, Arnold Hugh Martin. 1959. “Were Ancient Heresies National or Social Movements in Disguise?” The Journal of Theological Studies 10.2:280–​98. Kaegi, Walter E. 1995. Byzantium and the Early Islamic Conquests. Cambridge: Cam­ bridge University Press. Lamoreaux, John C. 2000. “Early Eastern Christian Responses to Islam.” In Medieval Christian Perceptions of Islam, edited by John Victor Tolan, 3–​31. London: Routledge. Leemhuis, Frederik. 2001–​6. “Readings of the Qurʾān.” In eq. Madigan, Daniel A. 2001. The Qur’ân’s Self-​Image: Writing and Authority in Islam’s Scripture. Princeton: Princeton University Press. McAuliffe, Jane Dammen. 1991. Qurʾānic Christians: An Analysis of Classical and Modern Exegesis. Cambridge University Press, electronic version 2007. Michel, Thomas F., trans. 1984. A Muslim Theologian’s Response to Christianity: Ibn Taymiyya’s Al-​Jawab al-​sahih. Delmar: Caravan. Modarressi, Hossein. 1993. “Early Debates on the Integrity of the Qurʾān: A Brief Survey.” Studia Islamica 77:5–​39. Motzki, Harald. 2001. “The Collection of the Qurʾān: A Reconsideration of Western Views in Light of Recent Methodological Developments.” Der Islam 78.1:1–​34. Mulligan, William E. 1977. “The Martyrs of Najran,” Catholic Near East Magazine, Fall. http://​​default.aspx?ID=101&pagetypeID=4&sitecode=HQ&pageno=1. Nasr, Seyyed Hossein, et al. 2015. The Study Quran: A New Translation and Commentary. New York: Harper Collins. Nasry, Wafik. 2008. The Caliph and the Bishop: A 9th Century Muslim-​Christian Debate; Al-​Maʾmūn and Abū Qurrah. Beirut: cedrac. Neuwirth, Angelika. 2001–​6. “Rhetoric and the Qurʾān.” In eq. Neuwirth, Angelika. 2003a. “Qur’an and History–​a Disputed Relationship: Some Reflections on Qur’anic History and History in the Qur’an.” Journal of Qur’anic Studies 5.1:1–​18. Neuwirth, Angelika. 2003b. “From the Sacred Mosque to the Remote Temple: Sūrat al-​Isrāʾ between Text and Commentary.” In With Reverence for the Word: Medieval Scriptural Exegesis in Judaism, Christianity, and Islam, edited by Jane Dammen McAuliffe, Barry D. Walfish, and Joseph W. Goering, 376–​407. Oxford: Oxford University Press. Penn, Michael Philip. 2015. Envisioning Islam: Syriac Christians and the Early Muslim World. Philadelphia: University of Pennsylvania Press.

54 Wilde Peters, Francis. 1997. “Alius or Alter: The Qur’anic Definition of Christians and Christianity.” Islam and Christian‐Muslim Relations 8.2:165–​76. Power, Timothy. 2012. The Red Sea from Byzantium to the Caliphate: AD 500–​1000. Cairo: auc Press. Rapp, Claudia. 2013. Holy Bishops in Late Antiquity: The Nature of Christian Leadership in an Age of Transition. Berkeley: University of California Press. Raven, Wim. 2001–​6. “Sīra and the Qurʾān.” In eq. Reynolds, Gabriel Said. 2004. A Muslim Theologian in the Sectarian Milieu: ʿAbd Al-​ Jabbār and the Critique of Christian Origins. Leiden: Brill. Reynolds, Gabriel Said. 2014. “On the Presentation of Christianity in the Qurʾān and the Many Aspects of Qur’anic Rhetoric.” Al-​Bayān: Journal of Qurʾān and Ḥadīth Studies 12.1:42–​54. Robin, C. 2015. “Ḥimyar, Aksūm and Arabia Deserta in Late Antiquity: The Epigraphic Evidence.” In Arabs and Empires before Islam, edited by Greg Fisher, 127–​71. Oxford: Oxford University Press. Safran, Janina M. 2003. “Rules of Purity and Confessional Boundaries: Maliki Debates about the Pollution of the Christian.” History of Religions 42.3:197–​212. Sellers, R.V. 1953. The Council of Chalcedon: A Historical and Doctrinal Survey. London: spck. Shahid, Irfan. 1960–​2007. “Ghassān.” In Encyclopaedia of Islam, 2nd ed., edited by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs. Leiden: Brill. Shahid, Irfan. 1971. The Martyrs of Najrān: New Documents. Brussels: Société des Bollandistes. Shahid, Irfan. 1972. “The Iranian Factor in Byzantium during the Reign of Heraclius.” Dumbarton Oaks Papers 26:293–​320. Shahid, Irfan. 1979. “Byzantium in South Arabia.” Dumbarton Oaks Papers 33:23–​94. Shahid, Irfan. 1984. Rome and the Arabs: A Prolegomenon to the Study of Byzantium and the Arabs. Washington: Dumbarton Oaks. Shahid, Irfan. 2001–​6. “Najrān.” In eq. Sinai, Nicolai. 2014. “When Did the Consonantal Skeleton of the Quran Reach Closure? Part I.” Bulletin of the School of Oriental and African Studies 77.2:273–​92. Sterk, Andrea. 2009. Renouncing the World Yet Leading the Church: The Monk-​Bishop in Late Antiquity. Cambridge, MA: Harvard University Press. Tanner, Norman, and G. Alberigo, ed. 1990. Decrees of the Ecumenical Councils. London: Sheed & Ward. Thomas, David, trans. 1992. Anti-​Christian Polemic in Early Islam: Abu ʿIsā al-​Warrāq’s “Against the Trinity.” Cambridge: Cambridge University Press. Torrance, Iain R. 1998. Christology after Chalcedon: Severus of Antioch and Sergius the Monophysite. Eugene: Wipf and Stock.

The QurʾĀnic Rūm


Tottoli, Roberto. 1999. “Origin and Use of the Term Isrāʾīliyyāt in Muslim Literature.” Arabica 46:193–​210. Van Rompay, Lucas. 2005. “Society and Community in the Christian East.” In The Cambridge Companion to the Age of Justinian, edited by Michael Maas, 239–​66. Cambridge: Cambridge University Press. Wakin, E. 1963. A Lonely Minority: The Modern Story of Egypt’s Copts. New York: Morrow. Wansbrough, John. 1977. Quranic Studies: Sources and Methods of Scriptural Interpretation. Oxford: Oxford University Press. Wigram, W. A. 1923. The Separation of the Monophysites. London: Faith. Wild, Stefan. 2006. Self-​Referentiality in the Qurʾān. Wiesbaden: Harrassowitz. Worth, Robert. 2010. “Muslim Sect Sees Struggle through Christian Lens.” New York Times, October 20. http://​​2010/​10/​21/​world/​middleeast/​21saudi .html.

­c hapter 3

Wine and Impurity in the Sura of the Bees: A Structuralist Interpretation of Qurʾān 16:67 Johan Weststeijn Islamic Law famously forbids the drinking of wine.1 In the Qurʾān, five verses express an opinion on this beverage, among them verse 16:67: “And from the fruits of the palms and the vines you take intoxicant and good provision. Most surely there is a sign in this for a people who ponder” (wa-​min thamarāt al-​nakhīl wa-​ l-​aʿnāb tattakhidūna minhu sakaran wa-​rizqan ḥasanan inna fī dhālika la-​āya li-​ qawm yaʿqilūn).2 Here the term sakar, “wine, intoxicant,” means an alcoholic drink made from fermented grapes or dates.3 The pairing of date palms and grapevines is common in the Qurʾān, for in hot and arid parts of the Middle East grapevines were often planted in the shadow of date palms, so the tendrils of the vine could twine around the trunks of the palms.4 Many Western studies on wine in the Qurʾān argue that verse 16:67—​despite Islamic Law’s ban of wine—​presents a positive assessment of this drink, an opinion shared by a number of medieval Muslim exegetes. Adherents to this positive view interpret sakar wa-​rizq ḥasan, “wine and good provision,” as a pair of equals: wine is on par with good provision. The phrase “most surely there is a sign in this for a people who ponder” is read to mean that wine “is praised as one of the signs of Allāh’s grace to mankind.”5 1 Wensinck 1960–​2007; McAuliffe 1984, 159; Tapper 1994, 216. 2 Translation adapted from Shakir 1999. The other verses that express an opinion on drinking wine in this life are Q 2:219, 4:43, 5:90–​91. 3 That the Qurʾān uses sakar, “intoxicant,” here to refer to khamr, “wine,” can be concluded from the fact that verses 16:65–​69 deal with water, milk, sakar, and honey, while verse 47:15 mentions, in the same order, water, milk, khamr, and honey. Note also that in the New Testament (Matthew 26:29; Mark 14:25; Luke 22:18) the expression “fruit of the vine” refers to “wine.” 4 Q 2:226, 17:91, 18:32, 23:19, 36:34; Albenda 1974; Tengberg 2012. 5 Wensinck 1960–​2007. Al-​Ṭabarī paraphrases in his Tafsīr to Q 16:67: “wa-​lakum ayḍan ayyuhā al-​nās ʿibra fīmā nasqīkum min thamarāt al-​nakhīl wa-​l-​aʿnāb” (There is also a lesson for you, people, in what We pour you from the fruits of the palms and the vines). Rāzī, Mafātīḥ al-​ ghayb, ad Q 2:219 quotes the opinion of Abū Ḥanīfa (translation Gätje 1976, 206): “When God said: ‘And (We give you) the fruits of the palms and the vines, from which you obtain an intoxicant as well as wholesome food’ then He granted a favor to us in that we (may) make an

© Johan Weststeijn, 2022 | DOI:10.1163/9789004500648_004

Wine and Impurity in the Sura of the Bees


However, some medieval and modern Muslim exegetes argue that Q 16:67 negatively assesses sakar, wine. Adherents to this view interpret sakar wa-​rizq ḥasan not as a pair of equals but as a pair of opposites.6 Wine is the opposite of good provision because wine is haram (impure, forbidden), while good provision is halal (pure, allowed). The famous early exegete Ibn ʿAbbās is reported to have said about this verse: “Wine is the one of the pair which has been forbidden (declared as impure), and good provision is the one of the pair which has been allowed (declared as pure).”7 The Encyclopaedia of Islam and the Encyclopaedia of the Qurʾan adhere to the first interpretation, which reads sakar wa-​rizq ḥasan as a pair of equals and therefore argues that this verse is positive about wine; these reference works do not mention that there also exists a negative reading.8 However, this positive reading raises a number of problems that need to be addressed. Because Islamic Law condemns wine, a positive interpretation of Q 16:67 would mean that the Qurʾān here starkly contradicts the Law. Even more problematic, a positive reading would make Q 16:67 the only verse in the Qurʾān that is unequivocally positive about this beverage. All other verses about drinking wine in this life—​as opposed to the Hereafter—​label it as a sinful, impure, or polluting potion, and therefore frown upon its consumption or condemn it outright.9 In wine there is more sin than benefit (Q 2:219); praying when intoxicated is forbidden because intoxication is like ritual uncleanliness (Q 4:43); wine is a pollution (rijs) caused by Satan (Q 5:90). Interpreting Q 16:67 as positive about wine necessarily leads to the conclusion that the Qurʾān contradicts itself on the topic of this drink. Some scholars therefore argue that God’s Book is fundamentally ambivalent towards earthly wine. In her monograph on wine in early Islam, Kathryn Kueny writes: “The potent liquid

intoxicating drink and wholesome food. (Therefore) that through which we have an intoxicating drink and wholesome food must be permitted, since a benefaction (which God grants) cannot be other than permitted.” See also Tapper 1994, 219; Kueny 2001, 10–​13; Kueny 2001–​6; Hallaq 2001–​6; Schmoll 2006. 6 Rāzī, Mafātīḥ al-​ghayb, ad Q 16:67: “innahu taʿāla nabbaha fī hādhihi al-​āya ayḍan ʿalā taḥrīmihā wa-​dhālika li-​annahu mayyiza baynahā wa-​bayn al-​rizq al-​ḥasan fī al-​dhikr, fa-​ wajaba an lā yakūna al-​sakar rizqan ḥasanan” (God points to the proscription of wine in this verse as well, for He makes a distinction between it and good provision; wine cannot be good provision); ʿAlī 1950, 738; Quṭb 1967, 5:82, 14:78; Mawdūdī 2006, ad Q 16:67, n55; Asad 2008, 2:451n76. 7 al-​Ṭabarī, Tafsīr, ad Q 16:67: “al-​sakar mā ḥurrima minhumā wa-​l-​rizq al-​ḥasan mā uḥilla minhumā.” 8 Wensinck 1960–​2007; Kueny 2001–​6; Hallaq 2001–​6. 9 On the relationship between sin and impurity in the Qurʾān see Freidenreich 2010, 12–​19.

58 Weststeijn that constitutes an abomination in one verse becomes a source of ‘good food’ in the other.”10 The classic strategy to explain this contradiction between a positive reading of Q 16:67 and the negative attitude of Islamic Law and the other Qurʾānic verses on wine argues that these verses were revealed at different moments in the Prophet’s career. At first the Revelation was positive towards wine, but it gradually grew more dismissive because some of Muhammad’s followers misbehaved as a result of drunkenness. The verses that frown upon and condemn wine would have been revealed later and would have abrogated the original positive message of Q 16:67. Islamic Law’s attitude towards wine would consequently have been derived from the verses that were revealed later.11 Traditional exegesis of the Qurʾān tends to take individual verses out of their immediate textual context and tries to interpret these verses by connecting them to specific events in the Prophet’s career. When verses from different parts of the Qurʾān appear to contradict each other, this is explained by pointing to changing circumstances in Muhammad’s life.12 Recently, however, scholars have stressed that the Qurʾān is not only the product of Muhammad’s direct environment in Mecca and Medina, but also of the wider culture of Late Antiquity.13 Others have argued, moreover, that the Qurʾān is not a loose collection of disparate utterances but a well-​structured text, and that insight into this structure helps our understanding of the individual verses.14 Consequently, to interpret wine verse 16:67, I will not try to connect it to a particular event in the Prophet’s career. Rather, I will place it in three different contexts: 1) ideas about wine, purity, and pollution in the larger worldview of Near Eastern Antiquity; 2) ideas about purity and pollution in the Qurʾān; and 3) the immediate textual context of Q 16:67, i.e., the verses that directly precede and follow this verse. Cultures in the late antique Near East attributed positive as well as negative qualities to wine. Wine can expand the mind and lead to rapture and ecstasy, but it can also obscure the mind and induce slurred speech and reckless

10 Kueny 2001, 1. See also Kueny 2001–​6. 11 Rāzī, Mafātīḥ al-​ghayb, ad Q 16:67: “hādhihi al-​sūra makkiyya, wa taḥrīm al-​khamr nazala fī sūrat al-​māʾida fa-​kāna nuzūl hādhihi al-​āya fī al-​waqt alladhī kānat al-​khamr fīhi ghayr muḥarrima” (Sura 16 is Meccan, and the proscription of wine was revealed in Sura 5 (al-​ Māʾida), so verse 16:67 was revealed at a time when wine was not forbidden); idem, ad Q 2:219 (= Gätje 1976, 200–​209); Wensinck 1960–​2007; Hallaq 2001–​6; Kueny 2001–​6. 12 Reynolds 2011. 13 For example, Hoyland 2012, 1072. 14 Cuypers 2011, 6; Ernst 2011.

Wine and Impurity in the Sura of the Bees


behavior.15 The positive and negative qualities of wine are related to a number of oppositions in the worldview of Late Antiquity that center on the opposition between purity and pollution.16 These pairs of oppositions and the corresponding worldview are also found in the Qurʾān.17 As a student of comparative religion, David Freidenreich argues that the Qurʾān should be considered a member of “a Near Eastern literary corpus that spans the boundaries separating one religious community from another.”18 He argues for example that to understand Qurʾānic ideas about impurity we should not only view them through the lens of later Islamic exegesis and law, but also compare them with biblical ideas about defilement. Although Qurʾānic and biblical notions about impurity might differ in detail, Freidenreich notes that in these two scriptures we find a common conceptual framework. The types of impurity addressed in the Quran more closely resemble those found in biblical literature than those addressed within Islamic legal literature. The significant differences in quranic and biblical discourse about holiness and impurity, therefore can best be understood as existing within the framework of a shared typology.19 In the late antique worldview, opposition between pure and impure corresponds to oppositions between life and death, growth and decay, eternal and transient, heaven and earth. On earth all things are transitory and eventually become polluted: fruit rots, people die. In heaven, by contrast, all things are eternal. They remain pure and never become polluted: people have eternal life and, as the Qurʾān states, maidens remain forever virgins, while fruit, water, and milk do not spoil.20 In this late antique worldview, to make pure is not only to clean, to clarify, and to enlighten, but also to cure, to let grow, and to bring to life. To make impure is to cloud, to obscure, and to stupefy, but also to intoxicate, to make ill, and to let die.21

15 16 17 18 19 20 21

Compare Judges 9:13 with Proverbs 31:4–​5. See further McAuliffe 1984, 160, 167; Kueny 2001, xi–​xvi, 89–​104; Schmoll 2006; Berenbaum and Skolnik 2007, 21:80–​81. Douglas 1966; Parker 1983; Choksy 1989. Lowry 2001–​6; Freidenreich 2010; Boisliveau 2014. Freidenreich 2010, 7. Freidenreich 2010, 6–​7. Q 44:56, 47:15, 56:17–​37, 76:13–​19; Tapper 1994, 220; Stetkevych 1997; Kueny 2001, 16. Cf. Revelation 7:16–​17, 21:4. Douglas 1966; Parker 1983; McAuliffe 1984; Choksy 1989; Freidenreich 2010; Boisliveau 2014.

60 Weststeijn Its positive, mind-​expanding quality allows wine to open a pathway to the divine, which makes it a sacred, heavenly drink.22 The negative quality of wine that befuddles reason and leads to foolish behavior (and eventually to a hangover) is a polluting aspect that makes its consumption dangerous for people on earth. In heaven, however, wine can be drunk without restraint, because in heaven the negative, polluting qualities of things do not exist.23 In heaven wine is not stupefying. The Qurʾānic paradise is an eternal banquet where the blessed drink a wine that does not contain anything noxious: it does not cloud the mind or induce headache (Q 37:45–​47; 56:18–​19; 76:21). Drinking wine on earth, however, is forbidden to Muslims because—​as explained by Islamic exegetes—​on earth wine clouds the mind and pollutes the drinker.24 1

The Debilitating and Polluting Aspects of Sakar

It is important to note that while all the other Qurʾānic verses about wine use either the highly positive raḥīq (Q 83:52, for the pure vintage of paradise) or the more common and neutral term khamr, only 16:67 refers to wine with a term that emphasizes its negative, polluting qualities: sakar, “intoxicant.”25 The verb askara, from the same root, carries the sense of “to deprive of reason,” and the noun sukr means not only “intoxication” but also “a state that intervenes as an obstruction between a man and his intellect.” The Qurʾān uses the expression sukkirat abṣārunā, “our eyes became clouded” (Q 15:15).26 That the root s-​k-​r stresses not only the stupefying but also the polluting qualities of wine is further shown by the Qurʾānic proscriptions of the form lā taqrabū … ḥatta … “Do not approach an object when you (or that object) are impure until you (or that object) are pure again.”27 In the following quotations, the state of impurity is printed in red, while the state of purity is printed in bold: (4:43a) lā taqrabū al-​ṣalāt wa-​antum sukārā ḥatta taʿlamū mā taqūlūn (4:43b) wa-​lā [taqrabū al-​ṣalāt wa-​antum] junuban … ḥatta taghtasilū (2:222) fa-​ʿtazalū al-​nisāʾ fī al-​maḥīḍ fa-​lā yaqrabuhunna ḥatta yaṭhurna

22 Kueny 2001, xi; Gutsfeld 2006a; Schmoll 2006. 23 Cf. Revelation 7:16–​17, 21:4. McAuliffe 1984; Tapper 1994, 220; Thomas 2006. 24 Rāzī, Mafātīḥ al-​ghayb, ad Q 2:219 (Gätje 1976, 205); McAuliffe 1984; Stetkevych 1997; Kueny 2001. 25 Khamr is used in Q 2:219, 5:90–​91, 12:36, 12:41, 47:15. 26 Lane 1863–​93, s.v. s-​k-​r. 27 Cf. Stetkevych 1993, 71.

Wine and Impurity in the Sura of the Bees


(4:43a) Do not approach prayer when you are intoxicated until you know what you are saying (4:43b) And [do not approach prayer when you are] ritually unclean … until you have washed yourselves (2:222) Keep away from women when they are menstruating, and do not approach them until they are ritually pure28 First, we notice that sukārā, the plural of sakrān, “intoxicated,” is presented as the opposite of “to know what one is saying,” i.e., to be wise, which implies that sakrān has the connotation of “foolish.” Second, we notice that sakrān is put on par with “ritually unclean” and “menstruating,” while its opposite is put on par with “to have washed oneself” and “to be ritually pure” (See Table 3.1). That the Qurʾān sees the pollution caused by intoxication as parallel to the pollution caused by the menses, is also shown by the fact that the injunction against intercourse with menstruating women of Q 2:222 follows only three verses after the injunction against drinking wine of Q 2:219. Islamic jurists came to the same conclusion when they argued that menstruation and intoxication have identical effects: both render the believer too polluted to pray.29 2

The Drinks of Life

Now that we have analyzed the role of wine within the context of late antique and Qurʾānic notions of purity and defilement, let us consider the immediate textual context of 16:67. This wine verse is one of the 128 verses of the Sura of the Bees (Sūrat al-​Nahl), and is part of a passage about the “drinks of life” (Q

table 3.1 Oppositions between impure and pure in Qurʾānic lā taqrabū … ḥatta … proscriptions

impure 4:43a 4:43b 2:222

28 29

sakrān ritually unclean menstruating

pure to know (what one is saying) to have washed oneself to be ritually pure

My translations. Reinhart 1990; Lowry 2001–​6. Cf. Leviticus 10:9.

62 Weststeijn 16:65–​69) beginning right after the exact middle of this sura.30 The verse preceding this passage concerns the sending down of the Qurʾān (16:64), while the verse following it concerns the growth and decay of man (16:70). (16:64) We have not sent down to you the Book except that you may make clear to them that about which they differ, and as a guidance and a mercy for a people who believe. (16:65) And God has sent down water from heaven and thereby given life to the earth after its death; most surely there is a sign in this for a people who would listen. (16:66) And most surely there is a lesson for you in the cattle; We give you to drink of what is in their bellies—​from between faeces and blood—​ pure milk, sweet to drinkers. (16:67) And from the fruits of the palms and the vines you take intoxicant and good provision; most surely there is a sign in this for a people who ponder. (16:68) And your Lord revealed to the bees saying: Make hives in the mountains and in the trees and in the trellises: (16:69) Then eat of all the fruits and walk in the ways of your Lord submissively. There comes forth from their bellies a drink of many colors, in which there is healing for men; most surely there is a sign in this for a people who reflect. (16:70) And God has created you, then He causes you to die, and some of you will be brought back to the vilest part of life, so that after having knowledge you do not know anything; surely God is knowing, powerful.31 The passage on the drinks of life starts with verse 16:65 about water. Verse 16:66 is about milk, and verses 16:68–​69 are about honey. Verse 16:66 also mentions another drink of life, blood. Although blood has a positive value in Christianity and the pagan religions, it is considered highly polluting in Judaism and Islam to the extent that its consumption is strictly forbidden. Classicist J. Andrew Foster and Arabist Kathryn Kueny analyze this passage in their study of the role of bees in Sura 16. They note that a repetition of sounds, words, and phrases ties together this section on the drinks of life, and 30 31

On water, milk, blood, and wine as “life-​fluids” see Drower 1956, 9–​11. On the evaluation of information placed in the middle of a sura see Cuypers 2011; Ernst 2011. Translation adapted from Shakir 1999.

Wine and Impurity in the Sura of the Bees


that—​at least concerning honey and bees—​it contains numerous echoes of classical and Christian thinking.32 Like wine, the other drinks of life were much more than just thirst quenchers in the late antique Near East. Compare the biblical land that “oozes with milk and honey,”33 and the rivers of water, milk, wine, and honey that flow through the Qurʾānic paradise (Q 47:15).34 The drinks of life had sacred qualities, related to pairs of opposites: pure and impure, halal and haram, allowed and forbidden, life and death.35 The passage in the Sura of the Bees deals with the provenance of the drinks of life: water is sent down by God from heaven to earth as rain, milk is formed in the intestines of cattle, wine is the result of the putrefaction of fresh fruit, and honey comes from the “bellies of bees.”36 In addition to their provenance, the passage also mentions the effects of these drinks: water revives that which is dead, wine intoxicates, honey heals.37 In each case, the information on the origin and effect of these drinks is accompanied by the statement that it contains a sign (āya) or a lesson for the audience (See Table 3.3). In 16:66, the verse about milk that immediately precedes the verse on wine, an obvious opposition emerges between pure and impure, and this opposition relates to the provenance of this vital liquid: From the intestines of cattle, between faeces and blood, two highly polluting substances whose consumption is strictly prohibited, comes a drink that is pure and tasteful.38 The Qurʾān uses a pithy ring structure to express this opposition: min bayna farthin wa-​ damin labanan khāliṣan sāʾighan, “from between shit and blood milk good and sweet.” This verse on milk starts with the words: “And most surely there is a lesson for you in the cattle.” Not only in the existence of milk, but especially in the paradoxical provenance of this drink of life there is a lesson for the people. (16:66) And most surely there is a lesson for you in the cattle; We give you to drink of what is in their bellies—​from between faeces and blood—​ pure milk, sweet to drinkers. 32 Foster and Kueny 2007, especially 145–​46, 152–​53. 33 For example, Exodus 3:8. 34 In Odyssey 11.27–​28 Odysseus pours a libation of milk, honey, wine, and water. 35 Deuteronomy 32:1–​14; Song of Songs 5:1; Joel 3:18; Ransome 1937; Drower 1956; Tapper 1994; Behr 2006; Gutsfeld 2006b; Schmoll 2006; Foster and Kueny 2007. 36 See also Quṭb 1967, 14:79. 37 On the medicinal qualities of honey see Waines 2001–​6; Gutsfeld 2006b; Foster and Kueny 2007, 156–​58. 38 Francesca 2012. For khāliṣ, “pure,” as the opposite of muḥarram, “forbidden, impure,” see also Q 6:139.

64 Weststeijn When we look closer we find that the other verses of this passage also deal with an opposition between pure and impure regarding either the provenance or the effect of the drink in question. Q 16:65 says that God sends down water from heaven to earth—​just as He sent down the Qurʾān (16:64).39 Rainwater and the Book are sent from a pure to an impure location, and make what is impure pure. The Book “makes clear,” which is a purifying action,40 and water brings that which is dead back to life.41 Elsewhere, the Qurʾān explicitly states that the effect of rainwater is purifying and that reviving the earth when it is dead constitutes a cleansing of that which is polluted. Water sent down from heaven removes the defilement (rijz) caused by Satan. (25:48) We sent down from heaven pure water (49) to revive a dead land. Wa-​anzalnā min al-​samāʾ māʾan ṭahūran li-​nuḥyiya bihi baldatan maytan (8:11) He sends down on you from heaven water to purify you and to remove from you the defilement of Satan.42 Wa-​yunazzilu ʿalaykum min al-​samāʾ māʾan li-​yuṭahhirakum bihi wa-​ yudhhiba ʿankum rijz al-​shayṭān In Q 16, the fact that water comes from a pure to an impure location and makes that which is impure pure constitutes “a sign for those who listen.” (16:65) And God has sent down water from heaven and thereby given life to the earth after its death; most surely there is a sign in this for a people who would listen. Wa-​llāh anzala min al-​samāʾ māʾan fa-​aḥyā bihi al-​arḍ baʿd mawtihā inna fī dhālika la-​āyatan li-​qawm yasmaʿūn Verses 16:68–​69, about honey, also deal with an opposition between pure and impure. The bee, which has given its name to this sura, is considered impure in

39 40 41


Cf. Quṭb 1967, 14:76. Lowry 2001–​6. Cf. Deuteronomy 32:1–​14, which compares rainwater that falls on the earth to teaching and words. Note that just as God sends the Qurʾān to make clear to people that about which they differ, Satan uses wine to divide people (Q 5:91). And then God sends rainwater to cleanse them from Satan’s defilement (Q 8:11). My translations.

Wine and Impurity in the Sura of the Bees


both Judaism and Zoroastrianism.43 Deuteronomy classifies all winged insects as unclean.44 In Rabbinic Judaism—​just as in Classical Antiquity and early Christianity—​it was believed that bees originate in the putrefying carcasses of dead oxen.45 In general, what originates from an impure location or from an impure animal is itself impure. The Mishnah states: “That which comes from something unkosher is itself unkosher, and that which comes from something kosher is itself kosher.”46 Honey, however, is considered kosher, even though it comes from the impure bee. The rabbis reasoned that honey is not produced by the bee but rather originates in the flower.47 Muhammad is reported to have said that it is forbidden to kill bees, along with ants, hoopoes, and shrikes.48 Killing the last two animals is unlawful because their meat is haram, i.e., impure, forbidden for human consumption,49 which suggests that the bee was once considered impure in Islam as well.50 Indeed, Muslim exegetes argued that in paradise there is honey that is perfectly pure because it does not come from bees.51 Thus, in verses 16:68–​69 about the origin of earthly honey we find an opposition between pure and impure regarding the provenance of this drink of life, just as with the vital liquids water and milk. From the intestines of an impure animal comes a pure, healing beverage. In this opposition between pure and impure there “most surely is a sign for a people who reflect.” (16:68) And your Lord revealed to the bees saying: Make hives in the mountains and in the trees and in the trellises: (16:69) Then eat of all the fruits and walk in the ways of your Lord submissively. There comes forth from their bellies a drink of many colors, in which there is healing for men; most surely there is a sign in this for a people who reflect. 43 44 45 46 47 48 49 50 51

Bechorot 7b; Choksy 1989, 14–​15; Kueny 2001, 12–​13. Deuteronomy 14:19. Leviticus 11:20–​23 makes an exception for locusts, crickets, and grasshoppers. Philo of Alexandria, De specialibus legibus (On the Special Laws), 1.291–​92; Foster and Kueny 2007, 153. More sources quoted in Ransome 1937, 112–​18, 153–​54, 205–​6, 277. Bechorot 1:2. Frymer and Rabinowitz 2007. Cf. Foster and Kueny 2007, 151. Juynboll 2007, 26; Foster and Kueny 2007, 166n77. Cf. Leviticus 11:13–​19; Deuteronomy 14:11–​18. Ibn Manzūr, Lisān al-​ʿArab, s.v. ṣurad, as quoted by Juynboll 2007. The consumption of creeping things that fly is prohibited in Islamic law, see Francesca 2012. al-​Zamakhsharī, al-​Kashshāf, ad Q 47:15 (= Gätje 1976, 181); Stetkevych 1997, 215; Waines 2001–​6.

66 Weststeijn Yakhruju min buṭūnihā sharāb mukhtalif alwānuhu fihi shifāʾ lil-​nās inna fī dhālika la-​āyatan li-​qawm yatafakkirūn After this passage about the drinks of life follows a verse that deals once again with pure/​impure oppositions. Verse 16:70 tells that what God has made pure He later makes impure: Whom He has granted life He then lets die, whom He has let grow into maturity He then returns to decrepit old age, whom He has made wise He then renders foolish. That the senility of old age is seen as ­impure is shown by the use of the word ardhal, “vilest,” in the expression ar­ dhal al-​ʿumur, “vilest part of life.” (16:70) And God has created you, then He causes you to die, and some of you will be brought back to the vilest part of life, so that after having knowledge you do not know anything; surely God is knowing, powerful. Wa-​llāh khalaqakum thumma yatawaffakum wa-​minkum man yuraddu ilā ardhal al-​ʿumur li-​kay lā yaʿlama baʿd ʿilm shayʾan inna Allāh ʿalīm qadīr 3

The Opposition between Pure and Impure in Wine Verse 16:67

Let us now analyze wine verse 16:67 within the context of this passage on the drinks of life. In their study of Q 16, Foster and Kueny give two somewhat conflicting interpretations of the role of wine in this passage, as if wavering between the positive and the negative reading of 16:67. At one point in their analysis, they approach the passage on the drinks of life as a unit about three drinks: milk, wine, and honey; a verse about rain precedes this unit.52 They seem to argue that because the verses about milk and honey are about good, pure drinks from an impure origin (cattle or bee intestines), the verse in between about wine must also be about a good, pure drink from an impure origin. Here Foster and Kueny apparently follow the common opinion that verse 16:67 equates sakar, “intoxicant,” with rizq ḥasan, “good provision,” and that this verse praises wine as one of God’s gifts to mankind.53

52 53

Foster and Kueny 2007, 152–​53. At the same time, they note that verses 16:66–​69 deal with the concepts of revelation (the main theme of 16:64) and generation and decay (the main theme of 16:70). Cf. Kueny 2001, 10–​13; Kueny 2001–​6.

Wine and Impurity in the Sura of the Bees


The three drinks, then, when clustered together into a single rhetorical trope, demonstrate God’s supreme ability to transform one substance into another, and perhaps more importantly, his ability to generate something useful and good from what is fleshly, putrid and rotten. … The quranic passages glorify these drinks as signs of God’s ability to transform substances into forms that are beneficial to humans.54 However, in the introduction to their study, Foster and Kueny approach the passage on the drinks of life as a unit about four drinks: water, milk, wine, and honey. Here they note that 16:67 contains an opposition between the two things made from fruit, wine and “a succulent juice,” and they compare the negative, stupefying, polluting effects of wine with the negative, polluting effects of death: “Like the parched, dead earth, the wine of dates and grapes can debilitate. … Yet also, … dates and grapes [produce] a succulent juice.”55 Here they apparently lean toward the negative reading of 16:67 which argues that there is an opposition between sakar “wine” and rizq ḥasan, interpreted by them as “succulent juice,” and that wine is an impure drink with polluting effects. Foster and Kueny’s first reading that the verse about wine must be about a pure drink from an impure origin—​because the surrounding verses about milk and honey are also about pure drinks from impure origins—​loses its strength when we consider the other verses of this passage: the verse about the very first drink of life, water, as well as the flanking verses about the Qurʾān and about man. Indeed, as we have seen above, not all verses in this passage deal with the creation of something pure from an impure origin. To the contrary, the Book (16:64) and water (16:65) do not come from an impure but from a pure origin. Nor do all verses in this passage deal with purifying effects. The verses about the Book, water, and honey (16:68–​69) indeed deal with purifying effects: the Book makes clear, water revives the dead earth, honey cures the sick. However, verse 16:70 about man deals with polluting effects: God turns mature wise men into decrepit old fools. What all of these verses have in common, however, is that they deal with an opposition between pure and impure (See Tables 3.2 and 3.3).56 Because all these verses deal with oppositions between pure and impure it is very probable that wine verse 16:67 also deals with such a pure/​impure opposition. This can only be the opposition between sakar and rizq ḥasan, and 54 55 56

Foster and Kueny 2007, 153. Foster and Kueny 2007, 145. Cf. Boisliveau 2014.

68 Weststeijn the qualifier ḥasan (good) leaves no doubt that rizq ḥasan is the pure element and sakar the impure. Thus, the reading of sakar wa-​rizq ḥasan as a pair of equals has become untenable. These elements are not equated but contrasted: “wine” is the opposite of “good provision.” It turns out that exegetes like Ibn ʿAbbās, who read sakar wa-​rizq ḥasan as a pair of opposites where sakar is impure and rizq ḥasan is pure, are correct. (16:67) And from the fruits of the palms and the vines you take intoxicant and good provision; most surely there is a sign in this for a people who ponder. Wa-​min thamarāt al-​nakhīl wa-​l-​aʿnāb tattakhidūna minhu sakaran wa-​ rizqan ḥasanan inna fī dhālika la-​āya li-​qawm yaʿqilūn It is unlikely that this opposition between sakar (impure) and rizq ḥasan (pure) concerns two opposite qualities of the same drink, wine. The phrase “from the fruits of the palms and the vines you take X and Y” implies that people use these fruits for two different purposes: not only do they process these fruits into an intoxicating drink, but also they eat these fruits to nourish themselves. The fact that the term sakar has as its primary meaning “wine” also makes it probable that sakar wa-​rizq ḥasan is about wine and something else opposed to wine. In addition, the verses about the other vital liquids water, milk, and honey are not about the opposite qualities of a drink but rather the opposition between pure and impure related to the provenance of that drink. If 16:67 is also about an opposition regarding the provenance of wine, it can only be the opposition between an impure drink (sakar) from a pure origin. Out of something pure (fresh, permitted fruit that provides excellent nourishment), man processes, by way of putrefaction, something impure: an intoxicating drink. As with honey and bees, this opposition contradicts the law that that which derives from something impure is itself impure, while that which derives from something pure is itself pure. It is again the contradiction between pure and impure related to the provenance of the vital liquid wine in which there “most surely is a sign for a people who ponder.” Foster and Kueny’s first reading, that 16:67 concerns a pure drink from an impure origin, is therefore incorrect, while their other reading, that 16:67 contrasts polluting wine with pure succulent juice, is correct. Wine is not a pure drink from an impure origin, but, to the contrary, an impure drink from a pure origin.

Wine and Impurity in the Sura of the Bees


table 3.2 Oppositions between pure and impure in Q 16:65–​70



heaven life milk fruit honey growth mature wise

earth death blood wine bee decay senile foolish

During the creation of these four vital liquids, water, milk, wine, and honey, something miraculous happens that appears to violate the laws of nature. Again and again, these miraculous origins contain a sign of God’s omnipotence to make the impossible possible. Although in everyday life there is no contact between heaven and earth, God sends down water from heaven to earth, like He sent down the Qurʾān. With this water God performs another miracle that brings to mind the Day of Resurrection: He revives that which is dead. Contrary to the law that that which originates from something impure is itself impure, clear milk originates from an impure location, between faeces and blood, and medicinal honey comes from the belly of an impure animal. Contrary to the law that that which originates from something pure is itself pure, reprehensible intoxicating wine is made from fresh, nourishing fruit. 4


Medieval Muslim exegetes tend to analyze Qurʾānic verses isolated from their immediate textual context. Despite the fact that Islamic Law condemns wine as impure, some medieval exegetes, as well as many Western scholars, argue that in 16:67 sakar wa-​rizq ḥasan, “wine and good provision,” is a pair of equals; that this verse praises wine as one of God’s signs to man; and is therefore positive about wine. To explain the contradiction with the prohibition of Islamic Law and with the other Qurʾānic verses that are negative about wine, these exegetes argue that 16:67 was revealed earlier than the negative verses, which abrogated its positive message. However, there has always been a number of

70 Weststeijn table 3.3 Structure of the passage on the drinks of life in the Sura of the Bees



16:66 16:67

Qurʾān From heaven (pure) to earth (impure) Makes clear (pure) that on which people are divided (impure) water From heaven (pure) to earth (impure) Brings that which is dead (impure) back to life (pure) milk From between faeces and blood (impure) a sweet drink (pure) wine

From fresh fruit (pure) man produces an intoxicating drink (impure)

16:68–​69 honey

From the bee (impure) a curative drink (pure)


From birth (pure) to death (impure); from wise (pure) to foolish (impure)


“As a guidance and a mercy for a people who believe.” “Most surely there is a sign in this for a people who would listen.” “And most surely there is a lesson for you in the cattle.” “Most surely there is a sign in this for a people who ponder.” “Most surely there is a sign in this for a people who reflect.” “Surely God is knowing, powerful.”

Muslim exegetes who argue that 16:67 contains an opposition between impure sakar and pure rizq ḥasan and is therefore negative about wine.57 When we analyze 16:67 “And from the fruits of the palms and the vines you take intoxicant and good provision” against the background of a) the ideas about purity and impurity as found in the late antique Near East and in the Qurʾān and b) the immediate textual context in which it is found, three arguments emerge that lead to the conclusion that 16:67 is indeed negative about wine:

1) All other Qurʾānic verses on drinking earthly wine are negative about this beverage and label it as something impure or polluting; 2) Verse 16:67 refers to wine with the term sakar, stressing its negative, polluting qualities; 3) 16:67 is part of a passage about oppositions between pure and impure.


See note 6.

Wine and Impurity in the Sura of the Bees


Therefore, it appears that exegetes who read sakar wa-​rizq ḥasan, “wine and good provision,” as an opposition between something impure and something pure, and therefore conclude that 16:67 is negative about wine, are correct. When interpreted in its immediate textual context and within the worldview of Late Antiquity, 16:67 does not contradict the other verses on wine or the ban of Islamic Law: the Qurʾān is internally consistent on this topic. To make sense of 16:67, we do not need the theory of abrogation.


EQ = McAuliffe, Jane, ed. 2001–​6. Encyclopaedia of the Qurʾān. Leiden: Brill. Albenda, Pauline. 1974. “Grapevines in Ashurbanipal’s Garden.” Bulletin of the American Schools of Oriental Research 215:5–​17. ʿAlī, Muḥammad. 1950. The Religion of Islām: A Comprehensive Discussion of the Sources, Principles and Practices of Islām. Lahore: Aḥmadiyyah Anjuman Ishāʿat Islām. Asad, Muhammad. 2008. The Message of the Qurʾān. Bitton: Book Foundation. Behr, Benita von. 2006. “Water.” In The Brill Dictionary of Religion, edited by Kocku von Stuckrad. Leiden: Brill. Berenbaum, Michael, and Fred Skolnik, eds. 2007. Encyclopaedia Judaica, 2nd ed. Detroit: Macmillan. Boisliveau, Anne-​Sylvie. 2014. “Self-​referentiality in the Qur’anic Text: ‘Binarity’ as a Rhetorical Tool.” Al-​Bayān: Journal of Qurʾān and Ḥadīth Studies 12:55–​74. Choksy, Jamsheed K. 1989. Purity and Pollution in Zoroastrianism: Triumph over Evil. Austin: University of Texas Press. Cuypers, Michel. 2011. “Semitic Rhetoric as a Key to the Question of the ‘Naẓm’ of the Qur’anic Text.” Journal of Qur’anic Studies 13:1–​14. Douglas, Mary. 1966. Purity and Danger: An Analysis of Concepts of Pollution and Taboo. London: Routledge. Drower, E. S. 1956. Water into Wine: A Study of Ritual Idiom in the Middle East. London: Murray. Ernst, Carl W. 2011. How to Read the Qur’an: A New Guide, with Select Translations. Edinburgh: Edinburgh University Press. Foster, J. Andrew, and Kathryn M. Kueny. 2007. “From the Bodies of Bees: Classical and Christian Echoes in Surat al-​Nahl.” Comparative Islamic Studies 3:145–​68. Francesca, Ersilia. 2012. “Dietary Law.” In Encyclopaedia of Islam, THREE, edited by Kate Fleet, Gudrun Krämer, Denis Matringe, John Nawas, and Everett Rowson. Leiden: Brill. Freidenreich, David M. 2010. “Holiness and Impurity in the Torah and the Quran: Differences within a Common Typology.” Comparative Islamic Studies 6:5–​22.

72 Weststeijn Frymer, Tivka S., and Louis Isaac Rabinowitz. 2007. “Honey.” In Encyclopaedia Judaica, 2nd ed., vol. 9, edited by Michael Berenbaum and Fred Skolnik, 517–​18. Detroit: Macmillan. Gätje, Helmut. 1976. The Qurʾān and its Exegesis: Selected Texts with Classical and Modern Muslim Interpretations. Translated by Alford T. Welch. Berkeley: University of California Press. Gutsfeld, Andreas. 2006a. “Wine: The Cultic Use of Wine.” In Brill’s New Pauly, edited by Hubert Cancik and Helmuth Schneider. Leiden: Brill. Gutsfeld, Andreas. 2006b. “Honey: Greece and Italy.” In Brill’s New Pauly, edited by Hubert Cancik and Helmuth Schneider. Leiden: Brill. Hallaq, Wael. 2001–​6. “Law and the Qurʾān.” In eq. Hoyland, Robert. 2012. “Early Islam as a Late Antique Religion.” In The Oxford Handbook of Late Antiquity, edited by Scott Fitzgerald Johnson, 1053–​77. Oxford: Oxford University Press. Kueny, Kathryn. 2001. The Rhetoric of Sobriety: Wine in Early Islam. Albany: suny. Kueny, Kathryn. 2001–​6. “Wine.” In eq. Juynboll, G. H. A. 2007. Encyclopedia of Canonical Ḥadīth. Leiden: Brill. Lane, Edward William. 1863–​93. An Arabic-​English Lexicon. London: Williams and Norgate. Lowry, Joseph E. 2001–​6. “Ritual Purity.” In eq. Mawdūdī, Abul AʿLā. 2006. Towards Understanding the Qurʾān. Translated and edited by Zafar Ishaq Ansari. Leicester: Islamic Foundation. McAuliffe, Jane D. 1984. “The Wines of Earth and Paradise: Qurʾānic Proscriptions and Promises.” In Logos Islamikos: Studia Islamica in Honorem Georgii Michaelis Wickens, edited by Roger M. Savory and Dionisius A. Agius, 159–​74. Toronto: Pontifical Institute of Mediaeval Studies. Parker, Robert. 1983. Miasma: Pollution and Purification in Early Greek Religion. Oxford: Clarendon. Quṭb, Sayyid. 1967. Fī ẓilāl al-​Qurʾān. Beirut: Dār al-​Turāth al-​ʿArabī. Ransome, Hilda M. 1937. The Sacred Bee in Ancient Times and Folklore. London: Allen & Unwin. Reinhart, A. Kevin. 1990. “Impurity/​No Danger.” History of Religions 30:1–​24. Reynolds, Gabriel Said. 2011. “Le problème de la chronologie du Coran.” Arabica 58:477–​502. Schmoll, Friedemann. 2006. “Drinking/​Drink.” In The Brill Dictionary of Religion, edited by Kocku von Stuckrad. Leiden: Brill. Shakir, M. H. 1999. The Holy Quran Translated. New York: Tahrike Tarsile Quran. Stetkevych, Suzanne Pinckney. 1993. The Mute Immortals Speak: Pre-​Islamic Poetry and the Poetics of Ritual. Ithaca: Cornell University Press.

Wine and Impurity in the Sura of the Bees


Stetkevych, Suzanne Pinckney. 1997. “Intoxication and Immortality: Wine and Associated Imagery in al-​ Maʿarrī’s Garden.” In Homoeroticism in Classical Arabic Literature, edited by J. W. Wright Jr. and Everett K. Rowson, 210–​ 32. New York: Columbia University Press. Tapper, Richard. 1994. “Blood, Wine and Water: Social and Symbolic Aspects of Drinks and Drinking in the Islamic Middle East.” In Culinary Cultures of the Middle East, edited by Sami Zubaida and Richard Tapper, 215–​31. London: Tauris. Tengberg, M. 2012. “Beginnings and Early History of Date Palm Garden Cultivation in the Middle East.” Journal of Arid Environments 86:139–​47. Thomas, Günther. 2006. “Heaven/​Sky.” In The Brill Dictionary of Religion, edited by Kocku von Stuckrad. Leiden: Brill. Waines, David. 2001–​6. “Honey.” In eq. Wensinck, A. J. 1960–​2007. “Khamr: Juridical Aspects.” In Encyclopaedia of Islam, 2nd ed., edited by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs. Leiden: Brill.

­c hapter 4

Historical-​Critical Research of the Sīra of the Prophet Muhammad: What Do We Stand to Gain? Harald Motzki † 1

The State of Affairs

In the twentieth century, Western scholarship has produced a great number of books and studies on the life of the Prophet Muhammad. Continuing that trend, the start of the twenty-​first century has already seen a few biographies of the Prophet, such as Hans Jansen’s De Historische Mohammed in two parts (published in 2005 and 2007), Vite antiche di Maometto by Michael Lecker (2007), Tilman Nagel’s monumental study Mohammed: Leben und Legende (2008), which is over 1,000 pages long, and from the same year Die ältesten Berichte über das Leben Muḥammads by Andreas Görke and Gregor Schoeler.1 These books can be divided into three groups. One group includes biographies that summarize and paraphrase the Islamic sources about the life of the Prophet and present it as a sort of historical narrative. These authors suggest that Muhammad’s life more or less followed the accounts in the Islamic sources. This category includes not only books by Muslim authors, such as the famous Muḥammad: His Life Based on the Earliest Sources, by Martin Lings, alias Abū Bakr Sirāj al-​Dīn,2 but also many works by non-​Muslim scholars, such as the biographies written by Frants Buhl, Montgomery Watt, and Maxime Rodinson.3 A second group of these books also recounts the Islamic sources, but occasionally adds critical or skeptical remarks without a thorough critical analysis of the sources. This group includes, for example, works about Muhammad published by Michael Cook and Hans Jansen.4 These authors are referred to as

1 Jansen 2005; idem 2007; Lecker 2007; Nagel 2008; Görke and Schoeler 2008. In addition, two brief German studies require mention: Bobzin 2000; Schöller 2008. 2 Lings 1991. 3 Buhl 1961; Montgomery Watt 1953; idem 1956; Rodinson 1967. Nagel 2008 also belongs to this group; his approach, however, distinguishes itself from that of his predecessors by using a much broader source base. 4 Cook 1983; Jansen 2005; idem 2007.

© Harald Motzki, 2022 | DOI:10.1163/9789004500648_005

Historical-​Critical Research of the Sīra of the Prophet Muhammad


“source-​skeptics.” A third group of books focuses on historical-​critical assessment of the sources available on the life of the Prophet, without the pretense of delivering a biography. These studies include, for example, Gregor Schoeler’s book Charakter und Authentie der muslimischen Überlieferung über das Leben Mohammeds and the collection I published, The Biography of Muḥammad: The Issue of the Sources.5 Fierce debates have taken place between some of the authors from the first two groups. Source-​skeptics among the Islamicists, such as Leone Caetani, Henri Lammens, and Ignaz Goldziher (early twentieth century ce) and Joseph Schacht, Régis Blachère, John Wansbrough, Patricia Crone, and Michael Cook (second half of the twentieth century ce), usually considered Islamic traditions about the Prophet as apocryphal and unreliable. The stories of the Sīra, which deal with the Prophet’s life, would then be a form of “Islamic salvation history,” without any historical value regarding the time of the Prophet itself. Some researchers, including Wansbrough, have even argued that the Qurʾān did not come into existence until two centuries after Muhammad’s death and that there is no historical connection between the text and the Prophet. The authors mentioned here faulted their colleagues in the first group for using the Islamic sources in a naive and uncritical way. Those scholars who had been branded as uncritical, by contrast, accused their counterparts among the source-​skeptics of hyper-​skepticism. In their view, the source-​skeptics could not prove that their own approach posed a genuine methodological improvement. Thus, in the twentieth century ce, research on the life of the Prophet Muhammad and the early days of the Islamic community reached a stalemate. It is this stalemate that authors from the third group are trying to resolve. I consider myself a part of this group. I have been researching the sources of early Islam for twenty-​five years. I have utilized several approaches to ending the stalemate. First, I critically review existing scholarship, focusing primarily on its premises and methods. Second, I have broadened the corpus of the sources for this research, which is now possible because in the last four decades many previously inaccessible sources have become available. Third, I have developed new methods to date the sources and the traditions they contain and to test their historical reliability. I have applied these methods to the different genres of traditions: legally relevant traditions (aḥādīth), maghāzī narratives, and tafsīr traditions.6 5 Schoeler 1996; Motzki 2000b. 6 Aḥādīth are the traditions about sayings or actions of the Prophet Muhammad. Maghāzī are the battles and conquests of the Muslims during and after the time of the Prophet. Tafsīr is defined as the exegesis of the Qurʾān.

76 Motzki I describe my method of research as historical-​critical. Many other authors have used the same term for their research without accounting for what exactly their method entails. The historical-​critical method has been developed most elaborately in the field of biblical scholarship, creating a standard to which other scholars should also be held.7 As the name suggests, this method has two characteristics: historical and critical. Historical means that, for example, a biblical story is not a text that fell from the sky or was revealed to people by God in the literal sense, but rather it is a text that grew over the course of centuries, and therefore has a history. Scholars attempt to reconstruct the story’s history to determine when and where it originated, its original historical context and intention, and how the story changed over time. Such a reconstruction is possible if variants of the story are available. This method is called critical because it uses generally accepted scientific criteria in the analysis of the text. These criteria include, on the one hand, textual criticism, which is done mainly by comparing the variants available to us. The aim is to reconstruct the original text and to identify subsequent additions and changes. On the other hand, the chains of authorities (asānīd, singular isnād) that precede the main text of Islamic traditions must be analyzed. The combination of textual criticism and the analysis of isnād often makes it possible to reconstruct the original text and to find out who distributed it. Another important criterion of the historical-​critical method is that a scholar’s statements must be verifiable by other experts. Critical therefore does not mean that the purpose of this method is to criticize the text and to question its content. However, such criticism may occur as a consequence of applying this method. 2

A Story of the Sīra

In the following I will use an example to demonstrate how I apply this method to sīra traditions8 and to illustrate the merits (and drawbacks) of the method. According to an Islamic tradition, the leaders of the Quraysh had serious doubts about their fellow tribesman Muhammad, who made public appearances in Mecca claiming to be a prophet of God. They believed he was not a prophet, and tried to label him a fortuneteller, madman, poet, and magician. Many modern books cite this tradition about the life of the Prophet, which

7 Müller 2010. 8 Muslims use the word sīra to indicate the biography of the Prophet Muhammad. The correct Arabic term is al-​sīra al-​nabawiyya (the biography of the Prophet).

Historical-​Critical Research of the Sīra of the Prophet Muhammad


can be found in the Sīra of Ibn Hishām, the oldest preserved biography of the Prophet (held in high regard among Muslims).9 A summary: As the Hajj season drew near, several prominent opponents of the Prophet approached al-​Walīd ibn al-​Mughīra, one of the elders of the Quraysh, to consult him about Muhammad. The purpose of the conversation was to come to an agreement on the following question: What image of Muhammad should we present to the delegations of the Arab tribes who come to Mecca for the Hajj, so that he will not make an impression on them? Various options were considered: Muhammad is a fortuneteller (kāhin), a possessed person (majnūn), a poet (shāʿir), or a sorcerer (sāḥir). Even though al-​Walīd admitted that none of these epithets really applied to Muhammad, he chose sorcerer, as Muhammad’s message sowed discord between family and clan members. The opponents of the Prophet followed this advice and went to the access roads to Mecca to warn the approaching Arabs about Muhammad. Following this event, so the story continues, God revealed to Muhammad Q 74 (Sūrat al-​Mudaththir):11–​25 about al-​Walīd and Q 15 (Sūrat al-​Ḥijr):90–​93 about the group of Qurayshis with whom the latter had had the discussion about Muhammad. Is this a report that qualifies as “historical”?10 Many Muslims believe that this ancient biography of the Prophet is a collection of reliable traditions. The non-​ Muslim historian of Islam will question it. How can we be certain this story is reliable? The Qurʾān tells us that the Meccans who opposed the Prophet have attacked him with these kinds of claims. In addition to Q 74:24–​25, which are mentioned in the story itself, Q 52 (Sūrat al-​Ṭūr):29–​30 serve as evidence. Q 74:24–​25 states: (74:24) He said, “This is naught but a trumped-​up sorcery (siḥr); (25) this is nothing but mortal speech.”11


10 11

Several editions of the Arabic text are available. I rely on Wüstenfeld (1860, 1:171–​72) and the more recent edition by al-​Saqqā (1955, 1:270–​72). For an English translation, see Guillaume 1967, 121–​22. Nagel 2008, 200 summarizes the story and considers it a reliable historical message. Of course, when referring to Qurʾānic verses, the word “historical” only means that according to Islamic tradition Muhammad proclaimed them as divine revelations. Note from the translator: the English translations of verses from the Qurʾān are adapted from Arberry’s translation. Transliterations were added to match their placement by the author in the original Dutch version of this chapter.

78 Motzki “He” in “he said” clearly refers to an unbelieving opponent of the Prophet; “this” in “this is naught but a trumped-​up sorcery” refers to Muhammad’s revelation. In Q 52, verses 29 and 30 read: (52:29) Therefore remind! by thy Lord’s blessing thou art not a soothsayer (kāhin) neither possessed (majnūn). (30) Or do they say, “He is a poet (shāʿir) for whom we await Fate’s uncertainty?” “They” in “Or do they say” refers to Meccan opponents of the Prophet. The key words of the story are present within these verses. The story itself claims that the revelation of the cited verses from Q 74 and Q 15 followed the event they describe. Muslim scholars usually view agreement between tradition and the Qurʾān as an indication that the tradition in question is reliable. The critical historian, however, wonders whether such a conclusion can be universally accepted. Is it not possible that the story was fabricated on the basis of the Qurʾānic verses? Can we verify that Mohammed actually came up with these Qurʾānic verses in the wake of the described events? To answer these questions, we must examine the story while keeping a critical eye on the sources. It is best to start with the oldest existing source preserving the story: al-​Sīra al-​nabawiyya (The life of the Prophet), composed by the aforementioned Egyptian scholar ʿAbd al-​Malik ibn Hishām (d. 218 ah/​833 ce). To a critical historian, it is not self-​evident to view Ibn Hishām’s work as a historically reliable source on the life of the Prophet: a gap of two centuries exists between the lives of the author and the Prophet Muhammad. How could Ibn Hishām have acquired reliable information about the Prophet’s life so long after the latter’s death? Here we encounter a problem that occurs in all knowledge areas of early Islam. All information about the early period can only be found in sources written one and a half centuries or more after the time on which they report. Therefore, some Western scholars have concluded that the information in these sources cannot be historically reliable. First, because a great amount of time passed before they were written. Second, because the information comes from Muslims and is therefore a priori one-​sided. This information is, after all, religiously biased. I consider this view to be overly skeptical and generalizing. A historian of Islam must ask on a case-​by-​case basis whether it is possible to find out where the information in the sources originates. That is part of the historical-​critical method. Was the information made up by the authors of these sources? Did they perhaps use previous sources that can still be found? Any researcher must take into account the problem of a possible religious bias in the information. However, the possibility of a religious inclination alone

Historical-​Critical Research of the Sīra of the Prophet Muhammad


should not lead to the complete dismissal of Islamic sources as unreliable. Adherents to a religion are capable of providing reliable information about their history. The historian’s task is to determine which information is historically reliable. After thorough investigation, the historian may conclude that it is impossible to reach certainty about the reliability of a data set. Despite such a failure in one or perhaps a few cases, the historian should not conclude that it is impossible, as Patricia Crone and Michael Cook have argued.12 We return to Ibn Hishām and his book about the life of the Prophet. Could it be that Ibn Hishām took his material from older sources and accurately reproduced it? Indeed, Ibn Hishām claims that his work is based on an older source: he often prefaces his stories with qāla Ibn Isḥāq (“Ibn Isḥāq said”). Ibn Isḥāq is known in biographical literature as a scholar from Medina who died in 150 ah/​767 ce; an unpreserved book about the life of the Prophet is attributed to him. Here we must keep in mind that in the early Islamic period, books were not distributed by publishers as they are today. In this early period, the books of scholars could not be bought in a bookstore or consulted in a library. A scholar generally related his knowledge and writings through lectures to his students, who in turn passed it down to their own students in the same manner. Thus, books were mainly distributed orally, although from the second Islamic century on, texts were usually also written down or copied. However, one was not allowed to quote from a book without having heard it from a teacher who himself possessed the authority to pass it down (in the early stages this infraction did sometimes occur).13 So, what about Ibn Hishām? How did he obtain the book of Ibn Isḥāq? Because he died 66 years after Ibn Ishāq, it is virtually impossible that Ibn Hishām received the material directly from Ibn Isḥāq. Indeed, that is not the case. At the beginning of his book, Ibn Hishām claims to have received the work of Ibn Isḥāq through one of his students, Ziyād ibn ʿAbd Allāh al-​Bakkāʾī (d. 183 ah/​799 ce). Is it possible to prove Ibn Hishām’s claims? Actually, it is. Fortunately, Ibn Hishām was not the only one to draw from the lectures and writings of Ibn Isḥāq. Additional fragments of Ibn Isḥāq’s work have been preserved in later sources, such as a book by Aḥmad ibn ʿAbd al-​Jabbār al-​ʿUṭāridī (177 ah/​793 ce–​270 ah/​883 ce).14 Aḥmad ibn ʿAbd al-​Jabbār al-​ʿUṭāridī came two generations after Ibn Hishām. He claims to have received his material from a student of Ibn Isḥāq, Yūnus ibn Bukayr (d. 199 ah/​814 ce), who is known as the author of a work 12 13 14

Crone and Cook 1977, 1. Motzki 1991, 276–​82; idem 2016, 149–​50. Ḥamīd Allāh 1981.

80 Motzki titled Kitāb al-​Maghāzī, about the battles and conquests of early Islam.15 Although only parts of al-​ʿUṭāridī’s book have been preserved, these parts contain many texts that can also be found in the Sīra of Ibn Hishām. It is thus possible to compare two versions of Ibn Isḥāq’s material. Because the narrators identify their sources as students of Ibn Isḥāq, we can determine which material originates from Ibn Isḥāq. That is, regardless of the differences between these versions, commonalities must have come from Ibn Isḥāq. This type of source reconstruction, based on variations in traditions, plays an important role in my research into early Islam.16 The story about the meeting of the Quraysh also occurs in the fragment that has been preserved from al-​ʿUṭāridī’s work. Comparison of his version of the story to Ibn Hishām’s text shows that they are almost identical, apart from a few copying errors and remarks added by Ibn Hishām. It is striking, however, that the two versions differ from each other regarding the isnād. Al-​ʿUṭāridī’s chain of authorities goes “Yūnus ibn Bukayr –​Ibn Isḥāq –​Muḥammad ibn Abī Muḥammad –​Saʿīd ibn Jubayr or17 ʿIkrima –​Ibn ʿAbbās.” Ibn Hishām, however, altogether omits the isnād from this story. How can this difference be explained? There are two possible answers. The isnād could have been added afterwards by al-​ʿUṭāridī or his teacher Yūnus ibn Bukayr. Alternatively, Ibn Hishām could have left out the isnād that originally accompanied Ibn Isḥāq’s story. Which of the two hypotheses is correct? This question can be answered by comparing these texts to other stories in the two works where this specific isnād does or does not occur. The isnād “Ibn Isḥāq –​Muḥammad ibn Abī Muḥammad –​Saʿīd ibn Jubayr or ʿIkrima –​Ibn ʿAbbās” can be found a second time in al-​ʿUṭāridī’s version of Ibn Isḥāq’s Life of the Prophet. Because this isnād does not appear more than those two times in al-​ʿUṭāridī’s fragment of Ibn Isḥāq’s Sīra, it can be concluded that it is not among Ibn Isḥāq’s standard chains of authorities. Moreover, its use is limited to certain topics, in this case a tradition on the change of the qibla, or the direction of prayer. According to the Sīra of Ibn Isḥāq in the tradition of Ibn Hishām, at a certain moment Muslims in Medina had to change their orientation during their daily worship (ṣalawāt, singular: ṣalāt) from Jerusalem to the Kaaba in Mecca.18 15 16 17 18

This work was probably delivered both orally and in writing, as became customary in the second half of the second century ah. See for example Motzki 1991; idem 1996; idem 1998; idem 2000a; idem 2001; idem 2005; idem 2015. This term indicates an uncertainty that is rarely found in Islamic traditions. I will return to this phenomenon later. Wüstenfeld 1860, 1:381–​82; al-​Saqqā, 1955, 1:550; Guillaume 1967, 258–​59.

Historical-​Critical Research of the Sīra of the Prophet Muhammad


As in our above example, Ibn Hishām does not provide an isnād in his story about the change of qibla. However, taken in context, it becomes apparent that the story about the change in the direction of prayer is embedded in a larger pericope about the relationship between Muhammad and the Jews of Medina. In the terminology of historical-​critical research, a pericope is a thematically defined story. The entire pericope of Ibn Hishām begins with the following isnād: “[Ibn Isḥāq said:] A client (mawlā) of Zayd ibn Thābit19 reported to me from ʿIkrima or Saʿīd ibn Jubayr from Ibn ʿAbbās.” This isnād is very similar to that of al-​ʿUṭāridī in his story about the change in the direction of prayer. Particularly striking and quite uncommon is the combination of the names ʿIkrima and Saʿīd ibn Jubayr, connected with the conjunction aw (or). Al-​ ʿUṭāridī’s and Ibn Hishām’s chains of authorities differ only in relation to one person: Ibn Isḥāq’s direct informant. Al-​ʿUṭāridī mentions him as Muḥammad ibn Abī Muḥammad; Ibn Hishām describes him as “a client (mawlā) of Zayd ibn Thābit.” Since the rest of the isnād is identical in both versions, we can presume that they are one and the same person: the mawlā of Zayd ibn Thābit whose name was probably Muḥammad ibn Abī Muḥammad. Comparable chains of authorities in text fragments that go back to other students of Ibn Isḥāq and that are found in later sources confirm this hypothesis.20 We can further conclude that this isnād must go back to Ibn Isḥāq himself. It was cited in the same story by two of Ibn Isḥāq’s students, in different forms,21 and other students of Ibn Isḥāq also mention this isnād in other texts. It is unlikely that al-​ʿUṭāridī or his teacher/​informant Yūnus ibn Bukayr fabricated




In early Islam, non-​Arabs were given the legal status of client (mawlā) to an Arab family (āl), usually because they were captured during the conquests; they were released after their conversion. Non-​Arabs then became mawālī (pl.) of the person who released them, a status inherited by future generations. Zayd ibn Thābit was a civilized Medinan companion of the Prophet Muhammad. He had studied in the Jewish school of Medina. According to some traditions he worked for Muhammad as a secretary and also wrote down verses of the Qurʾān. See Lecker 2002, 476. In a preserved fragmentary version of Ibn Isḥāq’s Maghāzī by Muḥammad ibn Salāma, a student of Ibn Isḥāq, the isnād “mawlā āl [family of] Zayd ibn Thābit –​Saʿīd ibn Jubayr aw [or] ʿIkrima –​Ibn ʿAbbās” is attached to those traditions in which Ibn Hishām does not provide an isnād (edited by M. Ḥamīd Allāh (1981), along with his edition of al-​ʿUṭāridī’s fragment of Ibn Isḥāq’s Maghāzī). In al-​Ṭabarī’s Tafsīr the isnād “Muḥammad ibn Ḥumayd (or Abū Kurayb) –​Salama ibn al-​Faḍl –​Muḥammad ibn Isḥāq –​Muḥammad ibn Abī Muḥammad –​Saʿīd ibn Jubayr aw ʿIkrima (sometimes only rajul [a man]) –​Ibn ʿAbbās” turns up in texts originating from Ibn Isḥāq’s Maghāzī. Al-​Ṭabarī (d. 310 ah/​923 ce) was a famous historian and commentator of the Qurʾān. The slight differences between the texts may be related to the teaching method that was sometimes oral, or to the accuracy of the students when writing or transcribing texts.

82 Motzki this isnād and added it to the story about the meeting of the Quraysh. It is more likely that Ibn Hishām must have omitted this isnād in his version of the story. We see this pattern more often in Ibn Hishām. The author himself refers to his book as merely a mukhtaṣar (excerpt, shortened version) of Ibn Isḥāq’s Maghāzī. It is very likely, then, that Ibn Isḥāq has named this Muḥammad ibn Abī Muḥammad, the client of Zayd ibn Thābit, as his informant of the story of the Quraysh meeting. But is this assessment accurate? Several indications in the books of Ibn Hishām and al-​ʿUṭāridī support this conclusion. For example, in the collections of both Ibn al-​Hishām and al-​ʿUṭāridī of Ibn Isḥāq’s Life of the Prophet, another series of traditions goes back to Muḥammad ibn Abī Muḥammad, the client of Zayd ibn Thābit, although neither author mentions the name of the narrator or his status. The traditions relate to negotiations between the clan leaders of the Quraysh and Muhammad during the early stages of his activity as a prophet. This pericope contains stories about several topics: Abū Jahl’s attempt on the Prophet’s life, a speech by al-​Naḍr ibn al-​Ḥārith addressed to the Quraysh, and a delegation sent by the Quraysh to ask the Jewish rabbis of Medina how to determine whether Muhammad really was a prophet.22 As in the narrative mentioned above about the meeting of the Quraysh, this series of stories is peppered with asbāb al-​nuzūl, references to Qurʾānic verses that, according to these traditions, were revealed in response to the events mentioned above. Two elements indicate that the aforementioned Muḥammad ibn Abī Muḥammad can be regarded as the narrator of these stories: the isnād and the stories’ contents. What do we learn from the isnād? In Ibn Hishām’s compilation, Ibn Isḥāq provides the following isnād for the pericope currently under discussion: A scholar of ḥadīth told me, from Saʿīd ibn Jubayr and [or?] ʿIkrima, Ibn ʿAbbās’s client, from Ibn ʿAbbās (ḥaddathanī baʿḍu ahl al-​ʿilm ʿan Saʿīd ibn Jubayr wa-​[aw?] ʿan ʿIkrima mawlā Ibn ʿAbbās ʿan Ibn ʿAbbās). Al-​ʿUṭāridī’s version: A scholar of the Meccans told me more than forty years ago from ʿIkrima, from Ibn ʿAbbās (ḥaddathanī shaykh min ahl al-​Makka qadīm mundhu biḍʿin wa-​arbaʿīn sanatan ʿan ʿIkrima ʿan Ibn ʿAbbās).


Wüstenfeld 1860, 1:187–​202; al-​Saqqā 1955, 1:294–​314; Guillaume 1967, 133–​41.

Historical-​Critical Research of the Sīra of the Prophet Muhammad


The chains of authorities in Ibn Hishām and al-​ʿUṭāridī are very similar to the asānīd that the two collectors connect to other stories with Muḥammad ibn Abī Muḥammad, the client of Zayd ibn Thābit. Once again, they differ especially in the name of the informant of Ibn Isḥāq. One calls him “a scholar of ḥadīth” (baʿḍu ahl al-​ʿilm), the other “a scholar of the Meccans” (shaykh min ahl al-​Makka).23 These indications point to the conclusion that Ibn Isḥāq attributed the pericope about the negotiations between the leaders of the Quraysh and Muhammad to a scholar who, in similar transmission chains belonging to other stories, is identified as Muḥammad ibn Abī Muḥammad or a client (mawlā) of Zayd ibn Thābit’s family. The contents of one of the stories that are part of this pericope confirm this conclusion: the speech addressed to the Quraysh by the aforementioned al-​Naḍr ibn al-​Ḥārith. This speech paraphrases the story about the meeting of the Quraysh with al-​Walīd, which is the pericope presented at the beginning of this study. This agreement between the two stories both in their contents and in the similarity of their chains of authorities shows that one and the same storyteller speaks here. When we compare all the traditions that Ibn Isḥāq attributes to this Meccan scholar, Muḥammad ibn Abī Muḥammad, the image emerges of a storyteller who is strongly interested in the reasons of Qurʾānic revelation (asbāb al-​ nuzūl). Likewise, the storyteller considers the relations between the Quraysh and the Prophet at the beginning of his activity in Mecca and the relations between the Jews and the Prophet in the first years of his stay in Medina.24 This narrator’s interest in Jewish matters, coupled with the fact that he is a mawlā, a client of the family of Zayd ibn Thābit, suggests that he or his father was originally of Jewish descent.25 This comparison establishes who authored the story about the meeting of the Quraysh and some other stories in Ibn Isḥāq’s book about the life of the Prophet. Thanks to al-​ʿUṭāridī’s narration of Ibn Isḥāq’s remark that the latter had heard this shaykh’s stories more than forty years earlier, these stories together with their isnād can be dated roughly to the end of the first Islamic century. That means that the time between the event and the narration of the story is now a mere century. Is it possible to get even closer to the time

23 24 25

The difference in the informants of this scholar is probably due to poor transmission. Instead of Saʿīd ibn Jubayr wa [and] ʿIkrima in Ibn Hishām and only ʿIkrima in al-​ʿUṭāridī, in most cases this isnād reads “Saʿīd ibn Jubayr aw [or] ʿIkrima.” A larger study on the many traditions that came from Muḥammad ibn Abī Muḥammad was published in 2017, after the present text was completed. See Motzki 2017. His patron, Zayd ibn Thābit, who comes from Medina, is known for his good relations with the Jews.

84 Motzki table 4.1 Transmission chains (asānīd) from the Sīra by Ibn Hishām and al-​ʿUṭāridī



Meeting at al-​Walīd’s Ibn Hishām dwelling about how to al-​ʿUṭāridī portray Muhammad to the tribes who make the pilgrimage Changing of the qibla Ibn Hishām al-​ʿUṭāridī

Negotiations between Quraysh and Muhammad, in the early years of Muhammad’s activity as a propheta

Ibn Hishām


Chain of transmitters No isnād Yūnus ibn Bukayr –​Ibn Isḥāq –​Muḥammad ibn Abī Muḥammad –​Saʿīd ibn Jubayr or ʿIkrima –​Ibn ʿAbbās Ibn Isḥāq –​Muḥammad ibn Abī Muḥammad –​Saʿīd ibn Jubayr or ʿIkrima –​Ibn ʿAbbās [Ibn Isḥāq said:] “a cliënt (mawlā) of Zayd ibn Thābit told me, from ʿIkrima or Saʿīd ibn Jubayr from Ibn ʿAbbās.” “A scholar of ḥadīth told me, from Saʿīd ibn Jubayr and [or?] ʿIkrima, cliënt of Ibn ʿAbbās, from Ibn ʿAbbās.” “A scholar of the Meccans told me more than forty years ago, from ʿIkrima, from Ibn ʿAbbās.”

a  In both versions, this story paraphrases the story about the al-​Walīd meeting.

of the Prophet? Not through the methods of historical-​criticism. None of the available variants of Muḥammad ibn Abī Muḥammad’s stories go back to his informants independently of him; that is, to Saʿīd ibn Jubayr or ʿIkrima. In other words, his story about the meeting of the Quraysh stands alone. Therefore, we cannot make a verifiable statement about the history of the story. Moreover, the isnād that this scholar mentions in his stories does not make a reliable impression, for the following reasons. First, this storyteller lists the same isnād (from Saʿīd ibn Jubayr or from ʿIkrima) for most or all of his stories. Second, it is strange that he no longer remembered whether he heard a story from Saʿīd ibn Jubayr or from ʿIkrima, both well-​known disciples of Ibn ʿAbbās. Third, the large number of mostly short texts that other people received from Saʿīd ibn Jubayr and ʿIkrima increases doubts that the long

Historical-​Critical Research of the Sīra of the Prophet Muhammad


stories Muḥammad ibn Abī Muḥammad attributes to these two people were really narrated by them. Another question related to Muḥammad ibn Abī Muḥammad’s story about the meeting of the Quraysh –​the first text I discussed in this study –​concerns the claim that Q 15:90–​93 was revealed in response to the event he narrates. At first glance, these verses contain nothing that may be connected to his story. Verses 90 through 93 are as follows: (15:90) So We sent it down to the partitioners (muqtasimīn), (91) who have broken the Qurʾān into fragments (ʿiḍīn). (92) Now by thy Lord, We shall surely question them all together (93) concerning that they were doing. This question can be answered by studying the early Qurʾānic exegesis of these verses.26 Indeed, the exegesis connects these verses to the epithets used by Muhammad’s Meccan opponents to label him as a false prophet. The old exegetes explain the phrase “the partitioners, who have broken the Qurʾān into fragments” as the Quraysh rejecting the Qurʾān as “poetry” (shiʿr) or “sorcery” (siḥr). The characterization sāḥir (wizard) is central to Muḥammad ibn Abī Muḥammad’s narrative that was presented at the beginning of this study. The interpretation of the Arabic word ʿiḍīn as “sorcery” instead of “fragments” is tied to ʿIkrima, one of the people Muḥammad ibn Abī Muḥammad cites as a source for his story.27 This connection suggests that he did not make up his story entirely by himself after all, but that it is based, at least in part, on information from the previous generation. Still, the claim that Ibn ʿAbbās is the source of the stories cannot be substantiated.28 If Ibn ʿAbbās cannot be reliably identified as the source of the story about the meeting of the Quraysh, we must ask ourselves where else Muḥammad ibn Abī Muḥammad could have obtained his story. The traditions about the explanation offered by ʿIkrima and others of his generation regarding Q 15:90–​ 91 point to the old Qurʾānic exegesis as the fertile soil out of which this story 26 27 28

Exegesis of the Qurʾān begins in the last quarter of the first Islamic century, the period in which ʿIkrima, Saʿīd ibn Jubayr, and Muḥammad ibn Abī Muḥammad lived. See Motzki 2010, 231–​303. Motzki 2010, 255–​67. Even if we supposed that Ibn ʿAbbās was the source of Muḥammad ibn Abī Muḥammad’s stories, there remains a chronological gap: Ibn ʿAbbās, a cousin of Muhammad, was not yet born at the beginning of Muhammad’s activity as a prophet. Moreover, he was an infant in the first years of Muhammad’s stay in Mecca. According to tradition, he was born three years before the hijra.

86 Motzki grew. Muḥammad ibn Abī Muḥammad himself belongs to this first historically identifiable generation of Qurʾānic exegetes. His stories thus appear to be more exegetical than historical in nature. That is, they try to interpret the Qurʾānic verses by providing them with a historical background. The question of whether the events really happened as he describes cannot be answered by the historical-​critical method. 3


This short analysis of a story about the life of the Prophet Muhammad shows that there is indeed a middle ground between Scylla and Charybdis, between extreme source skepticism and the more or less uncritical acceptance of Islamic traditions. The source-​skeptics view these kinds of stories as historically unreliable a priori, because they come from Muslim sources; source-​skeptics also date the origin of the stories in the second half of the second or even the third Islamic century. Most Muslim and less critical Western authors, by contrast, believe that the stories accurately reflect the facts of the Prophet’s life. My historical-​critical source analysis of the stories about the life of the Prophet, which Ibn Isḥāq narrated from his source Muḥammad ibn Abī Muḥammad, however, shows that the story is much older than the source-​skeptics think. Yet, it remains impossible to substantiate through the source-​critical method that the story goes back to a contemporary of the Prophet—let alone an eyewitness of the events—and reflects historical facts. The value of the story for a historical reconstruction of the life of the Prophet is therefore limited. The story only reflects the image, or one of the images, that Muslims formed during the first century after the Prophet’s death about the difficult phase of his activity in Mecca. Assuming that the Qurʾān, as it was composed in the first century of Islam, is based on Muhammad’s proclamations, we can conclude from this source that the Quraysh labeled Muhammad as a fortuneteller, madman, poet, and sorcerer. The historical context of how these events exactly transpired remains a mystery to us.


The Koran Interpreted. Translated by Arthur J. Arberry. London: Allen & Unwin, 1955. Bobzin, Hartmut. 2000. Mohammed. Munich: Beck. Buhl, Frants. 1961 [1930]. Das Leben Muhammeds. Translated into German by H. Schaeder. Darmstadt: Quelle und Meyer.

Historical-​Critical Research of the Sīra of the Prophet Muhammad


Cook, Michael. 1983. Muhammad. Oxford/​New York: Oxford University Press. Crone, Patricia and Michael Cook. 1977. Hagarism: The Making of the Islamic World. London/​New York: Cambridge University Press. Görke, Andreas, and Gregor Schoeler, eds. 2008. Die ältesten Berichte über das Leben Muḥammads: Das Korpus ʿUrwa ibn az-​Zubair. Princeton, NJ: The Darwin Press. Guillaume, Alfred. 1967 [1955]. The Life of Muhammad: A Translation of Isḥāq’s Sīrat Rasūl Allāh. Karachi: Oxford University Press. Ḥamīd Allāh, Muḥammad. 1981 [1976]. Sīrat Ibn Isḥāq: al-​Musammāt bi-​Kitāb al-​Mubtadaʾ wa-​l-​mabʿath wa-​l-​maghāzī. Konya: Maʿhad al-​Dirāsāt wa-​l-​Abḥāth lil-​Taʿrīb. Jansen, Hans. 2005. De historische Mohammed: De Mekkaanse verhalen. Amsterdam: Arbeiderspers. Jansen, Hans. 2007. De historische Mohammed: De verhalen uit Medina. Amsterdam: Arbeiderspers. Lecker, Michael. 2007. Vite antiche di Maometto: Testi scelti e tradotti da Roberto Tottoli. Milan: Mondadori. Lecker, Michael. 2002. “Zayd b. Thābit.” In Encyclopaedia of Islam, 2nd ed., edited by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs. Leiden: Brill. Lings, Martin. 1991 [1983]. Muhammad: His Life Based on the Earliest Sources. London: Islamic Texts Society. Montgomery Watt, W. 1953. Muhammad at Mecca. Oxford: Oxford University Press. Montgomery Watt, W. 1956. Muhammad at Medina. Oxford: Oxford University Press. Motzki, Harald. 1991. Die Anfänge der Islamischen Jurisprudenz: Ihre Entwicklung in Mekka bis zur Mitte des 2./​8. Jahrhunderts. Stuttgart: Steiner. (English translation 2002. The Origins of Islamic Jurisprudence: Meccan Fiqh before the Classical Schools. Translated by M. H. Katz. Leiden: Brill.). Motzki, Harald. 1996. “Quo vadis Ḥadīṯ-​Forschung? Eine kritische Untersuchung von G. H. A. Juynboll: ‘Nāfiʿ the Mawlā of Ibn ʿUmar, and His Position in Muslim Ḥadīth Literature.’ ” Der Islam 73:40–​80; 193–​231. (English translation 2010. “Whither Ḥadīth Studies?” In Analysing Muslim Traditions: Studies in Legal, Exegetical and Maghāzī Ḥadīth, edited by H. Motzki et al., 47–​124. Leiden: Brill.). Motzki, Harald. 1998. “The Prophet and the Cat: On Dating Mālik’s Muwaṭṭaʾ and Legal Traditions.” Jerusalem Studies in Arabic and Islam 22:18–​83. Motzki, Harald. 2000a. “The Murder of Ibn Abī l-​Ḥuqayq: On the Origin and Reliability of some Maghāzī-​Reports.” In The Biography of Muḥammad: The Issue of the Sources, edited by H. Motzki, 170–​239. Leiden: Brill. Motzki, Harald, ed. 2000b. The Biography of Muḥammad: The Issue of the Sources. Leiden: Brill. Motzki, Harald. 2001. “Methoden voor de datering van islamitische overleveringen” (inaugural address). Nijmegen: Katholieke Universiteit Nijmegen.

88 Motzki Motzki, Harald. 2005. “Dating Muslim Traditions: A Survey.” Arabica 52.2:204–​53. Motzki, Harald. 2010. “The Origins of Muslim Exegesis: A Debate.” In Analysing Muslim Traditions: Studies in Legal, Exegetical and Maghāzī Ḥadīth, edited by Harald Motzki, 231–​303. Leiden: Brill. Motzki, Harald. 2015, “Abraham, Hagar and Ishmael at Mecca: A Contribution to the Problem of Dating Muslim Traditions.” In Books and Written Culture of the Islamic World: Studies Presented to Claude Gilliot on the Occasion of his 75th Birthday, edited by Andrew Rippin and Roberto Tottoli, 361–​84. Leiden: Brill. Motzki, Harald. 2016. “Ibn Jurayj.” In Encyclopaedia of Islam, 2nd ed., edited by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs. Leiden: Brill. Motzki, Harald. 2017. Reconstruction of a Source of Ibn Isḥāq’s Life of the Prophet and Early Qurʾān Exegesis: A Study of Early Ibn ʿAbbās Traditions (Islamic History and Thought 3). Piscataway: Gorgias Press. Müller, Sascha. 2010. Die historisch-​kritische Methode in den Geistes-​und Kultur­ wissenschaften. Würzburg: Echter. Nagel, Tilman. 2008. Mohammed: Leben und Legende. München: R. Oldenbourg Verlag. Rodinson, Maxime. 1967 [1961]. Mahomet. Paris: Editions du Seuil. al-​Saqqā, Muṣṭafā, Ibrāhīm al-​Abyārī and ʿAbd al-​Ḥafīẓ Shalabī. 19552. al-​Sīra al-​ nabawiyya li-​Ibn Hishām. Cairo: Dār Iḥyāʾ al-​Turāth. Schoeler, Gregor. 1996. Charakter und Authentie der Muslimischen Überlieferung über das Leben Mohammeds. Berlin: Walter de Gruyter. (English translation 2010. The Biography of Muhammad: Nature and Authenticity, translated by Uwe Vagelpohl. London: Routledge.). Schöller, Marco. 2008. Mohammed. Frankfurt am Main: Suhrkamp Verlag. Wüstenfeld, Ferdinand, ed. 1860. Kitāb Sīrat Rasūl Allāh: Das Leben Muhammed’s nach Muhammed Ibn Ishâk bearbeitet von Abd al-​Malik Ibn Hischâm. Göttingen: Dieterichsche Verlagsbuchhandlung.

­c hapter 5

Arabicization, Islamization, and the Colonies of the Conquerors Kevin van Bladel Historians of the Mediterranean region and the Middle East used to treat the advent of Islam and the Arab conquests as a sudden break in history, dividing the past neatly into periods before and after Muhammad.1 In this way, the early seventh century ce became a threshold, before which there was “Antiquity” and after which there was “Islam” or “Islamic civilization.” At first glance, this distinction makes sense. In the earlier period, there were two great empires in the Mediterranean and the Middle East, the Romans and the Persians, each with several different concurrent languages. After Muhammad, one great Islamic empire extended across northern Africa and the Middle East, throughout which Arabic became the principal language. Historians have usually received different training to deal with the periods before and after. This dichotomy is itself a major factor for the generation of the two periods. Historians who specialize in the period before are typically trained in Greek and Latin, the languages that became classical in later European tradition; they sometimes study other important ancient languages as well, in order to consult sources and address public interest in early human civilizations. Historians who specialize in the period after use Arabic as their main source language and address public concerns with the formation and nature of Islam and the political makeup of the modern Middle East. The distinct kinds of training and public audiences have contributed to this sense that in the seventh century ce one major book of history ended and another began.2 There is no denying that the conquests of the seventh century ce had far-​ reaching effects. The events associated with the period transformed societies 1 I wish to express my gratitude to the organizers and hosts of the event “Muhammad and the End of Antiquity,” held by the Zenobia Foundation in Amsterdam, 14 November 2015, at which I presented a more general version of this paper. I thank Alexander Treiger and Arietta Papaconstantinou for helpful comments to drafts of this essay. 2 Van Bladel 2015 addresses institutional ways to move beyond this division. On the historical development of Islamic civilization as a framework for inquiry, see Van Bladel 2020.

© Kevin van Bladel, 2022 | DOI:10.1163/9789004500648_006


van Bladel

in many regions. Probably no other political change in the known history of the Middle East had such wide and lasting demographic ramifications. But the differences between the periods before and after appear especially distinct only because historians who study them have been divided into the two groups just described: those of Antiquity and of Islam. In recent decades, some scholars have learned to use the materials and languages of both periods, thus enabling them to address the character of the changes induced by the conquests and the period of transition without reinforcing an artificial distinction. Viewed with close scrutiny, the idea that these changes were very rapid turns out to be inaccurate, or at least dependent on perspective. It took more than a generation for newly united Arabian peoples to overturn the Roman and Persian political order by armed conquest. Archaeology has revealed that many regions were fairly unscathed by the conquest, and their populations continued as before but with new overlords.3 Some of the conquered regions with later importance, such as central Asia, were not closely integrated into the Islamic state until eighty years after Muhammad’s death in 632 ce. The fighters made substantial territorial acquisitions in the first decade, but the conquest continued for generations, carried out by institutionalized warfare.4 The concurrent development of a cohesive Arab identity, effacing past ethnic differences among Arabian peoples, took generations, too. The rise of the Arabic language to the status of a pervasive vernacular and the organization of Islam into forms recognizable today took centuries. The apparent abrupt shift in history is therefore a mirage created by hindsight from later centuries. New scrutiny of the transition from “Antiquity” to “Islam” has brought the very character of the conquests of the seventh century ce into intense debate. Historians use the terms “Arab conquests” or “Islamic conquests” to describe the events in different ways and with different emphases. The name used is significant. By the first term, historians mean that the Arabs were one of several different ethnic groups who gradually invaded and settled in the ancient empires of the region over several centuries. Thus the Arabs were—​like the Visigoths, who migrated from Eastern Europe into Spain, where they formed an independent kingdom—​a confederation of people from beyond the imperial frontiers who moved into richer and prosperous Roman lands and were there to stay, after and during a long period of ethnogenesis on the Roman frontiers and within Roman territory.5 The later and similar Arab conquest was thus a conquest by “Arabs” as such. By the second term, historians suggest that 3 See generally the contributions in King and Cameron 2015 as well as Schick 1995. 4 Crone 1999; Donner 1981, 221–​26. 5 E.g., Hoyland 2015.

Arabicization, Islamization, and the Colonies of the Conquerors


the new religion of Islam was a primary motivator for the conquest, and hence it was an “Islamic” conquest with a new religion initiating it.6 Neither term, however, seems perfectly adequate without qualification or explanation. The expression “Arab conquests” is not valid if we recognize that the people who carried out the conquests were generally not called “Arab” (by that name) at the onset of the conquests.7 Arab identity emerged as different peoples of the Arabian Peninsula participated in collective outward migration, cooperating in the conquests while competing for positions of power, wealth, and status. Developing a common Arab identity in the wake of their successes buttressed their unity. Before, during, and (for a long while) after the conquests, their primary ethnic identities were not as Arabs, but as members of groups of tribes such as Muḍar and Rabīʿa. Together these groups comprised the great northern people of Maʿadd, along with the Quḍāʿa and the mix of “southern” Arabian tribes and sedentary peoples such as Kinda, Azd, and Ḥimyar, not to mention members of the conquered populations who assimilated rapidly to the conquerors’s new culture. Even their language was not uniform. Through a process not yet sharply defined by modern philologists, the earlier varieties of Arabian languages gave way to Arabic as we know it in the midst of the conquests. There is no room here for a thorough discussion of this matter. It is highly unlikely, however, that all groups called “Arab” in hindsight spoke the same language at the onset of the conquests or that an “Arab” ethnicity of the sort that modern historians have projected onto the Arabian past existed. Despite these sound reservations, however, the expression “Arab conquests” can be justified as valid if we qualify it by allowing that it refers to the concurrent process of Arab ethnogenesis. On similar grounds, however, the conquests themselves were arguably not “Islamic” either, if we bear in mind the well-​known fact that Islam took centuries to be organized into its characteristic forms. For example, systematic Islamic law based on elaborated hermeneutic standards began to develop much later, in the eighth century ce.8 For another example, Sunni Islam, often considered today as the default mode of the religion, seems to have had its real inception only in the second half of the eighth century ce.9 Moreover,

6 E.g., Howard-​Johnston 2010, 461–​64. See also Donner 2010, for whom the new religion was the critical factor; Donner holds that the religion was not yet called Islam. 7 Bashear 1997, 112–​25; Donner 2010, 218–​20. See also the new contribution by Webb 2016 for thought-​provoking discussion of this problem, along with the tempered review of it by Wood 2017. 8 Hallaq 2005; El Shamsy 2013. 9 Tor 2015.


van Bladel

one may harbor doubts about how many of the early fighters actually knew the Qurʾān: manuscripts and reciters of the text and authoritative teachers may have been few while the colonists migrated from Arabian lands to cantonment cities on the fringes of subjugated territories. Some scholars have argued persuasively that the religion was not even called “Islam” for some time after the conquests began.10 However, many contemporary accounts written by non-​Muslims agree that a new religious movement was associated with the conquerors.11 Therefore, the conquests may just as well be called “Islamic,” provided that we understand that Islam was taking shape during this process, and that it has not stopped changing even to the present. To discuss these distinctions—​“Arab” versus “Islamic”—​within the current debate about how to characterize this drawn-​out, epoch-​making event, we must more precisely understand the transition between an apparent “Antiquity” and “Islam” and recognize that a simple periodization makes a poor guide for comprehending complicated historical processes.12 In the space remaining here, I address how historians have attempted to explain the gradual appearance of a large Arabic-​speaking Muslim population where one previously did not exist, and I offer my own suggestions for future research in this area. I intend to focus on two processes related to the previous terminological discussion: Arabicization and Islamization.13 Because of the success of Arabian peoples in the conquests, most sources available to historians for the subsequent period are in Arabic, by authors who were Muslims. The successful conquests alone, however, do not explain this fact. Arabic became the chief source language for historians from then on because of Arabicization. That is, Arabicization is the process by which the Arabic language was adopted as the medium of communication in public spaces and (increasingly) even in the homes of populations not originally Arabic speaking; it also includes the concomitant abandonment of languages previously used in these domains. Today, every country conquered by the Arabs, with the exceptions of the Iberian Peninsula and Iran, is an Arab country, in the sense that their populations almost entirely use forms of Arabic both in writing and in the daily

10 11 12 13

Donner 2010, 68–​74; Lindstedt 2015. Hoyland 1997, with relevant results summarized on pp. 548–​50. Hoyland 2017 reflects further on the choice of terms, showing that it is indeed more complicated than a choice between “Arab” and “Islamic.” On the history of these terms, at least in French, see Aillet 2011, 15. It is useful to distinguish “Arabization,” the process by which people generated an Arab ethnicity, from “Arabicization,” the process by which people came to use the Arabic language.

Arabicization, Islamization, and the Colonies of the Conquerors 93

vernacular.14 Their populations are also predominantly Muslim, even if the most rapid diminishment of Christian and other non-​Muslim populations took place in some regions only in recent times. How did the conquered people come to adopt both the language and the religion of the conquerors? 1


Over the last two decades, scholars concentrating on this period of transition have closely debated the reasons why Arabic became widespread or “successful” as they have sometimes put it. How is it now the chief language of all of northern Africa, the Levant, Iraq, and the Arabian Peninsula, while other languages used in those regions either disappeared or diminished to become the vernaculars of small, sometimes isolated minorities in remote villages, mountains, and the like? What would cause people to abandon the language of their parents, be it Greek, Aramaic, Coptic, Ḥimyarite, or Persian, and to favor the exclusive use of Arabic? The answers given so far by my colleagues in the field have largely been unconvincing. The problem is what linguists call “language shift,” the process by which a population abandons its old language and shifts to a new one.15 The language shift to Arabic among many different populations was extraordinary in its extent, and that requires an explanation. To discuss the matter in the limited space here, I have to sidestep the glaring problem of the genesis of Arabic itself, and focus on the outcome: a type of Arabic recognized by us as Arabic and, eventually, shared (with local variations) by all the conquered populations. Remarkably, many standard narrative accounts of the Arab or Islamic conquests do not even mention the subsequent shift in language. It is taken for granted that the conquest by Arabic speakers meant that the conquered people would come around to speaking Arabic sooner or later. In other cases, historians sometimes describe the process in metaphors. Languages “lose their vitality,” or become weak and die, like biological organisms. They are “ousted” and “give way” to one another, jostling around.16 Wasserstein writes about how Persia was “submerged” and the Persian language defied our expectation that it would “disappear into the maw of Arabic.”17 For others, languages may 14 15 16 17

Even the Iberian Peninsula and Iran would have become “Arab” countries eventually if not for subsequent changes: the Reconquista in the former case and the coming of the Seljuqs in the latter. Myers-​Scotton 2006, 67–​106; Fishman 2013; Potowski 2013. Hoyland 2004, 191. Wasserstein 2003, 265.


van Bladel

persist in a population merely because they had been there for centuries, as if longevity made something more durable. Such metaphors do not provide an explanation, yet they are not rare in modern scholarship concerned with the relative status of different languages in the ancient and medieval Near East. For example, as one scholar puts it, Greek disappeared quickly in the face of Arabic because it was part of a “Hellenistic veneer” on a “Semitic” population; Greek had put down “shallow roots.”18 It is not simply being obtuse to insist that languages do not have roots; it is a matter of disallowing a misleading metaphor. For no matter how long any language may have been used in a locale, it can disappear in one generation if the conditions are right. The only requirement for a language to become extinct is that no children learn it. The length of time in which a language was previously used in a region is, strictly speaking, irrelevant to language shift. Another problematic explanation for the shift to Arabic is that Arabic was a language of empire, a Reichssprache.19 Such languages, it is supposed, spread far and wide simply because an empire imposes the language of conquerors on conquered peoples. Political power, it would seem, accomplishes this change. An “imperial language” is thought to “filter down” into the subject population.20 Therefore, in our case people without power gradually conformed to the culture of the Arabs because the Arabs had power and prestige.21 It is perhaps intuitive that people learn the languages of their powerful conquerors as a matter of utility to better their own conditions. While surely a factor, as an explanation it does not suffice. Moreover, there have been historical cases in which conquering people who have created a state that we might call an empire adopted the language of their new subjects, rather than the other way around. Thus, the central Asian peoples known to the Chinese as Yuezhi conquered Bactria and created the kingdom of the Kushans, but they came to use the local Iranian language Bactrian for official purposes. The language they had previously used is unknown to historians and remains a matter of debate.22 Similarly, the original language of the Kassites, who ruled ancient Babylonia for centuries during the second millennium bce, remains unknown.23 Texts written under their rule were in Babylonian. In recent times, colonial empires of Europe have often left the languages of the subject peoples quite intact. Why 18 19 20 21 22 23

Donner 1981, 94 (as also critically noted by Papaconstantinou 2012, 59n3). Wasserstein 2003, 265–​66; Hoyland 2004, 190–​91. Wasserstein 2003, 265. Ferrando 2006, 265. Sims-​Williams 2002, 229–​30. Brinkman 1980.

Arabicization, Islamization, and the Colonies of the Conquerors 95

should it not have been so for the Arabs? Power alone does not explain it. Surely Greek was a language of empire, too, and the Byzantine Empire was reduced but did not vanish. Why did the newcomer Arabic speakers fail to learn the hitherto prestigious Greek in formerly Byzantine territories, and in formerly Greek-​speaking cities and towns? Indeed, in another account that reverses this “Reichssprache” argument, the use of Greek in writing survived in Syria and the Levant for two centuries after the conquest because it was “associated with” or “connoted” imperial power—​but here it is the former imperial power of the Byzantines that it connotes.24 The ease with which one can arrive at counterarguments clearly shows that it is not enough to say that Arabic spread because of its association with power and empire.25 For that matter, empire is itself not an analytical category, being rather an ideological contrivance used to aggrandize (or, today, to criticize) conquering states in a modern narrative. For some scholars today, “empire” means primarily a state presiding over separate, ethnically diverse populations. By itself the term “empire” (in any of its senses) does not explain language shift, but at best only some of its circumstances. Another explanation for the problem of language shift to Arabic, one that seems to have attracted the most support, suggests that it is a matter of identity. People maintained their traditional languages, the argument claims, because they felt a sense of ethnic pride or self-​confidence, or because their language was used in their religious worship.26 A robust sense of identity, the argument states, means that one retains traditional language; a lack of confidence in one’s identity leads to adopting another language.27 Thus one historian suggests that Arabic came to predominate because of the “strong cultural self-​confidence” of its speakers.28 But why should it not be the reverse, that Arabic speakers merely give the appearance of “strong cultural self-​confidence” because they retained their language? Another has suggested that “Greek was too closely identified with Christianity to succeed,” apparently in the face of the power of Islam.29 But there is no evidence of special discrimination against Greek speakers after the conquests. A related argument regards “Hellenism”—​used in the modern sense for ancient Greek culture and values—​as a force in language shift. “Hellenism” is said to have “enriched” Syriac and Coptic, without causing

24 Mavroudi 2014, 309. 25 For different reasons, Johnson 2015, 16 calls the “Reichssprache” argument a “red herring.” 26 Hodgson 1974, 449; Levtzion 1990, 303; Wasserstein 2003, 268; Hoyland 2004; Papaconstantinou 2012, 75–​76; Mavroudi 2014, 310. 27 E.g., Hoyland 2004, 191, regarding Persian. 28 Kennedy 1986, 117–​19. 29 Wasserstein 2003, 265.


van Bladel

their demise, so that Greek coexisted with these languages; Arabic did not.30 The modern scholars prone to this assessment have projected values born of modern nationalism and classicism back onto the period under consideration. Such values are not analytical. Although one’s language can be an integral part of one’s sense of identity, there are many cases in which language and identity are not neatly linked. For example, Egyptian Copts have not spoken Coptic for centuries, yet Copts today have retained a sense of identity as separate from Egyptian Muslims. Both groups speak Arabic. An eroding “sense of identity” (whatever that would mean in one’s experience) does not sufficiently explain language shift. Nobody abandons his mother’s language simply because he lacks self-​confidence. It is more complicated: after all, he has to learn another language instead. We should expect rather that one’s sense of identity follows the practical matter of which language one speaks or the community with which one associates, and not the reverse, and that people fashion identities following their everyday realities. Furthermore, research has indicated that pride in one’s identity is rarely effective in resisting language shift.31 It invalidates the argument that language shift is caused by changes in identity. Another claim is that Arabic came to be widely adopted because of the literary culture developed in it or, as it has been said, that Arabic was not only a Reichssprache but a Kultursprache.32 In this context some specialists cite the effort to translate learned texts from other languages into Arabic, particularly from Greek, and claim that it induced scholars to use Arabic, transferring “Hellenism” into Arabic and thereby rendering Greek irrelevant but Arabic prestigious.33 Leaving aside the construct of “Hellenism,” in the domain of the sciences Arabic did effectively come to occupy the earlier status of Greek as the medium of scientific inquiry from the eighth century bce onward.34 But, as Papaconstantinou points out, accounts based on the role of scholars and their translations cannot explain language shift.35 Put differently, no greengrocers, cobblers, masons, or farmers abandoned their language and adopted Arabic because the works of Galen and Aristotle were translated into Arabic. Nor did

30 31 32 33 34 35

Mavroudi 2014, 302–​3. Potowski 2013, 323: “positive attitudes are not enough for language maintenance to occur,” whereas “negative attitudes” toward the prior language of the community, as for example when one regards it as no longer useful, “lead to rapid shift.” Hoyland 2004, 191. Wasserstein 2003, 269–​70. Gutas 1998. Papaconstantinou 2012, 65.

Arabicization, Islamization, and the Colonies of the Conquerors


their masters, employers, or owners urge them to adopt Arabic because of a desire for them to participate in prestigious literary culture. Another suggestion is that wherever a conquered group accommodated the conquerors, it made itself useful, and became integrated into the ruling apparatus. The group succeeded not only in self-​preservation but also in preservation of its language. Wherever a conquered group resisted, its language disappeared.36 This proposal exactly opposes the norm known by linguists. Usually, language shift is more likely to occur when people speaking different languages are in closer contact. Turning to this last observation, I advocate another approach that does not begin with factors like identity or prestige. I consider prestige to be secondary; actual power comes before prestige. I argue 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. Therefore, we cannot understand the matter without close attention to the settlement patterns of the Arabic speakers in their conquered territories. Conquests alone do not cause a population to adopt a new language. Decrees by governors and official policies may induce an already bilingual population to prefer one language over the other. However, they cannot accomplish language shift in a monolingual population without an extensive program of implementation and the re-​education of that population, something unknown in premodern times except on relatively small scale (to the best of my knowledge). Orders that official documents be written in a preferred language really affect only the qualifications for secretaries. Aspiring to prestige does not confer a new language. The absolute prerequisite for language shift is close contact between speakers of different languages, and the language that becomes pervasive is usually the one that is most advantageous for its speakers in the circumstances at hand. Sometimes that utility is indeed a matter of prestige or power. But when populations do not interact, they retain their own languages, regardless of any other factors. This simple fact means that the primary and foremost criterion to understand the problem of the relative degrees of use of Arabic and Greek, Coptic, Aramaic, and Iranian languages must be the pattern of settlement by the Arabic speakers among the conquered peoples. The available literary sources, however, usually say little about demography. Ancient and medieval authors typically provide us with little demographic data that we can use effectively, but they have much to tell us about how they felt about themselves. Perhaps that is one reason why modern historians of the 36

Hoyland 2004, 193–​94.


van Bladel

seventh-​century ce conquests have generally neglected population contact in favor of identity as the chief issue. Nevertheless, attention to the survival and nonsurvival of the subject populations’ languages actually allows us to infer the degree of contact and the geographical pattern of contact between the Arabs and subject populations: language shift and contact between speakers of different languages are inextricably linked. Signs of language shift therefore may help us to ameliorate, to a limited degree, a shortcoming in our source data on settlement patterns. In recent discussions of Arabicization, only Papaconstantinou has touched on this most important factor, rather than lingering solely on ideological factors such as identity. Her special concern in these studies is the region of Egypt. She suggests that the Arabic-​speaking conquerors first considered Egypt an “exploitation colony,” in which the Arabs did not interact with the locals but only extracted goods and revenues from them. Later, under the Fatimid dynasty, beginning in the late tenth century ce, Arabs in Egypt had a “settlement colony,” in which the Arabs came to inhabit the countryside where they owned estates and lived among non-​Arabic-​speaking Egyptians.37 Papaconstantinou follows a model proposed by Salikoko Mufwene, a scholar in the growing field of contact linguistics. Ultimately, however, Mufwene’s distinction between settlement colonies and exploitation colonies is a simple observation.38 When settler colonists lived among conquered peoples, language contact intensified, and those colonized were much more likely to shift to the language of the conquerors. When colonists extracted payments or property while living apart, minimizing social contact, the subject peoples maintained their language. The distinction between the two sorts of colonies indicates that the degree and intensity of contact between speakers of different languages correlates to language change. Linguists will not disagree about this factor and Papaconstantinou is right to stress it. Occasionally, primary sources specifically note that contact is the basic condition for the shift to Arabic. For example, in the introduction to an Egyptian theological treatise composed around the year 1050 in Arabic (so that Egyptian Christians might understand it), the reason for this need is explicitly stated, with reference to a mystery of Christian theology:39

37 38 39

Papaconstantinou 2012, 73–​74; echoed by Mavroudi 2014, 304. Mufwene 2008, 216–​17. Kitāb al-​Īḍāḥ 14. See also Swanson 1996, 216 and Richter 2009, 427, whose reading of al-​ ḥunafāʾ in place of the editor’s ajānib i accept.

Arabicization, Islamization, and the Colonies of the Conquerors 99

I say that the reason for the concealment of this mystery from the believers at this time is their intermingling (ikhtilāṭ) with the pagans (ḥunafāʾ, meaning here the Muslims) and the loss of their original, Egyptian language, from which they used to know their doctrine. In short, this author complains that some Christians in Egypt no longer know Coptic because they have interacted with Arabic speakers to the point that they shifted to Arabic. The Christians therefore need their own theology explained to them in Arabic. The author blames the intermingling of the populations for the shift in vernacular from Coptic to Arabic. Al-​Balādhurī’s extensive collection of earlier reports about the conquest, his Kitāb Futūḥ al-​buldān (finished around 890), is one of our best sources for the different sorts of Arab settlement in conquered territories. According to his compilation, at first the conquerors deliberately planned to live in their own settlements separate from the conquered peoples. As al-​Balādhurī relates, “When Muʿāwiya ruled over Syria and Mesopotamia in the name of ʿUthmān ibn ʿAffān [r. 644–​56 ce], he was instructed by him to settle the Arabs in places far from the cities and villages, and to allow them to utilize the lands unpossessed by anyone.”40 This statement reflects what appears to have been a fairly consistent policy during the first decades of the conquests, as other reports and numerous sources attest. This situation minimized language contact and increased the isolation of the Arabic language. It also minimized the risk of Arab assimilation to the more numerous conquered peoples, including conversion to other religions.41 There were, however, important differences in settlements among formerly Sasanian territories and Byzantine territories, as well as local variations.42 The rule of separate settlement particularly applied to the Arabs who conquered the Sasanian Kingdom and Egypt. After initially occupying the former Sasanian capital of Ctesiphon (al-​Madāʾin), the Arabs of Iraq moved their headquarters to their new settlement of Kufa. The amṣār (sing. miṣr), or garrison cities, of Basra and Kufa remained the major Arab settlements on the southern fringes of Sasanian territory, adjoining the Arabian deserts.43 Garrisons appointed near

40 41 42


Hitti 1916, 278; al-​Balādhurī 1968, 178. Kister 1989; Levy-​Rubin 2011a, 58–​59. Donner 1981, 237–​50 clearly describes the differences between Iraqi and Syrian settlements of the conquerors. This differs from Al-​Sharkawi’s assumption (2010, 145) that the “post-​conquest socio-​demographic circumstances were somewhat uniform in all the provinces of the Arab Middle East, with the exception of North Africa.” Morony 1984, 236–​53.


van Bladel

other Iraqi and Iranian towns depended on these two colonies, and only gradually became independent.44 More than a generation later came other new remote foundations, such as Qayrawān (Kairouan) in North Africa in 670 ce and another in the oasis of Merv in central Asia in 671 ce. Wāsiṭ was founded in Iraq around 700 ce. (Baghdad was not founded until 762 ce, under a new caliphal dynasty.) For generations, the bases at Basra and Kufa dominated not only Iraq but also the Iranian plateau, and launched further conquests as far as the Caucasus and Afghanistan. Their dominance persisted because no organized power could capably challenge the Arabic-​speaking conquerors in the formerly Sasanian territory once the Persian dynasty had ended. In effect, after the whole Persian kingdom had been overcome, focus turned to managing local powers and suppressing local resistance.45 Some non-​Arabs moved into Basra and Kufa, but because they had heterogeneous origins, these two early amṣār grew as predominantly Arab cities where Arabic became the sole public medium. The situation differed in formerly Byzantine Palestine and Syria. The Byzantine state and its military remained an active threat in adjacent territory, particularly in Anatolia. Though reduced in power and territorial control, the Byzantines sought to reclaim cities and ports from the Arabs. Al-​Balādhurī makes numerous references to the conquerors’ placement of many garrisons in the ports and towns of these regions.46 Unlike the situation in the former Persian kingdom, these garrisons were necessary for ongoing defense against a lasting foe and for ensuring that locals did not rebel in favor of Byzantium. Such garrisons were mostly small settlements of Arabic speakers, but the Arabs lived alongside non-​Arabs, dispersed among the conquered population. In Syria and Palestine, the conquerors found that some of the population had evacuated their cities and fled into territory that was still under Byzantine control.47 These refugees, who fled especially from cities and coastal towns, probably included disproportionately large numbers of Greek speakers, leaving Aramaic-​speaking servile personnel and rural populations behind.48 In any case, vacant spaces in established cities became available for Muslim lords and their followers to claim. Cities such as Damascus, Aleppo, and Emesa

44 45 46 47 48

See, for example, Robinson 2000, 63–​89 on the garrison of Mosul. Crone 2012, 6–​7. al-​Balādhurī 1968, 107–​71; trans. Hitti 1916, 165–​265. See also Wheatley 2001, 116–​17; Howard-​Johnston 2010, 501–​5; Khalek 2011; and Carlson 2015, 797–​802; none of whom addresses language shift; see also Levy-​Rubin 2011b. Donner 1981, 245–​50; Levy-​Rubin 2011b, 169. Levy-​Rubin 2011b, 170.

Arabicization, Islamization, and the Colonies of the Conquerors 101

thus became important centers for the Arab Muslim population. Al-​Balādhurī records that in some Syrian cities and towns, residents had to give up half of their homes to the conquerors. Though we cannot be certain about the accuracy of the reported arrangements and concessions, there is no question that some Arabic-​speaking Muslims did move into established cities. Here, many Arabs lived side-​by-​side with Aramaic and Greek speakers in well-​developed urban centers. Umayyad rule from 661 ce followed this pattern. One wonders whether the initial success of the Umayyads against other factions was due in part to support from a more integrated population. Damascus was the dynasty’s capital until 744. This pattern of settlement contrasts notably with that of the separate garrison cities used to dominate the former Sasanian kingdom, in which Arabs established their own sites separate from the subject populations. In short, in Syria and Palestine it appears that the conquerors were dispersed among the locals more frequently than in Iraq and Egypt, creating regular and intense circumstances of contact between the locals and the speakers of Arabic. The integration of the Arabic-​speaking ruling minority populations with non-​Arabs in the old cities of Syria and Palestine as well as the countryside ought to have caused more rapid Arabicization in those regions. One plausible criterion measures this phenomenon: the adoption of Arabic by non-​Muslims to address members of their own non-​Muslim community. In fact, Christians in Palestine appear to be the first non-​Muslim community to have adopted Arabic for internal use. The earliest surviving Christian Arabic literature—​ Arabic texts about Christianity for Christians who communicated in Arabic—​ comes from Melkite monks in Palestine in the late eighth century ce.49 The appearance of compositions and translations in Arabic by Christians for Christians in the late eighth century ce in Palestine suggests that vernacular use of Arabic had become so widespread that it began to enjoy common currency even among non-​Muslims. That is, enough Christians had come to use Arabic as their vernacular to the extent that they could not easily understand the other vernaculars of the region (Greek, Aramaic) and required translations in Arabic. In Iraq, Christian authors began to use Arabic to address other Christians about 50 years later, in the first half of the ninth century ce. However, even the earliest examples, Abū Rāʾiṭa of Tikrīt and ʿAmmār of Basra, notably come from major Iraqi cities with Arabic-​speaking settler populations. Moreover, their works are defensive, responding to the challenge of Muslim theologians.50 The 49 50

Levy-​Rubin 1998, 153–​55; Griffith 2008, 48–​50; Treiger 2015, 195, 198–​203. Griffith 2008, 61–​63.


van Bladel

use of Arabic in these examples was probably not typical of the Christians in the Iraqi countryside, who must have continued to use dialects of Aramaic and Iranian languages. The Christians of Iraq apparently did not address their own communities in Arabic until about half a century later than the Christians of Palestine, at the earliest. Iraqi Christians probably did not use Arabic among themselves generally, at least in certain major population centers, until much later. The tenth century ce appears to be the period of rapid Arabicization in Iraq, well after the change in Palestine. Arabic-​speaking Christian Saracens immediately to the east of Palestine before the conquests complicate the picture: the Arabicization of Palestine could seem to have begun before the conquest, thereby making it seem more rapid in hindsight.51 Nevertheless, early evidence of Arabic in the region is not from within Palestine but immediately around it. It is difficult to accept that the colonization of the region of Palestine by Arabic-​speaking conquerors was not the chief stimulus to the shift to Arabic there. Indeed, we lack real evidence for the Arabicization of Palestine until after the conquest. On the fringes of Iraq, too, settled Christian people called Ṭayyāyē are usually understood, in hindsight, to have been Arabs.52 Whatever Arabian language they spoke, the Ṭayyāyē complicate the situation, for similar reasons. In the end, however, neither region’s metropolitan peoples or country folk shifted to Arabic until after the conquests by Arabic-​speaking Muslims, and that is the criterion proposed here to measure, however roughly, different rates of Arabicization. Other regions can be drawn into the comparison. The pattern of Arabicization moved even slower in Egypt. As in Iraq, the Arabic-​speaking conquerors established habitations largely separate from the conquered people, particularly the miṣr of Fusṭāṭ, having little to do with the locals for many decades.53 Eventually, of course, the Muslim Arabic speakers came to occupy many Egyptian sites, but Christians did not use Arabic for intra-​communal writing until the late tenth century ce, and particularly in the eleventh century ce.54 Spain (al-​Andalus) presents another case study that seems similar to the example of Palestine. It is more difficult to track rates of Arabicization in this region because few Christian Arabic texts survived the aftermath of the Reconquista. Surviving evidence nevertheless suggests a more rapid Arabicization of Christians in Spain arising from closer community contact. Arabs did not conquer and colonize the Iberian Peninsula region until the 51 52 53 54

See Fisher 2015, 347–​50 for the scanty but important linguistic evidence. Toral-​Niehoff 2014, 151–​211. Sijpesteijn 2013, 77–​84. Rubenson 1996; Sidarus 2013; Mikhail 2014, 95–​96.

Arabicization, Islamization, and the Colonies of the Conquerors 103

eighth century ce, but immediately they settled in existing towns and cities and on established country estates, probably creating many situations for contact between Arabic speakers and Romance-​speaking locals.55 Here, however, it seems that Christians wrote in Arabic for other Christians already in the second half of the ninth century ce,56 in a somewhat shorter amount of time than in Egypt or in Iraq. Such loose estimates constitute a clumsy yardstick, but they do permit initial comparisons between regions. The shift to Arabic among Jews ought to be considered, too. Unsurprisingly, the earliest known Jewish authors of Arabic works for other Jews seem to have originated in sites already inhabited by Arabic-​speaking Muslims. However, making local distinctions in the patterns of Arabicization among Jews is difficult because either the locations of these authors are not known for certain or they frequently migrated between regions. The earliest known Jewish authors of Arabic works wrote in the early tenth century ce, including Saʿadyah Gaon (who moved from his home in the Fayyūm in Egypt to Palestine to Baghdad) or the Karaite al-​Qirqisānī (the location of his career is not clearly known, but his family apparently came from a town along the Euphrates in Iraq). Various still undetermined factors must have played roles in the interregional discrepancies in Arabicization. Nevertheless, a difference of at least two generations in the general adoption of Arabic within Christian communities in Syria-​Palestine and Spain on the one hand and Egypt and Iraq on the other suggests corresponding differences in the rate of Arabicization. It is reasonable to hypothesize that the clear regional variation in settlement patterns is probably the primary and original factor in the different rates of Arabicization. Wherever Arabic speakers settled throughout the country alongside locals, even relatively sparsely, the locals came to adopt Arabic more quickly. Wherever Arabic speakers created new separate cities and effectively segregated themselves, the subjects in the lands around them were slower to shift to Arabic as the vernacular in quotidian use. Future refinements to this simple model of population contact will be most welcome. Of course, each region’s settlement patterns changed after the initial colonization by Arabic speakers, and these local complexities need to be taken into account where possible. For example, in 1938 Poliak considered the migration of country folk to the cities under ʿAbbāsid rule in the eighth and ninth centuries ce to be a decisive factor in Arabicization.57 In a given settlement, 55 56 57

Kennedy 1996, 16–​18. Griffith 2008, 67–​68, 153–​54. Poliak 1938, 53–​56.


van Bladel

the number of Arabic speakers compared to speakers of other languages must have been a factor in the choice between Arabic or another language, but data about these ratios are scarce. It will also be important to consider patterns of family building and intermarriage between Arabic speakers and others. The proposal I have offered here relates only to the initial settlements. This contribution certainly does not aim to provide a comprehensive explanation, but only an orientation for further research. We can sharpen our insight into the phenomenon of Arabicization by paying attention to patterns of language and cultural change identified by sociolinguists. In a landmark article, Milroy and Milroy argued that “Linguistic change is slow to the extent that the relevant populations are well established and bound by strong ties, whereas it is rapid to the extent that weak ties exist in populations.”58 That is, a dense, closely-​knit group will remain linguistically conservative. Specialists in language shift likewise report that such tightly-​knit groups with relatively few ties to outsiders resist the loss of their community language, even over long periods.59 This observation accords with common sense. Indeed, the common-​sense character of this observation may have led some scholars to view a strong sense of identity as a cause in the preservation of one’s language. However, the feeling of solidarity itself does not appear to cause resistance to language change, except perhaps where people are especially conscious of shibboleths that demarcate them. The feeling of solidarity and resistance to language change are simultaneous products of a network of social relationships characterized by close, multiplex bonds in a community with few external social ties. When learners from outside the population introduce only a few variations to a language, changes to the language as a whole will be correspondingly fewer and slower. Maintaining an internally-​used language by a small, closely-​knit group surely also facilitates their solidarity, but solidarity by itself does not affect language maintenance. The observation just cited from a pair of sociolinguists—​that closely-​knit ties slow language change—​receives support from the sociology of social networks in human relationships. In contrast to close, local networks in which every individual has a relationship to every other individual, the term “cosmopolitan network” has been used to describe loose social ties in sparse webs of connections among persons who do not associate closely. This looser network differs from tightly-knit networks in which every member may enjoy a close

58 59

Milroy and Milroy 1985, 375. Strictly speaking, their study describes changes induced by contact in an individual language, not by language shift, but the principles are the same. Fishman 2013, 473–​75.

Arabicization, Islamization, and the Colonies of the Conquerors 105

tie to every other member.60 These two kinds of networks represent relative ends on a scale, and are not mutually exclusive to the individual. Yet the so-​ called cosmopolitan networks better serve as conduits for the exchange of new information than dense, multiplex local networks: there is little new to report among groups with close ties; individuals already know all about each other.61 Information can include linguistic novelties that are adopted more readily when loose ties, such as extra-​familial ties between different towns, predominate. Under such circumstances, one is exposed to more social novelties or new information more frequently. Arabicization, like other instance of language shift, must have been conditioned by these social factors. 2


If subject peoples who had the most social contact with Arabic speakers adopted Arabic first, how did the conversion to Islam, or Islamization, occur? To begin, we must acknowledge with Thomas Carlson that “Islamization was a complex and multi-​dimensional process that spanned many centuries.”62 Yet, as with the study of Arabicization, most recent research on conversion in Late Antiquity and the early Islamic period has focused on identity, with the lines between Islamization and Arabicization blurred.63 One scholar has supposed that converts to Islam would “hasten to learn Arabic” as an “automatic” process, as if every Muslim needs to know Arabic.64 At times, too, historians have resorted to metaphorical explanations, in which a non-​Muslim community becomes “exhausted” and “declines” under Muslim rule;65 Islam then prevails. Historians have held sociological studies of conversion at arm’s length, partly because sociologists themselves have applied their models to conversion in ancient history with results that are not entirely persuasive.66 Nevertheless, the pattern of conversion is similar to that of language shift, 60 61 62 63 64 65 66

Stark 2007, 35–​39. Although one may relate “cosmopolitan” here to the “cosmopolitanism” recently a part of theoretical discussions in the domains of literature or empire, they are not strictly the same thing. Scott 2000, 35. Carlson 2015, 792. See Papaconstantinou 2015 for current approaches, and the studies collected in that volume. In the same volume, Cameron 2015 presents a largely noncommittal summary of recent research on conversion in the late ancient period. Anawati 1990, 244. E.g., Levtzion 1990, 305. E.g., Stark 1996.


van Bladel

while subtly distinct in important ways. The main difference arises from social constraints: in the societies in question, religious experts expected their followers to maintain exclusive allegiance to their one religion, but there were no rules against being multilingual. Moreover, dense, closely-​knit religious groups will tend to maintain their religion along with their language.67 Sociologists who have conducted empirical and descriptive studies of conversion in case studies in modern settings have arrived at conclusive results: one is likely to convert when the social ties to members of one’s former religion, often family members, become weak and advantages accrue through new social ties created by regular associations with members of another religion.68 In addition, religious differences expressed through conspicuous behaviors, such as distinctive attire or observances of prohibitions and obligations, are likely to generate social tension. People everywhere tend to prefer minimizing such tensions. This tension must have been especially present when non-​Muslims were required to indicate their religious community through attire and deferential conduct in public, standards called ghiyār in Arabic.69 The attenuation of old, close ties with coreligionists who had exerted pressure to conform, together with the growth of new pressures to follow different customs, makes individuals more likely to convert, especially when new social ties with members of other religions become socially and economically advantageous.70 Conversion to a religion also is likely after one creates new family relationships or other close relationships with members of that religion. To a degree, these observations reflect common sense. People rarely join the religions of strangers. Furthermore, it is much easier, and less uncomfortable, to convert to a new religion not held by one’s family when living far from that family. Those older, formerly close family ties can potentially be replaced by new ties to one’s immediate environment, where the new religion is normative.71

67 68 69 70


Levy-​Rubin 1998, 161 rightly infers the “close-​knit” character of the Palestinian Melkites from their retention of their religion despite their early Arabicization. Stark and Finke 2000, 116–​18 present a summary of this research. Levy-​Rubin 2011a. Bulliet’s analysis of conversion narratives concludes that “the initial decision to join the religious community of the rulers had more to do with attainment or maintenance of status than it did with religious belief” (1990, 131). This is not to say that a convert’s experienced feelings are not sincere. My considerations in this section are similar to those of Bulliet 2005, 246–​52, who refers to “information-​based innovation diffusion,” as information about Islam moved along networks of social relationships.

Arabicization, Islamization, and the Colonies of the Conquerors 107


Patterns of Arabicization and Islamization

After consideration of the two social patterns of relationships in which new languages are learned and old religions are maintained, I suggest that our seventh-​century ce subjects would more likely, and more quickly, adopt a new language through so-​called cosmopolitan network ties than a new religion. New, loose ties are easily formed, but it is personally risky to break established close and dense ties such as family bonds. The religions present when Islam took shape demanded exclusive commitment and exerted severe social consequences for apostasy, including ostracism, denial of inheritances, and sometimes death.72 These observations further suggest that the spread of Arabic would in the end outpace the spread of Islam. As it happens, present reality supports the theory: most non-​Muslim communities in countries once colonized by Arabs have come to use Arabic as the language of daily use, even in the home, but they retained Judaism, Christianity, and other religions, becoming what we would call non-​Muslim “minorities” in a generally Muslim population.73 A further factor, however, plays a role: the age of the individual subject. Children enjoy the biologically-​endowed ability to learn new languages, such as those used outside of the home in the neighborhood, with fluency, but adults past the critical language-​learning threshold seldom master new languages.74 At the same time, children likely cannot escape strong pressure to conform to the religion of their family elders. In societies like those of the Near East in the seventh century ce, when important matters of inheritances and the very laws to which one was subjected varied according to religious community, this pressure would be all the more significant.75 We should expect therefore that non-​Muslim children of non-​Arab background who grew up in Arabic-​speaking environments were the first in their families to enter Muslim society, but that they did so only late in their own lifetimes, as Arabic-​speaking converts to Islam. 72 73 74


Simohnson 2015. Cf. Versteegh 2014, 126. This factor mostly invalidates the argument that Aramaic speakers shifted to Arabic (as opposed to Greek) “with ease” because of the prehistoric genetic relationship of Aramaic to Arabic or because of the resulting similarity of the two (e.g., Levtzion 1990, 302; Levy-​ Rubin 1998, 157). All things in the context being equal, however, children seem to learn the languages to which they are regularly exposed equally well regardless of the historical relationship of those languages to one another. It may be, however, that adult learners acquire languages that are close cognates with their native tongues more easily. Morony 2012.


van Bladel

I suggest two models as typical. The first occurred mostly in the early decades, in the wake of the conquests. Many individuals, under the right circumstances, particularly in captivity among Muslims, converted to Islam despite lacking much of a grasp of Arabic or, perhaps, even a clear idea of what the religion meant. These were the “clients,” mawālī, who became Muslim and attached themselves to an existing tribe among the conquerors: conquered people who adopted their conquerors as patrons. Noted especially in the first decades of Umayyad rule (661–​750 ce), such converts were a distinct, relatively new group. It is difficult to estimate their actual numbers over time.76 Unsurprisingly, they possessed a notoriously poor ability to use Arabic.77 Once removed from their homes and families, they found it expedient to convert to Islam even with only a poor adult-​learner’s grasp of Arabic. Maintaining their former religion provided no advantages when their family and community ties were removed. Another telling case in which conversion preceded Arabicization comes from the accounts of the conquest and conversion of the inhabitants of the central Asian Sogdian city of Bukhārā. This case did not involve the relocation of conquered people. According to the surviving account, in 713 ce the general Qutayba ibn Muslim (d. 715 ce) subdued the city and attempted to force a mass conversion of the population to Islam. However, the converts quickly revealed themselves as insincere and continued to venerate idols in secret. To facilitate their participation in the religion, which meant, most publicly, attending the newly-​dedicated mosque for prayers, the people of Bukhārā required instruction in their own language about how to change postures during prayer. Qutayba ibn Muslim even offered cash for attendance at the mosque. The report exists because it was an exceptional case. Qutayba finally had to settle his Muslim followers among the Bukhārans as an attempt to ensure their fidelity to the religion. This case demonstrates that only close social contact, and here also an initial degree of coercion, could really effect conversion to Islam before Arabicization. Even so, the report suggests that it was a fairly unsuccessful effort in the short term.78 The second model is likely more typical than the first, when considered over time. In the centuries after Muhammad, non-​Muslim families who did not speak Arabic chose to move, or were brought, to Muslim settlements, surrounded by Muslim Arabic speakers who employed them or used their 76 77 78

Bulliet 2005 argues that the mawālī are over-​represented in Islamic historical sources and that consequently historians today tend to overestimate their relative numbers in the Muslim population. Versteegh 2014, 130–​31. Frye 1954, 48.

Arabicization, Islamization, and the Colonies of the Conquerors 109

services.79 Other non-​Muslim families remained in place, but came to have Muslim landlords and neighbors who settled in their villages and towns, over decades and centuries. In both cases, these families were now physically close to Arabic-​speaking Muslims. The children in these families learned Arabic in their neighborhoods with native fluency; their parents did not. Yet these children did not dare to convert to Islam so long as they remained under the guardianship of non-​Muslim parents and extended family, and under the jurisdiction of their community’s religious laws. Upon reaching maturity and self-​sufficiency (and when family elders died and properties were inherited), these Arabophone non-​Muslims of the second generation had the most to gain and the least to lose in becoming Muslim. The third generation, the grandchildren, spoke mostly only Arabic and were also Muslim. In other words, Arabic and Islam were first adopted as urban phenomena because the conquerors initially created urban residences, which became the main site of contact.80 This model of a three-​generation language shift is in fact typical, observed with some variations through many different modern case studies.81 In our era of large-​scale global migrations, it will be familiar to many readers from their personal experiences, too. Classical Arabic texts preserve numerous anecdotes (and medieval Arabic texts even preserve first-​person accounts) about non-​Muslim men who knew Arabic—​often scholars writing in Arabic—​and converted to Islam from another religion at an advanced age. These individuals no longer would incur the disappointment of their elders, who were now dead and gone. Moreover, through conversion they would enhance their standing in the Arabic-​speaking environment where they worked at the height of their careers.82 They learned the language first, joined the religion second, and not the other way around as one may have expected.83 Certainly there must have been other patterns, but we should expect those just described to have been typical. The relationship between Arabicization and Islamization would vary in other countries and in later centuries according to different demographic situations. 79 80

81 82 83

On such persons and what distinguished them from mawālī, see Crone 2003, 297–​300. On early Islam as an urban phenomenon, see Bulliet 1994, 67–​79 and, generally, Wheatley 2001. See also Al-​Sharkawi’s contribution (2013), which revives and modifies Ferguson’s hypothesis of 1959, that the modern vernacular Arabic dialects have common origins in the Arabic koine dialects generated in the first garrison cities created in the conquests. This study attempts to elucidate the evolution of Arabic dialects rather than the adoption of Arabic among the conquered peoples, but it remains relevant here. Tsunoda 2005, 73; Myers-​Scotton 2006, 100; Potowski 2013. Hackenburg 2015, Roberts 2017. Versteegh 2014, 126. Cf. Décobert 1992, 300.

110  4

van Bladel


The application of models created by sociologists is not history. That said, observing regularly occurring patterns is certainly a step in the right direction away from the ideas that historians of these processes in Late Antiquity have adopted, with the sense of one’s identity as a primary factor. Scholarship on Late Antiquity has produced many studies considering how people felt about themselves, but to understand language shift and conversion it is first necessary to understand some of the social and demographic changes behind those feelings. The initial process of Arabicization occurred rapidly where Arabs settled, especially where they distributed themselves across the land in many settlements, thus creating many sites of contact; Arabicization occurred more slowly everywhere else. Conversion to Islam followed Arabicization, except in the case of captives, for whom it was expedient to convert immediately. In areas where no Arabs settled at all, populations took a very long time to adopt Arabic or Islam. In places where non-​Arabic speakers moved into Arabophone settings, children soon adopted Arabic. Some rural populations never shifted to Arabic, as for example in Iraq, where numerous Jewish and Christian villagers continued to speak increasingly divergent dialects of Aramaic up to the present century. Except in special cases, mostly involving captives removed from their homeland, Islamization probably followed Arabicization. Surprisingly, these sorts of observations have infrequently entered discussions of Arabicization and Islamization.84 Each process required different kinds of interpersonal contact. That is, each region had its own path and its own pace of Arabicization and Islamization according to its own demography and the settlement patterns of the Arabic-​speaking Muslims in it. Therefore, there cannot be a single, uniform explanation for these two major outcomes of the seventh-​century ce conquests, nor can explanations based on ideological factors like sense of identity suffice. Demographic patterns must be the foundation for our explanations. The ideological factors invoked in primary sources are important, but they cannot be the starting point. If we must abandon the search for uniform patterns of Arabicization and Islamization, then much more research on the local populations of the post-​conquest period is needed.85 The study of Arabicization and conversion to Islam, taking population contact as 84 85

A rare exception is Mikhail 2014, 70–​71, 74–​75, 93–​95, who touches on the historical sociology of Islamization and Arabicization with sound insights. Cf. Versteegh 2014, 150: “Much more information is needed about the sociolinguistic context of the early Islamic empire and the pattern of settlement in each particular area. Even more help may be expected from general models to explain the evolution of languages.”

Arabicization, Islamization, and the Colonies of the Conquerors 111

its starting point, has barely begun. New indicators of the adoption of Arabic as a vernacular must be detected to refine the chronology of the shift. Using this kind of research, the changes occurring during the seventh century ce can be understood as more than a sudden break in history. Moreover, the impasse over designating the conquest as “Arabic” or “Islamic” can be more effectively negotiated. In the model I have proposed here, I have not emphasized the question of why the outcome should be language shift to Arabic, instead of a shift of Arabic-​speakers to another language or long-​term bilingualism. These questions raise complexities that I can address only briefly here. My suggestion is that utility was the deciding factor in a language shift to Arabic from a bilingual setting. Utility here includes not only necessary communication but also social and economic advantages.86 Mawālī, the early non-​Arabic-​speaking converts to Islam, mostly captives, would find indispensable practical advantages in being able to communicate with their new patrons. This view differs from the “Reichssprache” argument, because it is not about language imposed along with political power, but rather the social and economic advancement of the new learner. Most basically, however, utility can simply mean enabling communication with enough people of importance in one’s life. Therefore, the proportion of speakers in a given setting was critical. In other words, language shift was still primarily related to networks of social relationships. In the large cantonment cities founded as Muslim settlements, such as Basra and Kufa, the Arabic speakers probably always outnumbered the outsiders (who thus needed to learn the language pervasive in their new locale). Other languages became unnecessary when all members of a family used Arabic. The number of speakers of different languages must have mattered, but those numbers were meaningful only within domains of contact. For example, even in small garrisons, if Arabic speakers interacted with locals through a limited number of local representatives, then, in effect, the Arabic speakers outnumbered the speakers of other languages in the domain of their regular contact. This conclusion remains hypothetical because it is probably impossible to know the proportions of speakers in contact in any situation. It is clear, though, that foreign learners induced a degree of restructuring in the language. This restructuring has been proposed as the origin of the Arabic vernacular dialects. Holm’s model of partial restructuring through language contact would suggest that Arabic speakers were never a tiny fraction of the population anywhere in which a shift to Arabic occurred. If they had been, we would expect 86

Myers-​Scotton 2006, 91–​106.


van Bladel

a fully-​fledged creole, which the Arabic vernaculars are not.87 Bilingualism must have been prevalent among the populations on the demographic frontiers between conquerors and conquered—​frontiers that shifted ever outward in each successive generation. In this sense, then, long-​term bilingualism did exist, but moved steadily outward as usage grew and new generations of once bilingual families at the center of the shift became monolingual in Arabic. The obvious comparandum from the same region is ancient Greek. This comparison launched the recent wave of scholarship on this topic, following Wasserstein’s question: “Why did Arabic succeed where Greek failed?”88 Greek was the language of the dominant classes of the eastern Mediterranean during most of the millennium between Alexander and Muhammad. Although the extent of the shift to Greek during this period remains a subject of debate, it is clear that a complete shift did not take place. Coptic, Syriac, and other languages survived alongside Greek. Why, then, did the seventh-​century ce conquests not produce a long-​lasting bilingual society (like the communities that used Greek), but instead result in a complete language shift to Arabic in most of the conquered territories, causing the abandonment and demise of other languages? Answers partly emerge from a chronological perspective. Arabic has been used in the region for 13 centuries. Missing from these discussions, however, is almost any inquiry into the settlement and migration patterns of Greek speakers in Antiquity. Such inquiries could prove decisive in comparing the relative histories of Greek and Arabic (and other languages, such as Latin). The proposed “stable bilingualism” of ancient Greek with other languages probably just glosses over unstable demographic factors, such as attrition of urban populations, migrations to urban centers, patterns of intermarriage, and other factors that occurred over centuries and that have not yet been investigated with reference to language domains.89 It seems highly unlikely that any population has truly enjoyed “stable bilingualism” for hundreds of years, as has been supposed for Greek, but this question is certainly beyond the scope of this essay. Perhaps what is most remarkable about the Islamic or Arab conquests is that language shift and conversion to a new religion were closely connected outcomes of the same events. The co-​occurrence of these two major changes in the wake of the conquests distinguishes them from most other historical 87 88 89

Holm 2004; Versteegh 2004; McWhorter 2007, 165–​96. Cf. Al-​Sharkawi 2010, 169–​70. Wasserstein 2003. As J. Edwards (2013, 6) notes with respect to bilingualism, “even stability is relative.” Clackson (2012) offers an interesting perspective on bilingualism between Greek and Coptic, while relying perhaps too much on “stable bilingualism” as an explanatory tool.

Arabicization, Islamization, and the Colonies of the Conquerors 113

parallels, such as the Greek conquest of the Achaemenian Persian Empire. By using settlement patterns as the starting point, comparisons among different case studies can be made more effectively and thus we can shed light on differing outcomes following episodes of colonization in the wake of conquest.


Aillet, Cyrille. 2011. “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), edited by Dominique Valérian, 7–​34. Paris: Sorbonne. Al-​Sharkawi, Muhammad. 2010. The Ecology of Arabic: A Study of Arabicization. Leiden: Brill. Al-​Sharkawi, Muhammad. 2013. “Urbanization and the Development of Arabic: The Case of Garrison Towns in the Seventh and Eighth Centuries CE.” Journal of Ancient History 1.2:1–​35. Anawati, Georges C. 1990. “Christian Communities in Egypt in the Middle Ages.” In Conversion and Continuity: Indigenous Christian Communities in Islamic Lands, Eighth to Eighteenth Centuries, edited by Michael Gervers and Ramzi Jibran Bikhazi, 237–​51. Toronto: Pontifical Institute of Mediaeval Studies. Athamina, Khalil. 1986. “Arab Settlement during the Umayyad Caliphate.” Jerusalem Studies in Arabic and Islam 8:185–​207. al-​Balādhurī. 1968 [1866]. Kitāb Futūḥ al-​buldān. Edited by M. G. de Goeje, 2nd ed. Leiden: Brill. Bashear, Suliman. 1997. Arabs and Others in Early Islam. Princeton: Darwin. Bladel, Kevin van. 2015. “Graeco-​Arabic Studies in Classical Near Eastern Studies: An Emerging Field of Training in Its Broader Institutional Context.” Intellectual History of the Islamicate World 3:316–​25. Bladel, Kevin van. 2020. “A Brief History of Islamic Civilization from Its Genesis in the Late Nineteenth Century to Its Institutional Entrenchment.” Al-​ʿUṣūr al-​Wusṭā: The Journal of Middle East Medievalists 28: 150–​73. Brinkman, J. A. 1980. “Kassiten.” In Reallexikon der Assyriologie und vorderasiatischen Archäologie, vol. 5, edited by D. Edzard, 464–​73. Berlin: Walter de Gruyter. Bulliet, Richard. 1979. Conversion to Islam in the Medieval Period: An Essay in Quantitative History. Cambridge: Harvard University Press. Bulliet, Richard. 1990. “Conversion Stories in Early Islam.” In Conversion and Continuity: Indigenous Christian Communities in Islamic Lands, Eighth to Eighteenth Centuries, edited by Michael Gervers and Ramzi Jibran Bikhazi, 123–​33. Toronto: Pontifical Institute of Mediaeval Studies.


van Bladel

Bulliet, Richard. 1994. Islam: The View from the Edge. New York: Columbia University Press. Bulliet, Richard. 2005. “Conversion-​based Patronage and Onomastic Evidence in Early Islam.” In Patronate and Patronage in Early and Classical Islam, edited by Monique Bernards and John Nawas, 242–​62. Leiden: Brill. Cameron, Averil. 2015. “Christian Conversion in Late Antiquity: Some Issues.” In Conversion in Late Antiquity: Christianity, Islam, and Beyond, edited by Arietta Papaconstantinou, Neil McLynn, and Daniel L. Schwartz, 3–​21. Surrey: Ashgate. Carlson, Thomas A. 2015. “Contours of Conversion: The Geography of Islamization in Syria, 600–​1500.” Journal of the American Oriental Society 135.4:791–​816. Clackson, James. 2012. “Language Maintenance and Language Shift in the Mediterranean World during the Roman Empire.” In Multilingualism in the Graeco-​ Roman Worlds, edited by A. Mullen and P. James, 36–​57. Cambridge: Cambridge University Press. Crone, Patricia. 1999. “The Early Islamic World.” In War and Society in the Ancient and Medieval Worlds, edited by K. Raaflaub and N. Rosenstein, 309–​22. Cambridge: Harvard University Press. Crone, Patricia. 2003. “The Pay of Client Soldiers in the Umayyad Period.” Der Islam 80:284–​300. Crone, Patricia. 2012. The Nativist Prophets of Early Islamic Iran: Rural Revolt and Local Zoroastrianism. Cambridge: Cambridge University Press. Décobert, Christian. 1992. “Sur l’arabisation et l’islamisation de l’Égypte médiévale.” In Itinéraires d’Égypte: Mélanges offerts au père Maurice Martin s.j., edited by Ch. Décobert, 273–​300. Cairo: Institut français d’archéologie orientale. Donner, Fred. 1981. The Early Islamic Conquests. Princeton: Princeton University Press. Donner, Fred. 2010. Muhammad and the Believers: at the Origins of Islam. Cambridge, MA: Harvard University Press. Edwards, John. 2013. “Bilingualism and Multilingualism: Some Central Concepts.” In The Handbook of Bilingualism and Multilingualism, edited by Tej K. Bhatia and William C. Ritchie, 5–​25. Chichester: Blackwell. El Shamsy, Ahmed. 2013. The Canonization of Islamic Law: A Social and Intellectual History. Cambridge: Cambridge University Press. Ferguson, Charles A. 1959. “The Arabic Koine.” Language 35:616–​30. Ferrando, Ignacio. 2006. “The History of Arabic.” In The Encyclopedia of Arabic Language and Linguistics, vol. 1, edited by Kees Versteegh, 261–​68. Leiden: Brill. Fischer-​ Bovet, Christelle. 2011. “Counting the Greeks in Egypt: Immigration in the First Century of Ptolemaic Rule.” In Demography and the Graeco-​Roman World: New Insights and Approaches, edited by C. Holeran and A. Pudsey, 135–​54. Cambridge: Cambridge University Press.

Arabicization, Islamization, and the Colonies of the Conquerors 115 Fisher, Greg, ed. 2015. Arabs and Empires before Islam. Oxford: Oxford University Press. Fishman, Joshua A. 2013. “Language Maintenance, Language Shift, and Reversing Language Shift.” In The Handbook of Bilingualism and Multilingualism, edited by Tej K. Bhatia and William C. Ritchie, 466–​94. Chichester: Blackwell. Frye, Richard. 1954. The History of Bukhara. Translated from a Persian Abridgment of the Arabic Original by Narshakhī. Cambridge: The Mediaeval Academy of America. Griffith, Sidney H. 2008. The Church in the Shadow of the Mosque: Christians and Muslims in the World of Islam. Princeton: Princeton University Press. Gutas, Dimitri 1998. Greek Thought, Arabic Culture. The Graeco-​Arabic Translation Movement in Baghdad and Early ʿAbbāsid Society (2nd-​4th/​8th-​10th centuries). London: Routledge. Hackenburg, Clint. 2015. “Voices of the Converted: Christian Apostate Literature in Medieval Islam.” PhD diss., The Ohio State University. Hallaq, Wael B. 2005. The Origins and Evolution of Islamic Law. Cambridge: Cambridge University Press. Hitti, Philip Khuri. 1916. The Origins of the Islamic State, vol. 1. New York: Columbia University Press. Hodgson, Marshal. 1974. The Venture of Islam, Volume 1: The Classical Age of Islam. Chicago: University of Chicago Press. Holm, John A. 2004. Languages in Contact: The Partial Restructuring of Vernaculars. Cambridge: Cambridge University Press. Howard-​Johnston, James. 2010. Witnesses to a World Crisis: Historians and Histories of the Middle East in the Seventh Century. Oxford: Oxford University Press. Hoyland, Robert. 1997. Seeing Islam as Others Saw It: A Survey and Evaluation of Christian, Jewish and Zoroastrian Writings on Early Islam. Princeton: Darwin. Hoyland, Robert. 2004. “Language and Identity: The Twin Histories of Arabic and Aramaic (and: Why Did Aramaic Succeed Where Greek Failed?).” Scripta Classica Israelica 23:183–​99. Hoyland, Robert. 2015. In God’s Path: The Arab Conquests and the Creation of an Islamic Empire. Oxford: Oxford University Press. Hoyland, Robert. 2017. “Reflections on the Identity of the Arabian Conquerors of the Seventh-​Century Middle East.” Al-​ʿUṣūr al-​Wusṭā 25:113–​40. Johnson, Scott F. 2015. “Introduction: The Social Presence of Greek in Eastern Christianity, 200–​1200 CE.” In Languages and Cultures of Eastern Christianity: Greek, edited by Scott F. Johnson, 1–​122. Surrey: Ashgate. Kennedy, Hugh. 1986. The Prophet and the Age of the Caliphates: The Islamic Near East from the Sixth to the Eleventh Century. Essex: Longman Group. Kennedy, Hugh. 1996. Muslim Spain and Portugal: A Political History of al-​Andalus. Essex: Pearson Education Limited.


van Bladel

Khalek, Nancy. 2011. “Patterns of Habitation in Post-​Conquest Damascus and Its Hinterland.” In Le proche-​ orient de Justinien aux Abbassides: Peuplement et dynamiques spatiales, edited by A. Borrut, M. Debié, A. Papaconstantinou, D. Pieri, and J.-​P. Sodini, 139–​47. Turnhout: Brepols. King, G.R.D., and Averil Cameron, eds. 1994. The Byzantine and Early Islamic Near East II: Land Use and Settlement Patterns. Princeton: Darwin. Kister, M.J. 1989. “ ‘Do Not Assimilate Yourselves …’ Lā tashabbahū. …” Jerusalem Studies in Arabic and Islam 12:321–​53. Kitāb al-​Īḍāḥ = (Pseudo-​)Sāwīrus ibn al-​Muqaffaʿ, al-​Durr al-​thamīn fī īḍāḥ al-​dīn. Cairo: Madāris al-​Tarbīya al-​Kanasīya bi-​Kanīsat Raʾīs al-​Malāʾika al-​Jalīl Mīkhāʾīl. Levtzion, Nehemia. 1990. “Conversion to Islam in Syria and Palestine and the Survival of Christian Communities.” In Conversion and Continuity: Indigenous Christian Communities in Islamic Lands, Eighth to Eighteenth Centuries, edited by Michael Gervers and Ramzi Jibran Bikhazi, 289–​311. Toronto: Pontifical Institute of Mediaeval Studies. Levy-​Rubin, Milka. 1998. “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, edited by Arieh Kofsky and Guy G. Stroumsa, 149–​62. Jerusalem: Yad Izhak Ben Zvi. Levy-​Rubin, Milka. 2011a. Non-​Muslims in the Early Islamic Empire: From Surrender to Coexistence. Cambridge: Cambridge University Press. Levy-​Rubin, Milka. 2011b. “Changes in the Settlement Pattern of Palestine Following the Arab Conquest.” In Shaping the Middle East: Jews, Christians, and Muslims in an Age of Transition 400–​800 C.E., edited by Kenneth G. Holum and Hayim Lapin, 155–​72. Bethesda: University of Maryland Press. Lindstedt, Ilkka. 2015. “Muhājirūn as a Name for the First/​Seventh Century Muslims.” Journal of Near Eastern Studies 74.1:67–​73. Marsham, Andrew. 2009. “The Early Caliphate and the Inheritance of Late Antiquity (c. AD 610-​c. AD 750).” In A Companion to Late Antiquity, Philip Rousseau, 478–​92. Chichester: Wiley-​Blackwell. Mavroudi, Maria. 2014. “Greek Language and Education Under Early Islam.” In Islamic Cultures, Islamic Contexts: Essays in Honor of Professor Patricia Crone, edited by Behnam Sadeghi, Asad Q. Ahmed, Robert Hoyland, and Adam Silverstein, 295–​342. Leiden: Brill. McWhorter, John. 2007. Language Interrupted: Signs of Non-​Native Acquisition in Standard Language Grammars. Oxford: Oxford University Press. Mikhail, Maged S. A. 2014. From Byzantine to Islamic Egypt: Religion, Identity and Politics after the Arab Conquest. London: I.B. Tauris. Milroy, James, and Lesley Milroy. 1985. “Linguistic Change, Social Network and Speaker Innovation.” Journal of Linguistics 21:339–​84.

Arabicization, Islamization, and the Colonies of the Conquerors 117 Morony, Michael G. 1984. Iraq after the Muslim Conquest. Princeton: Princeton University Press. Morony, Michael G. 2012. “Religious Communities in the Early Islamic World.” In Visions of Community in the Post-​Roman World: The West, Byzantium and the Islamic World, 300–​1000, edited by Walter Pohl, Clemens Gantner, and Richard Payne, 155–​ 63. Surrey: Ashgate. Mufwene, Salikoko. 2008. Language Evolution: Contact, Competition and Change. London: Continuum. Myers-​Scotton, Carol. 2006. Multiple Voices: An Introduction to Bilingualism. Malden: Blackwell. Papaconstantinou, Arietta. 2012. “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, edited by A. Mullen and P. James, 58–​ 76. Cambridge: Cambridge University Press. Papaconstantinou, Arietta. 2015. “Introduction.” In Conversion in Late Antiquity: Christianity, Islam, and Beyond, edited by Arietta Papaconstantinou, Neil McLynn, and Daniel L. Schwartz, xv–​xxxvii. Surrey: Ashgate. Papaconstantinou, Arietta, Neil McLynn, and Daniel L. Schwartz, eds. 2015. Conversion in Late Antiquity: Christianity, Islam, and Beyond. Surrey: Ashgate. Poliak, A.N. 1938. “L’Arabisation de l’Orient sémitique.” Revue des études islamiques 12:35–​63. Potowski, Kim. 2013. “Language Maintenance and Shift.” In The Oxford Handbook of Sociolinguistics, edited by Robert Bayley, Richard Cameron, and Ceil Lucas, 331–​39. Oxford: Oxford University Press. Richter, Tonio Sebastian. 2009. “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, edited by Hannah M. Cotton, Robert G. Hoyland, Jonathan J. Price, and David J. Wasserstein, 401–​46. Cambridge: Cambridge University Press. Roberts, Alexandre. 2017. “Being a Sabian at Court in Tenth-​Century Baghdad.” Journal of the American Oriental Society 137.2:253–​77. Robinson, Chase F. 2000. Empire and Elites after the Muslim Conquest: The Tranformation of Northern Mesopotamia. Cambridge: Cambridge University Press. Rubenson, Samuel. 1996. “The Transition from Coptic to Arabic.” Égypte/​monde arabe 27–​28:77–​92. Schick, Robert. 1995. The Christian Communities of Palestine from Byzantine to Islamic Rule: A Historical and Archaeological Study. Princeton: Darwin. Scott, John. 2000. Social Network Analysis: A Handbook, 2nd ed. London: Sage. Sidarus, Adel Y. 2013. “From Coptic to Arabic in the Christian Literature of Egypt (7th-​ 11th Centuries).” Coptica 12:35–​56.


van Bladel

Sijpesteijn, Petra M. 2013. Shaping a Muslim State: The World of a Mid-​Eighth-​Century Egyptian Official. Oxford: Oxford University Press. Simonsohn, Uriel. 2015. “Conversion, Apostasy, and Penance: The Shifting Identities of Muslim Converts in the Early Islamic Period.” In Conversion in Late Antiquity: Christianity, Islam, and Beyond, edited by Arietta Papaconstantinou, Neil McLynn, and Daniel L. Schwartz, 197–​215. Surrey: Ashgate. Sims-​Williams, Nicholas. 2002. “Ancient Afghanistan and Its Invaders: Linguistic Evidence from the Bactrian Documents and Inscriptions.” In Indo-​Iranian Languages and Peoples: Proceedings of the British Academy CXVI, edited by Nicholas Sims-​Williams, 225–​42. Oxford: Oxford University Press. Stark, Rodney. 1996. The Rise of Christianity: A Sociologist Reconsiders History. Princeton: Princeton University Press. Stark, Rodney. 2007. Sociology, 10th edition. Belmont: Wadsworth Cengage Learning. Stark, Rodney, and Roger Finke. 2000. Acts of Faith: Explaining the Human Side of Religion. Berkeley: University of California Press. Swanson, Mark N. 1996. “The Specifically Egyptian Context of a Coptic Arabic Text: Chapter Nine of the Kitāb al-​Īḍāḥ of Sawīrus Ibn al-​Muqaffaʿ.” Medieval Encounters 2.3:214–​27. Tor, Deborah. 2015. “God’s Cleric: Al-​Fuḍayl b. ʿIyāḍ and the Transition from Caliphal to Prophetic Sunna.” In Islamic Cultures, Islamic Contexts: Essays in Honor of Professor Patricia Crone, edited by Behnam Sadeghi, Asad Q. Ahmed, Adam Silverstein, and Robert Hoyland, 195–​228. Leiden: Brill. Toral-​ Niehoff, Isabel. 2014. Al-​Ḥīra. Ein arabische Kulturmetropole im spätantike Kontext. Leiden: Brill. Treiger, Alexander. 2015. “Christian Graeco-​Arabica: Prolegomena to a History of the Arabic Translations of the Greek Church Fathers.” Intellectual History of the Islamicate World 3:188–​227. Tsunoda, Tasaku. 2005. Language Endangerment and Language Revitalization. Berlin: Mouton de Gruyter. Versteegh, Kees. 2004. “Pidginization and Creolization Revisited: The Case of Arabic.” In Approaches to Arabic Dialects: A Collection of Articles Presented to Manfred Woidich on the Occasion of His Sixtieth Birthday, edited by Martine Haak, Rudolf de Jong, and Kees Versteegh, 343–​57. Leiden: Brill. Versteegh, Kees. 2014. The Arabic Language, 2nd ed. Edinburgh: Edinburgh University Press. Wasserstein, David. J. 2003. “Why Did Arabic Succeed Where Greek Failed? Language Change in the Near East after Muhammad.” Scripta Classica Israelica 22:257–​72.

Arabicization, Islamization, and the Colonies of the Conquerors 119 Webb, Peter. 2016. Imagining the Arabs: Arab Identity and the Rise of Islam. Edinburgh: Edinburgh University Press. Wheatley, Paul. 2001. The Places Where Men Pray Together: Cities in Islamic Lands, Seventh through the Tenth Centuries. Chicago: University of Chicago Press. Wood, Philip. 2017. Review of Webb 2016. Al-​ʿUṣūr al-​Wusṭā 25:178–​83.

­c hapter 6

Continuity and Change: Elite Responses to the Founding of the Caliphate Peter Webb The wealth of recent scholarship on Late Antiquity demonstrates a seminal shift in perspective on the rise of Islam. The impression of most twentieth-​ century historians that Muhammad’s mission inaugurated a new epoch that definitively buried Antiquity is undercut by now prevailing views that Islam’s emergence did not trigger a rapid transformation of Middle Eastern life, politics, or even some of its religious systems. Late antique landholding elites, clerical establishments, and Christian and Zoroastrian rural populations greeted the emergence of the caliphate as a political change at the top that necessitated limited alteration to their local structures; the full transformation of the Middle East into a Muslim world appears to have unfurled more gradually. Scholarly emphasis on “continuity” does however carry latent risks of hypercorrecting the previous paradigm of “change” by overemphasizing some of early Islam’s continuities with Antiquity. Notwithstanding continuities identified in recent scholarship, key changes did occur, especially among those who authored the venture of Islam. This paper considers the scale of continuity and change by evaluating how communities of the early caliphate’s military elites responded to the success of the conquests and establishment of the caliphal system. To assess the scope of change, this paper’s focus is upon the social: the ideas that provide windows into the self-​identity of the early caliphate’s communities. Moving beyond the application of labels that define peoples/​sects/​groups in rough-​hewn generalized and totalizing outlines, there is a need to sharpen analysis by posing questions to our sources that can reveal how communities articulated their sense of self: how did they conceptualize issues of identity? What sort of world did they think they inhabited, i.e., what spatial narratives did they construct to express their senses of “home” and proprietary space? What roles did they accord religious rituals in setting communal boundaries? Along with these broad questions, we pay particular attention here to perhaps the period’s most salient issues of community and identity: Islam and Arabness and the employment of these terms in connection with articulations of the communal “inside” of the caliphate’s elite. By posing questions

© Peter Webb, 2022 | DOI:10.1163/9789004500648_007

Elite Responses to the Founding of the Caliphate


to the poetry from pre-​Islam and the Umayyad periods, we find important changes maturing in the Marwānid era (64–​132 ah/​684–​750 ce) revealing that the ever-​constant processes of constructing identity exerted particularly momentous force at this time. There is a wealth of source material, and in a paper of this scope we settle with revealing the contours of what poetry offers and invite more detailed studies. The paper closes with closer consideration of one case study—​the tribe of al-​Azd in Basra—​as an example of how elite responses to the rise of the caliphate engaged processes of Arab ethnogenesis that underline the difficulty of approaching social actors in the Umayyad period with preset and monolithic labels. The scope of change surrounding our historical figures situated them in processes of adapting and reorganizing themselves and their identities to negotiate the fluctuating political circumstances of their world. 1

Conquests and Conquerors?

First, a note on terminology. The rise of Islam is viewed through the prism of conquest history: Western scholarship labels the opening stages of the caliphate in terms of “Muslim conquests” or “Arab conquests,” and their narratives are organized as military history. Premodern Arabic sources about early Islam describe the process by which the Muslim polity expanded as fatḥ (pl. futūḥ/​ futūḥāt), and while fatḥ’s connotation of “opening” does not necessarily tally with military invasion,1 the association of fatḥ with specifically “military victory” may be a loan from Ethiopic contemporary with the development of Qurʾānic Arabic.2 The earliest layers of Arabic lexicography (i.e., from the late second ah/​eighth century ce onward) are clear that fatḥ is to be understood as “opening of the abode of war” (iftitāḥ dār al-​ḥarb) or “victory” (nuṣra).3 Hence, although fatḥ is not one of the most typical Arabic words for war/​conquest/​ invasion, Arabic writers chose to interpret fatḥ in explicitly military terms, 1 Donner 2016, 5 opts to define fatḥ as “some momentous event that is good for the Believers,” and thus argues to move away from conceptualizing the futūḥ as “conquest,” and instead as part of a “salvation-​historical agenda of nascent Islamic historiography” (2016, 8). 2 The Ethiopic evidence is described in Donner 2016, 3–​4, who considers it remote given the wider meanings of fatḥ in the Qurʾān. The evidence of the early Arabic lexicographic sources, outside of Donner’s purview, do support the military connotation of the word, and al-​Azharī places it at the opening of the word’s definition 2004, 3:457: Donner’s contention that “conquest” is but only the “secondary meaning” of the root (2016, 2) does not necessarily align with the word’s early interpretation. 3 See al-​Khalīl 1980, 3:194; al-​Jawharī 1984, 1:389; al-​Azharī 2004, 3:457.

122 Webb and thus primed modern scholarship to perpetuate the tradition of imagining Islam’s origins as imperial expansion. The difficulty with the imperial fatḥ/​conquest model is its dissonance with critical analysis of the sources. First, the earliest communities of “conquerors” do not appear to have referred to themselves as such: they are not recorded as having calling themselves fātiḥūn (i.e., the active participle of fatḥ),4 “raiders” (ghāzūn), “victors” (ghālibūn, ẓāfirūn, muẓẓafarūn, muntaṣirūn, etc.), or other combinations of terms that imply people with an imperial expansionist ideology. The first conquerors seem to have preferred the term muhājirūn,5 or “Emigrants.” This term also appears in Arabic poetry from the early period of expansion,6 and leads to the sense that nascent Muslims considered themselves to be “moving” rather than “conquering” with an imperial agenda. Second, and related to the first, the “Emigrants” did not mix with indigenous populations as they settled in new towns, the amṣār, nor did they seek to convert the locals. The pattern of settlement indicates groups intent on helping themselves, motivated by their own inward-​looking beliefs, rather than concerted efforts to establish and manage an integrated empire. Scholars have noted that indications of a “Muslim state” do not clearly appear immediately upon Muhammad’s death; a “state” becomes visible perhaps during Muʿāwiya’s caliphate, or even only during ʿAbd al-​Malik’s. In either case, the first twenty years of military expansion do not exhibit an imperial metropolitan “center” served by the amṣār. This observation accords with an impression that early Islam began with a diffusion of Emigrants whose goal was something other than imperial state creation. The Emigrant militarized advances were nonetheless different from “tribal migration.” Religious ideology is visible in the early records, along with indications of caliphal input in the organization of some advances, but the evidence for a state system is limited and a program of centrally coordinated campaigns is not uniformly attested across the geographical expansion that we label as the “Conquests.” Furthermore, the details of some individual “conquests” reveals an ephemeral nature. The Emigrants defeated the Byzantine

4 Fātiḥūn may not have been appropriate, since the Qurʾān associated fatḥ with God’s blessing to the Muslims, hence a related active participle muftatiḥūn might be more theologically sound—​i.e., “the people possessing fatḥ”; this term is likewise absent in the sources. 5 The importance of “Emigrant” identity in early Islam has recently been detailed by Lindstedt 2015. 6 See, for example, the line of ʿAbda ibn al-​Ṭabīb, who laments the departure of a beloved: “She set her dwelling in the military camp at Kufa,/​An Emigrant, and took all her love with her!” (al-​Anbārī 2003, 1:349).

Elite Responses to the Founding of the Caliphate


and Sasanian armies in short order, and they diffused into the territories left behind by the two collapsing empires. However, the expansions further afield, into the Caucasus, North Africa, Iberia, and Khurāsān/​Eastern Iran seem better understood as raids which yielded “treaties” of nominal fealty from the locals, and after which the Emigrant units soon returned to their amṣār towns in the Fertile Crescent. Only after a generation or two (depending on the region) did armies return to these far-​flung regions with more evident intentions of staying and managing. By the end of the first century ah/​seventh century ce, a more centralized caliphate was in operation; likewise, the will and interest in running an integrated empire becomes more evident. From the perspective of Arabic historiography, the moment at which the first Emigrant military force entered a region and extracted cash constituted the climactic event of fatḥ/​conquest, and Arabic historiographical narratives are structured to present such events as monuments of heroic action and provide mythic significance to communal origins that established the borders of the Muslim world.7 This narrative tendency inflates the significance of the early stages of raiding and expansion: the actual integration of lands into a (more or less) centrally planned caliphate unfurled gradually as Emigrants first filled the vacuum left by the Byzantines and Sasanians; later the caliphate assumed greater responsibility in local political management under its widening umbrella. To read Islam’s early history as a tableau of imperial military expansion accordingly falls prey to the epic of Arabic historiography and its tendency to glorify the origins of the caliphate in grand military terms via an edifice of “conquest.” After the initial, decisive battles with Byzantines and Sasanians, opportunity arose for someone to take control of the power vacuum left behind, and, over the course of a generation, the caliphate organized and finally “conquered” for itself the land previously occupied by the Emigrants. In short, the issue of “conquest” revolves around a semantic question. I do not mean that militarism was not central to the early Emigrants’ ideology, or that they were unaware of the powerful authority of the Qurayshite caliphs. However, the uneven contours of the expansion invite serious engagement with theoretical questions of what we mean by “conquest,” and narratological questions about how we can interpret the available sources. To engage in this critical scrutiny, this paper eschews the terms “conquest” and “conqueror.” To conceptualize the spread of the nascent Muslim community, this paper ventures the hypothesis that Muhammad’s immediate successors operated with

7 I use “mythic” here in the sense articulated by Eliade 1968, 8: an event of significance for communal senses of identity; it does not necessarily need be “untrue.”

124 Webb a Qurʾānic-​inspired belief that God would grant them blessed fatḥ, and that they expanded outward with the intention of founding new communities and extracting the resources of their Byzantine and Sasanian neighbors. These goals were not imperial, and perhaps the process of converting the occupied lands into a caliphal system run by Muslims was not even conceived of at the outset. Only after the dust of battle had settled, the spoils counted, and the opportunities for a new state grasped, could the Emigrants begin transforming into empire builders. The multiple fitna wars fought between the military groups and rival caliphs suggest that a unified venture of Islam needed to be forged, and that it was not robustly established at the outset. The descendants of the original Emigrants became the elite of the new system, and they monopolized the use of violence in the caliphate. Thus, I will refer to their communities as “elites” or “military elites,” since the term “conquerors” may misrepresent their self-​identity at all stages. Initially they were armed settlers, and once the Umayyad caliphate asserted itself as a state at the end of the first century ah/​seventh century ce, the elites then reimagined the first generation as the “conquerors;” the elites themselves were by this time keen to identify as “Muslims” and “Arabs.” Likewise, it seems misdirected to seek the elites’ responses to the process of conquest/​initial expansion: the important questions about the development of the society which created the Muslim culture we encounter in source literature really revolve around the elites’ responses to the maturation of the Umayyad caliphate, in particular how they responded to the evident efforts of imperial consolidation in the Marwānid period. 2

Continuity of Emigrant Communities?

It is a wonderful irony that the earlier scholarship which so emphasized the epoch-​making changes of Islam’s rise nonetheless persisted in a converse belief that the Emigrants themselves were little changed by the process of emigration and resettlement. Historians of Islam postulated that pre-​Islamic Arabia was populated by Arabs whom they imagined as a culturally cohesive community that embraced Muhammad’s mission, migrated across the Middle East, and established the caliphate, yet stayed much as they were before: they remained Arabs. Historical writing used one ethnonym for both the pre-​ Islamic and Umayyad eras, creating an impression of Arab ethnic continuity and unity. The scholarly opinions about Arabness enjoyed that pristine inertness which historians tend to project on peoples with an obscure history. There were few historical records about pre-​Islamic Arabia, and hence, from Edward

Elite Responses to the Founding of the Caliphate


Gibbon onward,8 European historians cast the “Arabs” as textbook barbarians: they emerged from the “Empty Ḥijāz”9 to destroy the remnants of the empires of Classical Antiquity and were but little civilized or changed in the immediate aftermath, until, that is, they were “civilized” by adopting a hybrid Persian-​Arab culture in the ʿAbbāsid era.10 Within these familiar paradigms, in the study of early Islam, a pre-​Islamic “Arab” was conceptually similar to an Umayyad-​era “Arab.” The latter differed predominately in having the trappings of an empire which made him wealthier and politically more organized than his pre-​Islamic forebears. The assumed lack of cultural change and the fixedness of Arab identity is epitomized in von Grunebaum’s 1963 observation that “pre-​Islamic Arabs” constituted a Kulturnation, which the “Muslim Arabs” converted into a Staatsnation:11 i.e., the wars of expansion and the establishment of the caliphate were primarily political in nature, and did not affect what it meant to be an “Arab.” Consequently, scholars set Arabness as the fixed pole from which emanated all the changes that they imagined swirled about the other late antique communities following the rise of Islam. The essential foreignness and difference of Arabness compared to late antique Middle Eastern cultures was postulated as the very driver of change: the newcomers were different; their language, culture, and religion were seen as different because the newcomers had been isolated in the deserts of pre-​Islamic Arabia. The putative spread of these “Arab ways” across the Middle East via the founding of the caliphate thereby neatly explained the end of “Greco-​Roman” Antiquity, the primary prism through which historians periodized the advent of Islam. Under this paradigm, “Arabs” had to play the role of game-​changers and conveyors of a new culture in order to sustain the narrative of Islam as an epoch-​making event. Hence, historians assumed that the core of “Arab culture” must have remained more or less


9 10


Gibbon 1994, 5:754–​55 marks a shift in European historiography by labelling Muhammad’s community as “Arab.” Prior to the late eighteenth century, European writing predominantly referred to Arabians and Muslims with the term “Saracen”: see Webb 2016, 46–​47, Donner 2018, 4. Montgomery 2006, 45–​50 provides a detailed study and critique of the “Empty Ḥijāz” theory, a construct that imagined pre-​Islamic Arabia as essentially isolated from the rest of the late antique world. In an authoritative source on the period’s history, the Cambridge History of Iran published in 1975, we read “Islam was rescued from a narrow Bedouin outlook and Bedouin mores primarily by the Iranians, who showed that Islam both as a religion and, primarily, as a culture, need not be bound solely to the Arabic language and Arab norms” (Frye 1975, xi; see also Zarrīnkūb 1975, 28, 42–​43, 56). Von Grunebaum 1963.

126 Webb constant in the centuries before and immediately after Islam’s rise. To this end, and as Fred Donner has shown, European scholars first coined the term “Arab conquests,” which from the middle of the nineteenth century became the popular explanation for the arrival of a barbarian ethnicity with the intent of establishing a new empire that caused the abrupt end of Antiquity.12 Similarly, Wellhausen dubbed the resultant caliphate under the Umayyads an “Arab kingdom.”13 On closer reflection, the modern European coining of the term “Arab conquests” raises difficulties. First, from a macro perspective, the grand narrative of the radical change which Islam wrought across the Middle East is problematic, and given the developments in scholarship which show the broad continuities in early Islam, there is no longer a need to imagine the early Muslims as a barbarian deus ex machina whose cultural difference forged a changed world in the wake of their expansion. The impression of the early Muslims’ distinctive separateness as ethnically unified “Arabs” is thus a carryover from an obsolete grand narrative. Second, turning to the early communities themselves, if the Emigrants themselves did not label their expansions as “Arab” (and they did not, as indicated by known premodern Arabic sources), by what authority did European historians cloak the historical movement in an ethnic guise?14 The Emigrants hailed from Arabia, but the Peninsula was, at that period, a fragmented region comprised of different cultural and political zones. The assumption that all Arabians were “Arabs” is a hasty melding of space with race into a rigid conception of what a community means, and it lacks sensitivity to how pre-​Islamic groups within the Peninsula imagined their own communal boundaries. The many wars, 12 13 14

Donner 2018, 4–​5, 11–​12. Wellhausen 1927. A recent and detailed study of culture and identity from the perspective of Christian communities in the Middle East under the early caliphate includes a short section evaluating the terminology for “conquests,” with the rather remarkable conclusion that “Arab conquests” is an appropriate term (Tannous 2018, 525–​31). The study argues that Syriac sources appear to treat the invaders as a cohesive ethnic group, and because it is difficult to determine quite what the Emigrants thought about themselves, “Arab” is preserved as a term of convenience (531n). However, Greco-​Latin writers and their late antique successors had a habit of viewing all peoples via an ethnic lens (see the essays in Gruen 2011). Therefore, the fact that Syriac writers thought about the “conquerors” as an ethnicity is not necessarily relevant to the question: they were predisposed to think in such ways, and they were outsiders to the Emigrant movement in any event. Moreover, the Syriac writers did not use the term “Arab” at all, so that the decision to interpolate “Arab” into their writing today perpetuates the Orientalist outlook and construct. For other critiques of the prevailing approach of “Arabizing” the past, see Millar 2013, 154–​58; Donner 2018, 12–​15.

Elite Responses to the Founding of the Caliphate


different language groups, and different kingdoms across pre-​Islamic Arabia are facets of cultural and communal divides. While contemporary Greek and Roman writers tended to generalize the panoply of Arabian identities in the centuries before Islam into a space/​race meld of Arabness, their undifferentiated views can now be improved. Moreover, theories of identity and ethnicity anticipate that groups at the center of major political and economic developments will exhibit changes, and it would be rare—​perhaps unprecedented—​ for groups to undertake widespread resettlement and reorganization of their communities into a novel state without altering core aspects of their own identities. The putative continuity of Arabness obscures the Marwānid achievement of establishing a centralized caliphate that substantially integrated the formerly disparate (and fractious) Emigrant communities. The early generations of European Orientalists can in some respects be excused for their fixed notions of Arabness, given the then-​prevailing notions of nation and race. Prior to the mid-​twentieth century, scholarship accepted that people’s racial and national identities were longue durée characteristics that remained constant across changes in context. In the Middle Eastern case, an “Arab” always embodied “Arab culture,” but now such notions of fixed “race” have been replaced with more nuanced ideas of “ethnicity.” Ethnos, and all forms of identity have been shown to be constructs—​ideas that are never whole. The monolithic national labels familiar to modern readers each fundamentally emerged as ways to imagine the cohesion of communities from time to time under their unique circumstances. Theories of ethnicity stress that changes in economic and political structure are the primary drivers of an ethnogenesis that exerts transformative power on how people imagine community and identity.15 The descendants of the Emigrant communities—​the people who created and established the caliphate in the mid-​to late first century ah/​seventh century ce—​indeed experienced various changes that ran deeper than simply establishing a new kingdom. The descendants distinguished themselves by their religion, with its associated scripture and rituals. They also established a new form of kingship: rather than referring to their rulers in the terms of the emperors/​shahs whom they replaced, the new office of caliph emerged 15

For the fundamental theories of ethnogenesis, see Weber 1996 and Barth 1969. These sources stress the central roles of political and economic factors in shaping communal boundaries within which ethnic identities develop. Subsequent theorists added further nuance, but the methodology of Weber and Barth remains largely intact: see Jenkins 2008 for more recent synthesis, and Pohl and Reimitz 1998 for a classic study of post-​Roman ethnogenesis.

128 Webb accompanied by new aspects of rulership.16 Geographically, the Emigrants also modified their world: they moved from various regions of the Arabian Peninsula and spread across the wider Middle East, where they primarily settled in newly-​constructed towns (amṣār) separated from old urban centers.17 From their perspective, therefore, the Emigrants uprooted their communities and built new homelands; formerly dispersed groups found themselves living in close quarters that remained spatially distinct from local populations. As an ironic twist, therefore, the amṣār—​situated outside of the Arabian Peninsula—​conceptually created the possibility of Arab unity: the amṣār first concentrated ex-​Arabian peoples into one space with common, shared interests, and thus nurtured the breaking down of the old senses of communal boundaries that had existed in pre-​Islamic Arabia. Economic changes also emerged: the Emigrants first enriched themselves with the spoils of war, and then established regular income through taxation of the conquered lands. Former traders and herdsmen found themselves in radically new economic relationships. In brief, the actions of the Marwānid-​Umayyad military elites directly oppose traditional models about the rise of Islam. Instead of bringing “Arabia” into the Middle East, they relinquished much of their physical ties to pre-​ Islamic Arabia and built new visions of community in new lands, where almost all aspects of their quotidian lives could be separated from former traditions. Prima facie, therefore, a study of elite responses to the establishment of the caliphate as a pan-​Middle Eastern system of government would expect the elite community to exhibit some of the most dramatic changes, and while conservative elements in societies seek to maintain traditions, the scope of change and the mixing of populations in the Umayyad period are of a magnitude which usually tip the scales against continuity and in favor of change.



Al-​Azmeh 2014, 96 refers to the caliphate as possessing “profound generic continuity with east Roman, Byzantine monarchism and imperialism,” with supporting discussion, and indeed it can be argued that the caliphate borrowed earlier traditions. However, al-​ Azmeh does not bring Sasanian traditions into the analysis; caliphal borrowings from that direction are significant. The Muslim rulers’ choice to designate themselves as “caliphs” was also new, and the caliphs’ imperial reach extended to a much different territory than the Romans. Beyond the caliph’s adaptions from Roman precedent, therefore, it does seem that early Muslims developed a new form of imperial rule. Archaeological work continues to reveal new towns established by the Emigrants, either as “greenfield” new builds or “brownfield” occupation of abandoned or lightly-​inhabited areas adjacent to the older cities, from central Asia to the Atlantic. Whitcomb 1994, 28 succinctly described the process of amṣār-​construction as “intentional reconstitution of the social organization of the conquered lands.”

Elite Responses to the Founding of the Caliphate




To evaluate the nature and extent of continuity and change experienced by Emigrant groups, we shall need to engage sources that furnish information capable of conveying the emotive senses of “self,” and compare such expressions of communal identities and practices in the pre-​Islamic period with expressions of the same kinds of sentiments in the Umayyad caliphate. Only once we establish a baseline notion of community in both pre-​Islam and the Umayyad era can we ground the comparison, and this poses two substantial methodological challenges. First, pre-​Islamic Arabia was a fragmented region with manifold cultures and peoples whose diverse identities resisted collection into one cohesive cultural block that can constitute a unified subject for analysis.18 Similarly, the Umayyad world included communities from the Atlantic to central Asia which had their own distinct fragmentations. The Umayyad-​era case is helped by the twin political and demographic facts that the caliphate was clearly developing centralization (even if it did not fully succeed, it achieved certain results), and the settlers in amṣār communities reshaped the demographic map of the Middle East and concentrated populations in new ways. We can thus discern Umayyad-​era acculturation among the caliphate’s military elite, but both ends of our comparative exercise rest on shifting foundations that will temper results. Second, scholars have often commented on the lack of detailed records for both pre-​Islamic Arabia and early Islam. Large numbers of pre-​Islamic inscriptions have been found in parts of Arabia, but (with the exception of south Arabian imperial inscriptions) these examples are predominately short texts, and extrapolating them to address big questions about identity, communal relations, and belief structures can be difficult to sustain, especially given the uneven distribution of finds. As for the Umayyads, we possess voluminous Arabic prose histories and collections of anecdotes, but these were written between fifty to two hundred years after the fact, and while they are more detailed than the inscriptions, especially with regard to the kind of information we need to think systematically about how people imagined their identities, the texts may project later, anachronistic discourses onto the past. At this point, it may seem that we shall never know enough about either pre-​ Islamic or Umayyad communities to establish meaningful criteria to evaluate the extent of continuity and change during the rise of the caliphate. However, 18

I have discussed the fragmentation of pre-​Islamic Arabia elsewhere (Webb 2016, 77–​88); see also an elaboration of these ideas with respect to Arabian Jewish communities in Hughes 2017, 9, 40–​53.

130 Webb the process of chipping away into the unknown can be ventured via an alternative source: Arabic poetry. The poetic corpus speaks to emotive subjects that offer uniquely rich insight into people’s identities and cultural values. We possess substantial quantities of pre-​Islamic, early Islamic, and Umayyad-​era poetry, and because the styles of poetic composition were, to an extent, stable across the period, the evidence’s form facilitates comparison. The corpus is also vast, providing ample material to observe patterns, making it possible to sift through poetic convention and artistic rhetoric to probe historical questions about the scope of social developments. Arabic poetry’s authenticity has been questioned because the extant collections were compiled from the late second century ah/​eighth century ce, and thus post-​date the pre-​Islamic and Umayyad-​era contexts by at least fifty, and up to 250 years. The gap is less significant than one might fear, however, since poetry circulated widely in oral (and likely written) form prior to the ʿAbbāsid era; poetry’s meters and rhymes assist memorization and protect textual integrity across its transmission, and while much poetry was lost, surviving verses are not necessarily ʿAbbāsid-​era Iraqi fakes anachronistically attributed to the mouths of pre-​Islamic Arabian poets. ʿAbbāsid-​era poetry narrators were themselves aware of poetry fabrication; they sought to identify spurious material and critiqued uncareful narrators,19 and as a result, we now possess collections of certain individual poets’ works and thematic anthologies that were endorsed by ʿAbbāsid-​era specialists. Other collections of poetry survive in historical and other sources, but such verses were neither transmitted by the specialists nor considered genuine. The latter poems often contrast with the tenor, syntax, and/​or vocabulary of poetry from the more established collections: the verses are akin to the poetry recorded in medieval popular literature, and more work is needed to determine to what extent the popularizing-​style Arabic poetry may trace to pre-​Islamic origins. In the 1920s, sweeping claims accused the whole poetic corpus of forgery, and while the specter of forgery remains, especially with problematic verses narrated outside of the main poetry collections, poetry scholars today are more confident that much of the material preserved in the specialized premodern collections is genuine. The presence of some persistent false attributions is an insufficient excuse to discount the material. Other scholars have defended poetry’s usefulness,20 and trends identifiable across large numbers 19 20

See, for example, Ibn Sallām al-​Jumaḥī’s comments on poetry fabrication, especially poems added to the Prophet’s biography (al-​Jumaḥī n.d., 1:4–​8). Bauer 2010; Agha 2011, 8 describes the twentieth-​century scholarly retreat from the “vigorous” doubts about authenticity in the 1920s; see also Webb 2016, 68–​70. Dmitriev 2017, 106 articulates a similar call to use the poetry.

Elite Responses to the Founding of the Caliphate


of verses very likely do present us with real sentiments from the earlier eras we seek to explore. The poetic corpus has been underutilized to examine the nature of community in pre-​and early Islam. Poetry specialists have asked literary questions about the development of Arabic poetry’s forms and style, but employment of poetry to answer historians’ questions about society and identity is limited. This circumstance is surprising since poetry collectors from the third century ah/​ ninth century ce considered poetry the “Archive of the Arabs” (dīwān al-​ʿarab), the primary repository of Arabian communal memories,21 and a uniquely Arabic cultural artefact: Ibn Qutayba (d. 276 ah/​889 ce) declared that “Poetry is the Arabs’ ” (lil-​ʿarab al-​shiʿr).22 Poetry therefore presents itself as the first source to pose our questions about community in early Islam. Because the amount of poetry is vast, and the issues relevant to depictions of identity are many, the following sections survey broad issues for which poetry provides pertinent insight into the nature of change amongst elite communities in early Islam. 4

Changes to Elite Communities: Poetic Indications

4.1 Terms of Communal Identity A striking aspect of Arabic poetry is the history of the emergence of the word “Arab” as an ethnonym definitive of a community. I have detailed this feature elsewhere:23 in brief, pre-​Islamic Arabic poetry is remarkable for the absence of the name ʿArab. Surviving poetry indicates that pre-​Islamic poets did not use Arabness as a term to define themselves, yet, in contrast, poets began using the term ʿArab almost ubiquitously to articulate community from the end of first century ah/​ seventh century ce onward, i.e., contemporary with the Marwānid-​era caliphate’s early articulations of its aspiring hegemonic polity across the Middle East. 21 22 23

See al-​Jumaḥī n.d., 1:24–​25; Ibn Qutayba 2017, §2.8.8, §2.8.16; al-​Yaʿqūbī 1883, 1:265. Ibn Qutayba 2017, §2.8.1. Webb 2016, 66–​70, 85–​96. Other studies have noted cases where “Arab” is used as a term of identifying groups and suggested alternative senses of Arab origins (Hoyland 2017, 126– 28; Al-Jallad 2020); nonetheless, the copious pre-Islamic poetry is essentially devoid of reference to “Arab” and this source is overlooked yet bears importantly upon the debate. Most references to “Arab” in other pre-Islamic sources are as an exonym of an “other” group; and the search for the meaning of Arab community should engage theory of ethnogenesis, constructivist or instrumentalist, to appraise the evidence which has been lacking in statements about putative Arab identity based on epigraphic and linguistic evidence alone. Theories about ethnicity do stress that a language group of itself does not necessarily coalesce into a cohesive community without other external factors.

132 Webb Von Grunebaum noted the lack of “Arabs” in pre-​Islamic poetry, yet sidestepped the issue by positing that because pre-​Islamic Arabic poets were aware of their Arab identity they had no need to call themselves “Arabs.”24 His theory, however, overlooks a crucial trend in pre-​Islamic poetry that reveals that pre-​ Islamic poets manifestly did intend to talk about their wider community, and with a particular term: they called themselves “Maʿadd.” Details of Maʿadd’s history in Arabic poetry have been traced elsewhere.25 The term appears as a byword for a broad community in nearly every collection of pre-​Islamic poetry. For an example, consider how al-​Aʿshā Maymūn ibn Qays (a northeastern Arabian poet whose poetry dates primarily to the years immediately preceding Muhammad’s prophethood) praises his people, the Qays: Our men, who, when the chargers of Maʿadd are gathered Are most respected and awed.26 When al-​Aʿshā articulates how his kin’s warriors stand above all others, we could expect (from a rhetorical perspective) that the poet would utilize a word that indicated the largest conceivable community as imagined by him and his audience. Upon the cusp of Islam, that term was “Maʿadd.” Likewise, the celebrated northern Ḥijāzi poet of the late sixth century ce, al-​Nābigha al-​ Dhubyānī, who travelled to Ghassānid, Lakhmid, and other notable groups in search of elite patronage, praised the tribal leader al-​Nuʿmān b. Wāʾil al-​Kalbī by invoking Maʿadd: You outstrip the nobles in nobility Like a stallion outstrips hunting dogs in the chase, You surpass all of Maʿadd as a patron sought and enemy feared, From the abundance of praise, you are its first recipient.27 Again, if a collective term greater than Maʿadd existed, would not the panegyrist’s voice use it instead? Similarly, al-​Akhnas b. Shihāb, poet of the northeast Arabian Taghlib tribe, marshals Maʿadd to boast his own clan’s primacy: All people of Maʿadd have their tribes And each have their safe havens. 24 25 26 27

Von Grunebaum 1963, 5–​7; see also Robin 2012, 58 and Hoyland 2015, 24–​25. See Webb 2016, 70–​77, and for a specific study of Maʿadd as an identity in pre-​and early Islam, see Webb 2021b. al-​Aʿsha 1974, 135. al-​Nābigha al-​Dhubyānī 1990, 140.

Elite Responses to the Founding of the Caliphate


But we have no mountain strongholds, Only swords of formidable repute.28 Michael Zwettler noted the use of Maʿadd in poetry, and interpreted the term as indicating militarized central Arabian groups—​“progressive northern Arabs”—​ whose adoption of the horse for warfare connected them to the Sasanian sphere of influence.29 The poets, however, never refer to their community as “Arab,” and the idea of Maʿadd spread beyond Zwettler’s model, as even non-​ horse raiding groups of the Sarāt mountains in the Ḥijāz use Maʿadd to express their sense of community. For example, the early seventh-​century ce poet of the Sarāt-​resident Hudhayl, Abū Dhuʾayb al-​Hudhalī, addresses his lover: Any woman of Maʿadd is dear to us, Yet you have been lavished with gifts!30 Similar citations of Maʿadd abound; from a rhetorical perspective, they consistently occur when poets intend to express the fullest extent of their community, or the concept of “all people,” the broadest imagined community. For the poets of the late sixth and early seventh century ce, Maʿadd was not a tribal identity: I have not found pre-​Islamic verses containing the construction “Banū Maʿadd,” and hence Maʿadd appears rather clearly to have denoted a super-​ tribal identity, and it is the only such broader identity found in pre-​Islamic poetry with any frequency and with the metaphorical rhetoric of pan-​communal inclusiveness.31 The fact that pre-​Islamic poets use the term Maʿadd in this way counters von Grunebaum’s suggestion that pre-​Islamic Arabians did not need to express their broad collective identity: they were in fact frequently 28

29 30 31

al-​Baṣrī 1999, 1:39–​40. The poem has multiple versions (see al-​Anbārī 2003, 1:513–​17 who presents the last line as “We follow the rains, conquering all in our path”), and several additional lines appear mentioning specific tribes. Later narrators may have inserted these lines to extol their own tribes and/​or patrons. In any case, the versions consistently present Maʿadd as the umbrella term used by an array of smaller groups. Zwettler 2000, 276–​80. al-​Sukkari 1963–​65, 1:88. Goldziher 1967–​71, 1:89 considered Maʿadd to be one of several super-​tribal identities, and suggested that Kinda and Ṭayyiʾ had similar connotations. Goldziher’s observation reflects its period; over the past century the collections of pre-​Islamic poetry have been improved and more material has come to light. From a survey of the now expanded corpus, we do find occasional references to Kinda and Ṭayyiʾ, but they are rare and not used in the same standalone metaphorical sense as Maʿadd. Maʿadd, moreover, can be found in almost any collection of pre-​Islamic verse, and clearly operated with grander rhetorical effect. Further research on super-​tribal names such as Kinda and Ṭayyiʾ, as well as Nizār and Muḍar, would serve to better evaluate their conceptual relationship with “Maʿadd.”

134 Webb engaged in communicating to a broad community via boasts, panegyrics, and lampoons, and they used the term Maʿadd to refer to the community of central Arabian groups who lived outside of the regular control of the Yemenis, Byzantines, Persians, and their frontier allies. The poetry also reveals that the term “Arab” lacked wide traction to unite these groups before Islam. In contrast to its absence in pre-​Islamic poetry, “Arab” appears in Muslim-​ era verse particularly in the Marwānid period. Poets from the late first century ah/​seventh century ce begin to describe their community as “Arab,” and by the second century ah/​eighth century ce virtually every poetry collection contains at least a few verses citing the name. Poets articulate the concept of “all people” with the binary term al-​ʿarab wa-​l-​ʿajam/​al-​ʿujm, and they express the sense of “all of us” with the term al-​ʿarab itself. Against the background of a community imagined as Maʿadd in pre-​Islam, poets in the Umayyad period established that their audience was known as “Arab” some two generations after the first Emigrants. As an example of the binary construction referring to “all people,” the Meccan poet Abū Dahbal al-​Jumaḥī (d. ca. 125/​743 ce) praised a nobleman of the Ashʿarī tribe, Abū al-​Fīl: Abū al-​Fīl’s virtues are innumerable, They have spread well-​known across the ʿArab and ʿUjm.32 This example contrasts with the preferred binary construction used to describe all people in pre-​Islamic poetry, which used Maʿadd, not ʿArab. Consider a pre-​ Islamic example of Abū al-​Ṭayyib ʿAbd al-​ʿUzzā’s elegy recorded in the prelude to the Prophet’s biography: Lament for him: he, the best of the barefooted and the sandal-​wearers of Maʿadd.33 Like the Islamic-​era ʿArab/​ʿAjam binary, Abū al-​Ṭayyib’s formula is a merism: the two opposites denote the totality of all people. Using the same rhetoric, Kaʿb ibn Mālik al-​Anṣārī, a Medinan poet contemporary with Muhammad, articulates the sense of “everybody” with another “Maʿaddite” merism: All of Maʿadd, their passionate and their serene (juhhāluhā wa-​ḥalimuhā) Altogether they shot at us with aggression.34 32 33 34

Abū Dahbal al-​Jumaḥī 1972, 78, see also 94. Ibn Hishām n.d., 1:175. Ibn Hishām n.d., 2:25.

Elite Responses to the Founding of the Caliphate


By contrast, in Marwānid-​era poetry, I have found only one Maʿaddite merism alluding to a sense of “all people,”35 whereas the ʿArab/​ʿAjam binary becomes common.36 “Arab” also displaces Maʿadd in stand-​ alone contexts where Marwānid-​era poets articulate a sense of their whole community. For example, al-​Rāʾī al-​Numayrī (d. 96 ah/​714 ce or 97 ah/​715 ce), who composed poetry in Basra among elite Iraqi circles, as well as northeastern Arabian tribes, boasts of his tribe’s supremacy: Numayr is the burning ember of the Arabs Burning all the brighter when war flares.37 “Arab” did not completely replace Maʿadd in the Marwānid era, however, since the name Maʿadd does appear in several contexts when poets addressed the caliphate’s military elite, and both “Maʿadd” and “Arab” continued in use through the Umayyad period.38 Yet the sudden appearance of “Arab” is striking. The stark contrast of pre-​Islamic silence on Arabness and the term’s pervasive presence beginning in the Marwānid period demands that studies of Arab ethnogenesis address why poets from the end of the first century ah/​seventh century ce would suddenly embrace the new label of “Arab.” The congruence of the emergence of express Arabness in poetry and the rise of the Marwānid caliphate has led me to propose that the rise of Arabness as an identity embraced by the caliphate’s military elite indirectly resulted from the organization of the caliphate, particularly from ʿAbd al-​Malik onward.39 ʿAbd al-​Malik faced the challenge of asserting authority over far-​flung militarized elite communities during the Second Fitna War, when rival caliphs were poised to fragment Middle Eastern political structures. ʿAbd al-​Malik defeated his rivals militarily, yet the challenge of how to maintain social authority in the wake of the war remained a problem. Mobilizing a new idea of ethnicity as Arabness, a term known to the Emigrants as the distinct and miraculous language of their proprietary scripture, the Qurʾān, logically could serve to identify the Muslim elite, particularly as a strategy to convince the militarized elite of their unity under the caliphates of ʿAbd al-​Malik and his sons.40 35 36 37 38 39 40

The Caliph Hishām is called “the Lord of Maʿaddites and non-​Maʿaddites,” i.e., “everyone” (al-​Iṣfahānī 1992, 20:408). See al-​Farazdaq 1987, 353; Dhū al-​Rumma 1972–​74, 1:23. al-​Rāʿī al-​Numayrī 1980, 18. See Webb 2021b, 80–90 for the gradual replacement of Maʿadd with ʿArab. See Webb 2016, 126–​52. For consideration of the emergence of the term “Arab” as the label for the caliphate’s elite, and difficulties with various interpretations, see Webb 2016, 110–​15, 141–​52.

136 Webb Whether ʿAbd al-​Malik’s policy was itself the driver for encouraging the notion of Arabness as the elite’s identity can be tested with further research. Whatever the precise case, a dramatic shift in the ways of imagining community occurred in the Marwānid period as a new sense of Arab community burst into poetic expression. Elements in late Marwānid Iraqi society tried to maintain Maʿaddite identity, but Arabness was ascendant, and by the ʿAbbāsid period, “Arab” displaced expressions of Maʿaddite community.41 Poetry indicates that a key communal response to the rise of the caliphate included the reorganization of disparate communities into a single entity under an “ethnic” idea of Arabness, which allowed larger numbers of groups to perceive a common identity than had ever been possible in pre-​Islamic Arabia. At the time of the first Emigrants, poets did not articulate community in terms of being “Arab,” which problematizes the familiar paradigm that labels the Emigrant migration as “Arab.” Self-​expressed “Arabs” emerge in poetry in subsequent generations, when Maʿaddite and non-​Maʿaddite peoples who had participated in the Emigrants’ wars, raids, and settlements were brought together thanks to new settlement patterns and the creation of a new military elite within the caliphate. Because some groups inside the elite, such as south Arabians, Omanis, and others, were not part of the pre-​Islamic Maʿaddite world of the central Arabian poets, the identity of “Maʿadd” would not have been readily amenable to provide a smooth transition of such groups into the new, broader constituency. In pre-​Islam, Maʿadd connoted a community of central Arabian groups; adding south and eastern Arabian newcomers would upset old balances. Would the “new Maʿaddites” hold the same status? Would they become subordinated? Would established Maʿaddite elites accept the new groups? Maʿadd’s former association with pre-​Islamic central Arabian power would stand in the way when communities had been reorganized in the amṣār and early caliphate, and it had evident disadvantages for a project that sought to unify the diverse populations of former Arabians. Hence a new identity, Arabness, could be more effective. 4.2 Spatial Considerations Through the amṣār settlements and the organization of the caliphate, the Emigrant groups (i) interacted with a broader array of peoples from various parts of Arabia and the Fertile Crescent, and (ii) became part of a geographical system larger than that of their pre-​Islamic forebears in Arabia. Given that senses of communal space and communal identity have theoretical 41

I trace the dwindling of Maʿadd in the later Umayyad era in Webb 2021b, 84–90.

Elite Responses to the Founding of the Caliphate


interconnections, the broadening of worldviews that followed the spread out of Arabia and the founding of the caliphate crossover with the sense of community, the articulations of “homeland” and the nature of communal self-​ perception would thus necessarily have expanded as a response to newfound unities between formerly distinct and distant pre-​Islamic groups. Questions therefore arise about how Emigrant populations (in their changed domiciles and with more cosmopolitan interactions) responded to the new construction of the Umayyad Middle East. Post-​emigration spatial narratives accordingly are fertile ground for investigating how the Emigrants and subsequent Umayyad-​era military elites changed as they settled in the amṣār, and as Arab ethnogenesis gathered momentum in the Marwānid period. Did the new community engender a new sense of “Arabia” to indicate a new collective homeland suitable for the newly-​formed elite collective, alongside greater awareness of pan-​Arabian geography? Poetry can assist this investigation: pre-​Islamic and Muslim-​era poets cite myriad toponyms in their verses and refer to peoples whom they imagined to be “foreign,” yet the poetic corpus has received little attention in this regard. As an example of the data provided by poetry, analysis of toponyms cited in the extant oeuvres of three pre-​Islamic poets from the late sixth century ce, Taʾabbaṭa Sharran, al-​Sulayk ibn al-​Sulaka, and Durayd ibn al-​Ṣimma, suggests that many pre-​Islamic poets did not stray far from their usual stomping grounds. Each of the three famous raiders mentions numerous place-​names in his poetry,42 but neither Taʾabbaṭa Sharran nor Durayd mention any known toponyms outside of their Ḥijāz homelands. The majority of the place-​names to which they refer were either obscure local places or locations outright unknown to the Muslim anthologists who studied the verses.43 Similarly, the collected poetry of the Hudhayl, a group in southwestern Arabia, contains over 472 place-​names across 4,600 lines of poetry, and a survey of these toponyms mirrors the findings from the individual poets noted above. Again, most place-​names are obscure or unknown, with few references to places beyond the mountains, wadis, and towns of the Ḥijāzi homeland of the Hudhayl.

42 43

In their extant oeuvres, al-​Sulayk mentions 10 toponyms, Taʾabbaṭa Sharran mentions 24, and Durayd mentions 47. In other words, therefore, place-​names occur once in every eight or nine verses of poetry. Specific analysis of these toponyms, cross-​references with geographical compendia and other poetry, and discussion of place-​names in the poetry of the Hudhayl is detailed in Webb 2021c.

138 Webb Intriguingly, more distant toponyms, such as Buṣrā in Syria and Aden in Yemen, only appear in Hudhalī poetry from the Islamic period.44 The limited geographical scope of the pre-​Islamic poets, the very scant crossover of toponyms even between poets who lived in the same region, and the lack of precise geographical knowledge about Arabia in Muslim-​era geographical compendia highlight the inferences about pre-​Islamic Arabian identity outlined above. Pre-​Islamic poetry’s spatial narratives indicate a fragmented map: groups did not necessarily travel very far from their familiar migratory paths, they did not share spatial vocabulary, and their worlds appear, in short, rather small and constricted. The representations of space hitherto uncovered indicate that poets were not possessors of pan-​Arabian horizons, harmonizing with the lack of other evidence for pan-​Arabian communal cohesion before Islam. The four groups of poetry considered come primarily from the Ḥijāz, and while further work is needed to consider whether more uniformity of toponyms existed in central Arabia, where most people of Maʿadd congregated, concepts of pan-​Arabian space do not appear in the pre-​Islamic poetry: “Arabia” (Ar. Jazīrat al-​ʿArab) is neither named expressly nor alluded to conceptually as a coherent geographical entity. Michael MacDonald’s conclusion that outsiders invented the very notion of “Arabia” as a geographical unit45 is thus supported by the poetry. Pre-​Islamic Arabians did not seem to know what “Arabia” meant and they did not use the term in poetry, notwithstanding the many other place-​names that they do identify. It follows that a pan-​ Arabian community spanning the Peninsula as we define it today was likely conceptually impossible for them to imagine. Mirroring the emergence of broader spatial awareness noted in the Hudhalī poetry from early Islam, other Arabian Umayyad-​era poets such as al-​Shammākh ibn Ḍirār (d. 30 ah/​650 ce)46 indicate similarly expanded horizons, spanning a triangle of Iraq, Syria, and Yemen.47 The later Marwānid-​era Medinan al-​Aḥwaṣ al-​Anṣārī (d. 110/​728–​29 ce) refers to places further afield, mentioning Syrian, Iraqi, western and central Arabian, and Omani toponyms.48 44

45 46 47 48

Only Mulayḥ ibn al-​Ḥakam, an Islamic-​era poet, mentions Aden (al-​Sukkarī 1963–​65, 3:1055). Maʾrib and other Yemeni toponyms appear very infrequently in Hudhalī poetry. Hudhalī poets mention Buṣrā three times: Abū Dhuʾayb (1:94 where he pairs it with Gaza), Ṣakhr al-​Ghayy (2:964), and Sāʿida ibn Juʾayya (3:1134); these three poets all lived in the early Islamic period. MacDonald 2009, 1–​2. Conflicting details surround al-​ Shammākh’s death and date. He was born during Muhammad’s early prophecy, and he participated in the Emigrants’ wars as a young man. al-​Shammākh 1977, 142–​43. al-​Aḥwaṣ 1990, 131, 132, 158, 166, 170, 181, 183.

Elite Responses to the Founding of the Caliphate


The frequency of such distant toponyms is noticeably greater than in our selection of pre-​Islamic poets, and while some pre-​Islamic poets did name distant places (al-​Aʿshā Maymūn ibn Qays is the most celebrated pre-​Islamic traveler-​ poet), survey of toponyms across the corpus points to a broadening idea of space and “home territory” as a consequence of emigration. The period when poets began calling themselves “Arabs” therefore coincides with the period when the poets became aware that the various elements of the Marwānid caliphate’s military elite shared an Arabian homeland bounded by the Red Sea, the Gulf, and the Euphrates. The Arabic conception of Jazīrat al-​ʿArab as a geographical unit may therefore be another result of Umayyad-​era elite responses to the consolidation of the Umayyad caliphate.49 4.3 The Hajj Comparison of pre-​Islamic and Umayyad-​era poetry about pilgrimage to Mecca offers another window into community from the perspective of ritual, with a perceptible change following the rise of Islam. The Hajj is mentioned in seemingly reliable and authentic pre-​Islamic poems, indicating the existence of a pilgrimage to Mecca before Muhammad which was sufficiently important for poets to marshal in oaths, and the poems indicate that the ritual focused on physical contact with Mecca’s central shrine (the Kaaba), worship of a single deity, and sacrifice. For example, al-​Nābigha al-​Dhubyānī mentions the Kaaba by name in a celebrated poem addressed to the Lakhmid king Nuʿmān ibn al-​Mundhir:50 I swear by the life of He, whose Kaaba I have touched,51 And I swear by the thick blood poured upon the sacrifice stones, By the Lord who preserves the birds in His sanctum, And the riders to Mecca, passing the wells of al-​Ghayl and al-​Saʿd.52 49

50 51 52

The Muslim-​era definition of Arabia is intriguingly fluid. Early indications suggest that the idea of Jazīrat al-​ʿArab did not extend far beyond Mecca and Medina, but the definition expanded outward to encompass the whole Peninsula by the end of the second century ah/​eighth century ce. Further study of the development of a Muslim concept of Arabia alongside the articulation of Arabness is needed: for considerations to date, see Munt 2015 and Webb 2016, 137–​38, 166–​67. al-​Nābigha 1990, 25; al-​Tibrīzī 2000, 1:528–​29 also narrates the poem. Another version of this line reads: “I swear by He whom I have visited on hajjes” (al-​ Nābigha 1990, 235). Debate surrounds these toponyms. The names appear to have been forgotten by the time Muslim collectors began commenting upon the poem in the late second century AH/​eighth century ce (see the commentary in al-​Tibrīzī 2000, 1:529). The fact that these

140 Webb Another celebrated Ḥijāzi poet of the late sixth and early seventh century ce, Zuhayr ibn Abī Sulmā, composed an oath that directly referred to the pilgrims’ campsite at Minā, as well as to another aspect of Hajj ritual—​the shaving of hair: I swore solemnly by the campsites of Minā, And by the shaven forelocks and lice-​laden hair.53 Al-​Sukkarī’s collection of Hudhayl poetry articulates similar expressions, such as the poem by Sāʿida ibn Juʾayya: I swear by my camel, and by every Sacrificial animal, covered in dust, Corralled at the narrows of Maʾzim And driven by al-​Akhshab.54 Maʾzim and al-​Akhshab are within Mecca’s landscape. Note also here the poet’s emphasis on the ritual of sacrifice. While several other pre-​Islamic verses about Meccan pilgrimage are similar to the above lines, their overall number is remarkably small within the vast corpus of pre-​Islamic poetry. Moreover, virtually all of the poets who mention the pilgrimage lived in (or, in the case of al-​Nābigha, originated from) the Ḥijāz and near Mecca.55 Poets from further afield are essentially silent on Mecca.56 To account for the discrepancy, the poetry supports a hypothesis that the pre-​Islamic Hajj was a local event particular to the Quraysh (the Quraysh are mentioned by name as custodians of the ritual in several pre-​Islamic poems), whereas the Hajj lacked importance for poets at a distance from Mecca. Muslim-​era narratives presenting the Hajj as the unifying event of the “Arabs” before Islam are accordingly anachronistic exaggerations of the ritual’s original ambit. If the Hajj had been an annual event attended by all Arabians, we could expect central and eastern Arabian poets to mention it, yet instead the

53 54 55 56

names were unknown to Muslims supports the argument that the verse was not composed in the Muslim era. Zuhayr 1982, 85. al-​Sukkarī 1963–​65, 3:1101. For a similar oath invoking the sacrifice of animals, see al-​ Sukkarī 1963–​65, 3:1172. See al-​Sukkarī 1963–​65, 1:39, 95, 144, 3:1101, 1172; al-​Anbārī, 2003, 1:431; al-​Marzūqī 1968, 4:1635; Zuhayr 1982, 23. An exception is the northeastern Arabian al-​Aʿshā, who twice mentions the Hajj: al-​Aʿshā 1974, 173, 241.

Elite Responses to the Founding of the Caliphate


poetic corpus reveals that while poets who lived near Mecca invoke the Hajj as a local ritual, even such local poets only infrequently mention it. By contrast, the Hajj becomes essentially ubiquitous in Umayyad-​era poetry. While aspects of the Hajj recall the tenor of the pre-​Islamic verses—​citation of Hajj in oaths, references to sacrifice, and emphasis on the Quraysh as the Hajj’s custodians—​Umayyad poets also put the Hajj in overtly communal contexts in which Hajj-​goers are synonymous with “Muslim” and the Hajj signifies communal solidarity. The association of the Hajj with such blatant communal connotations is evident, for example, in the poetry of Jarīr ibn ʿAtiyya (d. ca. 110 ah/​728–2​9 ce), one of the Umayyad-​era’s three most famous poets who composed poetry for Marwānid and Iraqi patrons, and among Basran tribal communities. Jarīr uses pilgrimage to disparage the kin of al-​Akhṭal (d. ca. 92 ah/​710 ce), Jarīr’s rival and a second member of the “big three” Umayyad poets (the third is al-​Farazdaq, whom we shall encounter below). Jarīr’s citation of Hajj is pertinent because al-​Akhṭal descended from the Taghlib, a Christian tribe in the Syrian desert: according to Jarīr’s lampoon, They [the Taghlib] make their Hajj to St. Sergius While the Meccan pilgrims exult praise of God.57 And similarly, Jarīr derides al-​Akhṭal’s people: Their hands never touch the Sovereign House (al-​Bayt al-​ʿAtīq)58 No, they stroke the Cross at Mass.59 Having chided al-​Akhṭal’s Christian kin for their absence at Hajj, Jarīr then praises his own people for their presence: they are the “group always camped in the plains at Minā.”60 Elsewhere, Jarīr repeats the praise of his kin using Hajj-​participation metaphors: Are we not the greatest of all people: The most noble at Minā, with the grandest tents!61 57 58 59 60 61

Jarīr 1969, 1:231. Al-​Bayt al-​ʿAtīq (the “house controlled by no authority other than God”) is a common sobriquet for the Kaaba. The Qurʾān twice refers to the Sanctum by the same term (Q 22:29, 33). Jarīr 1969, 2:840. For other examples, see 1:67, 158, 237. Jarīr 1969, 2:676, i.e., his people continuously practice the Hajj and can be located there. See also 1:248, 390. Jarīr 1969, 2:823.

142 Webb Jarīr constructs his boast by expressly referencing his kin’s preeminence at Mecca, revealing that when he sought to conceptualize the widest extent of his community’s boundaries, the Hajj presented itself as a device suitable to articulate communal demarcation, both positively for his own kin’s boast, and negatively as grounds for socially excluding al-​Akhṭal’s Christian kin. In precisely the same vein, the third great Umayyad-​era poet, al-​Farazdaq (d. 110 ah/​ 728 ce or 112 ah/​730 ce), refers to the Hajj in verses about communal inclusion and exclusion.62 The notable shift here is the emergence by the Marwānid period of the notion that Hajj participation of itself can operate to identify members of the Muslim community. The Hajj seems to have conceptually expanded from the domain of localized ritual into a very broad socio-​political function that signified loyalty to the caliph, the Quraysh, and membership within the Muslim community itself. That the Hajj adopted such a function intriguingly indicates the way in which its community of participants imagined their identity. Belonging to the “inside” of the group meant participating in a religious rite controlled by the ruling elite (the Quraysh), so that communities who had participated in the wars and raids of the Emigrants but did not participate in the religious rites of Hajj, such as al-​Akhṭal’s Christian Taghlib, could be chastised by Marwānid-​ era rivals as outsiders. The emergence of the expressly religious activity of Hajj as a marker of communal identity and the timing of this emergence in poetry of the Marwānid period again highlights that the end of the first century ah/​­seventh century ce was a seminal period for defining Umayyad elite social groups whereby religious affiliation became directly relevant to communal standing. The Hajj therefore straddles the continuum of “continuity” and “change” inasmuch as Muslims perpetuated the pre-​Islamic ritual with its same custodians, sacred locations, and rites, whereas by the Marwānid era, a reinterpretation of the Hajj emphasized its community-​defining aspect as a cornerstone of Muslim collective identity. The social role of the Hajj aligns with impressions that ʿAbd al-​Malik’s caliphate was a watershed moment in the development of Muslim orthodoxy, or at least the official endorsement of recognizably coherent boundaries of Muslim practice tied to the establishment of the Muslim communities as the caliphate’s elite.63 The emergence of the Hajj as a faith-​based marker of social belonging at this time starkly contrasts with pre-​Islamic poetry, where participation in religious ritual is muted and not tied to a sense of group identity. 62 63

al-​Farazdaq 1987, 34, 60, 108. Here, the non-​Hajj-​goers are called the Ṭayyiʾ. Donner 2010 emphasizes the role of ʿAbd al-​Malik in reorienting the parameters of the military elite community.

Elite Responses to the Founding of the Caliphate


4.4 Islam and Poetry Related to the new emphasis in Marwānid-​era poetry conjoining the Hajj and communal belonging, religious expression generally also exhibits analogous changes. Probing the nature of religious expression in Arabic poetry broaches questions that have long concerned scholars: How does “religion” manifest itself? What aspects of apparent religious expression in poetry can be linked to changes in belief systems? How can we discern the impact of Islam on poets of Muhammad’s generation and during early Islam? Generally, scholarship embraced the view that pre-​Islamic poetry contains little religious sentiment.64 Indeed, poetry infrequently named idols, and we have seen that pre-​Islamic verse only rarely attests to the Hajj; overall the kinds of poems associated with late antique religious groups, such as homilies or hymns, clearly did not feature in pre-​Islamic poetry. More recent work does not outright categorize the pre-​Islamic poetic corpus “profane”65 since poets contemporary with Muhammad, such as Umayya ibn Abī al-​Ṣalt and al-​Samawʾal, do engage mythico-​religious themes that touch on questions of faith and invoke similar topics found in the Qurʾān,66 and Muslim-​ era memories of some companions of the Prophet, such as Ṣirma ibn Abī Anas, note their shunning of polytheism before Muhammad’s mission, alongside some verses with religious tones.67 Nonetheless, religious themes are generally muted: the poets of Maʿadd often invoke a deistic notion of “the Fates” (al-​Manāyā) to explain the inevitability of Destiny, but even if the Maʿaddites worshipped these “Fates” as gods, their poetry makes little reference to trappings of systematic religious practice or ritual. Therefore, the poetry leaves us guessing about the precise beliefs of the Maʿaddites and how they might have practiced religion. Christian Arabian communities present a separate conceptual puzzle. Since the late nineteenth century, some scholars have maintained that many central Arabian Maʿaddite groups were Christian, however the collections of these purported “Christian poets,” such as that compiled by Cheikho, do not record 64

65 66 67

Brockelmann 1922 offered an early observation about the lack of religious sentiment in pre-​Islamic poetry. Montgomery 1997, 219–​22 raises important questions over such issues, and (221) he cites Wagner as an example of the traditional approach: “[I]‌n pre-​Islamic poetry, religion was accorded a minimal role. It was not part of the thematic register which poetry of an Islamic inspiration could replace.” See Dmitriev’s critique of the traditional view: 2017, 105–​6. For analysis of these poets, see Sinai 2011 and the summary of previous research on Umayya (398–​400); Dmitriev 2017. Ibn ʿAbd al-​Barr 1992, 2:747–​48. The historicity of such poetry and tropes of pre-​ Muhammadic believers needs further research.

144 Webb verses that expressly refer to Christianity.68 Broader searches of Arabic literature do reveal enigmatic Christian memories, such as a verse attributed to one ʿAmr ibn ʿAbd al-​Jinn al-​Tanūkhī: I swear by what the priests sanctify in all those sacristies (haykal) By Abīl al-​Abīliyyīn, Jesus son of Mary!69 The name Abīl is remembered variously as the “head of the Christians,” or a specific priest, or Jesus himself.70 The variations suggest “Abīl” was an ancient word associated with a form of Arabic Christianity largely forgotten in the Islamic period, but the poet who used the word, Ibn ʿAbd al-​Jinn, is cast back in an archaic history, and is almost entirely unknown,71 and the above lines comprise one of only four verses attributed to him. It is difficult to make conclusions with the available scarce evidence: beyond these unusual lines, Christian Arabic-​speaking communities in pre-​Islamic Arabia and the Syrian Desert did not leave an extant corpus of pietistic poetry or panegyric/​hagiographical material devoted to monks or church leaders. Within discussions of Christian poetry, ʿAdī ibn Zayd is sometimes presented as a poet-​spokesman of the Christian ʿIbād in al-​Ḥīra,72 though his religious poetry is one of the types of pre-​Islamic verse which arouse particular suspicions as to authenticity. In the early third century ah/​ninth century ce, the first known generation of pre-​Islamic poetry compliers voiced concerns about ʿAdī’s work: Ibn Sallām al-​Jumaḥī (d. 231 or 232 ah/​845 or 846 ce) notes 68

69 70 71 72

See Cheikho 1926. Cheiko classifies poets as “Christian” based on their tribe and a generous assumption about the number of Christian pre-​Islamic tribes. However, the poetry across the 800-​page collection does not express Christian themes. Shahid 2009, 295–​96 discusses Christian material in pre-​Islamic Arabic of Syria, but his description of a thriving corpus of Christian texts in Arabic follows his interpretation of two words in the poetry of al-​Nābigha al-​Dhubyānī, and the argument is framed as a “must-​have-​been,” without further evidence. His hypothesis that Christianity influenced Arabic poetry is likewise rather “must-​have-​been.” I would argue that his suggestion that references to monks and lamps in Arabic poetry imply Christian influence does not adequately demonstrate that poets were themselves Christian and influenced by Christian ideas (1989, 440–​41). Shahid himself admits that his proposal about the versification of the Bible is “pure speculation” (1989, 440). al-​Marzubānī 2010, 1:40–​41. Ibn Manẓūr 1994, 11:6–​7. This poet is also missing from Ibn Jarrāḥ’s (d. 296 ah/​909 ce) encyclopedic compendium of poets named ʿAmr, suggesting that the poet was not well remembered among Muslim-​ era poetry specialists. For the Christian community of the ʿIbād, see Toral-​Niehoff 2010, 333, who hesitates to assert the authenticity of the poetry.

Elite Responses to the Founding of the Caliphate


that “the narrators know little” of his oeuvre, and that much of the poetry attributed to him in Muslim-​era Iraqi literary circles should be considered spurious.73 Ibn Qutayba (d. 276 ah/​889 ce) more severely stated that “a very great detail of poetry has been forged in ʿAdī’s name, and our scholars (ʿulamāʾnā) do not narrate his poetry as philological proof (ḥujja).”74 Indeed, the style of his poetry containing biblical allusions notably differs from the pre-​Islamic Arabic verse in the major collections.75 The verses with biblical allusions ascribed to ʿAdī may be of the sort critiqued by the early poetry specialists, and hence the Christian sentiment should be understood as the result of Muslim-​era inhabitants of Kufa back-​projecting ideas already digested in exegesis of the Qurʾān onto the pre-​Islamic character of ʿAdī ibn Zayd. The verses too contain seeming anachronisms: for example, references to the collapse of the Byzantine (Banū al-​Aṣfar) and Persian empires as part of the ubi sunt motif more likely reflect worldviews of Muslim-​era poets, and references to characters in the Qurʾān, such as Tubbaʾ, also suggest that the verses emerged after a period of Muslim exegesis of the Qurʾān.76 Cheikho also noted the extent of spurious verses ascribed to ʿAdī.77 Intriguingly, the poems that are more widely accepted as genuinely by ʿAdī, and that circulated in early poetry collections, do not contain any biblical material or allusion. I have found only one line where ʿAdī swears by Abīl,78 but this line has varied recensions, and only one recension has an overtly Christian reference, casting some doubt on the “original” form of the verse composed by ʿAdī. Overall, ʿAdī’s poetry as preserved in the main collections speaks of “the Fates” (al-​manāyā) in terms very similar to the Maʿaddite poets noted above.79 73 74


76 77 78 79

Ibn Sallām n.d., 1:137, 140. Ibn Qutayba 1982, 1:219. Ibn Qutayba also notes that ʿAdī’s residence in al-​Ḥīra affected his language, which thus lost the “pure Bedouin” characteristic of other pre-​Islamic poets, an opinion he ascribes to the earlier Abū ʿUbayda (d. ca. 209 ah/​824–​25 ce). While perhaps a trope of ʿAbbāsid-​era taste, Abū ʿUbayda comments that “no Arabs narrate his verses because his language is not Najdi” (1:219). The other extant early commentator on poetry, Abū Ḥātim al-​Sijistānī (d. 255 ah/​869 ce), likewise decries his poetic prowess (1991, 114). These verses may have contributed to the poor opinion of ʿAdī’s non-​Najdi Arabic, but the surviving verses in Ibn Qutayba 1982, 1:219–​27 and Cheikho 1926, 439–​74 contain verses that do not radically diverge from Maʿaddite styles, however, these do not contain the expressly religious material. For the verses, see Cheikho 1926, 456, 458, 468. Cheikho 1926, 472 narrates a verse with the term kitāb Allāh (the Book of God): this intriguing reference could be a Muslim-​era projection, but further work is needed for its evaluation. Cheikho 1926, 465. Cheikho 1926, 453. Abīl is presumably Jesus, as discussed above, note 70. See Cheikho 1926, 442–​43, 468 for examples.

146 Webb To pursue the issue of authenticity, it is relevant to note that the third-​ century ah/​ninth-​century ce compilers of material about ʿAdī imagined the ʿIbād of Kufa as true “Arab Christians,”80 and because those Muslim-​era writers also associated Arabness with poetic composition, their worldview required them to reconstruct ʿAdī’s career via poetry, so that new poetry would then need to be created to sustain an acceptably Arabized narrative for the history of the Christian ʿIbād. In support of this hypothesis, the poetry with biblical references ascribed to ʿAdī is stylistically rather analogous to the large quantities of poetry forged by early ʿAbbāsid-​era narrators and spuriously attributed to pre-​ Islamic Yemenis to prove the Arabness of ancient southern Arabians. From the perspectives of language, style, and allusions to Qurʾānic material, ʿAdī’s religious poetry resembles these “Yemeni” verses,81 suggesting that the real pre-​ Islamic ʿAdī’s poetry likely had the same tenor as most other pre-​Islamic poets, and likely did not include overtly Christian references.82 The problematic character of ʿAdī’s verse with biblical allusions, the impression shared among early scholars that very little genuine poetry of ʿAdī survived, and the absence of Christian references in the verses that are considered authentic limit the extent to which a tradition of Arabic Christian literature can be postulated before Muhammad. If some poets (Maʿaddite and otherwise) were Christian, their faith left very faint marks in their poetry. Turning to the Islamic period, substantial research has investigated whether Arabic poetry changed as poets converted to Islam. While some scholars suggest a deviation from pre-​Islamic norms,83 the more common opinion holds that Islamic belief did not radically alter poetry.84 Poetic allusions to Islam are indeed indirect, but it is significant that studies largely focus on literary perspectives and such research into styles and motifs does not necessarily intersect with the paper’s questions about identity and community. The presence 80 81 82

83 84

As al-​Ṭabarī 1960–​69, 2:42–​43 makes explicit. See especially the “pre-​Islamic Yemeni” Arabic poetry in al-​Khuzāʿī 1997. Pseudo-​Wahb 1996 includes plentiful other poetry. Third-​century ah/​ninth-​century ce scholars criticized these forgeries (al-​Jumaḥī n.d., 1:10–​11). In some lines, ʿAdī mentions God (Allāh) and expresses monotheistic ideas (see Cheikho 1926, 454, 468), but these examples hardly differ from similar sentiments in other Maʿaddite poetry (the presence of Allāh in pre-​Islamic poetry is the subject of Brockelman 1922), and thus do not betray any specifically Christian influence. See Jacobi, 1960–​2007. Shahid 1989, 439 states that the rise of Islam “immediately affected the warp and woof of Arabic poetry,” though without discussion of the nature of the change. Montgomery 1997, 219–​20 discusses previous scholarship and draws similar conclusions that Islamic elements seamlessly mix with pre-​Islamic conventions without overtly privileging the Islamic (245–​53). See also S. Stetkevych 1994 and J. Stetkevych 2006.

Elite Responses to the Founding of the Caliphate


of religious themes in Islamic-​era poetry does reveal changes, and while more comprehensive survey is needed, here I outline some intriguing leads. Poetry from the generation of the original Emigrants refers to the name Allāh as the monotheistic deity. We have noted that Allāh also denotes God in pre-​Islamic poetry, but the early Islamic-​era verse clearly quotes from and reflects Qurʾānic language, indicating that scripture influenced poetic production. Modern poetry specialists have noted such Islamic influences, but they emphasize the “co-​existence of pre-​and early Islamic ideas and values.”85 In light of developments in our understanding of Islam’s history, this view could have been anticipated. Historians now stress that Islam did not enter history as a fully codified system,86 so that followers of Muhammad and the first caliphs cannot reasonably be expected to have transformed completely upon “conversion.” A mixture of old conventions influenced by new faith is likely to be as much as can be expected at such an early period, and the glimpses of a tentatively emerging Islam into poetry therefore support the historians’ thesis of Islam’s gradual maturation. Consider ʿAbda ibn al-​Ṭabīb’s invocation of God in his understanding of the sustenance of his nomadic lifestyle: Our hope is for the leftovers of the bounty Of a most generous Lord, all benefit is within Him, all is welcome! He is a Lord who granted us our possessions, All He bestows is sustaining.87 ʿAbda articulates straightforward monotheism within a poem otherwise replete with tropes familiar from pre-​Islamic convention. This complexity may indeed reflect the tenor of the nascent faith that an Arabian nomadic member of the early community was required to absorb: ʿAbda’s lines reflect awareness of the single deity, and the language in which he couches his “bestowing” Lord aligns with the signature Qurʾānic method of articulating God. Therefore, stylistically, ʿAbda’s line expresses the divine in a way that differs from pre-​ Islamic poets, but his verses do not intone more complex theology. Generally, most collections of early Islamic poetry contain at least some similar quotations, allusions, and resonances of the Qurʾān.88 Qurʾānic language clearly and 85 86

87 88

Montgomery 1997, 250. Goldziher 1890 and Izutso 1966 present theses about Islam as a program of moral reform that held considerable sway in mid-​to late twentieth-​century scholarship, prompting scholars to expect that early Islamic Arabians should have radically changed their ways after Muhammad. al-​Anbārī 2003, 1:368. For a discussion of the Qurʾān in early poetry, see Kadi 1994; Montgomery 1997, 230–​50.

148 Webb rapidly penetrated Arabic poetry, but the extent to which the Qurʾān, belief in the monotheistic God, and specific rituals combined to constitute a cohesive community of believers at such an early period poses questions of a different type, and this issue remains unresolved. As for other religious communities in early Muslim-​ era poetry, al-​ Shammākh ibn Ḍirār describes “the Christians” (al-​Naṣārā) as walking in black slippers in a metaphor,89 presenting them as a monolithic group that is visibly different and conceptually separate from al-​Shammākh’s own community.90 It is intriguing that al-​Shammākh uses a religious label to differentiate social groups: does this label indicate that the poet used faith to categorize others as well as to define his own identity within the fledgling caliphate?91 The character and communal notion of al-​Naṣārā as an “other” (an identity opposite and separate from “Muslim”) needs further investigation. In broad terms, ʿAbda, al-​Shammākh, and other early Muslim-​era poets such as Labīd, composed poetry in which they express their personal submission to a Qurʾānic-​sounding God, whereas poets praise the polity established by Muhammad in Medina as a political entity capable of military action, rather than as an apocalyptic group leading its believers to paradise. Kaʿb ibn Zuhayr’s celebrated ode to the Prophet Muhammad is much discussed for his emphasis on Muhammad and his community as a war band,92 and when al-​Shammākh refers to the caliphate of ʿUthmān ibn ʿAffān, he uses decidedly secular terminology too: Were not Ibn ʿAffān and the Government (al-​sulṭān) watching, I’d lampoon you hard!93 The notion of modern historians that the early caliphs projected their role via the lofty title of “God’s Successor on Earth” (khalīfat Allāh)—​the individual charged with the community’s salvation—​evidently is not part of al-​ Shammākh’s conception of authority here. As a voice from the generation following Muhammad, therefore, al-​Shammākh presents us with a nomadic poet from northeastern Arabia who knew enough of the Qurʾān to internalize

89 90 91 92 93

al-​Shammākh 1977, 83. Ibn Qutayba 2011, 1:259 explicitly argues that “Christian” is rhetorically structured as distinct from al-​Shammākh’s “Arab” people, who wore undyed leather sandals; this conclusion is logical. Elsewhere, al-​Shammākh invokes the names of “othered” communities of Hebrews (ʿibrānī) and Persians (al-​fārisī) (1977, 94, 124). See Sells 1990; Stetkevych 1994; Montgomery 1997, 222–​24. al-​Shammākh 1977, 122.

Elite Responses to the Founding of the Caliphate


its language occasionally in his poetry,94 but he neither refers to himself as a “Muslim” nor muses at length on Muslim dogma and identity. Al-​Shammākh refers to the nascent caliphate as a worldly power, yet he depicts “Christians” as a different community, and, by all accounts, he became an Emigrant and participated in combat in Iraq and Armenia. In the Marwānid era, the styles of poetry and references to the Qurʾān are similar to the above verses, but religious themes appear with clearer focus on matters related to status and community. More comprehensive comparison is needed before adducing firm conclusions, but a survey of Marwānid-​era poetry indicates increased religious allusions in lampoons and praise. For example, the Iraqi-​ domiciled Ziyād al-​Aʿjam95 (d. ca. 100 ah/​718 ce) ridicules his rivals by describing them with phrases such as “Their people are pigs of Iraqi peasants, shunned and impure,”96 or by chastising a rival’s mother’s vagina for being uncircumcised.97 He scorns the Jarm (an ancestral group of the Quḍāʿa who lived in the Syrian desert at the beginning of Islam) for continuing to drink wine “after its prohibition was announced.”98 Against the Yashkur (a major ancestral group of the Bakr ibn Wāʾil, also from the Syrian desert near Iraq), Ziyād is particularly uncharitable: If the robe of a Yashkurī should touch yours, Then you must not mention God’s name until you wash!99 94 95

96 97

98 99

See, for examples, al-​Shammākh 1977, 71, 142, 228. Ziyād al-​Aʿjam translates literally as “Ziyād the Non-​Arabic,” which is an unusual sobriquet for an early Arab poet! Aʿjam is a complex term with various meanings, one of which indicated residents of Khurāsān (easternmost Iran). Ziyād’s nickname may have signified his eastern birth, but it may also mean that his accent differed from the Iraqis (aʿjam also means non-​proper-​Arabic speaker, not non-​Arab per se; ʿajam became the preferred term to describe people of non-​Arab descent). Anecdotes about Ziyād show that he considered himself a full-​fledged member of the “Arab” community through his tribal affiliation and membership within the military elite groups who controlled the land between Khurāsān and Basra, although some of his poet-​rivals ridiculed him for his accent and eastern birth. For a detailed analysis of the anecdotes, see al-​Aʿjam, Dīwān 15–​22. al-​Aʿjam 1983, 67. al-​Aʿjam 1983, 64. Whether female circumcision is an “Islamic” practice is the subject of current polemical debate and anthropological research (for a balanced approach to the premodern sources, see Berkey 1996). Ziyād’s poem makes it clear that from his perspective, people who do not circumcise their girls were “outsiders” apart from his own community. The other examples from Ziyād’s poetry listed here reveal how Islamic taboos and injunctions could delineate communal boundaries. While Islamic law is equivocal on female circumcision, I consider it reasonable that in the first century ah/​seventh century ce Ziyād would consider it an important custom for his notion of Muslim/​Arab community. al-​Aʿjam 1983, 87. al-​Aʿjam 1983, 69.

150 Webb And he unequivocally derides Yazīd ibn Ḥabnāʾ: You have abandoned devoutness: religion is Muhammad’s For the people of devotion, for the Muslims.100 What emerges from Ziyād al-​Aʿjam’s collection is an instrumental use of Islam in the context of social interaction. Ziyād does not mean that his foes’ souls are in peril (his poetry is devoid of eschatology), but rather he casts them as politically and socially inferior because of their failure to meet an Islamic normative standard. Breaches of Islamic codes and engagement in Muslim taboos evidently served as an effective rhetoric for maligning foes, and I have not found such sentiments in earlier poetry. As such, it does not appear that earlier targets of poetic lampoon could be chastised for religious failings. For Ziyād’s rhetoric to be successful in his Iraqi environment, his target audience would need to have possessed awareness of a common set of religious norms expected from those inside the elite community, and to have adopted a worldview whereby those norms spoke to the heart of their self-​identity. Therefore, it seems that by the Marwānid period Muslim mores had rather well entrenched themselves as a basis for social inclusion; it is noteworthy that Ziyād was contemporary with Jarīr and al-​Farazdaq who likewise used the Hajj ritual to exclude Christian groups from the elite “inside” community. Moreover, Ziyād’s rhetoric is not unique: the Ḥijāzi resident al-​Aḥwaṣ al-​ Anṣārī chided an Umayyad appointee over Medina with a verse referring to the trappings of Islamic belief: I’m amazed to see Ibn Ḥazm mount a donkey, Yet more amazing still, he mounts the pulpit (minbar)101 Al-​Aḥwaṣ does not mean that Ibn Ḥazm is not Muslim, rather, it is Ibn Ḥazm’s inability to demonstrate the expected gravitas of sermon-​giving that forms the source of critique. Pre-​Islamic slander does not operate within this type of framework with a faith-​oriented tenor.102 In contrast, but with the same aim, Marwānid-​ era poets also utilize Islamic allusions for praise, whereby the upholding of Islamic values or the 1 00 al-​Aʿjam 1983, 50. 101 al-​Aḥwaṣ 1990, 93. See also 105. 102 For representative examples of the tone of pre-​Islamic hijāʾ (lampoon) with its focus on individual virtue as opposed to religious norms, see van Gelder 1988, 15–​20.

Elite Responses to the Founding of the Caliphate


performance of Muslim religious practice is correlated to social excellence. For example, Ziyād al-​Aʿjam praises the governor of Nishapur, ʿAbd Allāh ibn al-​Ḥashraj, in a four-​line fragment: the first two lines copy the established pre-​ Islamic poetic motifs of praising patrons for their generosity, virtue, hospitality, and “crowned kingship” (malikun agharru mutawwajun), but the third verse expresses a new sentiment. He is the most devout to ascend the pulpit (minbar) Second only to the sinless chosen Prophet.103 As in the above example by al-​Aḥwaṣ, mounting the pulpit to give the Friday sermon was the prerogative of Umayyad-​era governors, and here Ziyād uses the device to praise his patron. In the Ḥijāz, the poet Abū Dahbal al-​Jumaḥī praised Marwān ibn al-​Ḥakam in a similarly constructed two-​verse poem: the first line uses the pre-​Islamic praise language, but the second associates the new caliph’s rise to power with Moses’s command to his people to cease worship of the Golden Calf.104 Poetry specialists have already noted this duality of pre-​Islamic and Muslim norms in poetry, however, it is important also to highlight that citation of religious norms to establish social standing is a distinctly novel addition in Muslim-​era poetry. Ziyād also refers to Islam explicitly in a possessive way: “our religion,”105 which appears to be a new expression that emerges in the Umayyad era. Separate from questions of personal piety, therefore, religion emerges as a tool in social relations. For such verses to have been composed, the society as a whole would have been using faith as a means to self-​define and to imagine communal boundaries. Marwānid-​era poets were not necessarily any more devout than their forebears, nor did poetry become a medium for dedicated devotional expression at any point during the Umayyad era. Muslim poets perpetuated most pre-​Islamic tropes and metaphors, hence the rise of Islam itself cannot be said to have changed poetic composition, but Marwānid-​era poets did assign religious matters a social function that is absent in previous Arabic poetry. This observation fuels inference that the consolidation of the caliphate did change community, such that poets began using religion as a primary means to define meaningful social boundaries.

1 03 al-​Aʿjam 1983, 49. 104 Abū Dahbal 1972, 80. 105 al-​Aʿjam 1983, 56.

152 Webb The “religious” indicators chronologically align with the emergence of the self-​expression of people as “Arab” in poetry. Thus, from the perspective of communal identity, processes definitive of what would become iconic aspects of Muslim culture began to mature during the Marwānid period: the emergence of Arabness as an ethnic identity and Islam as a communal creed. The transformation of Arabness and Muslimness into markers of an elite identity stands as a seminal response of the Marwānid-​era military elites. 5

An Arabness Case Study: al-​Azd in Umayyad Politics

The above analysis of poetry offers important insights into understanding the military elites’ responses to the rise of the caliphate, including the process of ethnogenesis that reorganized the original Emigrants into an Arab ethnos. Because identities are social constructs, we should be able to detect fissures and contradictions when the array of formerly distinct peoples began to embrace Marwānid-​era Arabness as their identity. This final section explores the ramifications of Arab ethnogenesis using the case study of al-​Azd. In models of Arab genealogy developed during the early ʿAbbāsid era, al-​Azd were classified as a large tribe of “southerner Arabs” (al-​Yamāniyya), comprising three sub-​groups: one “eastern” group of the Azd ʿUmān and two “western” groups, the Azd Shanūʾa and Azd al-​Sarāt. ʿAbbāsid genealogists considered al-​Azd as among the genealogically purest Arabs: as members of the “southerner Arab” lineage, they were understood to be descendants of the “Arab Arabs” (al-​ʿarab al-​ʿāriba), the “original Arabs” whose blood was considered unmixed with that of non-​Arab peoples.106 The ʿAbbāsid genealogical models posited a common descent from four sons of an ancient eponymous ancestor, al-​Azd ibn al-​Gawth ibn Nabt, to unify various geographically dispersed peoples as “Azdīs”: Omanis (Azd ʿUmān), various ancestral groups in the Sarāt mountains of south-​central Arabia, the two main tribes of Medina, and the Ghassānids and Lakhmids (rulers of the Syrian and Iraqi deserts in the pre-​ Islamic period).107 Most of the Omani and Sarāt mountain al-​Azd groups were classified as descendants of Naṣr ibn al-​Azd, while the “nobles” of Ghassān and Lakhm, the Medinan Anṣār (the Aws and al-​Khazraj clans who comprised the first community that accepted Muhammad as their Prophet and leader), and 106 For al-​Azd’s genealogy, see Ibn al-​Kalbī 1988, 1:362–​508; Ibn Durayd n.d., 435–​95; for the “Arab Arabs” and the history of the idea in Arabic sources, see Webb 2016, 217–​22. 107 Genealogists note that al-​Azd was originally pronounced al-​Asd (Ibn Durayd n.d., 435); this accords with a tribal name appearing in pre-​Islamic inscriptions.

Elite Responses to the Founding of the Caliphate


the Muhallabid family (which claimed nobility over the Omani al-​Azd who had settled in Basra during the Umayyad era) were each identified as scion of Māzin ibn al-​Azd.108 While ʿAbbāsid-​era constructions of Arab genealogy grant unequivocal Arabness to al-​Azd, signs of the genealogies’ fabricated nature can be detected. The establishment of southerners as “Arab Arabs” at the top of an Arab family tree was almost certainly an invention of ʿAbbāsid-​era scholarship and bore little relation to how Umayyad-​era southerners imagined their identity, let alone pre-​Islamic groups.109 On the level of al-​Azd genealogical construction, the markedly disjointed genealogies connect widely dispersed groups, suggesting that various distinct entities in the early Islamic period had their ancestries rationalized to fit them within one Azdī identity. To explain the shared Azdī roots of such disparate groups, Muslim historiographers developed a migration story in which all the al-​Azd subgroups lived in Yemen in an ancient period until their leader, ʿAmr ibn Muzayqiya, was inspired to learn that a dam in Maʾrib was about to burst and flood his lands. Through a clever ruse he was able to sell out to unsuspecting locals shortly before the dam broke, and he led his people northward. Along the way, subgroups separated themselves from the migration at different points to settle new lands. Brian Ulrich has argued that the migration story is an anachronistic elaboration, constructed to retrospectively fit the various al-​Azd groups into one shared past.110 The articulation of “noble” Azdī lineage through the ancestor Māzin ibn al-​Azd, in contrast to the more numerous tribal blocs considered descendants from Naṣr ibn al-​Azd, is also curiously binary and streamlined. The genealogical issues led Crone and Blankinship to propose that the early caliphate created the entire existence of al-​Azd: a military (not genealogical) designation assigned to unify an array of people after they settled in the

108 Al-​Azd’s genealogical tree was established by the end of the second century ah/​eighth century ce: see the earliest version in Ibn al-​Kalbī (see note 89). Subsequent genealogists repeat the same genealogical model (e.g., Yāqūt 1987, 86; Ibn Ḥazm 1994, 330–​86). For discussion, see Webb, “al-​Azd,” ei3 and Ulrich 2019, 30–​48. 109 See Webb 2016, 213–​22 for the process by which southerners placed themselves at the top of the Arab family tree; and Webb 2021a for specific study of southerner Arabness as expressed in early ʿAbbāsid Iraq. Notwithstanding the complex nature of ethnicity in pre-​ Islamic Arabia, modern scholarship generally concurs that the pre-​Islamic south Arabian kingdom of Ḥimyar did not consider itself part of a wider Arab community, and their inscriptions express “Maʿadd” as an outsider group. Hence the “pure Arabness” of Yemenis in Muslim times was a novel construction. 110 See Ulrich 2008 for the migration mythology, and Ulrich 2019 for a broad study of how individual families manipulated al-​Azd genealogy to invent their own tribal identities.

154 Webb amṣār towns.111 In defense of the ʿAbbāsid-​era genealogists, pre-​Islamic epigraphy does include the name “As1dn” as a southwestern Arabian group located roughly within the homeland that Arabic historians call the Azd Sarāt.112 The name As1dn translates into Arabic as al-​Asd/​al-​Azd, but pre-​Islamic As1dn’s precise geographical extent and subdivisions remain unclear. Like many ancestral groups, al-​Azd’s boundaries were fluid and the social changes following the rise of Islam likely reshaped its identity, admitting new members and extensively reorganizing genealogical trees. The construction of a new al-​Azd genealogy encompassing Omanis, western Arabians, central Arabians, and even groups in the Syrian desert appears convincingly to be the result of Umayyad-​era politics; we can push further by investigating the construction of the very Arab identity of al-​Azd during the same period. The al-​Azd participated in the wars and raids of Muhammad’s successors, placing them at the epicenter of Emigrant settlement, the maturation of the caliphate, and subsequent Arab ethnogenesis. However, the Omani al-​Azd did not fit easily into Arabness, and these difficulties have attracted some attention in the study of the Muhallabids, a Basran family with roots from Oman. Their ancestor, Abū Ṣufra, led Omani contingents in invasions to southern Iran, then he settled in Basra. His son, al-​Muhallab ibn Abī Ṣufra (d. 83 ah/​ 702 ce), became a military leader during the mid-​first century ah/​seventh century ce,113 and the family’s third-​generation patriarch, Yazīd ibn al-​Muhallab, considered himself powerful enough to challenge caliphal authority. Yazīd was defeated and killed in 101 ah/​720 ce, but the family remained influential, and the Muhallabids and al-​Azd held positions at the top of the Basran social order during the ʿAbbāsid period. Some Arabic sources claim that Abu Ṣufra had a Persian origin,114 but the sources invite us to dig deeper because all Basran groups hailing from Oman and claiming Azdī ancestry appear to have been chided in the Marwānid period for being non-​Arabs. Arabness was thus not merely problematic for the Muhallabids—​other genealogies of the Azd ʿUmān 1 11 Crone 1980, 30–​31; Blankinship 1994, 44. 112 For further details of the pre-​Islamic al-​Asd/​al-​Azd, see Webb “al-​Azd,” ei3, and Ulrich 2019, 25–​28. 113 Arabic sources identify him as leader of al-​Azd, but Crone rejects this “al-​Muhallab b. Abī Ṣufra,” ei2. However, in what follows, we will see a strong connection between al-​Muhallab and Azdī tribal identity. It is likely that the Muhallabid family lacked Azdī roots, but perhaps many groups in Basra who claimed to be al-​Azd likewise hailed from different lineages, and in the Umayyad period they created a new Azdī lineage to organize themselves. 114 al-​Iṣfahānī 1992, 20:85–​88. The sources are not unanimous, however: see al-​ʿAsqalānī 1995, 9:185. See Hinds 1981 for discussion, and now Ulrich 2019, 117–​18.

Elite Responses to the Founding of the Caliphate


were involved, which adds perspective to Ulrich’s observation that the migration story used to explain how Omanis are Arabs from al-​Azd is a Muslim-​era fabrication designed to prove the Omani Azd’s connection to a wider Arab family tree.115 Proving Arab identity evidently challenged the Basran-​resident Azd ʿUmān, and their story presents us with the processes by which tribes “became Arab” in the Umayyad era. In what follows, we consider the evidence. The first piece of evidence comes from a poem by al-​Farazdaq, composed as a lampoon of the Muhallabid patriarch al-​Muhallab ibn Abī Ṣufra. Al-​Farazdaq worked during the early Marwānid period, contemporary with ʿAbd al-​Malik, when poets used the term “Arab” to speak about their own community. References to “Arabs” occur frequently throughout al-​Farazdaq’s poems when he speaks of his identity and other Umayyad elites, but in his treatment of al-​ Muhallab ibn Abī Ṣufra and al-​Azd, his tone differs. Al-​Farazdaq ran afoul of al-​Muhallab, and while al-​Muhallab had the support of the Umayyad governor of Basra, al-​Farazdaq nonetheless composed a lampoon of al-​Azd: Were it not for the strong-​arm of Bishr ibn Marwān,116 I’d have no concern for the brooding anger of al-​Muhallab. Let him shut the door in my face and hide, Neither of my parents stood by a jand tree!117 My clan are from Mecca and Ṭāʾif, They clambered down no Omani wadis. Nobles from Qays and Khindif: should I call out, They run to aid one waving his robes in distress. When I see al-​Azd, it’s their beards that flap About a ship’s deck, a miserable vessel. How I marvel to see them swap an anchor chain for reins,118 Who ever heard of such absurdity? They hide their noses—​not Arab noses those, Beards of Nabaṭīs,119 no clear Arabic from their mouths! How could they be—​they made no sacrifice at Mecca, They never worshipped idols at al-​Muḥaṣṣab.120 1 15 116 117 118 119

Ulrich 2008. The Umayyad governor of Iraq, appointed in 74–​75 ah/​693–​94 ce. The jand (ghāf) was a tree indigenous to Oman, the homeland of the Muhallabids. Here al-​Farazdaq chides al-​Azd for only recently becoming cavalrymen. In Umayyad-​era Arabic poetry, nabaṭ served as the preferred term to label the indigenous population of Iraq. At the time when Arabness became a potent marker for elite identity, nabaṭī took on the meaning of non-​Arab, much like the use of ʿajam and aʿjam. 120 Al-​Muḥaṣṣab is a prominent ritual toponym at Mecca.

156 Webb When a call for help was raised, they rode forth But it was on the boards of ships. No Azdī woman felt the sting of circumcision, Nor did she taste milk from a pail. Huntsmen never brought her eggs or truffles, Nor did she taste meat won by lot.121 Nor did a slave girl ever raise a tent above her, A Bedouin shelter propped on poles. She never lit a fire to guide night-​arriving guests, Nor did she ever hear the welcoming bark of dogs … 122 Al-​Farazdaq’s remarkable poem illuminates several issues connected with imagining communities in the Muslim cities of Iraq at the end of the first century ah/​seventh century ce. To lampoon al-​Muhallab, al-​Farazdaq focuses on the al-​Azd tribe as a whole, which aligns with Arabic prose narrative sources that identify al-​Muhallab as a military commander with authority over al-​Azd in Basra. To belittle al-​Azd, lines 13–​14 articulate the heart of al-​Farazdaq’s lampoon: the poet expresses al-​Azd’s non-​Arabness. From an ʿAbbāsid perspective, this accusation is nonsensical, since, by the third century ah/​ninth century ce, al-​Azd were firmly part the Arab family tree. However, al-​Farazdaq’s poem suggests that this position was not always the case. While Omanis claimed to be members of al-​Azd, participated in the wars of early Islam, settled in the Emigrant garrison town of Basra, and claimed to be Arabs along with the rest of the Marwānid-​era elite, their Omani home and different cultural heritage conflicted with the dominant Quraysh-​led groups who defined their sense of Arab identity with different parameters. Accordingly, opponents of al-​Azd could point to Azdī difference as “proof” to deny their Arabness. Al-​Farazdaq is helpfully specific in his justification: he focuses on al-​Azd’s ancestral home in Oman (distant from his central Arabian kin), their seafaring ways, and their lack of a Bedouin heritage. Al-​Azd, in al-​Farazdaq’s description, have no history of living like “Arabs” lived: al-​Azd toil with ropes on boats, Arabs work with horses’ reins (line 11); Arabs speak the clear Arabic language, al-​Azd do not (line 14); Arabs perform the Hajj, al-​Azd did not (line 15). Al-​ Farazdaq explains that the non-​Arabness of al-​Azd naturally emerges from their different past: Arabs used to worship idols, but the al-​Azd did not (line 16); the typical aspects of Bedouin life that Arabs performed in pre-​Islamic 121 Maysir, a form of gambling by throwing arrow lots, commonly appears in poetry as part of the typical repertoire of the pre-​Islamic Arabian tribesmen. 122 al-​Farazdaq 1987, 18–​19.

Elite Responses to the Founding of the Caliphate


times are likewise not part of al-​Azd’s past (lines 20–​26). Beyond their seafaring lifestyle, al-​Farazdaq does not specify how the pre-​Islamic al-​Azd lived. Further elaboration was unimportant for him: al-​Farazdaq’s rhetoric hinged on his emphasis of the non-​Arabness of al-​Azd, defining them as outsiders to the “inside” cultural system imagined for the Arabs.123 Ziyād al-​ Aʿjam, a contemporary of al-​ Farazdaq, articulated similar sentiments: Do al-​Azd hear what’s said about them In the palace’s courtyard—​or are they deaf? A people who’re circumcised once they reach senility, Misguided aspirants to be Arabs, though they be non-​Arab.124 The prominence of circumcision is pointed: Ziyād accuses the al-​Azd of being late joiners to Islam, i.e., because they only recently adopted Islamic practices, their elderly men had to be circumcised—​they were not born into Islam. He elaborates the same theme in a poem directed against al-​Muhallab ibn Abī Ṣufra: We were the ones who cut off Abū Ṣufra’s foreskin So that he could enter Basra, When ʿUthmān125 saw the foreskins, He set the knife upon them.126 123 The poem is also noted in Ulrich 2019, 77, 117 in the context of al-​Muhallab’s lineage and the standing of al-​Azd in Basra; Ulrich concurs in reading the poem as an example of the fluidity of Arab identity in the early period, though also considers the poem to evidence that the trappings of Bedouin identity were key to a sense of Arabness (76–​77). However, al-​Farazdaq does not make the claim that proper Arab identity is that of the Bedouin per se, since each of the examples he cites relating to the desert are past tense: for al-​Farazdaq, it appears that Arabs are those whose roots emanate from the Arabian desert, not necessarily those who currently live as desert nomads. The ramification of al-​Farazdaq’s poem also can be extrapolated further, given that very similar sentiments about the non-​Arab identity of al-​Azd are expressed in other sources—​notably al-​Farazdaq’s contemporary Ziyād al-​Aʿjam, whose poetry has not been studied in this regard, and will be discussed presently. 124 al-​Aʿjam 1983, 97. 125 The poet refers to ʿUthmān ibn Abī al-​ʿĀṣ, appointed ruler in Khorezm during the caliphate of ʿUmar ibn al-​Khaṭṭāb, who was apparently ordered to circumcise non-​Arabs in the city of Shahrak (or Shahrkand) (al-​Iṣfahānī 1992, 20:86–​87). In another story of Abū Ṣufra’s late circumcision, Ibn Ḥajar refers to the poetry specialist al-​Aṣmaʿī (d. 213 ah/​828 ce) who, specifically commenting on another of Ziyād’s poems, links it to an audience of Abū Ṣufra with the caliph ʿUmar (al-​ʿAsqalānī 1995, 9:186). 126 al-​Aʿjam 1983, 71.

158 Webb The emphasized absence of circumcision among the al-​Azd is significant, as we perceive Arabness and Islam as intertwined: the poem suggests that a “proper Arab” has a long tradition of complying with Islamic norms, and that the Bedouinized identity of Arabness cannot easily be separated from Muslimness. In their lampoons, both Ziyād and al-​Farazdaq also emphasize al-​Azd’s non-​ participation at Hajj, which, as we have seen above, was considered in the Umayyad era to be a central aspect of communal belonging. As al-​Farazdaq presents it, an “Arab” of Marwānid-​era elite understanding must ipso facto be a practicing Muslim too. This melding of religious and ethnic identity has perceptible relevance in the concept of community in the Marwānid period. In a third poem, Ziyād mirrors al-​Farazdaq in using fishing associations to chide al-​Azd: In came al-​Azd, tripping on their beards, Juwāf fish falling out of their nostrils.127 The rhetorical device suggests that because desert Arabs did not fish, al-​Azd’s alleged proclivity to fish indicates their non-​Arabness. I have not found such tropes used against other Basran tribal groups. Thus, the physical distance of al-​Azd’s Omani home from central Arabian heartlands, and the consequent cultural differences evidently handicapped al-​Azd’s efforts to assume an Arab identity when it became popular among the caliphate’s military elite. The al-​ Azd were easy targets for their detractors.128 Al-​Azd’s problematic Arabness further emerges in an anecdote from the end of the Umayyad era about a battle waged between a group of southerners, including al-​Azd, pro-​Umayyad, and pro-​Quraysh forces outside Medina in 130 ah/​748 ce. According to a memory of the battle recorded by Abū al-​Faraj al-​Iṣfahānī, A Qurayshite man beheld one of the Yemenis exclaim: “Praise be to God who has brought me contentment to see the Quraysh killed!” His son added, “And Praise be to God who has used our hands to disgrace them!” This was because the Quraysh did not consider the Azd from Oman as Arabs.129 1 27 al-​Aʿjam 1983, 83. 128 The poems noted above that chide Christian groups for their nonparticipation in the Hajj do not, as my research has so far revealed, deny the Arabness of those Christian groups to the extent expressed by al-​Farazdaq and Ziyād al-​Aʿjam. 129 al-​Iṣfahānī 1992, 23:241.

Elite Responses to the Founding of the Caliphate


Why the Quraysh in the early second century ah/​eighth century ce would have such a negative view about al-​Azd likely stems from the only recently silenced troubles in Basra, when Yazīd ibn al-​Muhallab sought to parlay al-​ Azd’s Basran power into a challenge against caliphal/​Quraysh authority in Iraq. The Quraysh’s response aligns with the poetry of al-​Farazdaq and Ziyād al-​ Aʿjam: the Omani al-​Azd possessed attributes that did not fit with the notions of Arab identity embraced by the center of the caliphate, and al-​Azd’s opponents could utilize ethnos to belittle them and deny them elite status. An anecdote from an entirely different context—​early Persian historiography from eastern Iran—​similarly uses al-​Azd’s perceived difference from other military elites of the Marwānid caliphate in political rifts. The Tārīkh-​i Harāt attributed to ʿAbd al-​Raḥmān Fāmī al-​Harawī (d. 546 ah/​1151–​52 ce) contains remarkable old material, including a story that the Muslim commander in the east, Qutayba ibn Muslim (d. 96 ah/​714–​15 ce), upon being appointed governor of Khurāsān, converted a garden constructed by the previous governor, Yazīd ibn al-​Muhallab, into a camel paddock. Qutayba and Yazīd were political rivals, and al-​Harawī’s text recounts how Qutayba remarked upon converting the garden that: “My father was a camel-owner (ushtur-​bān), Yazīd’s father was a garden-​owner (būstān-​bān).”130 Qutayba’s chiding of Yazīd precisely echoes the lines of al-​Farazdaq and Ziyād al-​Aʿjam: the Omani Muhallabids could be insulted for their lack of camel-​rearing roots, an overt insinuation of their place outside the growing sense of Arab-​Muslim elite identity in the Marwānid caliphate. The success of Abū Ṣufra and his son al-​Muhallab in Basra during the earlier periods of initial Emigrant expansion indicates that a lack of Arabian nomadic background was not a handicap to political power, nor did it prevent groups from participating in military expeditions of the first Muslim expansions. However, by the Marwānid era, the shifting expectations of elite identity found a new emphasis on Arabness (as evidenced above) and opened potential weaknesses in the social standing of the Muhallabid’s third-​ generation leader, Yazīd. Considering matters from the perspective of Arabness as it was conceived in early Islam, we can better understand the motivations of al-​Farazdaq, Ziyād, Qutayba ibn Muslim, and the Quraysh. Al-​Azd, especially its members from Oman, left almost no footprint in Muslim-​era texts about pre-​Islamic Arabia. The al-​Azd have no poets,131 Ibn al-​Kalbī’s Kitāb al-​Aṣnām (the Book of Idols) 1 30 al-​Harawī 2008, 11–​12. 131 Even the Azd Sarāt from central Arabia had few pre-​Islamic poets. Al-​Shanfarā and Ḥājiz al-​Sarawī are identified as Azdī, and while they were famous poets, they are shifty figures from outlaw lore. It can thus be argued that the central Arabian al-​Azd had at

160 Webb does not refer to any idols venerated by al-​Azd in Oman,132 and Azd ʿUmān are absent in the major collections of ancient Arabian lore, such as Ibn Qutayba’s Faḍl al-​ʿArab; Ibn Ḥabīb’s al-​Muḥabbar preserves only one curious story linking the Azd ʿUmān to a stray branch of the Quraysh.133 In contrast to the importance Arabic sources assign to al-​Azd as major participants in the nascent caliphate, Arabic memories about the cultural icons of Arabness palpably do not intersect with al-​Azd. Thus, although the al-​Azd in Basra were part of the Umayyad caliphate’s Muslim-​Arab elite by virtue of their physical participation and success in military ventures, they did not possess the same cultural and other intangible characteristics as the other “Arab” groups. Perhaps responding to the rhetoric of their rivals, the early sources suggest that al-​Azd appear to have made efforts to assert their Arabness by finding a way into Arab lore. Ibn al-​Kalbī’s Ansāb al-​khayl (The Genealogy of Horses) reports that al-​Azd claimed that the ancestor of their horses, Zād al-​Rākib, was a stallion from Solomon’s stables, which escaped and became the ancestor of all Arab horses.134 In this legendary story, the connection of Solomon to the origin of the Arabs’ horses strongly suggests a Muslim-​era creation meant to assert al-​Azd’s pedigree in horsemanship, which, in the Umayyad period, was closely associated with Arab identity. Similarly, the earliest elaborate stories of al-​Azd’s pre-​Islamic history and their spread across Arabia derive from the (likely apocryphal) text of Wahb ibn Munabbih (d. 110 ah/​728 ce), Kitāb al-​ Tījān,135 a book containing legends and epic expansion that have been called “extravagant flights of fancy” by one modern commentator.136 Premodern Arabic historians likewise criticized the historiography of Azdī and Yemeni Himyarite kings: the fourth-​century ah/​tenth-​century ce Ḥamza al-​Iṣfahānī remarked “there is nothing in all of history more corrupted and erroneous than the history of the governors and kings of Ḥimyar, for its dynasty lasted such

132 1 33 134 135 136

least some tentative connection to Arabic poetic culture, whereas any Azd ʿUmān poets seem unknown to Muslim collectors. The early ʿAbbāsid-​era poetry collector Ibn Sallām al-​Jumaḥī states that eastern Arabian Baḥrayn had “many excellent and eloquent [pre-​ Islamic] poets” (n.d. 1:271), but he did not mention any poets from Oman. Ibn al-​Kalbī 1924, 37–​39 refers to idols of Azdī groups in central Arabia, i.e., the Azd Sarāt, presumably related to the al-​Asd who appear in pre-​Islamic inscriptions but separate from the Azd ʿUmān. Ibn al-​Kalbī does not mention any idols in connection with Oman. Ibn Ḥabīb 1942, 168. Ibn al-​Kalbī 2009, 17–​19. Pseudo-​Wahb 1996, 273–​303. Vajda 1960–​2007. Khoury 1960–​2007 is more sympathetic to Wahb’s reliability as a narrator, and Wahb material mixed varied sources: the material in the Tījān better fits on Vajda’s side of the spectrum.

Elite Responses to the Founding of the Caliphate


a long time and so few of their kings are mentioned.” Ibn Khaldūn was even more scathing.137 This source was outside the ambit of Ulrich’s 2008 study of the al-​Azd migration stories, however it tallies with his findings and furnishes further support of his conclusion that the migration story was a Muslim-​era anachronism given that its first articulation occurs in pseudo-​Wahb’s very problematic text that contains considerable anachronistic material. No extant pre-​Islamic poetry grounds the migration tale within memories of tribal groups from pre-​Islamic Arabia, and the story did not circulate among the collectors of ancient Arabica. Muslim-​era storytellers seem the primary crafters of the story, using Qurʾānic exegesis to create a narrative specifically targeted to prove to a Muslim audience that Omani Azdis were “Arab.” Further support for this conclusion appears in another early source for the al-​Azd migration story, which so far has not been brought into the discussion. Muqātil ibn Sulaymān’s (d. 150 ah/​767 ce) Tafsīr, a text that resembles the pseudo-​Wahb storytelling genre and that has been criticized for extensive reliance on legendary material, includes the migration story.138 To summarize, the first references to the Arabness of Azd ʿUmān are very early, and nearly contemporary with the anti-​Azdī poetry of al-​Farazdaq and Ziyād al-​Aʿjam, suggesting efforts to situate the Basran Omanis within a family tree that included their compatriot central and western Arabian Emigrant groups. However, the pro-​Azdī stories come from contrived narratives of dubious reputation. Moreover, because they lack the usual means of corroboration through pre-​Islamic poetry, the stories are likely not vestiges of pre-​Islamic memories. We accordingly behold the fluidity of Arabness in the Marwānid period. Military elites articulated Arab identity, but the parameters of that Arabness were unclear. The Quraysh and Umayyad military elites from Maʿaddite central Arabian backgrounds controlled what seems the “hegemonic” voice in determining the hallmarks of Arab identity, which they crafted around both the Maʿaddites’ Bedouin heritage and the caliphate’s normative Islamic practice. Other groups, such as Yemenis and Omanis, attempted to join the newly emerging Arab family tree, and while they could acculturate to the current Islamic norms of the Marwānid elites, their different cultural and geographical origins jeopardized the legitimacy of their Arabness. Quite why al-​Azd faced the rhetoric that denied their Arabness likely stemmed from politics. Political tensions in Basra (where the Muhallabids and

1 37 Ḥamza al-​Iṣfahānī n.d., 106; Ibn Khaldūn 1998–​99, 1:16–​21. 138 Muqātil 1979–​89, 3:531.

162 Webb al-​Azd organized various acts of independence and rebellion, clashed with Umayyad governors, and played political power games with other ex-​Arabian groups who also claimed to be Arabs) pitted al-​Azd in particular against the other major stakeholders in the Umayyad system.139 Al-​Azd’s opponents used Arabness as a means to champion their own identity, and because al-​ Azd lacked the required cultural capital, rival ex-​Arabian groups could use Arabness against al-​Azd. The lampoons can therefore be understood as products of factional power politics, and the poets’ choices of Arabness as a weapon within that struggle is particularly telling: the al-​Azd evidently were less able than other ex-​Arabian groups to be accepted as “Arabs” in the Umayyad era. The challenges we have found that the al-​Azd faced regarding their membership to the elite’s “Arab” identity are anticipated in the anthropological theories of ethnogenesis of Weber and Barth, who both identify that political and economic factors are the main drivers for grouping people within ethnic boundaries. Al-​Azd, as central members of the Umayyad-​era military elite, would naturally seek to fit themselves within Arabness as it had become the preferred means for Marwānid-​era elites to articulate their identity, but al-​Azd’s “cultural stuff” traits140 differed and caused difficulties of assimilation. Al-​Azd, by virtue of political-​economic status, technically could locate themselves on the inside of the social boundary defined as the “Arab community,” but they lacked the cultural markers of other members of this community. When tensions flared, al-​Azd’s rivals could thus capitalize on cultural difference to exclude al-​Azd. Barth and Weber also offer theoretically grounded explanations for why the poets’ lampoons ultimately did not succeed: al-​Azd eventually integrated into the Arab family tree, leaving the anti-​Azd poetry in the margins of later historical sources. The theories of ethnogenesis stress that social groups form primarily because of economic/​political/​social realities, and not as a matter of cultural affinity. Basra’s al-​Azd were at the core of the caliphate’s military elite, so that once the Muhallabid factionalism subsided, especially in the early ʿAbbāsid period when almost all factional fighting ended in Iraq, all of Iraq’s military elite groups enjoyed high social status and wealth. Al-​Azd found themselves firmly inside the social boundary of the early ʿAbbāsid military elite, which by this time also meant “Arab elite.” The reality of al-​Azd’s influence in Basra demanded that the new system embrace them. To this end, the al-​Azd did at least share the common heritage with the other “Arab” elite groups of being 1 39 For discussion of these conflicts, see Ulrich 2019, 139–​52, 180–​87. 140 Barth 1969 coined the term “cultural stuff” to refer to attributes such as language, dress, history, food, and other visible attributes of an ethnic identity, which he considered secondary to political and economic boundaries in the formation of ethnic groups.

Elite Responses to the Founding of the Caliphate


part of the original Emigrant expansion, and since al-​Azd’s members were also allotted ʿaṭāʾ payments of the caliphal system,141 they had evident commonalities with the rest of the “Arab” elite. Theories of ethnogenesis urge us to take such material considerations seriously, and political/​economic factors seem to have been aligned such that the al-​Azd could be assimilated once the frictions of factional infighting had been resolved. In the ʿAbbāsid peace, the old cultural differences which had fueled the anti-​Azd lampoons and distanced the al-​Azd from Arab identity could then fade away as mere “cultural stuff,” and members of Basra’s al-​Azd would be welcomed into the Arab fold. Al-​Azd’s Arabness was approved in the ʿAbbāsid period, their genealogy cemented, the migration story widely narrated, and al-​Azd never again had to endure the slings and arrows of al-​Farazdaq and Ziyād’s slander. 6


The question of elite responses to the rise of Islam indicates that radical changes to the nature of society neither occurred during Muhammad’s lifetime nor the militarized settlement of the first Emigrants, but rather two generations later in the context of the maturing and centralizing caliphate. Poetry indicates that Islam clearly factored into the consciousness of Emigrant-​era Arabians, and the first Emigrants mobilized their faith into wars, raids, and resettlement, but the significance of these changes took time to crystallize into a real cultural transformation. Poetry supports the suspicions of modern historians that Islamic practices only became clear and sufficiently widely accepted to serve as an unambiguous marker of social boundaries in the Marwānid era when the development of the centralizing caliphate allowed for the early articulations of Islam as the property of one social group. It is most intriguing that this timing aligns exactly with the emergence of the name “Arab” in poetry to delineate elite social boundaries in the Marwānid caliphate. The Emigrants’ generation did not produce poems in which “Arab” connotes their sense of self-​identity, and likewise, their poets’ relationship with Islam does not serve as a rhetorical device for social inclusion or exclusion. The Marwānid-​era elites thus became “Muslim” and “Arab” simultaneously, so the knotty issue of how “Muslim identity” can be separated from “Arab identity” cannot easily be untied. The case of the Omani al-​Azd 141 ʿAṭāʾ refers to stipends paid by the government to the military elite groups in the amṣār in recognition of their contribution to the early expansions and as compensation for calls to fight new campaigns.

164 Webb brings into high relief that the hegemonic form of Arabness looked back to Bedouin culture as the signature of Arab identity. However, the various groups aspiring to be Arabs did not all have a Bedouin past, but rather they shared membership in the first Emigrant communities. As often occurs in many social groups, a contradiction emerges: members of the caliphate’s elite had a shared communal history as settlers in the amṣār (going back two or three generations), but when the Marwānid elites began expressly calling themselves Arabs, they looked to an older sense of the past to ground that identity. Other groups subsequently scrambled to produce genealogies, poetry, and migration stories to get on board. By contrast, the problems of al-​Akhṭal’s Christians reveal another side to Marwānid-​era elite identity: as it became increasingly important to be Muslim, al-​Akhṭal’s Christian Taghlib kin, despite their Bedouin origins, found their path complicated by their different faith, opening them to lampoons from Jarīr about their nonparticipation in the Hajj. The elite response to the caliphate therefore centers on a dual consolidation of faith and ethnos, which created both Arabness and Islam in a new codified form. The processes changed social identities and defined the faith in ways that the first Emigrants would not have been able to articulate. In the early period, Christians and Omani sailors could join the Emigrant armies, while Arabian poets would summon Allāh’s name in general terms as the venture of Islam began on broad foundations. After two generations, however, elite society in the amṣār understood their identity in terms of an Arab ethnos that looked to a new demographically-cohesive conception of Jazīrat al-​ʿArab, imagined as the homeland of an ancient unified people. These members of elite society developed a clear sense of normative Muslim culture that was distinct enough to demarcate social boundaries in the otherwise cosmopolitan caliphate. The responses to the organization of the caliphate in terms of ethnogenesis and religious codification opened new doors for Muslim-​era populations to conceptualize community around faith and ritual in new, starker ways, and the role of Arabness in Islam and in Middle Eastern culture embraced novel features that loomed significantly in the complex negotiation of identity for an ever-​growing percentage of the region’s population ever since.


Abū Dahbal al-​ Jumaḥī. 1972. Diwan. Edited by ʿAbd al-​ ʿAẓīm ʿAbd al-​ Muḥsin. Najaf: Maṭbaʿat al-​Qaḍā.

Elite Responses to the Founding of the Caliphate


Agha, Said Saleh. 2011. “ ‘Of Verse, Poetry, Great Poetry and History.” In Poetry and History: The Value of Poetry in Reconstructing Arab History, edited by Ramzi Baalbaki, Saleh Said Agha and Tarif Khalidi, 1–​35. Beirut: American University of Beirut Press. al-​Aḥwaṣ al-​Anṣārī. 1990. Shiʿr al-​Aḥwaṣ al-​Ansārī. Edited by ʿĀdil Sulaymān Jamāl. Cairo: al-​Khānjī. al-​Aʿjam, Ziyād. 1983. Shiʿr Ziyād al-​Aʿjam. Edited by Yūsuf Ḥusayn Bakkār. Beirut: Dar al-​Masīra. al-​Akhṭal, Abū Mālik Ghiyāth ibn Gawth al-​Taghlibī. 1996. Shiʿr al-​Akhṭal (Riwāyat al-​ Sukkarī ʿan Ibn Ḥabīb). Edited by Fakhr al-​Dīn Qabāwa. Damascus: Dar al-​Fikr. al-​Anbārī, Abū Muḥammad al-​Qāsim. 2003. Sharḥ Dīwān al-​Mufaḍḍaliyyāt. Edited by Muḥammad Nabīl Ṭarīfī. Beirut: Dar Sādir. al-​Aʿshā, Qays ibn Maymūn. 1974. Dīwān. Edited by Muḥammad Muḥammad Ḥusayn. Beirut: Dār al-​Nahḍa al-​ʿArabiyya. al-​ʿAsqalānī, Ibn Ḥajar. 1995. al-​Iṣāba fī tamyīz al-​Ṣaḥāba. Edited by ʿĀdil Aḥmad ʿAbd al-​Mawjūd and ʿAlī Muḥammad Muʿawwaḍ. Beirut: Dār al-​Kutub al-​ʿIlmiyya. al-​Azharī, Muḥammad ibn Aḥmad. 2004. Tahdhīb al-​lugha. Edited by Muḥammad ʿAbd al-​Raḥmān Mukhaymir. Beirut: Dār al-​Kutub al-​ʿIlmiyya. al-​Azmeh, Aziz. 2014. The Emergence of Islam in Late Antiquity. Cambridge: Cambridge University Press. Al-Jallad, Ahmad. 2020. “ʿArab, ʾAʿrāb, and Arabic in Ancient North Arabia: the first attestation of (ʾ)ʿrb as a group name in Safaitic.” Arabian Archaeology and Epigraphy 31:422–35. Barth, Fredrik. 1969. Ethnic Groups and Boundaries: The Social Organization of Culture Difference. Oslo: Universitetsforlaget. al-Baṣrī, ʿAlī Abī al-Faraj. 1999. al-Hamāsa al-Baṣriyya. Edited by ʿĀdil Sulaymān Jamāl. Cairo: al-Khānjī. Bauer, Thomas. 2010. “The Relevance of Early Arabic Poetry for Qurʾanic Studies Including Observations on Kull and on Q 22:27, 26:225, and 52:31.” In The Qurʾān in Context, edited by Angelika Neuwirth, Michael Marx, and Nicolai Sinai, 699–​732. Leiden: Brill. Berkey, Jonathan P. 1996. “Circumcision Circumscribed: Female Circumcision and Cultural Accommodation in the Medieval Near East.” International Journal of Middle Eastern Studies 28:19–​38. Blankinship, Khalid Yahya. 1994. The End of the Jihad State. Albany: New York University Press. Brockelmann, Carl. 1922. “Allah und die Götzen, der Ursprung des islamischen Monotheismus.” Archiv für Religionswissenschaft 21:99–​121. Cheikho, Louis. 1926. Kitāb Shuʿarāʾ al-​Naṣraniyya fī al-​Jāhiliyya. Cairo: Maktabat al-​Ādāb. Crone, Patricia. 1980. Slaves on Horses. Cambridge: Cambridge University Press.

166 Webb Crone, Patricia. 1960–​2007. “al-​Muhallab b. Abī Ṣufra.” In Encyclopaedia of Islam, 2nd ed., edited by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs. Leiden: Brill. Dhū al-​Rumma. 1972–​74. Dīwān. Edited by ʿAbd al-​Quddūs Abū Ṣāliḥ. Damascus: Majmaʿ al-​Lugha al-​ʿArabiyya. Dmitriev, Kirill. 2017. “Glory and Immortality: The Motif of Monumentum Aere Perennius by Samawʾal B. ʿĀdiyāʾ.” In Religious Culture in Late Antique Arabia, edited by Kirill Dmitriev and Isabel Toral-​Niehoff, 105–​21. Piscataway NJ: Gorgias Press. Donner, Fred. 2010. Muhammad and the Believers: At the Origins of Islam. Cambridge MA: Harvard University Press. Donner, Fred. 2016. “Arabic Fatḥ as ‘Conquest’ and its Origin in Islamic Tradition.” al-​ ʿUṣūr al-​Wusṭā 24:1–​14. Donner, Fred. 2018. “Talking About Islam’s Origins.” bsoas 81.1:1–​23. Eliade, Mercia. 1968. Myth and Reality. New York: Harper. al-​Farazdaq, Hammām ibn Ghālib. 1987. Dīwān. Edited by A. Fāʿūr. Beirut: Dār al-​Kutub al-​ʿIlmiyya. Frye, Richard N. 1975. “Preface.” In The Cambridge History of Iran: 4: From the Arab Invasion to the Seljuks, edited by Richard N. Frye, xi–​xiii. Cambridge: Cambridge University Press, 1975. Gelder, Geert Jan van. 1988. The Bad and the Ugly: Attitudes Towards Invective Poetry (Hijāʾ) in Classical Arabic Literature. Leiden: Brill. Gibbon, Edward. [1776–​89] 1994. Decline and Fall of the Roman Empire. London: Everyman. Goldziher, Ignáz. [1889–​90] 1967–​71, Muslim Studies (Muhammedanische Studien), edited by S. M. Stern, translated by C. R. Barber and S. M. Stern. London: George Allen & Unwin. Gruen, Erich S. ed. 2011. Cultural Identity: In the Ancient Mediterranean. Los Angeles: Getty Research Institute. Grunebaum, G. E. von. 1963. “The Nature of Arab Unity before Islam.” Arabica 10.4–​23. al-​Harawī, ʿAbd al-​Raḥmān Fāmī. 2008. Tārīkh-​i Harāt. Edited by Muḥammad Ḥusayn Mir Ḥusaynī and Muḥammad Riḍā Abūʿī Mihrīzī. Tehran: Markaz Paẓuhashi Mīrāth Maktūb. Hinds, Martin. 1981. An Early Islamic Family from Oman. Manchester: Journal of Semitic Studies Monograph, Manchester University Press. Hoyland, Robert. 2015. In God’s Path. Oxford: Oxford University Press. Hoyland, Robert. 2017. “The Identity of the Arabian Conquerors of the SeventhCentury Middle East.” Al-ʿUṣūr al-Wusṭā 25:113–40. Hughes, Aaron. 2017. Shared Identities: Medieval and Modern Imaginings of Judeo-​ Islam. New York: Oxford University Press.

Elite Responses to the Founding of the Caliphate


Ibn ʿAbd al-​Barr, Yūsuf ibn ʿAbd Allāh. 1992. al-​Istiʿāb fī maʿrifat al-​Aṣḥāb. Edited by ʿAlī Muḥammad al-​Bajāwī. Beirut: Dār al-​Jīl. Ibn Durayd, Abū Bakr Muḥammad. N.d. al-​Ishtiqāq. Edited by ʿAbd al-​ Salām Muḥammad Hārūn. Cairo: al-​Khānjī. Ibn Ḥabīb, Muḥammad. 1942. al-​Muḥabbar. Edited by Ilse Lichtenstadter. Hyderabad: Dāʾirat al-​Maʿārif al-​ʿUthmāniyya. Ibn Ḥazm, Abū Muḥammad ʿAlī. 1994. Jamharat ansāb al-​ʿArab. Edited by ʿAbd al-​ Salām Muḥammad Hārūn. Cairo: al-​Maʿārif. Ibn Hishām, Abū Muḥammad ʿAbd al-​Malik. N.d. al-​Sīra al-​Nabawiyya. Edited by Muṣṭafā al-​Saqqā, Ibrāhīm al-​Abyārī, and ʿAbd al-​Ḥafīz al-​Shalabī. Beirut: Dār al-​Maʿrifa. Ibn al-​ Kalbī, Hishām ibn Muḥammad. 1924. al-​Aṣnām. Edited by Aḥmad Zakī. Cairo: Dār al-​Kutub. Ibn al-​Kalbī, Hishām ibn Muḥammad. 1988. Nasab Maʿadd wa-​l-​Yaman. Edited by Nājī Ḥasan. Beirut: ʿĀlam al-​Kutub. Ibn al-​Kalbī, Hishām ibn Muḥammad. 2009. Ansāb al-​khayl. Edited by Aḥmad Zakī. Cairo: Dār al-​Kutub. Ibn Khaldūn. 1998–​99. al-​ʿIbar wa Dīwān al-​mubtadaʾ wa-​l-​khabar fī ayyām al-​ʿArab wa-​ l-​ʿAjam wa-​l-​Barbar. Cairo and Beirut: Dār al-​Kitāb al-​Miṣrī, Dār al-​Kitāb al-​Lubnānī. Ibn Manẓūr, Jamāl al-​ Dīn Muḥammad ibn Mukarram. 1994. Lisān al-​ʿArab. Beirut: Dār Ṣādir. Ibn Qutayba, ʿAbd Allāh ibn Muslim. 1982. al-​Shiʿr wa-​l-​Shuʿarāʾ. Edited by Aḥmad Muḥammad Shākir. Cairo: Dār al-​Ḥadīth. Ibn Qutayba, ʿAbd Allāh ibn Muslim.2011. Kitāb al-​Maʿānī al-​kabīr. Edited by Muḥammad Nabīl Ṭarīfī. Beirut: Dār Ṣādir. Ibn Qutayba, ʿAbd Allāh ibn Muslim. 2017. The Excellence of the Arabs. Edited by James Montgomery and Peter Webb, translated by Sarah Savant and Peter Webb. New York and Abu Dhabi: Library of Arabic Literature/​New York University Press. al-​Iṣfahānī, Abū al-​Faraj. 1992. al-​Aghānī. Edited by ʿAbd Allāh ʿAlī Muhannā and Samīr Jābir. Beirut: Dār al-​Kutub al-​ʿIlmiyya. al-​ Iṣfahānī, Ḥamza. N.d. Tārīkh sinī mulūk al-​arḍ wa-​l-​anbiyāʾ. Beirut: Maktabat al-​Ḥayāt. Izutso, Toshihiko. 1966. Ethico-​religious Concepts in the Qurʾān. Montreal: McGill University Press. Jacobi, Renata. 1960–​2007. “Mukhaḍram.” In Encyclopaedia of Islam, 2nd ed., edited by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs. Leiden: Brill. al-​Jawharī, Ismāʾīl ibn Ḥammād. 1984. al-​Ṣiḥāḥ. Edited by Aḥmad ʿAbd al-​Ghafūr ʿAṭṭār. Beirut: Dār al-​ʿIlm lil-​Malāyīn.

168 Webb Jenkins, Richard. 2008. Rethinking Ethnicity. London: Sage. al-​Jumaḥī, Ibn Sallām. N.d. Ṭabaqāt fuḥul al-​shuʿarāʾ. Edited by Maḥmūd Muḥammad Shākir. Cairo: al-​Khānjī. Kadi, Wadad, 1994. “The Limitations of Qurʾānic Usage in Early Arabic Poetry: The Example of a Khārijite Poem.” In Festschrift Ewald Wagner zum 65. Geburtstag, edited by Wolfhart Heinrichs and Gregor Schoeler, 162–​81. Beirut: In Kommission bei F. Steiner Verlag, Stuttgart. al-​Khalīl ibn Aḥmad al-​Farāhīdī. 1980. al-​ʿAyn. Edited by Mahdī al-​Makhzūmī and Ibrāhīm al-​Sāmarrāʾī. Baghdad: Wizārat al-​Thaqāfa wa-​l-​Iʿlām. Khoury, R.G. 1960–​2007. “Wahb ibn Munabbih.” In Encyclopaedia of Islam, 2nd ed., edited by P. Bearman, Th. Bianquis, C.E. Bosworth, E. van Donzel, and W.P. Heinrichs. Leiden: Brill. al-​Khuzāʿī, pseudo-​Diʿbil. 1997. Waṣāyā al-​mulūk. Edited by Nizār Abāẓa. Damascus: Dār al-​Bashāʾir. Lindstedt, Ilkka. 2015. “Muhājirūn as a Name for the First/​Seventh Century Muslims.” Journal of Near Eastern Studies 74:67–​73. Macdonald, Michael. 2009 [2001]. “Arabians, Arabias and the Greeks: Contact and Perceptions.” In Literacy and Identity in Pre-​Islamic Arabia, edited by Michael Macdonald, v:1–​33. Farnham: Ashgate. al-​Marzubānī, Muḥammad ibn ʿImrān. 2010. Muʿjam al-​Shuʿarāʾ. Edited by ʿAbbās Hāniʿ al-​Jarrāḥ. Beirut: Dār al-​Kutub al-​ʿIlmiyya. al-​Marzūqī, Aḥmad ibn Muḥammad. 1968. Sharḥ Dīwān al-​Ḥamāsa. Edited by Aḥmad Amīn and ʿAbd al-​Salām Hārūn. Cairo: Lajnat al-​Taʾlīf wa-​l-​Tarjama wa-​l-​Nashr. Millar, Fregus. 2013. Religion, Language and Community in the Roman Near East. Oxford: Oxford University Press. Montgomery, James. 1997. The Vagaries of the Qaṣīdah: The Tradition and Practice of Early Arabic Poetry. Cambridge: E.J. Gibb Memorial Trust. Montgomery, James. 2006. “The Empty Ḥijāz.” In Arabic Theology, Arabic Philosophy. From the Many to the One: Essays in Celebration of Richard M. Frank, edited by James Montgomery, 37–​97. Leuven: Peeters. Munt, Harry. 2015. “No Two Religions: Non-​Muslims in the Early Islamic Ḥijāz.” bsoas 78:249–​69. Muqātil ibn Sulaymān. 1979–​89. Tafsīr al-​Qurʾān al-​ʿaẓīm. Edited by ʿAbd Allāh Maḥmūd Shaḥāta. Cairo: al-​Hayʾa al-​Miṣriyya al-​ʿĀmma lil-​Kutub. al-​Nābigha al-​Dhubyānī. 1990. Diwan. Edited by Muḥammad Abū al-​Faḍl Ibrāhīm. Cairo: Dār al-​Maʿārif. Pohl, Walter and Helmut Reimitz. 1998. Strategies of Distinction: The Construction of Ethnic Consciousness in the Post-​Roman World. Farnham: Ashgate. al-​Rāʿī al-​Numayrī. 1980. Dīwān. Edited by Reinhard Weipert. Beirut: Franz Steiner.

Elite Responses to the Founding of the Caliphate


Robin, Christian, 2012. “The Birth of an Arab Identity.” In Museum Album, edited by Marie Foissy, Timothy Stroud, Simon Pleasance, David Hunter and Christ Atkinson (trans.), 47–​50. Paris: Institut du Monde Arabe. Sells, M. 1990. “Bānat Suʿād: Translation and Introduction.” Journal of Arabic Literature 21.2:140–​54. Shahid, Irfan. 1989. Byzantium and the Arabs in the Fifth Century. Washington: Dumbarton Oaks. Shahid, Irfan. 2009. Byzantium and the Arabs in the Sixth Century: Volume 2.2: Economic, Social and Cultural History. Washington: Dumbarton Oaks. al-​Shammākh ibn Dīrār al-​Dhubyānī. 1977. Dīwān. Edited by Salāḥ al-​Dīn al-​Hādī. Cairo: Dār al-​Maʿārif. al-​Sijistānī, Abū Ḥātim. 1991. Fuḥūlat al-​shuʿarāʾ. Edited by Muḥammad ʿAbd al-​Qādir Aḥmad. Cairo: Maktabat Nahḍat Miṣr. Sinai, Nicolai. 2011. “Religious Poetry from the Quranic Milieu: Umayya b. Abī l-​Ṣalt on the Fate of the Thamūd.” bsoas 74:397–​416. Stetkevych, Jaroslav. 2006. “A Qaṣīdah by Ibn Muqbil.” Journal of Arabic Literature 37:303–​54. Stetkevych, Susanne. 1994. “Pre-​Islamic Panegyric and the Poetics of Redemption,” in Reorientations/​Arabic and Persian Poetry, edited by Susanne Stetkevych, 1–​57. Bloomington IN: University of Indiana Press. al-​Sukkarī, Abū Saʿīd al-​Ḥasan ibn al-​Ḥusayn. 1963–​65. Sharḥ Ashʿār al-​Hudhaliyyīn. Edited by ʿAbd al-​Sattār Aḥmad Farrāj. Cairo: Maktabat Dār al-​ʿUrūba. al-​ Ṭabarī, Muḥammad ibn Jarīr. 1960–​ 69. Tārīkh al-​rusul wa-​l-​mulūk. Edited by Muḥammad Abū al-​Faḍl Ibrāhīm, 11 vols. Cairo: Dar al-​Maʿarif. Tannous, Jack. 2018. The Making of the Medieval Middle East: Religion, Society and Simple Believers. Princeton NJ: Princeton University Press. al-​Tibrīzī, al-​Khaṭīb. 2000. Sharḥ Dīwān al-​Ḥamāsa. Edited by Gharīd al-​Shaykh. Beirut: Dār al-​Kutub al-​ʿIlmiyya. Toral-​Niehoff. 2010. “The ʿIbād of al-​Ḥīra: An Arab Christian Community in Late Antique Iraq.” In The Qurʾān in Context, edited by Angelika Neuwirth, Michael Marx, and Nicolai Sinai, 323–​47. Leiden: Brill. Ulrich, Brian. 2008. “The Azd Migrations Reconsidered: Narratives of ʿAmr Muzayqiyā and Mālik b. Fahm in Historiographic Context.” Proceedings of the Seminar for Arabian Studies 38:311–​18. Ulrich, Brian. 2019. Arabs in the Early Islamic Empire. Exploring al-​Azd Tribal Identity. Edinburgh: Edinburgh University Press. Vajda, G. 1960–​2007. “Isrāʾīliyyāt.” In Encyclopaedia of Islam, 2nd ed., edited by P. Bearman, Th. Bianquis, C.E. Bosworth, E. van Donzel, and W.P. Heinrichs. Leiden: Brill.

170 Webb Pseudo-​Wahb ibn Munabbih. 1996. Kitāb al-​T ījān fī mulūk Ḥimyar. Cairo: al-​Hayʾa al-​ ʿĀmma li-​Quṣūr al-​Thaqāfa. Webb, Peter. 2016. Imagining the Arabs: Arab Identity and the Rise of Islam. Edinburgh: Edinburgh University Press. Webb, Peter. 2020. “al-​Azd.” In Encyclopaedia of Islam, THREE, edited by Kate Fleet, Gudrun Krämer, Denis Matringe, John Nawas, and Everett Rowson. Leiden: Brill. Webb, Peter. 2021a. “From the Sublime to the Ridiculous: Yemeni Arab Identity in Abbasid Iraq.” In Empires and Communities in the Post-​Roman and Islamic World, c. 400–​1000 CE, edited by Rutger Kramer and Walter Pohl, 283–​328. New York: Oxford University Press. Webb, Peter. 2021b. “Ethnicity, Power and Umayyad Society: The Rise and Fall of the People of Maʿadd.” In The Umayyad World, edited by Andrew Marsham, 65–102. London: Routledge. Webb, Peter. 2021c. “Desert Places: Poetry and Pre-​Islamic Arabian Toponyms,” Semitica et Classica 13:251–65. Weber, Max. 1996 [1922]. “The Origins of Ethnic Groups.” In Ethnicity, edited by John Hutchinson and Anthony D. Smith, 35–​39. Oxford: Oxford University Press. Wellhausen, Julius. 1927. The Arab Kingdom and its Fall. Translated by Margaret Graham Weir. Calcutta: University of Calcutta Press. Whitcomb, Donald. 1994. “Amsar in Syria? Syrian Cities after the Conquest.” Aram 6:13–​33. al-​Yaʿqūbī, Aḥmad ibn Abī Yaʿqūb. 1883. al-​Tārīkh. Edited by M. T. Houtsma. Leiden: Brill. Yāqūt, al-​Ḥamawī. 1987. al-​Muqtaḍab min Kitāb Jamharat al-​nasab. Edited by Nājī Ḥasan. Beirut: Dār al-​ʿArabiyya lil-​Mawsūʿāt. Zarrīnkūb, ʿAbd al-​Ḥusayn. 1975. “The Arab Conquest of Iran and its Aftermath.” In The Cambridge History of Iran: 4: From the Arab Invasion to the Seljuks, edited by Richard N. Frye, 1–​56. Cambridge: Cambridge University Press. Zuhayr ibn Abī Sulmā. 1982. Dīwān. Edited by Fakhr al-​Dīn Qabāwa. Beirut: Dār al-​Āfāq al-​Jadīda. Zwettler, Michael. 2000. “Maʿadd in Late-​Ancient Arabian Epigraphy and Other Pre-​ Islamic Sources.” Wiener Zeitschrift für die Kunde des Morgenlandes 90:223–​309.

­c hapter 7

Muhammad’s World in Egypt Petra M. Sijpesteijn The Arab takeover of the Byzantine province of Egypt in the mid-​seventh century ce introduced a new political regime in a new capital Fusṭāṭ, modern-​day Cairo. With it came a new language, Arabic, a new religion, Islam, and new cultural, administrative, and social practices. Much, however, also continued unaltered and especially at the level of daily life the routines of Egyptians living under Arab rule were in most cases not much affected. Greek, Coptic, and Arabic papyri preserved from this period catch both these continuities and changes. This chapter examines to what extent Arab-​ Muslim Egypt built on, reacted against, and broke with classical, Byzantine Egypt. The practices that the Arab conquerors brought to Egypt arguably lead us to customs common in Muhammad’s Arabia, but other influences are noticeable as well.1 1

Sources and Methods in the Study of Early Islam

Recent decades have witnessed a revolution in the historiography of early Islam. In the mid-​twentieth century ce, revisionist historians raised important concerns about Arabic narrative texts as historical sources. But their skepticism also left the field in a state of defeatist paralysis. At the end of the century, a new generation of scholars began to push beyond this impasse by drawing upon alternative sources to tackle the bias of the Arabic literary sources. Documentary evidence from archaeology, papyrology, diplomatics, numismatics, and epigraphy, as well as literary texts produced by communities outside the Muslim realm, used in conversation with the Muslim-​Arabic literary material, have opened new vistas. This combined with another important development—​the study of the connections with the cultures that preceded

1 This work was supported by the European Research Council under grant number 683194. This chapter builds on my article Sijpesteijn 2007. Here I pay additional attention to material culture and religious aspects and give a more elaborate discussion of where these practices might have originated.

© Petra M. Sijpesteijn, 2022 | DOI:10.1163/9789004500648_008

172 Sijpesteijn and partially overlapped with Islam, most notably Late Antiquity and ancient Arabia. Together, these methodologies have resulted in a multitude of vibrant multilingual, multidisciplinary research that offers new insights into the world of Islam’s rise and earliest history, and does so with a focus on different societal layers and historical processes. The study of early Islamic Egypt has been able to make especially important contributions to these scholarly developments thanks to its rich documentary record preserved in the form of papyri. Papyrus was the writing material of choice in the Mediterranean and Near East from its introduction in the third millennium bce until its replacement by paper in the tenth century ce. Horizontal and vertical layers of strips cut from the stem of the papyrus plant were glued together and smoothed to form a lightweight, durable, and versatile writing surface that was already in use in ancient Arabia.2 When the Arabs arrived in the Mediterranean they continued to rely upon papyrus as their main writing material, using it for every kind of text imaginable. Private and official letters, decrees and petitions, legal contracts, debt acknowledgements, tax assignments and receipts, lists of prisoners, converts, taxpayers, and accounts of estates or government departments, short notes requesting someone’s presence at court or at a private celebration, amulets, exercises and recipes, and fragments of theological, legal, and literary texts were all written on papyrus. Serving daily needs and practical purposes, the papyrus documents were discarded when they were no longer needed. As an organic material, papyrus does not generally fare well when exposed to the elements. In Egypt, however, the dry and unusable deserts in which unwanted papyri tended to be deposited offered conditions very nearly ideal to long-​term preservation. Thousands of papyrus documents survived there untouched, until they were dug up at the end of the nineteenth and early twentieth centuries ce as part of that era’s great hunt for antiquities. Retrieved from the sands, the papyri offer a uniquely direct view on the daily concerns and activities of Egyptians living under Muslim rule. By comparing the writing, language, movements, and preoccupations recorded in the documents from Arab-​Muslim Egypt with Egyptian papyri produced before the Arab takeover, it is possible to reconstruct what practices and customs the new rulers introduced in the province and how these interacted with local habits and traditions. Comparing the papyri with writings from the Near East and Arabia, especially inscriptions, offers moreover important information on the background of such practices and customs. 2 Sijpesteijn 2013, 1.

Muhammad’s World in Egypt


­f igure 7.1  Papyrus perf 558 recto, receipt for the delivery of 65 sheep from 643 ce Österreichische Nationalbibliothek


A New Regime

In the Islamic year 22 (643 ce), an Arab unit consisting of cavalry, heavily armed soldiers, and marines moved up the Nile in a campaign to subdue and collect taxes from the areas in the south of Egypt. Passing through Middle Egypt they obtained sixty-​five sheep from the local population. A receipt written in Greek and Arabic records the delivery of the sheep in minute detail (Figures 7.1–​7.2). The text, preserved on papyrus and now in the Austrian National Library, tells us much about the new rulers, their way of operating, their military and administrative organization, how they interacted with the local population, in what languages they functioned, and even their food preferences.3 The receipt was drawn up in name of the Arabic general in charge of the army that pacified Upper Egypt, ʿAbd Allāh b. Jābir, for Christophoros and Theodorakios, the Christian Egyptians who functioned as pagarchs, heads of 3 Grohmann 1952, 113–​15. For the following discussion, cf. Sijpesteijn 2013, 67–​69. For a recent interpretation of this document as recording the practice of murtabaʿ al-​jund, the seasonal grazing of animals in the Egyptian countryside in preparation of the campaigning season, see Bouderbala 2019, 375–​79.

174 Sijpesteijn

­f igure 7.2  Papyrus perf 558 verso

the administrative district of Ihnās (Gr. Heracleopolite). It is dated according to the Byzantine “indiction” calendar in the Greek section and the “hijra” calendar in the Arabic text. The Greek and Arabic parts of the document were written by two different scribes. The notary and deacon John wrote the Greek half, while Ibn Ḥudayd wrote the Arabic part, each according to his own scribal and legal traditions, with little coordination, it seems, between the two texts. Greek had been used in the Byzantine administration of Egypt and continued to be the main language of the Arab administration. Coptic, the last form of the Egyptian language, entered the chancery under the Arabs, whereas it had previously been used in private or semiprivate legal documents only.4 Arabic meanwhile was immediately added as an administrative language as in this document, although its use compared to Greek and Coptic was initially limited. The Arabic and Greek parts of the text on our papyrus, although conveying the same basic information, differ in focus and expression. The Greek text identifies the individuals in the document by their titles, including a Greek transliteration of the Arabic word for military commander, amīr (Gr. amiras), for ʿAbd Allāh b. Jābir. The Arabic scribe, however, used only names, but included 4 Cromwell 2017; Fournet 2020, 76–111.

Muhammad’s World in Egypt


patronymics, which are lacking in the Greek.5 In the Arabic a distinction is made between “fifty sheep for slaughter and fifteen other sheep” (khamsīna shāʾ min al-​jazar wa-​khamsa ʿashara shāʾ ukhrā), while the Greek lists the number of sheep delivered in numerals and spelled out in letters, followed by the qualification “only” (mona).6 The Arabic indicates that the sheep are given to ʿAbd Allāh’s troops for slaughter, while the Greek mentions that they are part of the dapanē, a technical fiscal term referring to the taxes paid in kind intended as maintenance of the Arabs. Finally, the Arabic is constructed in the third person according to the practice of Arabic legal documents, while the Greek follows a subjective first-​person style and specifies that the document functioned as evidence of the transaction having taken place.7 The receipt discussed in detail above records only one step in an extensive documentary procedure for the assignment and collection of levies imposed on the Egyptian population by the Arabs. Demand notes preceded the taking of goods and the issuing of receipts for the exacted products and services.8 These demand notes, as with the receipt discussed above, contain references to the practice of recording transactions for future reference, explain the purpose of the impositions, and in general provide evidence of a calculation and recording system that extended from the capital Fusṭāṭ over the whole province starting directly with the arrival of the Arabs. While the Romans had also collected taxes in kind, the Arab administrators perfected and expanded the practice, increasing the amount and the range of products requested from the local population. Arab administrators also put

5 Sijpesteijn 2013, 67. For other examples dating into the early eighth century ce, cf. Sijpesteijn 2013, 119n10. 6 The same practice occurs in Bactrian documents from central Asia that stand in the same Hellenistic tradition but date to the Islamic period, namely the eighth century ce. Cf. “one hundred, 100 dinars” in documents dated 478 (700 ce) and 527 (750 ce) both from Kadagstan and in a document dated 490 (712/​13 ce) from Rob (Sims Williams 2012, Document T line 19, p. 103; Document X line 29, p. 141; Document U line 26, p. 111); “one hundred 100 and twenty 20 local current Arab dirhams” in a document dated 525 (747 ce) (Sims Williams 2012, Document W line 31, p. 133) and “one hundred 100 and fifty 50 new Arab dirhams” in a document dated 490 (712/​13 ce) from Rob (Sims Williams 2012, Document v line 11, p. 119). 7 Whether this use of the third person in the document is motivated by Islamic legal prescriptions or the result of Arabian practice brought from the Peninsula to Egypt by the conquerors remains a question. To what extent the conquerors of the seventh century ce should be called “Arab,” “Muslim/​Islamic,” or “Arabian” remains a point of discussion. Cf. Donner 2003 and 2010; Hoyland 2014; Webb 2016. Cf. Sijpesteijn 2013, 68. Hoyland 2017 tries to reconcile the different views. 8 Four other Greek documents issued by ʿAbd Allāh b. Jābir dating to 642–​43 ce are in fact demand notes for contributions (Morelli 2010b, 162–​63).

176 Sijpesteijn into place arrangements to replace deliveries in kind by payments in coin via artisans and middlemen such as bakers and weavers.9 The Arabs’ decision not to divide the conquered lands among the soldiers, but to compensate them instead with a stipend and payments in kind provided for by the subjected population probably initiated the system that was then expanded to supply other groups and projects of the Arab administration. Papyri preserve large numbers of requests for fodder, food, clothes, and shelter for soldiers, travelling administrators, and their animals.10 One such Greek demand note was issued in the name of the conqueror and governor of Egypt ʿAmr b. al-​ʿĀṣ (d. 43 ah/​664 ce) himself. It commands that the pagarch of the same Ihnās/​Heracleopolite district mentioned before in January of 643 ce arrange to have a village in his district sell to ʿĀmir b. Asla,11 commander of a local unit of Arab troops, fodder at the fixed price of two dinar per four bundles, and to host ʿĀmir’s men for a period of three days. “Make sure to obtain a receipt for it,” ʿAmr b. al-​ʿĀṣ warns the pagarch and he adds that no other troops should be billeted in the town. On the back of the document a note specifies that the village of Kephalē indeed provided the requested fodder.12 Another set of Greek papyri records the demands imposed upon the Middle Egyptian district of Hermopolite (Ashmūnayn) in the years following the Arabs’ arrival. The central chancery transmitted numerous requests to the pagarch and his officers for materials, timber, stone, and ropes to construct public buildings in Fusṭāṭ. Greek and Coptic papyri record how such requests were received and handled by one local official, Senouthios, a scribe (Gr. notarios) and fiscal and administrative officer (Gr. anystes) who had served the Byzantines in the same position that he held under the Arabs. In one case he receives the order to close off the roads to his town and charge anyone trying to avoid paying the head tax, newly announced by the amīr. The head tax is called andrismos in the Greek papyrus, a new technical term for a new tax introduced by the Arabs.13 To what extent this tax should be equalled with the Qurʾānic jizya is discussed below. Another new element in the fiscal regime concerns the form of the tax demand notes as well as the receipts issued by the Arab chancery. Adopting the ancient Near Eastern legal documentary practice of double documents, the 9 10 11 12 13

Sijpesteijn 2013, 74. E.g., the texts in Grohmann 1957; Morelli 2001; Morelli 2010a; Sijpesteijn 2013. The Greek reads a-​s-​l-​a, allowing for a variety of readings in Arabic. The exact Arabic name intended remains unclear. Kießling, Rupprecht and Hengstl 1997, no. 14443. Morelli 2001, 19–​24.

Muhammad’s World in Egypt


Muslim fisc introduced in Egypt a practice of providing secured summaries of official documents that were at once irremovably attached to the document and hidden from unauthorized readers. Tax demand notes and receipts contained the necessary information twice: once in a visible and fully formulated text and, at the bottom, in an abbreviated form that was rolled up and closed with a seal. The seal would only be broken, and the secured summary revealed, in cases of disagreement or suspicion of corruption. Greek, Coptic, and Arabic documents from Islamic Egypt attest this practice.14 These papyri show that the Arabs introduced to Egypt their own well-​ developed documentary and scribal practice, which differed from the local tradition. They also initiated new administrative, and possibly legal, concepts and practices. It is also clear that this new regime extended into areas remote from the capital Fusṭāṭ and other garrisons where the bulk of the Arab troops were settled. Finally, the papyri witness the sophisticated written administration that was in place to record the kind and frequency of impositions levied on communities and to produce documentation that ensured the accountability and tracing of transactions. These administrative practices introduced by the Arabs to Egypt combine Near Eastern and Greek (Byzantine) elements. Moreover, a background in Levantine or Arab Hellenized customs seems likely, as will be discussed in the conclusion.15 3

Consumption Patterns and Material Culture

The papyri record not only how food was delivered to the Arabs, but also the kind of products they consumed. The Arabs introduced their own dietary preferences and material culture in Egypt. In the bilingual receipt discussed above, the Arab troops receive sheep to slaughter, presumably for their own consumption. Other papyri confirm the Arabs’ partiality to mutton. The papyri record requisitions of sheep meat and butter by the Arab administration that were intended for Arab soldiers and officials.16 The Arabs’ preference for butter contrasts with the consumption of oil in pre-​Islamic Egypt.17 The large amounts of hepsēma, must boiled to a syrup, that the Arab administration demanded from the Egyptian population remain somewhat of a mystery. The drink was known in Ptolemaic and Roman Egypt, but by far the most 14 15 16 17

Sijpesteijn 2012a; Sijpesteijn 2020a. See also the discussion in Sijpesteijn 2013, 70. Morelli 2001, 255; Morelli 1996, 111–​12; Sijpesteijn 2013, 408–​15; 428–​31. Sijpesteijn 2013, 66n115.

178 Sijpesteijn attestations occur in papyri from the Arab period, starting directly following the Arabs’ arrival.18 The Arabs ordered the syrup in large quantities for their sailors and workers on the fleet housed at the wharfs in Alexandria, Clysma, and Rhoda Island, and for administrators located throughout the province.19 The large numbers of wine amphorae found in Egypt, by contrast, show that wine consumption, especially of wine produced in Palestine, actually increased under the Arabs.20 The earliest mention of Egyptian lands administered for the caliph occurs in a labor contract for work on a vineyard located in the Fayyūm oasis.21 The papyri also contain information about a material culture that took shape in Egypt under the Arabs. Too little is known about the many buildings erected by the Arabs in the capital Fusṭāṭ immediately following the conquest to know whether new styles were introduced. Initially, the kinds of building materials and their intended use for administrative or ceremonial buildings were not distinctive. Only at the beginning of the eighth century ce do texts include requests for building materials and manpower specifically for the construction of mosques, albeit in Damascus and Jerusalem rather than in Fusṭāṭ. Attempts to connect structures in Fusṭāṭ with the supposed ethnic background of the Arab conquerors are difficult to support.22 It is, however, possible to trace changes in the material culture through the papyri in other areas. The Arabs used seals on documents for identification and to control access to the contents of the text, as was customary practice in the Near East as well as Egypt. Nevertheless, the application of seals on documents also witnessed innovations under the Arabs with Near Eastern, non-​Egyptian, practices such as that of the double document being (re)introduced into Egypt by the Arabs (see above). The forms of and decorations on seals similarly simultaneously showed innovation and continuation, resulting in new forms.23 Arab, Byzantine, and Sasanian practices were exchanged throughout the Muslim empire. Arab seals combined eastern and western iconography, joining 18 19 20 21 22 23

Cf. the order for hepsēma in a Greek papyrus dating to 644 from Ashmūnayn (Gr. Hermopolite), which belongs to the Senouthios archive (Kießling, Rupprecht and Hengstl 1997, no. 14219; cf. Morelli 2010a, 21–​27). Morelli 1996, 93–​96; 112–​13; 117–​18. Dixneuf 2011; Sijpesteijn 2013, 300–​1. See also the evidence in Palestinian vineyards of an increase in wine production for export in the late seventh and early eighth centuries (Avni 2014, 204–​7). Sijpesteijn and Worp 1983, 226–​29. Kubiak described the first layer of buildings in Fusṭāṭ as having Yemeni features, but an exclusively Yemeni origin for Egypt’s conquerors is untenable, something that Kubiak confirms (Kubiak 1987, 61–​64; 126). Sijpesteijn 2017, 655.

Muhammad’s World in Egypt


animals, celestial bodies, and texts in the local Egyptian context to form new imagery and decorative patterns. Thus, depictions of crescents and stars, which are commonly found on Sasanian seals, appear on seals used in the chancery of the Egyptian governor Qurra b. Sharīk (r. 90–​96 ah/​709–​14 ce), along with deer, other quadruped creatures, and even people (Figure 7.3). Similarly, the practice of adding multiple seal impressions onto clay sealings was introduced to Arab Egypt from the Sasanian realm. The multiple imprints on the sealings attached to Egyptian documents, however, do not belong to multiple individuals as they do in the Sasanian examples, but rather show the same seal being used repeatedly.24 The Sasanian practice moved to Egypt where it was adapted to local usages and given new meaning and form, through a process of “mutability of things through recontextualisation” as defined by Nicholas Thomas.25 At the beginning of the eighth century, administrative measures aimed at greater Islamicization and Arabicization were introduced throughout the Muslim empire. The measures are most clearly visible on the coins issued by the Muslim state where Arabic (Qurʾānic) phrases replaced images of rulers and religious symbols (Figure 7.4).26 Similarly, in the Muslim chancery, Arabic was introduced in addition to local languages as the caliphate promoted a multilingual policy.27 As discussed above, the use of local languages such as Greek and Coptic in the Egyptian chancery indeed continued for several generations. Eventually, however, the use of Arabic spread until all Egyptians used it for private as well as official, written and oral communication. By the ninth century ce, Arabic had become the written language of choice for most Egyptians and by the twelfth century ce Egyptian Christian literature flourished in Arabic. Another Islamicizing measure prohibited the use of crosses in public, including on official documents. The Greek text in the bilingual receipt for 65 sheep discussed above, is, however, preceded by a cross as were all Greek and Coptic documents of the time and another cross is used in the scribe’s signature. This symbol of the Christian faith was soon replaced by one neutral oblique stroke or two in Greek and Coptic documents produced by or for the Arab-​Muslim administration. The practice of beginning a text, including official ones, with a cross continued occasionally, however, into the eighth century ce.28 State-​sponsored religious buildings such as the great mosques in Jerusalem, Medina, and Damascus, which were erected in the early eighth century, did not 24 25 26 27 28

Sijpesteijn 2018. Thomas 1991, 28. Treadwell 2009. Sijpesteijn forthcoming. Cf. Rémondon 1952, 259.

180 Sijpesteijn

­f igure 7.3  Seal on a letter from governor Qurra b. Sharīk to the pagarch of Ishqūh/​ Aphrodito, Basileios dated Ṣafar 91 ah/​December 709–​January 710 ce (D. 13296: Papyrus Qurrah iii, detail of lead seal [oim E13756]) courtesy of the oriental institute of the university of chicago


Muhammad’s World in Egypt

­f igure 7.4  Coins of Hishām ʿAbd al-​Malik with Qurʾānic phrases from al-​Andalus, Marw, and al-​Jazīra omj 307–​f 03, omj–​307–​h 03 and omj–​308–​c 10. copyright universität jena, orientalisches münzkabinett

182 Sijpesteijn include imagery. By contrast, the seals that officials working for the Arab chancery in Egypt used on documents continued to depict images of people and animals. Clearly, administrative measures initiated in the political center of the empire were not implemented throughout the empire at the same time.29 Nor were all administrative domains considered to be alike, with the result that more private objects such as seals and documents resisted the impetus toward imperial standardization. 4

Muslim Rule?

The arrival of the Arabs thus had a clear impact on the documentation and material culture in Egypt, indicating that the new rulers brought their own customs and habits, which differed from but interacted with local practices. To what extent can the changes observed in the documents be connected to a new religious regime? The Greek part of the 22 ah/​643 ce-​dated bilingual Greek-​ Arabic receipt for 65 sheep discussed above conforms to much of the style and practices in place in pre-​Islamic Egypt, including the cross with which it begins. In other ways, too, the Greek part of this official document produced by and for the Muslim-​Arab and Egyptian-​Greek administration represents local Egyptian practices. The Greek text contains, however, one striking indication that the Arab-​Muslim context in which the receipt was written up did exercise some influence of a religious nature on the Greek text. The Arabic text starts with the well-​known invocation, “in the name of God, the Merciful, the Compassionate” (bi-​sm allāh al-​raḥmān al-​raḥīm). The Greek invocation, by contrast, reads: “in the name of God” (in onoma tou theou). While this expression is perfectly acceptable to a Christian-​Egyptian audience, it is a complete break with contemporary practice in Greek and Coptic documents, which customarily started with the mention of God, the Son, and the Holy Spirit, or a variation thereof.30 The Greek-​Arabic receipt for the sixty-​five sheep is dated in the Greek part according to the Byzantine indiction reckoning using a Coptic month name. The Arabic part is dated to “the later Jumādā of the year twenty-​two.” While there is no reference in the papyrus to the calendar used, the correlation of the two dates falling in 643 ce indicates that the Arabic part is dated according to the hijra calendar. The hijra year is, in Greek and Coptic papyri, sometimes 29 30

On the one hand, reform coins are found far and wide across the Islamic empire (Sijpesteijn forthcoming). On the other hand, a certain monetary independence can be observed in coins minted in different areas. Legendre 2020, 136. Bagnall and Worp 2004, 99–​102; 292–​95.

Muhammad’s World in Egypt


identified with the phrase “according to the Saracens” (kata sarakēnous) or “according to the Arabs” (kata arabas), but no corresponding way identifies the calendar in the Arabic texts.31 Yet the Arabic expression “according to the practice of the believers” (ʿalā sunna qaḍāʾ al-​muʾminīn), which occurs in several early legal documents in relation to the dating, suggests that the signalled legal regime had religious associations.32 The most important observation is, however, that the Arabs introduced a calendar throughout their empire established upon an event that took place in the year 622 ce, one important enough to signal a new time reckoning. Hijra years provided an absolute date in Greek, Coptic, and bilingual Arabic-​Greek documents related to the fiscal administration.33 That the hijra calendar was also considered to have an imperial and perhaps even religious meaning is suggested by a reaction in the dating systems that can be observed in Greek and Coptic papyri. The calendar based on Emperor Diocletian’s rule starting in 284/​5 ce, which later became known as the “era of the Martyrs,” was initially used exclusively in Egypt for horoscopes and gravestones.34 Only after the Arab conquest is the “era of Diocletian” attested in papyrus documents. A Greek papyrus dated 657/​8 ce is the earliest attestation of the use of this calendar in documents; it appears in Coptic documents from the eighth century ce onward.35 It implies that Egyptian non-​ Muslims juxtaposed “their” Diocletian calendar to the hijra calendar. The earliest reference to Islam/​Muslims (ahl al-​islām) and Muhammad occurs on a tombstone from Egypt dated 71 ah/​691 ce.36 Papyrological attestations, however, do not occur before the 720s ce. In the 22 ah/​643 ce-​dated bilingual receipt, the Arab-​Muslims are simply called “companions” (aṣḥāb) in the Arabic, while in the Greek they are identified as sarakēnoi and mōagaritai, a transliteration of the Arabic muhājirūn. Sarakēnos refers to individuals of supposed Arab descent in pre-​Islamic and Islamic Egyptian papyri.37 Muhājirūn is the word used in Syriac (mahgraye) and Greek (mōagaritai) sources from Syria and Egypt for the Arab newcomers. It refers to the migration that brought the conquerors from their hometowns to new territories.38 Muʾminūn, believers, is another 31 32 33 34 35 36 37 38

Worp 1985; Bagnall and Worp 2004, 300. Bruning 2015, 352–​74. The phrase has also been interpreted as referring to “the year (Ar. sana) of the believers’ rule.” For the most recent overview of the different interpretations, see Tillier and Vanthieghem 2019. Worp 1985, 114–​15. Bagnall and Worp 2004, 64. Bagnall and Worp 2004, 63–​87. El-​Hawary 1932. Mayerson 1989; De Jong 2017. Crone 1994.

184 Sijpesteijn term used to refer to the Arab-​Muslims. The use of the expression “according to the practice of the believers” has already been mentioned. The caliph is generally called “prince of the believers,” amīr al-​muʾminīn, also in Greek and Coptic transliteration. There is a unique attestation in a late seventh-​century ce Greek papyrus of prōtosymboulos, first counsellor, to refer to the caliph.39 The papyri mention palaces, maintenance, and other expenditures of the amīr al-​muʾminīn, mostly in relation to deliveries and contributions to be made by the Egyptians. They also include references concerning the caliph’s agricultural estates, but it is not clear whether these refer to the caliph personally or to caliphal rule in general. Finally, the imperial administration and the individuals working there are occasionally identified as the amīr al-​muʾminīn’s servants. The new rulers in Egypt were thus identified by their own names and terminology in the documents, but these terms are not exclusively or recognizably Muslim or even religious.40 These terms emphasize that the rulers differed from the local population, but do not indicate how that distinction was religiously defined. Muʾminūn might have referred in the seventh century ce to believers of different monotheistic religions.41 Muhājirūn, by contrast, indicates that the newcomers had migrated out of Arabia as part of the large conquests. How was their relationship vis-​à-​vis the Egyptian subject population defined? It is clear that a distinction was made not so much between Egyptian Christians and Jews on the one hand and Muslim Arabs on the other, but rather between ruled and rulers, or between taxpayers and recipients of income generated by the land and its people. In other words, the juxtaposition was defined in economic-​political rather than religious terms. This is confirmed by the way in which the papyri identify Egyptians. Egyptians occur not as “Christians” or “Jews,” but as “people of the land,” ahl al-​arḍ. The term aqbāṭ, Copts (qibṭ), or qibṭī (Coptic), occurs as well, but not with the purely miaphysite Christian connotation it would get in the high Middle Ages. Qibṭ, the Arabic rendition of the Greek Aegyptos is, after all, a reference to the inhabitants of the province regardless of their religion, and this is the meaning it has in the Arabic documents as well.42 Early references in papyri to the dhimma, the protection offered in the conquest treaties to the non-​Muslim inhabitants, has a purely

39 40 41 42

Sijpesteijn and Worp 1983, 226–​29. For the meaning of amīr, see Sijpesteijn 2013, 117–​23; Morelli 2001, 21; Morelli 2010a, 16–​17; Morelli 2010b. Cf. Donner 2010, who has argued for an inclusive monotheistic character of the earliest conquerors. Donner 2010. For a discussion of the term in the ninth-​century ce history of Ibn ʿAbd al-​Ḥakam, see Omar 2013. For the papyri, see Sijpesteijn 2020a.

Muhammad’s World in Egypt


political-​administrative rather than an explicitly religious meaning.43 A classification of different groups of inhabitants dating to 168 ah/​784 ce lists the ahl al-​dhimma, people of the covenant, as a remainder category of landholders and agricultural workers who were supposedly all Egyptian non-​Muslims.44 In a text from the finance director Sufyān b. Qurʿa for the governor ʿAbd Allāh b. Musayyab (in office 176 ah/​793 ce), the inhabitants of the districts of Ihnās, al-​Bahnasā, and its oasis are addressed as “Muslims and protected people (ahl al-​dhimma),” but here too the importance is their legal status, rather than their religious one.45 In other words, these texts identify the non-​Muslim Egyptian rural population by political-​economic category, rather than religious status, with only one group of non-​Muslims labelled as ahl al-​dhimma. It is only in the ninth century ce, two hundred years after the Arabs’ arrival in Egypt, that religious identity markers, as well as ethnic and geographical denominators, become prevalent in the documentary record from Egypt. The change is the result of a profound reconfiguration of the demographic landscape as a result of migration and conversion.46 The identification markers used for the Arab-​Muslim rulers indicate that they formed a group distinguishable and separate from the local population, with their own linguistic, cultural, and administrative customs. While the identification markers of the new rulers do not point to an exclusively Muslim identity, there are some indications that the Arabs also had identifiable religious ideas and practices that set them apart from the local population. The way in which the Arab emphasis on exclusively monotheistic expressions of belief found its way into the headings of documents has been discussed above. Some other elements in the Arabic papyri can be linked to Islamic legal prescriptions discussed in later law books. One example is the legal status of documents. The bilingual Arabic-​Greek receipt referred to above contains a phrase in the Greek which is lacking in the Arabic: “As evidence thereof [i.e., of the delivery of the sixty-​five sheep] I have drawn up this document.” According to Islamic law, it is not the written record of a legal transaction that constitutes the legal evidence thereof, but rather the observation of the transaction by two male (or four female) Muslim witnesses. The contrasting legal status of written documents in Roman and Islamic laws might explain why this phrase is absent 43 44 45 46

For a discussion of the term in legal sources as well as the earliest attestation of the term in a papyrus from Nessana, see Hoyland 2014, 55–​57. The other categories are landholders (anbāṭ), great landholders (jamājim anbāṭ), and fugitive landholders (jawālī): Diem 1984, 136–​41. Grohmann 1952, 132–​34. Sijpesteijn 2020a; Sijpesteijn 2011b.

186 Sijpesteijn in the Arabic part. The Arabic half of the receipt, in fact, conforms in another way to the same Islamic legal precept. While the Greek part of the document is written in the first person, the Arabic is in the third person, fitting the design of Islamic legal documents that are written from the point of view of the witnesses to the transaction.47 Although not many Arabic legal documents are known from the first two centuries of Islam, they overwhelmingly take this objective form.48 Other early Arabic documents contain references to Islamic legal institutions and concepts that are discussed in detail in later law books.49 Only a few papyri from the first two Muslim centuries refer to Muslim rituals. Egyptian contributions in kind, coin, and manpower are requested for the building of the great imperial mosques in Jerusalem and Damascus in the early eighth century ce, but no references to mosques in Egypt occur in the papyri before the ninth century ce.50 One exception is the Hajj, the Muslim pilgrimage to Mecca. An Arabic letter dating to 86–​99 ah/​705–​17 ce shows how two members of Egypt’s governing elite arranged their affairs to join the caliph’s pilgrimage caravan.51 Other eighth-​century ce letters contain further information of the practical and spiritual preparation for the Hajj.52 The papyri also show that Muslims in Egypt paid ṣadaqa/​zakāt, the alms tax obligatory for Muslims, from the second quarter of the eighth century ce onward. This tax on Muslim property seems at this time to have functioned as a fiscal imposition rather than a religious duty, although religious references do accompany the tax demands.53 Similarly, the poll tax introduced during the first generation of Muslim rulers in Egypt under the new (Greek) term andrismos can easily be explained as having been a relatively painless way for the new regime to raise money. It can also be interpreted, however, as implementation of the poll tax (jizya) mentioned in the Qurʾān collected from non-​Muslim monotheists.54 The Qurʾānic term jizya does appear in Arabic papyri from the early eighth century ce, but has the more extensive and general meaning of taxes collected in money rather than the poll tax per se.55 The Arabic terms used in the papyri for the poll tax are jizyat al-​raʾs and, from the early ninth century ce onward, jāliya, but again no definite religious character can be ascribed to 47 48 49 50 51 52 53 54 55

See also Sijpesteijn 2013, 68 for this observation. Only two legal documents written in the subjective, first person are known. Sijpesteijn 2013, 68–​69. Morelli 1998. Sijpesteijn 2014. Cf. Younes 2013, 220–​22. Sijpesteijn 2013, 172–​98. Morelli 2001, 19–​24. Cf. Sijpesteijn 2013, 72–​73. Frantz-​Murphy 2001, 143; Morimoto 1981, 53–​62, 136.

Muhammad’s World in Egypt


this otherwise commonly applied tribute on conquered peoples, neither in its application nor in its Arabic terminology.56 Terms, concepts, and regulations discussed in later Arabic Islamic theological and juridical treatises as part of a fully developed religious-​legal system are already present in the papyrological material from the first two centuries of Muslim rule in Egypt. It is not possible to determine if such references are part of a full-​fledged system based on the religion of Islam already present at the time. Alternatively, the mentions should be interpreted as part of the same Near Eastern religious and legal culture that formed the foundation of Islamic law and religion continually adjusting to and accommodating a new socio-​ political reality in the Arab-​Muslim empire and through interaction with the cultures encountered in the conquered lands.57 5

Patterns of Innovation and Influence

The Arab takeover of Egypt compares well with ancient and medieval Near Eastern patterns of regime change.58 The Arabs arrogated the highest offices, including governor, chief judge, supervisor of tax collection, and head of police in the capital, and also removed military responsibilities from local appointees, assigning their own military-​administrative officials. The bulk of the administrative offices remained unaltered, as well as the officials in charge. Continuation was further motivated by an uninterrupted use by the Arab chanceries in Fusṭāṭ and throughout the province of the administrative languages, terminology, and symbols that had been in place in Byzantine Egypt. The same people were thus in charge of the administration and executed their duties in a form and language that differed little from what had been in use in the preceding centuries. In short, a prevailing sense of continuity characterized the Arab takeover in the mid-​seventh century ce. This continuity, as I have argued elsewhere, resulted from both the strategic choices by the Muslim-​Arab regime and the conditions of the Arab conquest.59 The establishment of Arab rule in Egypt followed patterns of regime change in the Near East, but the prolongation of

56 57 58 59

Morimoto 1981, 176; Frantz-​Murphy 2001, 143; Legendre 2020, 138. Jizya becomes the commonly used term for poll tax in later medieval documents (Frantz-​Murphy 2001, 143). See also the idea of a shared late antique Near Eastern basis for the Qurʾānic text ( and for Islamic-​Arabic historiography, as well as ideas of piety and violence (Sizgorich 2004; 2009). Sijpesteijn 2007, 197–​98. Sijpesteijn 2013, 84; Sijpesteijn 2007, 186–​91.

188 Sijpesteijn this situation in Egypt was the result of specific historical circumstances. Only some sixty years after the Arabs established their rule over the province of Egypt were empire-​wide reforms, initiated by the Umayyad Marwānid caliphs in Damascus and aimed at greater Islamicization and Arabicization of administration and rule, implemented there. Partially motivated by internal developments in the caliphate, including financial pressure, but also brought on by external political circumstances, the Arab regime was only at that point strong and confident enough to execute a fundamental reorganization.60 The story of continuation as the prevailing approach to Egypt’s government and organization, with large-​scale change taking place two generations later, has one further dimension. The papyrological record shows that the Arabs, while initially maintaining most of the administrative, fiscal, and governmental structures and organization in place, did simultaneously introduce administrative, scribal, and fiscal innovations. The new rulers operated in a new language, collected new taxes based on a new economic infrastructure, and produced a documentation that was both more voluminous and different in its aims than that of their predecessors. Institutions such as the dīwān, the personal register on the basis of which Arabs in Egypt obtained a stipend, ʿaṭāʾ, the seasonal grazing of riding animals used in the military campaigns, and the cursus, the annual raids on Byzantine lands, required extensive documentation and record keeping. Lists of Arab “companions” (aṣḥāb) and other documents related to the organization and payment of the stipends to Arab troops and their dependents were recorded on papyrus.61 Censuses and large-​scale land surveys generated detailed reports of land holdings, descriptions of property, and individual circumstances.62 The new rulers and the administrative measures they undertook were obviously aimed at maximizing the exploitation of the earth and the income generated by it. At the same time, the documentation generated for this purpose served the Egyptians as much as it did the Arabs. From receipts for deliveries made to Arab troops travelling through the countryside to the tax demand notes and receipts related to the regular tax collection, the papyrus documents witness a striking culture of audit trails and accountability. Other early Arabic documentation revolved around regulated payments to orphans and the poor, seemingly of concern to the authorities.63 60 61 62 63

This concerned among others the relation with the Coptic church in Egypt, the Byzantine threat on the coastal cities and the Egyptian Arab regime’s preoccupation with fighting on its southern frontier (Sijpesteijn 2007). Sijpesteijn 2008; Sijpesteijn 2011a; Morimoto 1994; Bouderbala 2019. al-​Qāḍī 2015. Khan 1992, text 1; Sijpesteijn 2011a; Sijpesteijn 2012b.

Muhammad’s World in Egypt


These new administrative practices and institutions generated a new vocabulary, including Arabic titles and other words transliterated into Greek, and Coptic and Greek technical terms, among which were words translated from Latin or of Latin origin that were not in use in Egypt before the arrival of the Arabs. Symboulos for governor and prōtosymboulos for caliph are newly attested in Greek papyri from Arab Egypt, as are amiras and amiralmoumnin as Greek transliterations of Arabic terms for these officials. Andrismos for poll tax and cursus for the annual Muslim raid on Byzantium have already been mentioned. Plenty of other examples can be listed.64 Administrative practices such as the use of seals in documents to control access to the information described therein and the expansion of payments of taxes in kind, and the system that made that possible, were implemented by the new rulers as well. Administrative reforms such as the introduction of smaller agricultural tax units show a desire to rationalize the organisation of the countryside.65 Finally, the papyri show that the new rulers also had different dietary preferences and cultural customs. It is clear that the Arabs introduced an administrative, material, and perhaps even a religious and legal culture that set them apart from the local population. That the new rulers were experienced governors is not only clear from the administrative innovations described above and the extensive documentary record they generated, but also from the balance they struck in their rule between forcing change and accepting continuity, between cooperating with the local population and imposing their regime. Arab governance was, moreover, motivated by economic and administrative rather than religious concerns. Some Greek words that appear for the first time in the papyri of Arab-​Muslim Egypt point to a Levantine background.66 Administrative and scribal practices introduced at the conquest and in the decades following, show, moreover, a continuous interaction with Byzantine traditions.67 The Arabic words and names that appear transliterated in Greek papyri from southern Jordan and the Negev dating to before the Arab takeover show linguistic features that differ from those in the papyri dating to after the arrival of 64 65

66 67

See Sijpesteijn 2013, 69–​71. The introduction of the village (Gr. chorion; Ar. qarya) as a fiscal unit was introduced in Egypt by the conquerors (Legendre 2020, 135–36). The same fiscal division of the countryside based on villages (Gr. choria) can be found in the Byzantine Empire already in the seventh century (Haldon 1990, 138). Morelli 2002. See the discussion above about the introduction of the village as a fiscal unit in postconquest Egypt following Byzantine practice. Greek papyri of the seventh and eighth century CE attest scribal innovations that are clearly connected to Byzantine practices (Morelli 2001, 53).

190 Sijpesteijn the Arabs. The preconquest Arabic names and words in the Greek papyri are more Hellenized than those of the later papyri. This difference suggests that a new group of Arabic speakers, less integrated into the Greek-​speaking milieu than those who had been there before, became dominant in these areas after the conquest.68 At the same time, Greek papyri from postconquest Egypt attest Hellenized Arabic names, also suggesting a migration of those Hellenized Arabic speakers from the Levant into Egypt with the conquests.69 Other Arabic names in the postconquest papyri from Egypt show linguistic features pointing to a non-​Hellenized background, suggesting a simultaneous influx from Arabia.70 The script used by the earliest generation of Arab administrators and scribes in Egypt equally shows a connection with Arabian writing traditions as presented in inscriptions from the Peninsula and adjacent regions.71 It is striking that when Arab administrators appear in Egypt, at the end of the seventh century ce, they seem to have a Hellenized background as well. It is possible that these officials utilized administrative and governmental experience acquired in the Byzantine Empire, transferring and transmitting it into Egypt when they were appointed there. Arab but steeped in Byzantine cultural and administrative practice, these men would have been especially attractive executors of the administrative reforms. Sasanian administrative influences and iconographic traditions also reached Egypt under the Arabs as cultural practices were exchanged across the Muslim empire. Cultural and administrative practices were thus introduced in Egypt by the new rulers immediately after the conquest and in the decades following, building on Near Eastern and Roman-​Byzantine practices from Arabia and the Near East, including Egypt. The papyrological material from Arab Egypt shows how the newly established rule affected the daily life of the Egyptians and Arabs living in the country. By contrasting what changed with what remained in place, it is possible to reconstruct the administrative, material, and cultural baggage that the new rulers brought into Egypt potentially leading us to the governmental and administrative context of Muhammad’s time. At the same time, the papyri show us that the Arab rulers, with the aid of their Egyptian partners, struck a delicate balance 68 69 70


al-​Jallad 2017. See for example the Hellenized form Atias for Arabic ʿAṭiyya, the first Arab-​Muslim administrator attested in the papyri, who stood at the head of a local district in Egypt from 694 to 697 ce (Sijpesteijn 2013, 201). The name Ḥudayd, the scribe writing the Arabic part in the bilingual Arabic-​Greek papyrus dated 643 discussed above, is by contrast written with a wāw at the end, pointing to a non-​Hellenized background. I would like to thank Ahmad Al-​Jallad for clarifying this point for me. Some striking features of the script of the early Arabic papyri show that there was an influx of scribes who introduced different Arabian scribal conventions (Sijpesteijn 2021).

Muhammad’s World in Egypt


between conservatism and innovation, continuity and change, to facilitate a successful transition from a conquest society into a functioning Muslim empire.


Al-​Jallad, Ahmed. 2017. “The Arabic of the Islamic Conquests: Notes on Phonology and Morphology Based on the Greek Transcriptions from the First Islamic Century.” Bulletin of the School for Oriental and African Studies 80:419–​39. Avni, Gideon. 2014. The Byzantine-​Islamic Transition in Palestine: An Archaeological Approach. New York: Oxford University Press. Bagnall, Roger S., and Klaas A. Worp. 2004. Chronological Systems in Byzantine Egypt: Second Edition. Leiden: Brill. Bouderbala, Sobhi. 2019. “Murtabaʿ al-​Jund et manzil al-​qabāʾil: Pénétration militaire et installation tribale dans la campagne égyptienne au premier siècle de l’Islam.” In Authority and Control in the Countryside. From Antiquity to Islam in the Mediterranean and Near East (6th-​10th Century), edited by Alain Delattre, Marie Legendre, and Petra M. Sijpesteijn, 367–​91. Leiden: Brill. Bruning, Jelle. 2015. “A Legal Sunna in Dhikr Ḥaqqs from Sufyanid Egypt.” Islamic Law and Society 22:352–​74. Cromwell, Jennifer A. 2017. Recording Village Life. A Coptic Scribe in Early Islamic Egypt. Ann Arbor: University of Michigan University Press. Crone, Patricia. 1994. “The First‐Century Concept of Hijra.” Arabica 41:352–​87. Diem, Werner. 1984. “Einige frühe amtliche Urkunden aus der Sammlung Papyrus Erzherzog Rainer (Wien).” Le Muséon 97:109–​58. Dixneuf, Delphine. 2011. Amphores égyptiennes: production, typologie, contenu et diffusion (3e s. avant J.‐C.‐9e s. après J.‐C.). Alexandria: Centre d’Études Alexandrines. Donner, Fred M. 2003. “From Believers to Muslims: Confessional Self-​Identity in the Early Islamic Community.” Al-​Abhath 51 :9–​53. Donner, Fred M. 2010. Muhammad and the Believers. Cambridge, MA: Harvard University Press. El-​Hawary, Hassan M. 1932. “The Second Oldest Islamic Monument Known Dated AH 71 (AD 691) from the Time of the Omayyad Calif ʿAbd el-​Malik Ibn Marwan.” Journal of the Royal Asiatic Society 64:289–​93. Fournet, Jean-Luc. 2020. The Rise of Coptic. Egyptian versus Greek in Late Antiquity. Princeton: Princeton University Press. Frantz‐Murphy, Gladys. 2001. Corpus Papyrorum Raineri: Arabic Agricultural Leases and Tax Receipts from Egypt 148–​427 A.H./​765–​1035 A.D. Vienna: Hollinek. Grohmann, Adolf. 1952. From the World of Arabic Papyri. Cairo: Al-​Maaref. Grohmann, Adolf. 1957. “Greek Papyri of the Early Islamic Period in the Collection of Archduke Rainer.” Études de papyrologie 8:5–​40.

192 Sijpesteijn Haldon, John F. 1990. Byzantine in the Seventh Century. Cambridge: Cambridge University Press. Hoyland, Robert. 2014. “The Earliest Attestation of the Dhimma of God and His Messenger and the Rediscovery of P. Nessana 77 (60s AH/​680 CE).” In Islamic Cultures, Islamic Contexts: Essays in Honor of Professor Patricia Crone, edited by Asad Q. Ahmed, Behnam Sadeghi, Robert G. Hoyland, and Adam Silverstein, 51–​71. Leiden: Brill. Hoyland, Robert. 2015. In God’s Path: The Arab Conquests and the Creation of an Islamic Empire. Oxford: Oxford University Press. Hoyland, Robert. 2017. “Reflections on the Arabian Conquerors of the Seventh-​Century Near East.” Al-​ʿUṣūr al-​wusṭā 25:113–​40. Jong, Janneke de. “Arabia, Arabs and ‘Arabic’ in Greek Documents from Egypt.” In New Frontiers of Arabic Papyrology. Arabic and Multilingual Texts from Early Islam, edited by Sobhi Bouderbala, Sylvie Denoix, and Matt Malczycki, 3–​27. Leiden: Brill. Khan, Geoffrey. 1992. Arabic Papyri: Selected Material from the Khalili Collection. Oxford and New York: Nour Foundation in association with Azimuth Editions and Oxford University Press. Kießling, Emil, Hans-​Albert Rupprecht, and Joachim Hengstl. 1997. Sammelbuch griechischer Urkunden aus Ägypten 20, Wiesbaden: Harrassowitz. Kubiak, Władysław B. 1987. Al‐Fustat: Its Foundation and Early Urban Development. Cairo: American Research Center in Egypt. Legendre, Marie. 2020. “Aspects of Umayyad Administration.” In The Umayyad World, edited by Andrew Marsham, 133–57. London: Routledge. Mayerson, Philip. 1989. “The Word Saracen (Cαρακηνός) in the Papyri.” Zeitschrift für Papyrologie und Epigraphik 79:283–​87. Morelli, Federico. 1996. Olio e retribuzioni nell’Egitto tardo (V–​VIII d. C.). Florence: Istituto Papirologico “G. Vitelli.” Morelli, Federico. 1998. “Legname, palazzo e moschee: P.Vindob. G 31 e il contributo dell’Egitto alla prima architettura islamica.” Tyche 13:165–​90. Morelli, Federico. 2001. Corpus Papyrorum Raineri, Griechische Texte XV. Documenti greci per la fiscalità e la amministrazione dell’Egitto arabo. Vienna: Hollinek. Morelli, Federico. 2002. “Gonachia e Kaunakai nei papiri con due documenti inediti (P.Vindob. G 1620 e P.Vindob. G 18884) e uno riedito (P.Brook. 25).” Journal of Juristic Papyrology 32:55–​81. Morelli, Federico. 2010a. Corpus Papyrorum Raineri, Griechische Texte XV. L’archivo di Senouthios Anystes e testi connessi: lettere et documenti per la costruzione di una capitale. Berlin: De Gruyter. Morelli, Federico. 2010b. “Consiglieri e comandanti: i titoli del governatore arabo d’Egitto symboulos e amîr.” Zeitschrift für Papyrologie und Epigraphik 173:158–​66. Morimoto, Kosei. 1981. The Fiscal Administration of Egypt in the Early Islamic Period. Kyoto: Dohosha.

Muhammad’s World in Egypt


Morimoto, Kosei. 1994. “The Dīwāns as Registers of the Arab Stipendiaries in Early Islamic Egypt.” In Itinéraires d’Orient: hommages à Claude Cahen, Res Orientalis VI, edited by Raoul Curiel and Rika Gyselen, 353–​66. Leuven: Peeters. Omar, Hussein. 2013. “ ‘The Crinkly-​Haired People of the Black Earth’: Examining Egyptian Identities in Ibn ʿAbd al-​Ḥakam’s Futūḥ.” In History and Identity in the Late Antique Near East, edited by Philip Wood, 149–​67. Oxford: Oxford University Press. al-​Qāḍī, Wadād. 2015. “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, edited by Andreas Kaplony, Daniel Potthast, and Cornelia Römer, 59–​82. Leiden: Brill. Rémondon, Roger. 1952. “Ordre de paiement d’époque arabe pour l’impôt de capitation.” Aegyptus 32:257–​64. Sijpesteijn, Pieter J., and Klaas A. Worp. 1983. Corpus Papyrorum Raineri, Griechische Texte V. Vienna: De Gruyter. Sijpesteijn, Petra M. 2007. “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 Saddam Hussein, Proceedings of the British Academy 136, edited by Harriet Crawford, 183–​200. London: British Academy. Sijpesteijn, Petra M. 2008. “A Seventh/​Eighth‐Century List of Companions from Fustat.” In Sixty‐Five Papyrological Texts Presented to Klaas A. Worp on the Occasion of his 65th Birthday (P.Lugd‐Bat 33), edited by Francisca A.J. Hogendijk and Brian P. Muhs, 369–​77. Leiden: Brill. Sijpesteijn, Petra M. 2011a. “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, edited by Roxani E. Margariti, Adam Sabra, and Petra M. Sijpesteijn, 245–​68. Leiden: Brill. Sijpesteijn, Petra M. 2011b. “Building an Egyptian Identity.” In The Islamic Scholarly Tradition: Studies in History, Law, and Thought in Honor of Professor Michael Allen Cook, edited by Asad Q. Ahmed, Michael Bonner, and Behnam Sadeghi, 85–​106. Leiden: Brill. Sijpesteijn, Petra M. 2012a. “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, edited by Ilona Regulski, Kim Duistermaat, and Peter Verkinderen, 171–​82. Louvain: Peeters. Sijpesteijn, Petra M. 2012b. “Taking Care of the Weak: An Arabic Papyrus from the Tropenmuseum, Amsterdam.” In Inediti offerti a Rosario Pintaudi per il 65° compleanno (P.Pintaudi), edited by Diletta Minutoli, 289–​94. Florence: Edizioni Gonnelli. Sijpesteijn, Petra M. 2013. Shaping a Muslim State: The World of a Mid-​Eighth-​Century Egyptian Official. Oxford: Oxford University Press. Sijpesteijn, Petra M. 2014. “An Early Umayyad Papyrus Invitation for the Hajj.” Journal of Near Eastern Studies 73:179–​90.

194 Sijpesteijn Sijpesteijn, Petra M. 2017. “The Rise and Fall of Empires in the Islamic Mediterranean (600–​1600 CE): Political Change, the Economy and Material Culture.” In The Routledge Handbook of Archaeology and Globalization, edited by Tamar Hodos, 652–​ 68. Abingdon, New York: Routledge. Sijpesteijn, Petra M. 2018. “Expressing New Rule: Seals from Early Islamic Egypt and Syria (600–​800).” In Seals: Imprinting Matter, Exchanging Impressions, edited by Brigitte M. Bedos-​Rezak. The Medieval Globe 4:99–​148. Sijpesteijn, Petra M. 2020a. “Visible Identities: In Search of Egypt’s Jews in Early Islamic Egypt.” In Israel in Egypt: The Land of Egypt as Concept and Reality for Jews in Anti­ quity and the Early Medieval Period, edited by Alison Salvesen, Sarah Pearce, and Miriam Frenkel, 424–40. Leiden: Brill. Also to be published in Egypt and Empire: Religious Identities from Roman to Modern Times, edited by Elisabeth O’Connell. Turnhout: Peeters. Sijpesteijn, Petra M. 2020b. “The Arabic Script and Language in the Earliest Papyri: Mirrors of Change.” Jerusalem Studies in Arabic and Islam 49:433–94. Sijpesteijn, Petra M. Forthcoming. “Did the Muslim Empire have a Multilingual Language Policy?” In Navigating Language in the Early Islamic World, edited by Antoine Borrut and Alison Vacca. Turnhout: Brepols. Sims-​Williams, Nicholas. 2012. Bactrian Documents from Northern Afghanistan I: Legal and Economic Documents, revised ed. London: The Nour Foundation in association with Azimuth Edition. Sizgorich, Thomas. 2004. “Narrative and Community in Islamic Late Antiquity.” Past and Present 185:9–​42. Sizgorich, Thomas. 2009. Violence and Belief in Late Antiquity. Philadelphia: University of Pennsylvania Press. Thomas, Nicholas. 1991. Entangled Objects: Exchange, Material Culture, and Colonialism in the Pacific. Cambridge, MA: Harvard University Press. Tillier, Mathieu, and Naīm Vanthieghem. 2019. “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, edited by J. Tolan, 148–88. London: Routledge. Treadwell, Luke. 2009. “ʿAbd al-​Malik’s Coinage Reforms: The Role of the Damascus Mint.” Revue Numismatique 165:357–​81. Webb, Peter. 2016. Imagining the Arabs. Arab Identity and the Rise of Islam. Edinburgh: Edinburgh University Press. Worp, Klaas A. 1985. “Hegira Years in Greek, Greek‐Coptic and Greek‐Arabic Papyri.” Aegyptus 65:107–​15. Younes, Khaled M.M. 2013. “Joy and Sorrow in Early Muslim Egypt: Arabic Papyrus Letters, Text and Content.” PhD diss., Leiden University.

­c hapter 8

“May God be Mindful of Yazīd the King”: Further Reflections on the Yazīd Inscription and the Development of Arabic Scripts Ahmad Al-​Jallad In 2017,1 a team of Jordanian epigraphists2 discovered a truly unique inscription: a four-​word rock carving in Arabic bearing a cross. Shortly after its discovery, I was kindly invited by the team to identify and study the text. While

­f igure 8.1  The Yazīd inscription Photo: Z. al-​S alameen 1 This article is the edited, English version of Ahmad Al-Jallad. 2018. “‘Moge God Yazīd de Koning indachtig zijn.’ Nadere beschouwingen over de Yazīd-inscriptie en de ontwikkeling van de Arabische schriften.” In Mohammed en de Late Oudheid, edited by Josephine van den Bent, Floris van den Eijnde, and Johan Weststeijn, 198–208. Hilversum: Verloren. 2 The inscription was discovered during the Wadi Khḍerī archaeological and epigraphic survey project, led by Z. al-​Salameen, Y. Shdeifat, and R. Harahsheh. © Ahmad Al-​J allad, 2022 | DOI:10.1163/9789004500648_009

196 Al-Jallad undated, the inscription was clearly old—​paleographically, it fit comfortably in the sixth or seventh century ce.3 Its contents, however, made it remarkable. The inscription was a benediction for a certain yzydw ʾl-​mlk or yazīd-​w the king. Reading and interpretation, following the original publication: † dkr ʾl-​ʾlh yzydw ʾl-​mlk May God be mindful of Yazīd-​w the king At first glance, one is tempted to consider this text pre-​Islamic, placing it in the sixth century ce because of the following features. 1) The formula dkr ʾl-​ʾlh is known from sixth-​century ce Christian Arabic inscriptions. It appears to be an Arabicization of the Nabataean expression dkyr (“may so and so be remembered”), common in graffiti.4 A similar construction is known in sixth-​century ce Arabic inscriptions, e.g., the Zebed table 8.1 Inscriptions with the formula dkr ʾl-​ʾlh (may God be mindful of …)

Yazīd inscription Zebed inscriptiona

Dūmah inscriptionb

† dkr ʾl-​ʾlh yzydw ʾl-​mlk may God be mindful of Yazīd-​ w the king

dkyr (Nabataeo-​Arabic) dkr ʾl-​ʾlh ḥgʿ{b/​n}wbr slmh {b}y{r}[ḥ] šnt 4×100 +20+20+3 † may God be mindful of Ḥgʿnw son of Salamah, {in the month}… year 443 (= 548/​9 ce)

[d]kr ʾl-​ʾlh srgw br ʾmt-​mnfw w h{l/​n}nyʾ br mrʾlqys [roundel] w srgw br sʿdw w syrw w s{.}ygw may God be mindful of (personal names)

a For the latest treatment of this inscription, see Macdonald’s contribution to Fiema et al. 2015. b In the editio princeps, Laïla Nehmé read the first name as ḥgʿnw, a peculiar and unattested name, but with some plausible etymologies. It might be possible to suggest that the first letter of this name is in fact a cross carved too closely to the Arabic text. If correct, we are left with the more likely name gʿnw, attested four times in Safaitic (ociana, accessed 2 January 2018). However, this reading results in a unique formulation: an inscription with two crosses. Until another example of such a practice is found, I remain convinced of Nehmé’s original interpretation, but only wish to suggest this reading as a possibility, pending future discoveries. 3 Al-​Jallad et al. 2017. 4 Nehmé 2017, 129.

“May God be Mindful of Yazīd the King”


inscription (512 ce)5 and the Dūmah pre-​Islamic Arabic inscription (548/​ 549 ce).6 Moreover, the divine name ʾl-​ʾlh appears to be the Christian Arabic calque of Greek ho theós and Aramaic alāhā, both lit. “the god,” referring to the Christian god. This example seems to be the only spelling of the name of God in Arabic inscriptions of the sixth century ce. The spelling ʾllh—​with two lāms in a row and omitting the hamza—​appears first in the Qurʾān and Islamic Arabic texts in the seventh century ce.7 However, ʾal-​ʾilāh continues to be a way to refer to the Christian god in Islamic-​period Christian Arabic, even as late as the tenth century ce. The name ʾl-​ʾlh appears in the following pre-​Islamic inscriptions: –​ Ḥimà-​Sud PalAr 88 –​ The Zebed inscription (mentioned above) –​ The Dūmah inscription (mentioned above) –​ DaJ000NabAr1 (Nabataeo-​Arabic)9 –​ LPArab 110 2) The personal name yazīd is marked by wawation,11 a feature of Nabataean Aramaic, Nabataeo-​Arabic, and sixth-​century ce Arabic inscriptions, but almost entirely absent in seventh-​century ce Islamic texts. 3) The dāl carries a dot, a Nabataean Aramaic practice that continued into the Nabataeo-​Arabic script, but which is attested for the first time here in Arabic script proper. Note that this dot is not meant to distinguish dāl from dhāl (compare ‫[ د‬d]‌and ‫[ ذ‬dh]) as it occurs on both dkr and yzydw. While these features are relatively common in the sixth century ce, we are certainly not justified in restricting their appearance to that century. We can now establish that Arabic-​speaking Christians in the sixth century ce used variations of the Arabic script, with distinctive stylistic and orthographic characteristics. Moreover, there is no reason to assume that these practices would have ended immediately following the Islamic conquests. Therefore, the differences between the present inscription and seventh-​century ce Islamic Arabic inscriptions may not be due to chronology but rather tradition. The Yazīd inscription may 5 6 7 8 9 10 11

See Macdonald’s contribution to Fiema et al. 2015. Nehmé 2017. See Kiltz 2012 for a balanced discussion on the etymology of ʾallāh. See below for a further comment. Robin et al. 2014, 1099–​1102. Nehmé 2017, 130–​31. Nehmé 2017, 132; this example is known as the Umm al-​Jimāl inscription. Nehmé has reread the final two words of the first line as constituting a personal name: ʿbdʾl-​ʾlh, “worshipper of ʾl-​ʾlh,” comparable to ʿabdallāh of Islamic Arabic and Nabataean Arabic. Wawation refers to the nonetymological w that appears on most anthroponyms of an Arabic origin in Aramaic inscriptions, primarily Nabataean. For a full discussion of this feature, see Al-​Jallad forthcoming.

198 Al-Jallad continue a Christian Arabic writing tradition well-​attested in the sixth century ce. Eventually the official state script and writing tradition of the seventh century ce replaced it, as indicated by the Islamic inscriptions of this century and later. A major challenge to accepting the sixth-​century ce date for the inscription is identifying a suitable “Yazīd the king.” A few Yazīds known from this century could have plausibly borne the title mlk (“ruler,” “king”), but they ruled in distant south-​central Arabia. It therefore seems unlikely that one of their subjects would have carved the present text. In fact, the location of this text (near Qaṣr Burquʿ in the Jordanian panhandle) and its contents would best suit Yazīd i, the second Umayyad caliph who ruled at the end of the seventh century ce. As such, this inscription would be a precious contemporary testimony to his reign. The Christian character of the text accords with the close connections Yazīd reportedly had with the Christian Arab community. His mother and two of his wives were Christians, several high-​ranking officials in his court were Christians, and he was regarded favorably by the Christian community.12 Since the publication of this inscription in 2017, scholars have found further evidence to support a date in the seventh century ce and, therefore, its association with Yazīd i. In the following, I will summarize these points and conclude with some reflections on the development of Arabic script. 1) The Dot on the dāl A superscript dot occasionally distinguished d from r in the Nabataean script. Its use continued in the Nabataeo-​Arabic script, although by this time the r had a distinctive shape. It is attested for the first time in the Arabic script proper here.

­f igure 8.2  Superscript dot on UJadh 178 nehme 2010, 78

­f igure 8.3  dāl/​dhāl in the Yazīd inscription al-​j allad et al. 2017

Although certainly an archaic feature, the dot does not appear in any extant sixth-​century ce Arabic inscription. Therefore, it is not entirely accurate to consider it a paleographic marker of a pre-​Islamic inscription. This dotted dāl is also found in the Codex Parisino-​Petropolitanus (cpp) Qurʾān manuscript, dated to 12

Al-​Jallad et al. 2017, 320–​21.

“May God be Mindful of Yazīd the King”


the seventh or eighth century ce. The dotted dāl in the cpp was first recognized as such by François Déroche; in a recent study, Marijn van Putten connected it to the Yazīd inscription.13 The dot occurs in the word ʿibādihi (Q 24:25) “his worshippers.” Unlike the dot for the dhāl, which sits above the head of the dāl, this dot is positioned to the right just as the dāl of the Yazīd inscription.

­f igure 8.4  Dotted dāl in the Codex Parisino-​Petropolitanus

I have also identified dotted dāls in a seventh-​century ce Islamic Arabic inscription composed by Saʿd bin Dhakwān, client of Muʿāwiya. The text was discovered by Maysāʾ al-​Ghabbān in the Ḥismā region of northwestern Saudi Arabia and edited in her dissertation.14 A team of amateur Saudi Arabian explorers, known as farīq al-​ṣaḥrāʾ (“the desert team”), published a high-​quality photograph of this text in 2018.15

­f igure 8.5  Saʿd bin Dhakwān inscription Al-​S aʿīd and al-​B iṭār 2018, 127 13 14 15

See Déroche 2009 and the discussion by Marijn van Putten on his personal blog (Van Putten 2017). Al-​Ghabbān 2016. Al-​Saʿīd and al-​Biṭār 2018, 127.

200 Al-Jallad Translation: 1) Lord, convey blessings upon Muhammad Your servant and Messenger and make his reward great 2) and honor the place where he sets his feet. Written by Saʿīd son of Dhakwān, client of Muʿāwiya 3) son of Abū Sufyān, and he asks of God the best thing a servant can ask, whether of the first 4) or the last, that He grant him a noble death in His path.16 The Muʿāwiya mentioned in this text is most likely the first Umayyad caliph (r. 661–​ 90 ce), father of Yazīd i. Thus, the text would date to the same period that I suggest for the Yazīd inscription. In addition to its remarkable content, the inscription includes two instances of a dotted dāl, both on the first line. The first occurs with the name mḥmd (“Muhammad”) and the second with the word ʿbd (“servant”).

­f igure 8.6  Detail of section with the dotted dāls

Normalization: ‫على محمذ عٮذك‬ Translation: upon Muhammad, your servant The photograph makes it clear that these dots are intentional and not related to damage on the rock: the circular dots differ in color from the preexisting damage and match the patina of the main inscription. These dotted dāls are unambiguous, but admittedly rare. They were clearly not a feature of standard Islamic Arabic orthography, where dots were instead used to distinguish dhāl from dāl. Perhaps the authors of the above inscription and the cpp had been trained in another Arabic writing tradition, similar to that associated with the Yazīd inscription, and had erroneously employed a practice from this hand when composing their texts. If indeed the dotted dāl belonged to a writing style used by Christians, then the scribe of the cpp may have been (originally) a Christian. Indeed, Islamic tradition reports that Christian scribes were sometimes employed to copy early Qurʾāns, so that the dotted dāl of the former text may reflect just that situation. While the dotted dāl must certainly be a feature of the pre-​Islamic period (although it has not yet appeared in securely dated sixth-​century ce inscriptions), its use in these documents supports the hypothesis that it survived into the seventh century


Translation Sean Anthony (2018, via Twitter).

“May God be Mindful of Yazīd the King”


ce, and possibly the eighth. A closer investigation of related documents may turn up more examples of this practice. Finally, the position of the dot in both the Yazīd inscription and the cpp points to a Nabataean rather than Syriac origin. In Syriac, the supralinear dot marks the letter r, while a sublinear dot indicates d. In Nabataean, however, the d is sometimes marked by a dot above the letter.

­f igure 8.7  Syriac d

­f igure 8.8  Syriac r

The dotted dāl must therefore be regarded as inherited from the Arabic script’s Nabataean heritage rather than an imitation of a Syriac practice. 2) Wawation on the Name Yazīd Wawation is typically considered a feature of sixth-​century ce Arabic script. It is used, perhaps erroneously, to distinguish sixth-​century from seventh-​ century ce texts. While not generally employed in Islamic inscriptions of the seventh century ce, wawation is likewise sometimes absent in the sixth century ce as well, such as in the Jebel Usays inscription (528/​29 ce).17 Therefore, at least one writing tradition in the sixth century ce no longer marked personal names with a final wāw. Conversely, in at least one Islamic document from the seventh century ce, perf 558 (643 ce),18 the scribe added wawation to the personal name ʾibn ḥudayd, spelled as ʾbn ḥdydw.

­f igure 8.9  Detail of ʾbn ḥdydw in perf 558 larcher 2010 17 18

The most recent discussion of this inscription is Macdonald’s contribution to Fiema et al. 2015, 412–​13. This example is one of the earliest Islamic papyri, which records the delivery of sheep to conquering forces. See also Figures 7.1–​7.2 and the contribution by Petra M. Sijpesteijn to this volume.

202 Al-Jallad These two exceptions prove that wawation does not provide a definitive paleographic division between texts of the sixth and seventh century ce. Rather, its use indicates only that certain trends existed in both the pre-​Islamic and Islamic periods. After the publication of the Yazīd inscription, my colleague Mehdy Shaddel noticed another possible example of wawation from the seventh century ce. In a Sassanian-​style coin of Yazīd b. Muʿāwiya, the name Yazīd is spelled in Pahlavi as yzytw.19 While the rendering of Arabic d with Pahlavi t is standard, the presence of a final w on Yazīd is intriguing. Pahlavi sometimes uses w to denote the end of semantic units, but this practice is not found on other Arabo-​Sassanian coins.20 Because other words on this coin terminate in the w, it is possible that the feature reflects Pahlavi orthography rather than an Arabic original.21 Nevertheless, it remains an intriguing possible parallel that merits consideration, especially considering that it would refer to the same individual. 1

Two Other Early Christian Inscriptions and the Shape of the r

In addition to the features discussed above, the Yazīd inscription includes another peculiarity: the shape of the r is unlike examples in most seventh-​ century ce Islamic Arabic texts. Rather than a crescent, it has the form of a slightly curved line, a shape found in sixth-​century ce inscriptions.

­f igure 8.10  The r of the Zuhayr inscription (643/​644 ce) ghabban and hoyland 2018

­f igure 8.11  The r of the Yazīd inscription al-​j allad et al. 2017

I recently became aware of another early Christian Arabic inscription: an amulet discovered in the excavations of al-​Ḥīra (third season, 2010–​11), 19 20 21

Mochiri 1982. Van Putten (personal communication). Note that in the Paikuli inscription the Lakhmid king ʿamru is spelled ʾmrw, reflecting wawation in the Arabic original; see Humbach and Skjærvø 1983.

“May God be Mindful of Yazīd the King”


belonging to one ʿabd al-​masīḥ. Al-​Jumaili edited the text;22 while undated, the editor argued that the individual mentioned on the amulet should be identified as ʿAbd al-​Masīḥ bin Buqayla al-​Ghassānī, a prominent resident of al-​Ḥīra in the early to mid-​seventh ce century. The text contains several odd paleographic features, such as the y-​ḥ ligature in the word msyḥ (“Messiah”), but these irregularities may be the result of squeezing the entire text onto a single line. In other respects, the paleography would suit a date in the seventh century ce. Because the amulet includes only one name, and a particularly popular one at that, I believe it is impossible to identify the individual with much certainty.

­f igure 8.12  The ʿAbd al-​Masīḥ amulet al-​J umaili 2016

‫ىركه من الله‬ ‫عفرالله لعىد المسىح‬ Translation: Blessings from God May God forgive ʿAbd al-​Masīḥ. The amulet provides an interesting and late example of the r as a slightly curved line, here in the word ghafara. The shape compares to the r that appears in the 22

al-​Jumaili 2016.

204 Al-Jallad Yazīd inscription, the Dūmah inscription, and one example in the Ḥimà inscriptions, thus demonstrating that this letter shape continued into the seventh century ce. Even so, the ancient r occurs together with the more progressive lunate r, just as we find in texts of the sixth century ce, but never in Islamic Arabic inscriptions. table 8.2 The sletter r as a slightly curved line in sixth-​and seventh-​century inscriptions

ʿAbd al-​ Yazīd inscription Ḥimà-​Sud Masīḥ amulet (Shdaifat et al. PalAr 1 (al-​Jumaili 2016) 2017) (Robin et al. 2014)

Dūmah inscription (Nehmé 2017)

While this archaic letter shape coexists with the more evolved r of brkh on the first line, it nevertheless shows that the archaic form continued into the seventh century ce, at least optionally. Thus, the slightly curved r of the Yazīd inscription now has a parallel in the Islamic period. Unlike the Yazīd inscription, the name of the Christian god on the amulet is spelled ʾllh, a practice so far found only in texts of the Islamic period. Although the deity Allāh is known from the pre-​Islamic period, in Nabataean it is consistently spelled ʾlh; in other scripts the alif may be elided, e.g., Qaryat al-​Fāw lh, Safaitic whblh, “wahballāh.”23 Thus, the spelling here with two lāms is an isogloss shared with Islamic Arabic, although its exact significance is unclear. It is entirely possible that pre-​Islamic Christian Arabs at al-​Ḥīra—​a region for which we lack any proper documentation—​used this spelling as well. In 1936, Nelson Glueck published a Christian Arabic inscription from Kilwa in northwestern Arabia, which he dated to the end of the first millennium ce.24 Hoyland recently re-examined the text and concluded that it should date to the end of the seventh or early eighth century ce.25 While this date is certainly possible, I do not think any of its features rule out a date in the late eighth 23

24 25

For a recent discussion of the Rbbl bn Hfʿm grave inscription from Qaryat al-​Fāw, see Al-​ Jallad 2014. While the spelling lh could in theory represent an underlying lāh (without the article), a bilingual Safaitic-​Greek inscription containing the name whblh spelled as Ουαβαλλας in Greek confirms that lh reflects allāh (without an initial glottal stop); see wh 1860 and Greek 2 in Al-​Jallad 2015, 294. The Nabataean spelling ʾlh is found in theophoric names such as ʿbdʾlh, ʿabdʾallāh (cf. Islamic Arabic ʿbdʾllh). Glueck 1936. Hoyland 2018.

“May God be Mindful of Yazīd the King”


century ce, for example. Perhaps the single relevant piece of information provided by this text for the dating of the Yazīd inscription is the shape of the r in the word ḥfr-​h. If this reading is correct (Hoyland also provided an alternative where no letter appears between the f/​q and the h), then the r would have a shape similar to the examples in the Yazīd inscription and the ʿAbd al-Masīḥ amulet. Moreover, it would provide further support that this particular shape of the r survived in Christian Arabic texts well into the seventh century ce.

­f igure 8.13  Kilwa inscription hoyland 2018


Is Yazīd the King Really Yazīd i?

All of the peculiarities in the script of the Yazīd inscription now have convincing parallels in seventh-​century ce texts, so that any objections on paleographic grounds to a date in the seventh century ce cannot apply. Moreover, these peculiarities must belong to a nonstandard variety of the Arabic script—​ not the chancellery hand of the Islamic state but perhaps a traditional variant used by Christian Arabs. With the paleographic objections out of the way, the identification of “Yazīd the king” with Yazīd i comes down to this question: is it more likely that a Christian subject of Yazīd i, a pious man or soldier stationed at Burquʿ, carved this text saluting his leader, or did a wandering subject of one of our southern Yazīds of the sixth century ce carve it? One paleographic feature seems to support a date in the seventh century ce: the dotted alif (proto-​hamza), which is attested in this text, is now firmly established in seventh-​century ce texts, it is possible to suggest that the paleography slightly points toward the seventh century ce and thus a connection with Yazīd i. Unlike the dotted dāl, the proto-​hamza is a true innovation, and is not a feature of Nabataeo-​Arabic and Nabataean scripts. Therefore, this feature

206 Al-Jallad could have emerged at the earliest in the latter half of the sixth century ce, after the Ḥarrān inscription from 568 ce, the latest dated sixth-​century ce text. While the identity of our Yazīd must remain uncertain, the paleographic facts do not rule out a seventh century date and therefore a connection with Yazīd I. 3

The Writing School of Medina

That possibly a seventh-​century ce text uses formulae and orthographic features found primarily in sixth-​century ce Arabic inscriptions invites us to reconsider the development of the Arabic script. The great scholar of Arabian epigraphy and history Christian Robin suggested that a “reformation” of the Arabic script occurred during the career of the Prophet Muhammad as indicated by the orthographic differences between Arabic inscriptions of the sixth century ce and Islamic Arabic inscriptions of the mid-​seventh century ce and later.26 While he is indeed correct that Islamic writing differs considerably from the examples of Arabic writing known so far from the sixth century ce, this difference may not be strictly a chronological phenomenon. We must keep in mind that we do not yet have sixth-​century ce examples of the Arabic script from the central Ḥijāz, that is, the region of Medina and Mecca. Therefore, it is impossible to say if the writing schools in those towns had already experienced the changes we see in Islamic Arabic texts by the sixth century ce, or if indeed this reform resulted from Muhammad’s state building activities in the seventh century ce. Furthermore, all our examples of Arabic-​script writing from the sixth century ce come in the form of rock inscriptions; no Arabic writing on perishable materials has reached us from this period. Rock inscriptions, and even graffiti, tend to be archaizing. Two important texts underscore this point. The first is the Dūmah pre-​Islamic Arabic inscription. This text, written in the fully evolved Arabic script, is introduced by a single word dkyr, written in a more archaic Nabataeo-​Arabic script.27 This usage demonstrates that knowledge of older letter forms, or perhaps entire words, survived even after more evolved letter shapes had emerged. In other words, the evolution of the Arabic script from Nabataeo-​Arabic did not result in total disappearance of the latter. It is perhaps significant that such an archaism appeared in a formulaic graffito: although we tend to regard graffiti as informal, we must also bear in mind that this genre made use of rigid expressions and a fixed idiom. This fossilized 26 27

Robin 2006. Nehmé 2017 proposes the reading dkr. She argues that a small denticle present between the k and r is unintentional.

“May God be Mindful of Yazīd the King”


situation may extend beyond linguistic and idiomatic features to letter shapes and orthography as well. For this reason, we cannot infer, a priori, that the contemporary book hand was identical to rock inscriptions. The Jebel Usays inscription, a sixth-​century ce Arabic graffito from southern Syria dated to 528/​29 ce, further suggests a different register of writing. table 8.3 The Jebel Usays inscription, reading and interpretation following Macdonalda

ʾnh rqym br mʿrf ʾl-​ʾwsy ʾrsl-​ny ʾl-​ḥrth ʾl-​mlk ʿly ʾsys mslḥh snt 4x100+20+1+1+1

I, Ruqaym son of Muʿarrif the Awsite al-​Ḥārith the king sent me to Usays as a frontier guard, in the year 423 [= 528/​29 ce]

a Macdonald in Fiema et al. 2015, 412.

Comparison of this text to other examples of sixth-​century ce Arabic graffiti reveals its unique nature. Rather than conforming to the convention of a name plus date and benediction, the author describes his current condition. In addition, the inscription is the only pre-​Islamic example that does not use wawation. Both personal names, rqym and mʿrf, would normally take wawation in Nabataeo-​Arabic and sixth-​century ce Arabic. I would suggest that the occurrence of these features together is not a coincidence. Instead, the author utilized the contemporary orthography of the book hand, rather than the archaizing epigraphic spellings with wawation (and therefore deviated from the standard structure of a graffito). The Aramaicisms br and ʾnh28 further indicate that script “reform” was not a single, pan-​Arabian process. Rather, the features characterizing Nabataeo-​Arabic and Nabataean orthography were lost or reformed at different rates, and perhaps in different ways, across the Peninsula. Given that we lack any evidence of Arabic script from the area between Mecca and Medina, we may ask this question: was the Nabataean script used at Yathrib (present-​day Medina), or was a fully evolved Arabic script only introduced to these settlements in the sixth century ce? While a definitive answer must await a comprehensive survey of this region, an inscription from the region of Umm al-​Jadhāyidh in northwestern Saudi Arabia provides an intriguing hint. In 2002, S. al-​Theeb published the following text: 28

Larcher 2010 regards ʾnh as a phonetic spelling of a pausal form, but this reading now seems unlikely in light of an identical spelling in Nabataeo-​Arabic inscriptions, e.g., see Avner et al. 2013. Thus, the distinction may be due to conventions of the writing school—​that is, whether the Arabic form was spelled plene as ʾnʾ or the Aramaeogram ʾnh remained in use.

208 Al-Jallad

­f igure 8.14  Inscription #163 al-​t heeb 2002

šlm kʾr br ʾšlm may Kʾr son of ʾslm be secure dy mn ytrb who is from Yathrib The individual identifies himself as coming from Yathrib (the original name of Medina). If the author wrote in the script used at the oasis, rather than one he learned abroad, it would suggest that the Nabataean script was in use at Yathrib centuries before the rise of Islam, presumably for administrative purposes. The undated text is written in the classical Nabataean script rather than the Nabataeo-​Arabic script, suggesting that it predates the fourth century ce. This proposal could account for the peculiarities of the Arabic script found in seventh-​century ce Islamic inscriptions—​the lack of Aramaeograms, the loss of wawation, the use of dots, and the use of alif to mark internal long ā. These features would have developed at the oasis from a Nabataeo-​Arabic antecedent, rather than as a reform of the fully formed Arabic script. In fact, the archaic shape of the alif in Islamic Arabic inscriptions supports this hypothesis. The small tail at the base of the alif occurs in the Nabataeo-​Arabic inscriptions, but it is already absent from sixth-​century ce Arabic inscriptions, while it is the more common variant in seventh-​century ce Islamic Arabic. Therefore, the Islamic Arabic script developed independently from Nabataeo-​Arabic (preserving this feature while other varieties had lost it), rather than as a reform attested in sixth-​century ce Arabic inscriptions. table 8.4 The letter alif in sixth-​and seventh-​century inscriptions

S1, Islamic-period Nabataeo-​ inscriptions Arabic (Nehmé 2010, 71)

Zuhayr inscription, seventh-​century ce Islamic Arabic (see above)

Harrān inscription, 568 ce (Macdonald in Fiema et al. 2015, 414)

Jebel Usays inscription, 528 ce (see above)

“May God be Mindful of Yazīd the King”




The Yazīd inscription provides new insight into the development of the Arabic script. For the first time, we must recognize that different traditions of writing in Arabic coexisted. Further, these traditions developed alongside the Nabataeo-​Arabic script rather than in a sequential progression from Nabataean to Nabataeo-​Arabic to the Arabic script proper. The features typical of Islamic Arabic inscriptions likely emerged at the oasis of Medina through the development of an independent branch of Arabic script there from a local variety of Nabataeo-​Arabic. This form of Arabic script and its orthographic features became dominant following the rise of Islam, eclipsing other competing forms of Arabic.


“Coran” (Codex Parisino-​ Petropolitanus). Bibliothèque nationale de France, Département des manuscrits, Arabe 328, Gallica. Accessed 30 March 2018. http://​​ark:/​12148/​btv1b8415207g/​ Al-​Jallad, Ahmad. 2014. “On the Genetic Background of the Rbbl Bn Hfʿm Grave Inscription at Qaryat Al-​Fāw.” Bulletin of the School of Oriental and African Studies 77:445–​65. Al-​Jallad, Ahmad. 2015. An Outline of the Grammar of the Safaitic Inscriptions (SSLL 80). Leiden: Brill. Al-​Jallad, Ahmad. Forthcoming. “One Wāw to Rule Them All: The Origins and Fate of Wawation in Arabic.” In Scripts and Scripture, edited by Fred M. Donner and Rebecca Hasselbach. Chicago: Oriental Institute. Anthony, Sean W. 2018. Twitter, 12 March. https://​​shahanSean/​status/​ 973021331309686784. Avner, Uzi, Laïla Nehmé, and Christian J. Robin. 2013. “A Rock Inscription Mentioning Thaʿlaba, an Arab King from Ghassān.” Arabian Archaeology and Epigraphy 24:237–​59. Déroche, François. 2009. La transmission écrite du Coran dans les débuts de l’islam: Le codex Parisino-​petropolitanus. Leiden: Brill. Fiema, Zbigniew T., Ahmad Al-​Jallad, Michael C.A. Macdonald, and Laïla Nehmé. 2015. “Provincia Arabia: Nabataea, the Emergence of Arabic as a Written Language, and Graeco-​Arabica.” In Arabs and Empires before Islam, edited by Greg Fisher, 373–​433. Oxford: Oxford University Press. Ghabban, Ali I., and Robert G. Hoyland. 2008. “The Inscription of Zuhayr, the Oldest Islamic Inscription (24 AH/​AD 644–​645), the Rise of the Arabic Script and the Nature of the Early Islamic State.” Arabian Archaeology and Epigraphy 19:210–​37.

210 Al-Jallad al-​Ghabbān, Maysāʾ. 2016. “Al-​Nuqūsh al-​islāmiyya al-​mubakkira fī Haḍbat Ḥismā fī minṭaqat Tabūk: Dirāsa athariyya wa-​lughawiyya.” PhD diss., Riyadh: Jāmiʿat al-​ Malik Saʿūd. Glueck, Nelson. 1936. “Christian Kilwa.” Journal of the Palestine Oriental Society 16:9–​16. Hoyland, Robert. 2018. “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 (ssll 92), edited by Ahmad Al-​Jallad and Laïla Nehmé, 327–​37. Leiden: Brill. Humbach, Helmut, and P. Oktor Skjærvø. 1983. The Sassanian Inscription of Paikuli: Part 3.1 Restored Text and Translation. Wiesbaden: Ludwig Reichert Verlag. al-​Jumaili, Amir A. 2016. “Naqsh ʿarabi jadīd li-​Tamīma min al-​Ḥīra li-​ʿAbd al-​Masīḥ bin Buqayla al-​Ghassānī min al-​qarn al-​awwal al-​hijrī.” Majallat al-​Siyāḥa wa-​l-​Āthār 28:23–​31. Kiltz, David. 2012. “The Relationship between Arabic Allāh and Syriac Allāhā.” Der Islam 88:31–​50. Larcher, Pierre. 2010. “In Search of a Standard: Dialect Variation and New Arabic Features in the Oldest Arabic Written Documents.” In The Development of Arabic as a Written Language, edited by Michael C. A. Macdonald, 103–​12. Oxford: Archaeopress. Mochiri, Malek Iradj. 1982. “A Sasanian-​Style Coin of Yazīd b. Muʿāwiya.” Journal of the Royal Asiatic Society of Great Britain & Ireland 114:137–​48. Nehmé, Laïla. 2010. “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, edited by Michael C. A. Macdonald, 47–​88. Oxford: Archaeopress. Nehmé, Laïla. 2017. “New Dated Inscriptions (Nabataean and Pre-​Islamic Arabic) from a Site near Al-​Jawf, Ancient Dūmah, Saudi Arabia.” Arabian Epigraphic Notes 3:121–​64. ociana [Online Corpus of the Inscriptions of Ancient North Arabia]. 2014. Accessed 1 February 2018. http://​​ociana/​. Putten, Marijn. van. 2017. “Pre-​Islamic Dotted Dāl in a 7th c. Quran.” Phoenix’s Blog, 12 June. http://​​blog/​2017/​12/​pre-​islamic-​dotted-​d%C4%81l -​in-​a-​7th-​c-​quran.html. Robin, Christian J. 2006. “La réforme de l’écriture arabe à l’époque du califat médinois.” Mélanges de l’Université Saint-​Joseph 56:319–​64. Robin, C. J., A. I. al-​Ghabbān, and S. F. al-​Saʿīd. 2014. “Inscriptions antiques de la région de Najrān (Arabie saoudite 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 158:1033–​1128. al-​Saʿīd, Saʿd S., and Muḥammad Sh. Kh. al-​Biṭār. 2018. Nuqūsh Ḥismā: Kitābāt min ṣadr al-​islām shimāl gharb al-​mamlaka. Riyadh: Al-​Majalla al-​ʿArabiyya.

“May God be Mindful of Yazīd the King”


Shdaifat, Younis, Ahmad Al-Jallad, Zeyad al-Salameen, and Rafe Harahsheh. 2017. “An Early Christian Arabic Graffito Mentioning ‘Yazīd the King’.” Arabian Archaeology and Epigraphy 28:1–10. al-​Theeb, S. 2002. Nuqūsh Jabal Umm Jadhāyidh al-​Nabaṭiyya. Riyadh: Maktabat al-​ Malik Fahd al-​Waṭaniyya.

­c hapter 9

Of Siblings, Kingdoms, and the Days of the Messiah: Jewish Literary Responses to the New Order in the Land of Israel in the First Muslim Period Constanza Cordoni Research on the history of the Middle East during the late antique and early medieval “first Muslim period”1 has demonstrated that whereas the Arab conquests brought unpleasant changes to some groups in the region such as Zoroastrians and Chalcedonian Christians, for others, including Jews, the changes were not only less traumatic, but also in many respects brought an improvement with respect to the previous rule of Byzantium and Persia.2 Reuven Firestone has observed that the lack of solid evidence of especially polemical Jewish-​Muslim relations during this period3 may be due to the fact that when the Arabs conquered the regions of Byzantine Palestine and Sasanian Mesopotamia, Jews had been familiar for centuries with the notion of living under foreign rule in their ancestral homeland and in the diaspora.4

1 The expression “first Muslim period” is used in scholarly literature on the history of Palestine from a Jewish studies perspective to refer the period that stretches from the end of the Arab conquest of Byzantine Palestine, which culminated in 640 ce with the fall of Caesarea, to the end of the first Crusade in 1099 ce. See Lamdan, Erder, and Drory 2010 and Hirschberg 2007, 157–​61. The first Muslim period may be regarded as including part of Late Antiquity and the beginning of the Middle Ages or as belonging entirely within the long Late Antiquity. Even if I refer to some sources as “early medieval,” in the context of Jewish chronology most of the sources I will discuss do not belong in the Jewish Middle Ages proper, but rather in the Geonic period that coincides roughly with the second half of the first millennium of the Common Era and is itself part of the so-called “rabbinic period.” This period spans the second through the eleventh centuries, with its end marked by the movement of the rabbinic academy from Jerusalem in 1040 ce to Ramle and then to Tyre. On the rabbinic period, see Stemberger 2011, 1; on the Geonic period, see Brody 2013, 1–​11. 2 See Rustow 2013, 76–​78 and the literature cited in n. 4. 3 Firestone 2016. This statement is especially valid when we compare the evidence of Jewish-​ Muslim relations to Christian-​Muslim relations. In the latter case, a vast corpus of polemical literature is preserved. See Thomas and Roggema 2009; Thomas and Mallett 2010. 4 Jews had been a tolerated minority under Roman Christian rule since the fourth century and under Sasanian rule since the third century.

© Constanza Cordoni, 2022 | DOI:10.1163/9789004500648_010

Of Siblings, Kingdoms, and the Days of the Messiah


If we examine the Jewish literary sources of the first Muslim period to ask how Jews responded to the new circumstances inaugurated by the Arab conquests and the Islamization of the Middle East, we find but a handful of texts that more or less explicitly refer to Islam: these do not offer a consistent view into what in all probability constituted a new structure of feeling.5 It is important to bear in mind that the attitudes towards the rise of Islam to be discussed here are conveyed by members of literate elites. In this period, in which Arabicization was already underway, these anonymous authors did not write their texts in Aramaic, or for that matter in the mixture of Aramaic and Hebrew typical for the classical rabbinic literature of the previous period,6 or in Judeo-​Arabic, the language with which from the tenth century onwards Rabbanite and Karaite Jews would express themselves. They wrote in Hebrew.7 Carol Bakhos’s Ishmael on the Border (2006) was the first book-​length study to address the literary representation of Islam in the literature of the sages. Bakhos demonstrated that the rabbinic sources of both the pre-​Islamic and the Islamic period transmit different conceptualizations of Ishmael, the notional first Arab, as “the Other.” Bakhos argued that in pre-​Islamic sources Ishmael and his children function primarily as a hermeneutic Other, but from the seventh century onwards, they also represent an observable reality and come to be increasingly associated with Islam and its political hegemony.8 In the Islamic period, Ishmael is no longer just part of a merely theological argument that opposes him (the firstborn son of Abraham with Hagar) to Isaac (the son 5 Williams 1961 developed the concept of structure of feeling. 6 The so-​called amoraic period of rabbinic literature, i.e., fifth–​sixth century. 7 Hebrew is, incidentally, the language in which rabbinic literature, the formative literature of Judaism, first started to be composed in second-​century ce Palestine. On the process of Arabicization of the Near East see Levy-​Rubin 1998; Carlson 2015; and Van Bladel in this volume. On Hebrew as the characteristic language of late rabbinic literature, which displaces the earlier mixture of Aramaic and Hebrew, see Elbaum 1986; on Hebrew as an ideal language from a rabbinic perspective, see Stemberger 2010a; on Hebrew as the language to which Jews stayed loyal in changing circumstances, see Cordoni 2020. 8 See Bakhos 2006, ch. 2 and 3. Bakhos observed in this context that “despite recent studies on the nascent relationship between rabbinic Judaism and early Christianity that emphasize the intertwined, complex world of cultural and spiritual discourse in which both partook, rabbinic biblical interpretation of the classical period seems not to have been preoccupied with vying Christian theological claims as one would imagine. Of course, there are polemical midrashim, as well as midrashim that simply attest to the very interfacing of both groups. Surprisingly, however, the Christian supersessionist understanding of itself as the true Israel probably was not an immediate factor inducing rabbis to distinguish between Israel and marginalized biblical figures, although over centuries it, like Islam, certainly played other roles in the development of Jewish systems of thought, beliefs, and practices” (3).

214 Cordoni Abraham and Sarah conceived in spite of their advanced age and identified as progenitor by Jews). Expanding upon the work of Bakhos and other scholars, this article is concerned with the question of how a subset of Jewish literature that emerged in the period of transition between Late Antiquity and the Middle Ages formulated a response to the rise of Islam and to Muslim rule in the land of Israel. This literature comprises exegetical works of “post-​classical rabbinic literature”9 as well as roughly contemporary texts of Jewish late antique and medieval Near Eastern apocalyptic. The latter make use of rabbinic literature but have a different Sitz im Leben, which is why they may be referred to as “quasi-​ rabbinic” literature.10 Such an examination assumes that, even if the time and place of redaction of these anonymous texts can only be determined approximatively, it is because they were redacted at some specific point in time that they transmit views characteristic for that historical period. It is true that much of their material is taken from documents of earlier periods and their authors are not particularly concerned with “writing history.” However, these texts may be studied as witnesses to the history of ideas and culture because they convey bits and pieces of the cultural world of their redactors even if this world is seldom explicitly addressed. In what follows, I have grouped the texts to be discussed according to the aspect of the representation of the new order that is most salient, rather than to genre and chronology. In section 1, I will be concerned with how late rabbinic sources deal with the problem of the dissimilarity between Abraham’s two sons, bearing in mind that at some point their authors came to be acquainted with the Muslim post-​scriptural legends about Ishmael and his incorporation into an Arab genealogical discourse.11 In section 2, I turn to a group of sources that—​expanding on the apocalyptic texts of Daniel 2 and 7—​deal with the “world kingdoms” from the perspective of an Islamicized Near East, but do not emphasize the eschatological implications of Muslim rule. I discuss this last theme in section 3.

9 10 11

Thus designated because it emerged in the first Muslim period after the classical tannaitic and amoraic (third–​sixth century) periods. For this designation, see Alexander 2007, 227n2. Reeves 2006 suggested “post-​rabbinic.” This expression is problematic because it assumes that the rabbinic literary enterprise had ceased at the time when the apocalyptic genre of the seventh to ninth century flourished. See Bakhos 2006, 116–​19.

Of Siblings, Kingdoms, and the Days of the Messiah



The Dissimilar Siblings and Their Dissimilar Offspring

One of the literary texts of the early Muslim period most commonly interpreted as a Jewish reaction to Islam is the anonymous Hebrew narrative midrash Pirqe de Rabbi Eliezer (The Chapters of Rabbi Eliezer).12 It is assumed to have been composed in Palestine in the late eighth to early ninth century and contains evidence of its author’s acquaintance with Islamic traditions.13 A number of scholars understand the rabbinic depiction of the figure of Ishmael as a theological response to Islam’s claim to Abraham, following the identification of the figure of Ishmael with Islam.14 Chapter 30 in particular has been repeatedly read in this light.15 The first part retells the scriptural narrative of Abraham’s dismissal of Hagar and Ishmael, while the second focuses on Abraham’s two visits to Ishmael in the desert. In the second part, Pirqe de Rabbi Eliezer does not retell a scriptural story, but rather most likely presents an Islamic expansion on the narrative of Genesis 21. Three years after having reluctantly sent Hagar and Ishmael away, Abraham sets out twice to visit his son in Paran, but at his place of residence only finds his son’s current wife. The first time Abraham meets Ishmael’s first wife, Ayesha. She is so unwelcoming toward Abraham that he asks her to tell his son that an old man from the land of Canaan lets him know his threshold is not good. Understanding the coded language of his father, Ishmael remarries. His new wife, Fatima, hails from Hagar’s land—​a motif that brings the narrative closer to Scripture and distances it from the Muslim versions.16 Abraham meets Fatima the second time he attempts to visit his son. Fatima is as friendly toward Abraham as he 12 The following abbreviations are used throughout the article to refer to rabbinic sources: bSan = Babylonian Talmud, Sanhedrin; BerR = Bereshit Rabbah; DevR = Devarim Rabbah; EkhR = Ekhah Rabbah; MekhY = Mekhilta de Rabbi Yishmael; MidTeh = Midrash Tehillim; PesR = Pesiqta Rabbati; PesRK = Pesiqta de Rav Kahana; Pirqe R. El. = Pirqe de Rabbi Eliezer; SER = Seder Eliyahu Rabbah; TanB = Tanchuma Buber; ShemR = Shemot Rabbah; SivDev = Sifre Devarim; WayR = Wayiqra Rabbah; y = Yerushalmi. 13 The book may be seen as pseudepigraphic given that at some point its authorship, at least in its title and two biographical introductory chapters, came to be attributed to the second-​century Palestinian sage Eliezer ben Hyrcanos. See Börner-​Klein 2004, xxxix–​ xlviii; Keim 2017, ch. 2. 14 See Heller 1925; and Keim 2017, 192–​95, who observed that the retelling of Pirqe de Rabbi Eliezer is not simply about rejecting Ishmael and what he came to symbolize. 15 See Heller 1925; Heinemann 1974; and Schwarzbaum 1975 and their discussion in Bakhos 2006, 110–​14. 16 The Muslim versions have Ishmael marry into the Jurhum tribe. There are seventeen Muslim versions. See Bakhos 2006, 173n83; and Firestone 1990, ch. 10.

216 Cordoni can wish for Ishmael’s wife to be. Therefore, the coded language for his son this time is that the threshold is fine. As Katharina Keim noted, it is highly improbable that the story of Abraham’s visits to his expelled son is originally Jewish17 or that it was adapted to fit a Muslim agenda of “Heilsgeschichte” in which the bond between Abraham and Ishmael is most clear in the fact that they build the Kaaba together.18 Rather, it may be assumed that Pirqe de Rabbi Eliezer is here reclaiming for Judaism the positive depiction of the relationship between Abraham and Ishmael that is characteristic of Islamic tradition in general and of the narrative of Abraham’s visits in particular, even if the tenth-​century historian al-​Ṭabarī transmitted the earliest known version of it. As part of this “reappropriation” we may notice that Pirqe de Rabbi Eliezer omits the Islamic identification of Paran with the Ḥijāz and any hint about a possible involvement of Abraham in the building of the Kaaba with Ishmael. Further, here Abraham identifies himself as “an old man from the land of Canaan,” the land from which Ishmael has been and remains expelled. Is this particular Jewish counter-​narrative necessarily confrontational? Is Muslim continuation of the scriptural Ishmael narrative and its incorporation of Paran into the Arabian Peninsula polemical? According to Bakhos, to view the rendition of the story of the visits in Pirqe de Rabbi Eliezer as anti-​Islamic stems from projecting onto the text assumptions about the relations between the monotheistic faith communities in the work’s historical context, yet for which the text itself does not provide consistent evidence. In Pirqe de Rabbi Eliezer it is not Ishmael but rather his children, the Ishmaelites, who represent Islam.19 Abraham’s unequal treatment of his sons Ishmael and Isaac is already a major theme in the scriptural narrative. Several motifs in the Jewish interpretation of the dissimilarity of Abraham’s sons by Hagar and Sarah and the reasons for Abraham’s expulsion of his firstborn and that son’s mother found in sources of the classical rabbinic period (third–​sixth century)—​Ishmael’s sight is not as good as Isaac’s,20 Ishmael is sent away because of his bad behavior,21

17 18 19 20 21

As pointed out by Schussman 1980 it is not known in earlier Jewish literature. See Keim 2017, 191–​92 and the detailed treatment of the Muslim transformation of Ishmael into a Muslim prophet in Firestone 1990. See Bakhos 2006, passim. See also Bakhos 2007. See BerR 56:2. BerR 53:10: Ishmael is sent away because he practiced idolatry and sought to rob Isaac of his inheritance.

Of Siblings, Kingdoms, and the Days of the Messiah


his children will be robbers22—​take on new shape and meaning when they reappear in documents redacted after the Arab conquests. As becomes apparent,23 in addressing the dissimilarity between Abraham’s two sons, several later midrashim resort to depicting Ishmael in an increasingly negative light. It is claimed, for example, that Ishmael hated his brother24 or that by judging by Ishmael’s behavior Sarah can tell that his offspring will disappoint Abraham (Sarah is here compared to an astrologer who tells a father that his son will be a leader of robbers).25 Listening to her benefitted Abraham in that he is seen as father only of Isaac.26 Moses is depicted as foretelling that Ishmael’s descendants, again robbers, will provoke anger.27 Interestingly, neither here nor in the earlier texts that transmit this motif is it explicitly stated whose anger they will provoke. Whereas in the texts of the pre-​Islamic period it could be argued that it is God’s anger, in the sources of the Muslim period the anger may be that of the Christian rulers whom Muslims vanquished and not necessarily a specific Jewish anger. The line between the two sons of Abraham and their respective descendants is also drawn in Pirqe de Rabbi Eliezer’s retelling of the Akeda episode. Especially salient in the apparent hostility toward Islam is the interpretation of Gen 22:5, where Abraham tells Ishmael to remain with the ass while he takes Isaac to the place where he thought he would have to sacrifice him. The fact that Ishmael is unable to foresee, unlike Abraham and Isaac, leads the narrator to compare Ishmael to an ass and claim that a people who do not see resemble an ass: 22

23 24

25 26 27

The motif of the children of Ishmael as robbers is also found in a tannaitic source (SifDev 343) and in both Talmudim, where it is stated that the verse “The tents of robbers are at peace, and those who provoke God are secure, who bring their god in their hands” (Job 12:6) speaks of the Ishmaelites; see Bakhos 2006, 67–​68. With the exception of Pirq. R. El. 30, which emphasizes the undisturbed connection between Abraham and Ishmael despite the separation. See TanB Shem 24. A recurring motif in rabbinic literature found also in late midrashim is Isaac playing with Ishmael, interpreted as meaning that Ishmael sought to kill Isaac; see ShemR 1:1 (which also argues with Ishmael’s idol worship) and TanB Toledot 5. On these passages, see Bakhos 2006, 89–​93. The dissimilarity between Ishmael and Isaac is also addressed in Aggadat Bereshit. See below. See DevR 4:5. In Pirqe R. El. 41 it is claimed that the children of Ishmael cannot accept the Torah offered them because they cannot observe the commandment “You shall not steal” (Exod 20:15). TanB Toledot 1 on Gen 25:19 depicts Abraham as father to Isaac alone. The motif is also found and problematized in Pirqe R. El. 31, which contains an account of Ishmael coming from the desert to visit “Abraham his father.” See DevR Va-​etchanan.

218 Cordoni On the third day they reached Zophim, and when they reached Zophim they saw the glory of the Divine Presence resting on the top of the mountain, for it is said, On the third day Abraham looked up and saw the place far away (Gen 22:4). What did he see? A pillar of fire standing from the earth to the heavens. He said to his son Isaac: Son, do you see anything on one of the mountains? He said: Yes. He asked him: What do you see. He answered: A pillar of fire standing from the earth to the heavens. Abraham understood that the boy had been accepted for the perfect burnt offering. He said to Ishmael and Eliezer: Do you see anything on one of those mountains? They said to him: No. He considered them as [blind as] an ass and said to them, Stay here with the donkey (Gen 22:5). He said to them: Just as an ass does not see anything, so you do not see anything, for it is said, Then Abraham said to his young men, “Stay here with (‫עם‬, ʿim) the donkey” (Gen 22:5). A people (‫עם‬, ʿam) that resembles an ass.28 In comparing this Jewish counter-​narrative to Muslim traditions, it is noteworthy that Pirqe de Rabbi Eliezer brings Ishmael into a narrative where he did not appear in Scripture or in earlier rabbinic tradition29 and reappropriates for Judaism the sacrifice of Ishmael in Mecca30 as that of Isaac which in Scripture occurs at Mount Moriah.31 Even though it is not called Moriah, nothing in this rendering suggests that a setting other than the scriptural land of Canaan is envisioned. The place Abraham, Ishmael, Isaac, and Eliezer reach on the third

28 29

30 31

Pirqe R. El. 31. Translations are my own unless otherwise indicated. For Pirqe de Rabbi Eliezer I used the Hebrew text of Börner-​Klein 2004 and consulted the English translation by Friedlander 1916. In the earliest source of this midrash, BerR 56:2, which like the scriptural account in Gen 22 does not mention Ishmael, the two nameless slaves who accompany Abraham are compared to donkeys. The comparison of the nameless slaves or of Ishmael and Eliezer to a people who do not see and therefore to donkeys as part of the exegesis of Gen 22:5 is found in slightly later midrashim (PesRK 26:3 and WayR 20:2). The later TanB Va-​yeraʾ 46 follows the version without the name of Ishmael. PesR 34 interprets the ass of Zech 9:9 stating that “the ass represents the wicked who have no merit of their own and can manage to get along only by resorting to the merits of their fathers.” However, the passage appears to deal with no specific Others. As Keim 2017, 193 put it: “The very fact that Isaac is portrayed as the sacrifice negates the standard Muslim view that it was Ishmael.” That is, Moriah, since the book of Chronicles identified it with the Temple Mount in Jerusalem.

Of Siblings, Kingdoms, and the Days of the Messiah


day is called Zophim, a place name whose meaning (“those who see”) is evidently related to the events narrated.32 The ass motif reappears in a comparison of Isaac’s offspring and Ishmael’s offspring preserved in Aggadat Bereshit, an exegetical work probably composed in late Byzantine southern Italy in the ninth or tenth century:33 Another interpretation: These are the descendants of Isaac, Abraham’s son (Gen 25:19). But see, it was already said, These are the descendants of Ishmael [Abraham’s son] (Gen 25:12). … You might think that both are similar. God forbid! This is what Scripture says, For the Lord watches over the way of the righteous, but the way of the wicked will perish (Ps 1:6). This can be compared to the fat of a donkey that fell into rose oil. Even though its odor improved from the rose oil, in the end it returned to its stench like before. So the fat of the donkey is Hagar the Egyptian, for it is stated, Yet she increased her whorings, remembering the days of her youth, when she played the whore in the land of Egypt and lusted after her paramours there, whose members were like those of donkeys (Ezek 23:19–​20). She cleaved to Abraham and bore Ishmael, and it is written, These are the descendants of Ishmael, Abraham’s son. What is written after this?34 Whom Hagar the Egyptian, Sarah’s slave girl, bore to Abraham (Gen 25:12). Why does Scripture say, Whom Hagar the Egyptian … bore? To teach us that all mix with their own kind, and at the end everyone returns to the stench of their mother, for it is stated: And his mother got a wife for him from the land of Egypt (Gen 21:21). She returned to her roots: From the birthplace of her ancestors he took himself a wife. And similarly Abraham ordered his servant Eliezer, as is stated, I will make you swear … but will go to my country and to my kindred [and get a wife for my son] (Gen 24:3–​4). And also the servant said so when he delivered the word over to them, But you shall go to my father’s house, to my kindred, and get a wife for my son (Gen 24:38). Therefore the verses were put together: These are the descendants of Ishmael, Abraham’s son, whom Hagar the Egyptian, Sarah’s slave girl, bore to Abraham and These are the descendants of Isaac, Abraham’s son.35 32 33 34 35

Today Mount Scopus is called in Hebrew precisely Har ha-​Zophim. See Küchler 2013, 611–​13. See Kalahani 2003; Geula 2006; and Teugels forthcoming. This second part of the passage is transmitted in ms Oxford 2340. Aggadat Bereshit 37. I quote Aggadat Bereshit following Teugels 2001 with minor changes.

220 Cordoni The quoted text is remarkably concerned with distinguishing between the offspring of Abraham by Hagar and Sarah.36 To do so it resorts, like most late midrashim, to earlier rabbinic material, which it embeds in new arguments, and to scriptural proof texts. The motif of Ishmael associated with the ass is first explained as due to his mother’s immoral ways, rather than to Ishmael’s own ignorance, and illustrated with a rabbinic parable: Abraham is compared to rose oil which cannot neutralize the foul odor of Hagar’s immoral life in her homeland Egypt. In the second part of the passage, the ancestral homelands of Hagar and Abraham are said to be the reasons for the dissimilarity of Ishmael and Isaac and their respective genealogies. While the former have their origins in Egypt, the children of Isaac, and therefore of Israel, are at home (and even look for wives) in Abraham’s homeland.37 The text thereby constructs Isaac and his descendants as the actual Abrahamic people, while it expels the children of Ishmael to Hagar’s Egyptian homeland.38 Such Jewish traditions may have emerged in the Arabic milieu as a way of “articulating the competition between Judaism and Islam for the legacy of the Abrahamic covenant”39 before they wandered westward to Christian Europe where these late midrashim appear to have been composed. The sources of both early and late rabbinic literature needed to come to terms with the essentially problematic scriptural narrative of Gen 21, in which a father sends his own son and that son’s mother away. The sources of the first Muslim period thus far discussed, transmitted in Pirqe de Rabbi Eliezer and Aggadat Bereshit, share a concern for the biblical figure of Ishmael and


37 38 39

The argument is made on the basis of different verses in Aggadat Bereshit 73. See also Pirqe R. El. 29 which distinguishes between Ishmael, the son born before Abraham’s circumcision, and Isaac, the son of circumcision. The contrast between the two sons is also explicit in Pirqe R. El. 31 and 30. The name Ishmael is explained as having an eschatological meaning. Further discussion on this topic follows below. However, as I have argued elsewhere, the Abrahamic homeland is not unequivocally in the land of Israel. See Cordoni 2018. For the dissimilarity between Ishmael and Isaac in Aggadat Bereshit, see also c­ hapters 61, 73, and 81. As Mikva 2012, 282n23 pointed out with respect to the medieval Midrash va-​Yosha. Like several other late rabbinic sources (e.g., ShemR 1:1, MidTeh 11:4, and MidTeh 71:3), Aggadat Bereshit highlights the negative similarities between Ishmael and the archetypal scriptural villain Esau and their descendants. On the post-​biblical affiliation of Ishmael and Esau in pre-​Islamic rabbinic sources, see Bakhos 2006, 54–​63. The biblical anti-​heroes are imagined as designing plans against Isaac and Jacob respectively, while their descendants (the children of Ishmael) or associates (Esau and his chiefs) are conceived of as servants of the children of Israel. The motif is also found in post-​Islamic sources. See ShemR 1:1, Pirqe R. El. 38, Aggadat Bereshit 2, 43, and 81.

Of Siblings, Kingdoms, and the Days of the Messiah


his descendants. They not only update the post-​scriptural Jewish afterlife of Ishmael in rabbinic sources, arguing for a causal connection between this afterlife and the behavior of his descendants, but also reclaim Ishmael for Judaism from Muslim legends about Abraham’s visits. Even if one of the major goals of these midrashim may have been to emphasize, albeit in a new interpretive style and with new literary forms, the earlier scriptural and rabbinic notion that the seed of Isaac, the children of Israel, are God’s chosen people, the increasing animosity over time with which these sources treat Ishmael cannot be ignored. We may understand the antagonism depicted in these texts as evidence of a general rabbinic attitude toward the new religious and political order of the first Muslim period. That is, as a more specific response to the challenge posed by the rise of another Abrahamic faith and to the Muslim perception of Ishmael as Arabized progenitor of the umma. 2

Reordering the Kingdoms

The prophet Daniel recounted Nebuchadnezzar’s dream of a statue (Dan 2:31–​ 45) and interpreted it in terms of the four world kingdoms Babylonia, Media, Persia, and Greece, an interpretation which would itself be reinterpreted time and again in Jewish and Christian literature. Scholars refer to the ensuing “model” of world history as a succession of world kingdoms as the four-​ kingdom doctrine.40 One of the post-​classical rabbinic texts that adapts the four-​kingdom doctrine of earlier traditions is Pirqe de Rabbi Eliezer. This work reinterprets the list of kingdoms of classical rabbinic literature from the perspective of the early Muslim period, and adds the rule of the “children of Ishmael” and sovereign Israel (or the Messiah) to the lists of “future” rulers.41 In the context of Abraham’s seventh trial,42 for example, Pirqe de Rabbi Eliezer reports an 40

41 42

Dan 7, which relates Daniel’s vision of the four beasts, also connects to the interpretation of the dream in Dan 2 in terms of the kingdoms that will rule the world until the end of time. The apocalyptic genre of Daniel and this four-​kingdom topos are among the most richly received of the prophetic texts of the Old Testament. On the reception of Daniel see Steinschneider 1874 and 1875; Collins 1998; Collins and Flint 2001; and Stemberger 2010b. On the four-​kingdom doctrines, see Swain 1940 and Wendehorst 2015. One of the probable sources used and adapted by the author of Pirqe de Rabbi Eliezer for this passage is BerR 76:6 where reference is made to the three kingdoms and the wicked kingdom. Another is BerR 44:15. In ­chapters 26–​31 Pirqe de Rabbi Eliezer treats the Abraham narrative of Scripture as one marked by ten trials, including Abraham’s migration to the Promised Land, the expulsion of Ishmael and Hagar, and the binding of Isaac.

222 Cordoni interpretation of the animals God shows Abraham in Gen 15:9 in the name of the second-​century Rabbi Aqiba.43 Combining historical and eschatological atomizing exegesis of the verse,44 the sage interprets each of the animals to be sacrificed45 as referring to either a ruler over Israel or the children of Israel themselves: Rabbi Aqiba says: The Holy One, blessed be He, showed to our father Abraham (at the covenant) between the pieces four kingdoms, their dominion, and their downfall, for it is said, He said to him, “Bring me a heifer three years old (meshuleshet)” (Gen 15:9): this is the fourth kingdom, the kingdom of Edom, [which is] like a heifer on the grass (Jer 50:11); a female goat three years old (Gen 15:9): this is [refers to] the kingdom of Greece, for it is said, then the male goat grew exceedingly great (Dan 8:8); a ram three years old (Gen 15:9): this is [refers to] the kingdom of Media and Persia, for it is said, As for the ram that you saw with the two horns, these are the kings of Media and Persia (Dan 8:20); a turtledove (tōr) (Gen 15:9): this is [refers to] the children of Ishmael. This expression tōr is not said in the language of the Torah, but in the Aramaic language, in which tōr means ox (shōr), for when the male ox is harnessed to the female, they will open and break all the valleys; and a young pigeon (Gen 15:9): this is [refers to] Israel,46 who are compared to a young pigeon, for it is said, O my dove, in the clefts of the rock etc. (Song 2:19).47 On the surface the quoted text inexplicably reverses the order of the kingdoms of the probable source48 and replaces the kingdom of Babylon with the rule of the children of Ishmael, so that the list comprises three kingdoms and the 43

44 45 46 47 48

Other textual witnesses read Rabbi Eliezer instead, the same second-​century sage to whom the work at some point came to be attributed. However, the reading of the Venice print is not insignificant in this context, given that Rabbi Aqiba came to be especially known for his role in the second Jewish revolt against the Romans in 132–​35 ce, the so-​ called Bar Kokhba revolt. Bakhos 2006, 125 described this text of Pirqe de Rabbi Eliezer as combining historical predictive interpretation (which explains scriptural texts as foretelling events that are already known to have happened) with eschatological predictive interpretation. These animals are referred to as “pieces” and the narrative episode as the “covenant of the pieces.” ms jts Enelow 866 reads here “the people of the land of Israel.” Pirqe R. El. 28. The four kingdoms of BerR 44:15 are Babylon (= heifer), Media (= she-​goat), Greece (= ram), and Edom/​Rome (= turtledove and pigeon). On the probable sources of Pirqe R. El. 28, see Bakhos 2006, 126.

Of Siblings, Kingdoms, and the Days of the Messiah


children of Ishmael. In fact, what appears to be replaced is the period of foreign rule over Israel. As John Reeves has argued, Pirqe de Rabbi Eliezer thus covers the first eight centuries of the Common Era, alluding to Roman (Edom49), Byzantine (Greece), Sasanian (Media and Persia), and Arab (the children of Ishmael) rule.50 Later in the same chapter, the qualification of the heifer, the goat, and the ram in Gen 15:9 as meshuleshet, translated above as “three years old,” is interpreted twice, whereby a line is drawn between the kingdoms (Rome, Greece, and Media-​Persia) on the one hand, and the children of Ishmael and Israel, which are not called kingdoms, on the other.51 Once the distinction is made between kingdoms and the Abrahamic peoples, Pirqe de Rabbi Eliezer proceeds to draw an additional line: a line that separates the four first pieces (all standing for rulers over Israel) from the young pigeon that symbolizes Israel. In this retelling of the scriptural narrative, Abraham is said to have spared only this last sacrificial piece, while the rest he cut in two. After comparing the rule of the four kingdoms to a day of God when, during the evening it will be possible for the Messiah to come, the midrash depicts Abraham in evening prayer as follows: Abraham stood up and prayed before the Holy One, blessed be He, so that His children would not be enslaved by the four kingdoms. A deep sleep fell upon him and he slept, for it is said, [As the sun was going down,] a deep sleep fell upon Abram (Gen 15:12). Can a person who has lied down and

49 50


The names Edom, Esau, and Seir are used in rabbinic literature to refer to Rome. Reeves 2006, 15 also pointed out that under the rubric Aggadat Rabbi Ishmael, this exposition of Gen 15:9 enjoyed an independent circulation with different recensions in manuscript and print form. The text he quotes (ms huc 75 fol. 38b; see Reeves 2006, 14 and n35), for example, adds to the identification of the turtledove with the children of Ishmael a proof text, “not like his posterity” (Dan 11:4), which, according to Reeves, adds to the historical precision of the passage: In the scriptural context the mighty king (Alexander) will be able to oppose the aggression by a Persian king (Achaemenids), but those who succeed him (the Diadochi) will not rule as he did. In the late Hebrew apocalyptic context, the Greek king may be seen as Heraclius, whose reign does not endure after defeating the Sasanian Khosrow ii and who is in turn followed by Arab rulers who are not his descendants and who will rule differently. According to the fourth-​century sage Acha ben Jacob, “threefold” (meshuleshet) is said exclusively of powerful heroes, but his coeval Mesharshya sees the expression as meaning three future instances of foreign rule in the land of Israel, whereby it is the third rule that is understood as the one that will witness the return of a Davidic kingdom. Although both sage names are preceded by the expression “rabbi” and not “rav” as was usual in Babylonia, they still appear to refer to Babylonian sages.

224 Cordoni sleeps pray? From here you learn that Abram lay down and slept because of the strength of the prayer that his children overcome these four kingdoms, for it is said, and dread, a great darkness descended upon him (Gen 15:12). Dread: this is the fourth kingdom, for it is written, a fourth beast, terrifying and dreadful and exceedingly strong (Dan 7:7). Darkness: this is the kingdom of Greece, which will darken the eyes of the children of Israel with respect to the commandments of the Torah. Great: this is the kingdom of Media and Persia which was great in selling Israel for nothing. Descended: this is the kingdom of Babylon, for in their hand the crown of Israel fell. Upon him: these are the children of Ishmael, for upon them the Son of David shall break forth, for it is said, His enemies I will clothe with disgrace, but on him, his crown will gleam (Ps 132:18). Rabbi Azariah says: These kingdoms were only created as burning wood for Gehenna, for it is said, [When the sun had gone down and it was dark,] a smoking fire pot and a flaming torch passed [between these pieces] (Gen 15:17).52 A verse in Genesis is the point of departure for an exegetical narrative in which Abraham is concerned about whether Israel will be able to overcome and avoid being enslaved by the four kingdoms and the children of Ishmael.53 Even if in the atomizing reading of Gen 15:12 the children of Ishmael are again not referred to as a kingdom, and even if they are mentioned after a list of four kingdoms, they seem, notwithstanding their Abrahamic ancestry, to have more in common with the rest of the world rulers than with Israel. This assessment is suggested by the reference to “these kingdoms” in the statement of Rabbi Azariah that comes after the mention of the Ishmaelites. In both of the passages of Pirqe de Rabbi Eliezer, the children of Ishmael are the final item on each list, and therefore differ from the rest in belonging in the evening of time.54 Like their father Ishmael, these children are marginal to but still within Jewish tradition.55 52 53

54 55

Pirqe R. El. 28. In BerR 44:17, the probable source of Pirqe de Rabbi Eliezer, the verse is divided into four parts and the kingdoms are listed first according to a past-​to-​present chronology as Babylon (“dread”), Media (“darkness”), Greece (“great”), and Edom/​Rome (“descended upon him”), and second, as in Pirqe de Rabbi Eliezer, according to a present-​to-​past chronology as Edom (“dread”), Greece (“darkness”), Media (“great”), and Babylon (“descended upon him”). It is noteworthy that the first version of this list of world kingdoms and that interpreted from Gen 15:9 begin with Rome, but they refer to it as the fourth kingdom, which reveals that Rome’s position as last kingdom was an established topos. A pun on “evening” (erev) and Arab appears to be intended recurrently in this chapter. With respect to the ambivalence in the rabbinic attitudes toward Ishmael, Bakhos 2006, 1 observed: “since the medieval period, Ishmael has often, but not exclusively, symbolized

Of Siblings, Kingdoms, and the Days of the Messiah


The late midrash Seder Eliyahu (ninth century), a text of uncertain provenance, but probably composed outside of Palestine, preserves a statement concerning the scriptural Ishmael that acknowledges the phenomenon of Muslim rule as follows: “As a reward for the no more than slight token of respect which Ishmael showed his father, the Holy One, blessed be He, did not give to any people or kingdom the authority to rule over his children.”56 While on the surface the statement may appear to deal with the reward that awaits those who, like the scriptural Ishmael, show God a minimum of deference, the allusion to the exalted position of his children may likewise be read as a rationalization of the political success of the caliphate. Although neither the reward given to the offspring of Ishmael nor that given to Esau and his offspring57 are explicitly called kingdoms, this text in Seder Eliyahu appears to participate in the discourse of the world kingdoms theory, reducing the kingdoms to two: Rome and the Caliphate. It does not, however, manifest the animosity towards Muslim rule that is apparent in the second Pirqe de Rabbi Eliezer text quoted above.58 Although it does not address how kingdoms succeed one another, Aggadat Bereshit is very much concerned with the problem of Israel’s relation to the

56 57 58

Islam. Ishmael’s place nonetheless is marginal, but it is marginal within the Jewish tradition.” SER 14 (p. 65). In this context also Esau’s minimal repentance and sympathy for his brother is said to have been rewarded with Mount Seir and with a hundred provinces. Other aspects of Seder Eliyahu may be read as attesting to a similar neutral or even benevolent attitude toward the new late antique order. This midrash notably has a first-​person speaker who narrates encounters with different characters while walking on a journey. Because the wandering rabbi, who claims to hail from Yavne, talks in the opening chapter with a Zoroastrian priest in a city that may be identified as Ctesiphon and in other chapters with the sages “at the great academy which is in Jerusalem” the historical reality that the work may have alluded to has baffled its readers. Even if scholars have determined an approximate date of composition in the ninth century, the work is generally regarded as disconcertingly conspicuous for its lack of references to Islam or to Muslim rule. However, the rabbi’s pervasive mobility and the repeated reference to an academy in Jerusalem may be subtle allusions to Jewish life in the land of Israel under Muslim rule. After the Jewish revolt of 132–​35 ce, Hadrian prohibited Jews from settling in Jerusalem. Even if there is no evidence for the obliteration of Jewish life in the city, according to a number of sources a renewed Jewish settlement of the city appears to have been permitted shortly after the conquest of Palestine (see Gil 1992, 65–​74; Levy-​Rubin 2011, 53; Frenkel 2013, 108). In the middle of the tenth century, the Palestinian rabbinic academy moved from its traditional location in Tiberias to Jerusalem, where it gained renown and remained until 1040 ce. According to a version of Pirqe R. El. 30 preserved in ms huc 75, one of the things that the children of Ishmael will do in the future is that they will repair the house of study; in the standard reading (see below) it is the Temple that they will repair. See Reeves 2006, 71n30.

226 Cordoni world kingdoms, and in particular with the latter inhabiting Israel’s land.59 The following parable illustrates how Israel perceives foreign rule in the diaspora and in the Land: Rabbi Acha said: The Holy One, blessed be He, said, I only created a dwelling place in order that you will do my will and fear me on these conditions. And so it [Scripture] says: I said, Surely you will fear me, take my lesson and the dwelling place will not be destroyed (Zeph 3:7). He drove out the enemy before you, and said, Destroy! (Deut 33:27). At that time Israel lives in peace (Deut 33:28). Whenever the kingdoms of the world dwell in safety, Israel does not dwell in safety. This can be compared to a partridge that sang in the house of his master. When he sat down to dine, the partridge sang. After a while, his master brought a falcon. When the partridge saw it, he fled under the bed to hide itself, and did not open its mouth anymore. The king came to dine and said to his house-​servant, Why does the partridge not sing? He said, Because you brought a falcon to it; it sees it and is afraid and therefore does not sing. Take the falcon away and it will sing. Likewise it is with Israel in this world when they are placed outside of the land of Israel while the kingdoms of the world live in their land.60 … Also the Holy One, blessed be He, is, as it were, not visible in the world, until the moment that He uproots the kingdom of Edom, for it is said, God is king over the nations (Ps 47:9), and at that time God sits on his holy throne.61 The words of the amora Rabbi Acha tend to be interpreted as the expression of a Jewish wish for the end of Roman rule, a wish usually read in terms of anti-​ Christian sentiment.62 However, both the fact that the redaction of this work 59 60

61 62

On several occasions Aggadat Bereshit has especially harsh words for Ishmael and the children of Ishmael. It may be assumed that these figures refer to Muslim rule, but they are not, as in earlier tradition, identified as kingdom. ms Oxford 2340 reads: “As long as they are in the lands of their enemies, they are like a pigeon standing in fear in front of a falcon. … Moses said to Him: Master of the worlds, as long as Israel is among the peoples and the peoples are in Your Land, Israel is not visible in the world.” Aggadat Bereshit 58. In the introduction to the translation and discussion of the last chapter of Aggadat Bereshit, ­chapter 84, Teugels observed that “the nations”—​about which negative references abound in the work, as well as in connection with “the kingdoms of the world” and Esau/​Edom—​ are never explicitly identified as Christian nations. However, Teugels 2001, xxx viewed this identification as highly likely when she asked “but who else could be meant in the probable time and place when ab was composed?” It is likely that “the nations of the world who

Of Siblings, Kingdoms, and the Days of the Messiah


tends to be dated to the ninth century at the earliest and the wider context of the quoted text63 may suggest that this late midrash is not simply disclosing an anti-​Christian sentiment related to the Byzantine cultural context in which the work probably emerged.64 Rather, the midrash, written outside of the Land and of the Islamicate world, may rather express a desire for the end of any foreign rule in the Land, i.e., including that of the Muslim caliphate. The rule of the latter in their land, the midrash suggests, or any kingdom living in Israel’s land in past and present, causes Israel not only to be dispersed but also to be invisible. Even God appears to be invisible when foreign kings rule the Land. In Mishnat Rabbi Eliezer, a text attributed to the late first–​early second-​ century sage Eliezer ha-​Gelili but whose final redaction can be dated to our period of interest,65 we find three passages related to foreign rule and the concomitant dispersion of the Jews from their Land within a lengthy exposition of the meaning of the four species of the lulav. These species are said to correspond to “the four kingdoms that have scattered Israel and conquered their Land”66—​identified here as Babylonia, Persia, Greece, and Edom—​and to the four righteous men whom God appointed to save Israel and spread the Torah during the rule of each of these four kingdoms. The midrash adds that God does not forget his children after the time of the kingdom of Edom, even if no righteous men have been named for this period.67 So far, we cannot identify any allusion to Muslim rule, but later in the same chapter and as part of a description of the signs Israel may rely on to know of their salvation in the end of time, the midrash introduces another kingdom: Rather, three signs the Holy One, blessed be He, gave with them. And on all three of them Israel rely, to wit: The sign of the toes, for it is said, there was a man of great size, [who had six fingers on each hand, and six toes on each foot, twenty-​four in number; he too was descended from the giants.] (2 Sam 21:20); [the sign of] reproaching and reviling, for it is said, When he taunted Israel (2 Sam 21:21); and then [the sign of] the song, for it is said, David spoke to the Lord the words of this song (2 Sam 22:1). Israel will also

63 64 65 66 67

say that God has a son” refers to Christianity, but is the same reference necessarily the case with the alternative reading in Aggadat Bereshit 31 “the deceivers”? The quoted text is part of a homily based on Gen 37:1. Here we read that Jacob settled in the land of Canaan where Isaac had lived as an alien. Or, for that matter, ex post facto, the end of Roman rule in the land of Israel. On the date of composition of this midrash, also known as Midrash of 32 Hermeneutic Rules and as Midrash Agur, see Stemberger 2011, 33–​34. Enelow 1933, 103. Enelow 1933, 103–​4.

228 Cordoni be saved in the final days by these three signs. By the reign of the kingdom of the toes (malkhut etsbaʿot), this is the kingdom of Ishmael, for it is said, As the toes of the feet were part iron (Dan 2:42). By reproaching and reviling, for it is said, how your enemies flung abuse etc. (Ps 89:52), and after that, they will be redeemed and they will sing the song, for it is said, Blessed be the Lord forever. Amen and Amen (Ps 89:52).68 While the passages of Pirqe de Rabbi Eliezer quoted above do not refer to the children of Ishmael explicitly as a kingdom, by mentioning them at the end of a list of world kingdoms they acknowledged Muslim rule. The text in Mishnat Rabbi Eliezer goes a step further: The iron toes of the statue in the king’s dream of Daniel 2 are interpreted as referring to the “kingdom of Ishmael,” which thus takes up a place that earlier midrashim had reserved for Byzantine Rome. Although Mishnat Rabbi Eliezer does not appear to be complimenting Muslim rule with the wording “kingdom of the toes,” it concurs with Pirqe de Rabbi Eliezer because they both identify the time of Muslim rule as the one preceding the advent of the Messiah and the final deliverance of Israel.69 It is precisely in sources of the apocalyptic genre (discussed below) that Muslim rule is consistently acknowledged as the kingdom of Ishmael and as ruling in Israel’s Land in the end of time. The sources discussed in this section demonstrate that in diverse interpretive contexts and with different strategies, Jewish authors of the first Muslim period adapted what had evolved into a stable topos of world kingdoms, according to which Rome became the oppressing rule that precedes the days of the Messiah, to include the rule of their own time. 68 69

Mishnat Rabbi Eliezer 5. Translated following Enelow 1933. Another anomalous source related to the four-​kingdom doctrine that refers to Muslim rule as a kingdom is transmitted in the manuscript family and in the printed edition of the exegetical midrash Ekhah Rabbah. The passage in question (EkhR 1:42) interprets Lam 1:14 by means of a list of pairs of kingdoms—​Babylon and the Chaldeans, Media and Persia, Greece and Macedon, Edom and Ishmael. The pairs are depicted as alternating in rigidity, as suggested by Dan 2:42, according to which a kingdom is in part strong and in part fragile. The pairs Babylon/​the Chaldeans and Greece/​Macedon are portrayed as severe, while Media/​Persia and Edom/​Ishmael are characterized as lenient. In line with the fact that each of the first three pairs of names are pairs of synonyms that denote one nation in each case, the other textual witnesses of Ekhah Rabbah read Edom and Seir. However, even if the midrash is evidently not concerned with accurately characterizing foreign rules, but rather with making a point of the expressions nishqad (“bound”) and ishtaregu (“fastened together,” “interlaced”) of Lam 1:14, the fact remains that this version at least substitutes Seir, an allusion to Rome, with Ishmael, thus acknowledging him as king.

Of Siblings, Kingdoms, and the Days of the Messiah



The Kingdom of Ishmael and Its Role in the Jewish Eschaton

As we have seen in the previous section, rabbinic Jews of the first Muslim period saw the need to redefine material found in earlier rabbinic sources that depict foreign rule over Israel to accommodate Muslim rule in a scheme of Reicheschatologie.70 From a Palestinian perspective, Rome had been vanquished, so that even if it continued to be called “the fourth kingdom” and “the wicked kingdom,” it was necessary to account for the fact that another type of kingdom had succeeded it. There is, however, an apparent tension between referring to the Ishmaelites in terms of kingdom and as the children of Ishmael. Likewise, there is tension between considering them as just another oppressor and viewing them as announcing the time of the Messiah that precedes the end of time. To what extent this latter identification may be regarded as indicative of an actual widespread Messianic fervor is difficult to determine. However, ample evidence precisely during the first Muslim period indicates a resurgent interest in the events of the eschaton in the late antique and early medieval Jewish apocalyptic.71 Apart from the presence of other, Abrahamic historical dramatis personae unknown to the apocalyptic of the Hellenistic and Roman imperial periods and possible references to circumstances of the second half of the first millennium, what is most conspicuous for the late Hebrew apocalyptic is a language shared with the Christian and Muslim counterparts. This shared discourse may be described with Reeves as a “textual commerce” of claims and counterclaims about a certain set of topoi72 and of competing exegeses of scriptural texts.73 The remarkable commonalities in structure and style of a number of Jewish apocalyptic texts of the Islamic period attest to a sort of literary trend. Their more or less stable fabula in combination with a highly coded language are of interest for the history of the culture and ideas of the end of Late Antiquity in 70 71

72 73

The expression “imperial eschatology” is Paul Magdalino’s English rendering of the German “Reicheschatologie” (2014). According to Keim 2017, 139, Pirqe de Rabbi Eliezer “seems to have been written under the impression that the eschaton was imminent. This makes sense, given that it was probably composed during the so-​called apocalyptic revival of the 7th–​9th centuries ce.” See also Spurling 2017. For a synopsis of the elements of the fabula in Jewish, Christian, and Muslim apocalypses, see Reeves 2006, 18. Reeves 2006, 5 pointed out that unlike Jewish apocalyptic of the Hellenistic and Roman imperial periods, late antique and early medieval apocalyptic was preceded by a textualization of authority: “sacred scripture (écriture) supplies both the raw material and the ultimate rationale for the conceptual elaboration of Late Antique Near Eastern apocalyptic.”

230 Cordoni general and for an appreciation of Jewish responses to early Islam in particular.74 This corpus of literature turns to the collective character of the children of Ishmael and the kingdom of Ishmael and to individual figures of Muslim rulers (referred to as kings), all of whom are in different ways involved in shaping these Jewish end-​of-​time narratives.75 Before turning to sources of the quasi-​rabbinic late Hebrew apocalyptic, let us consider selected passages of Pirqe de Rabbi Eliezer that relate the name Ishmael and his children to the events of the eschaton. The name Ishmael—​to be more precise the challenge to the exclusiveness of Israel manifest in its morphology and semantic value—​is twice problematized in exegeses that interpret it as an allusion to ominous times. We read, for example: “Balaam said, In none of the seventy languages which the Holy One, blessed be He, created in His world, did He place His name apart from in [the name] Israel. Since He made the name Ishmael similar to that of Israel, woe to him who lives in his days, for it is said, Alas, who shall live when his name is El? (Num 24:23).”76 Elsewhere the meaning of the name Ishmael is explained as alluding to the fact that God will hear (ishmaʿ el) Israel’s complaint about the rule of the children of Ishmael, especially in the land of Israel: Ishmael. Whence do we know about Ishmael? For it is said, [Now you have conceived and shall bear a son;] you shall call him Ishmael (Gen 16:11). Why was his name called Ishmael? Because in the future the Holy One, blessed 74



Adolph Jellinek 1855, xvii–​xviii already noted this interest as one of his major motivations for producing an edition of a number of texts (so called minor midrashim) in the nineteenth century. As Reeves 2006, 4 warned, reading apocalyptic texts for the sake of extracting from them references or allusions to historical events does not do justice to the genre, but he conceded that, like cultural artifacts in general, the texts of this genre are embedded in a specific historical context, which to a certain extent they reflect: “Oppression, hardship, and perseverance under adverse circumstances were the tangible conditions of life for Jews under both Christian and Muslim rule, and being one of the approved cultural expressions of those experiences (among others), apocalyptic literature reflects the emotional peaks and valleys engendered by the seemingly hostile forces of history.” See also Hoyland 1997, 258, who observed that “though their usefulness in reconstructing events is limited, particularly as one needs to know the historical context in order to be able to cite and interpret them, apocalypses are extremely effective and sensitive indicators of a people’s hopes, fears and frustrations.” “The hope was generally entertained that the Arabs would accomplish what the Persians had failed to accomplish—​the overthrow of Edom, entrenched in Rome and Byzantium. It was also fervently hoped that they would break the power of Persia, thereby delivering the Jews from the religious intolerance of the Sassanian dynasty” (Silver 1959, 37). Pirqe R. El. 28.

Of Siblings, Kingdoms, and the Days of the Messiah


be He, will hear (ishmaʿ) the voice of the people groaning about what the children of Ishmael will do in the land of Israel77 in the last days. Therefore, his name is called Ishmael, for it is said, [God] will hear, and will humble them (Ps 55:20).78 Instead of reading here a clear distinction between the scriptural Ishmael and his children whose rule is said to precede the end of time, it may be the case that the author of Pirqe de Rabbi Eliezer attempted to link them. The argument appears to be as follows: Israel’s chosenness is evident even in the etymology of the name of Abraham’s firstborn. This theophoric name foreshadows how God will confirm Israel’s election by humbling Ishmael’s descendants in the land of Israel, the Jewish eschatological setting par excellence.79 Pirqe de Rabbi Eliezer is at its most explicitly eschatological in the last part of its c­ hapter 30: Rabbi Ishmael said: In the future the children of Ishmael will do fifteen things in the Land in the latter days, and they are: 1) They will measure the land with ropes;80 2) they will change a cemetery into a resting-place for sheep, into a dunghill; 3) they will measure with them and from them upon the tops of the mountains; 4) falsehood will multiply 5) and truth will be hidden; 6) the law will be removed far from Israel; 7) sins will be multiplied in Israel; 8) worm-​crimson will be like [white] wool, 9) and it will wither paper and pen; 10) the coinage of the government will be deemed useless;81 11) and they will rebuild the desolated cities 12) and sweep the ways; 13) and they will plant gardens and parks, 14) and fence in the broken walls of the Temple;82 15) and they will build a building in the Holy Place. And two brothers will arise over them, princes at the end; and in their days the Branch, the Son of David, will arise, for it is said: And

77 78 79 80

81 82

Reading of the Warsaw edition and ms jts Enelow 866. Pirqe R. El. 32. Pirqe de Rabbi Eliezer expands here on tannaitic and amoraic sources. See MekhY Pischa 1, BerR 45:8, and yBer 1:6 (4a). Bakhos 2006, 98. This item can also be seen as part of the set of measures of “consolidation of large areas under Islamic rule” that fostered geographic mobility within the realm of Islam. In this context, Rustow 2013, 78 pointed out: “Fresh trade routes opened; the Umayyad and Abbasid caliphs built roads and other transport infrastructure in the interest of taxation, information gathering, and communication with provincial officials.” Other textual witnesses read: “he will hew down the rock of the kingdom.” ms huc 75 reads here “the house of study.”

232 Cordoni in the days of those kings the God of heaven will set up a kingdom that shall never be destroyed (Dan 2:44). Rabbi Ishmael also said: Three wars of trouble will the sons of Ishmael in the future wage on the earth in the latter days, for it is said, For they have fled from the swords[, and from the stress of war] (Isa 21:15). Swords signify only wars, one in the forest of Arabia, for it is said, from the drawn sword (Isa 21:15); another on the sea, for it is said, from the bent bow (Isa 21:15); and one in the great city, which is in Rome, which will be more grievous than the other two, for it is said, and from the stress of war (Isa 21:15). From there the Son of David shall flourish and see the destruction of these and these, and from there will He come to the land of Israel, for it is said, Who is this that comes from Edom, from Bozrah in garments stained crimson? Who is this so splendidly robed, marching in his great might? (Isa 63:1).83 This passage, which is preserved only in some of the textual witnesses of Pirqe de Rabbi Eliezer,84 is attributed to the second-​century Rabbi Ishmael.85 Rabbi Ishmael is therefore depicted as predicting the future in the Land in the time of Muslim rule. While the items in the list related to the moral decay in the Land are comprehensible on their own as undesirable phenomena, others such as the reference to the rebuilding of destroyed cities or the planting of gardens do not appear to be necessarily worthy of disapproval. Read on its own, the list of things brought about by the agency of the children of Ishmael does not reveal a consistently polemical attitude towards Muslim rule in the land of Israel.86 It is in the second part of the passage, which may be read as a depiction of the Arab wars of conquest87 and which links the time of the children

83 84 85 86


Pirqe R. El. 30. For this passage I also consulted Reeves 2006, 70–​73, who provided a version based on four sources followed by a Hebrew synopsis on pp. 74–​75. This state of preservation is apparently due to self-​censorship. Like Rabbi Aqiba, Rabbi Ishmael was active in the generation that went through the second Jewish war, after which a rabbinic discourse of attachment to the land of Israel appears to have started taking shape. See Gafni 1997. A number of scholars have attempted to understand the meaning of some of the items in this list, as well as recurring motifs in other apocalyptic narratives of this period, by gleaning possible references to the work’s context in the obscure and vague language that is intrinsic to the apocalyptic genre. See Silver 1959, 36–​49; Lewis 1950, 320–​38; Wasserstrom 1995, 27–​33; Hoyland 1997, 307–​21; Rubin 1999, 33–​34; and Reeves 2006 (including the wealth of literature in the notes to his translations). See Silver 1959, 41–​42 for a possible interpretation of the “three wars.”

Of Siblings, Kingdoms, and the Days of the Messiah


of Ishmael to that of the Messiah, where the polemical impetus against the former is on par with the traditional invective against Rome. Whereas in the version quoted above the Messiah witnesses the mutual destruction of these (Byzantine rule, Christianity) and those (Arab rule, Islam) while still in Rome (Constantinople), and the land of Israel is not explicitly depicted as an eschatological battlefield, an alternative reading with a different order of the clauses has the Messiah witness the end of Israel’s antagonists in the Land: “and he will come to the land of Israel and behold the destruction of both these and those.”88 Several of the motifs of the passage quoted above are part of a more detailed treatment of the Messianic drama in the Secrets of Rabbi Simeon,89 one of a group of apocalyptic texts attributed to this prominent sage of the post-​Bar Kokhba period90 (over the centuries, Jewish literature imagined Rabbi Simeon as hero of mysticism).91 The Secrets, a dialogue between Rabbi Simeon and the angel Metatron, opens with a midrash of Num 24:21 that identifies the “Kenite” 88 89 90


See Reeves 2006, 73. For the text of the Secrets, see Jellinek 1855, 78–​82; for a translation see Reeves 2006, 78–​89. On Rabbi Simeon in this context see Gafni 1997, 64–​69 and more generally Rosenfeld 1999. Two other apocalyptic texts attributed to Rabbi Simeon are the Atidot (or “future things”) of Rabbi Simeon ben Yochai (a variant recension of the Secrets preserved as part of the larger Midrash of the Ten Kings) and the Prayer of Rabbi Simeon ben Yochai, where the historical account continues up to the time of the Crusades, when the final redaction of the texts appears to have taken place. The Prayer preserves a segment that describes in prophetic manner the Fatimid invasion of Egypt and Palestine and the brief Carmathian rule in Palestine and Syria (based on a tenth-​century apocalypse), as well as another segment with an account of the capture of Jerusalem by the Crusaders in 1099 ce. For the text of the Atidot, see Horowitz 1891, 51–​55; for that of the Prayer, see Jellinek 1857, 117–​26; and an English translation in Reeves 2006, 90–​105. In the mid-​twentieth century, Even-​Shmuel published an anthology of Jewish apocalyptic texts that includes the Secrets, Atidot, and Prayer; it is now the standard source book in its third edition. See Even-​Shmuel 2017. The Zohar, the major medieval work of Jewish mysticism, is attributed to Rabbi Simeon. Only two of the fifteen actions attributed to the children of Ishmael in Pirqe de Rabbi Eliezer are also quoted in the name of Rabbi Ishmael in the Secrets, namely, the measurement of the land with ropes and transformation of cemeteries into pastureland. Whereas the first may be an unwelcome action from a Jewish perspective, the second directly affects the ritual purity of the Land. The Secrets also incorporates material of earlier sources such as Sefer Zerubbabel, an apocalypse composed in all probability in Palestine before the Arab conquest. Himmelfarb 2017 argued that although Sefer Zerubbabel mentions “the kings of Qedar and the children of Qedem” as companions of the Jewish eschatological antagonist Armilos and although Qedar and Qedem allude elsewhere to Arab tribes so that their mention could be understood as an allusion to Islam, the text appears to have been composed before, albeit in all probability immediately before, the rise of Islam. See also the liturgical poem (piyyut) Oto ha-​yom (“That day”).

234 Cordoni with the “kingdom of Ishmael.”92 Metatron explains to Rabbi Simeon that he need not fear the Kenite because they act according to God’s will: they are his instruments as it were. God will conquer for them the Land—​Metatron goes on to explain—​and they will restore its splendor. Furthermore, and as anticipated by Isaiah, the kingdom of Ishmael does not precede the arrival of the Messiah but is contemporary with him and therefore comparable to him in their contribution to Israel’s salvation.93 In connection with the rule of a number of unnamed Ishmaelite kings, whose treatment appears to confirm the Palestinian perspective of the text,94 the Secrets’ prophecy includes deeds such as the repairing of the Temple95 and reshaping of Mount Moriah,96 the planting of saplings and the construction of irrigation systems,97 or the construction of a Jordan canal. It is further claimed that during the reign of yet another king the Land will be in peace. Statements pertaining to the rule of a king identified as Marwān have been interpreted as referring to the end of Umayyad rule, expressed in terms of the end of the kingdom of Ishmael: Another king will then arise—​a strong (king) and warrior. A dispute will erupt in the world during his reign. This will be a sign for you: When you see that the western gate has fallen—​(the one) at the western side of (the place of) prayer of the children of Ishmael in Damascus—​his dominion will have fallen. They will be assembled and marched out to do forced labor, and indeed the kingdom of Ishmael will collapse. Scripture affirms concerning them: The Lord has broken the rod of the wicked [the sceptre of the rulers (moshlim98)] (Isa 14:5), where rod means Ishmael.99 92 93 94

95 96 97 98 99

The later Prayer dissolves this identification, so that the Kenite stand first for a rule that precedes Muslim rule but then also for the Crusaders. It has been argued that this assessment reflects familiarity with a Muslim interpretation of Isa 21:7 as prophesizing the rise of Muhammad. See Reeves 2006, 7–​12 and Bashear 1991. The list of kings is merely a selection of rulers, probably representative ones in the eyes of our author, rather than, as Silver 1959, 44 put it, the “whole history of Islam from the rise of Muhammad through the line of Umayyad caliphs to the last of the dynasty.” Lewis 1950, 328 pointed out that the specific Palestinian perspective is manifest in a conflation of the patriarchal caliphs and Muʿāwiya. Prophesied of the “second king who will arise from Ishmael [who] will love Israel.” Prophesied of a king whose rule is described with recourse to an interpretation of Num 24:21: The king will “build for himself a place of prayer (lit. ‘prostration’) upon the site of the foundation stone, for it is said, and set your nest on the rock (Num 24:21).” This item is attributed to the sixth monarch in the list, a “great king,” whose physiognomy is described in detail: “reddish-​hued and cross-​eyed, with three moles, one on his forehead, one on his right hand, and one on his left arm.” The midrash may be playing on the possible reading of moshlim as “Muslim.” See Reeves 2006, 84n52. Text quoted following Reeves 2006.

Of Siblings, Kingdoms, and the Days of the Messiah


Interestingly, the quoted passage pertaining to the last king in a list that alludes to individual Umayyad rulers, and in one case to a group of four, associates the name Ishmael both with a faith, which has its own place of worship in Damascus, and with a rule that, unlike the previous statements concerning the improvement of the Land, is now perceived as wicked and justly punished.100 The fall of the Umayyad dynasty is alluded to with the motif of a gate falling, already known in the Babylonian Talmud.101 As Metatron explains to Simeon, the Messiah is expected to make his appearance and confront not a king of Ishmael, that is, an Umayyad king, nor for that matter those who vanquished the Umayyads,102 but rather the return of the wicked kingdom Edom and its leader Armilos.103 The Secrets leaves its relatively realistic account of the history of Muslim rule in the Land when it returns to the traditional Jewish Messianic narrative and “restores” Roman rule. This narrative had been transmitted over the centuries in Palestinian and Babylonian sources and it assumed that the next to last kingdom in the land of Israel is Edom (Rome) and the last that of the Messiah Son of David.104 For the apocalyptic imagination the Land and 100 The list of rulers in the Secrets is preceded by the mention of the measurement of the land with ropes and the defilement of the land as a consequence of turning cemeteries into pastureland, and a second midrash on Num 24:21. The gentile prophet Balaam, who in the original scriptural context predicts Israel’s conquest of the Edomites, Amalekites, and Kenites or Midianites, is here depicted as having rejoiced upon foreseeing the rise of the Kenite (i.e., Ishmael’s kingdom). However, the midrash appears to read his words for the Kenite—​“strong (etan) is your dwelling place” (Num 24:21)—​as conceding that Ishmael and Israel have in common a set of commandments, e.g., pertaining to eating customs, that go back to Abraham (Ethan being a name that rabbinic literature understands as referring to Abraham; see Lewis 1950, 24). As Reeves 2006, 81n33 put it, “this peculiar exegesis reflects a Jewish accommodation to [Muslim] rhetorical claims.” 101 The early second-​century Palestinian Rabbi Jose ben Qisma refers there to the fall of “this gate” as a sign for the imminent coming of the Messiah, whereby he may be understood as intending the gate of his hometown Caesarea Philippi whose fall would mean the end of Roman rule in the Land. See bSan 98a. The chronologically closer TanB Vayislach 8 sets the dialogue in Tiberias. 102 Therefore implying that the ʿAbbāsids were not perceived as perpetuating the kingdom of Ishmael. 103 On this figure, an equivalent to the Christian anti-​Christ and the Muslim Dajjāl, see Himmelfarb 2017, ch. 1. 104 Thus, the last chapter of tractate Sanhedrin in the Babylonian Talmud, concerned with the Messianic era as part of its elaboration on the theme of the world to come, and the late Hebrew apocalypse of pre-​Islamic times Sefer Zerubbabel. Even in the eleventh century in a responsum on redemption the Babylonian Hai Gaon would state: “Therefore when we see that Edom has attained ruling authority over the land of Israel, we can affirm that our redemption is beginning, for Scripture declares, deliverers will ascend Mount Zion

236 Cordoni more specifically Jerusalem as center of the eschatological events needed to be cleansed of outsiders, of the “uncircumcised foreigners and the impure,”105 Christians and Muslims alike, for a heavenly Jerusalem to descend and inaugurate a two-​thousand-​year period before the final judgement in the Valley of Jehoshaphat. The text appears to attest to a contradictory attitude to Muslim rule, which is either inherent to the Jewish perception of early Islam or evidence of the text’s composite character due to the use of sources of different periods.106 While those dating from the mid-​seventh century would reflect a relatively positive perception of the new order, those of the mid-​eighth century depict this order as deserving to be defeated.107 It is not with a king of Ishmael but with a king of the Arabs that our last example of a Jewish apocalyptic source is concerned. As part of the depiction of the events of the time of the Messiah, Pirqe Mashiach (The Chapters of the Messiah)108 transmits the following debate between the collective Israel and the king of the Arabs: Israel will say to the king of the Arabs, “The Temple is ours. Take the silver and gold but leave the Temple.” The king of the Arabs will reply, “There is nothing for you in this Temple. However, if you first choose for yourselves a sacrifice as you used to do in former days, we too will offer a sacrifice and [in accordance with] whoever’s sacrifice is accepted, we will become in order to judge the mountain of Esau, and sovereignty will belong to the Lord (Obad 1:21).” Quoted following Reeves 2006, 134–​35. 105 Secrets. 106 The same appears to be valid for the later Prayer, which, even though more consistent in its “invective against Islam” (Reeves 2006, 77), still depicts the sons of Ishmael and the kingdom of Ishmael as the chronological setting for the rise of the kingdom of the Messiah. Here we read: “And when he sees riders, horsemen in pairs, riders on asses, riders on camels: riders is the kingdom of Media and Persia, pairs is the kingdom of Greece, horsemen is the kingdom of Edom, riders on asses is the Messiah … , riders on camels is the kingdom of Ishmael, in whose days the kingdom of the Messiah will sprout.” This text, an atomizing exegesis of Isa 21:7, has a parallel in the Secrets. There, however, no variation of the four-​kingdom doctrine is used to explain the verse as referring to the fact that the Messiah and Muslim rule happen at the same time. This appears to be a Judaizing reading of an originally Muslim understanding of Isa 21:7. 107 By contrast, when each of the texts of the Rabbi Simeon ben Yochai’s complex is considered on its own as the coherent product of a different period, the distinction made by Lewis 1950, 323 is of interest: “While the text of the Secrets expresses a Messianic hope from these events [i.e., the Arab conquests] the others are subsequent and probably independent reflections of disillusionment.” 108 On Pirqe Mashiach see Spurling 2017 and Reeves 2006, 149. Text: Jellinek 1855, 68–​78.

Of Siblings, Kingdoms, and the Days of the Messiah


a people (ummah).” Israel will offer a sacrifice, but it will not be accepted because Satan will denounce them before the Holy One, blessed be He. The sons of Qedar will offer sacrifices and they will be accepted, for it is said, All the flocks of Qedar shall be gathered to you[, … they shall be acceptable on my altar] (Isa 60:7). At that time, the Arabs will say to Israel, “Come and believe in our faith.” But Israel will reply, “Either we kill or are killed, but we will not commit apostasy.”109 The collective Israel are depicted as willing to put their right to exclusive possession of the Temple (and Jerusalem) to a test that the king of the Arabs proposes. Once their sacrifice fails to be accepted but that of the Arabs is accepted, they are not willing to continue with the proposed deal, to become one nation and share the same beliefs as the Arabs. The text acknowledges that whereas the Temple Mount was of paramount significance in a time of Temple-​centered Judaism, in the early Muslim period Jews identify with their principles more than with the place where the Temple once stood; one principle is to stay a people other than the Muslim umma and its faith. Some of the texts discussed here and in the previous sections, assumed to have emerged in the land of Israel, are quite explicit with respect to what they depict as a future, transitional Muslim rule. Less troubled with (and hostile about) the changes brought about by this foreign rule in the Land are the sources generally identified as having originated in or being the product of a stage of redaction in the diaspora, such as Seder Eliyahu, Aggadat Bereshit, and to a certain extent texts of the Tanchuma-​Yelammedenu literature. In this connection we may discuss a last example cited by Robert Hoyland as evidence of a Jewish apocalyptic response to Islam. Following Bernhard Bamberger (1940), Hoyland suggested reading a passage in the rabbinic midrash Pesiqta Rabbati, a work scholars consider as representing in part Tanchuma-​Yelammedenu literature, as a Jewish apocalyptic response to Islam.110 In this passage in PesR 36, which is part of a larger group of homilies of messianic theme, mention is made of a war between the kings of Persia and Arabia in the year of the Messiah. On the basis of this mention, Bamberger had dated the entire group of homilies to the five-​year period between 632 and 637 ce when “Persia and Arabia were both world powers.” The same homilies had previously been dated by Jacob Mann to the ninth century because of their mention of the Mourners of Zion,111 understood as the earliest reference to a Karaite movement that flourished in the 1 09 Translated following Jellinek 1855, 71. 110 Hoyland 1997, 312–​13. 111 Mann 1970, 47ff.

238 Cordoni late ninth and tenth centuries in Jerusalem. Arnold Goldberg dated the four homilies to the fourth century and rejected any link to the Karaites in them.112 Philip Alexander mentions these homilies as evidence for a messianic oriented “shadowy group known as the Mourners for Zion” of the sixth/​seventh century whose relationship to the Karaites of Jerusalem of the ninth century is disputed.113 As is the case with other documents of Tanchuma-​Yelammedenu literature, dating material in Pesiqta Rabbati is no easy task. While the core of Pesiqta Rabbati, including those portions which are considered to represent a middle developmental stratum of Tanchuma-​Yelammedenu literature,114 appear to precede the advent of Islam and may have emerged in Byzantine Palestine, its final redaction may have taken place in Europe during the twelfth or thirteenth century.115 This type of geographical and cultural setting may help explain the apparent general lack of interest in Islam or Muslim rule in this literature. While most of the sources discussed in section 2 in the context of an adapted doctrine of world kingdoms situated the children of Ishmael at the end of a list of rulers, the texts examined in this section narrow the focus and address more in detail the time of Muslim rule and set it in relation to the events of the end time. While Muslim rule in an eschatological context is criticized in Pirqe R. El. 32, both Pirqe R. El. 30 and the Secrets of Rabbi Simeon depict it in a neutral if not positive light (at least as far as the Land itself is concerned) and as a meaningful tool in the divine plan in connection with Israel’s salvation in the days of the Messiah. The kingdom of Ishmael is not depicted as Israel’s antagonist, in any case not in the same way as Rome. 4


It is telling that most of the anonymous rabbinic and quasi-​rabbinic literary sources discussed in this article are either assumed to be of Palestinian origin or to relate in some way to Palestinian rather than to Babylonian tradition. It appears to have made more sense for the authors of these texts to argue with the aid of Palestinian authorities rather than with Babylonian ones: if rabbinic authorities are mentioned, they are primarily well-​known Palestinian sages 1 12 Goldberg 1978a, 20; Goldberg 1978b, 142–​43. 113 Alexander 2007, 231–​32. 114 Several developmental strata have been identified in the works of this genre. See Bregman 2003, 4–​5; Weiss 2017, 12. 115 See Ulmer 2017, 28–​30.

Of Siblings, Kingdoms, and the Days of the Messiah


of earlier times.116 None of the sources presents itself as emerging in the first Muslim period but as foreseeing the advent of Islam. In several cases these sources make use of earlier Jewish material and update it to fit the changed horizon of thought and expectations that the new order implied. Because these sources sometimes also appropriate motifs from Muslim tradition, some of the changes that we may ascertain with respect to their probable Jewish sources were possibly due to the new cultural landscape of the Islamicate Near East.117 The sources were selected and discussed according to three more or less distinct conceptual areas: 1) the scriptural Ishmael and his afterlife; 2) Ishmael’s descendants as one, generally the last, among several world powers; and 3) the kingdom of Ishmael or the kingdom of the Arabs as important actors in the Messianic drama. They do not attest to only one coherent attitude towards Islam and Muslim rule. The problem posed by the Ishmael story in Scripture and later by its expansion in Muslim tradition gave the midrashic authors the chance to respond to Islam with a series of counter-​narrative steps that argue with the essential dissimilarity of Abraham’s children by Hagar and Sarah. These hermeneutic shifts highlight not only the superiority of the second-​born Isaac and his offspring, but also the fact that Abraham’s firstborn belongs somewhere other than in the land of Israel. Whether in the desert of Paran or in the land of Hagar’s ancestors, the texts in Pirqe de Rabbi Eliezer and Aggadat Bereshit expel Ishmael anew from Canaan. In texts composed in the Islamic period this kind of rewriting of the expulsion is hardly insignificant. In the context of the discussion of the world kingdoms under whose rule Israel lives in the land of Israel and in the diaspora, the reflection on Ishmael is not linked to the scriptural narrative. In this connection it is as a last link in a chain of world rulers rather than as Israel’s opponents that Ishmael’s descendants, the children of Ishmael, are viewed. While on the one hand they participate in the worldly character of the kingdoms they succeed, that is, mainly Persia and Rome, on the other hand, their rule is understood as indicating the 116 Tellingly, these sages tend to be associated with the time of the Bar Kokhba revolt or with the period following the defeat. 117 One of the possible explanations for the new hermeneutic, more narrative style of works such as Pirqe de Rabbi Eliezer may be the exposure to the Muslim literary genre system. On the transitional character of Pirqe de Rabbi Eliezer, see Rubenstein 1996; more recently, focusing on the literary forms that make up the work, see Keim 2017. The Muslim panegyric genre known as faḍāʾil (“merits”) appears to be incorporated into Jewish apocalypses. See for example the praise of Jerusalem and the Temple in Pirqe Mashiach. On this genre in Muslim texts concerned with Jerusalem see Antrim 2012, 48–​55.

240 Cordoni imminence of a fundamentally new order, the Messianic era; thus, they are also themselves conceived of as a somewhat different order. They are part of the list of world kingdoms but seldom referred to as kingdom.


Alexander, Philip. 2007. “The Rabbis and Messianism.” In Redemption and Resistance: The Messianic Hopes of Jews and Christians in Antiquity, edited by Markus Bockmuehl, 227–​44. London: T&T Clark. Antrim, Zayde. 2012. Routes & Realms: The Power of Place in the Early Islamic World. Oxford: Oxford University Press. Bakhos, Carol. 2006. Ishmael on the Border: Rabbinic Portrayals of the First Arab. Albany: State University of New York Press. Bakhos, Carol. 2007. “Abraham Visits Ishmael: A Revisit.” Journal for the Study of Judaism 38:553–​80. Bamberger, Bernhard. 1940. “A Messianic Document of the Seventh Century.” Hebrew Union College Annual 15:425–​31. Bashear, Suliman. 1991. “Riding Beasts on Divine Missions: An Examination of the Ass and Camel Traditions.” Journal of Semitic Studies 37:37–​75. Börner-​Klein, Dagmar, ed. 2004. Pirke de-​Rabbi Elieser: Nach der Edition Venedig 1544 unter Berücksichtigung der Edition Warschau 1852. Berlin: De Gruyter. Bregman, Marc. 2003. Tanchuma-​Yelammedenu: Studies in the Evolution of the Versions. Piscataway: Gorgias Press (Hebrew). Brody, Robert. 2013. Saʿadyah Gaon. Oxford: The Littman Library of Jewish Civilization. Carlson, Thomas A. 2015. “Contours of Conversion: The Geography of Islamization in Syria, 600–​1500.” Journal of the American Oriental Society 135.4:791–​816. Collins, John J. 1998. The Apocalyptic Imagination: An Introduction to Jewish Apocalyptic Literature, 2nd ed. Grand Rapids: Eerdmans. Collins, John J., and Peter W. Flint, eds. 2001. The Book of Daniel: Composition and Reception. Leiden: Brill. Cordoni, Constanza. 2018. “Inheriting and Buying a Homeland: The Land of Israel and the Patriarchs.” Journal for the Study of Judaism 49:551–​80. Cordoni, Constanza. 2020 “ ‘For they did not change their language’ (MekhY Pischa 5): On Early Medieval Literary Rehebraicisation of Jewish Culture.” Medieval Worlds 11:165–​86. http://​​10.1553/​medievalworlds_​no11_​2020s165. Elbaum, Jacob. 1986. “Between Redaction and Rewriting: On the Character of the Late Midrashic Literature.” In Proceedings of the Ninth World Congress of Jewish Studies: Division B: The History of the Jewish People (From the Second Temple Period until the Middle Ages), vol. 1, 57–​62. Jerusalem: Magnes Press (Hebrew).

Of Siblings, Kingdoms, and the Days of the Messiah


Enelow, H. G., ed. 1933. The Mishnah of Rabbi Eliezer or the Midrash of Thirty-​Two Hermeneutic Rules. New York: Bloch Publishing Company. Even-​Shmuel, Yehuda. 2017. Midrashei Geula: Chapters of Jewish Apocalypse Dating from the Completion of the Babylonian Talmud until the Sixth Millenium, 3rd ed., with an introduction by Oded Irshai and Hillel Newman. Jerusalem: Bialik Institute. Firestone, Reuven. 1990. Journeys in Holy Lands: The Evolution of the Abraham-​Israel Legends in Islamic Exegesis. Albany: State University of New York Press. Firestone, Reuven. 2016. “Muslim-​Jewish Relations.” In Oxford Research Encyclopedia of Religion. Oxford: Oxford University Press. https://​doi-​​10.1093/​ acrefore/​9780199340378.013.17. Frenkel, Yehoshua. 2013. “Nota bene: Jerusalem.” In A History of Jewish-​Muslim Relations, edited by Abdelwahab Meddeb and Benjamin Stora, 108–​10. Princeton: Princeton University Press. Friedlander, Gerald, ed. 1916. Pirḳê de Rabbi Eliezer. London and New York: Kegan Paul /​ The Bloch. Gafni, Isaiah. 1997. Land, Center and Diaspora: Jewish Constructs in Late Antiquity. Sheffield: Sheffield Academic Press. Geula, Amos. 2006. “Lost Aggadic Works Known Only from Ashkenaz.” PhD diss., Jerusalem: Hebrew University (Hebrew). Gil, Moshe. 1992. A History of Palestine, 634–​1099. Cambridge: Cambridge University Press. Goldberg, Arnold. 1978a. Ich komme und wohne in deiner Mitte: Eine rabbinische Homilie zu Sacharja 2,14 (Pesiqta Rabbati 35). Frankfurt am Main: Gesellschaft zur Förderung judaistischer Studien. Goldberg, Arnold. 1978b. Erlösung durch Leiden: Drei rabbinische Homilien über die Trauernden Zions und den leidenden Messias Efraim (Pesiqta Rabbati 34.36.37). Frankfurt am Main: Gesellschaft zur Förderung judaistischer Studien. Heinemann, Joseph. 1974. The Aggadah and its Development. Jerusalem: Keter (Hebrew). Heller, Bernard. 1925. “Muhammedanisches und antimuhammedanisches in den Pirke Rabbi Eliezer.” Monatsschrift für Geschichte und Wissenschaft des Judentums 33:47–​54. Himmelfarb, Martha. 2017. Jewish Messiahs in a Christian Empire: A History of the Book of Zerubbabel. Cambridge: Harvard University Press. Hirschberg, Haïm Zʾew. 2007. “Israel, land of: History.” In Encyclopaedia Judaica, 2nd ed, vol. 10, edited by Michael Berenbaum and Fred Skolnik, 143–​79. Detroit: Macmillan. Hoyland, Robert. 1997. Seeing Islam as Others Saw It: A Survey and Evaluation of Christian, Jewish, and Zoroastrian Writings on Early Islam. Princeton: Darwin Press. Horowitz, Chaim M., ed. 1891. Bet Eqed ha-​Aggadot. Frankfurt am Main: Slobotsky (Hebrew).

242 Cordoni Jellinek, Adolph, ed. 1855. Bet ha-​Midrash: Sammlung kleiner Midraschim und vermischter Abhandlungen aus der ältern jüdischen Literatur, vol. 3. Leipzig: Vollrath. Jellinek, Adolph, ed. 1857. Bet ha-​Midrash: Sammlung kleiner Midraschim und vermischter Abhandlungen aus der ältern jüdischen Literatur, vol. 4. Leipzig: Vollrath. Kalahani, Ezra. 2003. Aggadat Bereshit: Introduction, Proposal for a Critical Edition and Discussion of Its Content and Structure. PhD diss., Hebrew University, Jerusalem (Hebrew). Keim, Katharina. E. 2017. Pirqei deRabbi Eliezer: Structure, Coherence, Intertextuality. Leiden: Brill. Küchler, Max. 2013. Jerusalem: Ein Handbuch und Studienreiseführer zur heiligen Stadt. Orte und Landschaften der Bibel: Ein Handbuch und Studienreiseführer zum Heiligen Land 4:2. Göttingen: Vandenhoeck and Ruprecht. Lamdan, Ruth, Yoram Erder, and Joseph Drory. 2010. “Palestine.” In Encyclopedia of Jews in the Islamic World, edited by Norman A. Stillman. Brill Online. Levy-​Rubin, Milka. 1998. “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, edited by Arieh Kofsky and Guy G. Stroumsa, 149–​62. Jerusalem: Yad Izhak Ben-​Zvi. Levy-​Rubin, Milka. 2011. Non-​Muslims in the Early Islamic Empire: From Surrender to Coexistence. Cambridge: Cambridge University Press. Lewis, Bernhard. 1950. “An Apocalyptic Vision of Islamic History.” Bulletin of the School of Oriental and African Studies 13.2:308–​38. Magdalino, Paul. 2014. “ ‘All Israel will be Saved’? The Forced Baptism of the Jews and Imperial Eschatology.” In Jews in Early Christian Law: Byzantium and the Latin West, 6th–​11th centuries, edited by John V. Tolan, Nicholas de Lange, Laurence Foschia, and Capucine Nemo-​Pekelman, 231–​42. Turnhout: Brepols. Mann, Jacob. 1970. The Jews in Egypt and in Palestine under the Fatimid Caliphs, vol. 1. New York: Ktav. Mikva, Rachel S., ed. 2012. Midrash vaYosha: A Medieval Midrash on the Song at the Sea. Tübingen: Mohr Siebeck. Reeves, John C. 2006. Trajectories in Near Eastern Apocalyptic: A Postrabbinic Jewish Apocalypse Reader. Leiden: Brill. Rosenfeld, Ben-​Zion. 1999. “R. Simeon b. Yohai: Wonder Worker and Magician Scholar, Saddiq and Hasid.” Révue des Études Juives 158.3–​4:349–​84. Rubenstein, Jeffrey. 1996. “From Mythic Motifs to Sustained Myth: The Revision of Rabbinic Traditions in Medieval Midrashim.” Harvard Theological Review 89:131–​59. Rubin, Uri. 1999. Between Bible and Qurʾān: The Children of Israel and the Islamic Self-​ Image. Princeton: Darwin Press.

Of Siblings, Kingdoms, and the Days of the Messiah


Rustow, Marina. 2013. “Jews and Muslims in the Eastern Islamic World.” In A History of Jewish-​Muslim Relations, edited by Abdelwahab Meddeb and Benjamin Stora, 75–​ 98. Princeton: Princeton University Press. Schussman, Aviva. 1980. “Abraham’s Visits to Ishmael: The Jewish Origin and Orientation.” Tarbiz 49:325–​45 (Hebrew). Schwarzbaum, Haim. 1975. From the Beginning Israel and Ishmael: Judaism and Islam as Mirrored in Folklore. Tel Aviv: Don Publishing (Hebrew). Silver, Abba Hillel. 1959. A History of Messianic Speculation in Israel. Boston: Beacon. Spurling, Helen. 2017. “A Revival in Jewish Apocalyptic? Change and Continuity in the Seventh–​Eighth Centuries with Special Reference to Pirqe Mashiah.” In Apocalypticism and Eschatology in Late Antiquity: Encounters in the Abrahamic Religions, 6th–​8th Centuries, edited by Hagit Amirav, Emmanouela Grypeou, and Guy G. Stroumsa, 163–​86. Leuven: Peeters. Steinschneider, Moritz. 1874. “Apokalypsen mit polemischer Tendenz.” Zeitschrift der Deutschen Morgenländischen Gesellschaft 28:647–​59. Steinschneider, Moritz. 1875. “Apokalypsen mit polemischer Tendenz.” Zeitschrift der Deutschen Morgenländischen Gesellschaft 29:163–​66. Stemberger, Günter. 2010a. “Hebräisch als ideale Sprache—​Konsequenzen für die Hermeneutik.” In Judaica Minora I: Biblische Traditionen im rabbinischen Judentum, 88–​102. Tübingen: Mohr Siebeck. Stemberger, Günter. 2010b. “Die jüdische Danielrezeption seit der Zerstörung des zweiten Tempels am Beispiel der Endzeitberechnung.” In Judaica Minora I: Biblische Traditionen im rabbinischen Judentum, 203–​20. Tübingen: Mohr Siebeck. Stemberger, Günter. 2011. Einleitung in Talmud und Midrasch, 9th ed. Munich: C.H. Beck. Swain, Joseph Ward. 1940. “The Theory of the Four Monarchies: Opposition History under the Roman Empire.” Classical Philology 35:1–​21. Teugels, Lieve M. 2001. Aggadat Bereshit: Translated from the Hebrew with an Introduction and Notes. Leiden: Brill. Teugels, Lieve M. Forthcoming. “The Provenance of Aggadat Bereshit: A Reassessment of the Origins of the Work as a ‘Tanchuma Satellite.’ ” In Tanchuma: The State of the Research, edited by Ronit Nikolsky. Leiden: Brill. Thomas, David, and Alex Mallett, eds. 2010. Christian-​Muslim Relations: A Biblio­ graphical History, vol. 2: 900–​1050. Leiden: Brill. Thomas, David, and Barbara Roggema, eds. 2009. Christian-​Muslim Relations: A Bibliographical History, volume 1: 600–​900. Leiden: Brill. Ulmer, Rivka, ed. 2017. A Bilingual Edition Pesiqta Rabbati, vol 1: c­ hapters 1–​22. Berlin: De Gruyter. Wasserstrom, Steven M. 1995. Between Muslim and Jew: The Problem of Symbiosis Under Early Islam. Princeton: Princeton University Press.

244 Cordoni Weiss, Dov. 2017. Pious Irreverence: Confronting God in Rabbinic Judaism. Philadelphia: University of Pennsylvania Press. Wendehorst, Stephan. 2015. “Four Kingdoms.” In Encyclopedia of Early Modern History Online, edited by Graeme Dunphy and Andrew Gow. Stuttgart: Metzler. Williams, Raymond. 1961. The Long Revolution. New York: Columbia University Press.

­c hapter 10

New Light on the Dark Ages: A Byzantine Perspective on the Arab Expansion Joanita Vroom 1


A period of flux characterizes seventh-​century ce Byzantine history. Many things were moving—​literally. First, the earth itself moved as a result of a series of destructive earthquakes. Second, rats moved, spreading a horrible plague epidemic. Third, foreign invaders moved, especially the Arabs, who managed to conquer large parts of the Byzantine Empire in the Near East and North Africa. This period from 600 to 950 ce is called the “Dark Ages” by historians, not only for these reasons, but also mainly because of the limited literary sources regarding daily life.1 Recent archaeological discoveries in the Byzantine world, especially in the Aegean region, allow new possibilities for forming a more nuanced image of daily life in this period: it is less dark than has often been assumed.2 What were the “Dark Ages” in the Byzantine world, exactly?3 How did this period manifest itself in the archaeological remains left behind? Is it right to speak of a “Dark Age”, or rather of a transformation period? In light of the scope of this volume, I offer a study of the impact of the Islamic expansion in the Byzantine world from an archaeological perspective. I begin with a general introduction to the “Dark Ages” in the Byzantine period, with an emphasis on the seventh and eighth centuries ce, followed by two archaeological case studies in the eastern Mediterranean. These case studies derive from recently excavated finds in two coastal cities: Athens, in central Greece, and Butrint, in southern Albania (Figure 10.1).

1 See for example Ostrogorsky 1959; idem 1969; Foss 1975; idem 1977; Aupert 1989. 2 Curta 2005; idem 2006; Ivison 2007; Vroom 2003, 49–​58, 137–​45; idem 2011; idem 2016; idem 2017. 3 For a discussion on the use of the term “Dark Ages,” see Vroom and Kondyli 2015, 329n1.

© Joanita Vroom, 2022 | DOI:10.1163/9789004500648_011

246 Vroom


Ravenna ITALY Rome




Corfu Syracusa NORTH AFRICA


Butrint Athens




Extent of the empire under Justinian I, 527-565 The empire in about 1020


The empire in about 1360

­f igure 10.1  Map of Byzantine Empire, with Athens and Butrint. Red line: size of the Empire during the reign of Justinian i (527–​65); region in orange: size ca. 1020; region in red: size ca. 1360 j. porck


Empires in Flux

My research focuses on excavated material culture, mainly ceramics, in the eastern Mediterranean and the Near East, from the Byzantine through the Ottoman Empire; that is, the period between the seventh and the twentieth century ce.4 The region of my research so far encompasses Greece, Albania, Turkey, Cyprus, Crimea, and Jordan, corresponding to the main areas of the Byzantine Empire and the region where East and West have met time and again—​in many different circumstances and with varying degrees of peace.5 Figure 10.1 shows the size of the Byzantine Empire in the sixth century ce, during the reign of the emperor known as Justinian i, or Justinian the Great.6 Justinian was emperor of Byzantium between 527 and 565 ce; the historian Procopius of Caeserea (ca. 500–​62 ce) is our main source on Justinian’s life and works.7 Justinian reigned primarily from Constantinople (modern Istanbul), the capital of the Byzantine Empire and the city where he commissioned the construction of the famous cathedral of Hagia Sophia. Thanks to a series of successful campaigns, and with the help of General Flavius Belisarius 4 5 6 7

Vroom 2003, 25–​26, table 1.1. Vroom 2015. Figure 10.1 shows the size of the Byzantine Empire between ca. 1020 and ca. 1360 ce. Procopii Caesariensis Anekdota.

New Light on the Dark Ages


(505–​665 ce), Justinian reconquered the western Mediterranean, and therefore reunified the former Roman Empire.8 However, he could not maintain this empire, and his reign is generally considered the end of the classical era.9 The seventh century ce included political uncertainty resulting from the fast emergence of another superpower in the Mediterranean: the Islamic caliphate. Muhammad and the first three caliphs (622–​56 ce) gradually conquered the Arabian Peninsula, Mesopotamia, Persia, Palestine, Syria, and Egypt.10 During the dynasty of the Umayyads, this area expanded, conquering for instance in its entirety the Sasanian, or Neo-​Persian, Empire. Some Byzantine provinces (i.e., North Africa and Spain) were overrun. Both empires (Byzantine and Sasanian) were heavily weakened by a series of all-​destroying wars, and the encroaching Arabs eagerly exploited this power vacuum. After the Umayyads had conquered large parts of the Middle East, they named Damascus, a former Byzantine governor’s seat, the capital of the new Arab empire.11 The Arab conquests around 700 ce therefore extensively weakened the political and military strength of the Byzantine Empire in the Mediterranean. Historians point to three different symptoms of Byzantium’s decline in this period.12 First, the monumental city centers fell into ruin. Second, the populations in the cities declined, and large parts of these cities consequently became deserted. Finally, long-​distance trade came to an end. The Byzantine emperors responded to these pervasive Arab invasions of their territory by militarizing their population and, from the mid-​seventh century ce onward, by reorganizing the Byzantine army into themata.13 Greek fire (also known as Byzantine fire or liquid fire) was another tool used in the Byzantine Empire to ward off the Arab invasions (Figure 10.2).14 It secured many military victories of the Byzantines and, moreover, is one of the reasons that the empire survived during the “Dark Ages.” This short historical overview of the Byzantine Empire during the “Dark Ages” highlights some important aspects concerning daily life in the Byzantine Empire. Although very little is currently known about this subject, excavations and an analysis of material culture provide useful information. Recent archaeological finds in Athens and Butrint, two important places in the Byzantine

8 9 10 11 12 13 14

Barker 1966, 75; Treadgold 1997, 246. Baker 2002; Decker 2016. Haldon 2005, map 2.8; Kennedy 2007. Kennedy 2007; Haldon and Kennedy 1980. Liebeschuetz et al. 2001. Haldon 2005, 68–​88; Burbank and Cooper 2010, 67. Ellis Davidson 1973; Pászthory 1986; Haldon 2006.

248 Vroom

­f igure 10.2  Miniature with the use of Greek fire, invented during the Byzantine-​Arab wars. Codex Skylitzes Matritensis, 12th century, Biblioteca Nacional de España, Madrid, Vitr. 26-​2, Bild-​Nr. 77, f 24v.b From: Pászthory 1986, 31

Empire during the rise of Islam, allow us to reassess the darkness of those “Dark Ages” in the Byzantine Empire and whether the supposed decline affected its entire population. 3

First Case Study: Athens

Most people know Athens from its classical temples on the Acropolis, but perhaps not all modern tourists realize that the city also had a rich history in the Byzantine, medieval, and Ottoman periods.15 Over the last five years, a small research team from the University of Leiden has studied the archaeological finds excavated at the Athenian Agora by the American School of Classical Studies at Athens (Figure 10.3).16 These finds date from the Byzantine to the Ottoman period.17 A map of Athens shows that the size of the walled area of the late antique city had considerably decreased from earlier phases 15 16 17

These periods are often disregarded in modern historical accounts, cf. Nikolakis 2015; Poulios and Senteri 2015. The current excavations of the Athenian Agora have been directed since 1994 by John McKesson Camp ii and his team. For other publications on the Agora, see Camp 1986; idem 2001; idem 2007; Camp and Mauzy 2009. Vroom 2013.

New Light on the Dark Ages


­f igure 10.3  Photograph of the antique Agora in Athens ascsa

(Figure 10.4).18 Until recently, it was believed that this smaller walled area reflected a reduction in population size. The only inhabited area lay within the late antique fortification walls,19 which had been built after the so-​called “Herulian invasion” in 267 ce. According to this view, the remainder of the earlier city, including the classical Agora, was entirely abandoned between the sixth and ninth centuries ce.20 Furthermore, the area of the classical Agora included small private residences only from the tenth century ce onward.

18 19 20

Bouras 2010, 28 fig. 1; Kazanaki-​Lappa 2002, fig. 1; Veikou 2012. On Figure 10.4, the fortification wall can be found within the red circle. Frantz 1988, 57–​94, 117–​22; Kazanaki-​Lappa 2002, 640–​41; Vroom and Kondyli 2015, 323.

250 Vroom

­f igure 10.4  Map of Athens, including the diminution of the area of the antique city in Late Antiquity until 1204 ce After: Bouras 2010, ­f igure 1

These buildings supposedly followed a uniform housing plan, with a group of simple rooms around a courtyard.21 For several years, my research team and I have conducted detailed studies on remains from the seventh and ninth centuries ce from this area of the classical Agora. Unexpectedly, we discovered much activity during this period: the construction of houses, graves, and roads, the development of industrial activities, and finds such as coins and glass objects.22 These remains mostly came from one of the many wells situated outside of the late antique city walls (in Figure 10.5 these finds are highlighted with a dark line and the wells with red circles).23 These finds clearly contradict formerly held views, which postulated that the area outside of the city walls was deserted at this time.24 21 22 23 24

Frantz 1961, fig. 34; Frantz 1988; Shear 1984, fig. 17; idem 1997, fig. 7–​9. See for example Charanis 1955; Tzavella 2008; Stern 2012. These features are primarily water wells, which were later filled with garbage. The dates of the four possible wells are unfortunately not presently clear. See Vroom 2013, fig. 17, more recently published in Vroom and Kondyli 2015, fig. 20. Vroom 2013, 97–​104; Vroom and Kondyli 2015, 323–​26.

New Light on the Dark Ages


­f igure 10.5  Map of the Agora with wells and building structures from the “Dark Ages,” including the late antique city wall (red line) J. Vroom, E. Tzavella, ascsa

The graph in Figure 10.6 provides an overview of the six wells containing objects from the so-​called “Dark Ages.” These objects include both locally produced and imported goods. One well from the western part of the excavations (H11:1 in Figure 10.5) stands out because of the large amount of material found in it. The well is 12 meters deep and contains ceramics, building material (such as tiles and stones), fragments with inscriptions, a bronze lamp, and a coin

252 Vroom

­f igure 10.6  Graph with wells with finds from the “Dark Ages” in the Athenian Agora J. Vroom

from the tenth century ce (Figure 10.7). The well was in use from the seventh until the ninth century ce and reused from the tenth century ce onward. We can distinguish three successive stratigraphic periods: late Roman (ca. fourth to sixth century ce, abbreviated lr), early Byzantine (ca. seventh to ninth century ce, abbreviated ebyz), and finally mid-​Byzantine (ca. tenth to twelfth century ce, abbreviated mbyz).25 We can identify a large variety of pottery shapes in this well. In the lower parts of the well, for example, we came across late Roman amphorae, while in the upper parts of the well we found mid-​Byzantine tableware.26 The contents of this well illustrate the continuous inhabitation of this outer area and show that the Byzantine settlement during the supposedly “Dark Ages” was much more resilient than is often assumed. In addition to the mid-​Byzantine tableware we also found fragments of cooking pots in the upper parts of the well. This pottery seems to indicate a shift from a commercial to a more domestic function of the well, which perfectly aligns with the evident (re)habitation of this part of the Agora outside of the late antique fortification walls in the mid-​Byzantine period.27 Because of its clear stratification, the well generally offers a reliable relative dating of the discovered objects. In the middle and lower layers, for example, 25 26 27

Vroom and Kondyli 2015, fig. 22. Vroom and Kondyli 2015, fig. 23; for amphorae from this period see Vroom 2014, 53–​61. See for example Shear 1984, 50–​57; idem 1997, 522–​24, 531–​32.

New Light on the Dark Ages


­f igure 10.7  Cross-section of well H11:1.2 from section Zeta in the Athenian Agora with finds, varying from the late Roman until the mid-​Byzantine periods J. Vroom, E. Tzavella, P. Doeve

254 Vroom we found some exemplary material from the “Dark Ages” that allowed us to identify the types of pottery (such as the abovementioned cooking pots) and to compare them with the rest of the excavated material from the Agora. These objects are usually quite difficult to recognize, not only in excavations, but also most of all in surveys. The origin of the material can be determined by studying the type of clay and the pottery shapes. The results show changes in Athens’s foreign trade over the course of the three consecutive periods outlined above: late Roman to early Byzantine (lr-​e byz), early Byzantine (eybz), and mid-​Byzantine (mbyz) (Figure 10.8).28 To begin, the ceramic finds—​especially the amphorae—​show that during the late Roman to early Byzantine period the Agora was part of an extensive economic network that extended from Gaza and Sinope (Black Sea) to Crete and southern Italy. Second, during the early Byzantine period this part of the city traded primarily within the Aegean region, but it maintained contact with southern Italy; moreover, the local production of pottery flourished. Third, in the mid-​Byzantine period the local production of ceramics increased

­f igure 10.8  Graphs and drawings of finds from well H11:1.2 in the Athenian Agora J. Vroom, F. Kondyli 28

Vroom and Kondyli 2015, fig. 21.

New Light on the Dark Ages


drastically, but the city maintained contact with the Aegean region, the northwest of Greece, and southern Turkey (Cilicia). In short, the economic networks did not collapse entirely in the wake of the Islamic conquests, even if the pottery finds from the Agora show that 1) its trade decreased to the shrinking Byzantine Empire and 2) Athens increasingly relied on its own ceramic production. 4

Second Case Study: Butrint

The second case study concerns the coastal city of Butrint in southwestern Albania, across from the island of Corfu (Figure 10.1). An international team directed by Richard Hodges and the Archaeological Institute of Albania have been conducting excavations here since 1994.29 Positioned on the east coast of the Adriatic Sea, across from Apulia, Butrint formed an important link between the Byzantine world and medieval Italy.30 The city had been inhabited since the archaic period, but it experienced a peak of wealth in the Roman and Byzantine periods.31 The administrative center of Butrint was situated on a peninsula, which is where the most important monuments have been excavated.32 The majority of Byzantine finds from recent excavations have likewise been found on this peninsula.33 In this section, I will focus on the early Byzantine (early medieval) finds from the excavation of two towers in the so-​called western defensive wall on the peninsula of Butrint.34 This fortification system, which can be dated to the late fifth century ce, consisted of a 106-​meter long wall with three towers. Two towers are square (wd 1 and wd 2), while the third is horseshoe shaped (Figure 10.9). The square towers measure 4.5 by 4.5 meters on the inside, with walls 1.4 meters thick. The highest point of the first tower (wd 1), which survives to this day, currently stands at 9 meters (Figure 10.10). For this reason, we know that both towers not only had a ground floor level, but also must have had a first floor (see Figure 10.11 for a reconstruction drawing of wd 1 in Butrint).35 At some point in the early Byzantine period a fire broke out at the bottom of the two square towers (wd 1 and wd 2), which caused the first floor and the roof of both towers to collapse. This event sealed everything on the inside of 29 30 31 32 33 34 35

See for example Hodges et al. 2004; Hodges 2006; idem 2008; Bowden and Hodges 2011. Soustal 2004, 20; Vroom 2018. Soustal 2004, 20–​22; Hodges 2006; idem 2008. Hodges et al. 2004, fig. 1.2; Martin 2004. Vroom 2004; idem 2006; idem 2008; idem 2021a; idem 2013; Hodges and Vroom 2007. Vroom 2012a, fig. 11, 13, 16–​18; idem 2012b; Vroom and Kondyli 2015, 318–​23. Vroom and Kondyli 2015, 393 fig. 5.

256 Vroom

­f igure 10.9  Drawing of the western defence wall in Butrint, Albania, with tower 1 (wd 1) and tower 2 (wd 2) depicted Butrint Foundation

the towers in a “Pompeian manner.”36 We are, in short, dealing with very well-​ preserved and sealed contexts. Initially, archaeologists believed that simultaneous fires in both towers had been caused by a military attack on Butrint.37 Careful study of the excavated objects from the towers indicated a much more complicated story.38 The pottery from the preserved contexts of the two towers cannot be dated to a single period. Furthermore, it is possible that a simple, “ordinary” kitchen fire had raged in both towers. In the first tower (wd 1; Figure 10.11) we uncovered a large collection of glass fragments, including sixty-​nine small drink chalices and broken glass from windows, as well as an odd-​looking metal object and an enormous jumble of broken early medieval pots.39 The pottery finds from this tower date mostly 36 37 38 39

Vroom and Kondyli 2015, 318 fig. 6. Hodges et al. 2009; Kamani 2011; idem 2013. Vroom and Kondyli 2015, 320, fig. 7–​9. Jennings 2010; Jennings and Stark 2013; Vroom and Kondyli 2015, 320–​21. Strangely, a small lekythos, a perfume bottle from the fourth century ce, was found in this same locked

New Light on the Dark Ages


­f igure 10.10  Photograph of tower 1 (wd 1) in Butrint, Albania J. Vroom

to the early Byzantine period, to the late eighth and early ninth century ce. Radiocarbon dating of the charcoal of the burned beams from the ceiling of context in the tower. This object probably belonged to a person from the early Middle Ages with an antiquary’s interest in beautiful objects.

258 Vroom

­f igure 10.11  Reconstruction drawing of the interior of tower 1 (wd 1) in Butrint, Albania, including photographs of finds J. Vroom, W. Euvermans, Butrint Foundation

this tower confirmed this date.40 The recovered material consisted primarily of amphorae, domestic wares (used for cooking and conserving food), and tableware (used for serving and consuming food and drinks).41 Looking at the function of the pottery it becomes clear that amphorae and domestic ware were dominant in the excavations from the early Byzantine layers in both towers, while tableware appeared less frequently.42 It is difficult to explain this phenomenon. Is this decline in tableware due to the fact that people in this (turbulent) period ascribed less value to ceramic bowls or dishes for consumption on the dinner table? Or is it because tableware in this period was made of other materials, such as wood or leather, which are generally poorly preserved in the Mediterranean? Additionally, we have tried to reconstruct the interiors of both towers before the fires. Through a detailed study of the archaeological records, pictures, 40 41 42

Vroom 2012a; idem 2012b, 291 fig. 5. Vroom 2012a, fig. 11, 13, 16–​18; idem 2012b, fig. 5–​12. Vroom 2012a; idem 2012b; Vroom and Kondyli 2015, fig. 11.

New Light on the Dark Ages


drawings, and maps, we have reconstructed, for example, the original position of twenty-​four complete and half-​complete pots and fourteen other artifacts in the first tower.43 The reconstruction drawing of the first tower in Butrint (see Figure 10.11) shows the tentative result of our research into the position of artifacts in this tower (wd 1).44 The amphorae were concentrated in the southeastern corner of the square tower. The glass artifacts, which were also excavated in a later stadium, were situated directly next to the door in a wicker basket and in a box of perishable material. The glass had likely been stored in this way because it was ready to be made into new glass products through melting, a common type of recycling in Byzantine times. Cutlery was kept close to the hearth in the middle of the tower. Two tall jugs, however, have been found near the eastern wall in the tower; the function and position of these jugs remains ambiguous. The peculiar-​looking metal object noted above turned out to be a door handle of the hatch leading to the first floor. Our research into these artifacts demonstrates that tower wd1 was not only used for military purposes during the early Byzantine period, but also at the same time functioned as a residence, where the storage of goods, industrial activities (such as the recycling of glass), the preparation of food, and cooking came together in one place.45 Finally, when we consider the origin of the pottery finds from both towers, it is clear that local products were used together with imported ceramics from southern and northern Italy and the northeastern Mediterranean Sea region.46 Some products came from the capital, Constantinople. This circumstance contradicts established theories about how the western part of the Byzantine Empire functioned following the Arab expansion. Historians and archaeologists have generally assumed that the Byzantine trade system in the Mediterranean, especially in the Byzantine areas around the Adriatic Sea, imploded after the seventh century ce.47 Furthermore, it is assumed that little contact existed between the different regions in the Mediterranean, and that economic activity and trade took place on an exclusively local level. Gradually, however, we are starting to understand, partly because of finds in excavated time capsules such as the towers in Butrint, that this purely local emphasis cannot have been the case and that everyday reality at the time likely was much more complicated. In particular, new trade networks probably developed in this period between different regions along the northern Mediterranean littoral. 43 44 45 46 47

Vroom 2012a; idem 2012b; Vroom and Kondyli 2015, fig. 18. See also Vroom and Kondyli 2015, fig. 18–​19. Vroom 2012a; idem 2012b. See Vroom and Kondyli 2015, 322–​23 for more literature. Vroom 2012b, 291–​94; Vroom and Kondyli 2015, fig. 16. See for this discussion Liebeschuetz et al. 2001.

260 Vroom 5


Through two archaeological examples, Athens in Greece and Butrint in southern Albania, I have presented two “Dark Age” assemblages that give a more nuanced picture of trade contact and daily life in Byzantine cities in the period of the seventh to the ninth century ce.48 This recent research into the material culture of the Byzantine Empire during the “Dark Ages” shows a more complex image than that posited by the prevailing view of this period. It must therefore lead to a shift in our thinking about the “Dark Ages,” especially the idea that the rise of the caliphate, from the seventh century ce onward, dealt a critical blow to the Byzantine economy. While the larger Mediterranean network ceased from the seventh century ce onward and certain long-​distance trade contacts indeed seem to have disappeared, the various regions of the Byzantine Empire were not reduced to complete autarky, as the findings from Athens show. New trade networks within the Byzantine Empire countered the loss of wider contacts. While these networks were in some cases admittedly smaller—​limited as they were to the Aegean region, Asia Minor, or the southern Adriatic—​they nevertheless show the adaptive capacities of the empire in this period. Furthermore, some critical parts of the city and suburbs were not abandoned during the Byzantine period, but remained inhabited even after the seventh century ce and the rise of the caliphate. At the Athenian Agora, which was situated outside of the late antique walls, all sorts of activities took place in the early Byzantine period (ca. seventh to ninth century ce), including the construction of houses, graves, roads, and wells. In this period, wells contained both locally produced and imported ceramics, such as amphorae from southern Italy. We can see a similar phenomenon at the two towers in the western defensive walls of Butrint, where the contexts of the late eighth until the early ninth century ce (apart from imported amphorae from southern Italy and the Aegean region) also included large quantities of recycled glass and domestic wares. While some of these finds point to a higher sense of self-​sufficiency, they also attest to a resilience of smaller-​scale trade networks within a diminished Byzantine Empire. What these two examples indicate above all is that Byzantine archaeology must invest in the near future in good material knowledge of both recent and older excavated finds, which will generate new questions and hypotheses about the “Dark Ages.”


These examples represent some of the many questions that my research team and I have attempted to solve in a creative and challenging way, as we work to establish new insights.

New Light on the Dark Ages



First of all, I would like to thank the directors of the archaeological projects in Athens and Butrint (John McKesson Camp ii and Richard Hodges, respectively) and their teams, as well as the Archaeological Institute in Albania and the American School of Classical Studies at Athens, for the pleasant cooperation during the excavations in recent years. Furthermore, I would like to thank the Dutch Netherlands Organisation for Scientific Research (nwo), the Packard Humanities Institute, and the 1984 Foundation for financial support for my research in Greece and Albania.


Aupert, Pierre. 1989. “Les Slaves à Argos.” Bulletin de Correspondance Hellénique 113:417–​19. Baker, G. P. 2002. Justinian: The Last Roman Emperor. New York: Cooper Square Press. Barker, John W. 1966. Justinian and the Later Roman Empire. Madison: University of Wisconsin Press. Bouras, Charalambos. 2010. Βυζαντινή Αθήνα 10ος-​12ος αι. Athens: Benaki Museum. Bowden, William and Richard Hodges. 2011. Butrint 3: Excavations at the Triconch Palace. Oxford: Oxbow Books. Burbank, Jane, and Frederick Cooper. 2010. Empires in World History: Power and the Politics of Difference. Princeton: Princeton University Press. Camp, John McKesson II. 1986. The Athenian Agora: Excavations in the Heart of Classical Athens. New York: Thames and Hudson. Camp, John McKesson II. 2001. The Archaeology of Athens. New Haven: Yale University Press. Camp, John McKesson II. 2007. “The Athenian Agora: Excavations of 1989–​1993.” Hesperia 76:627–​63. Camp, John McKesson II, and Craig A. Mauzy, eds. 2009. The Athenian Agora: New Perspectives on an Ancient Site. Mainz: Verlag Philipp von Zabern & The American School of Classical Studies. Charanis, Peter. 1955. “The Significance of Coins as Evidence for the History of Athens and Corinth in the Seventh and Eighth Centuries.” Historia 4:163–​72. Curta, Florin, ed. 2005. Borders, Barriers, and Ethnogenesis: Frontiers in Late Antiquity and the Middle Ages. Turnhout: Brepols Publishers. Curta, Florin. 2006. Southeastern Europe in the Middle Ages 500–​1250. Cambridge: Cambridge University Press. Decker, Michael J. 2016. The Byzantine Dark Ages. London: Duckworth.

262 Vroom Ellis Davidson, H. R. 1973. “The Secret Weapon of Byzantium.” Byzantinische Zeitschrift 66:61–​74. Frantz, Alison. 1961. The Middle Ages in the Athenian Agora. Princeton: American School of Classical Studies at Athens. Frantz, Alison. 1988. Late Antiquity A.D. 267–​700. The Athenian Agora: Results of Excavations Conducted by the American School of Classical Studies at Athens (xxiv). Princeton: American School of Classical Studies at Athens. Foss, Clive. 1975. “The Persians in Asia Minor and the end of Antiquity.” English Historical Review 90:721–​43. Foss, Clive. 1977. “Archaeology and the Twenty Cities of Byzantine Asia.” American Journal of Archaeology 81:469–​86. Haldon, John. 2005. The Palgrave Atlas of Byzantine History. Basingstoke: Palgrave Macmillan. Haldon, John. 2006. “ ‘Greek Fire’ Revisited: Recent and Current Research.” In Byzantine Style, Religion and Civilization: In Honour of Sir Steven Runciman, edited by Elizabeth M. Jeffreys, 290–​325. Cambridge: Cambridge University Press. Haldon, John, and Hugh Kennedy. 1980. “The Arab-​Byzantine Frontier in the Eighth and Ninth Centuries: Military Organisation and Society in the Borderlands.” Zbornik Radova Vizantoloskog Instituta 19:79–​116. Hodges, Richard. 2006. Eternal Butrint: A UNESCO World Heritage Site in Albania. London: General Penne. Hodges, Richard. 2008. Shkëlqimi dhe Rënia e Butrintit Bizantin /​The Rise and Fall of Byzantine Butrint. London, Tirana: Butrint Foundation. Hodges, Richard, William Bowden, and Kosta Lako, eds. 2004. Byzantine Butrint: Excavations and Surveys 1994–​99. Oxford: Oxbow Books. Hodges, Richard, Solinda Kamani, Matthew Logue, and Joanita Vroom, eds. 2009. “The Sack of Butrint, c. AD 800.” Antiquity 83.320. http://​​projgall/​hodges. Hodges, Richard, and Joanita Vroom. 2007. “Late Antique and Early Medieval Ceramics from Butrint, Albania.” In La circolazione delle ceramiche nell’Adriatico tra tarda antichità ed altomedioevo. III Incontro di Studio CER.AM.IS (Documenti di Archeologia 43), edited by Sauro Gelichi and Claudio Negrelli, 375–​88. Mantua: sap Società Archeologica. Ivison, Eric A. 2007. “Amorium and the Byzantine ‘Dark Ages’ (Seventh to Ninth Centuries).” In Post-​Roman Towns, Trade and Settlement in Europe and Byzantium, vol. 2: Byzantium, Pliska, and the Balkans (Millenium-​Studien 5.2), edited by J. Henning, 25–​29. Berlin: Walter de Gruyter. Jennings, Sarah. 2010. “A Group of Glass ca. 800 A.D. from Tower 2 on the Western Defences, Butrint, Albania.” In Glass in Byzantium: Production, Usage, Analyses. The Fascination of Glass Production (rgzm Tagungen 8), edited by Jörg Drauschke and Daniel Keller, 225–​35. Mainz: rgzm Publishing House.

New Light on the Dark Ages


Jennings, Sarah, and Karen Stark. 2013. “Appendix: The Glass from Tower 1 in the Western Defences.” In Butrint 4: The Archaeology and Histories of an Ionian Town, edited by Inge Lyse Hansen, Richard Hodges, and Sarah Leppard, 257–​59. Oxford: Oxbow Books. Kamani, Solinda. 2011. “Butrint in the Mid-​Byzantine Period: A New Interpretation.” Byzantine and Modern Greek Studies 35.2:115–​33. Kamani, Solinda. 2013. “The Western Defences.” In Butrint 4: The Archaeology and Histories of an Ionian Town, edited by Inge Lyse Hansen, Richard Hodges, and Sarah Leppard, 245–​56. Oxford: Oxbow Books. Kazanaki-​ Lappa, Maria. 2002. “Medieval Athens.” In The Economic History of Byzantium: From the Seventh through the Fifteenth Century, edited by Angeliki E. Laiou et al., 639–​46. Washington DC: Dumbarton Oaks Research Library and Collection. Kennedy, Hugh. 2007. The Great Arab Conquests: How the Spread of Islam Changed the World We Live In. London: Weidenfeld & Nicolson. Liebeschuetz, J. H. W. G. 2001. “The Uses and Abuses of the Concept of ‘Decline’ in Later Roman History.” In Recent Research in Late Antique Urbanism (Journal of Roman Archaeology Supplementary Series 42), edited by Luke Lavan, 233–​45. Portsmouth: Journal of Roman Archaeology. Martin, Sally. 2004. “The Topography of Butrint.” In Byzantine Butrint: Excavations and Surveys 1994–​99, edited by Richard Hodges, William Bowden, and Kosta Lako, 76–​ 103. Oxford: Oxbow Books. McEvedy, Colin. 1992. The New Penguin Atlas of Medieval History. London: Penguin Books. Nikolakis, Michalis. 2015. “Athens: The Tourist Capital of Post-​war Greece.” Pharos 21.1:21–​36. Ostrogorsky, George. 1959. “Byzantine Cities in the Early Middle Ages.” Dumbarton Oaks Papers 13:45–​66. Ostrogorsky, George. 1969. History of the Byzantine State. New Brunswick: Rutgers University Press. Pászthory, Emmerich. 1986. “Über das ‘Griechische Feuer’: Die Analyse eines spätantiken Waffensystems.” Antike Welt 17:27–​38. Poulios, Ioannis and Efthymia Senteri. 2015. “Branding Athens at a Time of Crisis: The Current Tourism Campaign of the Historic City.” Pharos 21.1:37–​56. Shear, T. Leslie, Jr. 1984. “The Athenian Agora: Excavations of 1980–​1982.” Hesperia 53:1–​57. Shear, T. Leslie, Jr. 1997. “The Athenian Agora: Excavations of 1989–​1993.” Hesperia 66:495–​548. Stern, E. Marianne. 2012. “Early Byzantine Glass from Athens (5th–​8th Centuries).” In Byzantine Small Finds in Archaeological Contexts (Byzas 15), edited by Beate Böhlendorf-​Arslan and Alessandra Ricci, 49–​60. Istanbul: Byzas.

264 Vroom Soustal, Peter. 2004. “The Historical Sources for Butrint in the Middle Ages.” In Byzantine Butrint: Excavations and Surveys 1994–​99, edited by Richard Hodges, William Bowden, and Kosta Lako, 20–​26. Oxford: Oxbow Books. Treadgold, Warren. 1997. A History of the Byzantine State and Society. Stanford: Stanford University Press. Tzavella, Elli. 2008. “Burial and Urbanism in Athens (4th–​9th c. A.D.).” Journal of Roman Archaeology 21:352–​68. Veikou, Myrto. 2012. “Byzantine Histories, Settlement Stories: Kastra, ‘Isles of Refuge’, and ‘Unspecified Settlements’ as In-​Between or Third Spaces.” In Oι βυζαντινές πόλεις (8oς—​15oς αιώνας), edited by T. Kioussopoulou, 159–​206. Rethimnon: Philosophical School of the University of Crete. Vroom, Joanita. 2003. After Antiquity: Ceramics and Society in the Aegean from the 7th to the 20th Century A.C: A Case Study from Boeotia, Central Greece (Archaeological Studies Leiden University 10). Leiden: Faculty of Archaeology. Vroom, Joanita. 2004. “The Medieval and Post-​Medieval Fine Wares and Cooking Wares from the Triconch Palace and the Baptistery.” In Byzantine Butrint: Excavations and Surveys 1994–​99, edited by Richard Hodges, William Bowden, and Kosta Lako, 278–​ 92. Oxford: Oxbow Books. Vroom, Joanita. 2006. “Corfu’s Right Eye: Venetian Pottery in Butrint (Albania).” In The Heritage of the Serenissima: The Presentation of the Architectural and Archaeological Remains of the Venetian Republic. Proceedings of the International Conference Izola—​ Venezia 4—​9.11.2005, edited by Mitja Guštin, Sauro Gelichi, and Konrad Spindler, 229–​36. Koper: Založba Annales. Vroom, Joanita. 2008. “Dishing Up History: Early Medieval Ceramic Finds from the Triconch Palace in Butrint.” Mélanges de l’École française de Rome—​Moyen Âge 120.2:291–​305. Vroom, Joanita. 2011. “The Other Dark Ages: Early Medieval Pottery Finds in the Aegean as an Archaeological Challenge.” In When did Antiquity End? Archaeological Case Studies in Three Continents (BAR I.S. 2268), edited by Redha Attoui, 137–​58, Oxford: Archaeopress. Vroom, Joanita. 2012a. “From One Coast to Another: Early Medieval Ceramics in the Southern Adriatic Region.” In From One Sea to Another: Trading Places in the European and Mediterranean Early Middle Ages, Proceedings of the III International SAAME Conference—​Comacchio 27th–​29th March 2009, edited by Sauro Gelichi and Richard Hodges, 375–​413. Turnhout: Brepols Publishers. Vroom, Joanita. 2012b. “Early Medieval Pottery Finds from Recent Excavations at Butrint, Albania.” In Atti del IX congresso internazionale sulla ceramica medievale nel Mediterraneo, Venezia, Scuola Grande dei Carmini, Auditorium Santa Margherita, 23–​27 novembre 2009, edited by Sauro Gelichi, 289–​96. Florence: All’Insegna del Giglio.

New Light on the Dark Ages


Vroom, Joanita. 2013. “The Medieval and Post-​Medieval Pottery Finds.” In “The Medieval Church and Cemetery at the Well of Junia Rufina”, edited by Alessandro Sebastiani et al., 243–​40. In Butrint 4: The Archaeology and Histories of an Ionian Town, edited by Inge Lyse Hansen, Richard Hodges, and Sarah Leppard, 215–​44. Oxford: Oxbow Books. Vroom, Joanita. 2014 [2005]. Byzantine to Modern Pottery in the Aegean: An Introduction and Field Guide, 2nd and revised edition. Turnhout: Brepols Publishers. Vroom, Joanita, ed. 2015. Medieval and Post-​ Medieval Ceramics in the Eastern Mediterranean: Fact and Fiction (Medieval and Post-​ Medieval Mediterranean Archaeology Series 1). Turnhout: Brepols Publishers. Vroom, J. 2016. “Byzantine Sea Trade in Ceramics: Some Case Studies in the Eastern Mediterranean (ca. Seventh—​Fourteenth Centuries).” In Trade in Byzantium: Papers from the Third International Sevgi Gönül Byzantine Studies Symposium, edited by P. Magdalino and N. Necipoğlu, 157–​77. Istanbul: Koç University’s Research Center for Anatolian Civilizations (anamed). Vroom, J. 2017. “Ceramics.” In The Archaeology of Byzantine Anatolia: From the End of Late Antiquity until the Coming of the Turks, edited by P. Niewöhner, 176–​93. Oxford: Oxford University Press. Vroom, J. 2018. “On the Edge: Butrint on the Western Frontier of the Byzantine Empire.” In The Archaeology of Imperial Landscapes: A Comparative Study of Empires in the Ancient Near East and Mediterranean World, edited by B. Düring and T. Stek, 272–​98. Cambridge: Cambridge University Press. Vroom, J. and F. Kondyli. 2015. “ ‘Dark Age’ Butrint and Athens: Rewriting the History of Two Early Byzantine Towns.” In Medieval and Post-​Medieval Ceramics in the Eastern Mediterranean: Fact and Fiction (Medieval and Post-​ Medieval Mediterranean Archaeology Series 1), edited by J. Vroom, 317–​42. Turnhout: Brepols Publishers.

Index of Names and Subjects ʿAbbāsids 13, 103, 125, 130, 136, 145n74, 146, 152–​54, 156, 159n131, 162–​63, 231n80, 235n102 ʿAbd Allāh b. Jābir 173–​75 ʿAbd Allāh b. Musayyab 185 ʿAbd al-​Malik (caliph) 122, 135–​36, 142, 155 ʿAbd al-​Malik ibn Hishām, 78 ʿAbda ibn al-​Ṭabīb 122n6, 147 Abraham 22, 213–​24, 229, 231, 235n100, 239 Abū Dahbal al-​Jumaḥī 134, 151 Abū Dhuʾayb al-​Hudhalī 133, 138n44 Abū al-​Faraj al-​Iṣfahānī 158 Abū al-​Fīl 134 Abū Rāʾiṭa of Tikrīt 101 Abū Ṣufra 154–​55, 157, 159 Abū al-​Ṭayyib ʿAbd al-​ʿUzzā 134 Achaemenid Empire 113, 223n50 Adam 39 Aden 138 ʿAdī ibn Zayd 144–​45 Adriatic Sea 21, 255, 259–​60 Aegean region 245, 254–​55, 260 Afghanistan 100 Africa 6–​7, 9, 47  See also North Africa ahl al-​dhimma 184–​85 Aḥmad ibn ʿAbd al-​Jabbār al-​ʿUṭāridī 79 al-​Aḥwaṣ al-​Anṣārī 138, 150–​51 al-​ʿajam/​al-​ʿujm 134–​35, 149n95, 155n119 al-​Akhṭal 141–​42, 164 al-​Akhnas b. Shihāb 132 Albania 21, 245–​46, 255–​58, 260–​61 Aleppo 100 Alexandria 2, 5n20, 44n47, 178 Allāh 45, 56, 66, 145n76, 146n82, 147–​48, 164, 182, 197n7, 204 Amalfi 7 amīr al-​muʾminīn 183–​84 ʿAmmār of Basra 101 ʿAmr ibn ʿAbd al-​Jinn al-​Tanūkhī 144 ʿAmr ibn al-​ʿĀṣ 176 ʿAmr ibn Muzayqiya 153 Anatolia 100 al-​Andalus 102, 181

Anṣār 152 apocalyptic texts 214, 221n40, 223n50, 228–​ 30, 232n86, 233, 235–​36  See also eschatology Apulia 255 Arabian Peninsula 6, 11, 15, 18, 20, 32, 35, 37, 40, 47–​50, 91–​93, 99, 124–​26, 128–​29, 136–​39, 144, 148, 152, 153n109, 157n123, 158–​61, 171–​72, 175n7, 184, 190, 198–​99, 204, 207, 216, 232, 237, 247 Arabic (language) 11, 13, 20–​21, 24, 33, 38n26, 47, 50, 76n8, 85, 89–​113, 121, 125n10, 134–​35, 139, 144–​45, 149n95, 154–​56, 171, 173–​77, 179, 182–​91, 195–​209, 213 Arabic script 21, 37, 38n26, 195–​209 Nabataeo-​Arabic script 196–​98, 201–​9 Arabicization 17, 19–​21, 24, 89–​113, 179, 188, 196, 213 Aramaic (language) 13, 93, 97, 100–​2, 107n74, 110, 197, 207, 213, 222 Syriac 13, 15, 32n5, 47, 95, 112, 126n14, 183, 201 Aristotle 96 Armenia 20, 149 asbāb al-​nuzūl 35, 82–​83 al-​Aʿshā Maymūn ibn Qays 132, 139 Ashʿarī 134 Ashmūnayn (Hermopolite) 176, 178n18 Asia Minor 33, 246, 260 ʿaṭāʾ 163, 188 Athens 21, 245–​55, 260 Athenian Agora 248–​55, 260 Aws 152, 207 al-​Azd 91, 121, 152–​63 Babylon 22, 94, 215n12, 221–​24, 227, 228n69, 235, 238 Bactria/​Bactrian (language) 94, 175n6 Badr, Battle of 38, 40n33 Baghdad 39n29, 39n31, 100, 103 al-​Balādhurī 99–​101 Basra 99–​101, 111, 121, 135, 141, 149n95, 153–​63 Believers 33–​40, 44–​45, 49, 99, 121n1, 143n67, 148, 183–​84

268  Belisarius, Flavius 246 Bible 35, 59, 63, 76, 144n68, 145–​46, 213n8, 220 bilingualism 97, 111–​12, 177, 179, 182–​83, 185, 190n70, 204n23 Black Sea 25, 254 Bukhārā 108 Burquʿ 198, 205 Buṣrā 138 Butrint 245–​47, 255–​61 Byzantine Empire 2, 4–​7, 9, 13, 15, 18, 20–​21, 24, 32–​34, 39, 40n33, 43–​50, 95, 99–​100, 122–​24, 128n16, 134, 145, 171, 174, 176–​78, 182, 187–​90, 212, 219, 222n48, 223–​29, 230n75, 232–​33, 235, 238–​39, 245–​ 48, 252–​60 art and culture 9–​10, 178, 227 thema 247 trade 4–​6, 9, 21 Cairo. See Fusṭāṭ calendar 48, 174, 182–​83 hijra 174, 182–​83 Diocletian 183 caliphate 1, 21, 36n21, 39, 120–​64, 179, 188, 225, 227, 247, 260 Carthage 3, 6–​7 Caucasus 33, 100, 123 central Asia 1, 90, 94, 100, 108, 128n17, 129, 175n6 Chalcedon, Council of 18, 41–​45, 50 Chalcedonians 18, 45–​49, 52, 212 Charlemagne 3 Children of Israel 35, 45, 220–​24 Chilperic ii 6 Christ 24n69, 34, 36, 40–​42, 235n103 Christianity and Christians 10, 12–​18, 20, 24, 32–​50, 62–​63, 65, 93, 95, 98, 107, 110, 120, 126n14, 200, 212, 213n8, 217, 220–​21, 226–​27, 229, 230n74, 233, 235n103, 236 Arabian 32n3, 101–​3, 143–​50, 158n128, 164, 196–​98, 202, 204–​5 Eastern 18, 42 Egyptian 98–​99 Syriac 16, 141–​42, 173, 179, 182, 184 Cilicia 255 circumcision 149n97, 156–​58, 220n36 coins 7, 14, 16, 179, 181, 182n29, 202, 250–​51

Index of Names and Subjects colonization 10, 25, 94, 98, 102–​3, 107, 113 exploitation colony 98 settlement colony 98, 100, 103 connectivity 3n10, 6–​9, 22–​23, 25 Constantinople 9, 42, 47–​49, 233, 246, 259 conversion 20, 81n19, 99, 105–​13, 147, 185 Cook, Michael 12–​13, 74–​75, 79 Coptic (language) 93, 95–​97, 99, 112, 171, 174, 176–​77, 179, 182–​84, 189 Copts 96, 184  See also Egyptians Corbey, Abbey of 6 Corfu 246, 255 Crete 47, 254 Crimea 246 Crone, Patricia 12–​13, 75, 79, 153, 154n113 Crusades 39–​40, 45, 212n1, 233n90, 234n92 Ctesiphon 34, 43n44, 99, 225n58 Cuypers, Michel 15 Cyprus 7, 246 Damascus 34, 100–​1, 178–​79, 186, 188, 234–​ 35, 247 Dark Ages 245–​60 Deuteronomy 65 Dhū Nuwās 47–​49 Donner, Fred 15, 126 Durayd ibn al-​Ṣimma 137 Eastern Roman Empire. See Byzantine Empire economic depression 4–​5 early medieval depression 3, 6–​8 Egypt 6, 9, 19–​21, 24, 33–​34, 41n36, 98–​99, 101–​3, 171–​91, 219–​20, 233n90, 246–​47 Egyptians 96, 98–​99, 171–​73, 175, 177, 179, 182–​86, 188, 190  See also Copts Eliezer 218–​19 Emesa 100 Emigrants 13, 122–​24, 126, 128, 134–​38, 142, 147, 152, 163–​64, 183–​84 eschatology 35, 150, 214, 220n36, 222, 229n70, 231, 233, 236, 238 eschaton 46, 229–​30  See also apocalyptic texts Ethiopians 18, 48 Ethiopic (language) 15, 121


Index of Names and Subjects ethnicity 16n59, 17, 20, 41, 90–​92, 95, 124, 126–​27, 131n23, 135–​36, 152, 153n109, 158, 162, 178, 185 ethnogenesis 90–​91, 121, 127, 131n23, 135, 137, 152, 154, 162–​64 Europe 2–​4, 13, 18, 34, 46, 56, 69, 74, 78, 86, 121, 178 (early) modern 2, 10, 94, 125–​27, 238 medieval 4, 7, 10, 90, 120 al-​Farazdaq 141–​42, 150, 155–​59, 161, 163 fatḥ (pl. futūḥ/​futūḥāt) 121–​24 Fatimid dynasty 98, 233n90 Fayyūm 103, 178 fitna 124, 135  See also war Fusṭāṭ 102, 171, 175–​78, 187 Galen 96 Gaul 3, 9 Gaza 138n44, 254 Geiger, Abraham 12 genealogy 152–​54, 163–​64, 214, 220 Ghassānids 47, 152 ghiyār 106 Gibbon, Edward 2, 9, 125 Goldziher, Ignaz 11, 75, 133n31, 147n86 graffiti 196, 206–​7 Greek fire 247–​48 Greek (language) 20, 47, 89, 93–​97, 100–​1, 107n74, 112, 127, 171, 173–​77, 179, 182–​86, 189–​90, 197, 204n23 Gregory the Decapolite 5 Hagar 213, 215–​16, 219–​21, 239 Hagia Sophia 246 Hajj 77, 139–​43, 150, 156, 158, 164, 186 Ḥamza al-​Iṣfahānī 160 heaven. See paradise Hebrew (language) 15, 213, 215, 223n50, 229–​ 30, 235n104 Hellenism 94–​96, 175n6, 177, 189–​90, 229 Heraclitus 33, 223n50 Herulian invasion 249 Ḥijāz 125, 132–​33, 137–​38, 140, 150–​51, 206, 216 Ḥimyar 48–​49, 91, 93, 153n109, 160 Ḥimyarite (language) 93

al-​Ḥīra 144, 145n74, 202–​4 homily 143, 227n63, 237–​38 houris 13 Ḥudaybiyya, Treaty of 39 Hudhayl 133, 137, 140 hymn 12, 143 hypostatic union 41–​42 Iberian Peninsula 92, 93n14, 102, 123 Ibn ʿAbbās 57, 68, 80–​82, 84–​85 Ibn Ḥabīb 160 Ibn Hishām 77–​84 Ibn Isḥāq 79–​84, 86 Ibn al-​Kalbī 159–​60 Ibn Khaldūn 161 Ibn Khurdādhbih 5 Ibn Qutayba 131, 145, 160 Ibn Sallām al-​Jumaḥī 130n19, 144, 159n131 identity 16, 19–​21, 25, 33–​34, 44, 90–​91, 95–​ 98, 104–​5, 110, 120–​27, 129–​33, 135–​36, 138, 142, 146, 148–​50, 152–​64, 237 idolatry 36n20, 108, 143, 155–​56, 159–​60, 216n21, 217n24 Ifrīqiyya 7 Ihnās (Heracleopolite) 174, 176, 185 ʿIkrima 80–​85 imperial language 94 incarnation 36 inscriptions 14, 129, 152n107, 153n109, 160, 172, 190, 196–​98, 200–​2, 204, 206–​9, 251 ʿAbd al-​Masīḥ amulet 203–​5 Dūmah inscription 196–​97, 204, 206 Ḥarrān inscription 206, 208 Iran 19, 92, 93n14, 100, 123, 125n10, 149n95, 154, 159 Iranian (language) 94, 97, 102 Iraq 13, 20, 24, 93, 99–​103, 110, 130, 135–​36, 138, 141, 145, 149–​50, 152, 153n109, 155–​ 56, 159, 162 Isaac 22, 213, 216–​21, 227, 239 Ishmael 22, 213–​25, 226n59, 228–​36, 238–​39 Ishmaelites 216, 217n22, 224, 229, 234 Islam 1, 3–​7, 10–​21, 24, 32–​38, 40, 44–​50, 62–​65, 69, 71, 75–​80, 81n19, 89–​92, 143–​ 47, 150–​52, 156–​61, 163–​64, 171, 183, 213, 215–​17, 220, 237–​40 Islamic law 20, 56–​61, 69, 71, 91, 185–​87

270  Islamization 17, 19–​21, 24, 92, 105–​10, 213 isnād-​cum-​matn method 19 Israel 22, 212–​14, 220–​40 Istanbul 246 Italy 7, 9, 219, 254–​55, 259–​60 Jarīr ibn ʿAtiyya 141 Jerusalem 18, 33–​36, 38, 40, 45–​50, 80, 178–​ 79, 186, 212, 218n31, 225n58, 233, 236–​38, 239n117 jizya. See poll tax John the Baptist 10 Jordan 45, 189, 195, 198, 234, 246 Judaism and Jews 12–​13, 15, 21–​22, 24n69, 35–​36, 42, 44, 47–​48, 62, 65, 81–​83, 103, 107, 110, 129n18, 184, 212–​40 Justinian i 246 Justinian Renaissance 2, 22 Kaaba 80, 139, 141n58, 216 Kaʿb ibn Mālik al-​Anṣārī 134 Kaʿb ibn Zuhayr 148 Karaite 103, 213, 237–​38 Kassite (language) 94 al-​Khazraj 152 Khurāsān 123, 149n95, 159 Kinda 91, 133n31 Koren, Judith 13 Kufa 99–​101, 122n6, 145–​46 Kultursprache 96 Kushans 94 Lakhmids 152 language shift 93–​98, 100n46, 104–​5, 109–​12 Latin (language) 89, 112, 126n14, 189 Levant 3, 9, 20, 39, 93, 95, 177, 189–​90 Lüling, Günter 12 Luxenberg, Christoph 13 Maʿadd 91, 132–​36, 138, 143, 145–​46, 153n109, 161 Marseille 2–​3, 6 Marwān ibn al-​Ḥakam 151 Marwānid era 121, 124, 127–​28, 131, 134–​39, 141–​43, 149–​52, 154–​56, 158–​59, 161–​ 64, 188 mawlā (pl. mawālī) 81–​84, 108, 109n79, 111 Māzin ibn al-​Azd 153

Index of Names and Subjects Mecca 18, 33–​34, 38–​39, 45n48, 46–​47, 49, 58, 76–​78, 80, 82–​86, 134, 139–​43, 155, 186, 206–​7, 218 Medina 33–​34, 58, 79–​83, 134, 138, 139n49, 148, 150, 152, 158, 179, 206–​9 Mediterranean Sea 1–​6, 8–​9, 11, 22–​25, 89, 112, 172, 245–​47, 258–​60 Merv 100 Mesopotamia 99, 212, 247 Messiah 42, 203, 221, 223, 228–​29, 233–​39 Messianic era 22, 229, 233, 235–​40 Middle East 56, 89–​90, 99n42, 120, 124–​29, 131, 135, 137, 164, 212–​13, 247 midrash 213n8, 215, 217, 218n24, 220–​21, 223, 225, 227–​28, 230n74, 233, 234n98, 235n100, 237, 239 Minā 140–​41 mishnah 65 miṣr (pl. amṣār) 99, 102, 122 Monophysites 42, 47–​50 Moses 151, 217, 226n60 Muʿāwiya 99, 122, 199, 200, 202, 234n94 Muḍar 91, 133n31 muhājirūn. See Emigrants al-​Muhallab ibn Abī Ṣufra 154–​55, 157 Muhallabid family 153–​55, 159, 161–​62 Muhammad, the Prophet 3, 18–​19, 32, 39, 46, 58, 74–​86, 132, 134, 143, 148, 151–​ 52, 206 biography of 18–​19, 34, 36, 75–​77, 130n19, 134 Muḥammad ibn Abī Muḥammad 80–​86 muʾminūn. See Believers Muqātil ibn Sulaymān 161 al-​Nābigha al-​Dhubyānī 132, 139, 144n68 Najrān 32, 36, 48–​50 Naples 6 Naṣr ibn al-​Azd 152–​53 Negus 36 Nestorian 42, 46 network theory 22–​25 Nevo, Yehuda 13 Nicea, Council of 41 Nöldeke, Theodor 11 North Africa 3, 6, 99n42, 100, 123, 245, 247 Nuʿmān ibn al-​Mundhir 139 al-​Nuʿmān ibn Wāʾil al-​Kalbī 132


Index of Names and Subjects Oman 136, 138, 152–​56, 158–​61, 163–​64 Ottoman Empire 246, 248 pagarch 176, 180 Pahlavi (language) 202  See also Persian (language) Palestine 33, 47, 100–​3, 178, 212, 213n7, 215, 225, 233n90, 233n91, 238, 246–​47 papyrus 2, 16–​17, 20–​21, 171–​74, 176–​78, 180, 182–​86, 188–​90, 201n18 paradise 13, 46n49, 49, 59–​60, 62–​65, 69–​70, 148, 218, 232, 236 Paran 215–​16, 239 Persia 4, 22, 32n1, 33–​34, 36–​39, 40n33, 45–​ 50, 89–​90, 93, 100, 125, 134, 154, 159, 212, 221–​23, 227, 228n69, 230n75, 236n106, 237, 239, 247 Persian Empire 18, 33–​34, 47, 99–​101, 113, 123–​24, 128n16, 133, 145, 178–​79, 190, 212, 222–​24, 228n69, 236n106, 247 Persian (language) 15, 93, 95n27  See also Pahlavi (language) pilgrimage. See Hajj piracy 5–​6 Pirenne, Henry 2–​3, 6–​9 Pirenne period, 3, 6–​7, 9 plague 4, 245 poetry 17, 19–​20, 25, 76–​78, 85–​86, 121–​22, 130–​52, 155–​64, 233n91 poll tax 176, 186, 187n56, 189 Procopius of Caeserea 246 prophet 34–​37, 216n18, 221, 233n90, 235n100 Prophet, the. See Muhammad, the Prophet proto-​hamza 205 Qaryat al-​Fāw 204 Qayrawān 100 qibla 46, 80–​81, 84 al-​Qirqisānī 103 Quḍāʿa 91 Qurʾān 11–​20, 32–​50, 56–​64, 67, 69–​71, 75, 77–​78, 81n19, 82–​83, 85–​86, 92, 121, 122n4, 124, 135, 141n58, 143, 145–​49, 161, 176, 179, 181, 186, 187n57, 197–​98, 200 early manuscripts 15, 34, 37, 200 exegesis 18, 35n12, 43–​44, 58–​59, 75n6, 85, 145, 161, 218, 222, 235n100, 236n106

Quraysh 76–​77, 80, 82–​86, 123, 140–​42, 156, 158–​61 Qurra b. Sharīk 179–​80 Qutayba ibn Muslim 108, 159 Rabbi Aqiba 222, 232n43 rabbinic literature 12, 21–​22, 213–​14, 217n24, 220–​21, 223n49, 235n100 Rabīʿa 91 al-​Rādhāniyya (Radhanites) 5 radiocarbon dating 257 al-​Rāʾī al-​Numayrī 135 rasm 37, 40 Reconquista 93n14, 102 Justinian Reconquista 9 Reichssprache 94–​96, 111 Revisionism 12–​14, 171 Roman Empire 2–​4, 8, 22, 23n65 Romania 2–​3, 8 Rome 23n65, 230n75, 246 Saʿadyah Gaon 103 ṣadaqa 186 Safaitic script 196, 204 Saʿīd ibn Jubayr 80–​85 Sāʿida ibn Juʾayya 138n44, 140 al-​Samawʾal ibn al-​ʿĀdiyāʾ 143 Saracens 102, 125n8, 183 Sarah 214, 216–​17, 219–​20, 239 Sasanian Empire. See Persian Empire seals 177–​80, 182, 189 settlement patterns 19, 24, 97–​99, 101, 103, 110, 112–​13, 122, 136 al-​Shammākh ibn Ḍirār 138, 148 Sharia. See Islamic law Sicily 7 Sinope 254 Ṣirma ibn Abī Anas 143 Sogdiana 108 Spain 1, 3, 9, 20, 90, 102–​3, 247 St. Willibald 5 Sufyān b. Qurʿa 185 al-​Sulayk ibn al-​Sulaka 137 Syria 4, 19, 24, 37–​38, 45, 47, 95, 99–​101, 103, 138, 141, 144, 149, 152, 154, 183, 207, 233n90, 246–​47 Taʾabbaṭa Sharran 137


Index of Names and Subjects

Taghlib 132, 141–​42, 164 Talmud 215n12, 217n22, 235, 237 Ṭayyāyē 102 Temple Mount 218n31, 237 thalassocracy 3, 7 trade 2–​7, 9, 21, 25, 47–​48, 50, 128, 231n80, 247, 254–​55, 259–​60 Trinity 36 Tunis 7 Turkey 32, 246, 255

al-​Walīd ibn al-​Mughīra 77 Wansbrough, John 12, 13n45, 37n23, 75 war 5–​6, 39, 44, 90, 121, 124–​26, 128, 133, 135–​ 36, 138n46, 142, 154, 156, 163, 232, 237 Persian-​Byzantine 18, 33, 49, 247–​48 Second Fitna 124, 135 Wāsiṭ 100 wawation 197, 201–​2, 207–​9 Weil, Gustav 11 wine 2, 18–​20, 56–​71, 149, 178

Umar i 4, 157n125 Umar ii 4 Umayya ibn Abī al-​Ṣalt 143 Umayyads 7, 16, 21, 101, 108, 121, 124–​26, 128–​ 30, 134–​39, 141–​42, 150–​55, 158, 160–​62, 188, 198, 200, 231n80, 234–​35, 247 Umayyad Mosque (Damascus) 10, 186  See also Marwānid era ʿUthmān ibn ʿAffān 99, 148

Yathrib. See Medina Yazīd I 21, 198, 200, 205–​6 Yazīd ibn al-​Muhallab 154, 159 Yemenites 18, 48, 134, 138n44, 146, 153n109, 158, 160–​61, 178n22 Yuezhi 94 Yūnus ibn Bukayr 79–​81, 84

Venice 7, 222n43 Visigoths 90 Wahb ibn Munabbih 160

Zayd ibn Thābit 81–​84 zakāt 186 Ziyād ibn ʿAbd Allāh al-​Bakkāʾī 79 Ziyād al-​Aʿjam 149–​51, 157–​59, 161 Zoroastrianism 15, 44, 65, 120, 212, 225n58 Zuhayr ibn Abī Sulma 140

Index of Biblical and Qurʾānic References Hebrew Bible  Gen 15:9 222, 224n53 Gen 15:12 223–​24 Gen 15:17 224 Gen 16:11 230 Gen 21 220 Gen 21:21 219 Gen 22 218n29 Gen 22:4 218 Gen 22:5 217–​18 Gen 24:3–​4 219 Gen 24:38 219 Gen 25:12 219 Gen 25:19 217n26, 219 Gen 37:1 227n63 Ex 3:8 63n33 Ex 20:15 217n25 Lev 10:9 61n29 Lev 11:13–​19 65n48 Lev 11:20–​23 65n44 Num 24:21 233, 234n96, 235n100 Num 24:23 230 Deut 14:11–​18 65n48 Deut 14:19 65n44 Deut 32:1–​14 63n35, 64n41 Deut 33:27–​28 226 Judg 9:13 59n15 2 Sam 21:20 227 2 Sam 21:21 227 2 Sam 22:1 227 Job 12:6 217n22 Ps 1:6 219 Ps 47:9 226 Ps 55:20 231 Ps 89:52 228 Ps 132:18 224 Prov 31:4–​5 59n15 Song 2:19 222 Song 5:1 63n35 Isa 14:5 234 Isa 21:7 234n93, 236n106 Isa 21:15 232 Isa 60:7 237 Isa 63:1 232

Jer 50:11 222 Lam 1:14 228n69 Ezek 23:19–​20 219 Dan 2 214, 221n40, 228 Dan 2:31–​45 221, 223 Dan 2:42 228 Dan 2:44 232 Dan 7 214, 221n40 Dan 7:7 224 Dan 8:8 222 Dan 8:20 222 Dan 11:4 223n50 Joel 3:18 63n35 Zeph 3:7 226 Zech 9:9 218n29 New Testament  Matt 26:29 56n3 Mark 14:25 56n3 Luke 22:18 56n3 Rev 7:16–​17 59n20 Rev 21:4 59n20 Qurʾān  1:6 44 1:6–​7 35 2:58 45 2:114 45n48 2:142–​50 46 2:219 56n2, 56n5, 57, 58n11, 60–​61 2:222 60–​61 2:226 56n4 2:259 45 3:19 42 3:42–​43 42 3:59 39 4:43 56n2, 57, 60–​61 4:156–​57 42n39 4:171 39 5:21 45 5:51 36 5:82 36, 42–​43 5:90 57 5:90–​91 56n2, 60n25

274  5:91 64n41 5:116–​17 42 6:139 63n38 7:137 45 8:11 64n41 9:31–​32 43 9:34 43 12:36 60n25 12:41 60n25 15:15 60 15:90–​93 77, 85 16 64–​65 16:64 62, 64, 66n52, 67, 70 16:64–​70 61–​71 16:65 62, 64, 70 16:65–​69 56n3, 62 16:65–​70 69 16:66 62–​63, 70 16:66–​69 66n52 16:67 18–​20, 56–​58, 60–​61, 66–​71 16:68–​69 64–​65, 67, 70 16:70 62, 66, 70 17:1 46–​47 17:91 56n4 18:9–​ 26 32 18:32 56n4 19:20 41 19:23 41 19:35 42n39, 43 19:37 40n34 21:91 42 22:40 43 23:19 56n4 23:50 45 24:25 199 24:35–​36 45

Index of Biblical and Qurʾānic References 25:48 64 30 33, 49 30:2 33, 38 30:2–​3 38 30:2–​5 38, 40, 45 30:4 40 30:30 43 30:32 44 30:31–​32 44 30:43 43 33:35 43 36:34 56n4 37:45–​47 60 44:54 13 44:56 59n20 47:15 56n3, 59n20, 60n25, 63, 65n51 47:51 63 50:41 46 52:29–​30 19, 77–​78 56:18–​19 60 57:13 46 57:27 43 66:12 41 70:43 46 74:11–​25 77 74:24 19 76:13–​19 59n20 76:21 60 79:14 46 83:52 60 85 48–​50 85:4 49n63 85:11 49 95:1 45n48 105 49–​50 112 39