Patronate and Patronage in Early and Classical Islam 9789047416524, 904741652X

Citation preview

PATRONATE AND PATRONAGE IN EARLY AND CLASSICAL ISLAM

ISLAMIC HISTORY AND CIVILIZATION STUDIES AND TEXTS edited by WADAD KADI AND

ROTRAUD WIELANDT

VOLUME 61

PATRONATE AND PATRONAGE IN EARLY AND CLASSICAL ISLAM EDITED BY

MONIQUE BERNARDS AND JOHN NAWAS

BRILL LEIDEN • BOSTON 2005

This book is printed on acid-free paper.

Library of Congress Cataloging-in-Publication Data A C.I.P. record for this book is available from the Library of Congress.

ISSN 0929-2403 ISBN 90 04 14480 3 © Copyright 2005 by Koninklijke Brill NV, Leiden, The Netherlands Koninklijke Brill NV incorporates the imprints Brill Academic Publishers, Martinus Nijhoff Publishers and VSP. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Brill provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910 Danvers MA 01923, USA. Fees are subject to change. printed in the netherlands

Josef van Ess

CONTENTS

Introduction ................................................................................ Monique Bernards and John Nawas

ix

Alliances in Islam ...................................................................... Ella Landau-Tasseron

1

Were the Jewish tribes in Arabia clients of Arab tribes? ...... Michael Lecker

50

The origin and development of the Islamic patronate .......... Ulrike Mitter

70

Patronate in Shì'ite law ............................................................ 134 Robert Gleave Mawàlì and the Prophet’s family: An early Shì'ite view ........ 167 Patricia Crone Mawàlì and muwalladùn in al-Andalus ...................................... 195 Maribel Fierro Conversion-based patronage and onomastic evidence in early Islam .......................................................................... 246 Richard W. Bulliet The economic status of the mawàlì .......................................... 263 Jamal Juda Mawàlì in the composition of al-Ma"mùn’s army: A non-Arab takeover? ............................................................ 278 Amikam Elad Mawàlì and music ...................................................................... 326 Hilary Kilpatrick

viii

contents

Mawàlì and Arabic poetry: Some observations ........................ 349 Geert Jan van Gelder The mawàlì and early Arabic historiography .......................... 370 Lawrence I. Conrad The contribution of mawàlì to the Arabic linguistic tradition .................................................................................. 426 Monique Bernards A profile of the mawàlì 'ulamà" .................................................. 454 John Nawas List of Contributors .................................................................... 481 Index ............................................................................................ 485

INTRODUCTION Monique Bernards and John Nawas

This book offers the reader two key features. It deals with the patronate/patronage system of early and classical Islam (seventh up to tenth centuries CE) from many perspectives and provides at the same time an overview of current methodologies used in Islamic Studies. This monograph uniquely presents as much as we currently know about the historical phenomenon of mawàlì (sing. mawlà), usually translated as “non-Arab Muslims” (but see below), during the first four centuries of Islam and the patronate and patronage systems associated with them, analyzed from fourteen different angles. In what follows, the meaning and scope of the difficult term “mawlà” are outlined. The themes covered in the book and the various methodologies are then discussed and thereafter a more detailed presentation of the contents and main findings of the fourteen chapters are given. A definition, as usual, is a first step toward understanding a difficult concept or word. Webster’s Unabridged Third New International Dictionary includes a definition for the word mawlà. Maula n, pl mawali [Ar. mawla (pl. mawàlì)]: a recent convert to Islam; esp: a non-Arab convert extended the status of a protected client by one of the Arab peoples. The term refers, then, to converts, non-Arabs, clients and their relationship vis-à-vis Arabs. After the initial surprise of encountering a technical term used in our field in a standard English dictionary, the Islamicist who has dealt with the term mawlà is disappointed because this definition does not cover the complexities surrounding the concept of mawlà in early and classical Islamic society, as the reader of this book will find out. It is not difficult to imagine how the social category of mawàlì came into being. In a tribal society such as the one that existed prior to Islam, an individual’s social role was defined within the context of his tribe so that the tribal unit was responsible for its members. The new religion of Islam, on the other hand, conveyed a

x

monique bernards and john nawas

message of universal equality for all believers irrespective of their ethnic background. Islam was revealed over the course of a few decades and society did not change overnight. A transitional period was called for in which non-Arabs who had converted to Islam affiliated themselves to an Arab tribe in order to fit into the society of the time. But now questions start to appear for which there are no simple answers. One of the themes dealt with in this book is whether this system of affiliation existed in pre-Islamic Arab society and was adopted by Islam for the new purpose of integrating non-Arab converts or whether it was an Islamic innovation. Arab Muslim society necessitated that newcomers entered the confines of traditional Arab societal bonds and as the Islamic Community expanded, becoming geographically more diverse, it was ultimately transformed into a society in which religious alliance rather than ethnicity had become the common denominator. We do not know when the phenomenon of being a mawlà disappeared (one article deals directly with this issue) but we do know that that took place some time in the course of the first four centuries of Islam. Additionally, we do not know whether the term “mawlà” had the same meaning throughout this period and whether it was used in the same manner in all parts of the Islamic empire. Once the patronate—the legal institution—was established between the two parties, the client could make use of the patron’s tribal affiliation and the patron’s patronage, i.e., the social implications of the bond between client and patron. In return for the patron’s tribal affiliation the client would, for instance, stipulate that his inheritance goes to his patron and thus a neat give-and-take situation could come into being. Linguistic matters further compound this situation, however. “Mawlà” in Arabic means both patron and client. If the sources wanted to make explicit which side of the relationship was in question, the author could add that the individual was mawlà min fawq—mawlà from above—, that is the patron, or a mawlà min ta˙t—a mawla from below, i.e., the client (alternatives include al-mawlà l-a'là for the patron and al-mawlà l-asfal for the client). Moreover, were all non-Arabs in early and classical Islam clients? This would imply synonymy and justifies aggregating nonArabs and mawàlì, as is sometimes done by modern researchers (including some contributors to this volume). Though not included in Webster’s definition, Islamicists know that manumitted slaves (sometimes or always?) became the mawàlì of their former masters. Here,

introduction

xi

too, some modern students equate freed slaves with mawàlì, lumping the members of these two categories together as well. In other words, were all non-Arabs mawàlì or the other way around, or were all freed slaves mawàlì or vice versa, or do modern studies simply suffer from analytical confusion for not making a distinction between mawàlì and non-Arabs? We simply do not have direct answers to these and similar questions. The articles in this book use at least one of these types or combinations of two or more of these types that offer us a wealth of insight into the important social phenomenon of walà" in early and classical Islam. Answers to the above questions are not found in one article; only in their entirety do the contributions provide partial answers and indicate where further research might help us to cope more adequately with these issues. The fourteen chapters of this book study the mawàlì from several angles. The first three contributions (Landau-Tasseron, Lecker and Mitter) deal with the earliest period and highlight continuities and/or changes in forms of alliance before and after the rise of Islam. The remaining eleven articles deal with later aspects of patronate and patronage. Two studies deal with walà" in Shì'ism (Gleave and Crone) and two discuss it in distant Islamic lands, the Eastern provinces (Bulliet) and al-Andalus (Fierro). The socioeconomic position of mawàlì is dealt with by Juda, while Elad studies the ethnic composition of the army in the early Abbasid period. The contributions of mawàlì to music (Kilpatrick), poetry (van Gelder), historiography (Conrad), Arabic grammar (Bernards) and to the Islamic religious sciences (Nawas) are then analyzed. The methodologies employed are as varied as these aspects: the traditional approach to history (sometimes including textual analysis), or the social science approach (including the application of social theory or of quantitative methods). The contributions of Landau-Tasseron, Lecker and Mitter deal with the origins of the institution of patronate. Landau-Tasseron canvassed many ˙adìths and combed through many historical and biographical sources in order to obtain as much information as possible about alliances before Islam and shortly after its emergence. An understanding of alliances amongst groups in this period helps situate and categorize the practices of walà". However, Landau-Tasseron makes it clear that there is great confusion from very early on about what “ally” (˙alìf ) and “mawlà” precisely meant. Lecker scrutinizes, through a re-reading of the primary sources, the position of the Jewish tribes

xii

monique bernards and john nawas

on the eve of Islam, arguing that, contrary to some views expressed in the secondary literature, they were not clients of Arab tribes. Mitter focuses on the regulations of two forms of patronate, the servile and the contractual, and shows how the rules of these two forms were regulated in legal ˙adìth. She uses information derived from the chain of transmission (isnàd ) and the text of the ˙adìth (matn) to date the institution of walà", concluding that the transfer of patronate might very well go back to before 58 AH/678 CE. Gleave’s article highlights the specificity which early Twelver Shì'ism gave to the term “walà".” His article provides insights into how Shì'ite law argued the issue and how this led to a particular coloring of patronage in Shì'ism making it quite different from the Sunnite interpretation of it (compare especially Mitter’s contribution). This involves the issue of whether it emerged from gratuitous manumission alone or from contract and matters of inheritance, liability and continuation of the bond after the death of both patron and client. Crone’s contribution demonstrates a Shì'ite retrojection, by way of a forged letter attributed to the caliph 'Umar (reg. 13–23 AH/634–644 CE) to his governor of Iraq, Ziyàd b. Abìhi (d. 53 AH/673 CE), on the caliph’s hostility toward the mawàlì and shows how a “natural” affinity was assumed between mawàlì and Shì'ism—a view once espoused by some Orientalists as well. Fierro addresses the issue of an “Andalusian uniqueness,” in comparison with the central Islamic lands, regarding the percentage of mawàlì encountered in biographical sources on al-Andalus: there it is far lower than in other areas of Islamic lands. The characteristics of ties of patronage in al-Andalus are studied and the additional phenomenon of muwalladùn (converts to Islam without patronage) is then analyzed. Bulliet deals with the issue of trying to discover when the institution of patronage became obsolete before it disappeared altogether. Using two different theoretical modes of innovation diffusion and testing them against onomastic information about the eastern parts of the empire, Bulliet concludes that, at least conversion based patronate must have become obsolete well before the advent of the Abbasids in 132 AH/750 CE. Juda examines the economic position of the mawàlì including that in the military, while Elad focuses on the composition of al-Ma"mùn’s (reg. 198–218 AH/813–833 CE) army. Drawing from a large range of sources, Juda provides an overview of positions held by the mawàlì, arguing that at times the mawàlì were economically more prosperous than Arabs. Elad used historical sources and traditions to scru-

introduction

xiii

tinize the widely held claim that there was considerable “Iranian” (i.e., non-Arab) influence after the Abbasids came to power. He shows that there is very little evidence in the sources about mawàlì units in the early Abbasid army and that one has indeed to conclude the opposite, namely that the Iranian mawàlì only gradually came to positions of influence within the army. Kilpatrick, van Gelder and Conrad offer us broader cultural perspectives on the mawàlì. Kilpatrick concentrates on the role of mawàlì in the development of music and highlights their contribution during the Umayyad and early Abbasid periods. Van Gelder starts by observing that, in traditional categorization schemes current in the classical period, no specific category of mawàlì poets existed. He looks for similarities and/or differences between Arab and non-Arab poets of Arabic but finds very little common ground either way. This leads him to conclude that literary patronage (e.g., payment for poetry) had very little to do with any ethnic patronage between Arabs and mawàlì. The role played by mawàlì in the development of ArabIslamic historiography is scrutinized by Conrad. He focuses on the merger of the literary tradition of Classical and Late Antiquity and shows how mawàlì may have helped weld a new tradition in its own right while maintaining cultural continuity. The last two contributions both use representative samples from biographical dictionaries to study the mawàlì. Bernards deals with a major issue in the history of Arabic grammar, namely to what extent non-Arabs were active in this field of inquiry and to what extent any non-Arab influence on its development can be verified. The underlying idea is that if one can establish a dominance of nonArabs this might provide direct evidence for foreign influences. Her findings show that there were no significant differences between Arabs and non-Arabs who were active in grammar on various counts. Nawas closes the volume by studying the social group of mawàlì 'ulamà". His findings demonstrate that the new “social group” of 'ulamà" constituted a real success story in terms of integration since almost no differences between Arabs and mawàlì 'ulamà" existed on a broad range of variables.

xiv

monique bernards and john nawas Acknowledgments

The articles contained in this book are the outcome of a conference held in the village of Berg en Dal, the Netherlands; Harald Motzki was instrumental in helping us organize the conference. He offered us both personal and institutional, moral and material support by putting at our disposal the services of the Department of Arabic and Islam at what was then the Catholic University Nijmegen (renamed Radboud University Nijmegen in 2004). Monique Bernards acquired further funding for the conference. The papers were first distributed among the participants and then they were discussed in detail and further openly debated. Josef van Ess (Tübingen) participated in the conference and provided the conference with an insightful summary of the main issues. Anja Welkenhuysen-Gybels (Arabic and Islamic Studies, Catholic University Leuven) provided excellent research assistance. Finally, our thanks go to the anonymous reader of the manuscript and to Wadad Kadi for a number of helpful suggestions. Antwerp, Belgium

ALLIANCES IN ISLAM* Ella Landau-Tasseron

Alliances between groups existed in pre-Islamic Arabian society as well as in the society that emerged after the advent of Islam. This article studies Islamic attitudes towards this institution and the changes it experienced once Islam became established. Attention is also given to the confusion between the terms ˙alìf (“guest ally,” explained below) and mawlà (client). The article argues that the institutions (alliance and clientage) and the terms that refer to them were partly converging in pre-Islamic times. However, clientage did not replace alliances in Islamic times. Pre-Islamic Arabian society was constituted of real or fictitious descent groups, that is, the bulk of each group considered themselves as having descended from a common male ancestor.1 Alliances between descent groups were common. Certain alliances amounted to treaties of cooperation for limited, clearly defined purposes, for example, specific military enterprises or consolidation vis-à-vis political rivals. The parties to such alliances never merged with one another nor did any of them use the nisba (adjective of relationship in an Arabic * In the research for this article I mainly relied on the digital libraries issued on cds in 1999 by al-Turàth, the Center for Computer Research in Amman: Al-Maktaba al-alfiyya, Maktabat al-ta"rìkh wa-l-˙a∂àra, Maktabat al-fiqh, Maktabat al-adab al-'Arabì and Maktabat al-sìra al-nabawiyya. I thank my colleague, Avraham Hakim, for his help with this technology. These libraries cover thousands of volumes, often in editions that were not available to me. This means that, more often than not, I could not check the references in the hardcopy books. In spite of this shortcoming and the textual errors in the cds, I still consider these libraries indispensable. 1 All units were descent groups, regardless how far removed the ancestor was. This means that smaller descent groups were part of larger ones. For example, the offspring of an ancestor three generations removed formed part of the offspring of the ancestor four (or any higher number of ) generations removed, in the same male line. It should be noted that not every link in any given male line became the focal point of a descent group; on the other hand, not every claimed ancestry was genuine. Groups that were formed in ways other than segmentation are also descent groups, albeit alleged, because they created for themselves common genealogies. Students of Islam commonly refer to the larger entities (such as Quraysh, Tamìm, Thaqìf, Asad, etc.) as tribes. There is no fixed term for the smaller entities. I shall follow the common usage and apply the term “section” to the constituent parts of tribes. The term descent group will be used as well, in the sense just defined.

ella landau-tasseron

2

name) of another. A second type of alliances consisted of agreements between descent groups upon general cooperation. Such alliances sometimes led to a merging of the groups involved, although, as far as I am aware, merging was never explicitly agreed upon. A third type of alliances was the attachment of people to descent groups other than their original ones. This type entailed mutual inheritance and shared liability, the latter including joint action in matters of blood-revenge and blood-money, and common accountability for offenses. As a rule, it was small descent groups that were accepted as allies by larger ones. More often than not, these smaller groups came to live with their allies, and also added the latter’s nisba to their own. I therefore call this kind of attachment “hosting alliances,” the smaller groups I call “guest allies,” and the larger groups “host allies.”2 As a rule, the sources do not explicitly make the distinction between these types of agreements. They do, however, distinguish between alliances that are tolerated, even commended, by Islam, and others that contradict Islamic principles and are therefore illegal. It would seem that the Islamic distinction just mentioned reveals an unequivocal attitude towards alliances, that is, Islam accepts the institution of alliance as such, but rejects certain types thereof. Notwithstanding, diverse practices and opinions about the subject coexisted in the Muslim community.

The Prophet’s words and conduct The Prophet is reported to have said “there must be no alliance in Islam, and (or: but) you should adhere to the pre-Islamic alliances.” Another version of the latter half of this ˙adìth is, “and all pre-Islamic alliances are bolstered by Islam.”3 The two parts of this ˙adìth are 2

See Landau-Tasseron, “Alliances among the Arabs,” Al-Qantara 26 (2005), 141–73. Henceforth I shall refer to this ˙adìth as “the alliance ˙adìth.” See al-Wàqidì, Kitàb al-maghàzì, ed. Marsden Jones, 3 vols. (London: Oxford University Press, 1966), II, 867–68; 'Abd al-Razzàq, Al-Mußannaf, ed. Óabìb al-Ra˙màn al-A'Ωmì, 11 vols. (Beirut: al-Maktab al-Islàmì, 1403/1983), X, 305–307; Abù Dà"ùd, Sunan Abì Dà"ùd, ed. Mu˙ammad Mu˙yì al-Dìn 'Abd al-Óamìd, 4 vols. (n.p.: Dàr al-Fikr, n.d.), III, 129; al-Nasà"ì, Al-Sunan al-kubrà, ed. 'Abd al-Ghaffàr Sulaymàn al-Bandàrì and Sayyid Kisrawì Óasan, 6 vols. (Beirut: Dàr al-Kutub al-'Ilmiyya, 1411/1991), IV, 90; al-Dàrimì, Sunan al-Dàrimì (Beirut: Dàr al-Kitàb al-'Arabì, 1407/1987), II, 316; Ibn Óibbàn, Ía˙ì˙ Ibn Óibbàn, ed. Shu'ayb al-Arna"ù†, 18 vols. (Beirut: Mu"assasat al-Risàla, 1414/1993), X, 211–15; al-ˇabarì, Tahdhìb al-àthàr, al-juz" al-mafqùd, ed. 3

alliances in islam

3

sometimes recorded independently of each other, indeed they may have originated autonomously.4 When separated, these two parts vindicate conflicting arguments. The first entirely rejects the institution of alliances, whereas the second allows pre-Islamic alliances to continue and even reinforces them. As a rule, Muslim scholars do not specify the type of alliances that they discuss in connection with this ˙adìth. Yet when they cite it (in full or only the first part) in refutation of inheritance by alliance, it is plainly hosting alliances that they have in mind. When they object to alliances that infringe upon Muslim unity, it is clearly military and political alliances that they have in mind. When this alliance ˙adìth is adduced as a whole, its second part qualifies the first, meaning that no new alliances are to be inaugurated but pre-Islamic ones must be kept (one version) or are bolstered (the other version). However, certain scholars sensed a contradiction in this precept. As a way of solving it the legal Óanafì scholar alJaßßàß (d. 370/981) offers a rather absurd interpretation of the text, that is, “[there should be no alliances in Islam], and Islam bolsters

'Alì Ri∂à b. 'Abdallàh b. 'Alì Ri∂à (Beirut: Dàr al-Ma"mùn li-l-Turàth, 1995), 19, 22, 25; idem, Tafsìr al-ˇabarì, 30 vols. (Beirut: Dàr al-Fikr, 1405/1985), V, 55–56; al-Jaßßàß, A˙kàm al-Qur"àn, ed. Mu˙ammad al-Íàdiq Qam˙àwì, 5 vols. (Beirut: Dàr I˙yà" al-Turàth al-'Arabì, 1405/1985), III, 197, 283–85; Ibn Kathìr, Tafsìr (Beirut: Dàr al-Fikr, 1401/1981), I, 490–91; II, 585; al-Óanafì, Mu'taßar al-mukhtaßar, 2 vols. (Beirut/Cairo: 'Àlam al-Kutub and Maktabat al-Mutanabbì, n.d.), II, 109; Ibn alJawzì, Zàd al-masìr fì 'ilm al-tafsìr, 8 vols. (Beirut: al-Maktab al-Islàmì, 1404/1984), II, 72–73; al-Haythamì, Majma' al-zawà"id wa-manba' al-fawà"id, 10 vols. (Cairo/Beirut: Dàr al-Rayyàn li-l-Turàth and Dàr al-Kitàb al-'Arabì, 1407/1987), VIII, 173; idem, Mawàrid al-Ωam"àn ilà zawà"id Ibn Óibbàn, ed. Mu˙ammad 'Abd al-Razzàq Óamza (Beirut: Dàr al-Kutub al-'Ilmiyya, n.d.), I, 504; Ibn Óajar al-'Asqalànì, Fat˙ al-bàrì shar˙ Ía˙ì˙ al-Bukhàrì, ed. Mu˙ammad Fu"àd 'Abd al-Bàqì and Mu˙ibb al-Dìn alKha†ìb, 13 vols. (Beirut: Dàr al-Ma'rifa, 1379/1959), IV, 473; X, 502; idem, AlIßàba fì tamyìz al-ßa˙àba, 4 vols. (Cairo: Ma†ba'at al-Sa'àda, 1328/1910), I, 509; al-'AΩìmàbàdì, 'Awn al-ma'bùd, 10 vols. (Beirut: Dàr al-Kutub al-'Ilmiyya, 1415/1995), VIII, 100–102. The order of the two parts is reversed (within a conglomerate of ˙adìths recorded as one piece under one isnàd ) in Ibn Óanbal, Musnad A˙mad, 6 vols. (Cairo: Mu"assasat Qur†uba, n.d.), II, 180; al-Bayhaqì, Al-Sunan al-kubrà, ed. Mu˙ammad 'Abd al-Qàdir 'A†à, 10 vols. (Mecca: Maktabat Dàr al-Bàz, 1414/1994), VIII, 29; al-ˇabarì, Tahdhìb al-àthàr, 24, 28–29. In al-Muttaqì al-Hindì, Kanz al'ummàl fì sunan al-aqwàl wa-l-af 'àl, ed. Íafwat al-Saqà, Bakrì Óayàtì, Nadìm Mar'ashlì and Usàma Mar'ashlì, 18 vols. (Beirut: Mu"assasat al-Risàla, 1981–86), V, 845, both versions of the second part are included. See also below, “Alliances in Islamic times.” 4 Ibn ManΩùr, Lisàn al-'Arab, 15 vols. (Beirut: Dàr Íàdir, n.d.), IX, 54, explicitly speaks of ˙adìthàni, two traditions but it is impossible to prove that this is the case. It is also impossible to separate the two parts for the purpose of isnàd analysis.

4

ella landau-tasseron

the negation of alliances.”5 Other scholars resolve the contradiction by one of the two devices, either abrogation or specification. It is sometimes argued that the injunction “there must be no alliance” was abrogated by “you should adhere to the pre-Islamic alliances” or that the prohibition only applies to specific alliances that contradict Islamic rules and principles, such as the rules of inheritance and the principle of unity among all Muslims.6 It should be noted that the abrogation theory is contradicted by the fact that, more often than not, the alliance ˙adìth is recorded in its full version; the Prophet would not have uttered in one breath two contradictory rules, least of all during the Farewell Pilgrimage.7 Conflicting attitudes towards alliances are expressed in terms of Qur"ànic exegesis as well. Certain verses are interpreted as endorsing alliances, for example, “be faithful to covenants” (Q.5:1) and

5 Al-Jaßßàß, A˙kàm, III, 147 ( fa-innahu yu˙tamal anna l-islàm man zàda shiddatan wataghlìΩan fì l-man' minhu wa-ib†àlihi fa-ka-annahu qàla idhà lam yajuzi l-˙ilf fì l-islàm . . . fa˙ilfu l-jàhiliyya ab'adu min dhàlika). As a Óanafì, however, al-Jaßßàß does not reject hosting alliances, as is evident from his discussion ibid., 145–48, see also below, section “Islamic objections.” 6 These possibilities are offered by al-Jaßßàß as well, A˙kàm, III, 283–84. See further Ibn ManΩùr, Lisàn, IX, 54; al-Mubàrakfùrì, Tu˙fat al-A˙wadhì bi-shar˙ jàmi' alTirmidhì, 10 vols. (Beirut: Dàr al-Kutub al-'Ilmiyya, 1410/1990), V, 174; al-'AΩìmàbàdì, 'Awn al-ma'bùd, VIII, 100, 101; Ibn Óajar al-'Asqalànì, Fat˙ al-bàrì, X, 502; alMunàwì, Fay∂ al-qadìr shar˙ al-jàmi' al-ßaghìr min a˙àdìth al-bashìr al-nadhìr, 6 vols. (Cairo: al-Maktaba l-Tijàriyya l-Kubrà, 1356/1937), III, 80; al-Haythamì, Majma' al-zawà"id, VIII, 29, 172–73; al-ˇabarì, Tahdhìb al-àthàr, 29–30; idem, Tafsìr, V, 56; Ibn Kathìr, Tafsìr, I, 491; al-Qur†ubì, Tafsìr al-Qur†ubì, ed. A˙mad 'Abd al-'Alìm al-Bardùnì, 20 vols. (Cairo: Dàr al-Sha'b, 1372/1952), VI, 33. Cf. al-Óanafì, Mu'taßar al-mukhtaßar, II, 294 where “there must be no alliance” is said to be the abrogator of the license to alliance, not the other way around. This order of abrogation is also implied by the arguments of al-Jaßßàß, A˙kàm, III, 283ult.–84, and al-ˇabarì, Tahdhìb al-àthàr, 25–26. Juda, Die Sozialen und Wirtschaftlichen Aspekte der Mawàlì in Frühislamischer Zeit (Ràfàt, 1983), 60–61 records the ˙adìth but ignores the contradiction between its two parts and the discussions precipitated by it. Goldziher, Muslim Studies, ed. by S.M. Stern, trans. C.R. Barber and S.M. Stern (London: George Allen and Unwin Ltd., 1967), I, 70, adduces the full version of the ˙adìth, interpreting the objection to alliances as an expression of the ideal of Muslim unity. He thus ignores the second part of the ˙adìth which endorses alliances. See also Schmucker, Untersuchungen zu Einigen Wichtigen Bodenrachtlichen Konsequenzen der Islamischen Eroberungsbewegung (Bonn: Selbstverlag des Orientalischen Seminars der Universität Bonn, 1972), 9–10. For a detailed discussion of the objections to alliances, see below, section “Islamic objections.” 7 Rabì'a b. 'Abbàd reports that he heard the Prophet utter this ˙adìth during the Farewell Pilgrimage, al-Wàqidì, Maghàzì, II, 867–68; Ibn Óajar al-'Asqalànì, Ißàba (ed. Cairo), I, 509.

alliances in islam

5

“cooperate for piety and the fear of God” (Q.5:2).8 Yet even a scholar who accepts such an interpretation of these verses may express reservations about alliances. Ibn Kathìr (d. 774/1373), having applied the verse “be faithful to covenants” to pre-Islamic alliances, adds that in fact all alliances are redundant because the bond of Islam supersedes them.9 Scholars would, occasionally, use a different tone, barely tolerating alliances at all and explicitly setting the conditions that, if they exist, they must accord with Islamic principles or serve the purpose of jihàd.10 In sum, whatever the interpretation of the alliance ˙adìth and the verses discussed here, no unequivocal allowance is made for inaugurating new alliances in Islamic times,11 and the congruence of the existing alliances with Islamic principles has always been considered as obligatory. As a model of a meritorious alliance, the pre-Islamic ˙ilf al-Fu∂ùl is adduced, wherein various sections of Quraysh cooperated in order to protect the rights of outsiders who came to trade in Mecca.12

8 Al-Jaßßàß, A˙kàm, III, 283–85; Ibn al-Jawzì, Zàd al-masìr, IV, 484; al-Shawkànì, Fat˙ al-qadìr al-jàmi' bayna fannay al-riwàya wa-l-diràya min 'ilm al-tafsìr, 5 vols. (Beirut: Dàr al-Fikr, n.d.), I, 462; II,7; al-Mubàrakfùrì, Tu˙fat al-a˙wadhì, V, 173. See also below, note 82. 9 Ibn Kathìr, Tafsìr, II, 584–85; cf. al-Jaßßàß, A˙kàm, III, 283–84, who argues the same, but apparently he is referring to political alliances because he accepts inheritance by hosting alliance, ibid., III, 145–48. 10 Al-Zara'ì, Óàshiyat Ibn al-Qayyim, 14 vols. (Beirut: Dàr al-Kutub al-'Ilmiyya, 1415/1995), VIII, 101. 11 The injunction “Do not inaugurate an alliance of mutual inheritance” (alMubàrakfùrì, Tu˙fa, V, 174), for example, implies that it is permissible to conclude alliances that do not entail mutual inheritance. 12 Al-ˇabarì, Tahdhìb al-àthàr, 20–22; al-Jaßßàß, A˙kàm, III, 284; al-Qur†ubì, Tafsìr, VI, 33; X, 169; al-Nawawì, Tahdhìb al-asmà" (Beirut: Dàr al-Fikr, 1416/1996), III, 181; al-Fàkihì, Akhbàr Makka, ed. 'Abd al-Malik Duhaysh, 6 vols. (Beirut: Dàr Kha∂r, 1414/1994), V, 190–96; al-Munàwì, Fay∂ al-qadìr, IV, 164–65. The political alliance of the Mu†ayyabùn (concluded among certain sections of Quraysh) is sometimes adduced instead of the Fu∂ùl, al-ˇabarì, ibid., 32; al-Jaßßàß, ibid.; Ibn Kathìr, Tafsìr, I, 491; Ibn Óibbàn, Ía˙ì˙, X, 216; al-Bayhaqì, Al-Sunan al-kubrà, VI, 367; alHaythamì, Majma' al-zawà"id, VIII, 172; idem, Mawàrid al-Ωam"àn, I, 504; al-Óanafì, Mu'taßar al-mukhtaßar, II, 375–76; Ibn Óajar al-'Asqalànì, Fat˙ al-bàrì, X, 502. Juda, Aspekte, 61, says that according to Ibn Is˙àq, “the Fu∂ùl alliance was the only one confirmed by the Prophet,” but this is not what the text says, see Ibn Hishàm, AlSìra al-nabawiyya, ed. Mu߆afà al-Saqà, 'Abd al-ÓafìΩ Shalabì and Ibràhìm al-Abyàrì, 4 vols. (Cairo: Maktabat Mu߆afà l-Bànì wa-Awlàduhu, 1355/1936), I, 141–42 (“the Prophet, peace be upon him, said: at the house of 'Abdallàh b. Jud'àn I witnessed the conclusion of an alliance, which I would not have exchanged for the most precious camels. Had I been called upon to participate in it in Islamic times, I would have agreed”).

ella landau-tasseron

6

In addition to the alliance ˙adìth there are reports that the Prophet himself concluded alliances and acted according to principles laid by them, suggesting to his followers to do the same. Such reports would seem to be serving the purpose of supporting the inauguration of new alliances in Islam. But, curiously enough, most of these reports are independent of the discussions of the subject, which lends at least some of them a measure of credibility. Even when such a report is recorded in conjunction with the ˙adìth “you should adhere to the pre-Islamic alliances,” it seems that the latter was adduced in order to justify the Prophet’s conduct and not that the narrative was concocted in order to vindicate the ˙adìth.13 Moreover, in spite of the widespread opposition to new alliances and to certain aspects of the pre-Islamic ones (see below), I did not come across inverse reports, that is, narratives portraying the Prophet as acting against this institution. This fact, too, lends credibility to the reports listed below. It also means that the consensus among Muslim scholars concerning the ban on newly inaugurated alliances is not based on the Prophet’s conduct, be it authentic or alleged. The reports associating the Prophet with alliances are as follows. 1) During the 'Aqaba meeting, in which the agreement between the Prophet and the future Anßàr was discussed, the Prophet used a pre-Islamic formula, “we are equal in matters of blood-revenge and unavenged blood,” and “we are together in peace and in war” (al-dam al-dam wa-l-hadm al-hadm, usàlimu man sàlamtum wau˙àribu man ˙àrabtum).14 2) In the early Medinan period Mu˙ammad concluded an alliance with a (pagan) chief of the descent group called Ashja', guaranteeing mutual assistance. Pre-Islamic formulae were used; the alliance was to last “as long as the mountain U˙ud stays in place, as long as the sea is capable of wetting a piece of wool.”15

13

E.g., al-Wàqidì, Maghàzì, II, 782, see below note 18. Ibn Hishàm, Sìra, II, 85; al-ˇabarànì, Al-Mu'jam al-kabìr, ed. Óamdì b. 'Abd al-Majìd al-Salafì, 20 vols. (Mosul: Maktabat al-'Ulùm wa-l-Óikam, 1404/1984), XIX, 44. Juda, Aspekte, 52–55 emphasizes that the Prophet resorted to the preIslamic political form of alliance, in the 'Aqaba as well as in the establishment of the brotherhood (he ignores the other cases listed here; on the brotherhood see below). See also below, section ˙alìf, 'adìd and mawlà. On the formula al-dam al-dam etc., see Landau-Tasseron, “Alliances among the Arabs,” 155–58. 15 Ibn Sa'd, Al-ˇabaqàt al-kubrà, 8 vols. (Beirut: Dàr Íàdir, n.d.), I, 274. 14

alliances in islam

7

3) A certain individual named Îumayra came to ally himself with the Prophet in the pre-Islamic manner. The Prophet referred him to 'Alì and changed the Jàhilì formula by replacing the name of the mountain mentioned in it. It is explicitly stated that the Prophet acted thus in order to differentiate between Jàhiliyya and Islam.16 4) Mu˙ammad offered one of his Companions, a Bedouin named Jàriya b. Óumayl, a choice between the status of a Muhàjir and that of an Anßàrì. The man chose the status of an Anßàrì, whereupon he concluded an alliance with the Anßàr.17 5) In the year 6/628 the Prophet and Quraysh concluded the treaty of Óudaybiyya. Within the framework of this treaty, [parts of ] the tribe called Khuzà'a allied themselves with the Prophet on the basis of the ancient alliance that they had had with the Prophet’s ancestor, Hàshim (or 'Abd al-Mu††alib). When these Khuzà'ìs were assaulted by allies of Quraysh, they appealed to Mu˙ammad, invoking that alliance. They had no other basis to appeal to Mu˙ammad because they had not yet been converted to Islam at that time. The Prophet’s response to that appeal resulted in the conquest of Mecca.18 6) Mu˙ammad incorporated Banù Ju'ayl, a section of the tribe called Baliyy, into his own descent group, Banù 'Abd Manàf. Apparently, they had been allies of Banù 'Abd Manàf since ancient times.19 7) When people of the tribe Thaqìf captured two Muslims, Mu˙ammad seized an 'Uqaylì who happened to have been allied to Thaqìf. When the captive protested that he was innocent of any offense, the Prophet explained that he had been seized for the

16 Ibn Óajar al-'Asqalànì, Al-Ißàba fì tamyìz al-ßa˙àba, ed. 'Alì Mu˙ammad alBijàwì, 8 vols. (Beirut: Dàr al-Jìl, 1412/1992), III, 494; Ibn ManΩùr, Lisàn, IX, 198. 17 Abù Nu'aym al-Ißbahànì, Óilyat al-awliyà" wa-†abaqàt al-aßfiyà", 10 vols. (Beirut: Dàr al-Kitàb al-'Arabì, 1405/1985), I, 354. 18 Al-Wàqidì, Maghàzì, II, 782–90 (the alliance ˙adìth is adduced here in its full version); al-ˇabarì, Ta"rìkh al-rusul wa-l-mulùk, 5 vols. (Beirut: Dàr al-Kutub al'Ilmiyya, 1407/1987), I, 503; II, 158; Ibn Abì Shayba, Al-Mußannaf, ed. Kamàl Yùsuf al-Óùt, 7 vols. (Riyà∂: Maktabat al-Rushd, 1409/1989), VII, 403; al-ˇa˙àwì, Shar˙ ma'ànì al-àthàr, 4 vols. (Beirut: Dàr al-Kutub al-'Ilmiyya, 1399/1979), III, 291, 312–16; al-Haythamì, Majma' al-zawà"id, VIII, 172; al-Qur†ubì, Tafsìr, VIII, 85; Ibn Óajar Ißàba (ed. Beirut), IV, 631. For the ancient alliance, see al-ˇabarì, ibid.; Ibn Sa'd, ˇabaqàt, I, 85. 19 Ibn Sa'd, ˇabaqàt, I, 270–71; Yàqùt al-Rùmì, Mu'jam al-buldàn, 5 vols. (Beirut: Dàr al-Fikr, n.d.), IV, 88.

8

ella landau-tasseron

crime of his allies.20 This act of the Prophet accorded with the pre-Islamic principle that partners to a hosting alliance shared liability. 8) As late as the year 9 AH the Prophet had recourse to the institution of alliance. He suggested to two of his followers, fugitives from ˇà"if, to ally themselves with their maternal uncle Abù Sufyàn. The two acted upon the Prophet’s advice.21 The fact that the Prophet accepted pre-Islamic institutions, principles and ceremonies is usually not commented upon. An explanation of the contradiction between the Prophet’s conduct and the alliance ˙adìth is also typically lacking. Even the explanation that alliances were necessary only when Islam was weak, and became redundant afterwards, is not meant as a reconciliation of the inconsistency between the Prophet’s words and deeds.22 Only rarely does one find a discussion such as Ibn Óazm’s (d. 456/1064), who explains that when the Prophet captured the 'Uqaylì, he was not acting according to the Jàhilì principle of shared liability (that is, he was not acting in accordance with the institution of ˙ilf ). Ibn Óazm insists that the Prophet was in fact guided by Islamic principles because the 'Uqaylì was an infidel and therefore guilty by definition and deserving of punishment. Other scholars solve the problem of the 'Uqaylì on a linguistic basis. They argue that when the Prophet said to him “I seized you for the crime of your allies,” he meant “I seized you in order to redress the crime of your allies,” that is, by exchanging captives. These explanations, however, are not meant to address the issue of alliances in Islam. They revolve around another difficulty, namely, the fact that the Prophet violated the basic reli-

20 Al-Shàfi'i, Kitàb al-umm, 8 vols. (Beirut: Dàr al-Ma'rifa, 1393/1973), II, 253–55; VII, 68; Abù Dà"ùd, Sunan, III, 239; Ibn Óanbal, Musnad, IV, 430, 433; Muslim b. al-Óajjàj, Ía˙ì˙ Muslim, ed. Mu˙ammad Fu"àd 'Abd al-Bàqì, 5 vols. (Beirut: Dàr I˙yà" al-Turàth al-'Arabì, n.d.), III, 1262; Ibn Óibbàn, Ía˙ì˙, XI, 198; al-Óumaydì, Musnad al-Óumaydì, ed. Óabìb al-Ra˙màn al-A'Ωamì, 2 vols. (Beirut/Cairo: Dàr alKutub al-'Ilmiyya and Maktabat al-Mutanabbì, n.d.), II, 365–66; al-Bayhaqì, AlSunan al-kubrà, I, 318; V, 175; VI, 320; IX, 72, X, 75; al-ˇabarànì, Al-Mu'jam al-kabìr, XVIII, 190–92; al-Munàwì, Fay∂ al-qadìr, V, 170–71; Ibn Kathìr, Al-Bidàya wa-l-nihàya, 14 vols. (Beirut: Maktabat al-Ma'àrif, n.d.), IV, 154; al-Shawkànì, Nayl al-aw†àr, 9 vols. (Beirut: Dàr al-Jìl, 1393/1973), VIII, 147. 21 Ibn Sa'd, ˇabaqàt, V, 504. See also below, section ˙alìf, 'adìd and mawlà towards the end. 22 Al-Jaßßàß, A˙kàm, III, 283–84. The explanation is meant to solve the contradiction between the two parts of the alliance ˙adìth.

alliances in islam

9

gious principle of personal responsibility, by seizing a person who had committed no crime.23 The only case where scholars deem it necessary to reconcile the inconsistency between the Prophet’s words and deeds is the mu"àkhàt (brotherhood), namely the special bond established by the Prophet between Muhàjirùn and Anßàr. Some reports about this matter refer to the Prophet’s initiative as “[he] created a brotherhood/a bond” (àkhà/allafa), others say “[he] concluded an alliance” (˙àlafa), between the Muhàjirùn and the Anßàr. The report of this event is sometimes juxtaposed with the ˙adìth “there must be no alliance in Islam,” followed by an explanation of the seeming contradiction. It is explained that brotherhood and alliance are not the same thing and that even if a narrator uses the verb ˙àlafa, he means àkhà or allafa. Therefore, there is no contradiction between the Prophet’s words (i.e., the alliance ˙adìth) and his conduct (the inauguration of the brotherhood).24 It is emphasized that the Prophet indeed opposed tribal alliances which infringe upon Muslim unity but endorsed brotherhood among Muslims as well as cooperation among them for pious purposes.25 Nonetheless, the mu"àkhàt is sometimes referred to as an alliance (˙ilf ) and is also adduced as a corroboration of the latter half of the alliance ˙adìth, “you should adhere to the pre-Islamic alliances.”26 23 Ibn Óazm, Al-Mu˙allà, ed. Lajnat I˙yà" al-Turàth al-'Arabì, 11 vols. (Beirut: Dàr al-Àfàq al-Jadìda, n.d.), XI, 59–60; al-'AΩìmàbàdì, 'Awn al-ma'bùd, IX, 103–104. See also above the references in note 20. On the problem of personal responsibility, see below, “Islamic objections.” 24 Al-Shàfi'i, Al-Umm, II, 253–55; VII, 68; al-Óumaydì, Musnad, II, 365–66; Abù Dà"ùd, Sunan, III, 239; Ibn Óanbal, Musnad, IV, 430, 433; Muslim, Ía˙ì˙, III, 1262; Ibn Óibbàn, Ía˙ì˙, XI, 198; al-Bayhaqì, Al-Sunan al-kubrà, I, 318; V, 175; VI, 320; IX, 72; X, 75; al-ˇabarànì, Al-Mu'jam al-kabìr, XVIII, 190–92; al-Munàwì, Fay∂ alqadìr, V, 170–71; Ibn Kathìr, Al-Bidàya wa-l-nihàya, IV, 154; al-Shawkànì, Nayl alaw†àr, VIII, 147. Al-ˇabarì, Tahdhìb al-àthàr, 25–26, resolves the contradiction by the device of abrogation: the brotherhood was inaugurated before the ban on alliances. See also below, “Alliances in Islamic times.” Juda, Aspekte, 54–55, ignores the differences among scholars and considers the brotherhood an alliance. 25 Al-Jaßßàß, A˙kàm, III, 283; Ibn Kathìr, Tafsìr, I, 491; II, 585; al-Óanafì, Mu'taßar al-mukhtaßar, II, 294; al-Zara'ì, Óàshiya, VIII, 102; al-Ràzì, Mukhtàr al-ßi˙à˙, ed. Ma˙mùd Khà†ir (Beirut: Maktaba Lubnàn, Nashirùn, 1415/1995), I, 63; see also Ibn Óajar al-'Asqalànì, Fat˙ al-bàrì, IV, 474; X, 502; al-'AΩìmàbàdì, 'Awn al-ma'bùd, VIII, 101–102. 26 Al-Bukhàrì, Ía˙ì˙ al-Bukhàrì, ed. Mu߆afà Dìb al-Baghà, 6 vols. (Beirut: Dàr Ibn Kathìr, 1407/1987), V, 2258; Abù Dà"ùd, Sunan, III, 129; Ibn Óanbal, Musnad, III, 111, 281; al-Bayhaqì, Al-Sunan al-kubrà, VI, 262; Ibn Óajar al-'Asqalànì, Fat˙ al-bàrì, IV, 472; X, 501; al-Haythamì, Majma' al-zawà"id, VIII, 172; see also Ibn Óazm, Al-Mu˙allà, XI, 61, and below, “Islamic objections” (Ibn Óazm on the

10

ella landau-tasseron Alliances in Islamic times

Despite the prohibition in the alliance ˙adìth, military and political alliances were inaugurated in Islamic times under various circumstances and in various times and places.27 This was not always done in complete disregard of the ˙adìth. When sections of Rabì'a (of 'Adnànì or so-called Northern origin) and Azd (of Qa˙†ànì or socalled Southern origin) forged an alliance in Baßra during the Umayyad period, they evoked ancient alliances between “Northern” and “Southern” tribes.28 Whether or not such alliances had existed, they were apparently evoked, among other things, in order to legitimize the relation between Rabì'a and Azd in accordance with the precepts of the alliance ˙adìth. Opponents to this particular alliance vindicated their position by the first half of the ˙adìth, namely, “there must be no alliance in Islam.”29 Pre-Islamic alliances, formally legal according to the pronouncement of the alliance ˙adìth in its full version, were sometimes evoked in contexts which were opposed to Islamic principles and to the interests of the community as a whole. For example, during the ridda wars Muslims gave safe conduct to apostates taken captive, on account of a pre-Islamic alliance.30 Another example is the Meccan pre-Islamic

payment of blood-money). According to Motzki (personal communication, July 2001, for which my thanks are due) the common link for the ˙adìth about the brotherhood being an alliance is 'Àßim b. Sulaymàn al-A˙wal, a Baßran scholar who fulfilled administrative functions in Kùfa and Madà"in and died between 141–43/758–60. I fail, however, to see a connection between this fact and the contents, and usage, of this ˙adìth. 27 For example, in the eleventh century, the Zaydì imàm of Ía'da in Yemen allied himself to certain tribes, in defiance of the authority of his uncle; al-Shawkànì, Al-Badr al-†àli' bi-ma˙àsin man ba'da al-qarn al-sàbi', 2 vols. (Beirut: Dàr al-Ma'rifa, n.d.), II, 156. Groups of Syrian Bedouins concluded an anti-Fà†imid alliance in the eleventh century, Ibn al-'Adìm, Bughyat al-†alab fì ta"rìkh Óalab, ed. Suhayl Zakkàr, 12 vols. (Beirut: Dàr al-Fikr, 1988), V, 2239. Contra Schmucker, Untersuchungen, 12. 28 Khalìfa b. Khayyà† al-'Ußfurì, Ta"rìkh Khalìfa b. Khayyà†, ed. Akram Îiyà" al'Umarì (Damascus/Beirut: Dàr al-Qalam and Mu"assasat al-Risàla, 1397/1977), I, 258; al-ˇabarì, Ta"rìkh, III, 370; Mu˙ammad b. Mu˙ammad 'Izz al-Dìn Ibn alAthìr, Al-Kàmil fì al-ta"rìkh, ed. Abù l-Fidà" 'Abdallàh al-Qà∂ì, 10 vols. (Beirut: Dàr al-Kutub al-'Ilmiyya, 1415/1995), III, 471–72; Ibn Óajar al-'Asqalànì, Ißàba (ed. Beirut), V, 90; al-Qalqashandì, Íub˙ al-a'shà fì ßinà'at al-inshà, ed. Yùsuf 'Alì alˇawìl, 8 vols. (Damascus: Dàr al-Fikr, 1407/1987), XIV, 270; cf. Ibn al-'Adìm, Bughyat al-†alab, IV, 2018. See also Elad, “The Southern Golan in the Early Muslim Period: The Significance of Two Newly Discovered Milestones of 'Abd al-Malik,” Der Islam 76 (1999), 53–54; Juda, Aspekte, 61. 29 Juda, Aspekte, 61. 30 Al-ˇabarì, Ta"rìkh, II, 325.

alliances in islam

11

alliance of the Fu∂ùl, which was evoked during a brawl between 'Alids and Umayyads in Medina. Members of that ancient alliance responded to the call of the 'Alids and were prepared to fight on their side.31 Although this particular alliance, the Fu∂ùl, is considered meritorious in Islamic terms, obviously its evocation under such circumstances contradicted the Islamic principle of unity among all Muslims.32 Certain hosting alliances continued from the Jàhiliyya, while others were inaugurated in Islamic times. The existence of the latter implies that the alliance ˙adìth was ignored. But, to the contrary, some of the evidence displays great concern for this very ˙adìth. The evidence is of two types, a) reports about incidents and discussions related to the topic of alliances and, b) references to Muslim allies. a) Incidents and discussions related to alliances It is reported that Banù al-'Aràba, a section of Bahz (which in turn was a section of Sulaym), killed a man of their own descent group, whereupon they fled and allied themselves to one 'A†à" b. Abì alJulayd. People from Sulaym came after Banù al-'Aràba demanding retaliation, but 'A†à" protected them and took upon himself to pay the blood-money on their behalf. Apparently he did not fulfill his undertaking because during 'Uthmàn’s reign (24–36/644–656) the Sulamìs brought the matter before the caliph, arguing that the alliance between the two parties had been concluded in Islamic times, during Mu˙ammad’s activity in Mecca, and was therefore not valid. The consequence of their argument was that the protection given Banù al-'Aràba by Abù al-Julayd was not valid and that the fugitives had to be handed over to them. This argument is obviously informed of the precepts of the full version of the ˙adìth, “there must be no alliances in Islam, but you should adhere to the pre-Islamic alliances.” After pondering the matter 'Uthmàn decided that the alliance was valid because it was pre-Islamic after all. The dividing line between Jàhiliyya and Islam was the Emigration and the alliance 31 Al-Dhahabì, Siyar a'làm al-nubalà", ed. Shu'ayb al-Arna"ù† and Mu˙ammad Na'ìm al-'Arqasùsì, 23 vols. (Beirut: Mu"assasat al-Risàla, 1413/1993), III, 276; alQur†ubì, Tafsìr, VI, 33, cf. Ibn Hishàm, Al-Sìra al-nabawiyya, ed. ˇahà 'Abd al-Ra"ùf Sa'd, 6 vols. (Beirut: Dàr al-Jìl, 1411/1991), I, 266–67. 32 This is not al-ˇabarì’s opinion, on the contrary, he concludes from this incident that it is permissible to summon up support from one’s allies (istißràkh or i'tizà") even though it is forbidden to do so from one’s descent group, Tahdhìb al-àthàr, 30–31.

12

ella landau-tasseron

had been concluded before that event. According to 'Uthmàn, the prohibition on inaugurating alliances only applied from the Emigration onwards. Different decisions about the issue of Jàhilì versus Islamic alliances are ascribed to 'Alì, 'Umar and Ibn 'Abbàs, who drew the line dividing Jàhiliyya from Islam at the revelation of sùra Li-ìlàf Quraysh, the treaty of Óudaybiyya, and the revelation of Qur"àn 4:33, respectively.33 'Alì made his decision in this matter in connection with the dispute over the Baßran chieftain (sharìf ), Rabì'a b. 'Àßim al-'Uqaylì. Both the original descent group of Rabì'a, 'Uqayl, and his allies, Ju'fì, claimed him for themselves. The point of the dispute was the validity of the hosting alliance between Rabì'a b. 'Àßim and Ju'fì, according to the time at which it was concluded.34 Thus, the issue of alliances not only attracted the attention of the Ràshidùn but also induced them to tamper with the line dividing between Jàhiliyya and Islam.35 Another case that necessitated caliphal intervention was that of an ally of the tribe Hudhayl. This tribe had renounced an ally but when that person was murdered they complained to 'Umar, demanding 33 Ibn Óazm, Al-Mu˙allà, XI, 60; Ibn Óajar al-'Asqalànì, Ißàba (ed. Beirut) V, 128; idem, Fat˙ al-bàrì IV, 474. According to Ibn Óabib, Munammaq, 261, the group is called Banù al-Ghazàla and the alliance was originally concluded with 'A†à"’s father, Abù l-Julayd. According to Lecker, Sulaym, 21, and Juda (read Abù l-Julayd for his Abù al-Jaland), Aspekte, 12–13, it was Banù l-'Aràba themselves who wished to end the alliance, but this interpretation is probably incorrect for two reasons. First, the Sulamì Bahzìs who appealed to 'Uthmàn to annul the alliance refer to Banù l-'Aràba in the third person. Secondly, the issue was the removal of protection from the offenders Banù l-'Aràba, so it cannot be this group who demanded it. Juda’s conclusion that bloodmoney had already been paid by Abù l-Julayd is not corroborated by the textual evidence. 34 Ibn Óabìb, Al-Munammaq fi akhbàr Quraysh, ed. Khùrshìd A˙mad Fàriq (Beirut: 'Àlam al-Kutub, 1405/1985), 262. Ibn Óabìb, however, interprets the incidents as an attempt on 'Alì’s part to prevent people from attaching themselves to Quraysh (although neither 'Uqayl nor Ju'fì were Qurayshì). The same is implied by Ibn Óazm, al-Mu˙allà, XI, 61. Cf. Ibn Óabìb, Mu˙abbar, 338–39, where the point of the dispute is not the alliance but the identity of Rabì'a’s father. On Rabì'a and his descendants, see Crone, Slaves on Horses. The Evolution of the Islamic Polity (Cambridge: Cambridge University Press, 1980), 106. 35 Juda, Aspekte, 58–59, mentions the Ràshidùn’s decrees as marking the end of the development from “loyalty to alliances” to “loyalty to the faith.” While naturally institutions were transformed on their way from Jàhiliyya to Islam, Juda’s interpretation in this case is contradicted by the fact that loyalty to alliances did not at all cease to exist, as this article shows (cf. Juda, Aspekte, 60, where he himself mentions the alliances which remained valid). The decrees of the caliphs were not an end of a process but responses to a specific issue, namely, which alliances were valid and which were not.

alliances in islam

13

their right for retaliation. The caliph made them swear a collective oath (qasàma) that they had not renounced the slain person, so that the alliance was still valid, and they were entitled to demand retaliation. 'Umar was also required to decide whether Shura˙bìl b. Óasana was entitled to choose an ally for himself. 'Umar decreed that Shura˙bìl’s alliance was legitimate, even though it was newly inaugurated during his own caliphate (13–23/634–644).36 Decisions about alliances were not confined to the period of the Ràshidùn. A legal problem pertinent to a certain alliance was brought before the caliph 'Umar b. 'Abd al-'Azìz (reigned 99–102/717–20). This caliph is also reported to have opposed the institution altogether, presumably because it both reflected and caused divisions inside the Muslim community.37 In later periods, legal problems related to alliances were discussed by jurists.38 The evidence adduced above indicates that the precepts of the alliance ˙adìth were followed from a very early period onward. The above-mentioned events from the period of the Ràshidùn appear to be historical precisely because the decisions of the caliphs have no basis in the conduct of the Prophet. Had someone wished to forge illustrations of the alliance ˙adìth, he would have invented incidents involving the ultimate authority, namely, the Prophet, rather than the Ràshidùn and Ibn 'Abbàs (d. 67–8/868–8). The historicity of the events, however, is no proof of the Prophetic origin of the ˙adìth upon which the caliphs are said to have based their decisions. In other words, the injunction of the alliance ˙adìth was functioning in the period of the Ràshidùn but not necessarily in the form of a ˙adìth. An isnàd-cum-matn analysis made by Harald Motzki reveals that the ˙adìth circulated in different centers of learning in the first quarter of the second century. Motzki concludes that the ˙adìth must have existed at least a generation earlier.39 On the basis of Motzki’s 36 Ibn Óazm, Al-Mu˙allà, XI, 65. Shura˙bìl’s story: 'Alì b. al-Óasan Ibn 'Asàkir, Ta"rìkh madìnat Dimashq, ed. 'Umar b. Gharàma al-'Amrawì, 80 vols. (Beirut: Dàr al-Fikr, n.d.), XXII, 471–72. 37 See Elad, “The Southern Golan,” 54 (the correct reference in al-'Iqd is IV, 436); Juda, Aspekte, 61. 38 On the legal problems (including the decree of 'Umar b. 'Abd al-'Azìz), see below, section “Islamic objections.” 39 Personal communications, July 2001. I thank Harald Motzki for this contribution. Motzki identified as common links the following: Zuhrì and Sa'd b. Ibràhìm in Medina, 'Amr b. Shu'ayb in ˇà"if, and al-Mughìra b. Miqsam al-Îabbì in Kùfa, who all died around 125/743.

14

ella landau-tasseron

method, nothing further can be said with certainty about the origin of the alliance ˙adìth but its precepts may be understood in the context of the Ràshidùn period. The great shufflings in the tribal system, precipitated by the conquests and the establishment of the Dìwàn, created great confusion. The caliphs’ decision to leave the pre-Islamic alliances intact reflects their realistic and practical attitude. A different decision would have caused even greater confusion in the system as well as fierce opposition. The caliphs could not annul the existing alliances just as they could not dissolve the boundaries between the descent groups. A ban on new alliances, on the other hand, was feasible. But why would the Ràshidùn be interested in banning new alliances? Perhaps they were trying their best to prevent the scheme of the tribal system, as registered in the Dìwàn, from being altered all too much. This explanation is mainly pertinent to hosting alliances, as are the disputed cases brought before the caliphs. Presumably, political and military alliances would also have been frowned upon. The Ràshidùn must have been aware of the disruptive potential of such alliances. The memory of the ridda wars, when the Muslims were challenged by allied tribes, was still fresh.40 The alliance ˙adìth was obviously present in the mind of the scholar Ibn Óabìb (d. 245/859). Reporting about the alliances of Quraysh, he states at a certain point, “[what I recorded hitherto] are all the pre-Islamic alliances of Quraysh; anything except these constitutes false claims [to alliances] originating in Islamic times, based on friendship, blood relationships, protection (or: neighborliness, jiwàr) or marital ties.”41 Following this statement is a chapter entitled “those who joined Quraysh in Islamic times not by an alliance, but by marital ties, friendship or blood relationship, protection (or: neighborliness, jiwàr) or clientage (walà").” Notwithstanding, some of the relationships recorded in this chapter indeed originated before the advent of Islam. Others, recorded by Ibn Óabìb as “non-allies,” are habitually defined as allies in other sources. Such are the families of Say˙àn, Óà†ib b. Abì Balta'a, Ya'là b. Munya, Shura˙bìl b. Óasana,

40 See e.g., Landau-Tasseron, “The Participation of ˇayyi" in the Ridda,” Jerusalem Studies in Arabic and Islam 5 (1984), 55, 59–60; idem, “Asad from Jahiliyya to Islam,” Jerusalem Studies in Arabic and Islam VI (1985), 20–22. For a detailed discussion of the Islamic objections to alliances, see below, “Islamic objections.” 41 Ibn Óabìb, Munammaq, 248. Ibn Óabìb also quotes the alliance ˙adìth (p. 262), in connection with Rabì'a b. 'Àßim discussed above.

alliances in islam

15

among others.42 The emphasis in the title of the chapter, “in Islamic times” as well as the statement preceding the chapter, seem to divulge Ibn Óabìb’s purpose. Apparently he denies certain tribal relationships the definition “alliance,” attempting to reduce the number of recorded alliances in general and to play down the alliances inaugurated in Islamic times, in accordance with the alliance ˙adìth. An aspiration to portray Quraysh, the Prophet’s tribe, as pure as possible may also have played a role in his decision to compile this particular chapter.43 Alternatively, Ibn Óabìb is right and false claims were made in order to legitimize newly inaugurated alliances. In such a case, too, the same ˙adìth is at work or else no such legitimation would have been needed. b) References to Muslim allies The evidence gleaned from references to Muslim allies is sometimes ambiguous. For example, the sources mention “'Abd al-Ra˙màn b. 'Abdallàh b. Mas'ùd, ally of [the Qurayshì section] Banù Zuhra.”44 This 'Abd al-Ra˙màn, son of a Companion, belonged to the second generation in Islam. His alliance with Zuhra may have been inaugurated in Islamic times, thereby contradicting the precepts of the alliance ˙adìth. But the term “ally” in this pedigree may refer to the father, the Companion 'Abdallàh b. Mas'ùd, designating his preIslamic status. In the latter case, the alliance was valid because it was pre-Islamic. Sometimes the evidence seems clear but is nevertheless misleading. “'Abd al-Ra˙màn b. 'Abd al-Ra˙màn b. 'Abdallàh b. 'Àmir al-Óa∂ramì, ally of Óarb b. Umayya,” was active in the last quarter of the first/seventh century. It could be argued that the term “ally” refers to the ancestor 'Àmir, contemporary of Óarb who died in pre-Islamic times. But alliances (as well as clientage) are often mentioned by the name of the original party (or patron), long deceased.45 “Ally” in this case is therefore ambiguous, applicable Ibn Óabìb, ibid., 249ff., see especially 252. Kister, “On Strangers and Allies in Mecca,” Jerusalem Studies in Arabic and Islam XIII (1990), 141, calls these ties “fake alliances,” following Ibn Óabìb (notwithstanding, Ibn Óabìb includes Ya'là b. Munya among the fakes, whereas Kister, “On Strangers and Allies,” 136, considers him as an important ally of Quraysh). 43 See below, “Islamic objections,” on the special status of Quraysh. 44 Ibn Sa'd, ˇabaqàt, VI, 181. 45 For example, 'Abd al-Ra˙màn b. Say˙àn, a poet in Mu'àwiya’s time, is defined 42

16

ella landau-tasseron

either to 'Abd al-Ra˙màn, who lived in Islamic times, or to one of his ancestors. Sometimes, however, the evidence is clear. There are many cases in which descendants of a party to a pre-Islamic alliance are clearly characterized as allies in their own right.46 In fact it appears that the permanency of hosting alliances was taken for granted and that alliances intrinsically included the descendants of the parties involved.47 This rule applied to women allies as well. 'Abdallàh b. 'Adì b. al-Óamrà" was a Khuzà'ì whose mother was an ally of Banù Zuhra. He was their ally as well.48 The precepts of the ˙adìth, “you should adhere to the alliances of the Jàhiliyya,” sanctioned the existing practice of perpetuating alliances for generations. The latest reference that I found to a hosting alliance dates from the thirteenth century CE.49 Certain alliances were nonetheless clearly inaugurated in Islamic times, contradicting the injunction of the ˙adìth “là ˙ilfa fì -l-Islàm” (“no alliance in Islam”). Among them are the alliances between groups of Persian soldiers, the Asàwira and Óamrà" al-Daylam on the one hand, and Arab tribes on the other. As a rule, Persians who joined the Muslims became clients of Arab descent groups, which means that their status was lower than that of the Arabs. These particular Persian troops were given the option to ally themselves with Arab tribes according to their own choice, which extricated them from the low status of clients. Presumably they were granted this privilege because of their military prowess.50

as an ally of “Óarb b. Umayya,” al-Ißfahànì, Kitàb al-aghànì, ed. Samìr Jàbir, 24 vols. (Beirut: Dàr al-Fikr, n.d.), II, 238. Nafi' b. 'Abd al-Ra˙màn, d. 169/785, was a mawlà of Ja'wana b. Sha'ùb who was an ally of Óamza, uncle of the Prophet (who died in 3/625), al-Dhahabì, Siyar, VII, 336. 46 See Appendix A for a list of allies. 47 See Landau-Tasseron, “Alliances among the Arabs.” 48 Al-Óàkim al-Naysàbùrì, Al-Mustadrak 'alà al-ßa˙ì˙ayn, ed. Mu߆fà 'Abd al-Qàdir 'A†à, 4 vols. (Beirut: al-Maktaba al-'Ilmiyya, 1410/1990), III, 315. 'Abdallàh’s mother was the sister of al-Akhnas b. Sharìq, see on him Appendix A no. 6. 49 The reference is to Rashìd al-Dìn 'Abd al-Wahhàb b. ¸àfir, a jurist from Alexandria, who died in 648/1250. Al-Dhahabì, Siyar, XXIII, 237, calls him a “Qurayshì by alliance.” The other biographers of 'Abd al-Wahhàb, whom I consulted, do not mention this alliance. 50 Yàqùt, Mu'jam al-buldàn, I, 366–68; IV, 343; V, 317; 'Athàmina, “Non Arab Regiments and Private Militias During the Umayyad Period,” Arabica 45 (1998), 348–55; Crone, Roman, Provincial and Islamic Law, Cambridge Studies in Islamic Civilization (Cambridge, London, N.Y., Sydney: Cambridge University Press, 1987), 54. See also below, section ˙alìf, 'adìd and mawlà.

alliances in islam

17

In addition to the peculiar case of the Persian troops, there were several other cases of hosting alliances that were evidently concluded in Islamic times.51 The case of the jurist Màlik b. Anas (d. 179/795) and his family merits some detailed discussion because it reflects the ambiguous position of the institution of alliance (˙ilf ). It is reported that Màlik’s grandfather, who belonged to the second generation in Islam (a Successor, tàbi' ), came to Medina from Yemen. 'Abd alRa˙màn b. 'Uthmàn b. 'Ubaydallàh (nephew of the Companion ˇal˙a b. 'Ubaydallàh, of the Qurayshì section Taym) offered him an alliance on the Jàhilì terms, “we take revenge together and we forego retaliation together as long as the sea is capable of wetting a piece of wool.” Màlik’s grandfather accepted. According to another version, he refused.52 The latter version indicates the difficulty posed by the status of Màlik’s family. They were commonly known as allies of Taym (or of 'Ubaydallàh) and it was apparently well known that their alliance was concluded in Islamic times (contrary to the precepts of the alliance ˙adìth).53 Various attempts were made to account for the family’s status. Certain biographers state that it was Màlik’s great-grandfather who had concluded an alliance with 'Uthmàn b. 'Ubaydallàh al-Taymì. This statement amounts to asserting that the alliance was pre-Islamic and therefore valid.54 Others claim that the close relations between Màlik’s family and Taym were “like an alliance, [but not a real one] because there is no alliance in Islam.”55 Certain scholars contend that Màlik was a client (mawlà), not an ally, of Taym. This claim solves the problem of an illegitimate alliance but it was generally rejected, probably because of the pejorative connotation of the term mawlà which habitually designated non-Arabs. Alternatively, it was explained as mawlà bi-l-˙ilf, a mawlà by alliance, which is the same as an ally.56 The latter explanation resolves the 51

See Appendix B. Ibn Far˙ùn, Al-Dìbàj al-mudhhab fì ma'rifat a'yàn 'ulamà" al-madhhab (Beirut: Dàr al-Kutub al-'Ilmiyya, n.d.), I, 17. On the formula, see Landau-Tasseron, “Alliances among the Arabs,” 155–58. 53 See Appendix C. 54 Ibn Óibbàn, Al-Thiqàt, ed. A˙mad Sharaf al-Dìn, 9 vols. (n.p.: Dàr al-Fikr, 1395/1975), VII, 459; idem, Mashàhìr 'ulamà" al-amßàr, ed. M. Fleischhammer (Beirut: Dàr al-Kutub al-'Ilmiyya, 1959), I, 140. For the great-grandfather, see also alKalàbàdhì, Rijàl ßa˙ì˙ al-Bukhàrì, ed. 'Abdallàh al-Laythì, 2 vols. (Beirut: Dàr alMa'rifa, 1407/1987), II, 692. 55 Al-Khalìl b. 'Abdallàh al-Khalìlì, Al-Irshàd, ed. Mu˙ammad Sa'ìd 'Umar Idrìs, 3 vols. (Riyadh: Maktabat al-Rushd, 1409/1989), I, 283. 56 Ibn Jamà'a, Al-Manhal al-rawiyy fì mukhtaßar 'ulùm al-˙adìth al-Nabawì, ed. Mu˙yì 52

18

ella landau-tasseron

difficulty of Màlik being a mawlà but reinstates the original problem of his involvement in an illegitimate alliance. Another way of unraveling the difficulty is the explanation that the family had only marital ties, but not an alliance, with the Taymìs.57 These various versions show how Muslim scholars grappled with the fact that the status of Màlik’s family contradicted a ˙adìth. Summing up the discussion up to now, it is evident that alliances existed in Islamic times. The alliance ˙adìth was disregarded as far as new alliances were inaugurated; it was taken into consideration as far as attempts were made to reconcile it with the facts of life. “Mà ma'nà baqà" al-˙ilf al-yawm”? (Ibn Óazm) Islamic objections to alliances Objection to alliances is often expressed in general terms, such as statements to the effect that an alliance which contradicts Islamic principles is unacceptable. The main specific reason for objection to alliances is their potential for creating divisions among Muslims. An alliance between descent groups, aimed at raiding other groups, is strictly banned, for it violates the principle of unity among all Muslims.58 Notwithstanding, the institution of military and political alliances is sometimes legitimized by the second half of the alliance ˙adìth (“Islam bolsters the pre-Islamic alliances”).59 Criticism of hosting alliances is often raised in connection with the Islamic laws of inheritance (see below). Yet another ground for objection to alliances is that they entail disregard for moral judgment. According to the custom of the Jàhiliyya a man was obligated to assist his ally even if he was acting iniquitously. This was unacceptable in Islam.60 Of course such an objection could have been raised against the whole tribal system, with its dictum “help your brother regardless of whether he al-Dìn 'Abd al-Ra˙màn Rama∂àn (Damascus: Dàr al-Fikr, 1406/1986), I, 135. On walà" bi-l-˙ilf, see below, section “˙alìf, 'adìd and mawlà.” 57 Ibn Far˙ùn, Al-Dìbàj al-mudhhab, I, 17. 58 Ibn Óajar al-'Asqalànì, Fat˙ al-bàrì, X, 502; al-'AΩìmàbàdì, 'Awn al-ma'bùd, VIII, 100. See also Elad, “The Southern Golan,” 53–55; Schmucker, Untersuchungen, 9–11. 59 Al-ˇabarànì, Al-Mu'jam al-kabìr, XVIII, 322. 60 Al-Jaßßàß, A˙kàm, III, 147, 283; Ibn Óajar al-'Asqalànì, Fat˙ al-bàrì, X, 502.

alliances in islam

19

is the oppressor or the oppressed” (unßur akhàka Ωàliman aw maΩlùman) but no real attempt was ever made to abolish the whole tribal structure. As shown above, social and legal problems precipitated debates about the validity of alliances as early as the period of the Ràshidùn. Such problems continued to emerge in later periods. For example, the Prophet’s descent group (ahl al-bayt), the Banù Hàshim, was entitled to allowances from the public treasury (sahm dhawì l-qurbà). The question arose whether this right of ahl al-bayt extended to their allies as well. It was argued that the Banù l-Mu††alib, who were both relatives and allies of Banù Hàshim, were entitled to allowances as well; the Prophet himself had granted gifts to members of the Banù l-Mu††alib, presumably on the basis of that alliance. The matter was brought before the caliph 'Umar b. 'Abd al-'Azìz (reigned 99–101/ 717–20), who did not invalidate the alliance, but decreed that the reason for the Prophet’s conduct had been different from the one argued. The caliph explained that the Prophet had granted presents to needy people, regardless of their genealogy and alliances. Consequently, the Banù l-Mu††alib’s alliance with the Banù Hàshim did not make them eligible for allowances payable to the ahl al-bayt.61 A contradictory version is recorded as well, stating that the caliph did consider the Banù l-Mu††alib as Hàshimites by alliance and consented to grant them the allowances of the ahl al-bayt.62 There seems, however, to be a consensus among the scholars regarding other allies (as well as the clients) of the Hàshimites. These did not partake of the special prerogatives of the ahl al-bayt.63 Allies (and clients) of ahl al-bayt were thus denied the right to partake of the privileges enjoyed by their noble partners. In other cases the matter was not as decisive. It is asserted that allies (and clients) of the Muhàjirùn and the Anßàr received the same pensions as genuine members of these groups.64 However, the report that the Anßàr 61 Al-ˇabarì, Tafsìr, X, 6; Ibn Óajar al-'Asqalànì, Taghlìq al-ta'lìq, ed. Sa'ìd 'Abd al-Ra˙màn al-Qazaqì, 5 vols. (Beirut/Amman: Al-Maktab al-Islamì and Dàr 'Ammàr, 1405/1985), III, 477–79; al-ˇa˙àwì, Shar˙ ma'ànì al-àthàr, III, 290–92; al-Bukhàrì, Ía˙ì˙, III, 1143; see also Ibn Qudàma al-Maqdisì, Al-Mughnì 'alà mukhtaßar al-Khiraqì, 10 vols. (Beirut, 1405/1985), VI, 316–17. 62 Ibn Sa'd, ˇabaqàt, V, 391–92; al-ˇabarì opts for this opinion, Tafsìr, VI, 10, because of the ˙adìth “the ally of a group is one of them” (˙alìf al-qawm minhum), see on this ˙adìth below. 63 E.g., al-ˇa˙àwì, Shar˙ ma'ànì al-àthàr, III, 292. 64 Ibn Óibbàn, Al-Thiqàt, II, 223; Ibn Sa'd, ˇabaqàt, III, 296; cf. Ibn Abì Shayba,

20

ella landau-tasseron

appealed to the Prophet to grant the status of Anßàrìs to their allies and clients reflects the uncertainty about the status of these social categories.65 Only a need created by such uncertainty could give rise to the ˙adìth “clients, allies and sons of tribeswomen (from men outside the tribe) are part of the tribe.”66 This ˙adìth again created a difficulty where the ahl al-bayt in particular and the Quraysh in general were concerned. This is evident from a statement by Ibn Óazm, insisting that only a genuine Qurayshì, not an ally of Quraysh, is eligible for the caliphate.67 Another scholar interprets the phrase “[allies etc.] are part of the tribe” as a metaphor, taking the family of the Prophet as an illustration. The ˙adìth, he says, does not mean that the dependants of the family of the Prophet are in reality part of that family but only that they should mould their conduct on that of the ahl al-bayt.68 The opinion commonly expressed in the sources is that pre-Islamic hosting alliances may persist if devoid of their legal consequences. In the Jàhiliyya, allies inherited from each other and they shared liability amongst themselves.69 The ˙adìth “. . . allies . . . are part of the tribe” notwithstanding, Muslim scholars usually argued that inheritance by alliance and shared liability were abrogated and pre-Islamic alliances continued only in the sense of companionship, mutual assistance and good counsel. By this position alliances ceased to be a legal institution.70 Al-Shàfi'ì (d. 204/820) maintains that this is the only possible stance because there was no report from the Prophet establishing the legal consequences of alliances, therefore they are not valid.71 There was more than evidence from silence regarding the question of inheritance. Many jurists interpreted the Qur"ànic verse 4:33 Mußannaf, VI, 457, here apparently the mawàlì and ˙ulafà" deserved the pension of Badrìs in their own right. 65 Ibn Óajar al-'Asqalànì, Fat˙ al-bàrì, VII, 114–15. 66 Al-Dàrimì, Sunan, II, 318; Ibn Óanbal, Musnad, IV, 340; Ibn Sa'd, ˇabaqàt, VII, 7; Ibn Abì Óàtim al-Ràzì, Al-'Ilal, ed. Mu˙ibb al-Dìn al-Kha†ìb, 2 vols. (Beirut: Dàr al-Ma'rifa, 1405/1985), II, 358–59; Ibn Qàni', Mu'jam al-ßa˙àba, ed. Íalà˙ b. Sàlim al-Mißràtì, 3 vols. (Medina: Maktabat al-Gharbà" al-Athariyya, 1418/1998), II, 180; al-Zayla'ì, Naßb al-ràya li-a˙àdìth al-hidàya, ed. Mu˙ammad Yùsuf al-Bannùrì, 4 vols. (Cairo: Dàr al-Óadìth, 1357/1938), IV, 148; al-Haythamì, Majma' al-zawà"id, I, 195; Ibn Óajar al-'Asqalànì, Fat˙ al-bàrì, VI, 552–53. 67 Ibn Óazm, Al-Mu˙allà, IX, 359–60; XI, 91. 68 Al-Munàwì, Fay∂ al-qadìr, II, 363; VI, 246. 69 See above, p. 2. 70 Al-ˇabarì, Tahdhìb al-àthàr, 30–31, is unique in stressing the legal duty of an ally to respond when required to help his partner. 71 Al-Shàfi'ì, Al-Umm, VI, 116; see further the references in notes 3 and 6.

alliances in islam

21

(“those with whom you have an agreement by oath, give them their due”) as confirming the Jàhilì practice of inheritance by hosting alliance and Q.33:6 (or Q.8:75, “relatives are closer to one another in the book of God”) as a later verse abrogating this practice. Following the latter verses (as well as several others) it was decreed that the legal heirs of a deceased person were his/her relatives rather than anybody else. These verses are in fact interpreted in many different, sometimes rather complicated ways, but the conclusion of most of the scholars, the Óanafìs excepted, is that no inheritance by alliance is allowed. The ˙adìth “there must be no alliance in Islam” is often adduced to corroborate this conclusion.72 A scholar may interpret the latter half of the alliance ˙adìth, “adhere to the preIslamic alliances,” as “adhere to the legal stipulations of the preIslamic alliances” (a˙kàm). Yet the same scholar neglects to mention the specific legal stipulations of inheritance and shared liability; moreover, he adds in the same breath that Islam abrogated the legal stipulations of alliances.73 Certain scholars accept inheritance by alliance under certain conditions. For example, according to 'Abd al-Razzàq (d. 211/826), 'Umar allowed inheritance from a deceased ally or from anyone counted with a descent group ('adìd, see below), on two conditions: first, that the deceased left no other legal heirs and, secondly, that the living ally had in the past paid blood-money on behalf of the deceased.74 The former condition was adopted by the Óanafìs. They see no problem with the verses “relatives are closer to one another in the book of God” (Q.33:6 and Q.8:75), arguing that this Qur"ànic statement gives precedence to relatives over allies but does not exclude the latter from inheritance. They understand 'Abd al-Razzàq, Mußannaf, X, 305–306; Abù Dà"ùd, Sunan, III, 128–29; alShàfi'ì, Al-Umm, IV, 80; idem, Al-Risàla, ed. A˙mad Mu˙ammad Shàkir (Cairo, 1358/1938), I, 589; al-Nasà"ì, Al-Sunan al-kubrà, IV, 90; al-ˇabarì, Tahdhìb al-àthàr, 25–27; idem, Tafsìr, V, 51–54; X, 57; al-Qur†ubì, Tafsìr, V, 166; Ibn Kathìr, Tafsìr, I, 489–92; Ibn al-Jawzì, Zàd al-masìr, II, 71–73; al-Óanafì, Mu'taßar al-mukhtaßar, II, 294–95; Ibn Óajar al-'Asqalànì, Fat˙ al-bàrì, IV, 472–74; VIII, 247–49; X, 502; al'AΩìmàbàdì, 'Awn al-ma'bùd, VIII, 96–101; al-Zìla'ì, Naßb al-ràya, IV, 149; Ibn Qudàma, Al-Mughnì, VI, 205. Note that some scholars hold that the “relatives” verses abrogate not the Jàhilì custom of inheritance between allies but the mutual inheritance between Muhàjirùn and Anßàr practiced in the early Medinan period. See also Elad, “The Southern Golan,” 54, on the decree ascribed to 'Umar b. 'Abd al-'Azìz, negating inheritance by a “client by oath,” a term presumably identical with “ally” (on this term, see below, section ˙alìf, 'adìd and mawlà). On the law of inheritance, see D. Powers, Studies in Qur"àn and Óadith: the formation of the Islamic law of inheritance (Berkeley: University of California Press, 1986). 73 Al-Munàwì, Fay∂ al-qadìr, V, 462. 74 'Abd al-Razzàq, Mußannaf, X, 305, 307. 72

22

ella landau-tasseron

the ˙adìth “là ˙ilfa fì l-islàm” as an objection to the (alleged?) Jàhilì practice of completely excluding kinsmen from inheritance, in favor of allies.75 Another possibility offered by some scholars was that mutual inheritance should continue in the pre-Islamic alliances but no new alliances entailing inheritance should be inaugurated.76 Discussing the boundaries of the groups sharing responsibility for the payment of blood-money, Muslim scholars do not usually adduce Qur"ànic verses but resort to other arguments. Ibn Óazm adduces the brotherhood (mu"àkhàt) between the Muhàjirùn and the Anßàr in order to vindicate his opinion. He argues that even though these two groups were allied to one another, they never shared the responsibility for paying blood-money. In the same vein, people have special ties with the agnates of their mothers (akhwàl ) but they do not share liability with them. Consequently, Ibn Óazm (based on this logic, on earlier authorities and on decrees ascribed to the Prophet) concludes that a group that may legally share the responsibility for blood-money should consist of agnates and exclude allies.77 Ibn Óazm also categorically rejects the related Jàhilì rule, namely, the common accountability for offenses, of agnates as well as of allies. This rule meant that people were held responsible for actions of others and were prone to suffer the consequences thereof. Ibn Óazm argues that no Muslim can be punished for the crime of another, be he his brother, son, father or ally.78 It is worthy of note that whereas al-Shàfi'ì held the same view on the basis of a formal justification (i.e., no sound ˙adìth from the Prophet establishing it), Ibn Óazm’s position presumably derived from theological considerations, namely the monotheistic principle of personal responsibility. Indeed this principle is fundamental in Islam or else there would be no sense in Divine justice in all its applications (including legal justice).79 The proof that certain Muslim jurists were aware of this problem is that they mention it when discussing the paying of blood-money. They argue that an offender’s group does not in fact share his responsibility for the offense but only helps him meet the consequences.80 75

Al-Jaßßàß, A˙kàm, III, 3–6, 145–48, 197, 283. Al-Mubàrakfùrì, Tu˙fat al-a˙wadhì, V, 174. 77 Ibn Óazm, Al-Mu˙allà, XI, 47, 58–62; see also Ibn Qudàma, Mughnì, VIII, 307. 78 Ibn Óazm, Al-Mu˙allà, XI, 60–61; a similar argument in Ibn Óajar al-'Asqalànì, Fat˙ al-bàrì, XII, 211. 79 Cf. Qur"àn 6:164; 17:15; 35:18; 39:7; 53:38, all including the phrase “no soul will bear the burden of another” wa-là tazaru wàzira wizra ukhrà. See also al-Qur†ubì, Tafsìr, VII, 157. 80 E.g., Zayn b. Ibràhìm Ibn Nujaym, Al-Ba˙r al-rà"iq shar˙ kanz al-daqà"iq, 7 vols. 76

alliances in islam

23

Certain scholars reject or ignore inheritance by allies but accept the validity of their shared responsibility for the payment of bloodmoney.81 This attitude also figures in the form of a ˙adìth, in which the Prophet explicitly states that the duties of allies include the paying of blood-money on behalf of the partner and assistance in general.82 The only school to accept both inheritance and shared liability is the Óanafì. Abù Óanìfa interpreted the full alliance ˙adìth literally and opted for retaining all the legal implications of the preIslamic alliances. The incident of Mu˙ammad seizing the 'Uqaylì, which is explained away by many scholars, is sometimes adduced by Óanafìs as proof of the shared liability of allies.83 The issues of inheritance and shared liability constitute the bulk of the Muslim legal discussions about alliances. Rarely, another legal issue is raised, namely, the participation in qasàma. This is a legal oath multiplied by 50, taken by members of a descent group in cases of homicide.84 The question is whether or not allies of a descent group may participate in such a procedure when performed by their partners.85

(Beirut: Dàr al-Ma'rifa, n.d.), VIII, 455, 456–57; al-Jaßßàß, A˙kàm, III, 194, who also adduces a ˙adìth (with no isnàd ), “nobody should be seized (or: punished) for the crime of his father nor of his brother” là yu"khadhu al-rajulu bi-jarìrati abìhi walà bi-jarìrat akhìhi, the phrasing seems to be echoing the story of the 'Uqaylì, see above, pp. 7–8. 81 'Abd al-Razzàq, Mußannaf, X, 305, 307; Sa'ìd b. Manßùr, Sunan Sa'ìd b. Manßùr, ed. Óabìb al-Ra˙màn al-A'Ωamì (India: al-Dàr al-Salafiyya, 1402/1982), I, 113; alˇabarì, Tafsìr, V, 54; al-Muttaqì al-Hindì, Kanz al-'ummàl, V, 845 (emphasizing that it is applicable to the pre-Islamic alliances). 82 The ˙adìth is rarely quoted. In fact, I found it in only one source (out of the hundreds covered by the databases which I used), al-Haythamì, Musnad al-Óàrith, II, 860. It is molded on Q.5:1 (awfù bi-l-'uqùd ) and Q.4:33 (alladhìna 'aqadat aymànukum) and it is transmitted by the Egyptian scholar Ibn Lahì'a—'Amr b. Shu'ayb—his father—his grandfather; 'Amr is one of the common links of the alliance ˙adìth, see above, note 39. 83 Al-Jaßßàß, A˙kàm, III, 197; al-ˇa˙àwì, Mukhtaßar ikhtilàf al-'ulamà", V, 100, 102; Al-Óanafì, Mu'taßar al-mukhtaßar, II, 109; Ibn al-Jawzì, Zàd al-masìr, II, 72. Cf. alSarakhsì, Al-Mabsù†, 30 vols. (Beirut: Dàr al-Ma'rifa, 1406/1986), VIII, 81, where the two practices are valid for mawlà l-muwàlàt, the Islamic institution akin to the Jàhilì ˙alìf (see below, section ˙alìf, 'adìd and mawlà). See also al-'AΩìmàbàdì, 'Awn al-ma'bùd, VIII, 97, 102. Rarely, Ibn Óanbal is also reported to have accepted inheritance by allies (but this is not al-mashhùr 'anhu), 'Awn al-ma'bùd, VIII, 102; Ibn Kathìr, Tafsìr, I, 491. Màlik is also reported, rarely, to have accepted common liability, Ibn Óazm, Al-Mu˙allà, XI, 59. 84 See Pedersen [and Linant de Bellefonds], art. “asàma,” EI 2, IV, 689–90; Crone, “Jàhilì and Jewish law: the Qasàma,” Jerusalem Studies in Arabic and Islam 4 (1984), 153–201. 85 Ibn Óazm, Al-Mu˙allà, XI, 89–91 (Ibn Óazm’s answer is negative).

24

ella landau-tasseron Óalìf, 'adìd and mawlà

The legal functions of the pre-Islamic hosting alliances were curtailed in Islamic times but one function was added, albeit of an administrative rather than a legal nature. Being an ally of a descent group meant being counted with that group for administrative purposes such as conscription and pension. 'Umar personally set the model by attaching his Qu∂à'ì ally, Ka'b b. 'Adì, to his own Qurayshì descent group, for the purpose of collecting the pension.86 The term for such an attached person is 'adìd, or huwa fì 'idàd, or 'idàduhu fì, all meaning the same, namely, “he is counted with” [a certain group of people].87 Although mostly meant to refer to Arabs registered in the Dìwàn, the term 'adìd could be applied to non-Arabs as well.88 It is not clear whether a formal alliance was always involved in these cases, in other words, it is not clear whether one could be an 'adìd without being a ˙alìf, the terms being often interchangeable. There are, however, scholars who explicitly distinguish between the two. “An ally is one who contracts with another to withstand [together] wrongdoing and to cooperate against anyone wishing them, or one of them, any harm, whereas an 'adìd is a person who is not a part of any descent group (là 'ashìra lahu), so he joins one and regards himself as part of them.”89 While this definition appears clear enough, it does not supply details on the rights and obligation of an 'adìd, which may be very similar to those of a guest ally. There are additional data indicating a distinction between the two institutions.90 In spite of the ambiguity outlined above, 'adìd remains a rather limited term. By contrast, the category of mawlà is by far more complex and its blending with that of ˙alìf is more confusing. The latter can safely be rendered by “an ally” (although there are several Ibn Óajar al-'Asqalànì, Ißàba (ed. Beirut), V, 601–604; Juda, Aspekte, 62. E.g, al-Nawawì, Tahdhìb al-asmà", III, 193. See also Juda, Aspekte, 62. The implication that the institution of 'adìd was pre-Islamic (al-Shaybànì, Al-Mabsù†, ed. Abù al-Wafà al-Afghànì, 5 vols. (Karatchi: Idàrat al-Qur"àn wa-l-'Ulùm al-Islàmiyya, n.d.), IV, 662) must be an anachronism since neither conscription nor pensions existed at that time. 88 See al-ˇabarì, Ta"rìkh, II, 401, where a Persian is converted to Islam and attaches himself to Zuhra b. Óawiyya al-Tamìmì (one of the Muslim commanders at Qàdisiyya) as an 'adìd. This case is peculiar first because such a person would normally become a mawlà, and, secondly, because 'adìd is as a rule attached to a descent group (or a town), not to an individual. See also above, p. 16. 89 Ibn Qudàma, Mughnì, VIII, 307. 90 See Appendix D. 86 87

alliances in islam

25

kinds of allies), whereas mawlà may refer to a client, a patron, an agnate (brother, son, father’s brother, father’s brother’s son), an affined kinsman (brother-in-law, son-in-law), a friend, a supporter, a follower, a drinking companion, a partner, a newly-converted Muslim attached to a Muslim and, last but not least, an ally.91 Most of these categories have legal implications. In Islamic times, the term mawlà mostly referred to Muslim freedmen and free non-Arabs who attached themselves to Arabs upon their conversion to Islam. In these senses mawlà is commonly translated as “a client.” The association of mawlà with non-Arabs and a low status imparted an increasingly pejorative connotation to it. Naturally, great confusion is caused by the multiplicity of meanings of mawlà. Confusion prevails also in the attempts to classify mawàlì. Alongside the fairly clear category of mawlà l-'itàqa (freedman) the sources mention hazy terms such as mawlà l-muwàlàt, mawlà l-'aqd, mawlà l-˙ilf, mawlà l-islàm, mawlà l-qabìla.92 In addition, the fact that an ally may be designated by both terms, mawlà and ˙alìf, reflects as well as precipitates confusion between the two. We may distinguish between the convergence of these categories and the confusion between the terms. As categories, mawlà and ˙alìf converge at least in two points. First, certain legal issues are pertinent to both ˙alìf (in the sense of a guest ally) and mawlà (in many of its senses).93 Secondly, both belong to the class of “followers” (tàbi'ùn) or “entourage”

91 Al-Bu˙turì, Ash'àr al-˙amàsa (Hamasae Carmina), ed. G.G. Freytag (Bonn, 1828), 187; al-Bukhàrì, Ía˙ì˙, IV, 1671; al-Nawawì, Tahdhìb al-asmà", III, 369; al-Munàwì, Fay∂ al-qadìr, II, 111; Ibn al-Athìr, Al-Nihàya fì gharìb al-athar, ed. ˇàhir A˙mad alZàwì and Ma˙mùd Mu˙ammad al-ˇanàjì, 5 vols. (Beirut: al-Maktaba al-'Ilmiyya, 1399/1979), V, 227; Ibn ManΩùr, Lisàn, XV, 408–409; Ibn Óajar al-'Asqalànì, Fat˙ al-bàrì, VIII, 248; idem, Muqaddimat fat˙ al-bàrì, ed. Mu˙ammad Fu"àd 'Abd al-Bàqì and Mu˙ibb al-Dìn al-Kha†ìb (Beirut: Dàr al-Ma'rifa, 1379/1959), I, 207, 340; alSuhaylì, Al-Raw∂ al-unuf fì tafsìr al-sìra al-nabawiyya li-Ibn Hishàm, ed. Majdì Manßùr al-Shùrà, 4 vols. (Beirut: Dàr al-Kutub al-'Ilmiyya, 1418/1997), II, 246; al-'AΩìmàbàdì, 'Awn al-ma'bùd, I, 260; XII, 138; al-Mubàrakfùrì, Tu˙fat al-a˙wadhì, X, 148; cf. Crone, Roman, 49 and Wael Hallaq, “The Use and Abuse of Evidence,” Journal of the American Oriental Society 110/I (1990), 85; Juda, Aspekte, 68–87. 92 E.g., al-Shàfi'ì, Al-Umm, IV, 126; al-Suyù†ì, Tadrìb al-ràwì fì shar˙ taqrìb alNawawì, ed. 'Abd al-Wahhàb 'Abd al-La†ìf, 2 vols. (Riyà∂: Maktabat al-Riyà∂ alÓadìtha, n.d.), II, 282–83; Ibn Jamà'a, Al-Manhal al-rawiyy, I, 135. I never encountered the term walà al-'aqìda mentioned by Juda, Aspekte, 18, 61; it seems to have been coined by the author, cf. p. 55 where he mentions ˙ilf shakhßì, another term apparently unknown to medieval Muslim scholars. It should also be noted that Juda’s work ignores the confusion between the terms, which is evident in the sources. 93 See e.g., Ibn Óazm, Al-Mu˙allà, XI, 58ff., 89; al-Shàfi'ì, Al-Umm, VI, 116.

26

ella landau-tasseron

(˙àshiya) of the person to whom they are attached. As such, certain legal and social implications apply to both ˙ulafà" and mawàlì; for example, the women of a household do not have to conceal themselves from the men who belong to these categories.94 As terms, mawlà is sometimes glossed by ˙alìf and, though very rarely, also vice versa. This occurs mainly in dictionaries and in commentaries on poetry.95 When mawlà is glossed by ˙alìf, it is in fact a shortened form of mawlà bi-l-˙ilf. One of the types of relationships between non-agnates is walà" bi-l-˙ilf, “a tie by oath/by alliance.”96 This term is usually recorded as one of the types of walà", without explanation. There is no reason to doubt that walà" bi-l-˙ilf is identical with ˙ilf, perhaps mainly in the sense of hosting alliance, precisely because mawlà some94 The same applies to a hireling (ajìr), see Ibn Qutayba, Ta"wìl mukhtalif al-˙adìth, ed. Mu˙ammad Zuhrì al-Najjàr (Beirut: Dàr al-Jìl, 1393/1972), I, 308. The embarrassment caused by this norm is apparent in the story of the Companion Abù Óudhayfa, his wife, and his ally (or mawlà), Sàlim, see Ibn Óajar al-'Asqalànì, Fat˙ al-bàrì, IX, 134. The case is discussed in detail by Motzki, “Der Fiqh Des Zuhri: Die Quellenproblematik,” Der Islam 68 (1991), 34–42, and Ulrike Mitter, “Unconditional manumission of slaves in early Islamic law: a ˙adith analysis,” Der Islam 78 (2001), 51–58. 95 Al-Daqìqì, Ittifàq al-mabànì wa-iftiràq al-ma'ànì, ed. Ya˙yà 'Abd al-Ra"ùf Jabr (Amman: Dàr 'Ammàr, 1405/1985), I, 140–41; Ibn ManΩùr, Lisàn, XV, 409. I found in fact only one instance of ˙alìf glossed by mawlà instead of the other way around, in al-Ràzì, Mukhtàr al-ßi˙à˙, I, 63: al-˙alìf al-mu˙àlif wa-l-mawlà. Other cases are misleading. Ibn Óazm, Al-Mu˙allà, XI, 90 says “˙alìf is called a mawlà,” but in fact he is commenting on the word mawàlì (not on ˙ulafà") in a certain ˙adìth, see below note 114. In al-Mizzì, Tahdhìb al-kamàl fì asmà" al-rijàl, ed. Bashshàr 'Awwàd Ma'rùf, 35 vols. (Beirut: Mu"assasat al-Risàla, 1400/1980), XIV, 307; and Ibn Khallikàn, Wafayàt al-a'yàn wa-anbà" abnà" al-zamàn, ed. I˙sàn 'Abbàs, 8 vols. (Beirut: Dàr al-Thaqàfa, 1969–72), VI, 392 (I owe both references to Monique Bernards), we find “wa-al-˙alìf 'inda al-'arab al-mawlà,” but this gloss is a commentary on the word mawlà in a verse by al-Farazdaq. The poet pejoratively calls the linguist 'Abdallàh b. Abì Is˙àq “mawlà mawàlì.” The explanation is that 'Abdallàh was a mawlà of the Óa∂ramì family, who was allied to Banù 'Abd Shams. Thus it is again mawlà (in the verse) which is glossed by ˙alìf (in the commentary) and not vice versa. Baghdàdì, in his commentary to this verse, uses ˙alìf and mawlà interchangeably, both when referring to 'Abdallàh and to the Ha∂ramì family, see al-Baghdàdì, Khizànat al-adab wa-lubb lubàb lisàn al-'arab, ed. 'Abd al-Salàm Mu˙ammad Hàrùn, 13 vols. (Cairo: Dàr al-Kàtib al-'Arabì li-l-ˇibà'a wa-l-Nashr and Maktabat alKhànjì, 1967–87), I, 236. See also Crone, Roman, 49–51. 96 Óilf meaning both oath and alliance; the connection between the two is obvious: the parties concluding an alliance take an oath of fidelity. Note that walà" in this phrase should not be translated as “clientage,” walà" being a multi-meaning term. See also Crone, Roman, 55, where she dismisses the walà" bi-l-˙ilf as irrelevant to the question of clientage. However, contrary to the impression given by Crone, walà" al-˙ilf is not merely a phrase used in poetry.

alliances in islam

27

times is explicitly glossed by ˙alìf.97 A term analogous to mawlà bil-˙ilf is mawlà l-yamìn, “one attached by oath,” which is also explicitly glossed by ˙alìf.98 Another synonymous term, mawlà l-'aqd, is recorded in a letter ascribed to the caliph 'Umar b. 'Abd al-'Azìz to one of his governors. In this letter 'Umar is said to have invalidated the practice of inheritance by mawlà l-'aqd. This is in keeping with the opinion of most jurists, who considered inheritance by allies as an abrogated pre-Islamic practice.99 The Islamic counterpart of ˙ilf/walà" bi-l-˙ilf/'aqd/yamìn appears to be walà" al-muwàlàt, contractual relationship entailing precisely the same rights and obligations as the Jàhilì ˙ilf and arousing the same debates among Muslim scholars.100 Even when ˙alìf and mawlà l-muwàlàt are mentioned as separate categories, it is difficult to see the differences.101 Presumably they were the same legally except that one was mainly assigned to Arabs, the other one to non-Arabs. In Islamic times, when an Arab was referred to as a mawlà, it was usually taken to mean mawlà bi-l-˙ilf, apparently because the term mawlà in its most common usage was associated with manumitted slaves and non-Arabs who had become Muslims. Thus when Zuhrì and Ibn Is˙àq define the jurist Màlik b. Anas and his family as

97

Al-ˇabarì, Tafsìr, XIV, 150; al-Munàwì, Fay∂ al-qadìr, II, 363; al-Dhahabì, Ma'rifat al-qurrà" al-kibàr, ed. Shu'ayb al-Arna"ù†, Bashshàr 'Awwàd Ma'rùf and Íàli˙ Mahdì 'Abbàs, 2 vols. (Beirut: Mu"assasat al-Risàla, 1404/1984), I, 105; Ibn Óajar al-'Asqalànì, “Nukhbat al-fikr,” in Kitàb subul al-salàm (Beirut: Dàr I˙yam" al-Turàth al-'Arabì, [1379 AH]), I, 233; idem, Fat˙ al-bàrì, IV, 86; al-Suyù†ì, Tadrìb al-ràwì, II, 382–83; Ibn Jamà'a, Al-Manhal al-rawiyy, I, 135; al-Mubàrakfùrì, Tu˙fat al-a˙wadhì, I, 62. Juda, Aspekte, 2, identifies walà" al-˙ilf with ˙ilf (but does not clearly distinguish between types of ˙ilf ). 98 Al-Bukhàrì, Ía˙ì˙, IV, 1671, and further references in Goldziher, Muslim Studies, I, 65 note 8, 101 notes 5–6. 99 See above on the abrogation of inheritance by allies. The terms yamìn, 'aqd and ˙ilf are equated in al-Jaßßàß, A˙kàm, III, 284. Óilf is commonly adduced as a synonym of the Qur"ànic 'uqùd in Q.5:1; e.g., al-Shawkànì, Fat˙ al-qadìr, II, 7. On the term mawlà l-'aqd (including references to 'Umar’s letter), see the discussion in Elad, “The Southern Golan,” 52, challenging Crone, Roman, 53–54. Crone argues that mawlà l-'aqd in 'Umar’s letter is not a ˙alìf but the Islamic contractual client, mawlà l-muwàlàt, see below. The sources, however, sometimes equate mawlà l-muwàlàt with ˙alìf as Crone herself admits, contrary to her own argument, on p. 53 and note 107 (additional sources are al-Jaßßàß, A˙kàm, III, 145–47; al-Sarakhsì, Mabsù†, VIII, 81, and al-Baghdàdì, Khizànat al-adab, I, 236). 100 See Hallaq, “The use and abuse,” 848–46, challenging Crone’s interpretation of walà" al-muwàlàt. 101 Ibn Qudàma, Mughnì, VIII, 307, and see also above note 83.

28

ella landau-tasseron

mawàlì of the Qurayshì section Taym, all other scholars interpret it as mawàlì bi-l-˙ilf or ˙ulafà", sometimes stressing that they were genuine Arabs of the tribe Aßba˙.102 Likewise, the official al-Walìd b. 'Abd al-Ra˙màn was classified as a mawlà but Ibn 'Asàkir asserted that this was impossible since this al-Walìd was an Arab. Ibn Óajar commented on Ibn 'Asàkir’s view saying, “it is possible that [alWalìd] was a mawlà bi-l-˙ilf.”103 Certain individuals are interchangeably called ˙alìf and mawlà.104 The confusion may easily be removed by interpreting mawlà in these cases as mawlà bi-l-˙ilf/bi-l-'aqd/al-yamìn which is tantamount to ˙alìf. However, this easy solution was not always adopted. Mawlà in such cases is sometimes interpreted as a client, not an ally, which results in a puzzle: was the individual in question, interchangeably called ˙alìf and mawlà, an ally or a client?105 The case of the Companion Íuhayb b. Sinàn is instructive. He was an Arab from al-Namir b. Qàsit, defined as “either a ˙alìf or a mawlà of 'Abdallàh b. Jud'àn.” It should be noted that, although an Arab, Íuhayb had been a slave manumitted in the Jàhiliyya by the Qurayshì nobleman 'Abdallàh b. Jud'àn.106 Obviously, mawlà in this case means “client,” and the confusion is not merely a matter of terminology but of legal and

Ibn 'Abd al-Barr, Al-Tamhìd, ed. Mu˙ammad 'Abd al-Kabìr al-Bakrì and Mu߆afà b. A˙mad al-'Alawì, 24 vols. (Morocco: Wizàrat 'Umùm al-Awqàf wa-lShu"ùn al-Islàmiyya, 1387/1967), XVI, 152; Ibn Far˙ùn, Al-Dìbàj al-mudhhab, I, 17; al-Suyù†ì, Tadrìb al-ràwì, II, 383; Ibn Jamà'a, Al-Manhal al-rawiyy, I, 135. See above, “Alliances in Islamic times” and below, Appendix C. 103 Ibn Óajar al-'Asqalànì, Tahdhìb al-tahdhìb, 14 vols. (Beirut: Dàr al-Fikr, 1404/1984), XI, 123. Cf. Juda, Aspekte, 68–71 on the increasingly narrowing down of the term mawlà to non-Arabs. 104 See Appendix E. 105 When 'A††àf b. Khàlid was asked, “was 'Ammàr b. Yàsir an ally of your descent group?” he replied, “no, he was our client (mawlà)”; al-Haythamì, Majma' al-zawà"id, IX, 292. 'Ammàr was an ally but his mother was a slave-girl, see below note 121. A˙mad b. 'Abdàn (died around the year 300/912), attached to Banù Zuhra, was asked whether he was a genuine tribesman, an ally, or a client. He answered that he and his family were clients of 'Abd al-Ra˙màn b. 'Awf; alDàraqu†nì, Su"àlàt Óamza, ed. Muwaffaq b. 'Abdallàh b. 'Abd al-Qàdir (Riyà∂: Maktabat al-Ma'àrif, 1404/1984), I, 185. Similar cases: Bujayr b. Abì Bujayr; Ibn 'Abd al-Barr, Al-Istì'àb fì ma'rifat al-aß˙àb, ed. 'Alì Mu˙ammad al-Bijàwì, 4 vols. (Beirut: Dàr al-Jìl, 1412/1992), I, 148; Îamra b. Tha'laba, ibid., II, 749; 'Anaza al-Sulamì, ibid., III, 1246. 106 Ibn Óibbàn, Al-Thiqàt, III, 194; idem, Mashàhìr 'ulamà" al-amßàr, I, 20. Ibn Óajar al-'Asqalànì, Ißàba (ed. Cairo), II, 195. In this case, there was clearly a bias and a material reason for the confusion, see below. See also Appendix E last section, on freedmen who became allies rather than clients. 102

alliances in islam

29

social status since the status of clients was lower than that of allies. This clearly emerges from the report that the caliph Hishàm (known for his frugality) ordered that the pension of the descendants of that same Íuhayb be reduced from the level of ˙ulafà" to the level of mawàlì.107 The confusion between the terms was rather annoying to the family of the Kindì Abù Tijràt. He is defined once as a mawlà bi-l-˙ilf of Shayba b. 'Uthmàn, another time as a ˙alìf of Shayba’s descent group, the Banù 'Abd al-Dàr (a section of Quraysh). Abù Tijràt was still Shayba’s ally at the time of Mu'àwiya. His family indignantly protested when people referred to them as mawàlì, presumably because they could be mistaken for clients, and insisted that they should be called ˙ulafà", allies.108 On the other hand, members of the family called Àl Hàni" apparently utilized the confusion between the terms (mawlà/mawlà bi-l-˙ilf/˙alìf ) to their own advantage. According to Ibn Óabìb they “falsely claimed to be allies of [the caliph] 'Uthmàn while they were merely his clients” ( yadda'ùna ˙ilf 'Uthmàn wa-innama hum mawàlin lahu).109 The terms mawlà and ˙alìf are interchangeably used in the 'ahd al-umma, the document drawn by the Prophet to define the relationships between the Anßàr, the Muhàjirùn and Jewish groups in Medina.110 In the same vein, Jewish groups that were allied to Yathribì descent groups are sometimes referred to as mawàlì.111 That the Jews were mawàlì bi-l-˙ilf, allies (but apparently not guest allies), rather than clients subordinate to the Yathribì Arabs, may be inferred not only from the usage of the terms in 'ahd al-umma but also from the discussion between the Yathribì delegation and the Prophet at al'Aqaba, prior to the Hijra. The Yathribìs hesitated to replace their alliances (˙ibàl ) with the Jews by a commitment to Islam. They 107 Ibn Sa'd, Al-ˇabaqàt al-kubrà, al-qism al-mutammim, ed. Ziyàd Mu˙ammad Manßùr (Medina: Maktabat al-'Ulùm wa-l-Óikam, n.d.), I, 96–97. 108 Ibn Óajar al-'Asqalànì, Ißàba (ed. Beirut), VII, 532; VII, 51; Ibn 'Abd al-Barr, Istì'àb, IV, 1793; al-Óàkim al-Naysàbùrì, Al-Mustadrak, IV, 79. There is, however, confusion regarding this family, see Ibn Óabìb, Munammaq, 250, 260, where they are mentioned as slaves of the Makhzùmì 'Umàra b. al-Walìd b. al-Mughìra. To add to the confusion, Abù Tijràt’s sister, Baraka b. Yasàr, was both an ally of Banù 'Abd al-Dàr, and a mawlàt of Abù Sufyàn, who belonged to a different Qurayshì section (note that both the names Baraka and Yasàr are typical of slaves). 109 Ibn Óabìb, Munammaq, 250, on Ibn Óabìb, see above “Alliances in Islamic times.” 110 Ibn Hishàm (ed. Beirut), Sìra, III, 33. 111 Al-ˇabarì, Tafsìr, II, 49; Ibn Kathìr, Tafsìr, III, 480; al-Qur†ubì, Tafsìr, II, 20; XIV, 139.

30

ella landau-tasseron

expressed the fear that the Prophet might at a later stage abandon them, whereupon they would be left with no allies at all. The Prophet reassured them of his intentions, using the Jahilì formula of alliance: al-dam al-dam wa-l-hadm al-hadm (that is, we will act together in matter of blood-revenge).112 This report does not reproduce the actual dialogue at al-'Aqaba. Its hindsight is apparent through the fact that it does not even allude to the possibility of the Yathribìs adhering to both their alliances on the one hand and to Islam on the other.113 At that early stage, before the Hijra and before the 'ahd al-umma, a severance from the Jews had not yet been planned. But the definition in this report of the relationship between the Yathribìs and the Jews is most probably correct, precisely because it is rather unusual. In Islamic times the memory of the former alliances between Arabs and Jews was not cherished. The Jews of Medina are therefore habitually mentioned as mawàlì rather than ˙ulafà" of the Anßàr because the former term could be associated with a lower status, as befitted the Jews in Islamic times. Conversely, the report in which Mu˙ammad calls the Anßàr “mawàlì of the Jews” is sometimes glossed by ˙ulafà", allies, in order to negate any implication that Arabs have been clients of Jews. It should be noted that the report calling the Anßàr mawàlì of the Jews could have been glossed by “patrons” but it is not.114 Obviously, it is not possible that the Yathribìs were allies of the Jews while at the same time the latter were the clients of the Yathribìs. The Jews of Yathrib must have been allies, not clients, of their Arab neighbors, the future Anßàr.115 The evidence for the Jàhiliyya and pre-conquest Muslim society in fact gives rise to the question whether or not ˙alìf and mawlà 112

See above, note 14. Such a possibility was not denied in other contexts, e.g., al-Qur†ubì, Tafsìr, IV, 58 (to Q.3:28). 114 Ibn Óazm, Al-Mu˙allà, XI, 90; al-Suhaylì, Al-Raw∂ al-unuf, II, 246, both gloss mawàlì with ˙ulafà"; see also (the report without the gloss) Ibn Hishàm (ed. Beirut), Sìra, II, 276; al-ˇabarì, Ta"rìkh, I, 558; Ibn Óibbàn, Al-Thiqàt, I, 90; Ibn Kathìr, Al-Bidàya wa-l-nihàya, III, 148. The purpose of this report is to vindicate Mu˙ammad’s prophecy. The Yathribìs, who had ties with their Jewish neighbors, recognize in Mu˙ammad the Prophet foretold by the Jews. For that matter, mawàlì in this report may have originally meant just “friends” or “neighbors.” 115 According to Crone, Roman, 56–57, and art. “Mawlà,” EI 2, VI, 875, the Jews were clients of, and therefore subordinate to, the Yathribì Arabs. This is the only point where Hallaq agrees with her, Hallaq, “The use and abuse,” 86. See also Michael Lecker’s contribution (next chapter) on the status of Jewish tribes just prior the arrival of Islam. 113

alliances in islam

31

where differentiated in these periods. This question is significant for the understanding of the development of the Islamic institutions of walà". If there existed in pre-Islamic times one institution, alliance, known by different names—walà" and ˙ilf —then there is ground for the interpretation of Islamic walà" as having developed out of alliances.116 If, on the other hand, walà" in the sense of clientage, and ˙ilf/walà" bi-l-˙ilf/walà" al-yamìn/walà" al-'aqd were two different institutions already in pre-Islamic Arabian society, then the development of the Islamic clientage must be interpreted differently.117 As mentioned above, the clearest category is that of freed slaves. In Islam, ˙alìf, guest-ally, and mawlà in the sense of freedman/client were certainly different categories. But the case for the Jàhiliyya is not that clear. If there was at that time a distinction between ˙alìf and mawlà, one would expect that freedmen would become mawàlì, clients, of their former masters, whereas free people would attach themselves to others as allies. Notwithstanding, some freed slaves are said to have become mawàlì whereas others, in apparently identical circumstances, became ˙ulafà".118 On the other hand, there is some evidence suggesting a pre-Islamic differentiation between mawàlì and ˙ulafà". The list of the Qurayshì women who followed Mu˙ammad includes separate categories for allies, mawàlì and strangers [married to Qurayshìs].119 This probably means women from descent groups allied to Quraysh, freedwomen, and women whose sole tie with Quraysh was their marriages with members thereof. Another indication of a differentiation is the report about the relatives of the Companion 'Urwa b. Mas'ùd. When 'Urwa was killed in ˇà"if by his fellowtribesmen, his son and nephew ran away to Medina and converted to Islam. The Prophet suggested that they become mawàlì of whomever they liked; it is not clear which of the many senses of mawlà is meant here. However, they chose to be only “mawàlì of God and the 116 This is a widespread interpretation, challenged by Crone in her Roman Provincial and Islamic Law. The book was severely criticized by Wael Hallaq in “The Use and Abuse of Evidence.” 117 The fact that Muslim scholars mention ˙alìf, mawlà and 'adìd as three separate categories cannot be used as evidence of such a differentiation in the Jàhiliyya, e.g., al-Sarakhsì, Mabsù†, XXVII, 125; Ibn Óazm, Al-Mu˙allà, XI, 89. 118 See Appendix E and Appendix A no. 8. Cf. Ibn al-'Arabì, quoted in alMunàwì, Fay∂ al-qadìr, VI, 111. He discusses the ˙adìth prohibiting the transfer of walà" (“man tawallà ghayr mawàlìhi . . .”), which is clearly about freedmen; but he uses ˙ilf and walà interchangeably in the discussion. 119 Ibn Sa'd, ˇabaqàt, VIII, 222.

32

ella landau-tasseron

Prophet.” Mu˙ammad then suggested that they attach themselves by alliance to their maternal uncle Abù Sufyàn, which they did. However, they resided with their fellow tribesman al-Mughìra b. Shu'ba, not with their ally.120 Another report pointing to a clear distinction between ˙alìf and mawlà is the following: Quraysh demanded from Mu˙ammad to return to Mecca the mawàlì and the allies of Quraysh, who had emigrated to Medina. They meant Bilàl, 'Ammàr and Íubay˙ who were mawàlì, Ibn Mas'ùd and al-Miqdàd who were allies.121 It is worthy of note that those designated as mawàlì were associated with slavery,122 whereas Ibn Mas'ùd and al-Miqdàd were not.123 Yet another pointer to a distinction between ˙alìf and mawlà in the Jàhiliyya is the fact that a partner in an alliance (a ˙alìf ) could be patron to a mawlà but apparently not vice versa.124 Indeed, it appears that a person entering an alliance brought his clients together with him.125 The only reverse case that I found was Zayd b. Óàritha, client of the Prophet, who is reported to have had an ally, Mu'ayqìb b. Abì Fà†ima al-Dawsì. However, this isolated case is doubtful, due to the confusion regarding the status of Mu'ayqìb.126 There is thus no unequivocal evidence proving a differentiation between clients and allies in the Jàhiliyya. Even if these two cate-

Ibn Sa'd, ˇabaqàt, V, 504. See also Juda, Aspekte, 58 on “mawàlì of God and the Prophet.” 121 Ibn al-Qaysarànì, Tadhkirat al-˙uffàΩ, ed. Óamdì 'Abd al-Majìd Ismà'ìl al-Salafì, 4 vols. (Riyà∂: Dàr al-Íamì'ì, 1415/1995), II, 460. 122 Bilàl was a slave, bought and freed by Abù Bakr, Ibn Óajar al-'Asqalànì, Ißàba (ed. Cairo), I, 165. The exact identity of Íubay˙ is not certain but all people bearing this name appear to have been clients, see ibid., II, 1757–56. 'Ammàr is a rare case of both a mawlà and an ally. He was a son of a slave girl and therefore inherited her status but he was at the same time an ally of the Qurayshì section Makhzùm (or of Abù Salama and Umm Salama or of Abù Óudhayfa), see Ibn Hishàm (ed. Beirut), Sìra, II, 96; al-Balàdhurì, Ansàb l-ashràf, ed. Mu˙ammad Óamìdullàh, vol. 1 (Cairo: Ma'had al-Makh†ù†àt and Dàr al-Ma'àrif, 1959), I, 157; Ibn Óazm, Al-Mu˙allà, XI, 89; Ibn 'Abd al-Barr, Istì'àb, III, 1135; al-Óàkim alNaysàbùrì, Al-Mustadrak, III, 434. 123 The case of al-Miqdàd is ambiguous, see below note 195. 124 See Appendix E. 125 Cf. Mu"arrij al-Sadùsì, Kitàb ˙adhf min nasab Quraysh, ed. Íalà˙ al-Dìn alMunajjid (Beirut: Dàr al-Kitàb al-Jadìd, 1396/1976), 59: the list of allies of Banù 'Abd al-'Uzzà includes Óà†ib and his client Sa'd. Similarly, Sa'ìd b. Abì Hind [and his family], clients of Samura who was an ally of Banù Khudra (also called Banù Abjar), were considered part of that descent group together with their patron (da'watuhum fì banì l-Abjar li-mu˙àlafat Samura b. Jundab iyyàhum), al-Mizzì, Tahdhìb alkamàl, XI, 93. See also Appendix E. 126 See Appendix E and note 182. 120

alliances in islam

33

gories were clearly marked off from one another, the sources do not allow a decisive characterization of them. In conclusion, it is evident that Muslims debated the validity of the pre-Islamic institution of alliance from a very early period onward. Eventually, pre-Islamic alliances of all types survived in Islam but the legal implications of hosting alliances were altered, to a greater or lesser degree. The ban issued by jurists on newly inaugurated alliances was not successful and alliances of all types continued to be initiated. Thus the precepts of the ˙adìth “there must be no alliance in Islam but you should adhere to the pre-Islamic alliances” were not always followed, yet they were taken into consideration since justifications were sought for keeping and inaugurating alliances. Confusion between ˙alìf and mawlà existed in pre-Islamic as well as in Islamic usage. This confusion arose from, and was reflected by, the term “mawlà bi-l-˙ilf/yamìn/'aqd,” which in fact means an ally, even though it is constructed with the word that also designates a client. As for the Islamic clientage, whatever its course of inception and development, it did not replace the institution of hosting alliance but coexisted with it. It is probable that clientage and hosting alliance had already coexisted in pre-Islamic times but the assignment of the latter to Arabs and the former status, both servile and contractual, to non-Arabs, is typically Islamic.

34

ella landau-tasseron Appendices

The following five appendices are lists of examples, to which many more can be added. Appendix A: List of allies perpetuating pre-Islamic alliances; other allies in Islamic times seem also to be perpetuating pre-Islamic alliances. Appendix B: Alliances inaugurated in Islamic times (followed by cases in which the time of the inauguration of the alliance cannot be ascertained) Appendix C: Màlik b. Anas and his family Appendix D: 'adìd ('adìd [“counted with”] and ˙alìf [guest ally; interchangeably applied; Indications of a distinction between the two categories) Appendix E: ˙alìf and mawlà (Clients of ˙alìfs; Freedmen who became allies) Appendix A List of allies perpetuating pre-Islamic alliances: 1) Mu˙ammad and Sulaymàn, sons of 'Uqba b. Dubayya, from Banù Shaybàn of the tribe Sulaym, were killed in the battle of Óarra (63/683). They are defined as “allies of Abù ˇàlib” (who died many decades before). The alliance had been concluded in pre-Islamic times.127 2) Ja'far b. 'Abdallàh b. Màlik, known as son of Bu˙ayna (daughter of al-Óàrith b. 'Abd al-Mu††alib), died in the year 63/683. He was an ally of Banù l-Mu††alib. The alliance had been concluded by his grandfather Màlik b. al-Qashib in pre-Islamic times.128 3) Ghazwàn b. 'Utba, son of the Companion 'Utba b. Ghazwàn, was an ally of the Qurayshì section Banù Nawfal. The alliance was pre-Islamic.129 127 Ibn Màkùlà, Al-Ikmàl fì raf ' al-irtiyàb 'an al-mu"talif wa-l-mukhtalif min al-asmà" wa-l-kunà wa-l-ansàb, 5 vols. (Beirut: Dàr al-Kutub al-'Ilmiyya, 1411/1991), III, 311, 392; Khalìfa, Ta"rìkh, I, 240. For the pre-Islamic alliance, see Ibn Hishàm (ed. Beirut), Sìra, I, 209; Lecker, The Banù Sulaym, The Max Shloessinger Memorial Series 4 ( Jerusalem: The Hebrew University of Jerusalem, 1989), 128–30. 128 Ibn Sa'd, ˇabaqàt, V, 177; Ibn Óajar al-'Asqalànì, Ißàba (ed. Cairo), II, 364 (s.v. 'Abdallàh b. Màlik). 129 Al-Bukhàrì, Al-Ta"rìkh al-kabìr, VII, 108; Ibn Óabìb, Munammaq, 243.

alliances in islam

35

4) Ya˙yà b. 'Abd al-Ra˙màn b. Óà†ib b. Abì Balta'a, who died in 104/722, was an ally of the Qurayshì section Asad. The alliance had been concluded in pre-Islamic times, perhaps by the grandfather Óa†ib.130 5) Sa'ìd b. Khalaf (or Khàlid) al-QàriΩì, who died after 130/747, was allied to the Qurayshì section Banù Zuhra. The alliance had been concluded in pre-Islamic times by the ancestor of the family, the poet Khàlid b. al-Óàrith.131 6) Ya'qùb b. 'Utba b. al-Mughìra b. al-Akhnas b. Sharìq, who died in the year 128/745, was an ally of the Banù Zuhra. His great grandfather al-Akhnas had been an ally of Zuhra, and enjoyed a high status in Mecca. Al-Akhnas’ sons, 'Umayr and al-Mughìra were allies of Zuhra as well.132 7) 'Abd al-Malik b. 'Abdallàh b. Abì Sufyàn, of the tribe Thaqìf, was an ally of Banù Zuhra. His grandfather was a brother of the Companion al-'Alà" b. Jàriya, who was an ally of Banù Zuhra and enjoyed a high status in Mecca.133 Al-'Alà", who preceded 'Abd al-Malik b. 'Abdallàh by two generations, was second or third generation to that particular alliance.134 8) 'Abdallàh b. Khabbàb b. al-Aratt, a Successor, was an ally of Banù Zuhra. The alliance originated in the servile status of alAratt in pre-Islamic times.135 9) 'Abdallàh b. al-Sà"ib of Kinda, who died 126/743, was an ally of the Prophet’s descent group, Banù Hàshim.136 According to Ibn Sa'd, however, his great grandfather, Sa'ìd b. Thumàma, had

Khalìfa b. Khayyà†, Al-ˇabaqàt, ed. Akram Îiyà" al-'Umarì (Riyadh: Dàr ˇayba, 1402/1982), 242; al-Mizzì, Tahdhìb al-kamàl, XXXI, 435; al-ˇabarì, Biographies of the Prophet’s Companions and Their Successors, The History of al-ˇabarì. An Annotated Translation, vol. 39, trans. and annot. Ella Landau-Tasseron (New York: State University of New York Press, 1998), 289. According to Ibn Óabìb, Munammaq, 253, Óà†ib is one of those who joined Quraysh without an alliance. 131 Khalìfa b. Khayya†, ˇabaqàt, I, 264; Ibn Óajar al-'Asqalànì, Tahdhìb, IV, 18. For the alliance, see Kister, “Strangers,” 139. 132 Al-Mizzì, Tahdhìb al-kamàl, XXXII, 350; Ibn Óibbàn, Al-Thiqàt, VII, 639; Ibn 'Abd al-Barr, Istì'àb, IV, 1444; Ibn Óajar al-'Asqalànì, Ißàba (ed. Beirut), IV, 710. Kister, “Strangers,” 127–33. 133 Al-Bukhàrì, Al-Ta"rìkh al-kabìr, ed. Hàshim al-Nadwì, 8 vols. (n.p., n.d.), V, 421; Kister, “Strangers,” 133. 134 Ibn Óabìb, Munammaq, 235. 135 Al-Bukhàrì, Al-Ta"rìkh al-kabìr, V, 78; al-Mubàrakfùrì, Tu˙fat al-a˙wadhì, VI, 331. On the father, Khabbàb, see also below Appendix E, and Kister, “Strangers,” 126–27. 130

36

ella landau-tasseron

been an ally not of Hàshim but of 'Abd Shams. His grandfather Yazìd b. Sa'ìd was known as Ibn Ukht al-Namir, ally of Banù 'Abd Shams in pre-Islamic times.137 10) The poet 'Abd al-Ra˙màn b. Ar†àt b. Say˙àn, who flourished during Mu'àwiya’s caliphate (40–60/660–680), was an ally of [the family of ] Óarb b. Umayya. So was his father, Ar†àt.138 11) 'Abdallàh b. 'Abdallàh b. 'Àmir b. Rabì'a al-'Anzì was an ally of the family of 'Umar b. al-Kha††àb. His grandfather had been ally and adopted son of al-Kha††àb.139 See also below, appendix E, on the Óa∂ramì family. Other allies in Islamic times seem also to be perpetuating pre-Islamic alliances: 1) Íàli˙ b. Óassàn of the Jewish tribe al-Na∂ìr, who was still alive at the time of the caliph al-Mahdì (died 169/785), was an ally of the tribe Aws. Al-Na∂ìr had been allied to Aws in pre-Islamic times.140 2) 'Abd al-Ra˙màn b. Sàlim, a judge and story-teller in Egypt who died in 143/760, was an ally of the tribe Jayshàn.141 His grandfather Sufyàn b. Hàni" (d. after 80/699) had already been allied to Jayshàn and had been known by their nisba.142 3) 'Abdallàh b. 'Uthmàn b. Khithyam (or Khuthaym) al-Qàrì, who died in 132/749, was an ally of the Qurayshì section Banù Zuhra,143 4) Ya'qùb b. 'Abd al-Ra˙màn b. Mu˙ammad b. 'Abdallàh b. 'Abd al-Qàrì l-Qurashì was an ally of Banù Zuhra,144 Khalìfa b. Khayyà†, ˇabaqàt, I, 258. Ibn Sa'd, ˇabaqàt, al-qism al-mutammim, I, 273. According to Ibn Óabìb, Munammaq, 250, this family was not formally allied to 'Abd Shams. 138 Al-Ißfahànì, Al-Aghànì, II, 237–38. 139 Ibn Óajar al-'Asqalànì, Ißàba (ed. Beirut), V, 19; Ibn Sa'd, ˇabaqàt, III, 386. 140 Khalìfa b. Khayyà†, ˇabaqàt, I, 274. On al-Na∂ìr and Aws, see V. Vacca, art. “al-Na∂ìr,” EI 2, VII, 852–53 (According to Vacca al-Na∂ìr were clients, but see above, section “˙alìf, 'adìd and mawlà”). 141 Ibn Màkùlà, Al-Ikmàl, II, 191. 142 Al-Mizzì, Tahdhìb al-kamàl, XI, 199; Ibn Óajar al-'Asqalànì, Tahdhìb, IV, 108; idem, Ißàba (Beirut), III, 260. 143 Ibn Óajar al-'Asqalànì, Tahdhìb, V, 275. 144 Al-Ißbahànì, Rijàl Muslim, ed. 'Abdallàh al-Laythì, 2 vols. (Beirut: Dàr alMa'rifa, 1407/1987), II, 372. According to Ibn Óabìb, Munammaq, 254, this family had marriage ties, but not an alliance with Banù Zuhra. 136

137

alliances in islam

37

5) Sufyàn b. 'Awf al-Qàrì, a Successor, was an ally of Banù Zuhra.145 The last-mentioned three belong to the section al-Qàra, which was allied to Quraysh in pre-Islamic times, and formed part of Mecca’s military force, the A˙àbìsh.146 Appendix B Alliances inaugurated in Islamic times: 1) Shura˙bìl b. Óasana, apparently left with no family after the death of his stepfather and stepbrothers, allied himself to the Qurayshì section Banù Zuhra. This happened during the caliphate of 'Umar b. al-Kha††àb. His son Rabì'a is also considered an ally of Zuhra.147 2) The Kùfan judge 'Abd al-Malik b. 'Umayr, who died in 136/753, was allied to the Qurayshì section Banù 'Adì and bore two nisbas, Lakhmì and Qurayshì.148 His brother, also called Qurayshì, was apparently included in the alliance.149 'Abd al-Malik originated in al-Óìra (southern Iraq), as indicated by the nisba Lakhmì. His origin far away from Mecca and the date of his death suggest that his alliance with the Quraysh originated not in preIslamic times but after the conquest of al-Óìra (in 13/634). 3) In the aftermath of the ridda wars various people emigrated to Medina and allied themselves with local descent groups.150

145 Ibn Óajar al-'Asqalànì, Ta'jìl al-manfa'a bi-zawà"id rijàl al-a"imma al-arba'a, ed. Ikràm Allàh Imdàd al-Óaqq (Beirut: Dàr al-Kitàb al-'Arabì, n.d.), I, 155. 146 Ibn Óabìb, Munammaq, 230. 147 Ibn 'Abd al-Barr, Istì'àb, II, 699; al-Mizzì, Tahdhìb al-kamàl, II, 413; Ibn 'Asàkir, Ta"rìkh, XXII, 471–72. 148 Ibn Sa'd, ˇabaqàt, VI, 315; al-Bukhàrì, Al-Ta"rìkh al-kabìr, V, 426; al-Kalàbàdhì, Rijàl ßa˙ì˙ al-Bukhàrì, II, 477; al-Bàjì, Al-Ta'dìl wa-l-tajrì˙, ed. Abù Lubàba Óusayn, 3 vols. (Riyadh: Dàr al-Liwà" li-l-Nashr wa-l-Tawzì', 1406/1986), II, 905; Ibn Óajar al-'Asqalànì, Lisàn al-mìzàn, ed. Dà"irat al-Ma'àrif al-NiΩàmiyya, 7 vols. (Beirut: Mu"assasat al-A'lamì li-l-Ma†bù'àt, 1406/1986), VII, 292; idem, Taqrìb al-tahdhìb, ed. Mu˙ammad 'Awwàma (Syria: Dàr al-Rashìd, 1406/1986), I, 364; al-Dhahabì, Siyar, V, 438. Ibn Óibbàn, Mashàhìr 'ulamà" al-amßàr, I, 110; idem, Al-Thiqàt, V, 116; al-'Ijlì, Ma'rifat al-thiqàt, ed. 'Abd al-'Alìm 'Abd al-'AΩìm al-Bustawì, 2 vols. (Medina: Maktabat al-Dàr, 1405/1985), II, 104. 149 Al-Bukhàrì, Al-Ta"rìkh al-kabìr, V, 160. 150 For example, Óußayn, Óißn and Sufyàn, three brothers from the tribe Judhàm, came to Medina after the ridda wars and allied themselves to the Anßàr, Ibn Óajar al-'Asqalànì, Ißàba (ed. Beirut), II, 174. See also Appendix D no. 2.

38

ella landau-tasseron

4) The caliph al-Mahdì (died 169/785) interfered in a case related to descendants of the royal family of Kinda. He transferred their alliance from the Qurayshì section Juma˙ to his own family, Banù l-'Abbàs. These Kindìs may have been still recognized as allies of the 'Abbàsid family in the lifetime of Ibn Sa'd (d. 245/859).151 In the following cases the time of the inauguration of the alliance cannot be ascertained: 1) The poet Yazìd b. al-Mufarrigh al-Óimyarì was allied to the family of Khàlid b. Asìd in the Umayyad period.152 2) A report on the battle of Fakhkh, which took place in the year 169/786 near Medina, was transmitted by an eyewitness, one alSariyy, who was an ally of the Qurayshì section Banù Zuhra.153 3) Ya˙yà b. Ya'mur al-'Adwànì, scribe of Yazìd b. al-Muhallab towards the end of the first century AH (beginning of the eighth century CE) was born in Ahwàz, apparently after the advent of Islam (because 'Adwàn would not have reached Ahwàz before Islam). He was an ally of the tribe Hudhayl.154 4) Among the casualties of the battle of Qudayd in the year 130/748 was one Mu˙ammad b. Muslim b. 'À"idh from the tribe of Hudhayl, an ally of Banù Munqidh.155 5) 'Ubayd b. Asbà† b. Mu˙ammad al-Hamdànì, a Kùfan who died in 205/820, was an ally of Quraysh.156

151 Ibn S'ad, ˇabaqàt, V, 13, says: “They are counted among them until this day,” but he may be reproducing an earlier authority. Al-Mizzì, Tahdhìb al-kamàl, XXIV, 129, says: “they are counted with Banù l-'Abbàs and their children are with Banù Juma˙,” which does not make much sense. 152 Al-Ißfahànì, Al- Aghànì, XVIII, 262, 281–94; al-Juma˙ì, ˇabaqàt al-shu'arà", ed. Joseph Hell (Leiden: Brill, 1916), 143–44; Ibn Qutayba, Al-Shi'r wa-l-shu'arà", ed. A˙mad Mu˙ammad Shàkir (Cairo, 1966), I, 360–65. 153 Al-ˇabarì, Ta"rìkh, IV, 601. 154 Al-ˇabarì, Ta"rìkh, III, 651. 155 Al-Bukhàrì, Al-Ta"rìkh al-ßaghìr, 2 vols. (Aleppo/Cairo: Dàr al-Wa'y and Maktabat Dàr al-Turàth, 1397/1977), II, 17. Banù Munqidh was a section of the Azdì tribe Daws, Ibn Óazm, Jamharat ansàb al-'arab, ed. 'Abd al-Salàm Mu˙ammad Hàrùn (Cairo: Dàr al-Ma'àrif, 1382/1962), 380. 156 Ibn Óibbàn, Al-Thiqàt, VIII, 432–33.

alliances in islam

39

Appendix C Màlik b. Anas and his family: Màlik and members of his family were known as allies of Banù Taym.157 Ibn Óazm, recording the names of Màlik’s cousins and other descendants of the family who lived in the second half of the second century, adds: “they were allies of Banù Taym.”158 Examples of clear references to members of the family as allies of Taym (or of 'Uthmàn b. 'Ubaydallàh al-Taymì) are the following: 1) Nàfi' b. Màlik b. Abì 'Àmir al-Aßba˙ì l-Madanì l-Taymì, died around 132/749.159 2) 'Abdallàh b. 'Abdallàh b. Uways b. Màlik b. Abì 'Àmir al-Aßba˙ì, died in 169/785. He was a cousin (ibn 'amm) of Màlik and married Màlik’s sister.160 3) 'Abd al-Óamìd b. Abì Uways (whose name was 'Abdallàh) b. Uways b. Màlik b. Abì 'Àmir, brother of Ismà'ìl b. Abì Uways al-Aßba˙ì, died in 202/817. He was Màlik’s nephew (ibn ukht).161 4) His brother Ismà'ìl b. 'Abdallàh b. 'Abdallàh b. Uways b. Màlik b. Abì 'Àmir, Abù 'Abdallàh, died in 226/840.162 5) 'Abd al-Óamìd b. 'Abdallàh b. Abì Uways al-Aßba˙ì, died in 230/844.163 Appendix D 'adìd 'adìd (“counted with”) and ˙alìf (guest ally) interchangeably applied: 1) The family of Màlik b. Anas “were counted with Banù Taym, they were allies of 'Uthmàn b. 'Ubaydallàh al-Taymì.”164 Uways,

157 Al-Kalàbàdhì, Rijàl ßa˙ì˙ al-Bukhàrì, II, 693; al-Ißbahànì, Rijàl Muslim, II, 220; Ibn 'Abd al-Barr, Al-Tamhìd, I, 89; al-Dhahabì, Siyar, VIII, 49. 158 Ibn Óazm, Jamharat ansàb, 436. 159 Al-Ißbahànì, Rijàl Muslim, II, 290; Ibn Óajar al-'Asqalànì, Tahdhìb, X, 366; al-Mizzì, Tahdhìb al-kamàl, XXIX, 290. 160 Al-Bukhàrì, Al-Ta"rìkh al-ßaghìr, II, 178; al-Baghdàdì, Ta"rìkh Baghdàd, 14 vols. (Beirut: Dàr al-Kutub al-'Ilmiyya, n.d.), X, 5. 161 Al-Kalàbàdhì, Rijàl ßa˙ì˙ al-Bukhàrì, II, 482. 162 Al-Bukhàrì, Al-Ta"rìkh al-ßaghìr, II, 354; al-Mizzì, Tahdhìb al-kamàl, III, 124. 163 Al-Ißbahànì, Rijàl Muslim, I, 441. 164 Al-Mizzì, Tahdhìb al-kamàl, XXVII, 93.

ella landau-tasseron

40

2)

3) 4)

5)

uncle of Màlik b. Anas, is sometimes referred to as 'adìd of Taym.165 The sons of al-Íalt who emigrated to Medina after the ridda “concluded an alliance with Banù Juma˙, they were registered in the Dìwàn with them and were counted with them” (kàna dìwànuhum wa-da'watuhum ma'ahum). In the next generation, al-Íalt b. Zubayd b. al-Íalt, judge in Medina, was still considered an ally of Juma˙.166 Yazìd b. Mufarrigh is defined once as an 'adìd, another time as a ˙alìf.167 Salama, brother of Ya'là b. Umayya who had been an ally of Banù 'Abd Shams in the Jàhiliyya, was “counted with” the Meccans in Islamic times ('idàduhu fì ahl makka).168 The judge Shuray˙, ally of Kinda who acquired a genealogy within that tribe, is sometimes defined as 'adìd.169

Indications of a distinction between the two categories: 1) Ibn Óabìb, in his efforts to reduce the number of the ˙alìfs of Quraysh, says that certain persons were 'adìds, not allies.170 2) Al-Shàfi'ì implies a distinction between the two categories when he argues that neither an ally nor an 'adìd are entitled to bloodmoney from their hosts.171 3) The same distinction is implied by 'Umar’s decree about inheritance, said to be applicable to ˙alìf and to 'adìd.172 4) 'Abd al-Ra˙màn b. Muljam al-Muràdì ('Alì’s assassin), originally from Óimyar, “was counted with” Muràd while he was allied to Banù Jabala of Kinda.173 It is worthy of note that the weakest of his ties was presumably those with Muràd, yet his nisba is coined after that tribe. Ibn Óajar al-'Asqalànì, Tahdhìb, I, 337; al-Nasà"ì, Al-Sunan al-kubrà, II, 65. Ibn Sa'd, ˇabaqàt, V, 13; and idem, Al-Qism al-mutammim, I, 280; al-Mizzì, Tahdhìb al-kamàl, XXIV, 129. 167 See above, note 152. 168 Ibn Óibbàn, Al-Thiqàt, III, 166. I am not sure whether “being counted” as an inhabitant of a certain city relates to the dìwàn or to the classification of scholars in the framework of 'ilm al-rijàl; but cf. al-Sharìd b. Suwayd, from Hamdàn or Óa∂ramawt, ally of Thaqìf and “counted with the people of ˇà"if ”; ibid., III, 188. 169 Al-Mizzì, Tahdhìb al-kamàl, XII, 436. 170 See above, “Alliances in Islamic times.” 171 Al-Shàfi'ì, Al-Umm, VI, 116, cf. al-Shaybànì, Mabsù†, IV, 663 (li-anna l-qawm kàna ya'qil ma'ahum ˙alìfuhum wa-'adìduhum). 172 'Abd al-Razzàq, Mußannaf, V, 305 (man kàna ˙alìfan aw 'adìdan . . .). On inheritance, see above, “Islamic objections.” 173 Al-Nawawì, Tahdhìb al-asmà", I, 319. 165 166

alliances in islam

41

5) Óalìf, 'adìd and mawlà are listed as separate categories of people attached to descent groups in the Jàhiliyya.174 Appendix E ˙alìf and mawlà: The terms interchangeably applied: 1) The expression mawlà mawàliya in a mocking verse by al-Farazdaq is explained as “'Abdallàh b. Abì Is˙àq, mawlà of the Óa∂ramì family who were allied to Banù 'Abd Shams.” Baghdàdì refers to this family, as well as to 'Abdallàh, as ˙ulafà" and mawàlì, interchangeably.175 2) 'Abdallàh b. 'Abdallàh b. 'Àmir b. Rabì'a l-'Anzì was “an ally of the family of 'Umar b. al-Kha††àb, a mawlà of that family” (˙alìf àl 'Umar b. al-Kha††àb mawlàhum.176 His grandfather 'Àmir had been an ally and an adopted son of al-Kha††àb.177 3) Sàlim was “a mawlà of Abù Óudhayfa, he was his ally.”178 4) The Tamìmì poet 'Utayba b. Mirdàs refers to himself as mawlà of Jamìl b. Ma'mar, when he was an ally of that person.179 5) It is said of Abù l-Badà˙ b. 'Àßim al-Balawì l-Anßàrì that he was “a mawlà of the Anßàr, there is no doubt about it, because his descent group, Baliyy, was allied to Banù 'Amr b. 'Awf from the Anßàr.”180 6) 'Amr b. 'Awf was a mawlà of Suhayl b. 'Amr according to one version, an ally of [Suhayl’s descent group], Banù 'Àmir b. Lu'ayy, according to another.181 7) Mu'ayqìb b. Abì Fà†ima, “ally of Banù 'Abd Shams, and some say: mawlà of Sa'ìd b. al-'Àß.”182 174 Al-Sarakhsì, Mabsù†, XXVII, 125. Al-Sarakhsì may have made a mistake in dating here. 175 Al-Baghdàdì, Khizànat al-adab, I, 236. 176 Ibn Óajar al-'Asqalànì, Ißàba (ed. Beirut), V, 19. 177 Ibn Sa'd, ˇabaqàt, III, 386. 178 Muslim, Ía˙ì˙, II, 1076. Sàlim is interchangeably defined as ˙alìf and mawlà, in different versions of one and the same story. See also above, note 94, and below, “Freedmen who became allies” no. 3. 179 Al-Ißfahànì, Al-Aghànì, XXII, 23. 180 Al-Zurqànì, Shar˙ al-Zurqànì 'alà Muwa††a" al-imàm Màlik, 4 vols. (Beirut: Dàr al-Kutub al-'Ilmiyya, 1411/1911), II, 493. 181 Al-Bukhàrì, Al-Mu'jam al-kabìr, XVII, 24. 182 Al-Îa˙˙àk, Al-À˙àd wa-l-mathànì, ed. Bàsil Fayßal A˙mad al-Jawàbira, 6 vols.

42

ella landau-tasseron

8) Sa'd b. Khawla, a Companion of the Prophet, is alternatively said to have been a member, an ally, and a mawlà of Banù 'Àmir b. Lu'ayy. He is also said to have been a Persian (in which case he could not have been a member of the tribe).183 9) The Successor Abù 'Alqama, judge in Ifrìqiyà, is alternatively said to have been a mawlà of Banù Hàshim, of Ibn 'Abbàs, and a ˙alìf of Banù Hàshim. He is also said to have been a Persian.184 The last two cases are perhaps exceptions to the rule that attached Arabs were ˙ulafà" whereas attached non-Arabs were mawàlì. Clients of ˙alìfs: 1) Khabbàb b. al-Aratt from Tamìm was a mawlà, a freedman, of a Khuzà'ì woman, ally of the Qurayshì section Zuhra. He was therefore “Tamìmì by genealogy, Khuzà'ì by walà", Zuhrì by alliance.”185 2) 'Abdallàh b. Abì Labìd was a client of al-Akhnas b. Sharìq, who was an ally of Banù Zuhra.186 3) Sa'ìd b. Abì Hind was a mawlà of the Fazàrì Samura b. Jundab, who was an ally of Banù Khudra, a section of Khazraj.187 4) Is˙àq b. Bakr was a mawlà of Shura˙bìl b. Óasana, who was an ally of the Qurayshì section Banù Zuhra.188

(Riyadh: Dàr al-Ràya, 1411/1991), I, 237; al-Kalàbàdhì, Rijàl ßa˙ì˙ al-Bukhàrì, II, 729; al-Bàjì, Al-Ta'dìl wa-l-tajrì˙, II, 747. 183 Ibn Óajar al-'Asqalànì, Ißàba (ed. Beirut), III, 53; idem, Fat˙ al-bàrì, V, 364. 184 Al-Mubàrakfùrì, Tu˙fat al-a˙wadhì, VIII, 293; Ibn Óajar al-'Asqalànì, Tahdhìb, XII, 191; al-Mizzì, Tahdhìb al-kamàl, XXXIV, 101. 185 There are several versions of Khabbàb’s status: an ally or a client of Zuhra, a client of a client of an ally of Zuhra, etc. See al-Óàkim al-Naysàbùrì, Al-Mustadrak 'alà al-ßa˙ì˙àn, III, 429–30; al-Nawawì, Tahdhìb al-asmà", I, 175; Ibn 'Abd al-Barr, Istì'àb, II, 437–38; Kister, “Strangers,” 126–27. Khabbàb’s son was considered an ally, see above Appendix A. 186 Al-Kalàbàdhì, Rijàl ßa˙ì˙ al-Bukhàrì, II, 861; Ibn Sa'd, ˇabaqàt, al-qism al-mutammim, I, 331; al-Mizzì, Tahdhìb al-kamàl, XV, 483. 'Abdallàh died in the time of the caliph al-Manßùr (d. 158/775). 187 Al-Kalàbàdhì, Rijàl ßa˙ì˙ al-Bukhàrì, I, 299; al-Dhahabì, Al-Kàshif, ed. Mu˙ammad 'Awwàma ( Judda: Dàr al-Qibla li-l-Thaqàfa al-Islàmiyya, 1413/1992), I, 445; alMizzì, Tahdhìb al-kamàl, XI, 93. His son belonged to the same status, Ibn Óajar al-'Asqalànì, Lisàn al-mìzàn, VII, 263. 188 Is˙àq b. Bakr b. Mu∂ar b. Mu˙ammad b. Óakìm b. Salmàn. He died in 118/736. Ibn Óibbàn, Al-Thiqàt, VIII, 113; al-Bukhàrì, Al-Ta"rìkh al-kabìr, I, 383; Ibn Óajar al-'Asqalànì, Taqrìb al-tahdhìb, I, 100; al-Mizzì, Tahdhìb al-kamàl, II, 413.

alliances in islam

43

5) Nàfi' b. 'Abd al-Ra˙màn was a mawlà of Ja'wana b. Sha'ùb, an ally of 'Abd al-Mu††alib.189 6) Ibràhìm b. 'Abdallàh was a mawlà of al-Qàra, allies of Banù Zuhra.190 7) Nàfi' b. Abì Nu'aym, from Ißfahàn, was a mawlà of Ja'wana b. Sha'ùb al-Laythì, ally of Óamza, uncle of the Prophet.191 8) Ayyùb al-Sakhtiyànì (d. 131/748) was a mawlà of [a section of ] Juhayna, allies of Banù l-Óarìsh.192 9) Khàlid b. 'Urfu†a, ally of Banù Zuhra, had a mawlà by the name of Muslim.193 10) 'Ubaydallàh b. Abì Yazìd was a mawlà of the Kinànì family of QàriΩ, allies of Banù Zuhra.194 Freedmen who became allies: 1) The Yarbù'ì Wàqid b. 'Abdallàh was bought in the Jàhiliyya and adopted by 'Umar’s father, al-Kha††àb. This case is analogous to that of Zayd b. Óàritha who was a freed slave, an adopted son and a mawlà of the Prophet. But Wàqid became an adopted son and an ally, not a mawlà, of al-Kha††àb.195 2) Al-Azraq b. 'Uqba was a slave of al-Óàrith b. Kalada in ˇà"if. During the siege of ˇà"if he escaped to Mu˙ammad who freed him and handed him over to Khàlid b. Sa'ìd b. al-'Àß, of the 189

Al-Mizzì, Tahdhìb al-kamàl, XXIX, 281, see also below no. 7. Ibn Màkùlà, Al-Ikmàl, IV, 140. 191 Al-Dhahabì, Siyar, VII, 336; Ibn Óibbàn, Mashàhìr, I, 141. Nàfi' died in 169/785. Ja'wana lived in the Prophet’s time and in spite of his relations with the Prophet’s uncle, he never met the Prophet, Ibn Óajar al-'Asqalànì, Ißàba (ed. Cairo), I, 262. 192 Al-Mizzì, Tahdhìb al-kamàl, III, 457. Ayyùb died in 131/748. 193 Al-Bukhàrì, Al-Ta"rìkh al-kabìr, VII, 260; Ibn Óajar al-'Asqalànì, Ißàba (ed. Cairo), I, 410. 194 Al-Bukhàrì, Al-Ta"rìkh al-kabìr, V, 403; al-Kalàbàdhì, Rijàl ßa˙ì˙ al-Bukhàrì, I, 471. 195 Ibn Hishàm (ed. Beirut), Sìra, II, 95. According to Ibn Óazm, the case of the Companion al-Miqdàd, who was adopted by al-Aswad, is the same; he was a slave because his mother was a slave-girl, Ibn Óazm, Al-Mu˙allà, XI, 89. Al-Miqdàd, however, is usually defined as an ally, and the status of his mother is not mentioned by the other sources which I consulted. According to Ibn Óajar al-'Asqalànì, Ißàba (ed. Cairo), III, 454–55, his mother was from Kinda. There is also a story that al-Miqdàd himself was an Abyssinian slave, e.g., al-Ißbahànì, Rijàl Muslim, II, 268, but Ibn 'Abd al-Barr, Istì'àb, IV, 1480, rejects this. Most sources record alMiqdàd’s pedigree in the Arab tribe Bahrà", see further Ibn Óabìb, Munammaq, 363; Ibn Óibbàn, Mashàhìr, I, 24; idem, Al-Thiqàt, III, 371; Ibn Óajar al-'Asqalànì, Tahdhìb, X, 254; al-Mizzì, Tahdhìb al-kamàl, XXVIII, 452. 190

44

ella landau-tasseron

Umayyad family. Al-Azraq, a freedman, did not become a mawlà but an ally of the Umayyads.196 3) Sàlim b. Ma'qil was a freedman, adopted in the Jàhiliyya by Abù Óudhayfa, who gave him his niece in marriage.197 Sàlim is usually known as a mawlà of the Companion Abù Óudhayfa, but some scholars insist that he was an ally.198 4) The Óa∂ramì family: the ancestor of the family was a Persian slave brought to Mecca from Óa∂ramawt. Ibn Óabìb denies that he and his descendants had an allied status in Mecca.199 Nevertheless, the Óa∂ramìs are commonly known as allies of the Umayyads, both before and after Islam. Among them were 'Abdallàh b. 'Amr al-Óa∂ramì, born during Mu˙ammad’s lifetime;200 'Abdallàh b. Mas'ùd b. Sa'ìd al-Óa∂ramì, active in 60/680;201 'Abd al-Ra˙màn b. 'Abd al-Ra˙màn b. 'Abdallàh b. 'Àmir al-Óa∂ramì, appointed by al-Óajjàj over Kùfa during the revolt of Ibn al-Ash'ath;202 seven and nine generations after al-Óa∂ramì his descendants, Abù 'Abbàd and his grandson, were still considered allies of Banù 'Abd Shams.203 See also above, “clients of ˙alìfs no. 1 and note 185; also above, section ˙alìf, 'adìd and mawlà, on Íuhayb b. Sinàn.

196 Ibn Óajar al-'Asqalànì, Ißàba (ed. Beirut), I, 46. Ibn Óajar narrates that he became ˙alìf but calls him mawlà. The family’s claim to alliance may indeed have been false, see ibid.; al-Balàdhurì, Ansàb, I, 157. 197 Ibn Óajar al-'Asqalànì, Ißàba (ed. Cairo), II, 6–7; al-Bayhaqì, Al-Sunan al-kubrà, VII, 137. See above, note 94. 198 Ibn Óajar al-'Asqalànì, Fat˙ al-bàrì, IX, 133; al-'AΩìmàbàdì, 'Awn al-ma'bùd, VI, 44. Al-Bayhaqì, Al-Sunan al-kubrà, VII, 137 notes: “This is a Qurayshì woman given in marriage to a freedman” ( fahadhihi qurashiyya min banì 'Abd Shams b. 'Abd Manàf zuwwijat min mawlà). According to one of the versions, he had been manumitted by his owner, a Medinan woman, and subsequently allied himself to Abù Óudhayfa, Ibn Qutayba, Tà"wìl mukhtalif al-˙adìth, I, 122. 199 Ibn Óabìb, Munammaq, 264–65. 200 Al-Nawawì, Tahdhìb al-asmà", I, 264. 201 Al-ˇabarì, Ta"rìkh, III, 279. 202 Al-ˇabarì, Ta"rìkh, III, 629. 203 Ibn Màkùlà, Al-Ikmàl, IV, 285.

alliances in islam

45

Bibliography Primary Sources 'Abd al-Razzàq, Mu˙ammad b. Hammàm al-Ían'ànì. Al-Mußannaf, ed. Óabìb alRa˙màn al-A'Ωamì, 11 vols. Beirut: al-Maktab al-Islàmì, 1403/1983. Abù Dà"ùd, Sulaymàn b. al-Ash'ath. Sunan Abì Dà"ùd, ed. Mu˙ammad Mu˙yì alDìn 'Abd al-Óamìd, 4 vols. n.p.: Dàr al-Fikr, n.d. Abù Nu'aym al-Ißbahànì, A˙mad b. 'Abdallàh. Óilyat al-awliyà" wa-†abaqàt al-aßfiyà". 10 vols. Beirut: Dàr al-Kitàb al-'Arabì, 1405/1985. al-'AΩìmàbàdì, Shams al-Óaqq. 'Awn al-ma'bùd. 10 vols. Beirut: Dàr al-Kutub al'Ilmiyya, 1415/1995. al-Baghdàdì, 'Abd al-Qàdir. Khizànat al-adab wa-lubb lubàb lisàn al-'arab, ed. 'Abd alSalàm Mu˙ammad Hàrùn, 13 vols. Cairo: Dàr al-Kàtib al-'Arabì li-l-ˇibà' a wal-Nashr and Maktabat al-Khànjì, 1967–87. al-Baghdàdì, al-Kha†ìb A˙mad b. 'Abdallàh. Ta"rìkh Baghdàd. 14 vols. Beirut: Dàr al-Kutub al-'Ilmiyya, n.d. al-Bàjì, Sulaymàn b. Khalaf. Al-Ta'dìl wa-l-tajrì˙, ed. Abù Lubàba Óusayn, 3 vols. Riyadh: Dàr al-Liwà" li-l-Nashr wa-l-Tawzì', 1406/1986. al-Balàdhurì, A˙mad b. Ya˙yà. Ansàb al-ashràf, ed. Mu˙ammad Óamìdullàh, vol. 1. Cairo: Ma'had al-Makh†ù†àt and Dàr al-Ma'àrif, 1959. al-Bayhaqì, A˙mad b. al-Óusayn. Al-Sunan al-kubrà, ed. Mu˙ammad 'Abd al-Qàdir 'A†à, 10 vols. Mecca: Maktabat Dàr al-Bàz, 1414/1994. al-Bu˙turì, Abù 'Ubàda. Ash'àr al-˙amàsa (Hamasae Carmina), ed. G.G. Freytag. Bonn, 1828. al-Bukhàrì, Mu˙ammad b. Ismà'ìl. Al-Ta"rìkh al-kabìr, ed. Hàshim al-Nadwì, 8 vols. n.d. id., Al-Ta"rìkh al-ßaghìr. 2 vols. Aleppo/Cairo: Dàr al-Wa'y and Maktabat Dàr alTuràth, 1397/1977. id., Ía˙ì˙ al-Bukhàrì. ed. Mu߆afà Dìb al-Baghà, 6 vols. Beirut: Dàr Ibn Kathìr, 1407/1987. al-Îa˙˙àk, A˙mad b. Mu˙ammad b. 'Amr. Al-À˙àd wa-l-mathànì, ed. Bàsil Fayßal A˙mad al-Jawàbira, 6 vols. Riyadh: Dàr al-Ràya, 1411/1991. al-Daqìqì, Sulaymàn b. Banìn b. Khalaf. Ittifàq al-mabànì wa-iftiràq al-ma'ànì, ed. Ya˙yà 'Abd al-Ra"ùf Jabr. Amman: Dàr 'Ammàr, 1405/1985. al-Dàraqu†nì, 'Alì b. 'Umar. Su"àlàt Óamza, ed. Muwaffaq b. 'Abdallàh b. 'Abd alQàdir. Riyadh: Maktabat al-Ma'àrif, 1404/1984. al-Dàrimì, 'Abdallàh b. 'Abd al-Ra˙màn. Sunan al-Dàrimì. Beirut: Dàr al-Kitàb al'Arabì, 1407/1987. al-Dhahabì, Mu˙ammad b. A˙mad. Al-Kàshif, ed. Mu˙ammad 'Awwàma. Judda: Dàr al-Qibla li-l-Thaqàfa al-Islàmiyya, 1413/1992. id., Ma'rifat al-qurrà" al-kibàr, ed. Shu'ayb al-Arna"ù†, Bashshàr 'Awwàd Ma'rùf and Íàli˙ Mahdì 'Abbàs, 2 vols. Beirut: Mu"assasat al-Risàla, 1404/1984. id., Siyar a'làm al-nubalà", ed. Shu'ayb al-Arna"ù† and Mu˙ammad Na'ìm al-'Arqasùsì, 23 vols. Beirut: Mu"assasat al-Risàla, 1413/1993. al-Fàkihì, Mu˙ammad b. Is˙àq. Akhbàr Makka, ed. 'Abd al-Malik Duhaysh, 6 vols. Beirut: Dàr Kha∂r, 1414/1994. al-Óàkim al-Naysàbùrì, Mu˙ammad b. 'Abdallàh. Al-Mustadrak 'alà al-ßa˙ì˙àn, ed. Mu߆fà 'Abd al-Qàdir 'A†à, 4 vols. Beirut: al-Maktaba al-'Ilmiyya, 1410/1990. al-Óanafì, Yùsuf b. Mùsà. Mu'taßar al-mukhtaßar. 2 vols. Beirut/Cairo: 'Àlam al-Kutub and Maktabat al-Mutanabbì, n.d. al-Óàrith b. Abì Usàma see al-Haythamì. al-Haythamì, 'Alì b. Abì Bakr, and al-Óàrith b. Abì Usàma. Bughyat al-bà˙ith 'an zawà"id musnad al-Óàrith, ed. Óusayn A˙mad Íàli˙ al-Bàkirì, 2 vols. Medina: Markaz Khidmat al-Sunna wa-l-Sìra al-Nabawiyya, 1412/1992.

46

ella landau-tasseron

id., Majma' al-zawà"id wa-manba' al-fawà"id. 10 vols. Cairo/Beirut: Dàr al-Rayyàn lil-Turàth and Dàr al-Kitàb al-'Arabì, 1407/1987. id., Mawàrid al-Ωam"àn ilà zawà"id Ibn Óibbàn, ed. Mu˙ammad 'Abd al-Razzàq Óamza. Beirut: Dàr al-Kutub al-'Ilmiyya, n.d. al-Óumaydì, 'Abdallàh b. al-Zubayr. Musnad al-Óumaydì, ed. Óabìb al-Ra˙màn alA'Ωamì, 2 vols. Beirut/Cairo: Dàr al-Kutub al-'Ilmiyya and Maktabat al-Mutanabbì, n.d. Ibn 'Abd al-Barr, Yùsuf b. 'Abdallàh al-Namarì al-Qur†ubì. Al-Istì'àb fì ma'rifat alaß˙àb, ed. 'Alì Mu˙ammad al-Bijàwì, 4 vols. Beirut: Dàr al-Jìl, 1412/1992. id., Al-Tamhìd, ed. Mu˙ammad 'Abd al-Kabìr al-Bakrì and Mu߆afà b. A˙mad al'Alawì, 24 vols. Morocco: Wizàrat 'Umùm al-Awqàf wa-l-Shu"ùn al-Islàmiyya, 1387/1967. Ibn Abì Óàtim al-Ràzì. Al-'Ilal, ed. Mu˙ibb al-Dìn al-Kha†ìb, 2 vols. Beirut: Dàr al-Ma'rifa, 1405/1985. Ibn Abì Shayba, 'Abdallàh b. Mu˙ammad. Al-Mußannaf, ed. Kamàl Yùsuf al-Óùt, 7 vols. Riyadh: Maktabat al-Rushd, 1409/1989. Ibn al-'Adìm, Kamàl al-Dìn 'Umar b. A˙mad. Bughyat al-†alab fì ta"rìkh Óalab, ed. Suhayl Zakkàr, 12 vols. Beirut: Dàr al-Fikr, 1988. Ibn 'Asàkir, 'Alì b. al-Óasan. Ta"rìkh madìnat Dimashq, ed. 'Umar b. Gharàma al'Amrawì, 70 vols. Beirut: Dàr al-Fikr, n.d. Ibn al-Athìr, al-Mubàrak b. Mu˙ammad. Al-Nihàya fì gharìb al-athar, ed. ˇàhir A˙mad al-Zàwì and Ma˙mùd Mu˙ammad al-ˇanàjì, 5 vols. Beirut: al-Maktaba al'Ilmiyya, 1399/1979. Ibn al-Athìr, Mu˙ammad b. Mu˙ammad 'Izz al-Dìn. Al-Kàmil fì al-ta"rìkh, ed. Abù al-Fidà" 'Abdallàh al-Qà†ì, 10 vols. Beirut: Dàr a-Kutub al-'Ilmiyya, 1415/1995. Ibn Far˙ùn, Ibràhìm b. 'Alì. Al-Dìbàj al-mudhhab fì ma'rifat a'yàn 'ulamà" al-madhhab. Beirut: Dàr al-Kutub al-'Ilmiyya, n.d. Ibn Óabìb, Mu˙ammad. Al-Munammaq fi akhbàr Quraysh, ed. Khùrshìd A˙mad Fàriq. Beirut: 'Àlam al-Kutub, 1405/1985. Ibn Óajar al-'Asqalànì, A˙mad b. 'Alì. Al-Ißàba fì tamyìz al-ßa˙àba. 4 vols. Cairo: Ma†ba'at al-Sa'àda, 1328/1910. id., Al-Ißàba fì tamyìz al-ßa˙àba, ed. 'Alì Mu˙ammad al-Bijàwì, 8 vols. Beirut: Dàr al-Jìl, 1412/1992. id., Fat˙ al-bàrì shar˙ ßa˙ì˙ al-Bukhàrì, ed. Mu˙ammad Fu"àd 'Abd al-Bàqì and Mu˙ibb al-Dìn al-Kha†ìb, 13 vols. Beirut: Dàr al-Ma'rifa, 1379/1959. id., Muqaddimat Fat˙ al-Bàrì, ed. Mu˙ammad Fu"àd 'Abd al-Bàqì and Mu˙ibb alDìn al-Kha†ìb, Beirut: Dàr al-Ma'rifa, 1379/1959. id., Lisàn al-mìzàn, ed. Dà"irat al-Ma'àrif al-NiΩàmiyya, 7 vols. Beirut: Mu"assasat al- A'lamì li-l-Ma†bù'àt, 1406/1986. id., “Nukhbat al-fikr,” in Kitàb subul al-salàm. Beirut: Dàr I˙yà" al-Turàth al- 'Arabì, [1379 AH]. id., Ta'jìl al-manfa'a bi-zawà"id rijàl al-a"imma al-arba'a, ed. Ikràm Allàh Imdàd alÓaqq. Beirut: Dàr al-Kitàb al-'Arabì, n.d. id., Taghlìq al-ta'lìq, ed. Sa'ìd 'Abd al-Ra˙màn al-Qazaqì, 5 vols. Beirut/Amman: al-Maktab al-Islamì and Dàr 'Ammàr, 1405/1985. id., Tahdhìb al-tahdhìb. 14 vols. Beirut: Dàr al-Fikr, 1404/1984. id., Taqrìb al-tahdhìb, ed. Mu˙ammad 'Awwàma. Syria: Dàr al-Rashìd, 1406/1986. Ibn Óanbal, A˙mad. Musnad A˙mad. 6 vols. Cairo: Mu"assasat Qur†uba, n.d. Ibn Óazm, 'Alì b. A˙mad. Al-Mu˙allà, ed. Lajnat I˙yà" al-Turàth al-'Arabì, 11 vols. Beirut: Dàr al-Àfàq al-Jadìda, n.d. id., Jamharat ansàb al-'arab, ed. 'Abd al-Salàm Mu˙ammad Hàrùn. Cairo: Dàr alMa'àrif, 1382/1962. Ibn Óibbàn, Mu˙ammad b. Óibbàn al-Bustì. Ía˙ì˙ Ibn Óibbàn, ed. Shu'ayb alArna"ù†, 18 vols. Beirut: Mu"assasat al-Risàla, 1414/1993.

alliances in islam

47

id., Al-Thiqàt, ed. Sharaf al-Dìn A˙mad, 9 vols. Dàr al-Fikr, 1395/1975. id., Mashàhìr 'ulamà" al-amßàr. ed. M. Fleischhammer. Beirut: Dàr al-Kutub al'Ilmiyya, 1959. Ibn Hishàm, 'Abd al-Malik. Al-Sìra al-nabawiyya, ed. ˇahà 'Abd al-Ra"ùf Sa'd, 6 vols. Beirut: Dàr al-Jìl, 1411/1991. id., Al-Sìra al-nabawiyya, ed. Mu߆afà al-Saqà, 'Abd al-ÓafìΩ Shalabì, Ibràhìm alAbyàrì, 4 vols. Cairo: Maktabat Mu߆afà al-Bànì wa-Awlàduhu, 1355/1936. Ibn Jamà'a, Mu˙ammad b. Ibràhìm. Al-Manhal al-rawiyy fì mukhtaßar 'ulùm al-˙adìth al-nabawì, ed. Mu˙yì al-Dìn 'Abd al-Ra˙màn Rama∂àn. Damascus: Dàr al-Fikr, 1406/1986. Ibn al-Jawzì, 'Abd al-Ra˙màn b. 'Alì. Íafwat al-ßafwa, ed. Ma˙mùd Fàkhùrì and Mu˙ammad Rawàs Qal'ajì, 4 vols. Beirut: Dàr al-Ma'rifa, 1399/1979. id., Zàd al-masìr fì 'ilm al-tafsìr. 8 vols. Beirut: al-Maktab al-Islàmì, 1404/1984. Ibn Kathìr, Ismà'ìl b. 'Umar. Tafsìr. Beirut: Dàr al-Fikr, 1401/1981. id., Al-Bidàya wa-l-nihàya. 14 vols. Beirut: Maktabat al-Ma'àrif, n.d. Ibn Khallikàn, A˙mad b. Mu˙ammad. Wafayàt al-a'yàn wa-anbà" abnà" al-zamàn, ed. I˙sàn 'Abbàs, 8 vols. Beirut: Dàr al-Thaqàfa, 1969–72. Ibn Màkùlà, 'Alì b. Hibat Allàh. Al-Ikmàl fì raf ' al-irtiyàb 'an al-mu"talif wa-l-mukhtalif min al-asmà" wa-l-kunà wa-l-ansàb. 5 vols. Beirut: Dàr al-Kutub al-'Ilmiyya, 1411/1991. Ibn ManΩùr, Mu˙ammad b. Makram. Lisàn al-'Arab. 15 vols. Beirut: Dàr Íàdir, n.d. Ibn Nujaym, Zayn b. Ibràhìm. Al-Ba˙r al-rà"iq shar˙ kanz al-daqà"iq. 7 vols. Beirut: Dàr al- Ma'rifa, n.d. Ibn Qàni', 'Abd al-Bàqì. Mu'jam al-ßa˙àba, ed. Íalà˙ b. Sàlim al-Mißràtì, 3 vols. Medina: Maktabat al-Gharbà" al-Athariyya, 1418/1998. Ibn al-Qaysarànì, Mu˙ammad b. ˇàhir. Tadhkirat al-˙uffàΩ, ed. Óamdì 'Abd alMajìd Ismà'ìl al-Salafì, 4 vols. Riyadh: Dàr al-Íamì'ì, 1415/1995. Ibn Qudàma al-Maqdisì, 'Abdallàh b. A˙mad. Al-Mughnì 'alà mukhtaßar al-Khiraqì. 10 vols. Beirut, 1405/1985. Ibn Qutayba, 'Abdallàh b. Muslim. Ta"wìl mukhtalif al-˙adìth, ed. Mu˙ammad Zuhrì al- Najjàr. Beirut: Dàr al-Jìl, 1393/1972. id., Al-Shi'r wa-l-shu'arà", ed. A˙mad Mu˙ammad Shàkir. Cairo, 1966. Ibn Sa'd, Mu˙ammad. Al-ˇabaqàt al-kubrà. 8 vols. Beirut: Dàr Íadir, n.d. id., Al-ˇabaqàt al-kubrà, al-qism al-mutammim, ed. Ziyàd Mu˙ammad Manßùr. Medina, Maktabat al-'Ulùm wa-l-Óikam. al-'Ijlì, A˙mad b. 'Abdallàh. Ma'rifat al-thiqàt, ed. 'Abd al-'Alìm 'Abd al-'AΩìm alBustawì, 2 vols. Medina: Maktabat al-Dàr, 1405/1985. al-Ißbahànì, A˙mad b. 'Alì b. Minjawayh. Rijàl Muslim, ed. 'Abdallàh al-Laythì, 2 vols. Beirut: Dàr al-Ma'rifa, 1407/1987. al-Ißfahànì, Abù al-Faraj. Kitàb al-aghànì, ed. Samìr Jàbir, 24 vols. Beirut: Dàr alFikr, n.d. al-Jaßßàß, Abù Bakr A˙mad b. 'Alì. A˙kàm al-Qur"àn, ed. Mu˙ammad al-Íàdiq Qam˙àwì, 5 vols. Beirut: Dàr I˙yà" al-Turàth al-'Arabì, 1405/1985. al-Juma˙ì, Mu˙ammad b. Salàm. ˇabaqàt al-shu'arà", ed. Joseph Hell. Leiden: Brill, 1916. al-Kalàbàdhì, A˙mad b. Mu˙ammad. Rijàl ßa˙ì˙ al-Bukhàrì, ed. 'Abdallàh al-Laythì, 2 vols. Beirut: Dàr al-Ma'rifa, 1407/1987. Khalìfa b. Khayyà† al-'Ußfurì. Ta"rìkh Khalìfa b. Khayyà†, ed. Akram Îiyà" al-'Umarì. Damascus/Beirut: Dàr al-Qalam and Mu"assasat al-Risàla, 1397/1977. id., Al-ˇabaqàt, ed. Akram Îiyà" al-'Umarì, Riyadh: Dàr ˇayba, 1402/1982. al-Khalìlì, al-Khalìl b. 'Abdallàh. Al-Irshàd, ed. Mu˙ammad Sa'ìd 'Umar Idrìs, 3 vols. Riyadh: Maktabat al-Rushd, 1409/1989. al-Mizzì, Yùsuf b. 'Abd al-Ra˙màn. Tahdhìb al-kamàl fì asmà" al-rijàl, ed. Bashshàr 'Awwàd Ma'rùf, 35 vols. Beirut: Mu"assasat al-Risàla, 1400/1980.

48

ella landau-tasseron

Mu"arrij al-Sadùsì. Kitàb ˙adhf min nasab Quraysh, ed. Íalà˙ al-Dìn al-Munajjid. Beirut: Dàr al-Kitàb al-Jadìd, 1396/1976. al-Mubàrakfùrì, Mu˙ammad b. 'Abd al-Ra˙màn. Tu˙fat al-a˙wadhì bi-shar˙ jàmi' alTirmidhì. 10 vols. Beirut: Dàr al-Kutub al-'Ilmiyya, 1410/1990. al-Munàwì, 'Abd al-Ra"ùf. Fay† al-qadìr shar˙ al-jàmi' al-ßaghìr min a˙àdìth al-bashìr alnadhìr. 6 vols. Cairo: al-Maktaba l-Tijàriyya l-Kubrà, 1356/1937. Muslim b. al-Óajjàj. Ía˙ì˙ Muslim, ed. Mu˙ammad Fu"àd 'Abd al-Bàqì, 5 vols. Beirut: Dàr I˙yà" al-Turàth al-'Arabì, n.d. al-Muttaqì al-Hindì. Kanz al-'ummàl fì sunan al-aqwàl wa-l-af 'àl, ed. Íafwat al-Saqà, Bakrì Óayàtì, Nadìm Mar'ashlì and Usàma Mar'ashlì, 18 vols. Beirut: Mu"assasat al-Risàla, 1981–86. al-Nasà'ì, A˙mad b. Shu'ayb. Al-Sunan al-kubrà, ed. 'Abd al-Ghaffàr Sulaymàn alBandàrì and Sayyid Kisrawì Óasan, 6 vols. Beirut: Dàr al-Kutub al-'Ilmiyya, 1411/1991. al-Nawawì, Mu˙yì al-Dìn Ya˙yà b. Sharaf. Tahdhìb al-asmà". Beirut: Dàr al-Fikr, 1416/1996. al-Qalqashandì, A˙mad b. 'Alì. Íub˙ al-a'shà fì ßinà'at al-inshà, ed. Yùsuf 'Alì alˇawìl, 8 vols. Damascus: Dàr al-Fikr, 1407/1987. al-Qur†ubì, Mu˙ammad b. A˙mad. Tafsìr al-Qur†ubì, ed. A˙mad 'Abd al-'Alìm alBardùnì, 20 vols. Cairo: Dàr al-Sha'b, 1372/1952. al-Ràzì, Mu˙ammad b. Abì Bakr. Mukhtàr al-ßi˙à˙, ed. Ma˙mùd Khà†ir. Beirut: Maktabat Lubnàn, Nashirùn, 1415/1995. Sa'ìd b. Manßùr. Sunan Sa'ìd b. Manßùr, ed. Óabìb al-Ra˙màn al-A'Ωamì. India: alDàr al- Salafiyya, 1402/1982. al-Sarakhsì, Mu˙ammad b. Abì Sahl. Al-Mabsù†. 30 vols. Beirut: Dàr al-Ma'rifa, 1406/1986. al-Shàfi'ì, Mu˙ammad b. Idrìs. Al-Risàla, ed. A˙mad Mu˙ammad Shàkir. Cairo, 1358/1938. id., Kitab al-umm. 8 vols. Beirut: Dàr al-Ma'rifa, 1393/1973. al-Shawkànì, Mu˙ammad b. 'Alì. Nayl al-aw†àr. 9 vols. Beirut: Dàr al-Jìl, 1393/1973. id., Fat˙ al-Qadìr al-jàmi' bayna fannay al-riwàya wa-l-diràya min 'ilm al-tafsìr. 5 vols. Beirut: Dàr al-Fikr, n.d. id., Al-Badr al-ˇàli' bi-ma˙àsin man ba'da al-qarn al-sàbi'. 2 vols. Beirut: Dàr al- Ma'rifa, n.d. al-Shaybànì, Mu˙ammad b. al-Óasan. Al-Mabsù†, ed. Abù l-Wafà l-Afghànì, 5 vols. Karatchi: Idàrat al-Qur"àn wa-l-'Ulùm al-Islàmiyya, n.d. al-Suhaylì, 'Abd al-Ra˙màn b. 'Abdallàh. Al-Raw∂ al-unuf fì tafsìr al-sìra al-nabawiyya li-Ibn Hishàm, ed. Majdì Manßùr al-Shùrà, 4 vols. Beirut: Dàr al-Kutub al-'Ilmiyya, 1418/1997. al-Suyù†ì, Jalàl al-Dìn. Tadrìb Al-Ràwì fì shar˙ taqrìb al-Nawawì, ed. 'Abd al-Wahhàb 'Abd al-La†ìf, 2 vols. Riyadh: Maktabat al-Riyà∂ al-Óadìtha, n.d. al-ˇabarànì, Sulaymàn b. A˙mad. Al-Mu'jam al-kabìr, ed. Óamdì b. 'Abd al-Majìd al-Salafì, 20 vols. Mosul: Maktabat al-'Ulùm wa-l-Óikam, 1404/1984. al-ˇabarì, Mu˙ammad b. Jarìr. Biographies of the Prophet’s Companions and Their Successors, trans. and annot. Ella Landau-Tasseron as The History of al-ˇabarì. An Annotated Translation, 39. New York: State University of New York Press, 1998. id., Ta"rìkh al-rusul wa-l-mulùk. 5 vols. Beirut: Dàr al-Kutub al-'Ilmiyya, 1407/1987. id., Tafsìr al-ˇabarì. 30 vols. Beirut: Dàr al-Fikr, 1405/1985. id., Tahdhìb al-àthàr, al-juz" al-mafqùd, ed. 'Alì Ri∂à b. 'Abdallàh b. 'Alì Ri∂à. Beirut: Dàr al-Ma"mùn li-l-Turàth, 1995. al-ˇa˙àwì, A˙mad b. Mu˙ammad b. Salàma. Mukhtaßar ikhtilàf al-'ulamà", ed. 'Abdallàh Nadhìr A˙mad, 5 vols. Beirut: Dàr al-Bashà"ir al-Islàmiyya, 1417/1997. id., Shar˙ ma'ànì al-àthàr. 4 vols. Beirut: Dàr al-Kutub al-'Ilmiyya, 1399/1979.

alliances in islam

49

al-Wàqidì, Mu˙ammad b. 'Umar. Kitàb al-maghàzì, ed. Marsden Jones, 3 vols. London: Oxford University Press, 1966. Yàqùt al-Rùmì. Mu'jam al-buldàn. 5 vols. Beirut: Dàr al-Fikr, n.d. al-Zara'ì, Mu˙ammad b. Ayyùb. Óàshiyat Ibn al-Qayyim. 14 vols. Beirut: Dàr alKutub al- 'Ilmiyya, 1415/1995. al-Zayla'ì, 'Abdallàh b. Yùsuf. Naßb al-ràya li-a˙àdìth al-hidàya, ed. Mu˙ammad Yùsuf al- Bannùrì, 4 vols. Cairo: Dàr al-˙adìth, 1357/1938. al-Zurqànì, Mu˙ammad b. 'Abd al-Bàqì. Shar˙ al-Zurqànì 'alà muwa††a" al-imàm Màlik, 4 vols. Beirut: Dàr al-Kutub al-'Ilmiyya, 1411/1911. Secondary Sources 'Athàmina, Khalil. “Non Arab Regiments And Private Militias During the Umayyad Period,” Arabica 45 (1998), 347–78. Crone, Patricia. Roman, Provincial and Islamic Law, Cambridge Studies in Islamic Civilization. Cambridge/London/N.Y./Sydney: Cambridge University Press, 1987. id., Slaves on Horses. The Evolution of the Islamic Polity, Cambridge: Cambridge University Press, 1980. id., “Jàhilì and Jewish law: the Qasàma,” Jerusalem Studies in Arabic and Islam 4 (1984),153–201. Elad, Amikam. “The Southern Golan in the Early Muslim Period: The Significance of Two Newly Discovered Milestones of 'Abd Al-Malik,” Der Islam 76 (1999), 33–88. Goldziher, Ignaz. Muslim Studies, ed. S.M. Stern, trans. C.R. Barber and S.M. Stern. London: George Allen and Unwin Ltd., 1967. Hallaq, Wael. “The Use and Abuse of Evidence,” Journal of the American Oriental Society 110/I (1990), 79–91. Juda, Jamal. Die Sozialen und Wirtschaftlichen Aspekte der Mawàlì in Frühislamischer Zeit, Ph.D. diss., Eberhard-Karls-Universitat Tübingen, Ràfàt ( Jordan), 1983. Kister, M.J. “On Strangers and Allies in Mecca,” Jerusalem Studies in Arabic and Islam 13 (1990), 113–54. Landau-Tasseron, Ella. “Alliances Among the Arabs,” Al-Qantara 26 (2005), 141–73. id., “Asad from Jahiliyya to Islam,” Jerusalem Studies in Arabic and Islam 6 (1985), 1–28. id., “The Participation of ˇayyi" in the Ridda,” Jerusalem Studies in Arabic and Islam 5 (1984), 53–71. Lecker, Michael. The Banù Sulaym, The Max Shloessinger Memorial Series 4. Jerusalem: The Hebrew University of Jerusalem, 1989. Mitter, Ulrike. “Unconditional manumission of slaves in early Islamic law: a ˙adìth analysis,” Der Islam 78 (2001), 35–72. Motzki, Harald. “Der Fiqh Des Zuhri: Die Quellenproblematik,” Der Islam 68 (1991), 1–44. Pedersen, J. [and Y. Linant de Bellefonds]. art. “asàma,” EI 2, IV, 689–90. Powers, David. Studies in Qur"àn and Óadith: the formation of the Islamic law of inheritance. Berkeley: University of California Press, 1986. Schmucker, Werner. Untersuchungen zu Einigen Wichtigen Bodenrachtlichen Konsequenzen der Islamischen Eroberungsbewegung. Bonn: Selbstverlag des Orientalischen Seminars der Universität Bonn, 1972. Vacca, V. art. “al-Na∂ìr,” EI 2, VII, 852–53.

WERE THE JEWISH TRIBES IN ARABIA CLIENTS OF ARAB TRIBES? Michael Lecker

The assumption that whole Jewish tribes in Arabia were clients of their Arab neighbors is an essential element in a recent discussion on the origins of the walà" institution in Islam:1 The Islamic institution of walà" is generally assumed to be of Arabian origin (cf. Goldziher, Muhammedanische Studien, 2 vols. [Halle: Max Niemeyer, 1889–1890], I, chap. 3; Juda, Die sozialen und wirtschaftlichen Aspekte der Mawàlì in frühislamischer Zeit [Tübingen, 1983], but this is scarcely correct. Leaving aside foreign merchants and colonists under imperial protection, the non-Arab population of pre-Islamic Arabia consisted of Jews, slaves and freedmen of African and Middle Eastern extraction, half-bred descendants of colonists, and presumably also ethnic and occupational pariah groups of the type attested in modern times (Qawàwila, Bayàdir, Íullubìs etc.). There is no reason to doubt that all were known as mawàlì in the sense of “kinsmen,” in so far as they were free and came under Arab protection (cf. the modern use of the word akh “brother”), but the question is, what this implied. Are we to take it that all non-Arabs were individually assigned to Arab patrons and acquired partial membership of Arab tribes through them, having no social organization of their own? Or did they form social groups of their own, so that they were collectively placed under the protection of Arab tribes in which they acquired no membership at all, merely becoming their satellites? The first solution is that enshrined in Islamic walà", but it is the second which is attested for Arabia. Thus it is well-known that the Jews of Arabia formed tribal groups of their own. In fact, Jewish tribes were sometimes strong enough to escape Arab protection altogether (and thus also the status of mawàlì ). But this was hardly the common pattern. The Jews of Fadak, for example, paid protection money to Kalb (Kister, “On the wife of the goldsmith from Fadak and her progeny,” Muséon 92 [1979], 321); the Jews of Wàdì l-Qurà similarly paid what would nowadays be known as khuwwa to Arab overlords (al-Bakrì, Mu'jam mà ista'jam, ed. F. Wüstenfeld [Göttingen, 1867–7, I, 30]; and those of Yathrib were reduced to client status by the Aws and Khazraj [q.vv.] some time before the rise of

1

Crone, art. “Mawlà,” EI 2.

the jewish tribes in arabia

51

Islam (Wellhausen, Skizzen und Vorarbeiten IV [Berlin, 1889), 7ff.]. Naturally, client status weakened the tribal organization of the Jews; the same is true of modern pariah groups. But the Jewish tribes were not dissolved, nor were the Jews assigned to individual patrons: clientage was a relationship between groups. Similarly, the Arabized descendants of the Persian workmen and prostitutes of Hajar clearly formed a quasi-tribal group of their own under 'Abdì protection, for all that they adopted the nisba of their protectors (al-ˇabarì, I, 986).

This article aims at establishing that on the eve of Islam the Jewish tribes of Yathrib (henceforward: Medina, even with reference to preIslamic times), and those of Fadak and Wàdì l-Qurà were not clients of Arab tribes and hence are irrelevant to the discussion of the origins of the walà" system in Islam.

The Jews of Medina on the eve of the Hijra Let us start with the status of the Jews of Medina at two points in time, namely when Mu˙ammad arrived there and some seventy years earlier. A passage in al-Wàqidì assumed by J. Wellhausen and others to reflect the Jews’ subordination by the Arabs is in fact garbled; in its correct form it states the opposite, namely that when Mu˙ammad came to Medina, the Jews were the strongest element in its population.2 With regard to the two main tribes, the Na∂ìr and QurayΩa, this can be corroborated by the evidence on their strong fortifications in Upper Medina or the 'Àliya.3 Roughly to the middle of the sixth century CE, the Na∂ìr and QurayΩa were “kings” operating under the Sassanians. In the last quarter of the sixth century Medina had an Arab king of the Khazraj appointed by the king of al-Óìra, which indicates that at that specific time the Na∂ìr and QurayΩa were not the dominant power in Medina.4 In general, the study of pre-Islamic Medina is fraught with difficulties because of the tendentious and apologetic nature of Anßàrì historiography. The tribal informants of the Anßàr were emotionally involved 2 Lecker, “Wàqidì’s account on the status of the Jews of Medina: a study of a combined report,” JNES 54 (1995), 15–32. 3 Lecker, Muslims, Jews and Pagans: Studies on Early Islamic Medina (Leiden: Brill, 1995), 10–15. 4 Lecker, “The levying of taxes for the Sassanians in Pre-Islamic Medina (Yathrib),” JSAI 27 (2002), 109–26; repr. idem, People, Tribes and Society in Arabia Around the Time of Mu˙ammad (Aldershot: Ashgate, 2005), no. 1.

52

michael lecker

in the pre-Islamic history of their tribes; after all, the tribal society in which they operated was an extension of the pre-Islamic one. Consequently their accounts not only preserved for posterity the memory of past events but also defended the reputation of their tribes. One should not expect their approach to have been impartial or academic, and hence their reports must not be taken at face value. This will be demonstrated by accounts on two events which took place in Medina in the decades leading to Islam, two events to which we now turn.

The aftermath of the Fi†yawn affair Al-Fi†yawn was a quasi-legendary Jewish king who reportedly used to deflower the brides of Medina before they were delivered to their lawful husbands. The tendentious report in the Aghànì (XIX, 97) on the aftermath of his affair is sometimes quoted in the secondary literature as if it came from an unprejudiced historian. For example, P. Crone wrote: Reputedly lords and masters of both Yathrib and its Arab immigrants at first, they [the Jews—M.L.] were defeated by the Arab tribes some time before the rise of Islam and reduced to client status. This did undermine their internal organization. On their defeat, we are told, “the Jews were weakened and lost their capacity to defend themselves; they were very afraid, so whenever a member of the Aws and Khazraj provoked them by [doing] something which they disliked, they would no longer go to one another as they had done in the past; rather, [every] Jew would go to the protectors among whom he lived ( jìrànihi lladhìna huwa bayna aΩhurihim) and say, “we are your protégés and clients ( jìrànukum wa-mawàlìkum)”; for every Jewish family (qawm) had sought refuge with a clan (ba†n) of the Aws or the Khazraj, seeking strength from them.” Defeat thus destroyed the tribal cohesion of the Jews who, unable to take joint action against the victors, were forced to seek protection from them; and apparently it reduced them to tenants too.5 5

Crone, Roman, Provincial and Islamic Law: The Origins of the Islamic Patronate (Cambridge: Cambridge University Press, 1987), 56; idem, art. “Mawlà,” EI 2, 875a; al-Bakrì, Al-Mamàlik wa-l-masàlik, MS Nur Osmaniye, 3034, 242b; ed. Van Leeuwen and Ferre (Tunis, 1992) I, 416–17, nos. 700–701; Jazìrat al-'arab min kitàb al-mamàlik wa-l-masàlik li-Abì 'Ubayd al-Bakrì, ed. 'Abdallàh Yùsuf al-Ghunaym (Kuwait: Dhàt al-Salàsil, 1397/1977), 94–95) says that following the actions of Abù Jubayla (who according to some helped the Arabs overcome the Jews), the Aws and Khazraj became strong, took the lands of the Jews and were dispersed in the 'Àliya of

the jewish tribes in arabia

53

Also Wensinck and Serjeant regarded the Aghànì report as an accurate account on the status of the Jews of Medina—viz. all of them— on the eve of Islam. Wensinck wrote: From this moment until the time of Mu˙ammad the Banù Qaylah [i.e., the Aws and Khazraj—M.L.] ruled Medina and were known in Arabia as mawàlì al-yahùd, patroni Judaeorum.

Serjeant argued: The Aghànì (XIX, 97) makes it clear that whatever position the Jews had held in former times, they had lost status and their power to defend themselves had diminished (dhallù wa-qalla mtinà'uhum), and by the immediately pre-Islamic era they were under the protection of the Arab tribes. . . . This loss of status followed after a slaughter of the Jews by Màlik b. al-'Ajlàn of Khazraj.6

Watt struck a more cautious tone: It is commonly suggested that the Aws and the Khazraj became rulers of Yathrib with all the Jews in subjection to them; but the sources do not support such a view.7

Medina. Only a small portion of the Jews remained, namely those who were prepared to live in humility and were content with ignominy ( fa-lammà fa'ala dhàlika 'azzati l-Aws wa-l-Khazraj wa-ghalabat 'alà diyàrihim, wa-tafarraqati l-Aws wa-l-Khazraj fì 'àliyati l-Madìna fa-lam yabqa mina l-yahùd illà aqalluhum mimman aqàma 'alà l-hùn wara∂iya bi-l-ßaghàr; the term ßaghàr is probably inspired by Qur"àn 9:29). Al-Ya'qùbì (in whose report the Jews were subdued by one of mulùk al-Yaman) is more restrained: the position of the Aws and Khazraj was improved, they planted palm-trees and built houses; al-Ya'qùbì, Ta"rìkh (Beirut: Dàr Íàdir and Dàr Bayrùt, 1379/1960), I, 204 ( fa-ßalu˙at ˙àlu l-Aws wa-l-Khazraj wa-gharasù [printed: wa-gh.r.s.] l-nakhl wa-ansha"ù l-manàzil ). 6 Wensinck, Muhammad and the Jews of Medina, with an excursus: Muhammad’s Constitution of Medina, by Julius Wellhausen, trans. and ed. W.H. Behn (Freiburg im Breisgau: Klaus Schwarz Verlag, 1975), 25; Serjeant, “The sunnah jàmi'ah, pacts with the Yathrib Jews, and the Ta˙rìm of Yathrib: analysis and translation of the documents comprised in the so-called ‘Constitution of Medina’,” BSOAS 41 (1978), 3. Beside the Aghànì report Serjeant (2) also refers to al-Wàqidì’s report discussed in Lecker, “Wàqidì’s account on the status of the Jews of Medina.” See also G. Levi della Vida, “Pre-Islamic Arabia,” in Nabih Amin Faris, ed., The Arab Heritage (Princeton: Princeton University Press, 1944), 51: “An example of a city which after a period of independence was obliged to yield to the nomads is afforded by Yathrib (the later Medina), a rich oasis in Hejaz where some Jewish tribes had settled, only later to fall under the control of two Bedouin tribes which occupied the town and reduced the old inhabitants to the condition of tributaries.” 7 Watt, Muhammad at Medina (Oxford: The Clarendon Press, 1956), 193. In a footnote he adds: “The general impression is that the Jews were independent.” See also idem, art. “al-Madìna,” EI 2, 995: “It is sometimes said that the Jews now became subject to these Arabs [i.e. the Aws and Khazraj—M.L.]. This is not borne

54

michael lecker

A sober approach was also adopted by F. Altheim and R. Stiehl who observed that the battle of Bu'àth which took place several years before the Hijra gave renewed significance to the QurayΩa and Na∂ìr, and hence Mu˙ammad included them in the treaty (i.e., the so-called Constitution of Medina) concluded before Badr in which he incorporated the whole population of Medina.8 The Fi†yawn affair should be dated to the second half of the sixth century CE, roughly three generations before the Hijra. This chronology is based on the fact that the Arab protagonist of this affair, Màlik b. al-'Ajlàn who killed the Jewish king was the great-grandfather of two Companions of the Prophet.9 For the apologetic Anßàrì historiography, the affair was a watershed in the history of pre-Islamic Medina, separating an era of Jewish domination from an era of Arab domination; the slaughter carried out at that time by a foreign prince turned the Jews into underlings and punished them for the sexual excesses of their corrupt king. Rather than setting forth a process of gradual decline in the power of the Jews, tribal tradition adopted a simplistic concept in which Jewish domination came to an abrupt end. The Fi†yawn affair no doubt weakened the Jews, including the Na∂ìr and QurayΩa tribes. But whatever effects it had in the short run, by the time of the Hijra the main Jewish tribes restored their power. Al-Ya'qùbì reports on an attempt by the Anßàr on the eve of the Hijra to seal an alliance with Quraysh. It was motivated by

out, however, by the historical accounts of the period up to 5/627. The main Jewish groups, though doubtless now weaker than the Arabs, retained a measure of independence and continued to occupy some of the best lands.” 8 Altheim and Stiehl, Finanzgeschichte der Spätantike (Frankfurt am Main: V. Klostermann, 1957), 160: “Das Treffen von Bu'àΔ hatte aber QuraiΩa und al-Na∂ìr neue Bedeutung gegeben, und Mohammed nahm die Juden Medinas in den Brudernbund auf, in dem er, noch vor der Schlacht bei Badr, die ganze Stadt zusammenschloss.” (For a discussion of the Jewish participation in this treaty, see now Lecker, The Constitution of Medina (Princeton: Darwin Press, 2004). Watt, Medina, 195, remarks that in the story of Bu'àth the Na∂ìr and QurayΩa “seem to be acting as sovereign bodies, making alliances with Arab clans as equals, not politically subordinate to any of them, but perhaps tending to become relatively weaker” (this is a reference to the hostages handed over by the Jews to the Khazraj prior to the battle). 9 Ibn al-Kalbì, Nasab Ma'add wa-l-yaman al-kabìr, ed. Nàjì Óasan (Beirut: 'Àlam al-Kutub-Maktabat al-Nah∂a l-'Arabiyya, 1408/1988), I, 415; Ibn Óazm, Jamharat ansàb al-'arab, ed. 'Abd al-Salàm Hàrùn (Cairo: Dàr al-Ma'àrif, 1382/1962), 353–54. The statement that Màlik was in command of the Khazraj in the battle of Bu'àth is erroneous; cf. Abù l-Faraj al-Ißfahànì, Kitàb al-aghànì (Bùlàq, 1285/1868), XV, 161–65; Wellhausen, Skizzen IV, 7.

the jewish tribes in arabia

55

the fact that “the Na∂ìr, the QurayΩa and other Jews became emboldened against them” (ijtara"at 'alayhim).10 In sum, when it is reported that “the Jews were weakened and lost their capacity to defend themselves,” this could only be true of some of them, certainly not of the main tribes. This exaggerated statement represents the tendentious and apologetic approach of Anßàrì informants. The battle of Óusayka The ways of Anßàrì historiography can also be demonstrated by the reports on the small scale battle between the Salima (a subdivision of the Khazraj) and the Jews living in the Óusayka village in the north of Medina. At the nucleus of these reports there is historical fact: on the eve of Islam certain Jewish tribes—not including the main ones living in the 'Àliya—were losing ground to their Arab neighbors. But we are essentially concerned here with the way in which the battle was described in Anßàrì historiography. The reports naturally come from the interested party itself, namely the Salima. They claimed that a proper review of their troops had been carried out before the battle, precisely where the Muslim warriors of Badr were later reviewed on their way to the battlefield. This was a good omen: the Salima implied that the victory of Badr was heralded by their own victory at Óusayka. According to the tribal report, the battle was anything but small. After the review, we [the Salima] marched on the Jews of Óusayka who were the strongest Jewish group at that time [!] and destroyed them [literally: we killed them in whichever way we wanted]. Since then and to this very day all of the Jews [!] were subordinate to us (. . . wa-hum a'azz yahùd kànù yawma"idhin fa-qatalnàhum kayfa shi"nà fa-dhallat lanà sà"ir yahùd ilà l-yawm).

Samhùdì adduces verses by a poet whose identity is not specified— in fact it is Ka'b b. Màlik of the Salima: We have attacked them in the morning at the foot of the mountain on the day of Óusayka with the wide swords of Bußrà and the well-straightened brown spear 10 Al-Ya'qùbì, Ta"rìkh, II, 37. Al-Ya'qùbì also reports of a later Anßàrì attempt to make an alliance with the Thaqìf.

56

michael lecker No one of them stood up to fight us and they did not carp at us on the day we drove them away.11

The expulsion of the Jews of Óusayka, which was a local event, was connected to the Prophet’s obtaining of a foothold or “territorial basis” in Medina.12 While some of the Jewish clans in Lower Medina or the Sàfila were giving way to military pressure from their Arab neighbors, those living in Upper Medina or the 'Àliya were only marginally affected. Hence the claims of the Salima should be taken with a pinch of salt.

Anti-Anßàrì polemics Political and intertribal polemics form the context of the apologetic and exaggerated statements made by the Anßàr with regard to their pre-Islamic history. Regardless of their actual political and military weight, the Anßàr drew fire from their adversaries.

11 Al-Samhùdì, Wafà" al-wafà bi-akhbàr dàr al-mu߆afà, ed. Mu˙ammad Mu˙yì lDìn 'Abd al-Óamìd (Cairo, 1374/1955), IV, 1191, s.v. Óusayka: ßaba˙nàhum [printed ßafa˙nàhum] bi-l-saf˙i yawma Óusaykatin\ßafà"i˙a Bußrà wa-l-Rudayniyyata l-sumrà; fa-mà qàma minhum qà"imun li-qirà'inà\wa-là nàhabùnà yawma nazjuruhum zajrà; Lecker, “Mu˙ammad at Medina: a geographical approach,” JSAI 6 (1985), 39–40; Hirschberg, Yisra"el be-'arav (Tel-Aviv: Mossad Bialik, 1946), 127–28; al-Wàqidì, Kitàb al-maghàzì, ed. Marsden Jones (London: Oxford University Press, 1966), I, 23–24; al-Fìrùzàbàdì, al-Maghànim al-mu†àba fì ma'àlim ˇàba, ed. Óamad al-Jàsir (Riyà∂: Yamàma, 1389/1969), s.v. = Yàqùt, Mu'jam al-buldàn (Beirut: Dàr Íàdir-Dàr Bayrùt, 1957), s.v.; Naßr, alAmkina wa-l-miyàh wa-l-jibàl wa-l-àthàr, MS Br. Lib. Add. 23, 603, 54b; Ibn Shabba, Ta"rìkh al-Madìna al-munawwara, ed. Fahìm Mu˙ammad Shaltùt (n.p., [1399/1979]), I, 158 (instead of yà bunayya, innà 'turi∂nà hàhunà bi-l-Suqyà ˙attà qàbalnà l-yahùd biÓusayka, read: . . . ˙ìna qàtalnà . . .). For the reading ßaba˙nàhum, see Óamad al-Jàsir’s comments on this edition, “Qirà"a sarì'a fì kitàb Ta"rìkh al-Madìna li-Ibn Shabba,” al-'Arab 18 (1983–84), 314. On the swords in question, see Schwarzlose, Die Waffen der alten Araber aus ihren Dichtern dargestellt (Leipzig: J.C. Hinrichs’sche Buhchhandlung, 1886), 131, 218. In an elegy on 'Abd al-Mu††alib, Qußayy’s offspring are likened to al-Rudayniyya l-sumr; Ibn Hishàm, al-Sìra al-nabawiyya, ed. Mu߆afà l-Saqqà, Ibràhìm al-Abyàrì and 'Abd al-ÓafìΩ Shalabì (Beirut: I˙yà" al-Turàth al-'Arabì, 1391/1971), I, 185 (trans. Guillaume, The Life of Muhammad: A Translation of Ibn Ishaq’s Sìrat Rasùl Allàh [Oxford: Oxford University Press, 1955], 77: “. . . like very spears”); for other attestations of this combination, see e.g., al-Bakrì, Mu'jam mà sta'jam, IV, 1265, s.v. Manbij; Aghànì, X, 34. For ßa'da (or spear) Rudayniyya, see Ibn Kathir, Bidàya (Cairo: Ma†ba'at al-Sa'àda, al-Ma†ba'a l-Salafiyya and Maktabat al-Khànjì) 1351–58/1932–39), II, 355:5. 12 Lecker, “Mu˙ammad at Medina,” 29–30, 39–42; idem, The Constitution of Medina, Index.

the jewish tribes in arabia

57

The weakness of the Aws and Khazraj vis-à-vis the Jews when the former settled in Medina and in particular the excesses of alFi†yawn, were used against the Anßàr in intertribal polemics. This already began as early as the first century of Islam: Raw˙ b. Zinbà' al-Judhàmì (84/703)13 mentioned al-Fi†yawn in order to tease his Anßàrì wife.14 Considering the predominantly negative image of the Jews in the early Islamic period, the claim that the Anßàr were of Jewish stock was meant to revile them. According to Ibn Is˙àq, the Aws and Khazraj descended from four hundred scholars of the Children of Israel whom the Óimyarì king Tubba' settled in Medina; a late compilation describes the said genealogy of the Anßàr as a result of a Jewish plot.15 The military activity of a tribe was always a central component of its remembered history and the Aws and Khazraj were no exception. The Qurashì scholar al-Zubayr b. Bakkàr, quoting his uncle Muß'ab b. al-Zubayr, said that with the exception of the battle of Bu'àth, the pre-Islamic wars between the Aws and Khazraj were merely stone-throwing and fighting with clubs.16 The distinguished Qurashì scholars who descended from 'Abdallàh b. al-Zubayr17 should have known better: they were born in Medina where their family lived and prospered since the time of Mu˙ammad.18 An apocryphal dialog between Mu˙ammad and the Anßàr allegedly took place at Óunayn following Anßàrì complaints over the division of spoils. The Prophet addressed them with the following rhetorical question: “Had I not come to you when you were not riding horses and [only later] you began riding them?” They humbly replied: “You came to us when we were subordinate and few, and God strengthened

13 Hawting, art. “Raw˙ b. Zinbà' al-Ju£àmì,” EI 2, s.v.; Hasson, “Le chef judhàmite Raw˙ ibn Zinbà',” Studia Islamica 77 (1993), 95–122. 14 She was the daughter of al-Nu'màn b. Bashìr al-Anßàrì; al-Bakrì, Sim† al-la"àlì fì shar˙ amàlì l-Qàlì, ed. 'Abd al-'Azìz al-Maymanì (Cairo, 1354/1936), I, 179; Aghànì, VIII, 139:22 (where the verses are ascribed to a cousin of Raw˙); al-Jà˙iΩ, Rasà"il, ed. 'Abd al-Salàm Hàrùn (Cairo: al-Khànjì, 1384–99/1964–1979), II, 359 (Kitàb al-bighàl ). 15 See Kister, “Óaddithù 'an banì Isrà"ìl,” Israel Oriental Studies 2 (1972), 233. 16 Wellhausen, Skizzen IV, 30 = Aghànì, II, 162:16. 17 Gibb, art. “'Abd Allàh b. al-Zubayr,” EI 2, s.v. 18 Sezgin, Geschichte des arabischen Schrifttums (GAS) (Leiden: Brill, 1967–), I, 271f., 317f.

58

michael lecker

us through you.”19 A common polemical technique is used here: one party endorses the claim of another. Admittedly, there were not many horses in pre-Islamic Medina since its agriculturalists employed less delicate beasts. But the supposed exchange between the Prophet and the Anßàr was polemical and aimed at undermining their prestige. The Anßàrì point of view is demonstrated, for example, by Sa'd b. 'Ubàda20 who is supposed to have given at Óunayn the following answer: “The Arabs know that there is no tribe more capable of defending its families and property (amna' li-mà warà" Ωuhùrihim) than us.”21 At the 'Aqaba meeting (622 CE) al-Barà" b. Ma'rùr22 supposedly told the Prophet: “[W]e shall defend you as we shall defend our womenfolk . . ., for we are men of war and men of coats of mail which we inherited from father to son.”23 The pre-Islamic history of the Anßàr was, as we are told elsewhere, one of defiance, not of submission: the Anßàr alighted on the fringes of Medina (khàrij al-Madìna), then they became strong enough to expel the Jews from it and alight inside it, while the Jews were pushed to the fringes.24 The Anßàr “who have the strongest hearts and the noblest ambitions” never paid tribute (itàwa)25 to any of the kings.26 Such claims and counter-claims are at the background of the reports on the state 19 Wa-ji"tanà yà rasùla llàhi wa-na˙nu adhilla qalìlùna fa-a'azzanà llàhu bika; Aghànì, XIII, 67–68. 20 Watt, EI 2, s.v. 21 Al-ˇabarì, Tafsìr, X, 71:14; cf. the commentaries of Q.42:23: ibid., XXV, 16:–9; al-Qur†ubì, al-Jàmi' li-a˙kàm al-qur"àn3 (Cairo: Dàr al-Kutub, 1354–87/1935–67), XVI, 24. 22 Zettersteen, EI 2, s.v. 23 Al-ˇabarì, I, 1220. 24 Al-Balàdhurì, Futù˙, ed. M.J. de Goeje (Leiden: Brill, 1863–66), 17. 25 Cf. Cahen, EI 2, s.v. 26 Wa-hum a'azzu l-nàs anfusan wa-ashrafuhum himaman lam yu"addù itàwa qa††u ilà a˙ad mina l-mulùk; Ibn 'Abd Rabbihi, al-'Iqd al-farìd, ed. A˙mad Amìn, A˙mad alZayn and Ibràhìm al-Abyàrì (Cairo: Lajnat al-Ta"lìf wa-l-Tarjama wa-l-Nashr, 1359–72/1940–1953), III, 334:12. The Anßàr were a branch of the Azd; for the boastful tribal tradition of another branch of the Azd in connection with its settlement in 'Umàn, see Kashf al-ghumma al-jàmi' li-akhbàr al-umma, ed. A˙mad 'Ubaydalì (Nicosia: Dalmùn li-l-Nashr, 1405/1985), 211f. (quoting al-Kalbì). For a most outspoken pro-Anßàrì bias (and pro-Yemenì bias in general), see al-Hamdànì, Kitàb qaßìdat al-dàmigha, ed. al-Akwa' (Cairo, [1384/1964]), 218f., who also deals with the status of the Aws and Khazraj. E.g., 221: wa-lam tujàwiri l-Aws wa-l-Khazraj a˙adan illà wa-dàna lahà, kànat Muzayna wa-Ghifàr wa-ghayruhumà min qabà"il Nizàr lahum shibha l-jìra wa-adhallù QurayΩa wa-l-Na∂ìr fa-kànù lahum khawalan (“. . . and they became their slaves”). On the Dàmigha cf. El-Shami and Serjeant, “Regional literature: the Yemen,” in J. Ashtiany et al., eds., 'Abbasid Belles-lettres (Cambridge: Cambridge University Press, 1990), 448.

the jewish tribes in arabia

59

of affairs in pre-Islamic Medina. The claims should be carefully weighed against more solid evidence.

The nomad’s share in the annual crop Grants in kind made to various Arab tribes were adduced as evidence that the bestowers were clients of the receivers: “The Jews of Fadak, for example, paid protection money to Kalb . . . the Jews of Wàdì l-Qurà similarly paid what would nowadays be known as khuwwa to Arab overlords . . .; and those of Yathrib were reduced to client status by the Aws and Khazraj (above, 50).” I submit that these grants did not entail client status. Each settlement had an intricate network of relationships with different tribes, close and remote. These relationships should not be lumped together since there were different kinds of tribes, and consequently different relationships (see the chapter by Landau-Tasseron on the many alliances). Medina had close links with strong tribes such as the 'Àmir b. Ía'ßa'a and the Gha†afàn, and occasional contacts with the Tamìm who lived in eastern Arabia. It also had very close ties with relatively small tribes such as the Juhayna, Ghifàr, Aslam, Khuzà'a and Ashja' whose territories were on one to three days’ journey from Medina. The small tribes were dependent on Medina, had military alliances with one of its tribes, provided various services to the people of Medina and in general posed a negligible military threat. The large tribes could be dangerous and had to be bought off. The examples regarding Medina which are given below relate to the 'Àmir b. Ía'ßa'a and the Gha†afàn. At the time of Mu˙ammad four of the relatively small tribes, namely the Aslam, Ghifàr, Muzayna and Juhayna (or certain groups among them) were assigned the same tax collector, Ka'b b. Màlik al-Anßàrì;27 their taxes reached Abù Bakr after Mu˙ammad’s death. Among the small tribes which remained loyal to Medina after his death we also find the Ka'b of Khuzà'a whose tax collector was Bishr b. Suyfàn al-Ka'bì, and the Ashja' whose tax collector was Mas'ùd b. Rukhayla al-Ashja'ì.28 27 28

203.

Whose verses were quoted above, 55. Al-Diyàrbakrì, Ta"rìkh al-khamìs (Cairo: al-Ma†ba'a l-Wahbiyya, 1283/1866), II,

60

michael lecker The kayla of Abù Barà"

Medina granted an annual share of its produce of dates to the strong leader of the 'Àmir b. Ía'ßa'a, Abù Barà" 'Àmir b. Màlik (nicknamed mulà'ib al-asinna or “the one who plays with spears”). He received from the people of Medina (to wit, from all of them, not only from the Jews) a kayla of dates, namely an annual grant, in return for a safe conduct given to the Medinans traveling in Najd.29 The term kayla is derived from the root k-y-l which denotes a measure of capacity.30 Abù Barà"’s kayla protected the people of Medina when they traveled in Najd; it did not render them the clients of Abù Barà" or his tribe.

The ja'àla of Fadak In connection with the conquest (or rather temporary takeover) of Fadak by the Kalb around 570 CE it is reported that the tribal leader of the Kalb, al-Óàrith b. Óißn nicknamed al-Óarshà, was entitled to a ja'àla from the people of Fadak. According to M.J. Kister, al-Óàrith “had the right to the pay ( ja'àla) imposed on the people

29 Wa-kàna li-'Àmir b. Màlik min ahl Yathrib kayla min tumrànihim 'alà anna man †ala'a Najdan minhum fa-huwa fì khafàrat 'Àmir b. Màlik; Óassàn b. Thàbit, Dìwàn, ed. W. 'Arafat (London: Luzac, 1971), II, 176; quoted in Lecker, The Banù Sulaym: A Contribution to the Study of Early Islam ( Jerusalem: Institute of Asian and African Studies, 1989), 106 (in note 34, read kayla instead of kìla). This arrangement is of crucial importance in connection with the battle of Bi"r Ma'ùna. Cf. Caskel, ]amharat anNasab: Das genealogische Werk des Hi“àm ibn Mu˙ammad al-Kalbì (Leiden: Brill, 1966), II, 160, who says about 'Àmir that he came to Medina at the beginning of 4/625 “zu unbekannten Zwecken. . . . Er verpflichtete sich dort, den Muslimen Schutz zu gewähren, wir kennen aber weder den Grund noch den Umfang dieser Verpflichtung.” On the safe conduct offered by Abù Barà" to the Companions of the Prophet cf. Kister, “The expedition of Bi"r Ma'ùna,” in G. Maqdisi, ed., Arabic and Islamic Studies in Honor of H.A.R. Gibb (Leiden: Brill, 1965), 337–38, 355–57. Cf. also Bosworth, art. “Bi"r Ma'ùna,” EI 2. 30 On the way in which the Medinans used to measure dates, see Lisàn al-'Arab (Beirut: Dàr Íàdir, 1374–76/1955–56), s.v. k-y-l, 605a:–2. In another context it is reported that the kayla of the family of caliph 'Umar, or the food quantity assigned to it, was overlooked by 'Umar’s mawlà, Màlik b. 'Iyà∂; Ibn Óajar, al-Ißàba fì tamyìz al-ßa˙àba, ed. 'Alì Mu˙ammad al-Bijàwì (Cairo: Dàr Nah∂at Mißr, 1392/1972), VI, 274–75, no. 8362 (wallàhu 'Umar kaylat 'iyàl 'Umar fa-lammà qadima 'Uthmàn wallàhu l-qasm, fa-summiya Màlika l-Dàr). According to 'Alì b. al-Madìnì, Màlik was 'Umar’s storekeeper (kàna . . . khàzinan li-'Umar).

the jewish tribes in arabia

61

of Fadak.”31 This is a most unusual case of a settlement subdued by a nomadic tribe.32 Now j(a/i/u)'àla does not mean a tribute but a payment for services such as the return of a missing camel or a fugitive slave. Ja'àlat al-gharaq is the reward of one who dives to rescue a person or some goods.33 In the story of Joseph a ja'àla was promised to his brothers for returning the gold or silver cup (which also served Joseph as an official measure, siqàya, ßuwà' al-malik) planted in Binyàmìn’s34 effects.35

31 Kister, “On the wife of the goldsmith from Fadak and her progeny,” 321. Kister (330) linked the conquest of Fadak to the decline in the power of the Jews of Medina. He argued that the payment by Fadak of “some tribute” indicated, “that the power of the Jewish agricultural settlements in that period, the end of the third quarter of the sixth century, began to decrease.” Kister refers to Abù l-Baqà", al-Manàqib al-Mazyadiyya, MS Br. Mus., Add 23, 296, 72b–73b = I, 287 (al-Óàrith b. Óißn b. Îam∂am b. 'Adì b. Janàb al-Kalbì al-ma'rùf bi-l-Óarshà, kànat lahu ja'àla 'alà ahl Fadak fa-dafa'ùhu 'anhà fa-aghàra 'alayhim . . .). 32 Al-Óàrith wa-huwa l-Óarshà wa-qad ra"asa wa-ßàra lahu saby Fadak ˙ìna ftata˙ahà [sic] Kalb fì l-jàhiliyya; Ibn al-Kalbì, Nasab Ma'add, II, 562 (quoted by Kister from the MS). See also al-Jàsir, Fì shimàl gharb al-jazìra (nußùß, mushàhadàt, in†ibà'àt)2 (Riyà∂, 1401/1981), 297, quoting Ibn Màkùlà, al-Ikmàl, ed. 'Abd al-Ra˙màn b. Ya˙yà lYamànì (Hyderabad: Dà"irat al-Ma'àrif al-'Uthmàniyya, 1381/1962), II, 433. Obviously, the Kalb did not settle in Fadak following their victory. The raison d’être of the reports on Fadak is the fact that a Jewess captured by the Kalb in Fadak later gave birth to the last king of al-Óìra, al-Nu'màn b. al-Mundhir. 33 Mà yuj'alu li-man yaghùßu 'alà matà' aw insàn ghariqa fì l-mà"; Abù l-Qàsim Ma˙mùd b. 'Umar al-Zamakhsharì, al-Fà"iq fì gharìb al-˙adìth, ed. 'Alì Mu˙ammad al-Bijàwì and Mu˙ammad Abù l-Fa∂l Ibràhìm (Cairo: 'Ìsà l-Bàbì l-Óalabì, 1971), I, 174, s.v. th-m-n. In connection with 'Abdallàh b. 'Atìk’s raid against Abù Ràfi' Sallàm b. Abì l-Óuqayq in Khaybar (see now Motzki, “The murder of Ibn Abì l-Óuqayq: on the origin and reliability of some Maghàzì-reports”, in idem, ed., The Biography of Mu˙ammad: The Issue of the Sources [Leiden: Brill, 2000], 170–239) it is reported that the latter offered the Gha†afàn and the idol worshippers living around him (i.e., in the area surrounding Khaybar) al-ju'l al-'aΩìm in return for their fighting against the Prophet; al-Maqrìzì, Imtà' al-asmà' bi-mà li-l-rasùl mina l-anbà" wa-l-amwàl wa-l-˙afada wa-l-matà', ed. Ma˙mùd Mu˙ammad Shàkir (Cairo: Lajnat al-Ta"lìf wal-Tarjama wa-l-Nashr, 1941), I, 186. Cf. M. Bonner, “Ja'à"il and holy war in early Islam,” Der Islam 68 (1991), 45–64. At 46, n. 8, Bonner has ji'àla in the sense of “‘protection money’ much like modern Arabian khuwwa”, for which sense he quotes Crone, Roman, Provincial and Islamic Law, 56n, who in turn refers to Kister’s study on Fadak. See similarly in Bonner, Aristocratic Violence and Holy War: Studies in the Jihad and the Arab-Byzantine Frontier (New Haven: American Oriental Society, 1996), 14 note 7. At the beginning of the ridda, 'Uyayna b. Óißn and al-Aqra' b. Óàbis demanded a ju'l in return for the protection of Medina from the menacing Bedouin under their command. The prominent Muslims who supported their demand suggested that they receive a †u'ma; al-Diyàrbakrì, Khamìs, II, 202:5. 34 See Wensinck and G. Vajda, EI 2, s.v. 35 Ibn Kathir, Bidàya, I, 213.

62

michael lecker

A man gave a ja'àla to caliph Sulaymàn b. 'Abd al-Malik’s eunuch in return for private audience with the caliph.36 Most relevant for us here is an account relating to pre-Islamic Arabia: in Ibn al-Kalbì’s missing book Kitàb Óammàd al-Ràwiya it is supposedly reported with regard to Yawm al-Íafqa that Khusro’s caravan, having traveled from Madà"in via al-Óìra and Yamàma, was transported by the Sa'd (of the Tamìm) from the boundaries of the Óanìfa territory to the Yemen in return for a ja'àla. The Kalb may well have earned the ja'àla in return for similar services. In this case Fadak’s ja'àla is comparable to Medina’s kayla. Another term relevant for us here is 'ulfà or the share of the harvest given to a guard or a friend. The Banù l-Sharìd of the Sulaym were entitled to an 'ulfà from every harvest of barley reaped by the agriculturalists of the Iran settlement.37 The terms, the crops and the tribes varied, but in all three cases mentioned above we have an annual grant in kind made by the settled to the nomads at harvest-time. The grant can best be conceived of as a playing card in the game of Arabian politics. It should not lead to the conclusion that the settled, be they Jewish, Christian or idol worshippers, were the clients of the nomads. The †u'ma or ukl of Wàdì l-Qurà In Wàdì l-Qurà38 there was another type of relationship between the settled and their tribal neighbors who may well have engaged in agriculture themselves. The Jewish inhabitants of the settlement had a treaty with the 'Udhra (a tribe of the Qu∂à'a federation) which 36 Ibn 'Asàkir, Ta"rìkh madìnat Dimashq, ed. 'Umar b. Gharàma al-'Amrawì (Beirut, 1415/1995–), LXVIII, 176. Ja'àla could also mean the wages of a prostitute (qa˙ba); Aghànì, XIII, 84:1–1. In the context of the eunuch ja'àla may also be rendered “bribes.” 37 Al-Jàsir, Abù 'Alì l-Hajarì wa-ab˙àthuhu fì ta˙dìd al-mawà∂i' (Riyà∂: al-Yamàma, 1388/1968), 189: al-'ulfà . . . an yaj'ala [note the verb] l-insàn 'inda ßaràm sha'ìrihi wajazz qa∂bihi li-khafìr aw li-ßadìq shay"an yu'†ìhi iyyàhu. Wa-li-banì l-Sharìd min banì Sulaym 'alà zurrà' Iran 'ulfà 'inda ˙aßàd kull sha'ìr ilà l-yawm; quoted in Lecker, The Banù Sulaym, 225. Lane, Arabic-English Lexicon (Cambridge, 1984), has under al-'alfà: “What a man assigns, on the occasion of the reaping of his barley, to a guardian [thereof ] from the birds, or to a friend” (Tàj al-'arùs, quoting Hajarì). The entitlement of the nomads to the grants is expressed by similar phrases: kànat lahu ja'àla 'alà ahl Fadak; wa-li-banì l-Sharìd min banì Sulaym 'alà zurrà' Iran 'ulfà; wa-kàna li-'Àmir b. Màlik min ahl Yathrib kayla min tumrànihim. 38 Lecker, EI 2, s.v.

the jewish tribes in arabia

63

had nothing to do with the religion of the former. The treaty secured for the 'Udhra a portion of the annual crop of Wàdì l-Qurà (the terms used are †u'ma and uk[u]l ) in return for their driving away from Wàdì l-Qurà the Balì (another tribe of the Qu∂à'a) and other tribes.39 So this specific case was not connected to safe conduct but to the protection of the settlement itself. The 'Udhra were in immediate proximity to the settlement, while the large nomadic tribes which received the Kayla/ja'àla/'ulfà were roaming vast territories and only came closer to the settlements in the summer time, when water supply became scarce. Consequently, 'Udhra’s share in the crops which amounted to one-third must have been larger than the one allotted to the nomadic tribes. In any case, 'Udhra’s share does not point to its ascendancy over the settled population of Wàdì l-Qurà. The advent of Islam improved 'Udhra’s position. Before Islam, the remaining two-thirds of the annual harvest remained in the hands of the Jewish cultivators. When Mu˙ammad conquered Wàdì l-Qurà (7 AH), he took half of the Jews’ share, i.e., one-third of the total, while the Jews kept one-third to themselves. When the Jews were expelled by 'Umar b. al-Kha††àb (r. 13–23/634–644), they received in cash the estimated value of their share, namely ninety thousand dinars. 'Umar then offered the 'Udhra an additional sixth of the crops in return for one-sixth of the value, i.e., forty-five thousand dinars. They accepted his offer, becoming the owners of half (onethird plus one-sixth) of Wàdì l-Qurà’s produce.40 Some comparative material on the relationship between settled and nomads is linked to al-Suwàriqiyya, the central settlement in the land of the Sulaym. The settlement belonged to the Sulaym alone and “every Sulamì had a share in it.” The following passage found in a geographical treatise by 'Arràm al-Sulamì (fl. third/ninth century)

39 Wa-kàna lahum fìhà [i.e. in Wàdì l-Qurà] 'alà l-yahùd †u'ma wa-ukl fì kull 'àm wamana'ùhà lahum mina l-'arab wa-dafa'ù 'anhà qabà"il Balì b. 'Amr b. Il˙àf b. Qu∂à'a waghayrahum mina l-qabà"il; al-Bakrì, Mu'jam mà sta'jam, I, 43; see similarly Yàqùt, Buldàn, s.v. al-Qurà. See also Kister, art. “Qu∂à'a,” EI 2, at 317b; reprinted in idem, Concepts and Ideas at the Dawn of Islam, no. III, 7. 40 One third of the remaining half was included in the charitable endowments or ßadaqàt of the Prophet, while one-sixth of the same half belonged to all the Muslims; al-Màwardì, al-A˙kàm al-sul†àniyya wa-l-wilàyàt al-dìniyya, ed. A˙mad Mubàrak al-Baghdàdì (Kuwait: Maktabat Dàr Ibn Qutayba, 1409/1989), 219 (read instead of wàdì l-qarya: Wàdì l-Qurà); idem, Les statuts gouvernementaux, trans. E. Fagnan (Alger: Typographie Adolphe Jourdan, 1915), 362; idem, The Ordinances of Government, trans. Wafaa H. Wahba (London: Garnet, 1996), 187.

64

michael lecker

probably relates to his own time; but it can shed some light on conditions in pre-Islamic Arabia: They [i.e., the Sulamìs] are bàdiya, except those born in it [i.e., in alSuwàriqiyya] who live there. The others roam around it and supply food along the two pilgrim roads, namely the road of the Óijàz and the road of Najd. And the limit is Îariyya which is the border point, at a distance of a seven days’ journey [from al-Suwàriqiyya].41

In this case the settled and the nomads belonged to the same tribe. The farmers used their expertise to till the land, while the nomads took care of the beasts, above all the camels which require extensive grazing grounds and hence cannot be pastured by the farmers. The produce of al-Suwàriqiyya was transported on camelback and was sold to the pilgrims in their stations as far as Îariyya on a seven days’ journey from al-Suwàriqiyya. This cooperation between the settled and the nomads among the Sulaym is comparable to the pre-Islamic one between the Jewish settlement Khaybar and its nomadic neighbors, who for a large fee transported and sold Khaybar’s surplus of dates. During the building of his mosque in Medina Mu˙ammad himself reportedly carried bricks, reciting two rajaz verses. The first of these verses is: I would rather carry this load [of bricks] than the [fruit] load of Khaybar this, our Lord, is purer and more moral (hàdhà l-˙imàlu là ˙imàla Khaybar\hàdhà abarru rabbanà wa-a†har).

It is explained on the authority of al-Zuhrì (d. 124/742) that when the Jews cut off the fruit of their palm trees, the nomads would come to them with their camels and carry it for them to the villages, one camel load after the other ('urwa bi-'urwa, literally, “one loop of the camel load after the other”). They would sell (the fruit), 41 'Arràm al-Sulamì, “Asmà" jibàl Tihàma,” in 'Abd al-Salàm Hàrùn, ed., Nawàdir al-makh†ù†àt2 (Cairo: Mu߆afà l-Bàbì l-Óalabì, 1393/1973), II, 431–32; the variants between square brackets are from Yàqùt, Buldàn, s.v. al-Suwàriqiyya: al-Suwàriqiyya . . . libanì Sulaym khàßßa wa-li-kull [add. min] banì Sulaym minhà [ fìhà] shay" . . . wa-hum bàdiya illà man wulida bihà fa-innahum tànùna [thàbitùna; lectio facilior] fìhà wa-l-àkharùna bàdùna ˙awàlayhà wa-yamìrùna †arìqa l-Óijàz wa-Najd fì †arìqayi l-˙ajj, wa-l-˙add [wa-ilà ˙add] Îariyya wa-ilayhà yantahì ˙adduhum 'alà sab' marà˙il; quoted in Lecker, The Banù Sulaym, 222. Cf. al-Bakrì, Mu'jam mà sta'jam, I, 100, where there is no reference to trade: al-Suwàriqiyya . . . wa-˙adduhà yantahì ilà Îariyya. Also the journey from Îariyya to Medina took seven days; Ibn Sa'd, ˇabaqàt, II, 78.

the jewish tribes in arabia

65

keeping to themselves half of the return.42 Precisely like the people of al-Suwàriqiyya under Islam, the Jews of Khaybar did not have camels of their own and hence the surplus of their produce was transported for them by the nomads; the latter were not mere carriers but also handled the transactions on behalf of the Jewish growers.43 'Uyayna’s itàwa During the siege of Medina (or the battle of the Khandaq, 5/627) Mu˙ammad is supposed to have promised 'Uyayna b. Óißn and alÓàrith b. 'Awf, the respective leaders of the Fazàra and Murra tribes (both of which belonged to the Gha†afàn federation) one-third of Medina’s produce of dates. In return, the two were to retreat and to induce the other nomads to abstain from fighting (tarji 'àni bi-man ma'akum [sic] wa-tukhadhdhilàni bayna l-a'ràb). They demanded one half of the produce but finally settled for a third. However, Anßàrì opposition called the deal off. The leaders of the Anßàr told the Prophet that before Islam the Bedouin wretches (literally, 'ilhiz-eaters, 'ilhiz being a dish eaten at times of famine which was made of camelhair mixed with blood) could not have had one single date unless they purchased it or were offered it as guests (bi-shiran aw qiran).44 The fiery speeches of the Anßàrì leaders are a fine example of Anßàrì historiography; whether they reflect historical fact is another matter

42 Kànat yahùd idhà ßaramat nakhlahà jà"athumu l-a'ràb bi-rakà"ibihim fa-ya˙milùna lahum 'urwa bi-'urwa ilà l-qurà fa-yabì'ùna, yakùnu li-hàdhà [sic] nißfu l-thaman wa-li-hà"ulà"i nißfuhu; al-Samhùdì, I, 328. Cf. Gil, “The origin of the Jews of Yathrib,” JSAI 4 (1984), 204. Ibn Óajar, Muqaddimat fat˙ al-bàrì shar˙ ßa˙ì˙ al-Bukhàrì (Bùlàq, 1301/1884), 106:7, explains the Prophet’s words by saying that in the Afterlife these building stones [he refers to the above mentioned bricks—M.L.] are better than the dates carried from Khaybar. 43 The mention of bricks indicates that this was not the mosque which the Prophet built shortly after the Hijra with palm boughs stripped off; cf. Kister, “‘A booth like the booth of Moses . . .’,” BSOAS 25 (1962), 150–55. Rather, it was the mosque which according to one source was built of bricks four years after the Hijra. Alternatively it is reported that this second stage in the building took place after the conquest of Khaybar (7 AH/628 CE); al-Suyù†ì, al-Óujaj al-mubìna fì l-taf∂ìl bayna Makka wa-l-Madìna, ed. 'Abdallàh Mu˙ammad al-Darwìsh (Damascus/Beirut: al-Yamàma, 1405/1985), 52. 44 Al-Wàqidì, II, 477–78. On 'ilhiz, see Lisàn al-'Arab, s.v.; Kister, “O God, tighten Thy grip on Mu∂ar . . .,” JESHO 24 (1981), 246–47.

66

michael lecker

altogether. Elsewhere there is evidence that 'Uyayna received an annual itàwa from the date produce of Medina (again there is no specific mention of the Jews). The report on 'Uyayna’s itàwa merits trust because it is at the background of an account which is critical of 'Uyayna. He is said to have ridden to Medina (the wording seems to suggest that he was in a hurry) to collect his annual itàwa, rather than mediate in an internal strife among the Dhubyàn (another tribe of the Gha†afàn federation).45 In this context itàwa does not mean a tribute or tax but an annual grant in kind made to a nomadic leader, precisely like the abovementioned kayla and ja'àla. The size of this and similar grants appears to have been humble. Medina and the other settlements had a huge surplus of dates and hence could afford to grant part of it to the leaders of large nomadic tribes in order to secure their good will. The size of the grants presumably varied according to the harvest and the political situation; but even if they were to amount to onethird or one-half of the annual produce, this would not make the inhabitants of the settlements, be they Jews or Arabs, clients of the nomadic beneficiaries.

Bibliography Primary Sources Abù l-Baqà" Hibat Allàh al-Óillì. Al-Manàqib al-mazyadiyya, ed. Íàli˙ Mùsà Daràdika and Mu˙ammad 'Abd al-Qàdir Khrìsàt. Amman: Maktabat al-Risàla l-Óadìtha, 1404/1984. Aghànì—Abù l-Faraj al-Ißfahànì. Kitàb al-aghànì. Bùlàq 1285/1868. 'Arràm b. al-Aßbagh al-Sulamì. “Asmà" jibàl Tihàma,” in 'Abd al-Salàm Hàrùn, ed. Nawàdir al-makh†ù†àt2, vol. 2. Cairo: Mu߆afà l-Bàbì l-Óalabì, 1393/1973, 373–441. al-Bakrì, Abù 'Ubayd. Al-Mamàlik wa-l-masàlik, MS Nur Osmaniye. Ed. A.P. Van Leeuwen and A. Ferre. Tunis, 1992. id., Mu'jam mà sta'jam, ed. Mu߆afà al-Saqqà. Cairo: Lajnat al-Ta"lìf wa-l-Tarjama wa-l-Nashr, 1364–71/1945–51; repr. Beirut: 'Àlam al-Kutub, n.d. id., Sim† al-la"àlì fì shar˙ amàlì l-Qàlì, ed. 'Abd al-'Azìz al-Maymanì. Cairo 1354/1936. al-Balàdhurì, Abù l-'Abbàs A˙mad b. Ya˙yà. Futù˙, ed. M.J. de Goeje. Leiden: Brill, 1863–66. al-Diyàrbakrì, Óusayn b. Mu˙ammad. Ta"rìkh al-khamìs. Cairo: al-Ma†ba'a l-Wahbiyya, 1283/1866; repr. Beirut: Mu"assasat Sha'bàn, n.d. 45 Wa-yuqàlu innahu kànat lahu itàwa 'alà ahl Yathrib ya"khudhuhà fì kull 'àm . . .; Ibn Shabba, Medina, II, 538–39. A verse included in this report specifically refers to dates.

the jewish tribes in arabia

67

Fà"iq—Abù l-Qàsim Ma˙mùd b. 'Umar al-Zamakhsharì. Al-Fà"iq fì gharìb al-˙adìth, ed. 'Alì Mu˙ammad al-Bijàwì and Mu˙ammad Abù l-Fa∂l Ibràhìm. Cairo: 'Ìsà l-Bàbì l-Óalabì, 1971. al-Hamdànì, al-Óasan b. A˙mad. Kitàb qaßìdat al-dàmigha, ed. al-Akwa'. Cairo [1384/1964]. Óassàn b. Thàbit. Dìwàn, ed. W. 'Arafat. London: Luzac, 1971. Ibn 'Asàkir. Ta"rìkh madìnat Dimashq, ed. 'Umar b. Gharàma al-'Amrawì. Beirut 1415–19/1995–98. Ibn Óajar, Abù l-Fa∂l A˙mad b. 'Alì al-'Asqalànì. Al-Ißàba fì tamyìz al-ßa˙àba, ed. 'Alì Mu˙ammad al-Bijàwì. Cairo: Dàr Nah∂at Mißr, 1392/1972. id., Muqaddimat fat˙ al-bàrì shar˙ ßa˙ì˙ al-Bukhàrì. Bùlàq 1301/1884; repr. Beirut: Dàr al-Ma'rifa, n.d. Ibn Óazm al-Andalusì. Jamharat ansàb al-'arab, ed. 'Abd al-Salàm Hàrùn. Cairo: Dàr al-Ma'àrif, 1382/1962. Ibn Hishàm, Abù Mu˙ammad 'Abd al-Malik. Al-Sìra al-nabawiyya, ed. Mu߆afà alSaqqà, Ibràhìm al-Abyàrì and 'Abd al-ÓafìΩ Shalabì. Beirut: I˙yà" al-Turàth al'Arabì, 1391/1971; English trans. A. Guillaume, The Life of Muhammad: A Translation of Ibn Ishaq’s Sìrat Rasùl Allàh. Oxford: Oxford University Press, 1955. Ibn al-Kalbì, Hishàm b. Mu˙ammad. Nasab Ma'add wa-l-yaman al-kabìr, ed. Nàjì Óasan. Beirut: 'Àlam al-Kutub-Maktabat al-Nah∂a l-'Arabiyya, 1408/1988. Ibn Kathìr, Ismà'ìl b. 'Umar. Al-Bidàya wa-l-nihàya fì l-ta"rìkh. Cairo: Ma†ba'at alSa'àda, al-Ma†ba'a al-Salafiyya and Maktabat al-Khànjì, 1351–58/1932–39; repr. Beirut: Maktabat al-Ma'àrif, 1974. Ibn Màkùlà, 'Alì b. Hibat Allàh. Al-Ikmàl, ed. 'Abd al-Ra˙màn b. Ya˙yà al-Yamànì. Hyderabad: Dà"irat al-Ma'àrif al-'Uthmàniyya, 1381/1962. Ibn Sa'd, Mu˙ammad. Al-ˇabaqàt al-kubrà. Beirut: Dàr Íàdir-Dàr Bayrùt, 1380–88/1960–68. Ibn Shabba, 'Umar. Ta"rìkh al-Madìna al-munawwara, ed. Fahìm Mu˙ammad Shaltùt, n.p. [1399/1979]; repr. Beirut: Dàr al-Turàth and al-Dàr al-Islàmiyya, 1410/1990. 'Iqd—Ibn 'Abd Rabbihi, al-'Iqd al-farìd, ed. A˙mad Amìn, A˙mad al-Zayn and Ibràhìm al-Abyàrì. Cairo: Lajnat al-Ta"lìf wa-l-Tarjama wa-l-Nashr, 1359–72/ 1940–53. al-Jà˙iΩ, 'Amr b. Ba˙r. Rasà"il, ed. 'Abd al-Salàm Hàrùn. Cairo: al-Khànjì, 1384–99/1964–79. Jazìrat al-'arab min kitàb al-mamàlik wa-l-masàlik li-Abì 'Ubayd al-Bakrì, ed. 'Abdallàh Yùsuf al-Ghunaym. Kuwayt: Dhàt al-Salàsil, 1397/1977. Kashf al-ghumma al-jàmi' li-akhbàr al-umma, ed. A˙mad 'Ubaydalì. Nicosia: Dalmùn lil-Nashr, 1405/1985. Lisàn al-'Arab—Ibn ManΩùr Abù l-Fa∂l Jamàl al-Dìn. Lisàn al-'Arab. Beirut: Dàr Íàdir, 1374–76/1955–56; repr. Beirut: Dàr Íàdir, 1968. Maghànim—al-Fìrùzàbàdì, Majd al-Dìn Mu˙ammad b. Ya'qùb. Al-Maghànim almu†àba fì ma'àlim ˇàba, ed. Óamad al-Jàsir. Riyà∂: Yamàma, 1389/1969. al-Maqrìzì, A˙mad b. 'Alì Taqì al-Dìn. Imtà' al-asmà' bi-mà li-l-rasùl mina l-anbà" wa-l-amwàl wa-l-˙afada wa-l-matà', ed. Ma˙mùd Mu˙ammad Shàkir, vol. 1. Cairo: Lajnat al-Ta"lìf wa-l-Tarjama wa-l-Nashr, 1941. al-Màwardì, 'Alì b. Mu˙ammad. Al-A˙kàm al-sul†àniyya wa-l-wilàyàt al-dìniyya, ed. A˙mad Mubàrak al-Baghdàdì. Kuwayt: Maktabat Dàr Ibn Qutayba, 1409/1989; English trans. Wafaa H. Wahba, The Ordinances of Government. London: Garnet, 1996; French trans. E. Fagnan, Les statuts gouvernementaux. Alger: Typographie Adolphe Jourdan, 1915. Naßr b. 'Abd al-Ra˙màn al-Fazàrì al-Iskandarànì Abù l-Fat˙. Al-Amkina wa-l-miyàh wa-l-jibàl wa-l-àthàr, MS Br. Lib. Add. 23, 603. Qur†ubì, Tafsìr—Abù 'Abdallàh Mu˙ammad b. A˙mad al-Qur†ubì. Al-Jàmi' li-a˙kàm al-qur"àn3. Cairo: Dàr al-Kutub, 1354–87/1935–67.

68

michael lecker

al-Samhùdì, 'Alì b. A˙mad. Wafà" al-wafà bi-akhbàr dàr al-mu߆afà, ed. Mu˙ammad Mu˙yì l-Dìn 'Abd al-Óamìd. Cairo 1374/1955; repr. Beirut: Dàr I˙yà" al-Turàth al-'Arabì, 1401/1981. al-Suyù†ì, Jalàl al-Dìn. Al-Óujaj al-mubìna fì l-taf∂ìl bayna Makka wa-l-Madìna, ed. 'Abdallàh Mu˙ammad al-Darwìsh. Damascus/Beirut: al-Yamàma, 1405/1985. al-ˇabarì—Abù Ja'far Mu˙ammad b. Jarìr al-ˇabarì. Ta"rìkh al-rusul wa-l-mulùk, ed. M.J. de Goeje et al. Leiden: Brill, 1879–1901. al-ˇabarì, Tafsìr—id., Jàmi' al-bayàn fì tafsìr al-qur"àn. Bùlàq, 1321–30/1903–12. al-Wàqidì, Abù 'Abdallàh Mu˙ammad b. 'Umar. Kitàb al-maghàzì, ed. Marsden Jones. London: Oxford University Press, 1966. al-Ya'qùbì, A˙mad b. Abì Ya'qùb. Ta"rìkh. Beirut: Dàr Íàdir and Dàr Bayrùt, 1379/1960. Yàqùt al-Óamawì. Mu'jam al-buldàn. Beirut: Dàr Íàdir and Dàr Bayrùt, 1957. Secondary Sources Altheim, F. and R. Stiehl. Finanzgeschichte der Spätantike. Frankfurt am Main: V. Klostermann, 1957. Bonner, M. Aristocratic Violence and Holy War: Studies in the Jihad and the Arab-Byzantine Frontier. New Haven: American Oriental Society, 1996. id., “Ja'à"il and holy war in early Islam,” Der Islam 68 (1991), 45–64. Caskel W. and G. Strenziok. ]amharat an-Nasab: Das genealogische Werk des Hi“àm ibn Mu˙ammad al-Kalbì. Leiden: Brill, 1966. Crone, P. Roman, Provincial and Islamic Law: The Origins of the Islamic Patronate. Cambridge: Cambridge University Press, 1987. GAS—Sezgin, F. Geschichte des arabischen Schrifttums. Leiden: Brill, 1967–. Gil, M. “The origin of the Jews of Yathrib,” JSAI 4 (1984), 203–24; repr. F.E. Peters, ed., The Arabs and Arabia on the Eve of Islam, (The Formation of the Classical Islamic World, 3.) Aldershot: Variorum, 1999, 45–66. Hasson, I. “Le chef judhàmite Raw˙ ibn Zinbà',” Studia Islamica 77 (1993), 95–122. Hirschberg, H.Z. Yisra"el be-'arav. Tel-Aviv: Mossad Bialik, 1946. al-Jàsir, Ó. Abù 'Alì l-Hajarì wa-ab˙àthuhu fì ta˙dìd al-mawà∂i'. Riyà∂: al-Yamàma, 1388/1968. id., Fì shimàl gharb al-jazìra (nußùß, mushàhadàt, in†ibà'àt)2. Riyà∂, 1401/1981. id., “Qirà"a sarì'a fì kitàb Ta"rìkh al-Madìna li-Ibn Shabba,” al-'Arab 18 (1983–84), 289–356. Kister, M.J. “‘A booth like the booth of Moses . . .’,” BSOAS 25 (1962), 150–55; repr. idem, Studies in Jàhiliyya and Early Islam, no. 8. id., Concepts and Ideas at the Dawn of Islam. Aldershot: Variorum, 1997. id., “The expedition of Bi"r Ma'ùna,” in G. Maqdisi, ed., Arabic and Islamic Studies in Honor of H.A.R. Gibb. Leiden: Brill, 1965, 337–57; repr. idem, Studies in Jàhiliyya and Early Islam, no. 10. id., “‘Óaddithù 'an banì isrà"ìla wa-là ˙araja’,” Israel Oriental Studies 2 (1972), 215–39; repr. with additional notes, idem, Studies in Jàhiliyya and Early Islam, no. 14. id., “O God, tighten Thy grip on Mu∂ar . . .,” JESHO 24 (1981), 242–73; repr. idem, Society and Religion, no. 7. id., “On the wife of the goldsmith from Fadak and her progeny,” Le Muséon 92 (1979), 321–30; repr. idem, Society and Religion, no. 5. id., Society and Religion from ⁄àhiliyya to Islam. Aldershot: Variorum, 1990. id., Studies in Jàhiliyya and Early Islam. London: Variorum, 1980. Lane, E.W. Arabic-English Lexicon. Cambridge, 1984 (repr. in two vols.). Lecker, M. The Banù Sulaym: A Contribution to the Study of Early Islam. Jerusalem: Institute of Asian and African Studies, 1989. id., The Constitution of Medina. Princeton: Darwin Press, 2004.

the jewish tribes in arabia

69

id., Jews and Arabs in Pre- and Early Islamic Arabia. Aldershot: Variorum, 1998. id., “The levying of taxes for the Sassanians in Pre-Islamic Medina (Yathrib),” JSAI 27 (2002), 109–26; repr. idem, People, Tribes and Society in Arabia Around the Time of Mu˙ammad (Aldershot: Ashgate, 2005), no. 1. id., “Mu˙ammad at Medina: a geographical approach,” JSAI 6 (1985), 29–62; repr. idem, Jews and Arabs, no. 8. id., Muslims, Jews and Pagans: Studies on Early Islamic Medina. Leiden: Brill, 1995. id., “Wàqidì’s account on the status of the Jews of Medina: a study of a combined report,” JNES 54 (1995), 15–32; repr. Rubin, U. The Life of Mu˙ammad, 23–40; also repr. Lecker, M. Jews and Arabs, no. 7. Levi della Vida, G. “Pre-Islamic Arabia,” in Nabih Amin Faris, ed., The Arab Heritage. Princeton: Princeton University Press, 1944, 25–57. Motzki, H. “The murder of Ibn Abì l-Óuqayq: on the origin and reliability of some Maghàzì-Reports,” in idem, ed., The Biography of Mu˙ammad: The Issue of the Sources. Leiden: Brill, 2000, 170–239. Rubin, U. ed. The Life of Mu˙ammad (The Formation of the Classical Islamic World, 4.) Aldershot: Variorum, 1998. Schwarzlose, F.W. Die Waffen der alten Araber aus ihren Dichtern dargestellt. Leipzig: J.C. Hinrichs’sche Buhchhandlung, 1886. Serjeant, R.B. “The sunnah jàmi'ah, pacts with the Yathrib Jews, and the Ta˙rìm of Yathrib: analysis and translation of the documents comprised in the so-called ‘Constitution of Medina’,” BSOAS 41 (1978), 1–42; repr. Rubin, U. The Life of Mu˙ammad, 151–192. El-Shami, A. and R.B. Serjeant. “Regional literature: the Yemen,” in J. Ashtiany et al., eds., 'Abbasid Belles-lettres, The Cambridge History of Arabic Literature, II. Cambridge: Cambridge University Press, 1990, 442–68. Watt, W.M. Muhammad at Medina. Oxford: The Clarendon Press, 1956. Wellhausen, Skizzen IV—Wellhausen, J. Medina vor dem Islam, Skizzen und Vorarbeiten, IV/1. Berlin 1889. Wensinck, A.J. Muhammad and the Jews of Medina, with an excursus: Muhammad’s Constitution of Medina, by Julius Wellhausen, trans. and ed. W.H. Behn. Freiburg im Breisgau: Klaus Schwarz Verlag, 1975; 2nd ed. Berlin: Adiyok, 1982 (= Mohammed en de Joden te Medina. Leiden: Brill, 1908).

ORIGIN AND DEVELOPMENT OF THE ISLAMIC PATRONATE Ulrike Mitter

Two kinds of patronate (walà") exist in Islamic law: servile patronate (walà" al-'itq) and contractual patronate (walà" al-muwàlàt). Walà" al'itq arises as a consequence of manumission of a slave and is acknowledged by all Sunnì schools. Walà" al-muwàlàt comes about through a contract between two free parties; in classical Islamic law, this form of patronate is acknowledged only by the Óanafìs and by some Shì'ì schools.1 This article deals with both kinds of walà" according to the Sunnì legal literature.2 Mainly, the development of the legal theory is dealt with, though toward the end of the article some information on the historic practice is given as well. The first part of the article offers an outline of the discussion on the origins of walà" and an interpretation of the sources. The second part examines the regulations of walà" as expressed in early legal traditions (a˙àdìth, sing. ˙adìth).3

1 For the Shì'ì schools, see Crone, Roman, provincial and Islamic law. The origins of the Islamic patronate (Cambridge: Cambridge University Press, 1987), 38. 2 Shì'ì traditions often have an Imàm as their oldest authority. Because of the lack of intersection of authorities, a comparative isnàd analysis of Shì'ì and Sunnì a˙àdìth is almost impossible. On the problem of Shì'ì asànìd (in the sìra literature), see Jarrar, “Sìrat Ahl al-Kisà". Early Shi'ì Sources on the Biography of the Prophet,” in Motzki, ed., The Biography of Mu˙ammad: The Issue of the Sources, Islamic History and Civilization. Studies and Texts (Leiden/Boston/Cologne: Brill, 2000), 3 and 6–8. Jarrar assumes that a “significant corpus of materials” was circulating in the Óijàz, Iraq and Syria in the first half of the second century and that this corpus was also used by Shì'ì scholars. Often, their mutùn are similar to Sunnì traditions and do not reveal a “particularly Shì'ì tenet” (ibid., 6). By an analysis of the mutùn, the relation between Shì'ì and Sunnì material can be described to a certain degree. 3 When speaking of “early a˙àdìth,” I mean the a˙àdìth which according to the results of a detailed isnàd-cum-matn analysis stem from the first and second Islamic century.

the origin and development of the islamic patronate

71

The origins of walà" Theories on the origins of walà" Controversy exists in the secondary literature about the origins of walà". There can be no doubt that pre-Islamic Arabs knew the institution of patronate,4 but we have to ask if the patronate they were acquainted with was the same institution as the one known in classical Islamic law. Did the classical Islamic institution emerge out of an old-Arabian institution or did early Muslims adopt and modify a non-Arab institution? A short overview of the most important secondary literature follows. In the nineteenth century, Alfred von Kremer mentioned similarities between Roman patronate and Islamic walà", namely the act of manumission of slaves, testamentary manumission and the mutual right of patron and client to inherit from each other if there were no heirs.5 By mentioning manumission, he obviously thought of walà" al-'itq but by mentioning mutual right of inheritance, he could have had only walà" al-muwàlàt in mind.6 Despite the similarities, von Kremer did not postulate any Roman influence on the Islamic patronate. He argued that slavery and patronate are very old Arabian and Semitic institutions which go back to a time much earlier than the foundation of the Roman empire7 thereby implying that the Islamic institution has its roots in old-Arabian customary law. Pre-Islamic Arabian roots for walà" al-'itq are also proclaimed by Jamal Juda.8 4 For details on pre-Islamic walà", see Juda, Die sozialen und wirtschaftlichen Aspekte der Mawàlì in frühislamischer Zeit (Tübingen: REFO-Druck Hans Vogler, 1983), 1ff.; Reinert, Das Recht in der altarabischen Poesie (Ph.D. diss., Cologne, 1963), 20–22. 5 Von Kremer, Culturgeschichte des Orients unter den Chalifen, 2 vols. (Vienna: Wilhelm Braumüller, 1875–77), I, 525–26. 6 Almost all Muslim jurists assume succession to be unilateral in the case of walà" al-'itq (cf. below). Traditions on walà" al-muwàlàt, too, do not mention the right of the client to succession but in classical law, the right to succession is mutual if this was stipulated. See for example al-Sarakhsì, Al-Mabsù†, 30 parts in 15 vols. (Cairo, 1324–31/1906–13), XXX, 45, ll. 2–3. For additional references, see Crone, Roman law, 39 note 52; Hallaq, “The Use and Abuse of Evidence: The Question of Provincial and Roman Influence on Early Islamic Law,” Journal of American Oriental Society 110/I (1990), 8 note 23. The question is, however, why von Kremer made this comparison in the first place since in Roman law, the client is not the heir of his patron. For the status of Roman freedmen, see Hausmaninger and Selb, Römisches Privatrecht (Vienna/Cologne: Böhlau Verlag, 1981), 126–27. 7 Von Kremer, Culturgeschichte, I, 547. Cf. Hallaq, “Use and Abuse,” 28. 8 “Walà" al-'itàqa bestand nach den Eroberungen fort, ohne einen neuen Namen zu erhalten.”; Juda, Aspekte, VI.

72

ulrike mitter

As to walà" al-muwàlàt, Ignaz Goldziher, a contemporary of von Kremer, claimed in 1889 that it had its roots in pre-Islamic alliance (˙ilf ).9 Goldziher was followed by Emile Tyan (1971),10 Juda (1983) and Wael Hallaq (1989). According to Juda, walà" al-tibà'a (another term for muwàlàt) “ist also dem Wesen nach (als freiwilliger Zusammenschluß zweier Parteien) nichts anderes als eine Fortführung des vorislamischen walà" al-˙ilf.”11 Hallaq argues that ˙ilf and walà" almuwàlàt are both contracts for mutual help between two equal parties and neither requires separation from their natal group. Hallaq adds that in both institutions, an individual could attach himself to a clan or to a single member of the clan. The new client (mawlà, pl. mawàlì )12 had the same rights and duties as other members of the tribe.13 Alongside these similarities differences do exist between the two institutions. This is what leads Patricia Crone to conclude that the origins of contractual walà" cannot be found in ˙ilf. She points out that the right to succession of an ally (˙alìf, pl. ˙ulafà") was based on his obligation to avenge the blood of his ally whereas the entitlement of an Islamic patron is based on his obligation to pay bloodmoney on behalf of his client. Furthermore, the patron does not receive a fixed share of the inheritance of his client as the pre-Islamic ˙alìf is said to have had but he was entitled to the entire estate.14 Crone argues that the individual Islamic walà" (both contractual walà" and the servile tie) does not have any roots in pre-Islamic Arabia since the pre-Islamic Arabs did not know an individual patronate but only collective relationships.15 On the contrary, the legal relationship which existed between a manumitter and a freed-

9 Goldziher, Muhammedanische Studien, 2 vols. (Halle: Max Niemeyer, 1889–90), I, 63 and 106. 10 Tyan, art. “Óilf,” EI 2, III, 388. 11 Juda, Aspekte, 75. 12 When used in this article without specification, mawlà means client, although in the a˙àdìth, the term mawlà was also used for the patron. 13 Hallaq, “Use and Abuse,” 13–14. For more references on individual ˙ilf and walà", see below. Also see Landau-Tasseron’s chapter in this book about alliance and clientage. 14 Crone, Roman law, 53–54 and 84. Crone’s first argument against possible roots of walà" in ˙ilf (i.e., ˙ilf only resembles contractual walà" and not walà" over freedmen) is, however, not effective because servile and contractual walà" are two different institutions anyway. 15 Ibid., 43.

the origin and development of the islamic patronate

73

man resembled more a labor contract than a patronate.16 According to Crone, the source of Islamic walà" is to be found in the lateRoman patronate which contains certain similarities with pre-classical walà": both institutions could be renounced, sold and given away, and were inherited like ordinary property.17 Crone suggests that it was an early Umayyad caliph in Syria—she mentions Mu'àwiya (d. 60/680)—who adopted the late-Roman institution. From the beginning of the second Islamic century onwards (eighth century CE), the early regulations were modified by the lawyers. The institution became an inalienable agnatic tie and walà" on free converts was rejected.18 These opposite conclusions point to a methodological problem. When comparing legal institutions (of different periods and/or legal systems) to find out if one institution is dependent on or has been developed from another, it is difficult to determine and interpret the differences and parallels. As Jonathan Brockopp puts it: “How one defines an institution greatly affects the parallels one sees.”19 I will argue in this article that an analysis of the a˙àdìth can be useful when trying to settle this point.

16 Ibid., 58. According to Crone, ibid., 67–68, the existence of pre-Islamic labor contracts are an indication of the acquaintance of the Arabs with the Hellenistic paramonè (conditional manumission of slaves which obliged the former slave to serve his master for several years before gaining full freedom on the death of his master). The paramonè, Crone contents, influenced the development of the Islamic kitàba (emancipation contract for a slave who buys his freedom). For counterarguments to Crone’s thesis, see Hallaq, “Use and abuse,” 17–21, who stresses the differences between kitàba and paramonè. 17 Crone, Roman law, 84–86. See also id., Slaves on Horses, the Evolution of the Islamic Polity (Cambridge: Cambridge University Press, 1980), 49. As Crone points out, it is highly unlikely that the Arabs introduced elements of classical Roman law into Islamic law (Roman law, 103). If there were Roman elements in Islamic law, it should be elements of late-Roman law. These elements were not introduced directly into Islamic law, however, but through provincial practice which influenced both Islamic law and Roman law (ibid., 91–92). 18 Ibid., 91. According to Joseph Schacht, The Origins of Muhammadan Jurisprudence (Oxford: The Clarendon Press, 1950), 161, walà" al-muwàlàt was created during the Umayyad period. He does not tell us where he thinks this institution came from but his theory is compatible with Crone’s dating. 19 Brockopp, Early Màlikì Law. Ibn 'Abd al-Óakam and his Major Compendium of Jurisprudence, Studies in Islamic Law and Society (Leiden/Bosten/Cologne: Brill, 2000), 169 note 25. For more details on this methodological problem, see Mitter, “Problemen van het onderzoek naar ontleningen aan niet-Arabische rechtsstelsels in het ontstaan en de ontwikkeling van het islamitisch recht,” Sharqiyyàt 9/II (1997), 107–23.

74

ulrike mitter

The problem of the sources The reason for the difficulties in determining the origins and the developent of walà" is the Quellenlage or “layers of sources.” There are a˙àdìth dealing with walà" which claim to stem from early authorities. If these a˙àdìth were authentic, the origin of walà" would be obvious: the institution would simply go back to Mu˙ammad and his Companions (ßa˙àba, sing. ßa˙àbì) who followed old-Arabian rules in the absence of new Qur’ànic regulations. But in the view of many Western scholars, most of the a˙àdìth reflect a later stage in the development of Islamic law. They believe that the lawyers gradually ascribed their own opinion (ra"y) to older and older authorities: first to the Successors (tàbi'ùn, sing. tàbi' ), then to the ßa˙àba and finally to the Prophet himself. The older the authority, the younger the tradition.20 Not only the authenticity of the a˙àdìth is called into doubt but also the authenticity of the Qur"àn itself.21 Without Qur"àn and ˙adìth, we are left with almost no material on Islamic law in the first one and a half centuries. Every statement on the development of legal institutions in this period would then be pure speculation. This is why Schacht says that the first century will remain a dark spot as far as Islamic law is concerned. Arguing on the basis of his theory of the late development of a˙àdìth, Schacht states that Muslims in the first Islamic century (seventh century) were indifferent towards

20 In 1890, Ignaz Goldziher explained the origin of traditions as the consequence of juridical discussions between aß˙àb al-ra"y and aß˙àb al-˙adìth. In his view, the latter were “less honest” because they were ready to invent traditions, if necessary (Studien, II, 73ff.). Joseph Schacht referred to Goldziher’s theory in 1950 and developed it further. Until the contrary is shown, Schacht proclaimed that Prophetic a˙àdìth should be treated as spurious (Origins, 3–5, 15ff., 149 and passim). In 1980, Noth spoke about “Fälschungen” and “pia fraus” (“Die Scharìa, das religiöse Gesetz des Islam—Wandlungsmöglichkeiten, Anwendung und Wirkung,” in Wolfgang Fikentscher, Herbert Franke and Oskar Köhler, eds., Entstehung und Wandel rechtlicher Traditionen (Freiburg and Munich: Karl Alber, 1980), 419). In 1983, Gautier Juynboll tried to provide answers as to where, when and by whom Prophetic traditions were put into circulation (Muslim Tradition. Studies in Chronology, Provenance and Authorship of Early ˙adìth (Cambridge: Cambridge University Press, 1983), 71–74). Because of their problematic nature, legal a˙àdìth are still dismissed as a source for early fiqh by Brockopp in 2000 (Law, 124–26). 21 Cf. Brockopp, Law, 119–24, who refers to the arguments of critical scholars as Wansbrough and Crone and explains why he still uses the Qur"àn as evidence for early Islamic law.

the origin and development of the islamic patronate

75

legal issues and Islamic law “in the technical meaning of the term, did not yet exist.”22 If we were to accept Schacht’s premises, how could we then proceed? We could search for other material, such as Überreste (“remains”) but there is not much.23 There are also non-Islamic sources, but why should they be more reliable than Arabian-Islamic sources?24 It is true that from early manuscripts, authentic second-century texts can be distilled.25 But there is nothing which can fill in the gap of the first one and a half centuries (seventh-eighth century). That is to say, there would be nothing if we continued to disregard ˙adìth as a valid source. In my opinion, the exclusion of a˙àdìth is no longer necessary. If analyzed carefully, a˙àdìth can be used as a source for the development of Islamic jurisprudence ( fiqh) in the first century. The first steps in this direction were undertaken by the same Schacht who otherwise disqualified the a˙àdìth as spurious. He developed the theory of the “common link,” that is the scholar in whom all transmission strands (asànìd, sing. isnàd ) of one ˙adìth come together. According to Schacht, the common link was responsible for the lower, older part of the isnàd (which often consists of a single strand) and for the text (matn, pl. mutùn), which means that this common link was the one who brought about the circulation of the tradition.26 Gautier H.A. Juynboll refined Schacht’s isnàd analysis and introduced the terms “partial common link” (pupils of the common link who are common links themselves), “inverted partial common link,”

22 Schacht, An Introduction to Islamic Law (Oxford: The Clarendon Press, 1964), 19. Cf. also id., Origins, 5. For a summary of Schacht’s thesis, see Powers, Studies in Qur"an and Óadìth: The Formation of the Islamic Law of Inheritance (Berkely/Los Angeles/London: University of California Press, 1986), 1–2. 23 Cf. Motzki, “Quo vadis, ˙adìΔ-Forschung? Eine kritische Untersuchung von G.H.A. Juynboll: ‘Nàfi', the mawlà of Ibn 'Umar, and his position in Muslim ˙adìΔ literature’,” Der Islam 73 (1996), 40. 24 Non-Islamic sources are less affected by Islamic myth-making but they could well be affected by negative feelings of their authors towards Arabs and Islam. Often they were written to refute Islam. In addition, the authors were no insiders. For problems with Armenian, Syrian and Byzantine sources, see Kaegi, Byzantium and the early Islamic conquests (Cambridge: Cambridge University Press, 1992). 25 Recently, Brockopp, Law, 95–100, extracted authentic dicta of Màlik b. Anas from fourth-century manuscripts. The manuscripts contain copies of al-Mukhtaßar alkabìr fì-l-fiqh of the Egyptian scholar 'Abdallàh b. 'Abd al-Óakam (d. 214/829). 26 Schacht, Origins, 171ff.

76

ulrike mitter

“seeming common link,” “real common link,” “spider” and “dive.” Juynboll argued that the common link needed several partial common links to be regarded as the historic transmitter of a tradition.27 Gregor Schoeler28 and Harald Motzki29 combined the analysis of the mutùn with the analysis of the asànìd. The results of this “isnàd-cummatn analysis” are encouraging. It could be shown that certain a˙àdìth are much older than assumed until now. Isnàd-cum-matn analysis The first step in an isnàd-cum-matn analysis is the collection of all available versions of the ˙adìth which is to be analyzed.30 It is important to use the pre-canonical ˙adìth collections in addition to the canonical ones because the former contain many non-Prophetic a˙àdìth and isnàd versions which are not found in the cannonical collections. Secondly, the asànìd of all versions are presented in an isnàd bundle which reveals the common links amongst the different generations of scholars. The third step is the matn analysis. The mutùn of each common link (starting with the youngest ones) have to be compared word by word with each other and with the versions of other common links, and differences and common features are accurately noted down. By doing this, it will become obvious that there is a correlation between asànìd and mutùn which can hardly be the consequence of a later fabrication of a˙àdìth. Usually, the text of each authority shows its own characteristics which can consist of different fillers and synonyms, of different word order, and even of different motives. Had scholars attached new asànìd to an existing matn,31 this would nor27 Juynboll, “Nàfi', the mawlà of Ibn 'Umar, and his position in Muslim ˙adìth Literature,” Der Islam 70 (1993), 207–216; id., “Early Islamic society as reflected in its use of isnàds,” Le Muséon 107 (1994), 153–59. 28 Schoeler, Charakter und Authentie der muslimischen Überlieferung über das Leben Mohammeds (Berlin/New York: Walter de Gruyter, 1996), especially 60. 29 Motzki, “Quo vadis”; “The Prophet and the Cat. On dating Màlik’s Muwa††a" and Legal Traditions,” Jerusalem Studies in Arabic and Islam 22 (1998), 18–83. For the analysis and dating of a maghàzì report, see id., “The murder of Ibn Abì l-Óuqayq: On the Origin and Reliability of some Maghàzì-Reports,” in Harald Motzki, ed., The Biography of Mu˙ammad: The issue of the sources (Leiden: Brill, 1999). 30 If a ˙adìth is well-known, however, and appears in countless number in the canonical ˙adìth collections, it is not necessary to anaylze in detail all versions if they differ only slightly in the younger part of the isnàd and in the matn. 31 This is what Western scholars generally assume when talking about fabrica-

the origin and development of the islamic patronate

77

mally have led to nearly identical mutùn with different asànìd, something which can only be observed in a few cases.32 Since it is not reasonable to assume that the scholars would as well have invented a new matn for each new isnàd in order to conceal the true origin of a text, deliberate modification of the matn can be ruled out as an explanation for all cases of different texts. Also unintended changes which occured through the transmission process cannot serve as an explanation because the differences are simply too considerable. A better explanation for the pecularities of the texts is their origin from different transmitters. In the case of a single strand—that is an isnàd which is not supported by a common link but that goes back directly to the oldest authority (the Prophet, a ßa˙abì or a tàbi' )—this phenomenon might be explained by sheer coincidence although in my opinion, the possibility of coincidence must not be flogged to death. If the authorities in question are common links, the pattern assumes another dimension, anyway. It is obvious then, that, for example, the versions of all pupils of common link A have special characteristics which do not appear in the versions of the common links B and C, and the other way around. Even without looking at the isnàd, it is possible to determine whether a matn stems from A, B or C. This pattern is repeated in almost every analyzed tradition. It is highly improbable that it came into being by fabrication, because it is not conceivable that every falsifier would have followed the unwritten rule that all texts of an authority must have typical elements which do not turn up in the texts of other authorities.33 The pattern rather tion of a˙àdìth. See Schacht, Origins, 163 (“the isnàds were often put together very carelessly”); Motzki, “Quo vadis,” 193–94, on Juynboll’s interpretation of an isnàd bundle. 32 Nearly identical mutùn are normal from the second half of the second century onwards. The pupils of Màlik b. Anas (d. 179/795), for instance, have very similar texts. But in the first century, the versions usually differ considerably from each other because of still unfixed transmission methods and the lacking need for literal transmission. If very similar versions are ascribed to two early authorities (ßa˙àba or tàbi'ùn), we can conclude that one version is the copy of the other. In such a case, the attribution of the text to one of two transmitters would indeed be falsified or at least defective. Thus the isnàd-cum-matn analysis can be used as a method to detect spurious traditions. 33 To give an example: Fraudulent scholar A took a tradition from his collegue B. He does not want to mention B in the isnàd and ascribes the text directly to B’s authority X, pretending that he himself got the text from X. At the same time he modifies the text to make it look different from a B-text but also different from a C-text and a D-text, if there are more transmitters of the tradition of X than A

78

ulrike mitter

should be regarded as a structure which has been formed by the transmission of a text from teachers to pupils. Inasmuch as the results of the isnàd-cum-matn analysis are essential to the argumentation presented in this article, I considered it necessary to present a detailed example of this technique which is found in the appendix to this article. I chose the Barìra tradition because of its importance for the development of walà" and because this ˙adìth is particularly widespread and has many different versions. Although I can present only part of the analysis due to space limitations, this part does suffice to illustrate how the method works. As a result of the ˙adìth analysis, basically two groups of a˙àdìth can be distinguished. First, there are traditions which I call “anecdotes.” Such anecdotes report on legal decisions of early authorities by telling a whole story including dialogues and names of persons. They are ascribed almost exclusively to the ßa˙àba and Mu˙ammad. The second group consists of general statements on fictitious cases. They are ascribed mainly to the tàbi'ùn and later scholars. Next to these two groups, there are “maxims,” brief phrases containing only the essence of an anecdote or a juridical statement. These maxims are ascribed to the Prophet, ßa˙àba, tàbi'ùn and later lawyers, mostly using exactely the same words. Often the maxims turn out to be later summaries of the longer traditions.34 In “collective traditions,” several ßa˙àba are named together as authorities for one maxim. In what follows, I summarize the highlights of the results of the analysis. The origins of walà" reconsidered By applying the isnàd-cum-matn analysis to traditions on walà" al-'itq (walà" al-muwàlàt will be dealt with subsequently), it becomes obvious that many traditions expressing what later became the classical and B. Furthermore, he has to make sure that some elements of the traditions of B, C and D are still in his text to keep the characteristic look of the X-text. If the tradition is also transmitted by authorities Y and Z, the handiwork of A would have to be even more complicated, because Y-texts and Z-texts also have their pecularities. For a more practical setting, fill in the following names and compare with the analysis of the Barìra tradition (in the appendix): A = Ibn Shihàb al-Zuhrì; B = Hishàm b. 'Urwa; C = Abù l-Zubayr; X = 'Urwa b. al-Zubayr; Y = Amra bint 'Abd al-Ra˙màn; Z = al-Qàsim b. Mu˙ammad. 34 Schacht, however, explains it the other way round: (Anonymous) maxims are the oldest, most rudimentary mutùn which, in time, became enlarged and embellished with spurious details (Origins, 188–89).

the origin and development of the islamic patronate

79

doctrine can be found early in the first century, together with traditions expressing doctrines which were later rejected. This is true for all aspects of the institution like alienation of walà", the right of succession by the patron, unconditional manumission and the transition of walà" to the heirs of the patron, as will be seen in detail in the second part of this chapter. Since both doctrines existed at the same time, they obviously represented different solutions to a legal problem and should not be seen as successive stages in the development of fiqh.35 In some respects, early walà" seems to have had a quite “classical” Roman appearance which means that the resemblance to late-Roman patronate diminishes and with it the probability of an adoption of this institution by the Muslims. But what is more important is that the individual walà" which confered to the manumitter an agnatic title to his freedman’s estate and imposed the obligation of paying blood-money on him was a wellknown institution since the very first decades of Islam. There are some traditions to this effect which most probably go back to 'Umar b. al-Kha††àb (r. 13–23/634–644) and even one or two traditions that perhaps have their origin in a ruling by Mu˙ammad himself. Therefore, we can rule out the possibility that a non-Arab institution had been adopted in Umayyad times as suggested by Crone. Theoretically, the development of walà" al-'itq could have been influenced by a Roman institution at the time of the first caliphs, because even at this early time, Muslims could have been in contact with Byzantine institutions. It is, however, unlikely that the Arabs should have paid attention to the legal system of a people just subdued. In my view, the decisions of Muslim authorities in the first decades of Islam must have been based on pre-Islamic Arabian law rather than on Roman law. Still, the possibility of Roman influence on the development of walà" is not precluded but it is brought forward to pre-Islamic times as suggested already by Gotthelf Bergsträsser36 and von Kremer,37 though not with regard to walà" but in general. This means that the Arabian customary law of Mecca and Medina must have contained an institution similar to Islamic walà". When freeing slaves, the pre-Islamic Arabs seem to have attached freedmen 35

But see Crone, Roman law, 78ff. and 91. Bergsträsser, “Anfänge und Charakter des juristischen Denkens im Islam. Vorläufige Betrachtungen,” Der Islam, 14 (1925), 80. 37 Von Kremer, Culturgeschichte, I, 534. 36

80

ulrike mitter

individually to their manumitters through a walà" tie which granted the patron a title to his freedman’s estates.38 According to the ˙adìth analysis, walà" al-'itq developed continuously from the time of Mu˙ammad onwards. This result matches more or less the Muslim view of things but it was achieved by using a text-critical technique of analysis without presupposing or precluding foreign influence.39 Turning now to walà" al-muwàlàt, it can be stated that around the year 100/718–9, different kinds of individual and semi-collective contractual walà"40 were widely accepted: walà" with and without conversion, walà" over Arab and non-Arab strangers41 and walà" over foundlings. What are the roots of this institution? This question is not easy to answer because of the general scarcity of traditions on walà" al-muwàlàt. It is difficult just to work out the different doctrines, let alone to ascribe a certain doctrine to a certain scholar and to work out a chronology. But with due caution the conclusion can be drawn that the traditions on wàlà" over Arab and non-Arab strangers without conversion go back to 'Umar b. al-Kha††àb and other very early Muslims. The mere mention of Arab mawàlì shows that the traditions are very old because it did not take long to restrict the term mawàlì to non-Arabs. As walà" al-'itq, this kind of contractual walà" is therefore more likely to go back to pre-Islamic Arabian law than to have been inspired in Islamic times by late-Roman law. This conclusion supports the view about the origin of Islamic walà" al-muwàlàt in pre-Islamic ˙ilf, as suggested by Juda, Hallaq and others. Yet, perhaps, it is not in ˙ilf where we should look for the origins 38 For individual walà" and ˙ilf in pre-Islamic Arabia, see also Smith, Kinship and Mariage in Early Arabia (Cambridge, 1885; London: New edition of Stanley A. Cook, 1907; reprint London, 1990), 54; Reinert, Recht, 21; Hallaq, “Use and Abuse,” 15–16; Juda, Aspekte, 11. For individual walà" in pre-Umayyad time (thus before the adoption of the late-Roman patronate by Umayyad caliphs as suggested by Crone), see Motzki’s review of Crone’s Roman law, 343–44. 39 It is not necessary to presuppose foreign influence on the development of fiqh. Muslims did not start from scratch as they developed their legal system but they could fall back in part on old-Arabian law. For old-Arabian regulations as expressed in poetry, see Reinert, Recht. On the poet Zuhayr b. Abì Sulmà (d. 609 CE) Reinert says that he actually deserves to be called a “lawyer” (ibid., 1). 40 In the a˙àdìth it is always one single stranger, never a group, who attaches himself mostly to “people” (qawm) amd sometimes to individuals. For the latter, see e.g., 'Abd al-Razzàq, Mußannaf, IX, no. 16179. 41 The distinction between an Arab stranger (là yu'raf lahu aßl ) and a non-Arab stranger (sàqi†) is mainly made by the Meccan 'A†à" b. Abì Rabà˙ (d. 114/732) and his pupil Ibn Jurayj (d. 150/767). See 'Abd al-Razzàq, Mußannaf, IX, nos. 16175 and 16168.

the origin and development of the islamic patronate

81

of walà" al-muwàlàt but in pre-Islamic contractual walà". Although later Muslim scholars tend to identify walà" al-muwàlàt with ˙ilf,42 and although there surely are many resemblances between ˙ilf and walà" in terminology and legal implications,43 the correspondence between ˙ilf and walà" was not total.44 The terms walà" and ˙ilf had been differentiated (and confused) already in pre-Islamic times.45 Perhaps it is correct to state that ˙ulafà" had a better position in the tribe than mawàlì, although part of the pejorative meaning of walà" originated only later, when the term mawàlì was restricted to freedmen and non-Arabs. In ˙ilf, both parties are said to have been equal, for example regarding the right of succession which was reciprocal.46 In walà", they may have been not.47 At any rate, in the early a˙àdìth the relationship between contractual patron and client is not described as equal. There is, as far as I could ascertain, not one ˙adìth claiming

42

For example al-Sarakhsì, Mabsù†, VIII, 81. There is a tradition which says that Islamic walà" took the place of pre-Islamic ˙ilf (al-walà" bi-manzilat al-˙ilf ) This tradition, which is combined with the prohibition of sale of walà", cannot be dated back further than to the Kùfan lawyer Sufyàn b. 'Uyayna (d. 198/811). See Mitter, Das frühislamische Patronat. Eine Untersuchung zur Rolle von fremden Elementen bei der Entwicklung des islamischen Rechts (Ph.D. diss., Nijmegen, 1999), 269. For this tradition cf. also Hallaq, “Use and Abuse,” 14; Juda, Aspekte, 158. 43 See Ella Landau-Tasseron’s article in this book. 44 For the differences between ˙ilf (hosting alliance) and walà" (clientage), see ibid., According to Reinert, Recht, 22, the purpose of a walà" contract was the creation of a fictitious kinship tie. The partners actually lived together and shared the pasture. The point of a ˙ilf contract was political alliance while maintaining the individuality of the partners. For the differences between ˙ilf and walà", see also Juda, Aspekte, 20–21. 45 According to Landau-Tasseron, in her contribution to this book, the problems started with the term mawlà bi-l-˙ilf/yamìn/'aqd which signifies ˙alìf (ally) although the word mawlà (client) is used. Reinert specifies that mawlà l-yamìn was used in Medina and ˙alìf in Mecca (Recht, 21). As just stated in the previous footnote, Reinert distinguishes between contracts of walà" (ibid., 20–22) and ˙ilf (ibid., 22) but he deals with ˙ulafà" in the section on walà". 46 Landau-Tasseron, above; Hallaq, “Use and Abuse,” 13–14; Tyan, “Óilf,” EI 2, III, 388; Crone, Roman law, 53; Juda, Aspekte, 4; Reinert, Recht, 22. 47 Falaturi, “Der Koran: Zeugnis der Geschichte seiner Zeit,” in Albrecht Noth and Jürgen Paul, eds., Der islamische Orient. Grundzüge seiner Geschichte (Würzburg: Ergon, 1998), 54, states that “Anders als im Falle des ˙ilf-Verhältnisses handelt es sich aber [bei walà"] nicht um ein Bündnis zweier gleichgestellter Partner. Der Pakt geht mehr von dem Schutzbedürftigen aus.” Nevertheless, he cites a formula to describe the relationship of walà" which includes reciprocal succession. One may wonder, then, what Falaturi exactly means by saying that the parties were not on the same social level. As to walà" al-jiwàr, Juda mentions “die einseitige Leistung eines Schutzgewährenden” and the fact that mawàlì were second-class members (Aspekte, 21). But then, also socially weak ˙ulafà" had existed (ibid., 13ff.).

82

ulrike mitter

the right of the client to inherit from his patron. Equality is expressed sometimes in traditions on payment of blood-money, but scarcely in traditions on succession or mutual help.48 Only in later fiqh books do we encounter the idea of equality between patron (al-mawlà la'là ) and client (al-mawlà l-asfal ) but this could well be related to changed social conditions and/or changed opinion of later fuqahà". Inasmuch as the parties in pre-Islamic ˙ilf were regarded as equals and in Islamic walà" al-muwàlàt they were not, the origin of the latter in ˙ilf must at least be reassessed. What has just been said on the origins of walà" al-muwàlàt does not apply to the origins of walà" with conversion (walà" al-islàm). Needless to say, this kind of contractual walà" does not go back to pre-Islamic times. But it is even doubtful whether it stems from Mu˙ammad as a very famous Prophetic ˙adìth claims. The ˙adìth, which is classified as “∂a'ìf ” by various Muslim scholars,49 cannot be traced back further than to the Umayyad prince 'Abd al-'Azìz b. 'Umar (d. after 147/764–5). Existing a˙àdìth of the ßa˙àba turned out to be not reliable. Besides, on walà" al-islàm there are no anecdotes, the existence of which could be an indication for the old age of the institution.50 (There are anecdotes on walà" without conversion). Another indication for the late origin of walà" al-islàm is the specification of the converts as non-Arabs. This specification must be a late one because at the outset, converts had been Arabs, too. Judging from the existence of traditions from the tàbi'ùn, walà" alislàm was well-known at the end of the first century/eighth century. From the fact that there are almost no traditions of Meccan and Medinese tàbi'ùn but many of Syrian and Iraqi scholars, we can infer

48 There is a statement by Ibn Jurayj (d. 150/767) that “they [the patrons] pay blood-money for him [the stranger] and he for them, and they help him,” with no mention of mutual help. 'Abd al-Razzàq, Mußannaf, IX, no. 16168. 49 Cf. Juda, Aspekte, 86. 50 This statement—which contradicts the theory of Schacht and others who think that anecdotes serve the purpose to create a spurious sphere of authenticity—should not be misunderstood. I do not conclude from the mere existence of an anecdote that a regulation is old but the ˙adìth analysis which I carried out shows time and again that anecdotes of the ßa˙àba existed at a very early time. Therefore, the assumption that anecdotes are an indication for the old age of an institution is not a premise of the analysis but its result. This allows the conclusion that the total lack of anecdotes is an indication for the late origin of a regulation.

the origin and development of the islamic patronate

83

that walà" al-islàm could have been developed in Syria51 and Iraq at some point in the first century (the a˙àdìth do not allow more precise conclusions), based on the model of walà" al-muwàlàt without conversion.52 The fact that there are barely Syrian asànìd on walà" al-'itq is, by the way, an additional argument against the origin of walà" al-'itq in Syria as Crone suggests. I will now move from the question about the origins of walà" and possible non-Arab influence on this institution and deal with the regulations of early walà" as recorded in the a˙àdìth. Regulations of walà" according to early ˙adìth Aspects of walà" treated by early lawyers can be divided mainly into two groups. First, the rights and duties of patron and client and, second, the possibilites of passing on the patronate to a third party. Inasmuch as not all possible legal incidents of walà" are mentioned in the a˙àdìth, some aspects will be missing in the following presentation, as the right of the patron to act as marriage guardian for a freedman’s daughter and to have custody of the freedman’s children. I will try to reconstruct the early doctrines, that is to say the state of affairs in the first and second century of Islam (seventh and eighth centuries). The following comments are based on the isnàd-cum-matn analysis of a great number of a˙àdìth. Again for reasons of space it is impossible to provide the details of even one example in each section, so in most cases I stuck to merely presenting the results of the analysis. The references, too, are not given exhaustively but selectively. On the whole, walà" al-muwàlàt is treated much more detailed than walà" al-'itq, because earlier I have studied the latter in great detail.53 The sources I used are the pre-canonical ˙adìth collections in the 51 The large number of Syrian asànìd are an indication of the preoccupation of Syrian scholars with walà" al-islàm. The attachment of the Syrians to the question of walà" on conversion is also mentioned by Crone, Roman law, 155 note 9, who does not, however, distinguish walà" al-islàm from other kinds of walà al-muwàlàt. 52 The regulations mentioned in the a˙àdìth are virtually identical for both kinds of contractual walà". 53 On all aspects of walà" al-'itq, see Mitter, Patronat. On tasyìb, see id., “Unconditional manumission of slaves in early Islamic law: a ˙adìth analysis,” Der Islam 78/I (2001), 35–72.

84

ulrike mitter

first place54 but also the canonical collections55 and later works.56 Sometimes it is difficult to say if a tradition deals with servile or contractual walà". If a manumitter or a slave is mentioned, clearly concerned with walà" al-'itq. Rules on a stranger or convert apply only to walà" al-muwàlàt both with and without conversion. But in some cases, both kinds could be at issue, for example, if we read that a patron inherits as the last heir of the client or that sale of walà" is forbidden. Although the traditions do not always make a clear distinction between walà" al-'itq and walà" al-muwàlàt, the two are entirely different institutions with a different history. In what follows, I will analyze the two forms of walà" separately, always starting with walà" al-'itq. Furthermore, walà" al-muwàlàt without conversion will be analyzed separately from walà" al-islàm in order to enhance the understanding of the history and development of contractual walà". Rights and duties of patron and client In early ˙adìth, the rights and duties of the patron are not mentioned in a systematic and comprehensive way57 and the rights and duties of the client are rarely mentioned at all. The latter were propably well-known to everybody. It could have been loyality— another translation of walà", by the way—which was expected of clients in the first place. a) Right of the patron to the patronate Walà" over the freedman—whether Arab or non-Arab is not stipulated58—belongs to the manumitter. This rule is encountered in the Barìra tradition, an anecdote which circulated in the middle of the

54 The two Mußannaf of both 'Abd al-Razzàq al-Ían'ànì (d. 211/827) and Ibn Abì Shayba (d. 235/849), the two Sunan of Sa'ìd b. Manßùr (d. 227/842) and alDàrimì (d. 255/869), Muwa††a" of Màlik b. Anas (d. 179/796), Aßl of al-Shaybànì (d. 189/805), Mudawwana of Sa˙nùn (d. 240/854), Umm of al-Shàfi'ì (d. 204/820) and Musnad of Ibn Óanbal (d. 241/855). 55 Above all the two Ía˙ì˙ of al-Bukhàrì (d. 256/870) and Muslim (d. 261/875). 56 Sunan of al-Bayhaqì (d. 458/1066). In exploring the legal background, I used Mabsù† by al-Sarakhsì (d. 483/1090), Bidàya of Ibn Rushd (d. 595/1198) and Mughnì of Ibn Qudàma (d. 620/1223). 57 The missing points were made up for in later fiqh works. 58 Since during Mu˙ammad’s life Arabs were still enslaved (cf. Juda, Aspekte, 64), the very early decisions on freedmen could be meant to have applied to Arabs, too.

the origin and development of the islamic patronate

85

first century (end seventh century) in many different versions, both in the Óijàz and in Iraq (see the detailed analysis in the appendix). The common elements of all versions go back to 'À"isha (d. 58/678) who is the oldest common link of the isnàd bundle. To be sure, the exact wording of 'À"isha’s tradition cannot be reconstructed because the versions of her transmitters are too different but it is possible to reconstruct the rough outlines of her report. In time, Muslim lawyers extracted dozens of regulations from this tradition. With the Barìra tradition they buttressed the prohibition of tasyìb59 (i.e., unconditional manumission, by saying that the manumitter cannot renounce walà" because it comes automatically into being) and the prohibition of walà" al-muwalàt (arguing that if walà" exclusively belongs to the manumitter, there could be no contractual walà").60 Problems with the Barìra tradition started when selling a mukàtab (a slave who buys himself free) became forbidden because Barìra was known to have been a mukàtaba—a fact that did not bother either 'À"isha nor Mu˙ammad because in their time, the selling of a mawlà and a mukàtab was still allowed.61 While in servile walà", the right of the patron to the patronate is based on the favor [of manumission] (ni'ma) which is granted to the freedman, things are much more complicated in walà" al-muwàlàt. The right of the contractual patron is based either on a contract or on the conversion of a client or on paying blood-money or on the care for a foundling or on a combination of the factors. Whereas the right of the manumitter to the patronate was never doubted, the right of the contractual patron was called into doubt by some lawyers and this last case was modified in time and, finally, (partially) abolished. Generally speaking, the contractual patron is supposed to be an Arab Muslim, usually referred to as a “rajul ” or a “rajul muslim.” Only in later sources do we encounter the opinion that the patron can be a dhimmì.62 The potential clients are non-Arab converts “at

59

For tasyìb, see below. See for instance al-Bayhaqì, Al-Sunan al-kubrà, 10 vols. (Hyderabad, 1344–55 AH), X, 298. Cf. also below the traditions of al-Óasan and al-Sha'bì. 61 On the sale of walà", see below. 62 According to al-Saraskhsì, a dhimmì can take as patron a Muslim or a dhimmì. The contract will even be valid if the client converted to Islam (Mabsù†, VIII, 96, l. 12). Cf. also Hallaq, “Use and Abuse,” 8. In addition, al-Sarakhsì discusses the possibility of a minor being patron (with and without his father’s permission), of a woman, a slave and a mukàtab (Mabsù†, VIII, 95–96). 60

86

ulrike mitter

the hands” of a person (aslama 'alà yadayhi ),63 Arab strangers (là yu'lam lahu aßl ),64 non-Arab strangers (sàqi†, ibn al-ar∂, rajul min ahl al-ar∂ )65 who came to live with a tribe—both groups of strangers are probably supposed to be Muslims but without conversion being mentioned66—and foundlings (laqì† or manbùdh) whose ethnicity is not specified.67 Strictly speaking, walà" over foundlings does not belong to walà" al-muwàlàt but forms a separate category.68 In the following, I will always treat it subsequent to the rules of contractual walà". Traditions on walà" al-muwàlàt without conversion over Arab and non-Arab strangers are recorded almost exclusively in the pre-canonical collections, which makes clear how important those collections are for the reconstruction of the development of a doctrine. The a˙àdìth state that the person or the tribe who pays blood-money for a newcomer or who is accepted by him as a patron is indeed his patron. If a person dies without relatives ('aßaba) and without a patron, his inheritance is generally held to fall to the Muslims at large. A respective tradition of 'Umar b. al-Kha††àb who allegedly wrote a letter on the treatment of an Arab stranger is transmitted by the 63 'Abd al-Razzàq, Mußannaf, IX, nos. 16271–75. The same traditions are included in vol. VI, nos. 9872–9876. Al-Sarakhsì, Mabsù†, VIII, 96, ll. 14ff., declares that walà" al-islàm can only be concluded with a non-Arab convert. An Arab convert remains attached to his tribe. He cannot become a contractual mawlà, no more than a freedman can, because both are supposed to have a familiy which, in the case of the freedman, is his patron. In the very beginning of Islam, Arab converts were regarded as mawàlì of Mu˙ammad ( Juda, Aspekte, 57–58) which means that, in a sense, there was some form of walà" al-islàm even then. But surely, this was rather meant in a general way, without having the usual legal implications of walà". (For “mawàlì of God and the Prophet,” see also Landau-Tasseron’s article above.) In the earliest a˙àdìth on walà" al-islàm, which go back to the turn of the first century, converts are frequently specified as non-Arabs. 64 'Abd al-Razzàq, Mußannaf, IX, nos. 16174–81. On the problem of Arab mawàlì there are only few traditions, almost all of them recorded in 'Abd al-Razzàq and ascribed to 'Umar b. al-Kha††àb or 'A†à" b. Abì Rabà˙. The latter and his pupil Ibn Jurayj equate the situation of an Arab stranger with the one of a non-Arab stranger (ibid., nos. 16174 and 16175). 'Abd al-Razzàq (or his pupil Ma'mar) does also specify the stranger in a tradition of Ibn Mas'ùd as an unknown Arab, whereas Sa'ìd b. Manßùr, who has several versions of this tradition, does not say anything about the ethnic origin of the stranger in question. 65 Ibid., nos. 16168–73. 66 Normally, no mention is made of the religion of a stranger but in a few traditions it is explicitely stated that he is a Muslim. For example ibid., no. 16170. 67 Ibid., nos. 16182–90. 68 In Ibn Rushd’s Bidàya, the foundling is mentioned in the book on found property (luq†a) but mostly he is mentioned in the books on farà"i∂ or walà" together with other forms of walà".

the origin and development of the islamic patronate

87

four tàbi'ùn Ibn Shihàb al-Zuhrì (Medina, d. 124/741), 'A†à" b. Abì Rabà˙ (Mecca, d. 114/732), 'Amr b. Shu'ayb (Mecca, d. 118/736) and Abù Qilàba (al-Baßra, d. 104/722–3)—the authority of 'Amr b. Shu'ayb and Abù Qilàba is 'Amr b. al-'Àß (d. 43/663)—and by an anonymous scholar.69 If there are four or five tàbi'ùn transmitting different versions of an 'Umar tradition, the source for the elements which all traditions have in common could be a ruling of 'Umar.70 The 'Umar tradition has one element which does not appear in other traditions on walà" al-muwàlàt: the listing of the stranger in the dìwàn of the tribe.71 At the turn of the first century/early eighth century, several traditions of the ßa˙àbì Abdallàh b. Mas'ùd (d. ca. 32/653) were circulating in Mecca and in Iraq, all expressing Ibn Mas'ùd’s approval of walà" al-muwàlàt in an indirect way by having him solve the 69 'Abd al-Razzàq Mußannaf, IX, nos. 16174 ('A†à"), 16177 (Abù Qilàba—'Amr b. al-'Àß), 16178 ('Amr b. Shu'ayb—'Amr b. al-'Àß), 16172 and 16181 (al-Zuhrì) and 16176 (al-Zuhrì and the anonymous scholar with one matn). There are two other versions of the 'Umar tradition, one in Ibn Abì Shayba (Mußannaf, XI, no. 11624), the other one in Sa'ìd b. Manßùr (Al-Sunan, ed. Óabìb al-Ra˙màn al-A'Ωamì, 3 vols./1,2 [in 1] [Bombay, 1403/1982], III/1, no. 209) but both rank the client among converts. Ibn Abì Shayba (al-Zuhrì) does it implicitly by the title of the chapter, Sa'ìd b. Manßur ('Amr b. Shu'ayb—'Amr b. al-'Àß) mentions a convert explicitly in the matn. Nevertheless, the text of Sa'ìd has many elements in common with the other matn of 'Amr b. al-Shu'ayb and undoubtedly is a version of our 'Umar tradition on contractual walà" without conversion. Obviously, the element of conversion was introduced later into the matn. (Of course, the stranger could have been a convert but the original text of the 'Umar tradition does not mention conversion). The texts of 'Amr b. Shu'ayb are similar to each other and different from Abù Qilàba’s version. Again, there is the transmission pattern which suggests that the common authority of both transmission strands ('Amr b. al-'Àß) is the historic transmitter of the tradition. Ibn Abì Shayba’s Zuhrì text is similar to 'Abd al-Razzàq’s Zuhrì text. Since conversion is not mentioned in the matn of Ibn Abì Shayba’s text, its arrangement in the chapter on walà" al-islàm could be a mistake. 70 'A†à", al-Zuhrì and the anonymous scholar do not mention an authority for their text. Therefore, we cannot entirely exclude the possibility that they transmitted circulating versions of the tradition of the older 'Amr b. al-'Àß. A direct dependence on his text seems, however, very unlikely because of the different aspect of the texts. Either way, if the tradition stems from 'Amr b. al-'Àß, it existed in the first half of the first century (late seventh century). Intentional fabrication of the tradition is unlikely because at that early time, it was unnecessary to support one’s personal authority by citing an older authority. If 'Amr b. al-'Àß, a combatant for Islam who was held in high esteem by 'Umar, reports on one of 'Umar’s legal decisions, his report is likely to have been based on a historic event. 71 The dìwàn is explicitly mentioned by Abù Qilàba and in the matn of al-Zuhrì and the anonymous scholar. In the version of 'Amr b. Shu'ayb on the converted stranger, the affiliation of the stranger to the dìwàn is expressed differently: “yu'àddu al-qawm” (he is counted with the people). On the 'adìd, see Landau-Tasseron, above.

88

ulrike mitter

problem of a stranger who died without relatives and patron. According to Ibn Mas'ùd, a non-Arab stranger is free to choose the patron he wants ('A†à"—Ibn Mas'ùd).72 If the stranger who has permanent residence dies without relatives and without a patron, his estate falls to the treasury (Ziyàd b. al-Jarrà˙—Ibn Mas'ùd and Mu˙ammad b. alMuntashir—Masrùq—Ibn Mas'ùd).73 Another tradition has him say, however, that if a stranger, who perhaps is supposed to be an Arab,74 does not have relatives [and no patron], he is free to dispose of his entire estate by will (common links are Ibràhìm al-Nakha'ì and Abù Is˙àq).75 I do not think that Ibn Mas'ùd treated Arab and non-Arab strangers differently (he would be the only one to do so). More probably, here we are apparently dealing with two different legal problems which can be harmonized as follows: a stranger (Arab or not) is allowed to dispose of his estate but if he died without having done this, the inheritance then falls to the state. The rough outlines of this doctrine are likely to go back to Ibn Mas'ùd. Scholars still accepted walà" al-muwàlàt without conversion around the turn of the first century.76 Traditions rejecting this institution are scarce. In fact, I only know of one such tradition. It is ascribed to 'Alì b. Abì ˇàlib (d. 40/661) and its common link is Rabì' b. Abì Íàli˙77 who quite surely spread the tradition.78 His authority is “a

'Abd al-Razzàq, Mußannaf, IX, no. 16168. Ibid., no. 16169 and Ibn Abì Shayba, al-Mußannaf, ed. Mu˙ammad A˙mad al-Nadwì, 15 vols. (Bombay, 1973–83), XI, no. 11626 (Masrùq, d. 63/682–683); 'Abd al-Razzàq, Mußannaf, IX, no. 16170 (Ziyàd b. al-Jarrà˙, d. c. 100/718–9). The text of the latter was perhaps not been transmitted independently from Ibn Mas'ùd but may also stem from the older Masrùq. On the problematic identity of Ziyàd b. al-Jarrà˙, see Ibn Óajar, Tahdhìb at-tahdhìb al-kamàl fì ma'rifat ar-rijàl, 12 vols. (Beirut: Dàr Íàdir, 1968; repr. Hyderabad, 1325/1907), III, no. 701. 74 In some versions, Ibn Mas'ùd speaks in general on a person who has no 'aßaba and no walà". In 'Abd al-Razzàq (for reference see next footnote), he speaks explicitly about an Arab whose origin is unknown (là yu'lam anna aßlahu min al-'àrab walà yudrì mimman huwa). 75 In this tradition, Ibn Mas'ùd addresses the people of the Yemen (Ibràhìm, in Sa'ìd b. Manßùr, Sunan, III/1, no. 217 and 218) or Hamadan (Abù Is˙àq, in ibid., nos. 215 and 216 and 'Abd al-Razzàq, Mußannaf, no. 16180). 76 For example 'A†à" (Ibn Abì Shayba, Mußannaf, XI, no. 11617; 'Abd al-Razzàq, Mußannaf, IX, nos. 16152, 16168 and 16175) and al-Zuhrì (ibid., nos. 16172 and 16173). 77 On Rabì' b. Abì Íàli˙, see al-Shaybànì, Kitàb al-aßl al-ma'rùf bi-l-mabsù†, 5 vols. (Karachi: Idàrat al-Qur"àn wa-l-'Ulùm al-Islàmìya, n.d.), IV, 186 note 1. Judging from his teachers and pupils, Rabì' must have died around the middle of the second century/later half of the eighth century. 78 Three different versions are ascribed to him, all transmitted by the Kùfan 72 73

the origin and development of the islamic patronate

89

man” who is identified (perhaps emended) in al-Shaybànì as Ziyàd. According to this scholar, 'Alì did not want a non-Arab stranger (rajul min ahl al-ar∂ or rajul min ahl al-sawàd ) to take him as his patron. Therefore, the stranger went to Ibn 'Abbàs (d. 67–8/868–8) who accepted him. It is not at all likely that 'Alì did reject walà" almuwàlàt because in his time this kind of walà" was commonly accepted and widely practised. The tradition was perhaps fabricated by Rabì'. But there is also another possibility. The text could have been misunderstood and it was not 'Alì’s intention to reject walà" al-muwàlàt altogether but he simply did not want this particular man to be his mawlà. Walà" al-islàm was controversially discussed around the turn of the first century. Traditions supporting walà" al-islàm are ascribed to the Kùfan Ibràhìm an-Nakha'ì (d. 96/717)—his tradition perhaps being the oldest tradition on walà" al-islàm79—and to other lawyers of Kùfa and Baßra who died at the beginning of the second century/early eighth century.80 Generally speaking, they say that walà" (or the right

scholars Sufyàn al-Thawrì ('Abd al-Razzàq, Mußannaf, IX, nos. 16157 and 16171), Wakì' (Ibn Abì Shayba, Mußannaf, XI, no. 11628) and Abù Yùsuf (al-Shaybànì, Aßl, IV, 186). Although these scholars are no common links, the differences of their texts lead to the conclusion that their versions were transmitted independently from each other. 79 The tradition of Ibràhìm could be authentic since there are three transmission strands: Manßùr—Ibràhìm ('Abd al-Razzàq, Mußannaf, IX, nos. 16 272 and 16273), Abù Màlik—Ibràhìm (Sa'ìd b. Manßùr, Sunan, III/1, no. 205) and Óammàd— Ibràhìm (al-Shaybànì, Aßl, IV, 182) and one of the transmitters, Manßùr, is a partial common link. Traditions of earlier authorities are problematic. The Prophetic ˙adìth on walà" al-islàm seems to be late (more on this below). Some traditions of the ßa˙àba are single strands which cannot be dated (for example, the one of Zayd b. Thàbit, in Ibn Abì Shayba, Mußannaf, XI, no. 11630). One of the few mutùn on walà" al-islàm, ascribed to 'Umar b. al-Kha††àb, originally dealt with walà" without conversion, the asànìd of two other versions are not trustworthy because of weak transmitters (Layth b. Abì Sulaym and Ash'ath b. Siwàr, in al-Shaybànì, Aßl, IV, 185) or an interrupted transmission (Layth b. Sa'd from Abù l-Ash'ath, in Ibn Abì Shayba, Sunan, III/1, no. 11627). 80 There are the single strands of Óammàd b. Abì Sulaymàn (Sa'ìd b. Manßùr, Sunan, III/1, no. 214) and 'Abd al-Karìm b. Abì l-Mukhàriq ('Abd al-Razzàq, Mußannaf, IX, no. 9876) and the (however, not very strong) common links al-Óasan al-Baßrì and 'Àmr al-Sha'bì (al-Dàrimì, Sunan, ed. Fawàz A˙mad Zamalì and Khàlid al-Saba' al-'Alamì, 2 vols. (Beirut, 1407/1989), II, no. 3032). The tradition of 'Umar b. 'Abd al-'Azìz, transmitted by his son, might not stem from 'Umar but from 'Abd al-'-Azìz. My dating of all these traditions could be disputed but since the tradition of the older Ibràhìm seems to be authentic, and since there had to be some scholars at that time endorsing walà" al-islàm, it could well be that the scholars named are the authorities in these asànìd.

90

ulrike mitter

to inherit and the duty to pay blood-money) belongs to the person “at whose hands” ('alà yadayhi ) another man converted. Traditions rejecting walà" al-islàm circulated more or less at the same time. If the traditions ascribed to the Iraqi scholars 'Àmir al-Sha'bì (Kùfa, d. 103/722) and al-Óasan al-Baßrì (d. 110/728) were not authentic, they at least stemmed from the common links Mu†arrif b. ˇarìf (Kùfa, d. between 133–143/ 750–761) and Yùnus b. 'Ubayd (Baßra, d. 138 or 139/755–7).81 One generation later, the Umayyad caliph 'Abd al-'Azìz b. 'Umar was very much interested in supporting walà" al-islàm. He not only transmitted the famous tradition of the Prophet,82 but he also transmitted a tradition from his father 'Umar b. 'Abd al-'Azìz to the effect that a contractual patron even inherits as an agnate.83 The rule that conversion may be realized “at the hands” of one person while walà" may be given to another is encountered only in later sources.84 This rule is based on the assumption that conversion does not by itself require a walà" tie but that a contract for walà" is 81 For Mu†arrif—al-Sha'bì, see e.g., 'Abd al-Razzàq, Mußannaf, IX, no. 16274. For Yùnus—al-Óasan, see e.g., Ibn Abì Shayba, Mußannaf, XI, no. 11634. As mentioned above, al-Sha'bì and al-Óasan are also said to have supported walà" al-islàm. Strangely enough, there is one and the same isnàd (Sa'ìd b. Manßùr—Hushaym— Yùnus—al-Óasan) supporting both views (see Sa'ìd b. Manßùr, Sunan, III/1, no. 199; and ibid., no. 296). Inasmuch as it would be pointless that a scholar faked opposite opinions for his authority, there must be a different explanation. Maybe alÓasan changed his opinion or perhaps he had been misunderstood. 82 'Abd al-'Azìz b. 'Umar is a strong common link for the ˙adìth of Mu˙ammad. He transmits the tradition from the qà∂ì 'Abdallàh b. Mawhab or Wahb from Tamìm al-Dàrì from Mu˙ammad (see for example in 'Abd al-Razzàq, Mußannaf, IX, no. 9872; Ibn Màja, Sunan, ed. Mu˙ammad Fu"àd 'Abd al-Bàqì, 2 vols. (Beirut: Dàr al-Kutub al-'Ilmiyya, n.d.; repr. 1373/1954), II, no. 2752; al-Tirmidhì, Sunan (al-Jàmi' al-Ía˙ì˙), ed. A˙mad Mu˙ammad Shàkir, Mu˙ammad Fu"àd 'Abdalbàqì and Ibràhìm 'Atwa 'Iwa∂, 5 vols. (n.p., 1382–98/1962–87), IV, no. 2112). One pupil of 'Abd al-'Azìz (Ya˙yà b. Óamza) introduces one or two other authorities between 'Abdallàh b. Mawhab and Tamìm (al-Bayhaqì, Sunan, X, 296–97). There are also asànìd which lead directly to 'Abdallàh b. Mawhab without naming 'Abd al-'Azìz (Sa'ìd, Sunan, III/1, no. 202 and al-Bayhaqì, Sunan, X, 298) but since the mutùn of these versions are the same as the one of 'Abd al-'Azìz, different authorities for these texts are not plausible. There are also two entirely different transmission strands but they, too, are not likely to be historic because of the similarity of their mutùn to the matn of 'Abd al-'Azìz and to the mutun of other authorities for walà" al-islàm. 83 Ibn Numayr—'Abd al-'Azìz (Ibn Abì Shayba, Mußannaf, XI, no. 11625) and Ismà'ìl b. 'Ayyàsh—'Abd al-'Azìz (Sa'ìd b. Manßùr, Sunan, III/1, no. 210). More on this tradition below. 84 Al-Sarakhsì, Mabsù†, VIII, 92, l. 12–13; al-Marghìnànì, Al-Hidàya, 4 vols. (Cairo, 1980), III, 274.

the origin and development of the islamic patronate

91

also necessary. The convert has to accept a Muslim as his patron, otherwise there will be no walà" between them. It seems, though, that the additional element “take as patron” (wàlà) is older than the above-mentioned rule because it already appears in some early traditions like the one of Ibràhìm.85 In the beginning, the element may have been mentioned only in passing86 but later the convert’s explicit consent to walà" was absolutely required and this element became an indispensable condition for walà" al-islàm. The question is, however, if in the beginning, when the convert’s explicit consent did not necessarily constitute part of the conclusion of walà" al-islàm, the lawyers thought that conversion automatically led to walà". I do not think so. First, because the phrase actually states that the convert took a Muslim as his patron, however incidental this statement may have been. Secondly, there is another element which indicates the lack of automatism of walà" namely the statement that the walà" tie can be dissolved “as long as no bloodmoney has been paid.” This element, which is an integral part of the later doctrine,87 is part of Ibràhìm’s tradition as well. The Kùfan scholar Óammàd (d. 120/738) who transmits the blood-money element from Ibràhìm, is said to have promoted this rule himself.88 A contemporary of Óammàd, 'Abd al-Karìm b. Abì l-Mukhàriq (Baßra, d. 126 or 127/743–745), is said to have rejected the idea that a convert might take another person as his patron.89 This points to an ikhtilàf (difference of opinion regarding a legal matter) of the scholars on this question at the beginning of the second century/early eighth century. The third reason for my hypothesis that walà" al-islàm did not arise automatically as a result of conversion is the fact that Muslims

85 In several texts of Manßùr’s version of Ibràhìm’s tradition (e.g., 'Abd al-Razzàq, Mußannaf, IX, no. 16272), in Abù Màlik’s version (Sa'ìd b. Manßùr, Sunan, III/1, no. 205) and in Óammàd’s version (al-Shaybànì, Aßl, IV, 182). 86 The wàlà element crops up even in the tradition of 'Abd al-Karìm who was strictly against the possibilty that the convert might change patrons. 87 Al-Marghìnànì, Hidàya, III, 274; al-Sarakhsì, Mabsù†, VIII, 91, l. 12. 88 Óammàd said: “He [the convert] is allowed to change [the patron], if he wants to, if [the patron] has not paid yet blood-money for him. If he has paid, he is not allowed to change.” Sa'ìd b. Manßùr, Sunan, III/1, no. 214. The dating of the tradition is difficult because I only have this single text. Inasmuch as the older Ibràhìm already has the blood-money element, it can be assumed that Óammàd also mentioned it. 89 'Abd al-Razzàq, Mußannaf, IX, no. 9876.

92

ulrike mitter

knew the concept of walà" al-muwàlàt without conversion first and that this kind of walà" always occurred on a voluntary basis. It is clear that some rules of the old type of walà" without conversion reappear in the new walà" al-islàm: Since in traditional walà" almuwàlàt the mere paying of blood-money could lead to a walà" tie, walà" al-islàm can no longer be dissolved once blood-money had been paid. Since the consent of a free person created a walà" tie, the consent of the convert is simply required.90 This voluntary form of walà" al-islàm can already be seen in the oldest a˙àdìth though not yet in a very strict and regulated way. During the first half of the second/eighth century, things changed. There were some scholars who disliked the idea of voluntary walà" al-islàm. They rejected the changing of patrons and stuck to the idea of walà" being the automatic consequence of conversion. In the Prophetic tradition which probably came into being at this time, there is no trace of voluntariness. But at the same time there were others, who disliked the idea of walà" over a free person and rejected walà" al-islàm. From the second half of the century onwards, the scholars who endorsed walà" al-islàm promoted the voluntary form of this institution.91 Later, the formula of voluntary walà" al-islàm was fixed.92 Regarding the foundling, very different opinions existed but far too few traditions to allow us to establish convincing common links 90 That the consent of the convert is taken for granted is confirmed by the inverted word order in some versions. See for example 'Abd al-Razzàq, Mußannaf, IX, no. 16272 (Manßùr—Ibràhìm): “He takes a man as his patron and converts at his hand” ( yuwàlì ar-rajul wa-yuslim 'alà yadayhi ). Here, it is not conversion which leads to walà" but the intention of the convert to enter a walà" tie with this particular man. 91 Ibn 'Uyayna, Abù Óanìfa, Abù Yùsuf and al-Shaybànì; e.g., al-Shaybànì, Aßl, IV, 183. 92 For the lack of automatism in the creation of walà" by conversion, see also Juda, who asserts that the impact of the Prophetic ˙adìth on walà" al-islàm has been overestimated by Western scholars (Aspekte, 86). Not every convert became the mawlà of an Arab. This is clear from the example of Naba†ì farmers who were reviled by the Arabs and therefore not accepted as their clients (ibid., 159–60). In the biographical and historiographical literature there are many examples for walà" independent from and later than conversion (ibid., 73–74). Since “in the eyes of the authorities, a convert who had failed to acquire a patron [. . .] did not count as a Muslim at all” (Crone, Roman law, 125 note 7), non-Arab converts, nevertheless, might have been keen to conclude a walà" contract with a Muslim. The discrepancy between historic reality and fiqh is described by Daniel Pipes, Slave Soldiers and Islam. The genesis of a Military System (New Haven/London: Yale University Press, 1981), 172, as follows: Although clients “had a legal right to change patrons . . . mawlas who changed allegiance were never trusted again.”

the origin and development of the islamic patronate

93

except for the Medinese al-Zuhrì (d. 124/742) who, from Abù Jamìla, transmits an 'Umar tradition in the form of an anecdote. This tradition belongs to the group of traditions granting walà" to the finder of the child. 'Umar says that walà" belongs to the finder and the state has the obligation to pay for the maintenance of the child. There are five transmitters of al-Zuhrì’s tradition,93 all having fairly different texts which means that the tradition was surely transmitted by al-Zuhrì. In all probability, al-Zuhrì did not invent the text and the lower part of the isnàd but did transmit it from his authority Abù Jamìla.94 But no possibility exists to check if the tradition really can be traced back to 'Umar because there are no other transmitters except for al-Zuhrì. According to al-Óakam, a Kùfan contemporary of the Medinese al-Zuhrì, the finder, who had obtained the walà", has to pay maintenance for the foundling himself. He has no right to be helped by the state.95 A third opinion is that of the Kùfan Óammàd or his authority Ibràhìm (on an illegitimate child which is commonly put on the same level as a foundling)96 stating that [the finder could receive financial help from the state] if his debts for the foundling are testified.97 It is clear that among the supporters of the rule that walà" over a foundling belongs to the finder, there was a discussion at the beginning of the second/eighth century about the maintenance of the foundling. 93 The transmitters are Màlik ('Abd al-Razzàq, Mußannaf, IX, no. 16182), Ma'mar (ibid., no. 16183), the partial common link Ibn 'Uyayna (Ibn Abì Shayba, Mußannaf, XI, no. 11 615 and al-Bayhaqì, Sunan, X, 298. The latter has Ibn al-Musayyab between Abù Jamìla and 'Umar which is surely a later addition), Ibn Abì Dhi"b (Ibn Abì Shayba, Mußannaf, XI, no. 11 620) and 'Amr b. Dìnàr ('Abd al-Razzàq, Mußannaf, IX, no. 16190). Ibn Dìnàr mentions an illegitimate child instead of a foundling. 94 The isnàd 'Abd al-Razzàq—Ma'mar—al-Zuhrì—ßa˙àbì is reliable which means that the material goes back to the generation preceeding al-Zuhrì. See Motzki, “Der Fiqh des- Zuhrì: die Quellenproblematik,” Der Islam 68 (1991), 8 (Ma'mar’s transmission of a˙àdìth) and 33, 40 and passim (al-Zuhrì’s transmission). 95 'Abd al-Razzàq, Mußannaf, IX, no. 16189. Al-Óakam transmits the tradition from Shuray˙. Whereas al-Óakam’s transmission is quite secure (he is the common authority of both Ibn Jurayj and al-Óasan), the authorship of Shuray˙ cannot be proven. 96 For example by Ibràhìm or one of his successors in the isnàd (Ibn Abì Shayba, Mußannaf, IX, no. 11618) and by 'Amr b. Dìnàr—al-Zuhrì—an anonymous scholar— 'Umar ('Abd al-Razzàq, Sunan, IX, no. 16190). 97 Abù Óanìfa, the transmitter of this tradition, adds to this statement that the finder can only get subsidy if he is obliged by the state (sul†àn) to pay for the maintenance. 'Abd al-Razzàq, Mußannaf, IX, no. 16188.

94

ulrike mitter

Other scholars apparently loosened the bond between finder and foundling. In this connection, we frequently encounter the opinion that a foundling is “free.” But the meaning of this seemingly clear statement is not at all clear. Sometimes, it means that the foundling is free to choose as a patron his finder or somebody else.98 This would imply an endorsement of walà" over a foundling in a modified form. The tie does not rise automatically but only after the agreement of the weaker party. Sometimes, the statement means that walà" belongs to “them,”99 but it is not clear whether this is a group of people, as in the case of strangers, or the Muslim state. In the first case, a kind of contractual walà" over the foundling would be endorsed, in the second case, walà" would be rejected. The latter is the opinion of the Màlikis.100 Sometimes, however, the statement is even combined with the opinion that walà" belongs to the finder.101 Here, the statement can only mean that the foundling is a free person in contrast to a slave. All this implies that the isolated statement “the foundling is free”102 does not indicate the absence of walà" over a foundling. An explicit objection to such a walà" tie and the advocation of the foundling’s protection by the state are not very often mentioned and mostly by scholars from the middle of the second century onwards.103 At that time, also, 'Umar’s words “the foundling is free” were taken out of their context, leaving the impression that 'Umar rejected walà" over a foundling,104 which was not the case, as we have seen. 98 This opinion is ascribed to 'Alì b. Abì ˇàlib. Ibn Abì Shayba, Mußannaf, XI, no. 11616. 99 Also ascribed to 'Alì. 'Abd al-Razzàq, Mußannaf, IX, no. 16184. 100 In support of his argument, Màlik cites 'Alì with the words “the foundling is free.” Sa˙nùn, Al-Mudawwana al-kubrà, 5 vols. (Beirut, 1410/1994), II, 577–78. 101 In Màlik’s and Ibn 'Uyayna’s version of the 'Umar tradition, for which see above. 102 As it is transmitted, for example, by Sufyàn al-Thawrì (Kùfa, d. 161/778) from al-Sha'bì and Ibràhìm. 'Abd al-Razzàq, Mußannaf, IX, no. 16187. 103 For example, by Sufyàn al-Thawrì ('Abd al-Razzàq, Mußannaf, IX, no. 16185) and Màlik. Also al-Óasan al-Baßrì (d. 110/728) is said to have awarded walà" of a foundling to the state. Hishàm (Baßra, d. 147/764) is the common link of two different versions of al-Óasan’s tradition (Ibn Abì Shayba, Mußannaf, XI, no. 11619; al-Bayhaqì, Sunan, X, 298, ll. 13–14). In the version of al-Bayhaqì, it is additionally stated that the “ßà˙ib” of the foundling does not receive anything. The textual basis is scarce but the differences between the two texts constitute an argument for their independent transmission from Hishàm. The possible attribution of the tradition to al-Óasan cannot be proved because Hishàm is the only transmitter of alÓasan’s tradition. 104 Al-Bukhàrì ( farà"i∂, no. 19) cites 'Umar’s words adding a version of the Barìra tradition, to confirm his opinion that there is no walà" except walà" al-'itq.

the origin and development of the islamic patronate

95

In his Origins, Schacht mentions the 'Umar tradition and interprets it as a later combination of two different doctrines, namely giving walà" to the finder and giving walà" (or the protection of the foundling) to the state. He argues that it is illogical to award walà" to the finder but places the obligation to maintenance onto the state.105 Schacht mentions that the asànìd of this tradition converge in al-Zuhrì implying that al-Zuhrì is responsible for combining the two doctrines. But as we have already seen, the splitting of walà" and maintenance is just one of three different possibilities which were discussed at the beginning of the second century. It is not illogical to charge the community with the payments, if the finder would otherwise run into debt. b) Hereditary title of the patron In classical law, the manumitter takes the place of an agnatic relative of his freedman and inherits as his last agnate ('aßaba). He is excluded from inheritance by agnatic blood relations but in his turn he excludes the cognates (dhawù l-ar˙àm).106 In early law, however, scholars debated the question whether the manumitter should inherit as the last agnate or rather as the last cognate who would be excluded even by the most distant blood relative of the freedman.107 When

105

Schacht, Origins, 161–62. First of all, the shares ( farà"i∂ ) of the Qur"ànic heirs are deducted from the inheritance. Subsequently, the residue is awarded to the 'aßaba, the manumitter being the last of them. Cf. Bergsträsser and Schacht, Grundzüge, 91. 107 The cognate doctrine is also examined by Crone in her article “Two Legal Problems Bearing on the Early History of the Qur"àn” ( Jerusalem Studies in Arabic and Islam 18 [1994], 1–37). In Crone’s view, this rule—which she calls “DAEP rule” (dhawù l-ar˙àm exclude patrons)—is evidence for the conclusion that in the first century, Qur"ànic heirs (like mother and daughter) and non-Qur"ànic heirs (dhawù l-ar˙àm in classical sense, like maternal aunt or daughter of a paternal uncle) were regarded relatives of the same type or, to put it differently, that the Qur"ànic shares were unknown or ignored (ibid., 34). I think, however, that Qur"ànic and nonQur"ànic heirs were not regarded as heirs of the same kind, although the term dhawù l-ar˙àm covered both groups in the early period; they only had the same effect in excluding a patron from inheritance. If the adherents of the DAEP rule did not distinguish between Qur"ànic heirs and dhawù l-ar˙àm properly, this was because the distinction did not matter at all in this respect. If the patron was regarded as the last heir, the entire estate would fall to the blood relations, no matter if they were Qur"ànic heirs or not. The distinction was important only in the case of an agnatic patron. In the presence of dhawù l-ar˙àm, the agnatic patron would get the whole estate, in the presence of Qur"ànic heirs, he would get only the residue after deduction of the farà"i∂—as it happens in the Bint Óamza tradition. The early date of existence of the Bint Óamza ˙àdìth is another argument against Crone’s theory that Qur"ànic shares were unknown in the first century. 106

96

ulrike mitter

did the doctrine of the manumitter as an agnatic heir come into being? The dating of the traditions has shown that this doctrine was known already in the second half of the first century or earlier. The most prominent tradition which expresses this doctrine is the Bint Óamza tradition.108 It says that Mu˙ammad gave half of the inheritance to the daughter of the deceased freedman and the other half to the manumitter which is a female one, namely Bint Óamza. That means that the daughter received her obligatory Qur"ànic share and the female patron, as the only agnate of the freedman, received the residue. Had the patron been regarded as the last heir of all, the residue would have been given to the daughter of the freedman as well because of the principle of radd (return of the residue of the estate to the sharers).109 This tradition goes back to the Kùfan 'Abdallàh b. Shaddàd (d. 81 or 82/700–702), the halfbrother of Bint Óamza. It cannot be proved that Ibn Shaddàd transmitted an authentic Prophetic tradition but experience shows that a˙àdìth which were transmitted in families are often trustworthy.110 The tradition could have a historic kernel.111 A fact to which attention should be drawn is that most traditions on the patron’s right of succession have Iraqi asànìd. This is valid for traditions on both doctrines (patron inherits as last agnate or as last cognate). It is valid also for traditions from the tàbi'ùn and for traditions from the ßa˙àba and Mu˙ammad. Obviously, the question of the status of the patron was a legal problem for the Iraqis. The reason for the absence of this discussion in Medina is probably the opinion of most Medinese scholars that ar˙àm are no heirs anyway. If a manumitter inherits at all, something which was never questioned, 108

For example, Ibn Abì Shayba, Mußannaf, XI, no. 11184. Scholars who reject radd—as the Màlikis do—would have given the residue to the public treasury. Cf. Coulson, Succession in the Muslim Family (Cambridge: Cambridge University Press, 1971), 49–52. 110 Schacht, however, concludes the opposite (Origins, 170); but see, for example, the two versions of the Barìra tradition transmitted by the common links Hishàm b. 'Urwa and al-Zuhrì from 'Urwa (from 'À"isha from Mu˙ammad). There are no grounds for the assumption that Hishàm might have transmitted the tradition less trustworthly than al-Zuhrì, just because 'Urwa was his father (for the Barìra tradition, see the analysis in the appendix). As a rule, a ˙adìth transmitted with a family isnàd contains more details than other versions. If a family tradition does not add anything at all to the text of other transmitters, fabrication of the isnàd can be taken into consideration. For an example of a suspicious familiy isnàd, see Mitter, Patronat, 194. 111 For details, see ibid., 146–56. 109

the origin and development of the islamic patronate

97

he must necessarily inherit as an agnate.112 A contemporary of Ibn Shaddàd, the Kùfan Ibràhìm al-Nakha'ì (d. 96/717), was one of the most fervent advocates of the cognate doctrine.113 He even interpreted the Bint Óamza tradition in this sense by saying that Bint Óamza received her part merely as an exeptional means of subsistence (†u'ma).114 Despite the opinion of many early lawyers that a manumitter is no agnatic heir, the agnate doctrine finally gained general acceptance. This might be explained by the influence of an important lawyer as Abù Óanìfa (d. 150/767) who was a supporter of the agnatic patron115 (presumably because it fitted better in the developing systematic structure of walà"). Be this as it may, the agnate doctrine was known—also outside Medina—long before the turn of the century. It can be asked why the Iraqis adhered to the cognate doctrine in the first place, if the Prophet had much earlier promoted the agnate doctrine. But it could well be that the Iraqis did not 112

From the Medinese al-Zuhrì, there are traditions about the exclusion of dhawù l-ar˙àm by the patron ('Abd al-Razzàq, Mußannaf, IX, nos. 16208 and 16215) and exclusion of ar˙àm in general (id., X, no. 19110). For the opinion of Màlik and alShàfi'ì, see Ibn Rushd, Ibn Rushd. The distinguished jurist’s primer. Bidàyat al-mujtahid, trans. Nyazee, 2 vols. (Reading: Garnet Publishing Limited, 1996), II, 412. But 'Umar b. al-Kha††àb is credited with the opposite: a tradition has him award the inheritance of an accidentally killed boy ( ghulàm) to his maternal uncle and probably has a historic kernel. The link of this tradition with walà", however (in some versions the boy is said to have been a freed slave), and the introduction of a Prophetic ˙adìth into the matn seem to be later additions (Mitter, Patronat, 175–79). For examples of the 'Umar tradition, see al-Bayhaqì, Sunan, VI, 214, ll. 3–8; 'Abd al-Razzàq, Mußannaf, IX, no. 16198; Ibn Óanbal, Musnad al-imàm A˙mad (Beirut, 1413/1993), I, nos. 190 and 325. 113 Others are al-Qàsim b. 'Abd al-Ra˙màn (Kùfa, d. 120/737; see 'Abd alRazzàq, Mußannaf, IX, no. 16205), the early Kùfan tàbi' 'Alqama b. Qays (d. between 61–73/680–693; see ibid., no. 16196) and Ibn ˇàwùs (Yemen, d. 132/750; see ibid., no. 16216). 114 Ibn Abì Shayba, Mußannaf, XI. no. 11212. According to Crone, Ibràhìm wished to award the entire estate to the daughter of the freedman because he thought that a blood relative had to receive everything, thereby ignoring the Qur"ànic law of inheritance which awards to the daughter a maximum of half of the estate (“Two legal problems,” 28–29). In my view, however, Ibràhìm’s opinion should not be understood as his ignorance of the Qur"ànic shares but as his intention to exclude the patron (Bint Óamza) from succession, which is in compliance with his opinion that a patron should be regarded as the last heir and never as an agnatic heir. The result may be the same but what Ibràhìm criticized was not the part of the inheritance which the daughter of the freedman missed but the part which the patron received. When denotating the patrons half as an exceptional “tu'ma,” Ibràhìm played down the legal effects of Mu˙ammad’s words (which he, obviously, did not dare to deny altogether) by implying that the Prophet did not consider the patron to be an agnatic heir. 115 Al-Shaybànì, Aßl, IV, 166.

98

ulrike mitter

know about his decision and it was only Ibn Shaddàd who introduded the agnate doctrine in Iraq. Concerning the right of the contractual patron to the inheritance, in most of the early a˙àdìth which are explicitly about walà" almuwàlàt, it is merely stated that the patron indeed has this right.116 This goes for walà" al-muwàlàt with and without conversion. There are only a few traditions—all of them on walà" al-islàm—that clearly state whether the patron inherits as an cognate117 or, very rarely, as an agnate.118 But there are also the traditions just mentioned on the cognate doctrine in the case of walà" al-'itq. Many of them do not specify whether they refer to a manumitter or a contractual patron. A few texts do mention a freedman, though,119 but the traditions without specification must have been applied to the contractual patron as well.120 It would be illogical if they applied only to the manumitter, who was regarded by some lawyers even as an agnatic heir, and not to the contractual patron who in later fiqh ended up as the last heir. According to al-Sarakhsì, the contractual patron is considered to be the last heir because the right of the blood-related

116 In later sources we find the rule that the contractual patron inherits, if the client has no heirs (al-Marghìnànì, Hidàya, III, 274). But in the a˙àdìth, no mention is made as to whether the client has relatives or not. According to Juda, Aspekte, 185, the familiy of a client was entiteld to inherit, provided that they were Muslims. 117 Al-Sha'bì and al-Óasan are said to have treated the patron as last cognate. Al-Sha'bì gave the inheritance to the mother of the client (Ibn Abì Shayba, Mußannaf, XI, no. 11632), al-Óasan gave it to the sister (ibid., no. 11629). 118 I have only one text which explicitly treats the contractual patron as an agnatic heir (although there could have been more scholars who regarded him as an agnate of his client, cf. below. It is transmitted by Ibn Abì Shayba (Mußannaf, XI, no. 11625) from Ibn Numayr from 'Abd al-'Azìz b. 'Umar from his father. 'Umar b. 'Abd al-'Azìz is said to have divided the inheritance of a deceased convert between the daughter and the patron. In a variant of this tradition, the inheritance is divided between the daughter and the sons of the patron (Sa'ìd b. Manßùr, Sunan, III/1, no. 210, from Ismà'ìl b. 'Ayyàsh from 'Abd al-'Azìz). In this variant, however, contractual walà" is not mentioned at all. Normally, this could be interpreted as an abbreviation of Sa'ìd b. Manßùr or a later editor of the Sunan, who introduced a shortened version of this tradition in the chapter on walà" al-muwàlàt. (Very frequently, one has to take the heading of a chapter into consideration when trying to understand the mutùn which are often not more than abbreviated versions.) But since Ibn Abì Shayba’s version is, as far as I can tell, the only tradition which applies the agnate doctrine to a contractual patron, we also have to consider the possibility that the conversion element of Ibn Abì Shayba’s version was introduced later into a matn on walà" al-'itq. 119 For example, 'Abd al-Razzàq, IX, nos. 16203 (collective tradition) and 16205 (al-Qàsim b. 'Abd al-Ra˙màn). 120 Cf. also Crone, “Two legal Problems,” 22–23 and 34.

the origin and development of the islamic patronate

99

heirs of the client—ar˙àm being heirs—cannot be taken away from them through a contract of their deceased relative.121 The problem of the importance which such a contract might assume, reveals a possible reason for the rejection of contractual walà" by Medinese lawyers (like Màlik and al-Shàfi'ì): perhaps they were not willing to grant the contractual patron the position of an agnatic heir, which was the only position they knew for a patron, since they did not recognize the right of ar˙àm to inherit.122 In the first century, however, adherents of the agnate doctrine in the case of manumitters might have applied this doctrine to contractual patrons, too.123 As far as walà" al-muwàlàt without conversion is concerned, the problem of an Arab or a non-Arab stranger who came to live with a tribe and died without having a patron is dealt with. In such a case, the inheritance should be given to the Muslims at large. This is the opinion of 'Umar b. al-Kha††àb, Ibn Mas'ùd, 'A†à" and alZuhrì.124 The problem was not discussed in traditions on walà" alislàm.125 Of course, scholars who advocated automatic walà" out of conversion did not deal with such a situation because, according to them, it could not arise. But the problem of a convert living with Muslims without being their mawlà is implied in those traditions on walà" al-islàm which ask for an additional contract of walà", as mentioned above. If the convert did not enter such a contract he was 121 Al-Sarakhsì, Mabsù†, VIII, 92, l. 15. See also al-Marghìnànì, Hidàya, III, 274; al-Shaybànì, Aßl, IV, 191 (detailed) and id., Al-Jàmi' al-ßaghìr (Karachi, 1407/1987), 207 (summarized). 122 According to Màlikì law, if there are no 'aßaba the inheritance falls to the bayt al-màl after having subtracted possible farà"i∂. Only if the state is corrupt, the inheritance may be given to the ar˙àm. Cf. Th.W. Juynboll, Handleiding tot de kennis van de Mohammedaansche wet, volgens de leer der Sjàfi'itische school (Leiden: Brill, 1930), 249. 123 In several traditions on a patron’s agnatic right of inheritance there is no specification of the type of patron. See, for example, the traditions of Zayd b. Thàbit who is said to have generally excluded cognates (ar˙àm) from succession and to have awarded the inheritance of a client to the patrons (mawàlì) in the presence of cognates. 'Abd al-Razzàq, Mußannaf, IX, no. 16207; al-Bayhaqì, Sunan, VI, 241, ll. 16–18. Different versions of the traditions circulated toward the end of the first century. See also the version of the tradition of 'Umar b. 'Abd al-'Azìz which has been mentioned above. 124 For the traditions of 'Umar and Ibn Mas'ùd, see above. For 'A†à", see 'Abd al-Razzàq, Mußannaf, no. 16168, on the sàqi†. For al-Zuhrì, see ibid., no. 16172. 125 I know of one matn, where the question of the inheritance of converts without patrons is mentioned explicitly, though unvoluntarily. It is the version of the 'Umar tradition on walà" al-muwàlàt to which the element of conversion was added later. There, 'Umar b. al-Kha††àb awards the inheritance of converts, who are not attached to anybody, to the public treasury.

100

ulrike mitter

exactly in the same situation as the stranger without walà" (who, in Islamic times, likewise was a Muslim). In fact, non-Arab strangers without patrons and converts without a walà" contract form the same group of people because converts are normally supposed to be nonArabs. Here, we can see the transition from walà" al-muwàlàt without conversion to walà" al-islàm: the rule that the inheritance of Arab and non-Arab strangers without patrons was given to the community, probably stems from pre-Islamic times. In Islam, it was applied first to Arab and non-Arab (Muslim) strangers alike, later to nonArab converts. Some scholars, however, did not apply it to converts because they presuppose a convert to enter a walà" contract with the Muslim at whose hands he converted. Almost always in walà" al-muwàlàt, the right of succession is connected with the duty of paying blood-money for misdeeds committed by the client, be he a convert or not. Apparently, the hereditary title of the contractual patrons rests on the payment of blood-money. In case of the manumitter, occasionally inheritance and blood-money are linked, too, but more often, inheritance is linked with the favor (ni'ma) of manumission which he granted to the freedman. Traditions on the foundling rarely contain a statement on the right of inheritance at all, let alone on the status of the patron as last agnate or last cognate. The subject of almost all traditions is the walà" tie of the foundling in general. The traditions are concerned with this basic question without going into details. Of course, having the patronate means having the right to be the heir of the mawlà.126 c) Hereditary title of the client As far as I know, there are only two Meccan traditions to the effect that a freedman can inherit from his manumitter if the latter dies without any heirs, one ascribed to Mu˙ammad,127 the other to 'Umar 126 Three texts mention the inheritance of a foundling explicitly. One short version of the 'Umar tradition (Ibn Abì Shayba, Mußannaf, XI, no. 11620), one version of the Ibràhìm tradition where the foundling is compared to an illegitimate child (ibid., no. 11618) and a tradition (on the inheritance of a women) which says that a woman inherits from her foundling, her freedman and her illegitimate child. This matn is ascribed to Wathìla b. al-Asqa' (Ibn Abì Shayba, Mußannaf, XI, no. 11621) and via Wathìla to the Prophet (in the canonical collections of Abù Dàwùd, Sunan, ed. Mu˙yì l-Dìn 'Abd al-Óamìd (Beirut: Dàr al-Kutub al-'Ilmiyya, n.d.), III, no. 2906; and al-Tirmidhì, Sunan, IV, no. 2115). This might be an example for “growing backwards” of asànìd which is postulated by Schacht, e.g., Origins, 165. 127 For example, in 'Abd al-Razzàq, Mußannaf, IX, no. 16191.

the origin and development of the islamic patronate

101

b. al-Kha††àb.128 Both traditions are anecdotes which were circulating in Mecca around the turn of the first century. We have to conclude that either the transmitters of the traditions thought that patron and client should have equal rights of succession129 or that the traditions are authentic because somebody must have been responsible for spreading this doctrine. In classical law, the right of succession was assumed to be unilateral. What about contractual walà"? Contrary to what might be expexted, the right of a client to inherit is not touched upon at all in early ˙adìth. Invariably, we read of the right of the patron or of the man who paid blood-money to inherit but nothing about the client’s right. Even the right of succession of an Arab stranger who is registered in the dìwàn of a tribe and has to pay blood-money for it’s members,130 is not mentioned anywhere in the early a˙àdìth. In later sources, the client does have the right to stipulate that succession should be mutual, and walà" al-muwàlàt seems to have been considered a relationship of equals in every respect.131 It seems unlikely, however, that early scholars should have taken the right of succession into consideration only for a former slave and not for a freeborn person. All the more since payments of blood-money were considered to be mutual in walà" al-muwàlàt, which is not the case in walà" al-'itq. The reason for the missing discussion could be that a contractual patron normally had many relatives and it would practically be impossible that his client could ever come into the situation to inherit from him. Besides, contractual walà" without conversion is frequently described as a tie between one single stranger and a group of people or a tribe (qawm) whose members, of course, would inherit first among themselves. The stranger, on the other hand, is supposed to be alone.132 It fits in that scenario that 'Umar b. al-Kha††àb was

128

For example, ibid., no. 16194. The common link of the Prophetic ˙adìth is 'Amr b. Dìnàr (d. 126/744) who transmits the tradition from 'Awsaja—Ibn 'Abbàs—Mu˙ammad. The 'Umar tradition is transmitted by 'Ikrima b. Khàlid (d. after 114/732) and 'A†à" (d. 114/732). They do not mention an authority for their different texts. 130 'Abd al-Razzàq, Mußannaf, IX, no. 16177. 131 For references, see Hallaq, “Use and Abuse,” 35 note 23. 132 Cf. the tradition of Ibn Mas'ùd on the inheritance of a stranger who died without heirs, presented above. In later fiqh, the client is definitely supposed to have no family. That is clear from al-Sarakhsì’s statement on an Arab convert who belongs to his tribe and therefore cannot enter a walà" contract (Mabsù†, VIII, 96, l. 14ff.). Cf. also Hallaq, “Use and Abuse,” 10–11 and references given there. 129

102

ulrike mitter

asked once about the inheritance of a stranger which had to be distributed (see above), not about the inheritance of a patron. It could now be asked, why the right of a freedman to the inheritance of his manumitter had been discussed in the first place if patrons are supposed to have had many relatives. Perhaps, with regard to a manumitter, things were different. Obviously, slaves were sometimes freed in order to secure their patrons old-age maintenance,133 which could mean that those patrons did not have many relatives who would take care of them. Another reason for the mentioning of the freedman’s, but not the contractual client’s right of succession, could be the much closer relationship between a manumitter and his freedman.134 Concerning contractual walà", there were some lawyers who even thought that the patron should not have the right to inherit from his client.135 d) Payments of blood-money The obligation of the manumitter to pay blood-money on behalf of his freedman is not mentioned very often in early ˙adìth but this does not mean that he was not considered to have to pay.136 The silence of the sources should be taken to mean that there was no juridical problem. The situation must have been obvious: if the freedman had no relatives of his own, the 'àqila (clan who pays the bloodmoney) of his patron had to pay. This conclusion is not a total conclusio e silentio since it is corroborated by the existence of traditions on special cases. These special cases are, for example, payment of blood-money by female patrons,137 blood-money for sawà"ib138 and

133

Manumission was sometimes linked with the condition of monthly payments as long as the patron lived ('Abd al-Razzàq, Mußannaf, VIII, no. 15611) and with the condition of staying with the patron (ibid., no. 15600). Such conditions were criticized by lawyers around the turn of the century. Cf. also Crone, Roman law, 67ff. 134 The manumission of a slave is compared with the procreation of a son. See Forand, “The Relation of the Slave and the Client to the Master or Patron in Medieval Islam,” International Journal of Middle East Studies 2 (1971), 61. 135 Juda, Aspekte, 185–86. 136 This is, however, the conclusion of Juda, ibid., 188. The Arabs, he maintains, were supposed to pay blood-money only for contractual clients, because only they were registered in the dìwàn together with members of a tribe. 137 'Abd al-Razzàq, Mußannaf, IX, nos. 16258 and 16291. Both are traditions of the Meccan 'A†à" (d. 114/732). See below on kubr. 138 See below on tasyìb.

the origin and development of the islamic patronate

103

the problem of patrons who try to avoid paying blood-money.139 All these traditions which imply that a manumitter normally had to pay blood-money, were current at the beginning of the second century. It is reasonable to think that they came into being at the end of the first century. The Íafiyya tradition which transmits a decision of 'Umar in the case of a female patron is even older and probably goes back to a historic event.140 In the a˙àdìth, scholars hardly ever discussed whether the freedman has to pay blood-money on behalf of his patron. According to later lawyers, the freedman does not have this obligation since he is no 'aßaba and no heir of the manumitter.141 Probably, the tàbi 'ùn did not deal with this issue because patrons normally had an 'àqila. Why should they bother, then, with a non-existant problem? Typically in later law, which was more systematic and less practical, this aspect was discussed, too, however theoretical it might have been. Concerning walà" al-muwàlàt, the situation is different. Payments of blood-money by the patron on behalf of his client (stranger or convert) are not only mentioned frequently in the sources but almost always the duty of paying blood-money is connected with the right of succession. Payments of blood-money are considered to be so important that they even can enable the payer to receive the inheritance of a deceased stranger without being his patron or, to put it differently, paying the blood-money leads to walà" al-muwàlàt.142 As far as an Arab and a non-Arab stranger is concerned, the payments are assumed to be reciprocal, especially if the stranger is registered in the dìwàn of the tribe. This is the opinion, for example, of 'A†à" on a non-Arab stranger143 and of 'Umar b. al-Kha††àb on an Arab 139 'Abd al-Razzàq, Mußannaf, IX, nos. 17852 ('A†à") and 17855 (al-Thawrì, d. 161/778). For more examples, see Crone, Roman law, 62 note 183. 140 This ˙àdìth will be analyzed in some details below (kubr). 141 Ibn Qudàma, Mughnì, VIII, no. 6831. Al-Shàfi'ì thought that the duty of paying blood-money was reciprocal if there were no relatives of the freedman who could pay (Umm, VI, 116, ll. 21–22) but his opinion was not generally accepted. 142 E.g., the tradition of 'Umar b. al-Kha††àb ('Abd al-Razzàq, Mußannaf, IX, nos. 16178 and 16220). On the connection of inheritance and blood-money, see above. Since the payment of blood-money for a crime committed by the client is essential for the walà" tie, the Imàmìs do not speak of walà" al-muwàlàt but of walà" ta∂ammun al-jarìra. Crone, Roman law, 127 note 50. See also the article by Gleave in this book. 143 “They pay blood-money for him and he for them.” More precisely, these are the words of Ibn Jurayj when he asked 'A†à" about the inheritance of a man who

104

ulrike mitter

stranger.144 Regarding walà" al-islàm, however, blood-money payments of the convert for his patron are never mentioned.145 Concerning a foundling, payments of blood-money are no matter of interest. I came across only one tradition which mentions bloodmoney for a foundling. It is an 'Alì tradition which says that the foundling is free and payments of blood-money and walà" belong to “them.”146 What mattered much more in connection with foundlings was the payment of their maintenance. Transfer and dissolution of walà" In classical law, walà" al-'itq is regarded an agnatic kinship tie. It arises automatically on manumission and it is impossible to dissolve it. In pre-classical law, it had been easier to change patrons but also in classical law, there are certain situations in which walà" must be passed on to a third person, for example, when the manumitter dies. With regard to walà" al-muwàlàt, the subject of changing patrons was a matter of discussion, too. The different opinions of early lawyers will be described here. a) Changing of patrons If a patron manumitted his slave without claiming the patronate, the slave, who is called a sà"iba (pl. sawà"ib), was free to go wherever he wanted and to choose another patron.147 The ˙adìth analysis shows that until the turn of the first century, this practice (tasyìb) was widely accepted.148 If the sà"iba died without relatives and without a new patron, his inheritance was given to the manumitter despite his renunciation. This is expressed in two anecdotes which are ascribed to joined a group of people without being their mawlà. 'Abd al-Razzàq, Mußannaf, IX, no. 16168. 144 Ibid., nos. 16174, 16177 and 16178. 145 Juda’s statement “Es wird nie erwähnt, daß ein maulà [meaning contractual client] für einen Araber Blutgeld zahlen mußte,” (Aspekte, 186) must be referring only to traditions on walà" al-islàm because, as we have seen, this statement would not be correct for walà" al-muwàlàt without conversion. 146 'Abd al-Razzàq, Mußannaf, IX, no. 16184 with the isnàd al-Óasan b. 'Umàra— al-Óakam—Ya˙yà b. al-Jazzàr—'Alì. It is not clear, however, who “them” refers to. 147 In a sense, tasyìb had the same legal consequences as walà" al-muwàlàt. A sà"iba was allowed to enter a walà" contract with the patron of his choice on a par with a free person. In Shì'ì law, tasyìb and contractual walà" were linked much closer than they were in Sunnì law; see Gleave’s contribution to this book. 148 The development of tasyìb is described in detail in Mitter, “Manumission.”

the origin and development of the islamic patronate

105

'Umar b. al-Kha††àb. One is transmitted in different versions by six scholars of the generation of the tàbi'ùn, two of them having died before the turn of the first century.149 If so many different versions of this tradition were current at the end—and two of them even in the middle—of the first century, it is probable that they are reflecting a historic event. For the other tradition, the only transmitter mentioned in the sources is the Meccan 'A†à" (d. 114/732).150 He does not name any authority for the tradition but there is no reason to assume a priori that he invented the whole story. More probably, he transmitted an 'Umar tradition that was circulating around the turn of the first century.151 If the sà"iba did not take a new patron and did not have any relatives, the manumitter was also held responsible for paying the bloodmoney. An early example for a tradition dealing with this problem is, again, an 'Umar tradition which surely existed at the time of the Medinese Sulaymàn b. Yasàr (d. 104 or 107/722–726) and perhaps goes back to a historic event.152 We should note that scholars who awarded the inheritance to the manumitter and held him responsible for the blood-money did not reject tasyìb, although later lawyers took such traditions as examples for rejection of tasyìb.153 Until the turn of the first century, tasyìb was still valid and even regarded a religious virtue154 but the authorities had to find a solution to the problem of a person who was not integrated into a family or a patronate. So they were sometimes forced to ignore the renunciation of the patronate by the manumitter. At the beginning of the second century, the lawyers in general still did accept tasyìb but they started to approach the problem in a more systematic and theoretical way. They decided on the basis of theoretical cases, whether the manumitter or the state must be 149 Namely 'Abdallàh b. Wadì'a b. Khidàm, the son of an eyewitness to the event, who died in 63/682–683 (al-Bayhaqì, Sunan, X, 300, ll. 14–19) and 'Abdallàh b. Shaddàd who died 81 or 82/700–701 (e.g., al-Dàrimì, Sunan, II, no. 2983). 150 E.g., 'Abd al-Razzàq, Mußannaf, IX, no. 16226. 151 For 'A†à"s transmission of 'Umar traditions, see Motzki, Die Anfänge der islamischen Jurisprudenz. Ihre Entwicklung in Mekka bis zur Mitte des 2./8. Jahrhunderts (Stuttgart: Franz Steiner, 1991), 134–36. 152 E.g., 'Abd al-Razzàq, Mußannaf, VIII, no. 18425. 153 Both al-Bayhaqì (Sunan, X, 299–302) and Ibn Qudàma (al-Mughnì, 10 vols. [Cairo, 1968–69], VI, no. 4938) cite the 'Umar traditions together with anti-tasyìb traditions. 154 Tasyìb is compared with ßadaqa in an 'Umar tradition which existed at the end of the first century. See al-Dàrimì, Sunan, II, no. 3119.

106

ulrike mitter

held responsible for a sà"iba without a new patron.155 Whereas during the first century decisions on matters of inheritance of the sà"iba and pending claims of blood-money of the sà"iba were made separately, now, both aspects were put together in general pro or contra statements. From the middle of the second century onwards, tasyìb became forbidden. Abù Óanìfa and al-Shàfi'ì declared it void altogether156 but Màlik still accepted some of the legal implications of tasyìb. He conceded the right of renunciation of walà" to the manumitter but he deprived the sà"iba from his right to choose a new patron. Instead, he asigned the walà" to the Muslims at large.157 In my opinion, the change of mind could have been initiated by problems arising from sawà"ib without relatives and patrons as described above. With the beginning of systematic reasoning of the tàbi'ùn, inconsistencies (such as forcing the manumitter to act as a patron although he was allowed to renunciate walà") must have become more and more obvious. The isnàd-cum-matn analysis leads to the conclusion that as early as the time of 'Umar, the inheritance of a deceased sà"iba was given to the manumitter even against the latter’s will when there was no new patron. On the one hand, this would suggest that already in the first century the entitlement to the inheritance was considered to be a strong right of the manumitter and, on the other hand, that tasyìb meant renunciation of this right (although the authorities could not always show consideration for this renunciation).158 At the time of the tàbi'ùn there also was another possibility of freeing a slave without obtaining his walà", namely by granting him a kitàba contract (that is a contract which allows the slave to buy himself free after the payment of some installments)159 and by stipulat155 The Meccan 'A†à" (d. 114/732) awarded the inheritance of a sà"iba, who died without a patron, to the manumitter ('Abd al-Razzàq, Mußannaf, IX, no. 16152). According to the Medinese al-Zuhrì (d. 124/742), the patronate belongs to the treasury and by no means to the manumitter of a sà"iba (ibid., no. 16228). 156 Al-Shàfi'ì, al-Umm, 8 vols. in 4 (Beirut, n.d.), IV, 127. For Abù Óanìfa, see al-Marghìnànì, Hidàya, III, 271. 157 Màlik, al-Muwa††a", riwàyàt Ya˙yà b. Ya˙yà al-Masmùdì, ed. Mu˙ammad Fu"àd 'Abd al-Baqì (Cairo, n.d.), chap. 38, no. 25. 158 In Crone’s view, Roman law, 68, a manumitter did not have the right to inherit from his freedman in the first century but he had the right to claim further service. Accordingly, tasyìb would not mean renunciation of walà" but renunciation of service. 159 On kitàba and the status of a mukàtab (slave who buys himself free), see Brockopp, Law, 165ff.

the origin and development of the islamic patronate

107

ing in that contract that the former slave may choose the patron he wants after the payments. Later, this stipulation was forbidden and walà" always fell to the one who granted the kitàba contract.160 Under certain circumstances, an automatic change of patrons— not concerning the freed person, however, but his or her children— can be achieved within the institution of walà" al-'itq. If a slave-woman is manumitted, but still married to a slave who has another master, walà" over the children belongs to the patron of the mother. However, if her husband is freed afterwards, he pulls ( jarra) the patronate of the children to his patron. This principle is called jarr al-walà". There is an 'Uthmàn tradition to this effect of which the common link is al-Zubayr b. al-'Awwàm (d. 36/656).161 There had been other opinions, too, like walà" of the children remains with the patron of the mother162 or the father “pulls” walà" only during his lifetime.163 But eventually, the right of the father over his children was stronger than the right of the mother and of her manumitter,164 although the change of patrons contradicts the rule of walà" as an agnatic relationship that cannot be transfered to another patron.165 Contractual walà" was regarded as a significantly less binding relationship than walà" al-'itq and, therefore, changing of patrons was basically possible. As to walà" al-muwàlàt without conversion, its conclusion was totally voluntary.166 A stranger could live with a tribe

160 'A†à" (Mecca, d. 114/732) and Qatàda (Baßra, d. 117/735) allowed the changing of patrons. See 'Abd al-Razzàq, Mußannaf, IX, no. 16217 ('A†à") and nos. 16218, 16219, 16158 and 16159 (Qatàda). For the classical doctrine see Ibn Qudàma, Mughnì, VI, no. 4943. Cf. Crone, Roman law, 86. 161 The tradition is transmitted by three scholars and stems possibly from alZubayr. 'Abd al-Razzàq, Mußannaf, IX, nos. 16281–16284. 162 This opinion is, for example, ascribed to Shuray˙ (Kùfa, d. between 76–79/ 695–698), who is said to have changed his opinion when he heard that Ibn Mas'ùd (Kùfa, d. ca. 33/653) applied the jarr rule. 'Abd al-Razzàq, Mußannaf, IX, nos. 16278 and 16279. 163 Ibid., no. 16299. Sa'ìd b. Jubayr (Kùfa, d. 94 or 95/712–714) and Mujàhid (Mecca, d. between 100–104/718–723). 164 Al-Sarakhsì, Mabsù†, VIII, 87. The Shì'ìs argue for the priority of the father with the Qur"ànic verse on adoption (Q.33:5 “call them after their fathers”) which they interpret by analogy as applying to walà", too. Gleave discusses this in his article. 165 The principle of jarr al-walà" is discussed by al-Sarakhsì, Mabsù†, VIII, 93, ll. 1–2—but not by early lawyers—in connection with contractual walà". 166 For 'Atà"s statement on the matter, see 'Abd al-Razzàq, Mußannaf, IX, no. 16168 and Ibn Abì Shayba, Mußannaf, XI, no. 11617. 'A†à" names Ibn Mas'ùd as an authority for this opinion.

108

ulrike mitter

without entering a walà" contract with anybody.167 From walà" alislàm we know the rule that as long as no blood-money had been paid, the client was free to take another patron. Regarding walà" without conversion, there are not many traditions expressing a similar rule but it seems logical that this rule applied to the older type of walà" al-muwàlàt, too.168 As I argued above, in walà" al-islàm it has always been possible to change patrons. The two rules which in later sources express the voluntariness of the contract (first, the rule just mentioned that walà" could be dissolved as long as no blood-money had been paid and second, the rule that the convert had to explicitly express his will to take as a patron the man at whose hands he converted), were already known at a very early stage. There have also been scholars who rejected such voluntariness but their opinion was just one of many. There is one tradition which grants a foundling the possibility of choosing his patron. It is not possible to discern the date of origin of this tradition, which is ascribed to 'Alì, even approximately because the isnàd is a single strand and there are opposite views ascribed to 'Alì.169 Usually, walà" of the foundling is said to have belonged either to the finder or to the state. The mentioned 'Alì tradition expresses an opinion in between: Walà" does not rise automatically because the foundling is regarded as a free person. Therefore, his walà" is equated to walà" al-muwàlàt, the foundling can choose a patron. But the later doctrine, that the state is responsible for foundlings, had not yet been reached. Finally, there is a famous Prophetic ˙adìth in which changing of patrons or clients (without permission) is condemned. There are two variants of this ˙adìth: one does prohibit the changing of patrons without permission,170 which, in fact, means that it does allow the chang167 This is clear from traditions which solve the problem of the inheritance of Arab and non-Arab strangers who died without relatives and patrons, discussed above. 168 One matn of Ibràhìm al-Nakha'ì (“on a man who takes a man as his patron”), actually words this rule. 'Abd al-Razzàq, Mußannaf, IX, no. 16160 (with the isnàd Ibn 'Uyayna/Ma'mar—Manßùr—Ibràhìm). Since all other traditions of Ibràhìm on walà" al-muwàlàt are, however, traditions on walà" al-islàm, the text mentioned here could originally have dealt with a convert, too. 169 The 'Alì traditions have been listed above. 170 “Là ya˙illu [li-muslim] an yatawàlà mawlà rajul muslim bi-ghayri idhnihi (it is not allowed for a Muslim to take the client of an[other] Muslim as his client without [the patron’s] consent).” 'Abd al-Razzàq, Mußannaf, IX, no. 16154. “Man tawallà mawlà qawm bi-ghayri idhni mawàlìhi (who takes the client of people as his client with-

the origin and development of the islamic patronate

109

ing. The other variant prohibits it outright.171 In what follows, the two types will be called “permission variant” and “prohibition variant.” The mutùn belonging to the permission variant are usually shorter and do not contain as many details and combinations with other rules as those of the prohibition variant.172 It is not totally clear whether the ˙adìth applies to walà" al-'itq or walà" al-muwàlàt or both since most mutùn do not specify walà". There are some indications of a connection with walà" al-'itq.173 But since the ˙adìth seems to have developed from a clause of the Constitution of Medina on allies,174 it could have been applied to contractual clients as well.175

out the permission of his patrons),” ibid., no. 16309. In other traditions, the verbal forms wàlà, tawàlà, tawallà, translated here as “take as client,” mean “to take as patron.” For example: “an yu"dhina lahu an yatawàlà man shà"a (to allow [the client] to take as a patron whom he wants,” ibid., no. 16151. For the ambiguitiy of the terms, see Gleave’s article in this book, who translates them with the neutral “to associate with.” 171 “Man tawallà/tawàlà ghayra/ilà ghayri mawàlìhi (who takes as clients other than his own clients).” 'Abd al-Razzàq, Mußannaf, IX, nos. 16305 and 16307. 172 The permission variant is sometimes combined with a rule on blood-money and a rule on accepting an adulterer or somebody who is legally impure into Islam (àwà mu˙dithan fì l-islàm). For the meaning of mu˙dith, see Lane, Arabic-English Lexicon, 8 vols. (New York, 1863–93), II, 528. The prohibition variant is combined with several rules on killing, testaments, the maxim al-walad li-l-firàsh (the child belongs to the [marriage] bed) etc. There also are details on the Prophet’s speech, in which the prohibition variant is embedded, and on a camel next to which the alleged transmitter of the tradition stood while listening to the speech. The camel motive is part of the version of the common link Shahr b. Óawshab (Iraq, d. 100/718–9). 'Abd al-Razzàq, Mußannaf, IX, no. 16307. 173 One text of the prohibition variant does specify patrons as manumitters (alDàrimì, Sunan, II, no. 2864). Muslim, Ía˙ì˙, V, 407, who collects versions of the permission variant, has a chapter on the “prohibition for the freedman to choose the patron he wants.” (As an aside, in 'Abd al-Razzàq, Mußannaf, 6, there are similar texts in a chapter on “permission for the client to choose the patron he wants.”) According to Landau-Tasseron, the Prophetic ˙adìth clearly deals with freedmen (see her contribution). 174 In some versions of the permission variant, a ßa˙ìfa (piece of paper) is mentioned—which is a reference to the Constitution of Medina—which is said to have contained the rule on changing patrons. The prohibition variant, however, does not mention such a ßa˙ìfa. Instead, it mentions a speech made by Mu˙ammad. According to Crone, Roman law, 32–33 and 136 note 141, both versions of the tradition developed from a clause of the Constitution on allies (“là yu˙àlifu mu"min mawlà mu"min dùnahu”) which had been misinterpreted by the lawyers as a rule for changing of patrons. First, the permission doctrine came into being, followed by the prohibition doctrine which was developed on a par with the doctrine of walà" as an inalienable tie. 175 In his chapter on the changing of patrons, Juda refers to both walà" al-'itq and walà" al-muwàlàt (Aspekte, 155–58), thereby giving the impression that Prophetic ˙adìth applies to both types of walà" (ibid., 156 and references note 61).

110

ulrike mitter

Both variants were current at the end of the first century. The permission variant was used by 'A†à" (Mecca, d. 114/732) in his argumentation on changing patrons.176 The prohibition variant was more often used by later scholars in connection with prohibition of sale of walà"177 which does not necessarily mean that it originated later. The ˙adìth shows that at the turn of the first century the question of changing patrons and the problems evolving from such changes,178 kept the lawyers busy. b) Selling and giving of walà" During the first century, servile walà" could be sold and given away. There are traditions of Mu˙ammad’s wife Maymùna (giving away)179 and 'Urwa b. al-Zubayr (selling)180 to this effect which are probably authentic. At the turn of the century, some scholars in Óijàz ('A†à")181 and Iraq (Ibràhìm)182 still accepted giving of walà". Even selling of walà" was tolerated in special cases.183 There is no convincing indication for an early opposition to this practice. Mu˙ammad and his ßa˙àba 'Umar, 'Alì and Ibn Mas'ùd are said to have forbidden selling of walà" but we must doubt that these traditions existed in the first half of the first century. Also, the total lack of anecdotes indicates that Mu˙ammad and the early ßa˙àba did not deal with the problem of selling walà". Rejected—but not forbidden—was the selling of walà" in the second third of the first century by some later ßa˙àba as Ibn 'Abbàs (d. between 69–70/688–690),184 Jàbir b. 'Abdallàh (d. 78/697)185 and 'Abdallàh b. 'Umar (d. 74/693).186 Ibn 'Abbàs argued that it should 'Abd al-Razzàq, Mußannaf, IX, nos. 16151 and 16152. Ibn Qudàma, Mughnì, VI, 409 and no. 4936. 178 On these problems, cf. Juda, Aspekte, 155–158. 179 Ibn Abì Shayba, Mußannaf, XI, no. 11663. 180 Ibn Qudàma, Mughnì, VI, no. 4936. There is no isnàd to this tradition but why should it have been invented since it contradicts the classical doctrine? 181 'Abd al-Razzàq, Mußannaf, IX, no. 16151. In the same tradition, 'A†à" is said to have later changed his mind. 182 Sa'ìd b. Manßùr, Sunan, III/1, no. 282. This tradition deals with giving of walà" of a sà"iba. 183 Ibràhìm tolerated it in the case of walà" of a sà"iba (Ibn Abì Shayba, Mußannaf, XI, no. 11666). Sa'ìd b. al-Musayyab tolerated it in the case of a mukàtab (ibid., no. 11669). 184 Al-Dàrimì, Sunan, II, no. 3158. 185 'Abd al-Razzàq, Mußannaf, IX, no. 16143. 186 Ibid., no. 16150. 176 177

the origin and development of the islamic patronate

111

not be allowed to sell a free man. At the end of the first century a tradition of Ibn Mas'ùd (d. 32 or 33/652–654) circulated to the effect that kinship (nasab) cannot be sold.187 There is no mention of prohibiting the giving of walà" yet. From the turn of the century onwards, selling and giving of walà" was mentioned in one breath.188 The prohibition maxim “là yubà'u l-walà" wa-là yùhab” (walà" must not be sold and given away) was coined and ascribed to earlier authorities who were known or supposed to have had a similar opinion. Very likely, the maxim goes back to the famous Prophetic ˙adìth which was put into circulation by 'Abdallàh b. Dìnàr (d. 127/744–745).189 The connection of the prohibition maxim to the question of nasab, too, was effected in this time. This should not be taken to mean, however, that the idea of walà" as a kinship tie had been unknown before this date.190 Expressing this idea in the form of a maxim (“al-walà" lu˙ma ka-l-nasab”), however, and combining it with the prohibition maxim was new. In later times, the nasab maxim was accepted without question as part of the Prophetic tradition, although the matn of the Prophetic tradition transmitted by Ibn Dìnàr did not contain it. Towards the middle of the second century, it was a common doctrine that alienation of walà" was forbidden. Perhaps the acceptance of this doctrine was encouraged by the wide spread of Ibn Dìnàr’s Prophetic ˙adìth. At any rate, there are no traditions of later scholars endorsing the practice of selling walà" and giving it away. Since the argument that walà" cannot be sold, because it is a kinship tie, is relatively late and by far not always mentioned, rejection of alienation of walà" must originally have had another reason. I think it can be found in the argument of Ibn 'Abbàs. After all, selling a mawlà comes to treating him like a slave. In the course of time and with increasing influence of the mawàlì, this could no longer be tolerated. 187

Sa'ìd b. Manßùr, Sunan, III/1, no. 278. Similarly, inheritance and blood-money of a sà"iba were mentioned in one breath around the same time. 189 E.g., 'Abd al-Razzàq, Mußannaf, XI, no. 16138. 190 The analysis of the tasyìb traditions shows that the Muslims had been acquainted with the idea of walà" as nasab since the first century. Even at the time of 'Umar b. al-Kha††àb, the manumitter was considered to have a title to succession. The right of the patron to inherit as an agnatic relative was known at least at the time of 'Abdallàh b. Shaddàd who died 81 or 82/700–701. At the end of the first century, Ibn Mas'ùd was said to have prohibited selling of walà" because walà" equaled nasab. 188

112

ulrike mitter

If selling and giving of servile walà" were problematic, this is even more true for contractual walà". And that is exactly what al-Sarakhsì says.191 In the first century, when selling and giving of servile walà" were still tolerated, the case of walà" al-muwàlàt was never discussed explicitly. There are not many traditions endorsing giving and selling of walà", anyway. If in these traditions the type of mawlà is specified, it always is a freedman, frequently a mukàtab or a sà"iba. I did not come across traditions in favor of selling or giving of contractual walà". The traditions which prohibit this practice, do not always mention the kind of walà", either. If they do, it is walà" al'itq.192 Because of their silence on the subject, early a˙àdìth give the impression that selling and giving of walà" were practised only in the case of the servile tie. Furthermore, the traditions on walà" al-muwàlàt without conversion convey the impression that Arab and non-Arab strangers joined a tribe in order to receive protection and social integration, not in order to be handed over to arbitrariness and to run the risk of their walà" being sold like walà" of a former slave.193 Concerning walà" al-islàm, things are different. For one thing, the traditions do not reveal anything on the kind of relationship between Arabs and non-Arabs. Although one could likewise ask why nonArab converts should have entered into a walà" contract with an Arab if there was the risk that their walà" could have been sold, probably, many non-Arab mawàlì did not have much choice because of their humble social standing.194 Still, the early traditions do not 191 Al-Sarakhsì, Mabsù†, VIII, 98, ll. 7–8. He explains that walà" al-muwàlàt is based on consent (al-tarà∂ì), and that the client (al-asfal ) would never agree to the transfer of his walà" to another person than the one with whom he entered into a contract. The Óanafìs approve of alienation of walà" by the client himself (ibid., ll. 9ff.; al-Shaybànì, Aßl, IV, 201–202). 192 For example, 'Abd al-Razzàq, Mußannaf, IX, no. 16144 ('A†à"); Ibn Abì Shayba, Mußannaf, XI, no. 16664 (Ibràhìm al-Nakha'ì). 193 In pre-Islamic times, where contractual walà" without conversion seems to have originated, two groups associated by walà" “weiden und zelten zusammen [. . .] und betrachten sich in ihren Unternehmungen als Einheit” (Reinert, Recht, 22). Given such a situation, it is not conceivable that even the stronger party could have had the idea of selling their mawàlì. 194 As Pipes, Slave Soldiers, 171, states with regard to mawàlì in the army, “free mawlas had nothing to lose when they went over to Islam.” Even when stressing the fact that the free “fighting mawlas were entirely in the hands of an Arabian patron” Pipes does not, however, mention the possibility that the mawàlì could have been sold by their patron (ibid., 172). But he points out that “walà" of the freedman and that of the free man differed only in minor details, and their social standing was comparable” and “for mawlas who fought, their status was servile” (ibid., 108). On the similiarities between the status of free mawàlì in military service and military slaves, see also ibid., 184–85. In his chapter on transfer and alienation of

the origin and development of the islamic patronate

113

mention sale of walà" contracts, they neither endore it nor do they reject it.195 c) Bequeathing walà" In classical law, walà" is not inherited as an ordinary piece of property. Instead, the kubr doctrine is applied, which is expressed in the rule “al-walà" li-l-kubr.” This rule means that after the death of a patron, walà"—or rather the rights and duties emerging from the walà" tie196—passes on to the next agnatic relative (kubr) of the deceased. In some cases, this rule matches normal inheritance rules, in others it does not. Normal possessions which were inherited by the son of a deceased, would be inherited by the grandson after the death of the son. Not so in the case of walà". If the deceased had a brother, he would exclude the grandson from succession. In other words, walà" would return to the brother of the deceased. According to the kubr rule, women are excluded from succession except in the case of their own freedmen and mukàtabs. The majority of scholars are said to have supported the kubr rule. This does not only hold for tàbi'ùn from Iraq197 and the Óijàz,198 but also for the ßa˙àba, for example 'Uthmàn199 and 'Umar b. al-Kha††àb.200 To the latter the Íafiyya tradition is ascribed, which says that 'Alì b. Abì ˇàlib and his cousin al-Zubayr b. al-'Awwàm argued about a mawlà of Íafiyya, the deceased mother of al-Zubayr and aunt of 'Alì. Both thought that they were entitled to the walà" of her mawlà. 'Umar decided that the walà" tie should be divided between them. Al-Zubayr received the right of succession whereas 'Alì received the duty to pay blood-money. This tradition circulated in the second walà", Juda, Aspekte, 155ff., does not distinguish between servile and contractual walà", thereby implying that selling and giving of walà" could also have included walà" alislàm. 195 One argument against the assumption that walà" contracts were commonly sold is the rejection of some people as mawàlì by the Arabs (see the example of the Nabatean farmers, Juda, ibid., 159–60). If the Arabs had regarded walà" contracts as alienable, they would probably have accepted everyone as their mawlà without exception. 196 Walà" is not inherited but one inherits [the estate of the mawlà] through walà" as in the case of nasab. Al-Sarakhsì, Mabsù†, VIII, 98, l. 5. 197 The Kùfan scholars Ibràhìm al-Nakha'ì ('Abd al-Razzàq, Mußannaf, XI, no. 16257) and 'Àmir al-Sha'bì (ibid., nos. 16261 and 16262). 198 The Meccan ˇàwùs (ibid., no. 16241) and the Medinese Sa'ìd b. al-Musayyab (Ibn Abì Shayba, Mußannaf, XI, no. 11555). 199 Màlik, Muwa††a", chap. 38: 12, no. 22. 200 Sa'ìd b. Manßùr, Sunan, III/1, 274.

114

ulrike mitter

half of the first century and can be traced back to a historic event. It may be asked, why al-Zubayr, who is the kubr of his mother, does not get the whole patronate. The reason for the division of walà" is the interference of blood-money rules. The problem is that a mother does not have the same blood-money group ('àqila) as her son. The 'àqila of a woman consists of members of her father’s tribe but her son’s 'àqila consists of members of her husband’s tribe. Since it is always the patron’s 'àqila that has to pay blood-money for mawàlì freed by him or her, 'Alì, as a member of Íafiyya’s 'àqila, receives the duty of paying blood-money.201 Al-Zubayr, as Íafiyya’s son and her kubr, basically has the right to inherit the walà". But additionally giving him the duty of paying blood-money would have amounted to a change of responsible 'àqila and to a break of the tie between the woman and her mawàlì.202 The opposite doctrine to the effect that walà" should be inherited in the same way as property (màl ) did not have as many supporters. It was propagated above all by the Kùfan judge Shuray˙ (d. between 76–79/695–699). There is a set of traditions, all of them expressing Shuray˙’s opinion, which are transmitted by three tàbi 'ùn.203 The common kernel of their considerably different (and therefore independently transmitted) mutùn must go back to a common source which could be no other than Shuray˙ himself. This judge seems to be less “legendary” than Western scholars used to believe.204 Not only Shuray˙ but also the Meccan al-Zubayr b. al-'Awwàm205 (the one of the Íafiyya tradition just discussed) and his son, the anti201 'Alì must have thought that he was entitled to succession because payments of blood-money constituted an essential element of walà". 202 Nevertheless, some lawyers awarded the whole walà" to the son, including the duty of paying blood-money. Undoubtedly, they did so in order to preserve the wholeness of walà" which consists of the right to inherit and the obligation to pay blood-money. But their opinion did not gain general acceptance. For traditions of the Kùfan scholars 'Àmir al-Sha'bì (d. 103/721) and Ibn Abì Laylà (d. 148/765), see 'Abd al-Razzàq, Mußannaf, IX, nos. 16256 and 16296. 203 Ibràhìm ('Abd al-Razzàq, Mußannaf, IX, no. 16251), al-Sha'bì (al-Bayhaqì, Sunan, X, 303, l. 22–24) and Abù 'Awn al-Thaqafì (Ibn Abì Shayba, Mußannaf, XI, no. 11614). One example for a Shuray˙ matn is “kàna yujrì l-walà" majrà l-màl ” (he treated walà" as possession). This text is recorded by Ibn Abì Shayba with the isnàd Wakì'—Abù 'Àßim—al-Sha'bì—Shuray˙ (ibid., no. 11613). 204 Schacht Origins, 228–29; Crone, Roman law, 82; Motzki, Anfänge, 152, refers to a Shuray˙ tradition from the Meccan 'A†à" and points out that the tradition can be considered to stem from Shuray˙ since there is no reason for a Meccan scholar to have invented an Iraqi authority. 205 Al-Bayhaqì, Sunan, X, 305, l. 5ff.

the origin and development of the islamic patronate

115

caliph 'Abdallàh b. al-Zubayr,206 were known advocates of the màl doctrine. There are reasons for assuming the old age of these traditions and their possible authenticity.207 Obviously, when claiming the inheritance of his mother’s mawlà, al-Zubayr did not think that he should inherit because of the kubr rule but because of his opinion that walà" has to be inherited like property, which of course is inherited in the first place by a son. Both doctrines were current at the turn of the century in the Óijàz and Iraq. Because of the large number of Iraqi asànìd which are attached to Medinese authorities (and because of the old age of most traditions), it can be assumed that the kubr doctrine originated in Medina early in the first century and was then exported to Iraq. The early date of existence of the kubr rule supports the theory that this rule has its roots in an old-Arabian inheritance rule.208 The màl doctrine seems to be younger than the kubr doctrine and might well have been a reaction to the Qur"ànic inheritance rules according to which women have the right of succession (the kubr rule excludes them from succession) and the direct descendants of the deceased always inherit in the first place (in the second generation, the kubr rule prefers a brother to a son).209 Does the kubr rule apply to walà" al-muwàlàt, too? As is often the case, in many traditions there is no specification of the kind of walà" but in some a˙àdìth words like ni'ma (grace) or mu'taq (freedman) point to an application of the kubr rule to servile walà". That does not necessarily mean that all kubr traditions are meant to be applied to walà" al-'itq, exclusively. Yet at the very least, I did not encounter 206 'Abd al-Razzàq, Mußannaf, XI, no. 16245. This tradition is an anecdote which circulated in Mecca and Iraq. 207 Mitter, Patronat, 332–35. 208 According to Robert Brunschvig, the kubr rule was part of Arabian customary law and originally stemmed from the Yemen. For a discussion of Brunschvig’s theory, see Powers, Studies, 90–93. But see Crone, Roman law, 82, who assumes that the Islamic rule of kubr was developed by Muslim jurists in the second century independently from Yemeni rule. 209 According to Powers, the Qur"ànic inheritance law replaced the old Arabian principle of the kubr by that of direct descent (Studies, 210–11). This old kubr rule seems to have been maintained in the institution of walà" but it had been subjected to some changes. Whereas the old-Arabian kubr always was the “eldest” (that is why Powers calls it the “principle of seniority,” ibid., 210), the Islamic kubr is, in the first place, the son of the deceased patron. That difference can probably be attributed to the influence of the Qur"ànic law.

116

ulrike mitter

any traditions which deal with bequeathal of walà" and which unmistakably focuses on walà" al-muwàlàt. Furthermore, the maxim on women (“Women do not inherit through walà" except in the case of their own freedmen and the freedmen of the latter or in the case of their own mukàtabs and mukàtabs of the latter”),210 which is connected with the kubr rule, clearly refers to walà" al-'itq. According to al-Sarakhsì, however, contractual walà" equals walà" al-'itq. If the patron dies, his male descendants inherit the walà" of his client (and of the freedman of his client) to the exclusion of the female.211 This late regulation does not, however, say much about the situation in the first and second centuries. Inasmuch as Shì'ì law often seems to have preserved pre-classical regulations,212 and since in Imàmì law the death of either party puts an end to the contract,213 early lawyers could well have thought walà" al-muwàlàt to end with the death of one party. But without a˙àdìth on that subject, we cannot know for sure if the kubr traditions were supposed to apply to walà" al-muwàlàt, too, or only to walà" al-'itq. Development of walà" in the first and second century AH Having dated the traditions on each aspect of walà", it is now possible to give an outline of the development of walà" in the first and second century, again starting with walà" al-'itq. During the first fifty years of the first century, walà" al-'itq was not as binding a tie as in later times. The tie could still be sold and given away. But at the same time, it was common opinion that the manumitter was entitled to the patronate and to the inheritance of the freedman. Manumission by tasyìb, which granted the freedman the possibility to enter into a walà" contract with a third person, was not unusual. But if the sà"iba died without a new patron and without relatives, his inheritance was given to the manumitter despite his renunciation. It is clear that by freeing a slave, the manumitter acquired an hereditary title which was considered to be so strong that it remained valid in certain situations even when walà" had been renounced. The principle of kubr, according to which walà" would pass on to the next 210 211 212 213

For example, Ibn Qudàmà, Mughnì, VI, no. 4962. Al-Sarakhsì, Mabsù†, VIII, 97, ll. 18ff. Crone, Roman law, 23. Ibid., 40.

the origin and development of the islamic patronate

117

agnate of a deceased patron, was known and widespread but it was not accepted by all scholars. Some thought that walà" should be inherited in the same way as property was. It is clear that walà" between two individuals was known to the early Muslims, even if this tie was not exactely the same institution as it was in later times. In the last third of the first century, the opinion of the lawyers towards the sale of walà" changed. Some lawyers rejected this practice but they did not yet clearly prohibit it. Giving of walà" and manumitting a slave as a sà"iba was still possible. The doctrine of the patron inheriting as an agnate of the mawlà was known by this time or earlier. At the same time, the doctrine of the patron inheriting as last cognate was still widespread. Some scholars granted the mawlà the right to inherit from his patron, if he died without heirs, but this opinion did not gain widespread acceptance. It was agreed upon that the 'àqila of the patron had to come up with the bloodmoney for the mawlà. At the turn of the first century, lawyers of the tàbi'ùn generation in Óijàz and Iraq approached the juridical problems in a more systematic way on the basis of the decisions of the ßa˙àba. This is true for all aspects of walà". Tasyìb, for example, was still considered valid but, now, lawyers provided general statements on tasyìb, combining in one tradition the rules on the inheritance of a sà"iba and on the blood-money which had to be paid for him. The same applies to the kubr doctrine which was accepted now by most scholars. They thought the kubr rule through as they applied it to different theoretical constellations of heirs. Giving walà" away was still tolerated and even selling walà" was accepted by some scholars in special cases. At the same time, however, the prohibition maxim was coined, which combined the prohibition of selling and giving of walà" in one general statement. At that time, the Prophetic maxim al-walà" li-man a'taqa was used as basis of argumentation without mentioning the original context (the Barìra tradition) and without naming authorities.214 Also the collective traditions, which list the opinion of several companions of the Prophet, came into vogue. From the second century onwards, the doctrine of walà" as an agnatic kinship tie took shape. To be sure, the idea was not new but in the course of the general systematization of walà" it was applied 214 For example al-Zuhrì, in 'Abd al-Razzàq, Mußannaf, IX, no. 16165; and ibid., VIII, no., 15784.

118

ulrike mitter

to all aspects systematically. In the middle of the second century, the prohibition of tasyìb gradually dominated (respectively, its acceptance in a restricted form by the Màlikìs) and the prohibition of giving and selling of walà" was not questioned anymore, probably due to the now circulating ˙adìth of Mu˙ammad to this effect; classical walà" came into being. According to the results of the ˙adìth analysis, walà" developed from an institution which was still disputed in some ways and which could be renounced, sold and given away, into a quasi-agnatic tie in all respects. This development started in Medina at the time of the ßa˙àba. The characteristics of what later became classical walà"— agnatic succession of the patron, his obligation to pay blood-money, bequething walà" according to the rule of the kubr, rejection of the sale of walà"—existed in the first century and some elements even at the time of the first caliphs. Alongside these opinions, also nonclassical opinions were expressed such as exclusion of the patron from succession by every relative of the client, acceptance of tasyìb and bequething walà" as normal property. What we have here, is early ikhtilàf. The view of walà" as a kinship tie existed already in the first half of the first century without having been elaborated yet in all aspects. The full nasab doctrine was the result of the systematization of the different aspects of walà" by the tàbi'ùn. The development of walà" al-muwàlàt, on the other hand, cannot be narrated in such detail as we did for walà" al-'itq since not many ahàdìth on the subject exist. Probably, individual contractual walà without conversion existed in the Óijàz early in the first century. This institution was still accepted at the beginning of the second century. From the start, scholars were particularly interested in the right to succession of the patron and his duty to pay blood-money and they dealt with the problem of the inheritance of a stranger without a patron. Typical for the juridical systematization of walà" in the generation of tàbi'ùn is a tradition of al-Zuhrì to the effect that the walà" contract of a stranger needs the evidence of a witness (bayyina 'àdila) to be valid.215 Rejection of walà" al-muwàlàt over strangers is a rare theme in the a˙àdìth. It seems to have been criticized—in the 215 'Abd al-Razzàq from Ma'mar from al-Zuhrì (Mußannaf, IX, no. 16173). I came across only one single version of this tradition but considering Motzki’s analysis of this particular isnàd mentioned above, there is no reason to assume a forgery of this tradition.

the origin and development of the islamic patronate

119

case of a non-Arab stranger—around the turn of the century and consequently fell into disuse without much ado or, in a sense, it converged with walà" al-islàm. Arabs were no longer mawàlì anyway,216 and walà" over non-Arabs soon became explicitly connected with the question of conversion. In case that conversion did not immediately lead to walà" and allowing the convert to enter into a walà" contract with whom he liked, this, in fact, amounted to walà" al-muwàlàt without conversion. The origins of walà" al-islàm are apparently not to be found in the Óijàz—despite certain traditions of Mu˙ammad and some ßa˙àba— but in Syria and Iraq. Nothing precise can be said about the development of the institution during the first century because the oldest traditions seem to stem from the †àbi'ùn. As in the case of other kinds of contractual walà", the spectrum of topics debated in early a˙àdìth is not very large. Specifications about the patron’s right of inheritance or the discussion of jarr al-walà" and the kubr rule are mentioned only much later. Walà" al-islàm has always been individual and the consent of the client seems to have been required from the very beginning, though at the time not in a regulated way. In the first half of the second century, some scholars, such as the Umayyad caliph 'Abd al-'Azìz b. 'Umar, rejected the voluntary form of walà" on conversion. The traditions transmitted by this caliph are an indication of the interest political authorities had for this institution since, otherwise, traditions of caliphs on walà" are rather seldom. In the long run, the doctrine of the advocates of voluntary walà" al-islàm gained acceptance. Inasmuch as scholars of the Óijàz never dealt much with walà" al-islàm, it comes as no surprise that mainly Iraqi scholars rejected the institution.217 Opposition to walà" al-islàm started at the outset of the second century, by analogy with rejection of special aspects of servile walà" like selling and giving of walà". Rejection of walà" over a non-Arab stranger is ascribed to 'Alì at the same time, also by Kùfan scholars. In the first century, walà" over a foundling belonged to the finder. At the beginning of the second century, the scholars discussed the financial responsibility for the foundling which some placed on the 216 On the gradual limitation of the term mawàlì to non-Arabs—a process started right after the first conquests—see Juda, Aspekte, 68–71. 217 On the special attachment of Iraqi and Syrian lawyers to walà" al-muwàlàt, see Juda, ibid., 85–86.

120

ulrike mitter

finder, others on the state, and again others thought that the finder might get some help from the state. It is not possible to precisely date and to properly interpret the opinion that a foundling is regarded as a free person. What we can say is that this rule seems to stem from the first century and that at the beginning, it did not necessarily mean that there was no walà" tie between foundling and finder. It was not until the middle of the second century that a foundling was really supposed to be free in the sense that only the state has his walà".

Conclusions on the nature of walà" and on legal theory and practice It is clear from the ˙adìth analysis that the institution of walà" consists of a very complex set of regulations. To begin with, we have to differentiate between walà" al-'itq (servile patronate) and walà" almuwàlàt (contractual patronate) which were in fact different institutions from the time of the ßa˙àba onwards, despite the fact that some a˙àdìth apply to both kinds of walà". Whereas walà" al-'itq is more or less a monolithic block, walà" al-muwàlàt has to be divided into walà" with and without conversion. Walà" without conversion is to be divided again in walà" over Arab and non-Arab strangers. Further, there is walà" over a foundling which is in fact different from both servile and contractual walà"; due to the fact that no contract and no manumission is involved. In the following, however, it will suffice to deal with walà" al-'itq on the one hand and walà" al-muwàlàt and walà" al-islàm on the other. Walà" al-'itq was always widely accepted and its abolition was never suggested by anybody. It was always considered to be a kinship tie and, in time, became regarded as a full agnatic relationship which could not easily be dissolved. Contractual walà", however, was rejected by most later jurists and it never became an agnatic relationship. The contractual patron inherits as last cognate and dissolution of walà" al-muwàlàt was considered possible even in classical law as long as no blood-money had been paid. Blood-money plays a much more important role in contractual than in servile walà". The mere payment of blood-money can establish a relationship of patronate and some lawyers considered the payments to be reciprocal, different from walà" al-'itq. On the whole, the freedman is much closer attached to his patron than the contractual client.

the origin and development of the islamic patronate

121

From the outset, the two institutions were different. Walà" al-'itq can come into existence in only one way, that is by manumission of a slave. It arises automatically and only the Màlikìs allow its renunciation by the manumitter. Actually, the relationship between manumitter and freedman is just the modified and, in some respects, loosened tie which already existed between the slave and his master. The former slave gains freedom and rights which he did not have as a slave. Walà" al-'itq has always been an individual relationship between the freedman and the manumitter. Walà" al-muwàlàt is a newly created tie that can be concluded in different ways: by payment of blood-money, by conversion “at the hands” of a Muslim and by mutual agreement. It can be concluded at any time and does not have to be concluded at all. (Except for walà" al-islàm which some scholars thought to have been an automatic consequence of conversion.) Its conclusion limits the freedom of the client. He receives protection and social integration but in exchange, he has to grant his partner the right of succession, apparently without having the right of being the heir of his patron himself. There are semi-collective and individual forms of walà" al-muwàlàt. Whereas walà" al-islàm is always described as an individual tie between the convert and the Muslim at whose hands he converted, in the case of other forms of walà" al-muwàlàt, the client, too, is always described as a stranger, Arab or not, but the patronal party often was supposed to be a “group of people” (qawm). As to the legal status of mawàlì in the first two centuries, it can be stated that according to early ˙adìth, walà" does not create a relationship of equals. A freedman does not inherit from his manumitter and does not pay blood-money for him but the manumitter has that right and that obligation. In the case of a former slave it is just normal that the manumitter is the more active party. But the contractual client is not supposed to inherit from his patron, either. At least, his right of succession is not stated explicitly in the ahàdìth. Payings of blood-money are supposed to be reciprocal only in traditions on walà" without conversion. Since the client mentioned in the traditions is a (converted) stranger and, as such, in a socially weaker position, it comes as no surprise that he does not have equal rights. The unequality between contractual patron and client gives rise to the theory that contractual walà" did not originate in preIslamic ˙ilf but in some other kind of pre-Islamic walà". The fairly detailed picture of walà" in the first and second cen-

122

ulrike mitter

tury presented in this article, basically represents the legal theory. It reflects the different opinions of the lawyers and the development of their juridical regulations. When we look for information on mawàlì in the army and on their other social functions, not much is to be found in the a˙àdìth. Do we have to conclude, therefore, that the legal theory has nothing to do with legal practice? Do the a˙àdìth not tell us anything about the situation of mawàlì in real life, about the real rights and duties of patrons and clients? It is true that a˙àdìth never reflect the whole reality. Obviously, many aspects are not touched upon. But in the light of the results of the ˙adìth analysis, it can be suggested that the traditions which go back to the generation of the ßa˙àba do reflect (a part of ) the legal practice. This conclusion is based on the finding that some traditions (mainly transmitted in the form of anecdotes), go back to the time of the ßa˙àba and probably preserve authentic decisions of the ßa˙àba in historic cases. Summing up the information of these traditions, we can state that early in the first century, there were male and female manumitters who inherited from their freedmen in the absence of certain heirs and who paid blood-money for them. Some manumitters sold the rights of walà" and gave them away, others renounced walà" which enabled the freedman to choose another patron. After the death of the manumitter, walà" was bequeathed to his heirs according to the kubr rule or in the same way as property. Arab and non-Arab strangers came to live with tribes, some of them concluding walà" contracts, others remaining without a patron. Payments of blood-money enabled the payer to receive the inheritance of the stranger. Such was the legal practice before the turn of the century. The generation of the tàbi'un, however, started systematizing Islamic law by discussing fictitious cases. Therefore, the a˙àdìth that go back to these jurists cannot be taken as an indication of the practice. To give an example: despite the prohibition of selling of walà" by the lawyers, selling walà" might well have endured in practice. Unlike judges and caliphs, the jurists were not very interested in legal practice and changing living conditions, although their decisions must have been influenced to a certain degree by the social circumstances.218

218 The prohibition of selling of walà" by the lawyers could have been a reaction to the growing significance of the mawàlì who were no longer willing to be treated like slaves.

the origin and development of the islamic patronate

123

They were concerned above all with the religious and moral side of law and continued to concentrate on the judgments of the Prophet and his Companions, interpreting and re-interpreting the traditions and applying them to every possible situation.219 The judges, on the other hand, were forced to deal with reality and had to find compromises that satisfied people and did not contradict the spirit of Islam. Typically, the traditions which are ascribed to judges show a tendency to practical solutions.220 To conclude, a˙àdìth that go back roughly to the middle of the first century reflect the legal practice, if selectively. From the end of the first century onwards, jurisprudence became more and more theoretical and the famous gap between legal theory and legal practice started to open up.221

219 The Medinese lawyer Ibn Shihàb al-Zuhrì (d. 124/742) is an example for such a systematic thinker. Arguing on the basis of the a˙àdìth, he extracted their legal essence and applied it to analogous cases (see above and Motzki, “Fiqh”). Of course, there are many aspects of law which are not covered by a˙àdìth and Qur"ànic rules and which had to be elaborated by the lawyers anyway. In the field of tax law, this work has been done by Abù Yùsuf (d. 182/798) in his Kitàb al-Kharàj. (Surely, it is no coincidence that Abù Yùsuf also was a judge!) From the third/ninth century onwards, the lawyers reacted to the changing needs of the society by developing legal devices (˙iyal ) and the science of documents (shurù†) in order to handle given facts of commercial law which was “ruled by customary law” (Schacht, Introduction, 78 and 82). If in Islamic law we look for the influence of non-Arab legal systems, we may find it mainly in the fields of commercial law and taxation as well as in the fields of criminal law and martial law. 220 The Basran judge Hishàm b. Hubayra (d. 75/694) decided that a woman’s giving away of the walà" of her mawlà to her husband was valid but that the walà" would return to the agnates of the woman after her husband’s death (Ibn Abì Shayba, Mußannaf, XI, no. 11667). So in this case, the judge permitted the giving of walà" but at the same time he prevented the change of patron to be final. For the difference in approach to a legal problem by judges and jurists, see also Motzki, “Der Prophet und die Schuldner. Eine ÓadìΔ-Untersuchung auf dem Prüfstand,” Der Islam 77/I (2000), 76–77. 221 Schacht devoted a whole chapter of his Introduction, 76–85, to the relation between theory and practice. See also id., “Fiqh,” 890; Coulson, “Islamic Law, Sharì'a (part of “The Culture of Islam),” The New Encyclopaedia Britannica (Chicago, 198715), XXII, 32; Heffening, Das islamische Fremdenrecht bis zu den islamisch-fränkischen Staatsverträgen. Eine rechtshistorische Studie zum Fiqh (Hannover: Heinz Lafaire, 1925), 4; Schmidt “Occupatio,” 300–302; Becker, “Die Entstehung von 'U“r- und Óaràland,” in Carl Heinrich Becker, ed., Islamstudien, 2 vols. (Leipzig, 1924), 218–19.

124

ulrike mitter Appendix: Analysis of the Barìra ˙adìth

The Barìra tradition is a very famous Prophetic ˙adìth which occurs very often in the ˙adìth collections. There are versions ascribed to the tàbi'ùn 'Urwa b. al-Zubayr (d. 99/717–718), 'Amra bint 'Abd alRa˙màn (d. 98/716–717), al-Qàsim b. Mu˙ammad (d. 106/724–725), al-Aswad (d. 73/692–693), 'Amr b. Shu'ayb (d. 118/736), Ibn Abì Mulayka (d. 118/736), Abù Salama (d. 94 or 104/712 or 723), alZuhrì (d. 124/742) and Ayman (d. about 100/718–719) and to the ßa˙àba Abù Hurayra (d. 59/678) and 'Abdallàh b. 'Umar (d. 73/693). Except for al-Aswad, who is a Kùfì, all of them are Óijàzì scholars. The oldest common link of the Barìra tradition is 'À"isha who reports on an event involving the Prophet, herself and the slave-girl Barìra. The most prominent version of the Barìra tradition is the one of 'Urwa b. al-Zubayr which, in the following, will be analyzed in detail. Four Medinese transmitters allege to have the tradition from 'Urwa: His son Hishàm (d. 145 or 146/762–763), Ibn Shihàb alZuhrì (d. 124/742)—both are partial common links—Abu l-Zubayr (d. between 126–132/743–750) and Yazìd b. Rùmàn (d. 130/747–748). Let us have a look at the version of Hishàm first. Hishàm’s version As an example of his version, I will present the matn in translation, which is transmitted by Ibn Jurayj and recorded by 'Abd al-Razzàq.222 ['À"isha] said: (1) Barìra came [to me] and said: (2) “I have the emancipation contract from my owners (kàtabtu ahlì ) on condition of [the payment] of nine ounces (awàq), every year an ounce. (3) Help me [with it].” (4) 'À"isha said: “If your owners want me to pay them [the amount] at one time and your walà" belongs to me, then I will do it.” (5) [Barìra] went to her owners but they refused [this offer]. (6) When she came back from her owners, the Messenger of God (eulogy) was sitting [next to 'À"isha]. [Barìra] said [to 'À"isha]: “I proposed it to them but they refused, unless the walà" belongs to them.” (7) The Messenger of God (eulogy) heard this and asked 222 'Abd al-Razzàq, Mußannaf, IX, no. 16164. The numbers in brackets indicate narrative elements.

the origin and development of the islamic patronate

125

[about it]. (8) ['À"isha] told him [what had happened]. (9) He said: “Take [Barìra] (10) and grant them the stipulation that the walà" belongs to them (wa-shtari†ì lahum al-walà").223 (11) The walà" belongs to the manumitter [anyway] (innamà l-walà" li-man a'taqa).” (12) She did so. (13) The Prophet (eulogy) rose among the people preaching. (14) He praised God and lauded him and said: (15) “Now then, (16) what is in the minds of people who make stipulations which are not in the book of God the Almighty? They are void even if they were one hundred conditions. (17) God’s decree has more authority and his condition is stronger.” Alongside Ibn Jurayj (Mecca, d. 150/767), four other scholars transmit Hishàm’s text: Màlik b. Anas (Medina, d. 179/795), Abù Usàma (Kùfa, d. 201/816–817), Jarìr [b. 'Abd al-Óamìd] (Kùfa, d. 188/804) and Wakì' (Kùfa, d. 196 or 198/811 or 814). Their versions are slightly different from that of Ibn Jurayj. Màlik is a partial common link of three transmission strands. There are only a few trivial differences between the three versions.224 Some of the differences between Ibn Jurayj’s text and Màlik’s reconstructed text are the following (differences in italics): Ibn Jurayj:

Màlik:

(4) 'À"isha said: If your owners want me to pay them [the amount] at one time . . .

'À"isha said: If your owners want me to pay them [the amount]

(5) [Barìra] went to her owners [but] they refused.

[Barìra] went to her owners and told them [but] they refused.

(13) The Prophet rose among the people preaching

The Messenger of God rose among the people

(16) What is in the minds of people

What is in the minds of men

—

(18) Walà" belongs to the manumitter [repetition of the maxim]

223 Mu˙ammad’s advice to 'À"isha was a problem for later lawyers because it sheds bad light on the Prophet. Cf. the commentary of al-Nawawì in Muslim, alJàmi' al-Ía˙ì˙ (Cairo, 1963), V, 402; and Schacht, Origins, 173. I do not agree, however, with Schacht’s conclusion that later lawyers mitigated this critical point in the shorter versions which are ascribed to 'Amra and Ibn 'Umar. According to my analysis, the versions of 'Amra and Ibn 'Umar are not younger than the version of Hishàm. Furthermore, the Prophet’s advice does not really appear in a more favorable light in their versions. 224 Al-nabì instead of rasùl Allàh, addition or omission of 'À"isha or zawj al-nabì.

126

ulrike mitter

The version of the common link Abù Usàma has peculiarities (in italics) which cannot be found in other versions of Hishàm’s text.225 – 'À"isha’s proposal to Barìra (4) contains an element of manumission not found elsewhere: “If your owners want me to pay them [the amount] at one time, and I emancipate you and your walà" belongs to me, then I will do this.” – Only in the text of Abù Usàma, does 'À"isha speak in the first person when saying: (7) The Messenger of God (eulogy) heard this and asked me [about it]. (8) I told him [what happened].” In all other Hishàm texts this is reported about 'À"isha in the third person. – In Mu˙ammad’s statement (9), an element of manumission appears which is not existent in all other texts: “Buy226 [Barìra] and emancipate her.” – After Mu˙ammad’s speech (13–17) and before the repetition of the walà" maxim (18), both versions of Abù Usàma have a special repetition of Mu˙ammad’s rhetorical question: What is in the mind of a man of you who would say [to another man]: “Emancipate soand-so and the walà" belongs to me”? The versions of the two remaining transmitters Jarìr227 and Wakì'228 are useless for dating and reconstructing Hishàm’s text. Jarìr’s text is incomplete. The direct transmission of Wakì' from Hishàm must be doubted because Wakì' is also mentioned as a transmitter of Jarìr’s text. As has been shown, the mutùn of Hishàm which are transmitted by Ibn Jurayj, Màlik and Abù Usàma are different from each other but, at the same time they have many elements in common. The differences point to an independent transmission of the tradition from the same source: Hishàm b. 'Urwa. On the whole, the reconstructed

The extract given below consists of the elements which appear in two or more versions. For Màlik’s text see Muwa††a", 'itq wa-walà" (38), 10, no. 17. 225 Two scholars transmit the text from Abù Usàma. They are Abù Kurayb (Muslim, Ía˙ì˙, 'itq (20), 2, no. 8) and 'Ubayd b. Ismà'ìl (al-Bukhàrì, al-Djàmi' alÍa˙ì˙ (Leiden, 1862–1908), (50), no. 3). Their versions differ much from each other. Nevertheless, there are significant similarities. 226 In all other versions of Hishàm’s text it is not “buy her,” but “take her.” 227 Muslim, Ía˙ì˙, V, 399, no. 9. 228 Ibn Óanbal, Musnad, VI, no. 25840.

the origin and development of the islamic patronate

127

Hishàm text tallies with the version of Ibn Jurayj given above plus two additional elements which are missing only in Ibn Jurayj’s text: first, the explicit statement that Barìra let her owners know about 'À"isha’s suggestion (addition to 5) and, second, the repetition of the maxim at the end of the speech (18). al-Zuhrì’s version Two partial common links transmit the tradition from al-Zuhrì: Ma'mar (Baßra, d. 179/795) and Yùnus [b. Yazìd] (Medina, d. 159/ 775–776). Furthermore, there are the versions of Shu'ayb [b. Abì Khamza] (Óumß, d. 162/778–779), al-Layth (Egypt, d. 175/791), 'Uqayl (Medina, d. between 141–144/758–762) and al-Zuhrì’s nephew. As starting point for the analysis of al-Zuhrì’s version serves, again, a text which is recorded in the pre-canonical collection of 'Abd alRazzàq.229 The authority of 'Abd al-Razzàq is Ma'mar. ['À"isha]230 said: (1) Barìra came to 'À"isha (3) asking her for help (2) with her kitàba. (4) 'À"isha said: “What do you think, if I would pay them at one time what they request, would they sell you, so that I could emancipate you?” ['À"isha] said: “[I encouraged her] to ask them (˙attà tas"alahum).” (5) [Barìra] went and asked them. (6) [Her owners] said: “Yes, if the walà" belongs to us.” (8) The Messenger of God (eulogy) came [to see 'À"isha] and she told him what had happened. (9) He said: “Buy her and emancipate her. (11) The walà" belongs to the manumitter [anyway].” (12) So, she bought and emancipated [Barìra]. (13) ['À"isha] said: Then the Prophet (eulogy) rose preaching and said: (16) “What is in the minds of people who make stipulations which are not in the book of God the Almighty? If somebody makes a stipulation which is not in the book of God, this stipulation is void even if he stipulated it one hundred times. (17) God’s decree has more authority and his condition is stronger.” When reconstructing al-Zuhrì’s 'Urwa text in the same way as Hishàm’s text and then comparing it with the latter, it becomes 'Abd al-Razzàq, Mußannaf, IX, no. 16161. Although 'À"isha allegedly transmitted the tradition, she is mentioned in the matn in the third person. This inconsistency between isnàd and matn is typical for 'Urwa’s version of the Barìra tradition. 229 230

128

ulrike mitter

obvious that the texts of al-Zuhrì and Hishàm have many elements in common. The sequence of events is more or less the same in both versions and some phrases are formulated in a similar way, above all the words of Mu˙ammad. On the other hand, we can distinguish without difficulty between a Zuhrì text and a Hishàm text although there are different versions of their texts. Without looking at the isnàd, one can tell which matn belongs to Hishàm and which belongs to al-Zuhrì. – Hishàms text starts with the words: (1) Barìra came [to me] and said: (2) “I have the emancipation contract from my owners on condition of [the payment] of nine ounces, every year an ounce. (3) Help me [with it].” A Zuhrì text starts with a shorter sequence: “(1) Barìra came to 'À"isha (3) asking her for help (2) with her kitàba.” – In Hishàm’s version, Barìra plays an important role. She explains in detail the problem of her emancipation contract (2, 3) and reports on the meeting with her owners (6). In al-Zuhrì’s version, it is the owners of Barìra whose words are cited in direct speech. – The elements 7 and 10 and 14 and 15 are missing in al-Zuhrì’s version which means that Mu˙ammad has a smaller speaking part. – On the whole, the versions of al-Zuhrì’s text are formulated much more uniformly than the versions of Hishàm’s text. Abù l-Zubayr’s version The matn of Abù l-Zubayr, the third transmitter of 'Urwa’s tradition, is a short version and very different from the versions of Hishàm and al-Zuhrì. It is only ascribed to 'À"isha, not to Mu˙ammad. She [Barìra] belonged to people of the Banù 'Àmir b. Ía'ßa'a. They granted her the emancipation contract on condition of [payment of ] nine ounces. Then they sold her to 'À"isha, while her emancipation contract (mukàtaba) was as it was and she had not paid anything of her kitàba.231

231 'Abd al-Razzàq, Mußannaf, VIII, no. 15783 and IX, no. 16162 (a shorter version with the same isnàd ).

the origin and development of the islamic patronate

129

Abù l-Zubayr is particularly interested in the kitàba which could be due to the fact that in his time, selling a mukàtab(a) became problematic. He stresses the fact that Barìra had not yet paid anything of her kitàba installments (which amounts to still being a slave who could be sold without problems). Since Abù l-Zubayr is the only one who transmits this detailed statement on the kitàba, we cannot know if it stems from 'Urwa. Abù l-Zubayr could have asked 'Urwa especially about this problem or he could have thought himself that this must have been the most logical explication and, therefore, added it to the text.232 Despite the very different aspect of his text, there are elements which Abù l-Zubayr’s tradition has in common with Hishàm’s and/or al-Zuhrì’s tradition: the selling of Barìra to 'À"isha and the mentioning of Barìra being a mukàtaba, including the amount of the kitàba, nine ounces. Yazìd b. Rùmàn’s version The matn of Yazìd b. Rùmàn233 is falsely ascribed to 'Urwa. It only consists of 'À"isha’s statement: “The husband of Barìra was a slave,” which is an alien element in 'Urwa’s version of the Barìra tradition. It is part of the versions of al-Qàsim b. Mu˙ammad and al-Aswad.234 Perhaps, the Baßran transmitter Wuhayb (d. 165 or 169/781–786) is responsible for the matn and its attribution to the Medinese scholars 'Ubaydallàh b. 'Umar—Yazìd—'Urwa—'À"isha. Obviously, the discussion about the status of Barìra’s husband, whether he was a slave or a free person, took place in Iraq and not in Medina. Conclusion Due to the differences between the text of Hishàm, al-Zuhrì and Abù l-Zubayr, it can be stated that the three scholars transmitted 232 The kitàba element is not likely to go back to 'À"isha, anyway, because in her time, selling a mukàtab was perfectly normal and there would have been no reason to emphasize it. 233 Muslim, Ía˙ì˙, 'itq, 2, no. 13. 234 Al-Qàsim b. Mu˙ammad (Ibn Óanbal, Musnad, VI, no. 24242) and al-Aswad (ibid., no. 25588) combine three “sunnas” connected with Barìra. 1. Our Barìra tradition which culminates in Mu˙ammad’s walà" maxim. 2. Mu˙ammad gives Barìra the choice to stay with her husband after her emancipation or to leave him. 3. Mu˙ammad approves of Barìra’s giving some meat, which she received as ßadaqa, to 'À"isha.

130

ulrike mitter

the tradition independently from each other from 'Urwa. The common elements of two or three versions constitute the original 'Urwa text. This text cannot be reconstructed literally but it is clear that 'Urwa spread a version of the Barìra tradition which contained the following elements: (1) Barìra goes to see 'À"isha; (2) specification of the kitàba; (3) Barìra asks 'À"isha for help; (4) 'À"isha’s advice; (5) Barìra’s proposal to her owners; (6) condition made by her owners; (7) 'À"isha’s report to the Prophet; (9) statement of Mu˙ammad: “buy her”; (11) first walà" maxim; (16) rhetorical question of Mu˙ammad; (17) his statement on the condition of God. As has been said, 'Urwa is not the only person who is said to have transmitted the Barìra tradition from 'À"isha but he has the longest version with most dialogues and characters. When analyzing the texts of his contemporaries in the same way, starting always with the youngest common link, it becomes clear that the texts of four scholars cannot be regarded as having been transmitted independently from 'Urwa’s text,235 whereas the texts of six scholars can.236 They all have different combinations of the elements which appear in the 'Urwa tradition. Forgery can be excluded as an explanation for the existence of these versions because it is not likely that six scholars should have taken a text of 'Urwa, omitted his name in the isnàd, and decided to shorten (why never extend?) the matn, most coincidently always in a different way, sometimes changing the wording, sometimes maintaining the original formulation. A more probable explanation for the differences of the mutùn is: The transmitters heard 235 1. The tradition which is ascribed to Abù Óurayra. Its matn is a copy of Màlik’s Nàfi'—Ibn 'Umar text (Muslim, Ía˙ì˙, V, 401, no. 15). 2. The tradition of Ayman which is quite enigmatic (al-Bukhàrì, Ía˙ì˙, mukàtab, no. 5; and ibid., shurù†, no. 10). It is as long as 'Urwa’s text and includes elements of Hishàm’s and alZuhrì’s 'Urwa version but also differences from both. At first sight, it seems to be an independent transmission but there are elements which are very typical for 'Urwa and there is a strange frame story. 3. The tradition of al-Zuhrì which only contains the walà" maxim of Mu˙ammad. Al-Zuhrì also transmits the full version from 'Urwa, as we have seen, and uses the maxim for his legal argumentation ('Abd alRazzàq, Mußannaf, VIII, no. 15784). 4. The tradition of 'Amr b. Shu'ayb, whose matn, too, consists of the maxim only. Ibn Shu'ayb does not name 'À"isha in the isnàd and probably transmits the maxim from one of the versions current in the Óijàz (ibid., IX, no. 16165). 236 'Amra bint 'Abd al-Ra˙màn (Màlik, Muwa††a" (10), 38, no. 19), 'Abdallàh b. 'Umar (ibid., no. 18), al-Qàsim b. Mu˙ammad (Ibn Óanbal, Musnad, VI, no. 24242), al-Aswad (al-Bukhàrì, Ía˙ì˙, farà"i∂, no. 19), Ibn Abì Mulayka ('Abd al-Razzàq, Mußannaf, IX, no. 16163) and Abù Salama (Ibn Óanbal, Musnad, VI, no. 24776).

the origin and development of the islamic patronate

131

or remembered different versions of the Barìra tradition from their common authority 'À"isha. We do not know for sure whether 'À"isha, who spread the tradition, invented the whole story and put the famous words in the mouth of Mu˙ammad or whether she reported on a real event. But what we have here is an example of a very early tradition ascribed to the Prophet. It existed before 58/678, the year in which 'À"isha died.

Bibliography Primary Sources 'Abd al-Razzàq al-Ían'ànì. Al-Mußannaf, ed. Óabìb al-Ra˙màn al-A'Ωamì, 11 vols. Beirut, 1430/1983; 1st ed. 1391/1972. Abù Dàwùd, Abù Bakr 'Abdallàh b. Abì Dàwùd Sulaymàn b. al-Ash'ath al-Sijistànì. Sunan, ed. Mu˙yì l-Dìn 'Abd al-Óamìd. Beirut: Dàr al-Kutub al-'Ilmiyya, n.d. al-Bayhaqì, Abù Bakr A˙mad b. al-Óusayn. Al-Sunan al-kubrà. 10 vols. Hyderabad, 1344–55 AH. al-Bukhàrì, Abù 'Abdallàh Mu˙ammad b. Ismà'ìl. Al-Djàmi' al-Ía˙ì˙. Leiden 1862–1908. al-Dàrimì, 'Abdallàh b. 'Abd al-Ra˙màn. Sunan, ed. Fawàz A˙mad Zamalì and Khàlid al-Saba' al-'Alamì, 2 vols. Beirut, 1407/1989. Ibn Abì Shayba. Al-Mußannaf, ed. Mu˙ammad A˙mad al-Nadwì. 15 vols. Bombay 1973–83. Ibn Óajar al-'Asqalànì, Abù l-Fa∂l A˙mad b. 'Alì. Tahdhìb at-tahdhìb al-kamàl fì ma'rifat ar-Rijàl. 12 vols. Beirut: Dàr Íàdir, 1968; repr. Hyderabad 1325/1907. Ibn Óanbal, A˙mad b. Mu˙ammad. Musnad al-imàm A˙mad. Beirut 1413/1993. Ibn Màja, Mu˙ammad b. Yazìd al-Qazwìnì. Sunan, ed. Mu˙ammad Fu"àd 'Abd al-Bàqì, 2 vols. Beirut: Dàr al-Kutub al-'Ilmiyya, n.d.; repr. 1373/19541. Ibn Qudàma, 'Abdallàh b. A˙mad. Al-Mughnì. 10 vols. Cairo 1968–69. Ibn Rushd/Nyazee: see Nyazee Màlik b. Anas. Al-Muwa††a", riwàyàt Ya˙yà b. Ya˙yà al-Masmùdì, ed. Mu˙ammad Fu"àd 'Abd al-Baqì. Cairo, n.d. al-Marghìnànì, 'Alì b. Abì Bakr. Al-Hidàya. 4 vols. Cairo, 1980. Muslim b. al-Óajjàj. Al-Jàmi' al-Ía˙ì˙. Cairo, 1963. Nyazee, Imran Ahsan Khan trans. Ibn Rushd. The distinguished jurist’s primer. Bidàyat al- Mujtahid. 2 vols. Reading (UK): Garnet Publishing Limited, 1996. Sa˙nùn, 'Abd al-Salàm b. Sa'ìd al-Tanùkhì. Al-Mudawwana al-kubrà. 5 vols. Beirut, 1410/1994. Sa'ìd b. Manßùr. Al-Sunan, ed. Óabìb al-Ra˙màn al-A'Ωamì, vol. 3/1, 2 in 1. Bombay, 1982/1403. al-Sarakhsì, Shams al-A˚imma. Al-Mabsù†. 30 parts in 15 vols. Cairo, 1324–31/1906–13. al-Shàfi'ì, Mu˙ammad b. Idrìs. Al-Umm. 8 vols. in 4. Beirut, n.d. al-Shaybànì, Abù 'Abdallàh Mu˙ammad b. al-Óasan. Kitàb al-Aßl al-ma'rùf bi-l-mabsù†. 5 vols. Karachi: Idàrat al-Qur"àn wa-l-'Ulùm al-Islàmiyya, n.d. al-Shaybànì, Abù 'Abdallàh Mu˙ammad b. al-Óasan. Al-Jàmi' al-ßaghìr. Karachi 1407/1987. al-Tirmidhì, Mu˙ammad b. 'Ìsà. Sunan (al-Jàmi' al-Ía˙ì˙), ed. A˙mad Mu˙ammad Shàkir, Mu˙ammad Fu"àd 'Abdalbàqì and Ibràhìm 'Atwa 'Iwa∂, 5 vols. n.p., 1382–98/1962–87.

132

ulrike mitter

Secondary Sources Becker, Carl Heinrich. “Die Entstehung von 'U“r- und Óaràland,” in id., Islamstudien. 2 vols. Leipzig, 1924; repr. Hildesheim: Georg Olms, 1967, I, 218–33. Bergsträsser, Gotthelf. “Anfänge und Charakter des juristischen Denkens im Islam. Vorläufige Betrachtungen,” Der Islam 14 (1925), 76–81. Brockopp, Jonathan E. Early Màlikì Law. Ibn 'Abd al-Óakam and his Major Compendium of Jurisprudence, Studies in Islamic Law and Society. Leiden/Bosten/Cologne: Brill, 2000. Coulson, Noel James. “Islamic Law, Sharì'a (part of “The Culture of Islam”),” The New Encyclopaedia Britannica. Chicago, 198715, XXII, 31–35. id., Succession in the Muslim Family. Cambridge: Cambridge University Press, 1971. Crone, Patricia. Roman, provincial and Islamic law. The origins of the Islamic patronate. Cambridge: Cambridge University Press, 1987. id., Slaves on Horses, the Evolution of the Islamic Polity. Cambridge: Cambridge University Press, 1980. id., “Two Legal Problems Bearing on the Early History of the Qur"àn,” Jerusalem Studies in Arabic and Islam 18 (1994), 1–37. Falaturi, Abdoljavad. “Der Koran: Zeugnis der Geschichte seiner Zeit,” in Albrecht Noth and Jürgen Paul, eds., Der islamische Orient. Grundzüge seiner Geschichte. Würzburg: Ergon, 1998. Forand, Paul G. “The Relation of the Slave and the Client to the Master or Patron in Medieval Islam,” International Journal of Middle East Studies 2 (1971), 59–66. Goldziher, Ignaz. Muhammedanische Studien. 2 vols. Halle: Max Niemeyer, 1889–90. Hallaq, Wael B. “The Use an Abuse of Evidence: The Question of Provincial and Roman Influence on Early Islamic Law,” Journal of American Oriental Society 110/I (1990), 79–91, New York, 1989, 1–36; repr. in id., Law and Legal Theory in Classical and Medieval Islam. Aldershot/Great Britain and Brookfield/USA: Variorum, 1995, article IX. Hausmaninger, Herbert and Walter Selb. Römisches Privatrecht. Vienna/Cologne: Böhlau Verlag, 1981. Heffening, Willi. Das islamische Fremdenrecht bis zu den islamisch-fränkischen Staatsverträgen. Eine rechtshistorische Studie zum Fiqh. Hannover: Heinz Lafaire, 1925. Jarrar, Maher. “Sìrat Ahl al-Kisà". Early Shi'ì Sources on the Biography of the Prophet,” in Harald Motzki, ed., The Biography of Mu˙ammad: The Issue of the Sources, Islamic History and Civilization. Studies and Texts. Leiden/Boston/Cologne: Brill: 2000, 86–156. Juda, Jamal. Die sozialen und wirtschaftlichen Aspekte der Mawàlì in frühislamischer Zeit. Tübingen: REFO-Druck Hans Vogler, 1983. Juynboll, Gautier H.A. “Early Islamic society as reflected in it’s use of isnàds,” Le Muséon 107 (1994), 151–94. id., Muslim Tradition. Studies in Chronology, Provenance and Authorship of Early ˙adìth. Cambridge: Cambridge University Press, 1983. id., “Nàfi', the mawlà of Ibn 'Umar, and his position in Muslim ˙adìth Literature,” Der Islam 70 (1993), 207–44. Juynboll, Th. W. Handleiding tot de kennis van de Mohammedaansche wet, volgens de leer der Sjàfi'itische school. Leiden: Brill, 1930; 1st ed. 1903. Kaegi, Walter E. Byzantium and the early Islamic conquests. Cambridge: Cambridge University Press, 1992; repr. 1995. Kremer, Alfred von. Culturgeschichte des Orients unter den Chalifen. 2 vols. Vienna: Wilhelm Braumüller, 1875–77. Lane, Edward William. Arabic-English Lexicon. 8 vols. New York, 1863–93. Mitter, Ulrike. Das frühislamische Patronat. Eine Untersuchung zur Rolle von fremden Elementen bei der Entwicklung des islamischen Rechts. Ph.D. diss., Univ. Nijmegen, 1999.

the origin and development of the islamic patronate

133

id., “Problemen van het onderzoek naar ontleningen aan niet-Arabische rechtsstelsels in het ontstaan en de ontwikkeling van het islamitisch recht,” Sharqiyyàt 9/II (1997), 107–23. id., “Unconditional manumission of slaves in early Islamic law: a ˙adìth analysis,” Der Islam 78/I (2001), 35–72. Motzki, Harald. Die Anfänge der islamischen Jurisprudenz. Ihre Entwicklung in Mekka bis zur Mitte des 2./8. Jahrhunderts. Stuttgart: Franz Steiner, 1991. id., The Biography of Mu˙ammad (vgl. Anm. 2). id., “The Collection of the Qur"àn. A Reconsideration of Western Views in Light of Recent Methodological Developments,” Der Islam 78/I (2001), 1–34. id., “Der Fiqh des-Zuhrì: die Quellenproblematik,” Der Islam 68 (1991), 1–44. id., “The murder of Ibn Abì l-Óuqayq: On the Origin and Reliability of some Maghàzì-Reports,” in id., ed., The Biography of Mu˙ammad: The issue of the sources. Leiden: Brill, 1999. id., “Patricia Crone: Roman, Provincial and Islamic Law,” Der Islam 65/II (1988), 342–45. id., “The Prophet and the Cat. On dating Màlik’s Muwa††a" and Legal Traditions,” Jerusalem Studies in Arabic and Islam 22 (1998), 18–83. id., “Der Prophet und die Schuldner. Eine ÓadìΔ-Untersuchung auf dem Prüfstand,” Der Islam 77/I (2000), 1–83. id., “Quo vadis, ˙adìΔ-Forschung? Eine kritische Untersuchung von G.H.A. Juynboll: ‘Nàfi', the mawlà of Ibn 'Umar, and his position in Muslim ˙adìΔ literature’,” Der Islam 73 (1996), 40–80 and 192–231. Noth, Albrecht. “Die Scharìa, das religiöse Gesetz des Islam—Wandlungsmöglichkeiten, Anwendung und Wirkung,” in Wolfgang Fikentscher, Herbert Franke and Oskar Köhler, eds., Entstehung und Wandel rechtlicher Traditionen. Freiburg/Munich: Karl Alber, 1980, 415–37. Pipes, Daniel. Slave Soldiers and Islam. The genesis of a Military System. New Haven/London: Yale University Press, 1981. Powers, David S. Studies in Qur"an and Óadìth: The Formation of the Islamic Law of Inheritance. Berkely/Los Angeles/London: University of California Press, 1986. Reinert, Werner. Das Recht in der altarabischen Poesie. Ph.D. diss., Cologne, 1963. Schacht, Joseph. “Foreign Elements in Ancient Islamic Law,” Journal of Comparative Legislation and International Law 32 (1950), 9–17. id., An Introduction to Islamic Law. Oxford: The Clarendon Press, 1964. id., The Origins of Muhammadan Jurisprudence. Oxford: The Clarendon Press, 1950. Schoeler, Gregor. Charakter und Authentie der muslimischen Überlieferung über das Leben Mohammeds. Berlin/New York: Walter de Gruyter, 1996. Smith, W. Robertson. Kinship and Marriage in Early Arabia. Cambridge, 1885; New edition of Stanley A. Cook, London, 1907; repr. London, 1990. Tyan, Emile. “Óilf,” EI 2, III, Leiden/London, 1971, 388–89.

PATRONATE IN EARLY SHÌ'ITE LAW1 Robert Gleave

Introduction In the Twelver Shì'ì (or Imàmì) tradition, a bundle of terms have been used to describe the believer’s relationship to the imàm. Included in these are the cognate terms wilàya (or walàya), walì and walà".2 The strictly legal employment of these terms was, then, loaded with a theological resonance which would not be lost on the reader. The terms were often used in works of fiqh to indicate a relationship between legal institutions which was analogous to the central devotional relationship between the Shì'a and their imàm. The doctrine of wilàyat al-faqìh, as used in later Imàmì fiqh and developed by Rù˙ Allàh al-Khumaynì is perhaps the most pertinent example, though similar discussions occurred under the rubric of marriage law and endowment law with regard to walì.3 This article concerns the Shì'ite discussions of the patronate tie (termed walà" ) which is created between the master and his former slave on manumission. The issues which most taxed Imàmì jurists revolved around models for the unusual patron-client/freedman relationship. Was it to be viewed as a type of kinship tie (nasab) with all the legal consequences of such an analogy, or was it analogous to the relationship between the imàm and his community of followers, or (under certain circumstances) was an exchange contract the most appropriate paradigm with which to discuss walà"? In Imàmì Shì'ite law, as in Sunnì law, the patronate ties between freedman and former master (or client and patron) are viewed as 1 I would like to thank the participants at the colloquium for their helpful suggestions on an earlier draft of this article. In particular Ulrike Mitter, both during the colloquium and afterwards, commented on the article’s content in a most constructive manner. 2 On which see Corbin, En Islam Iranien, 4 vols. (Paris: Buchet and Chastel, 1971–72), III, 9–10. 3 For a general discussion of the enduring power and use of the terms the wilàya and walì, see Sachedina, The Just Ruler in Shi'ite Islam (Cambridge: Cambridge University Press, 1988), especially 173–231.

patronate in early shì'ite law

135

having a primary analogy with the ties of kinship. The Prophetic ˙adìth cited in Sunnì works of fiqh, “al-walà" lu˙ma kalu˙mat al-nasab” (patronate is a tie, like the tie of kinship) is commonly cited (though through a different isnàd which includes the imàms Ja'far and Bàqir). Also cited is the Prophetic ˙adìth (with an isnàd which includes imàm Ja'far): “al-walà" li-man a'taqa” (the patronate belongs to the manumitter), though this general rule is restricted by subsequent reports from the imàms. Apart from the legal matters, dealt with below, there is an underlying theological emphasis. A believer’s most important patronate tie is to the imàm, be he hidden or manifest. The patronate tie arising from manumission (or, as we shall see, from a contract) is the introduction of an intervening layer of patronate between the individual and the imàm. The patron takes on certain duties in relation to the client, and, in a sense, replaces the imàm in doing so. A person with no (created) patronate tie (technically termed sà"iba in Imàmì fiqh works) has, in its place, a patronate tie to the imàm. In other respects, the patron is conceived of as a protector, who takes responsibility for the transgressions of the individual in a manner not dissimilar to the imàm’s intercession for the individual with God. This protective role is reflected legally in the patron becoming the blood-money group ('àqila) of the client. The tie between them is likened to kinship in that the patron inherits from the client on the latter’s death. If however a freedman (or indeed any person) has neither kin nor patron, then he is not without a protector. The imàm acts as security for all such unfortunate persons. Imàm Bàqir is reported as having said of the sà"iba, “Whatever walà" he has to God, he has to his Prophet. Whatever walà" he has to [the Prophet], he has to the imàm. His delicts are the responsibility of the imàm, and his inheritance goes to him also.”4 Reports such as these emphasize the derivative nature of human walà" ties; they are ultimately human replicas of the link between the Shì'a and the imàm.5 On other occasions, a walà" tie can be seen as a contract between two independent legal agents, comprising of a promise to pay each

4 Al-Kulaynì, Al-Kàfì fì 'ilm al-dìn, 8 vols. (Tehran: Dàr al-Kutub al-Islàmiyya, 1367 AH), VII, 171; Ibn Bàbùya, Al-Muqni', in: Al-Jawàmi' al-fiqhiyya (Qum: Kitàbkhàna-yi Àyatallàh Mar'ashì, 1404 AH), III, 136. 5 On the guide/protector role of the imàm in early Imàmì thought, see AmirMoezzi, The Divine Guide in Early Shi'ism (New York: Suny, 1994).

136

robert gleave

other’s delicts, on the condition that they also inherit from each other. The non-contractual walà" also shares (in part) this characteristic, as the benefit of gaining inheritance from a former slave is offset against the burden of paying compensation for his delicts. These analogies give one some idea of the quasi-sacred nature of the patronate relationship. It stands in place of the practical manifestations of devotion to the imàm, and in legal works, this quasisacred character is expressed through legal stipulations such as the tenacity of the walà" tie (it can never be obliterated by the client, though it can be rendered legally irrelevant) and the restrictions present in Imàmì fiqh (but lifted in other legal schools) on the occasions on which the tie can be formed. However, the obvious Shì'ì emphasis on kinship ties (linked as they are to the primacy of the family of the Prophet) tempers over enthusiasm for the patronate tie. In what follows, I concentrate on those areas of the patronate in which the Imàmì theory is distinctive, examining how the final (orthodox) doctrine in these areas emerged out of early intra-Imàmì (and inter-Imàmì-Sunnì) dispute.6

A case of Shì'ite walà" The following case illustrates the distinctive features of the Imàmì conception of walà": Al-Óasan b. Ma˙bùb relates from Abù Ayyùb, from Burayd al-'Ijlì, who said: I asked Abù Ja'far [imàm al-Bàqir], “A man upon whom there was [a duty] to free a slave, dies before he frees the slave. His son inherits and then buys a slave from his own money7 and frees him for his father. The freedman then earns some wealth and dies [with no heir]. Who has the inheritance?” [The imàm] said, “If the manumission of the slave was an obligatory duty which had been placed upon his father due to Ωihàr,8 or in

6 The two most cited texts in what follows are al-ˇùsì, Al-Mabsù†, 10 vols. (Beirut: Dàr al-Kutub, 1412 AH) and al-Mu˙aqqiq, Sharà"i' al-islàm, 4 vols. in 2. (Najaf: Dàr al-Kutub al-Islàmì, 1389 AH). They represent Imàmì fiqh in its developmental and classical phases respectively, and hence a comparison between them appropriately demonstrates the formation of legal orthodoxy in Imàmì fiqh. 7 Al-Kulaynì (d. 329/939) and Ibn Bàbùya record this as min kasbihi; al-ˇùsì in his al-Tahdhìb, 10 vols. (Tehran: Dàr al-Kutub al-Islàmiyya, 1364 AH) and in his alIstibsar, 4 vols. (Tehran: Dàr al-Kutub al-Islàmiyya, 1390 AH) record this as min kisihi. 8 In place of Ωihàr (the vow which effects divorce by declaring one’s wife to be “one’s mother”), Ibn Bàbùya records nadhr (a vow more generally).

patronate in early shì'ite law

137

thanks [to God, i.e., in the fulfillment of a vow] or some other obligation, then the freedman is sà"iba and he has no [patronate] tie to anyone. If [the freedman], before he dies, makes an association [of patronate] with one of the Muslims, then [that person] is liable for the freedman’s delicts and actions. [That person] is his patron and heir as long as the [freedman] has no relative who can inherit from him. If [the freedman] makes no such association with a Muslim before he dies, then inheritance goes to the imàm of the Muslims, providing he has no relative9 who can inherit from him. If the manumission of the slave, which the father had taken upon himself, was merely voluntary, and the father had ordered the slave to be freed for him [after his death], then the patronate tie of the freedman is inherited by all the10 descendants of the deceased. The one who bought the slave and freed him on the orders of his father is just one of the inheritors, providing [once again] that the freedman has no free male relatives who may inherit from him. If the son bought the slave, and freed him for his father, using his own money, and did so voluntarily without any orders from his father, then the patronate tie and the inheritance of the freedman belongs to the one who bought him with his own wealth, and then freed him for his father, providing that the freedman has no heir from amongst his relatives.11

The detailed exposition by the imàm indicates a relatively sophisticated juristic setting in which this putative exchange originated. The various combinations are entertained, and the rules decreed. However, there is no expression here of abstract rules relating to walà": rather the rules emerge from the exploration of the implications of a specific case. The dynamic of the case can perhaps best be presented in a flow diagram (figure 1). Some of the legal principles which emerge from this case govern the institution of walà" in a manner not dissimilar to the legal principles found in works of the Sunnì schools. However there are a number of distinctive elements which are not shared with the Sunnì schools (or are not shared with all of the Sunnì schools). The most important of these are:

9

Ibn Bàbùya inserts “no relative from amongst the Muslims” here. Ibn Bàbùya inserts the term “male” here. The variants listed here are not insignificant since, for example, the debate over whether women can inherit walà" from a male relative drew on the variety of views present in early Imàmì fiqh. 11 Al-Kulaynì, Kàfì, VII, 171; Ibn Bàbùya, Faqìh, III, 137; al-ˇùsì, Tahdhìb, VIII, 254; id., Istibßàr, IV, 23. 10

138

robert gleave

1) Walà" cannot emerge from manumission carried out as an obligation. The obligation to free a slave arises from a transgression of the law (for example, failing to fulfill the fast during Rama∂an), or from the failure to fulfill a vow. For a walà" tie to be established, the manumission must be gratuitous, and not “required” to remain in obedience to the law. 2) Walà" can be created by a contract between two free individuals (this contractual walà" is also permitted by most Óanafì jurists). One takes on the rights and duties of a patron, and the other those of the client (i.e., the freedman). This contractual patronate shares many of the legal effects of the manumission patronate. 3) Though for Imàmìs both contractual walà" and manumission walà" are legally valid, both are rendered irrelevant by the existence of the family members of the slave (quràba). In these circumstances, the patron is excluded from inheritance and has no duties of blood-money payment. Whilst this principle is present in Sunnì formulations of walà", its execution in Imàmì law is, as we shall see, distinctive. 4) The walà" tie survives the death of both patron and client. Whilst its legal effects may be limited in subsequent generations, it is not extinguished by death. The families of the freedman and the patron are bound together in perpetuity. This principle, also, is found in Sunnì law, though, once again, its execution is distinctive. Under an imaginary application of Imàmì law, there would probably be fewer patronate ties, since walà" only arises from gratuitous manumission. However, patronate ties can be more easily created by contract, which counterbalances the limitation to gratuitous manumission. Furthermore, patronate ties are more easily excluded from legal consideration as the presence of a family member of the freedman (but not a wife) nullifies the patronate tie. Hence, under Imàmì law, walà" (by contract or manumission) may or may not be more common, but it is certainly a fragile institution compared with the ties envisaged in the Sunnì systems. Part of the reason for the fragility of the walà" institution in Shì'ite law is that human patronate is of secondary importance, compared with the patronate due to the imàm from all true believers.

patronate in early shì'ite law

139

Figure 1: The distribution of walà" in the case cited in Mu˙ammad b. Ya'qùb al-Kulaynì al-Kàfì fì 'ilm al-dìn, VII, 171.

Did the manumission come about due to an obligation to free a slave placed upon the master?

No

Yes Did the freedman establish an agreement of walà" with another Muslim?

Was manumission carried out by the patron in-person?

Yes Yes

No

No

The walà" is established with respect to the partner in the agreement

Walà" is established

No walà" is evident here.

Did the manumitter give orders for the manumission to be carried out, either in his will or otherwise?

No

The walà" is established with respect to the person who performed the manumission but not the person originally conceived of as the manumitter.

Yes

Walà" is established

140

robert gleave Walà" in Imàmì akhbàr and fiqh

The material found in the collections of akhbàr (mainly a˙àdìth from the imàms, but included are some reports from the Prophet) can be seen as developing in tandem with the writings of early jurists. Indeed I have argued elsewhere that some of the Imàmì “collections” are not collections as such, but belong to other genres of Muslim writing.12 This is not to say that all the rules found in works of fiqh can be evidenced in the akhbàr directly. Legal reasoning plays a significant role in the elaboration of the law. Furthermore the link between akhbàr as source and fiqh work as description of the law took some time to be explicitly expressed. The earliest Imàmì works of fiqh read like mukhtaßars. Evidence for a particular rule is only occasionally adduced and the reasoning behind a rule left to later commentators to describe (or on some occasions, to create). With respect to walà", the most relevant sections of the akhbàr collections are found mainly in four chapters or sub-sections: kitàb al'itq (on manumission), kitàb al-mìràth ( or kitàb al-farà"i∂, on inheritance), kitàb al-nikà˙ (on marriage)13 and kitàb al-qißàß (or kitàb al-diyàt, on compensation for delicts). Mu˙ammad b. Óasan al-ˇùsì (d. 460/1067) in his fiqh work al-Mabsù†, recognizes the limited legal relevance of walà": Three sets of rules are associated with walà": marriage, blood-money ('aql ) and inheritance. All of these are also associated with kinship (nasab).14

Here the link is explicitly made between ties of the patronate (walà") and ties of kinship (nasab) echoing the famous Prophetic dictum alwalà" lu˙ma kalu˙ma al-nasab. It was the points at which walà" was analogous to nasab (and the points where it failed to be so) that formed the major points of discussion in Imàmì and Sunnì law, as exemplified by the four principal distinctive elements of Imàmì walà" listed above.

12 Gleave, “Between Hadìth and fiqh: The “Canonical” Imàmì Collections of Akhbàr,” Islamic Law and Society 8 (2001). 13 I cite fewer examples from the section on marriage law, since the influence of walà" on Imàmì marriage law cannot be said to be “distinctive” in a manner comparable with inheritance and delict-liability. 14 Al-ˇùsì, Mabsù†, IV, 93.

patronate in early shì'ite law

141

1. Walà" emerges from gratuitous manumission alone As has been recognized by other writers, Imàmì law recognizes the patronate tie as only emerging from cases where the master frees a slave gratuitously. That is, for walà" to exist as a legally relevant relationship, the master must not have been compelled by a legal obligation to free the slave. Manumission must, therefore, be a supererogatory act for the walà" tie to emerge. In this respect, Imàmì law differs from most variants of Sunnì doctrine where any manumission, be it obligatory or supererogatory, creates a patronate tie. The justification for this stipulation in Imàmì law is primarily based upon akhbàr, reports from the imàms in which those freed through obligatory manumission are excepted a walà" relationship with their former master. A freedman exempt from the tie is termed sà"iba, and in a number of akhbàr the term sà"iba is linked with Qur"ànic regulations in which manumission forms an essential element. For example: Ibn Ma˙bùb from 'Ammàr b. Abì al-A˙waß from imàm al-Bàqir: 'Ammàr asked imàm al-Bàqir about al-sà"iba. [The imàm] said “Look in the Qur"àn and in it, there is [the phrase] ta˙rìr raqaba. That, O 'Ammàr, is al-sà"iba: he who has no walà" to any of the Muslims [but has walà"] to God, the mighty the eternal.”15

The stipulation that obligatory manumission does not lead to walà" cannot be obviously justified by reference to the Qur"àn. However, a juridical link between the language of akhbàr and fiqh (sà"iba) and the phraseology of divine revelation (ta˙rìr raqaba) is established through reports such as this. On those occasions when the Qur"àn uses the phrase ta˙rìr raqaba (“free a slave”), no walà" tie is established. When one examines the employment of this phrase in the Qur"àn, one can construct a list of circumstances in which sà"iba status is achieved: Q.4:92: (at three points during this verse): “He who kills a believer unintentionally must free one believing slave (ta˙rìr raqaba mu"mina)”; “If the victim is from a hostile, but Muslim, tribe, then free a believing slave (ta˙rìr raqaba mu"mina)”; “If the victim is from an allied tribe, then [pay] the blood-money and free a believing slave (ta˙rìr raqaba mu"mina).” 15 Al-Kulaynì, Kàfì, VII, 171; Ibn Bàbùya, Faqìh, III, 136; al-ˇùsì, Tahdhìb, VIII, 252; id., Istibßàr, IV, 26.

142

robert gleave

Q.5:89: “The penalty for a broken oath is feeding ten paupers, or clothing ten paupers . . . or freeing one slave (ta˙rìr raqaba).” Q.58:3: “Those who divorce their wives by using [the Ωihàr 16 formula] shall free a slave (ta˙rìr raqaba) before being reunited.” In all these circumstances, the manumission (of a believing, or a non-believing, slave) is seen either as a penalty for transgressing the law (as in the cases of unintentional suicide or breaking an oath), or as a consequence of the operation of the law (as in the case of the Ωihàr formula).17 Whilst in works of usul al-fiqh the Imàmì jurists denounced the use of qiyàs, in the elaboration of the law identifying underlying reasons for Qur"ànic (and other revelatory) rulings was commonplace. In this circumstance, the establishment of sà"iba status was not restricted to occasions of unintentional homicide, breaking an oath and using the Ωihàr formula. Instead the jurists concluded that these cases all concerned manumission for “obligatory reasons” (wàjibàt), and the result was a general ruling that on those occasions where manumission was required in order for the master to restore his legal rectitude before God, there could be no walà" tie, and the freedman is sà"iba. Hence in fiqh works, like the Mabsù† of ˇùsì, one finds such general declarations as: Walà" is not established for anyone who frees a person for an obligatory reason, such as a vow or as penance. [ The freedman] is sà"iba. There is no walà" tie upon him from anyone.18

Here the meaning of sà"iba includes those who are freed in fulfillment of a vow (which should be distinguished from a slave freed as a penalty for failure to fulfill a vow). In addition to the above instances of “obligatory” manumission, sà"iba status could be granted at the whim of the manumitter, as indicated by the following report:

16

This refers to the controversial vow of sexual abstinence in which the man (and, according to some, the woman) vows to abstain from sexual relations for a limited period of time. 17 It is debated whether the Ωihàr formula is a transgression of the law, or merely disapproved. It is not clear from the Qur"àn, though there is nearly unanimously held opinion that it has legal effects. See Hawting, “An ascetic vow and an unseemly oath? ila and zihar in Muslim Law,” BSOAS 57 (1994). 18 Al-ˇùsì, Mabsù†, VI, 70.

patronate in early shì'ite law

143

Ibn Ma˙bùb from Khàlid b. Jarìr from Abù Rabì' who asked imàm Ja'far about sà"iba. He said, “[It is when] a man frees his slave and then says, ‘Go wherever you please. I will have nothing from your inheritance and I have no duty regarding your delicts.’ and he has two witnesses to this effect”.19

There was not inconsiderable debate over reports such as this. Did it merely describe the formula to be used in cases of obligatory manumission, or was it opening up the possibility of tasyìb (the unilateral declaration of an absence of walà" on occasions when one might expect the patronate to be created)? Resolution of this debate was settled in works of fiqh and tasyìb became a legally accepted practice.20 Here the walà" tie can be obliterated in cases of voluntary manumission, providing that the master declares the slave free of walà". Other possible means of establishing sà"iba status include kitàba manumission (an arrangement whereby a slave purchases his own freedom by installments). Whilst this is not a case of obligatory manumission (it does not act as penance for a transgression by the master, for example), the slave freed by kitàba is sà"iba. Another report is transmitted via imàm al-Bàqir, in which imàm 'Alì states: Whoever mistreats his slave, that slave is free with no walà". He is sà"iba and may go.21

This is probably not a case of penitential recompense (kaffàra) since the master is not to free any slave, but specifically the one he mistreated. It is best viewed as a compensation for the slave and a penalty for the crime of the master. For the crime of mistreating a slave, the master loses the ownership of the slave; as compensation for the mistreatment, the slave is freed with no walà" relationship. Other legal issues emerge, however, for the slave (being a person of

19 Al-Kulaynì, Kàfì, VII, 171; Ibn Bàbùya, Faqìh, III, 136; al-ˇùsì, Tahdhìb, VIII, 252; id., Istibßàr, IV, 26. 20 According to Crone (Roman, provincial and Islamic Law, 81), tasyìb was present in the doctrines of pre-classical law, but was abandoned in classical law. The Imàmì acceptance of tasyìb opens up the interesting question of the relative antiquity of elements of Imàmì jurisprudence in comparison with Sunnì law. Crone calls them “archaisms” (Crone, Roman, provincial and Islamic Law, 83). With respect to inheritance law, this has been explored by Richard Kimber in his “The Qur"ànic Law of Inheritance”. On tasyìb in pre-classical law, see Mitter, “Unconditional Manumission in Early Islamic Law: a ˙adìth analysis,” Der Islam (2001). 21 Al-Kulaynì, Kàfì, VII, 172.

144

robert gleave

limited legal capacity) cannot receive compensation for a crime committed against him by another. Legal niceties such as this were discussed, at times at great length, in works of fiqh. For ˇùsì, sà"iba status is portrayed as normally emerging from obligatory manumission. Mu˙ammad Ibn Idrìs (d. 558/1202), on the other hand, established a different mode of presentation by placing unilateral declaration as the norm, with other possibilities listed afterwards: Whenever a man frees a slave and exempts himself from liability for the [freedman’s] delicts, then the freedman is sà"iba. He has no walà". This also applies to those who are freed in fulfillment of a vow, as penance or due to some other obligatory action. There is no walà" for anyone who frees a slave in this way.22

Whilst revelatory material provided the immediate context (and justification) for Imàmì juristic rulings concerning walà", there ran parallel to these discussions a more exploratory exposition of why walà" was only established in cases of gratuitous or voluntary manumission (ta†awwu'an as it is termed in the literature). The reasoning here is related to the theological importance of walà" indicated at the outset of this essay. Being partner to a walà" relationship introduces an intermediate patronate relationship between the believer and his imàm. The imposition of this intermediate tie, whilst operating purely in the legal realm (related to inheritance, marriage and compensation), was, in a sense, a privilege for the patron. The patron was adopting some of the functions of the imàm, and hence the motivation behind the manumission had to be pure. This could only be achieved by supererogatory (and not obligatory) actions. Hence the patron earned the role of mawlà of the freedman, not through his mere obedience to the law, but through his efforts to go beyond what is legally required. This proposed explanation of walà" to cases of gratuitous manumission is always linked to the Imàmì report “When one is freed for [the sake of ] Allàh, then He [i.e., God] is patron to the one who is freed.”23 The phrase “for Allàh” is taken to refer to those occasions when manumission is required for the master to obey the law. 22 Ibn Idrìs, Sarà"ir al-islàm, 3 vols. (Najaf: Mu"assasat al-Nashr al-Islàmì, n.d.), III, 25. 23 Al-Kulaynì, Kàfì, VI, 198/139; al-ˇùsì, Tahdhìb, VIII, 250.

patronate in early shì'ite law

145

These reflections represent the established position in works of fiqh from the fifth/eleventh century onwards. However, there remained some inconvenient akhbàr which required interpretation. ˇùsì, whose commentary on selected akhbàr (entitled al-Istibßàr) was forced to deal with reports such as that attributed to imàm al-Bàqir: “the sà"iba and the non-sà"iba are equal on manumission”.24 His response was, first, to reject the report as mursal (a standard technique for disposing of difficult reports which contradict the fiqh). Next he states: That the walà" of the sà"iba is identical with the walà" of the non-sà"iba is not present in the obvious reading of the report (Ωàhir al-khabar). The report only establishes that the two are equal in terms of manumission.25

That is, they are equal in the sense that they have both been freed, but this does not imply that they are identical in terms of patronate status. A further problematic report is found in the Man là ya˙∂uruhu alfaqìh of Ibn Bàbùya al-Shaykh al-Íadùq (d. 381/991): 'Àßim b. Óamìd from Abù Baßìr who asked imàm Ja'far, “A man frees another man in penance for an oath or for Ωihàr. Who has the walà" here?” [The imàm] said, “The one who performed the manumission”.26

The obvious interpretation of this khabar is (surely?) that the imàm’s reply is a response to Abù Baßìr’s question. The assumption of relevance, presented in discussions of pragmatics, would lead to the conclusion that manumission for an oath or the use of the Ωihàr formula in divorce leads to a patronate tie.27 This is in direct contradiction, not only with other akhbàr, but also (and perhaps more importantly) the established legal doctrine. ˇùsì seeks to alleviate the problematic content of this report through an explanation which might be considered unconvincing: With respect to this report, we interpret it in this way: his walà" is due to [the former master] because the slave made an agreement with

24

Al-ˇùsì, Tahdhìb, II, 230; id., Istibßàr, IV, 27. Al-ˇùsì, Istibßàr, IV, 67. 26 Ibn Bàbùya, Faqìh, III, 133. 27 I explore one means of employing the “assumption of relevance” in tafsìr of the Qur"àn in Gleave, “The ‘First Source’ of Islamic Law: Muslim Legal Exegesis of the Qur"àn,” in R. O’Dair and A. Lewis, eds., Law and Religion (Oxford: Oxford University Press, 2001), 154–55. 25

146

robert gleave him after manumission [i.e., a case of contractual walà"], because if the slave had not made such an agreement, then he would be sà"iba.28

ˇùsì is arguing, then, that a contractual walà" tie must be implied (though not explicitly stated) in this case, otherwise the report is unable to be harmonized with other akhbàr, and the law in general. At the very least, the akhbàr found in early Imàmì collections represent legal doctrines current amongst Imàmì jurists of the ninth and tenth centuries, though this need not commit one to the opinion that the akhbàr can all be traced back to the imàms. From the efforts exercised by ˇùsì to bring the akhbàr into line with fiqh, one can conclude that there was a body of opinion which considered walà" to be created on obligatory in addition to gratuitous manumission (as it is in Sunnì doctrine). The significance of this body of opinion is unclear, though it was sufficiently influential to have its views enshrined in akhbàr attributed to the imàms. By rejecting the reports as unsound, or through apparently “forced” interpretation, the potential of these reports to disrupt legal doctrine was reduced. The rule concerning walà" via gratuitous manumission had, of course, further legal ramifications. In the case cited at the outset, the son freed a slave for his dead father. The imàm decreed that a slave freed for another, on the latter’s orders, leads to the same legal effects as if he had freed a slave himself. It is dependent upon his intention in freeing the slave—is it to expiate a transgression or is it a gratuitous act? Only in the case of a gratuitous manumission does walà" arise. In his Mabsù†, however, ˇùsì puts forward the following argument concerning voluntary manumission by proxy after death: Manumission by one person for another [who is dead] is the creation of a patronate tie for that person after his death. “The Patronate is a tie like the tie of kinship” said the Prophet. Just as it is not permitted to create a kinship tie to someone [after his death], so it is not permitted regarding the patronate.29

The issue at stake here is when (and if ) walà" is analogous to nasab. ˇùsì points out (accurately) that a person born after the death of a 28 Al-ˇùsì, Istibßàr, IV, 26–27. This explanation is also found in Tahdhìb, VII, 256. The editor of the Faqìh offers another unconvincing explanation: that we interpret the verb “he frees” as indeterminate, meaning no one freed the slave as such, and hence he has walà" to himself. Ibn Bàbùya, Faqìh, III, 133. 29 Al-ˇùsì, Mabsù†, VI, 210.

patronate in early shì'ite law

147

relative will never inherit from that relative. There must be some shared experience of kinship for that relationship to emerge and inheritance to become a possibility. This shared experience has rather limited possibilities when one of the parties is dead. Analogously a patronate tie emerges from a shared experience of manumission, which is similarly impossible if one party is deceased. This argument is used in reference to those who die intestate, and ˇùsì does not rule it out when discussing those who die leaving a will in which voluntary manumission is stipulated. Just as one cannot create a kinship tie through a will, so one cannot create a patronate tie by the same means. It would seem to imply (and ˇùsì is less than unambiguous on this point) that the legal principles laid out in the case cited earlier concerning the son and father are invalid. In that report, the freedman’s walà" is shared amongst the (male) heirs of the deceased manumitter, providing the manumission was voluntary (I discuss the manner in which patronate tie can be inherited below). Later tradition was resistant to ˇùsì’s reasoning here and argued that, in this aspect, walà" was not analogous to nasab: it could be established after the death of the manumitter, just as walà" is established for the mudabbar freedman.30 Indeed the case cited above could well be a case of tadbìr. Mu˙ammad Óasan al-Najafì (d. 1266/1860) reflects the attitude of the Imàmì tradition well when he states: What is related in the Mabsù† in the section on “penance and the establishment of the manumitter’s patronate” is clearly weak.31

In sum, the rule that walà" could not be created through legally required manumission was not left unchallenged in Imàmì fiqh (and there are akhbàr which could be interpreted as supporting opposing opinions). From the earliest fiqh works, however, it does seem to have been the dominant opinion. Furthermore, much of the detail of the law relating to this rule (such as walà" created after the death of the manumitter) took some time to take canonical form.

30 The manumission known as tadbìr refers to an agreement whereby the slave becomes free on the death of his master. See Brunschvig, art. “'Abd,” EI 2, I, 24. 31 Ja'far al-Najafì, Jawàhir al-kalàm fì shar˙ sharà"i' al-islàm, 15 vols. (Beirut: Mu"assasat al-Murta∂à, 1412 AH), XIII, 629.

robert gleave

148 2. Contractual Walà"

Imàmì jurists, like the Óanafìs, also permitted walà" to be established by contract. In the report cited above concerning the son freeing a slave for his dead father, this was indicated by the phrase “making an association before death” (tawàlà qabl an yamùt). The validity of contractual walà" is also raised in other reports. When asked about a slave, freed as sà"iba, imàm Ja'far replied: He associates with whomever he wishes ( yatawallà man shà"a). Liability for his [delicts] is upon the one with whom he has associated, and his inheritance is also for him.32

Imàm Ja'far is also reported as having said: When a man associates (wàlà) with another man, [the latter] has [the former’s] inheritance, and upon [the latter] is the duty of delict payment.33

The variety of verbal forms used here (wàlà, tawallà, tawàlà) have been translated as “to associate” or “to make an association”. It is possible to interpret this variation in linguistic use as representing a certain lack of clarity in relation to the contract. Whilst tawàlà (and possibly wàlà also) hint at a degree of mutuality in the contract, tawallà indicates more firmly a patron-client type relationship. Another ambiguity present in the akhbàr is whether this contract of the patronate is open to all free persons, or restricted according to manumission history. At times in the akhbàr collections, contractual walà" appears open only to those who have been previously freed from both slavery and walà" simultaneously. The ambiguity in the sources concerning the details of contractual walà" reflected (or caused) the variety in both presentation and substance of the presentation of walà" in early Imàmì fiqh. Ibn Bàbùya’s al-Muqni', for example, follows the wording of reports quite closely. On contractual walà" he writes: Whoever frees a slave as sà"iba has no liability for his [delicts] and has no inheritance from him. He should testify to this effect. Whomever the [sà"iba] wishes to associate with (tawallà), if that person agrees, then liability [for delicts] is a duty upon him and [the sà"iba’s] inheritance is upon him also.34 32 33 34

Al-Kulaynì, Kàfì, VII, 171. Ibid. Ibn Bàbùya, Al-Muqni', 38.

patronate in early shì'ite law

149

Here contractual walà" appears, if not restricted to, then certainly the norm for a sà"iba and the contract is modeled on the patronclient relationship, with no hint of mutuality. ˇùsì in his Mabsù†, however, presents contractual walà" in a quite different manner: Two men make a pact (ta'àqada 'alà al-ta'àdud ) and they say, “I make a contract with you that you protect me and I protect you. You defend me and I defend you. I will be your 'àqila, and you mine. You inherit from me, and I from you.” For some of [the Sunnìs], when they do this, it has no legal effect. For us, if one of them has an heir, then there is no legal effect, whether the heir be close or distant. If there is no heir, and each one has made it a condition that he is the 'àqila of the other to the exclusion of all others, then walà" is established between them. Each one inherits from his friend, according to the rules of walà".35

Here the emphasis is quite clearly on the mutual nature of the contract, and there is no clear patron-client dynamic in operation. It is, perhaps, significant that the technical term for contractual walà" (ta∂ammun al-jarìra) is not mentioned here. It may not yet have appeared in Imàmì fiqh, though ˇùsì uses it in his chapter on farà"i∂ in the Mabsù†: As for walà" ta∂ammun al-jarìra, this is [applied to] anyone who frees a person due to an obligatory matter like an oath (nadhr) or penance (kaffàra). There is no walà" established. [The slave] is sà"iba and has no walà" to anyone. If he makes an association (tawàlà) with a man, then that person will be responsible for his delicts and actions, and he will also receive [the freedman’s] inheritance, if he leaves no heir. This contract is sound. This is the case also for one who has no heir at all. He makes an agreement with a man with this condition and his walà" is then too [this man]. This is also sound and the inheritance goes to the [latter], just as in gratuitous walà".36

ˇùsì here presents a contract more obviously based upon the patronclient model, with superior (a'là) and inferior (asfal ) parties. Furthermore, whilst a non-sà"iba might make such a contract, the norm for contractual walà" is that the client party will be a sà"iba freedman. ˇùsì has yet another description of contractual walà" in his Nihàya:

35 36

Al-ˇùsì, Mabsù†, VI, 70. Ibid., IV, 108.

150

robert gleave Walà" is of two types. One comes about through manumission and the other through ta∂ammun al-jarìra . . . as for the latter type of walà", it applies to when a person undertakes the liability for the delicts of the freedman. When a man frees a slave and disassociates himself from the liability for the freedman’s delicts, then the former slave is sà"iba. This is also the case when he mistreats his slave . . . or if he frees him due to an obligation of penance or Ωihàr or unintentional homicide or a vow or breaking the fast of Ramadan or other obligations. The freedman in these cases is sà"iba. He has no walà" to the one who freed him, nor to anyone else. If this freedman makes an association (tawàlà) with his former master that the latter is liable for his delicts, then walà" is restored between them. If he makes an association with another man, then walà" is established here also.37

Here, as in the kitàb al-farà"i∂ of the Mabsù† the emphasis is on the creation of a patron-client rather than a mutual relationship. From this brief survey, one observes a number of discrepancies (both between authors, but also within one author’s writings) in the manner in which contractual walà" is presented: At times contractual walà" is a mutual contract in which the two parties are equals. At other points, contractual walà" is clearly presented as an unequal contract on the patron-client basis. At times the contractual walà" is viewed solely as an avenue for a sà"iba to re-establish the patronate after being freed due to an obligation on his former master. At other times, this is presented as the norm, though a non-sà"iba is also permitted to make such a contract. At yet other times, the contract is described as open to all individuals without walà", without conditions as to the status of the parties. In earlier works, and in the akhbàr collections, the technical term for contractual walà" (ta∂ammun al-jarìra) is not used. In later works, it has become standard. By the time Ibn Idrìs wrote his Sarà"ir, the juristic tradition had settled upon a more coherent presentation: Walà" ta∂ammun al-jarìra is when two people, one or both of whom is of unknown genealogy or has no heirs, agree (ta'àqada) that one will be the 'àqila for the other and liable for his delicts and errors. He will also receive the inheritance from the other.38

Here one sees the incorporation of the possible mutual nature of the contract (“one or both of whom . . .”), though the rule is expressed 37 Al-ˇùsì, Al-Nihàya, in: Al-Jawàmi' al-fiqhiyya (Qum: Kitàbkhàna-yi Àyatallàh Mar'ashì, 1404 AH), 367–68. 38 Ibn Idrìs, Sarà"ir, III, 23.

patronate in early shì'ite law

151

in the patron-client model. Furthermore the parties are described not in terms of their status history (i.e., one or both as sà"iba) but in terms of genealogy (“of unknown genealogy”). By the time of al-Mu˙aqqiq al-Óillì ( Ja'far b. Óasan, d. 676/1277), Imàmì presentation of contractual walà" had attained a stability which was dominant in subsequent centuries: Walà" ta∂ammun al-jarìra: someone makes an association (tawàlà) with another that the latter is liable for the former’s actions and that the former has walà" to the latter.39

The points of emphasis here are that anyone, regardless of status history or genealogy may enter into a contractual patronate relationship. The relationship mimics that of the patron and client through manumission, and has a designated term: ta∂ammun al-jarìra. This is further confirmed in his discussion of liability for delicts in contractual walà": The ∂àmin al-jarìra is liable, but the ma∂mùn is not. He [the ∂àmin] is not liable with the 'aßaba, and not with a patron by manumission because his agreement is dependent upon the [ma∂mùn] being of unknown genealogy or having no patron.40

Here Mu˙aqqiq hints that the validity of the contract is dependent upon the client-party being of unknown genealogy or having no patron himself. The emphasis once again is on the unequal nature of the contract. It is this presentation which became the standard expression in Imàmì fiqh, and Mu˙aqqiq’s achievement in encapsulating Imàmì law in all sorts of areas explains why his Sharà"i' al-Islàm was the most popular choice for later commentary from amongst the early fiqh works. The discrepancies concerning mutual vs. patron/client patronate, sà"iba vs. non-sà"iba participants and terminology had been resolved: Contractual walà" is the creation of a patron-client relationship through an agreement.41

39

Al-Mu˙aqqiq, Sharà"i' al-islàm, IV, 39. Ibid., IV, 290. 41 It should be noted that this was a norm of presentation, rather than an exclusion of mutual contractual walà". Later commentators reincorporated ˇùsì’s acceptance of mutual contractual walà" in their commentaries on the Sharà"i'. See for 40

152

robert gleave

Contractual walà" is open to all free persons. Individuals with no heirs or genealogy will obviously find the arrangement most attractive as it provides both an heir and an 'àqila. Such people may be freed slaves (sà"iba), but those of unknown genealogy (majhùl al-nasab), such as a foundling (laqì†)42 would also qualify. Contractual walà" is given a term whereby it might become part of the fiqh canon: ta∂ammun al-jarìra. Walà" was seen as a fictitious kinship tie in both Sunnì and Imàmì fiqh. It functions as such in the area of marriage, inheritance and blood-money. Contractual walà" (valid according to the Imàmìs) is perhaps best seen as a fictitious walà"-tie. In Imàmì legal writings, it functions in an identical manner to manumission-acquired walà", except (it appears) it cannot be inherited. 3. Inheritance and liability through walà" For Imàmì jurists, the patron can only inherit if there are no relatives (the dhawì al-ansàb) of the freedman living, or able to inherit. Similarly the patron is only liable for blood-money payment if the freedman has no relatives who can act as an adequate 'àqila for him. Here one sees the relationship between the rules of inheritance and 'aßaba (blood-money group) interacting with the rules of the patronate. In both areas (mìràth and 'aql ), the Imàmì system differed from that found in Sunnì fiqh. The majority Imàmì view, shared by most Sunnì jurists, is that the client cannot inherit from the patron, but a minority Imàmì view, preserved in a passage in Ibn Bàbùya’s Man là ya˙∂uruhu alfaqìh, indicates that the client may also inherit: If a man dies leaving a patron (mawlà mu'minan) or a client (mu'minan 'alayhi ), and leaves no other heir, then that person has the wealth.43

This view, as with some of the views expressed over contractual walà", was rejected, and ˇùsì and Mu˙aqqiq restrict inheritance to the patron alone. As Mu˙aqqiq unambiguously points out, “The example Mu˙ammad al-MuΩaffar, Taw∂ì˙ al-kalàm fì shar˙ sharà"i' al-islàm, 10 vols. (Baghdàd: Dàr al-Ta'lìm, 1400 AH), IV, 334, where he states in commentary on this passage, “If it is a mutual agreement (min al-†arafayn), then the person says, ‘You are my 'àqila, I am yours. You inherit from me, and I from you.’” 42 A foundling has no walà" to his finder in Imàmì fiqh (al-ˇùsì, Mabsù†, VI, 70). 43 Ibn Bàbùya, Faqìh, IV, 305.

patronate in early shì'ite law

153

freedman does not inherit from the patron. There is no dispute [!] over this.” As is well-known the Shì'ite law of inheritance does not accept male agnatic relatives as a separate class of heir; that is, the 'aßaba, as defined in Sunnì fiqh, do not have an exclusive right to the residue of the estate after the Qur"ànic portions have been allocated. In their place in Imàmì fiqh are placed the blood relations of the deceased (dhawì al-quràba), both agnate and cognate, who succeed according to their proximity to the deceased. The enumeration of the members of this group is not so relevant here, except to say that this group is considerably more extensive than that envisaged by Sunnì law, including ascendants and descendant relatives in the absence of a nearer relative. This group (also given the term dhawì al-ansàb), according to a number of reports, excludes the patron of the individual from any inheritance: Imàm Ja'far said that imàm 'Alì never took inheritance from any of his freedmen if they had relatives (quràba). He gave it to relatives of the freedman.44

Under Sunnì law, the patron is excluded from inheritance by the deceased’s agnates, but “inherits together with the Qur"ànic heirs and himself excludes remoter relatives”.45 In Imàmì law, he is excluded much more easily, and inherits nothing if the deceased has relatives, be they remote or near. He does however inherit with the spouse of the deceased in the absence of relatives. ˇùsì states the position well: The patron may be in only two situations for us: he either takes all the wealth in the absence of relatives or asbàb (i.e., a wife), or he inherits nothing. There are no circumstances in which he inherits when there is a single relative, be they a Qur"ànic heir or not. For the Sunnìs, the patron may be in two situations also: either he inherits everything, or he inherits a portion. . . . For us, he only takes a portion if there is a wife alone. The wife takes a quarter [as Qur"ànic heir] and the rest goes to the patron.46

The ease whereby a patron is excluded from inheriting is based upon the akhbàr of the imàms and implies that the patronate is of importance (in symbolic terms), but of reduced legal relevance (in practical

44 45 46

Al-Kulaynì, Kàfì, VII, 135. Crone, art. “Mawlà,” EI 2, VI, 874. Al-ˇùsì, Mabsù†, IV, 94.

154

robert gleave

terms). Part of this may be the theological emphasis on the sanctity of the family unit, stemming from a devotion to the ahl al-bayt. Compensation and blood-money, the delicts for which the patron is liable, are also affected by the presence of relatives. Ta'ßìb, or agnacy, functions in Sunnì law as both a means of distributing inheritance, but also as a means of distributing (financial) liability for the delicts of the relative. The burden is distributed amongst the 'àqila— that is, firstly, the male relatives ('aßaba) of the guilty party. This definition excludes female relatives, who under Sunnì law do not inherit as part of the 'aßaba in any case. Under Shì'ì law, however, the female agnates (and cognates) inherit since the ta'ßìb principle is irrelevant. The issue is, then, how does Imàmì law define 'àqila when the 'aßabà have no place in inheritance law? The answer lies in the uncoupling of inheritance regulations from the idea of an 'àqila. “It is not a condition of being a [member of the 'àqila] that one also be one of the people who inherit the wealth of the deceased” states Mu˙aqqiq.47 The 'àqila is presented as consisting of first the 'aßaba and second the patron. 'Aßaba is used here in a manner very similar to its usage in Sunnì fiqh. There was dispute over which relatives were included in the term. ˇùsì, for example, argues that the son and father of the guilty party were not included,48 whilst Mu˙aqqiq (and subsequent Imàmì fiqh) asserted that they were.49 Both views were supported by Imàmic reports. Explicitly excluded from the 'àqila, also on the basis of Imàmic reports and in agreement with the Sunnìs were women, minors and the insane, even though they may be amongst the heirs. The Óanafì inclusion of the ahl al-balad (fellow countrymen) and the ahl al-dìwàn (military comrades) of the culprit is also not considered valid by the Imàmìs. In the absence of 'aßaba, the patron bears the liability; furthermore, if there is 'aßaba (the definition of which is described in the next section) then its liability is limited by minimum and maximum payments, assessed by a means test. If there is insufficient wealth from these parties, then the patron should also be liable. ˇùsì expresses the principle thus:

47 48 49

Al-Mu˙aqqiq, Sharà"i', IV, 288. Al-ˇùsì, Mabsù†, VII, 173. Al-Mu˙aqqiq, Sharà"i', IV, 288.

patronate in early shì'ite law

155

[The patron] is liable after the relatives ('aßabàt). The order (of liability) is as follows: when the diya [i.e., the amount] has been determined . . . we allocate a third [each year] to the brothers and sons [of the culprit]. Then [if the amount is not reached], the uncles and their sons; then the uncles of the father and so on. If there are no relatives remaining who can make the payment, the patron must make up what is left.50

Mu˙aqqiq puts it more simply: Are patrons taken from when there are 'aßabàt? There is debate, but “yes”, when the diya is too much for the 'aßaba.51

The underlying principle here is then that the patron by manumission is not entirely excluded from liability by the existence of relatives. Rather, his liability is dependent on the amount of diya due, and the ability of the 'aßabà to pay. Once again he acts as a safeguard ( just as the imàm himself does) against injustice. 4. Inheriting walà" The patronate tie survives the death of both client and patron. The lineage of the two families are bound by walà" in perpetuity, though its legal relevance lessens as the client gains relatives. This certainly applies to manumission-acquired walà" (the jurists rarely discuss this as an issue in relation to contractual walà", though it seems it does not survive death). However, the walà" tie is made legally irrelevant quite easily as it is excluded by the presence of the client’s relatives. It is important to remember that the tie is not extinguished by the presence of the client’s relatives, but merely temporarily suspended. ˇùsì explores 28 different, and increasingly complex, cases of how walà" (and hence the duties and benefits associated with it) are passed on after the death of both patron and client. He recognizes however that most of his discussion is theoretical as walà" is so easily excluded from consideration, saying “we only outlined these cases so that one might know the school of the Sunnìs and practice [one’s legal skills] through them.”52 If a patron dies, the client’s walà" does not disappear but is passed on to the heirs of the patron. If the patron has no heirs, then the 50 51 52

Al-ˇùsì, Mabsù†, VII, 179. Al-Mu˙aqqiq, Sharà"i', IV, 290. Al-ˇùsì, Mabsù†, IV, 108.

156

robert gleave

walà" of his freedman is passed on to the patron’s patron. If this earlier patron has died, it is passed on to his heirs and so on. This complex system whereby walà" is inherited inspired some of the most excruciating cases of patronate and inheritance in the works of fiqh. It is clear that these were aimed, rather than providing useful guidance to judges or muftis, at exercising the legal minds of students. The term used for heirs in this context is 'aßaba. As we have already seen, ta'ßìb was excluded from the Imàmì inheritance system, but made reappearance in modified form in the rules concerning compensation for delicts. Inheritance of walà" is another area where ta'ßìb was re-activated, and it was a matter of some dispute here also. The issue centered on whether a woman (normally excluded from the definition of 'aßaba) might inherit the patronate tie of a deceased relative. Here the usual definition of 'aßaba lies in tension with the Imàmì inheritance principle regarding the dhawì al-ansàb. Once again the extent to which walà" is analogous to nasab is debated. In this case, the question concerned whether the system of inheriting wealth is the model for inheriting the patronate. Mu˙aqqiq, standing at the end of the formative period of Imàmì fiqh, describes the differences of opinion succinctly: If there is no patron-benefactor [i.e., he has died], then Ibn Bàbùya says, “The walà" goes to the male and female descendants [of the patron]”. This view is approved (˙asan) and in found in [ˇùsì’s] alKhilàf, with the condition that the patron was a man. Mufìd said, “the walà" goes to the male descendants, excluding the female ones, whether the patron was a man or a woman.” Shaykh [al-ˇùsì] in al-Nihàya said, “walà" goes to the male descendants, excluding the female ones if the patron was a man.” If [the patron] was a woman, then the walà" goes to her 'aßaba. He said that the reports [of the imàms] support this view.53

To the two opinions related from ˇùsì (found in his Khilàf and Nihàya), one can add the view expressed in the Mabsù† (which seems identical to that expressed by Mufìd here).54 ˇùsì also defines 'aßaba in a manner I have not seen challenged (nor extensively commented upon) in later fiqh:

53 54

Al-Mu˙aqqiq, Sharà"i', IV, 36. Al-ˇùsì, Mabsù†, IV, 98.

patronate in early shì'ite law

157

'Aßaba means firstly two people: the father and the son, and with respect to walà", this means the 'aßaba is the father of the patron or the son of the patron. The son and father of the patron inherit the walà" together because they are of the same degree [of distance from the patron]. For the Sunnìs, the son excludes the father. The son excludes the son’s son. There is no dispute over this. The father excludes the grandfather. For the Sunnìs, the grandfather excludes the brother, but for us they share [the patronate]. The brother excludes the paternal uncle. There is no dispute over this. The brother’s son shares the same level with the grandfather, according to us. For the Sunnìs, the grandfather excludes [the brother’s son]. The brother’s son excludes the paternal uncle and the paternal uncle’s son. There is no dispute over this. The paternal uncle excludes the paternal uncle’s son and so on. The brothers and sisters via the mother [i.e., cognate relatives], and those related to them do not inherit walà". There is no dispute over this.55

The last section of this passage establishes clearly the Imàmìs’ uncoupling of the inheritance of walà" (where cognates are excluded) from the inheritance of wealth (where according to the Imàmì system, cognates are included). This, then, is a definition of 'aßaba-priority which is entirely irrelevant to the inheritance of wealth, but activated in the inheritance of the patronate. It is clear that the patronate can be divided between people, and it follows that its duties and benefits are similarly divided. One has here the Shì'ite idea of 'aßaba which would probably have been utilized had not Imàmì law rejected the concept of ta'ßìb in its inheritance law.56 Figure 2, below shows the order of inheritors of walà", the higher categories excluding completely the lower categories. One can see how it differs from (ˇùsì’s conception of ) Sunnì 'aßaba:

55

Ibid., IV, 94. Of course, the Sunnì system of inheritance of walà" is not identical with the system of wealth inheritance, though the systems are related in a manner not found in the Imàmì systems. The crucial point for the Sunnìs was the kubr rule, whereby walà", unlike wealth returns to the brother of the deceased patron if there is no son, rather than being passed on to the grandchildren of the deceased (as wealth would) if the son is deceased. I thank Ulrike Mitter for making this point clear. On this, see Crone, Roman and Provincial Law, 82–83. 56

robert gleave

158

Figure 2: Those listed together share walà" equally. Imàmì Inheritance of walà"

Sunni Inheritance of walà" (according to al-ˇùsì)

Class I: Father and Son

Class I: Son Class II: Father

Class II: Grandfather, Grandson and Brother*

Class III: Grandson and Grandfather

Class III: Brother’s son

Class IV: Brother Class V: Brother’s son

Class IV: Paternal Uncle

Class VI: Paternal Uncle

Class V: Paternal uncle’s son

Class VII: Paternal Uncle’s Son

* ˇùsì is not entirely clear on whether the grandfather and grandson exclude the brother from inheriting walà" or they share it.

The difference of opinion over whether a woman descendent of a deceased patron can inherit the patronate of his freedman was settled, as in so many other areas of walà" law, not through extensive juristic argumentation, but by the dominating influence on later generations of Mu˙aqqiq’s Sharà"i' al-Islàm. His description of the manner in which walà" is inherited from the patron is dependent on the gender of the manumitter: If the manumitter is male, then the walà" is transferred to the male descendents excluding the female (awlàd al-dhukùr dùna al-inàth). If there are no male descendents, the patron of the manumitter inherits the walà", be he male or female. If the patron is dead, then the 'aßaba of the patron inherit the walà" of their relative’s client’s client. If tracing the walà" back through the generations ever comes to a stop, then the imàm himself inherits the walà", and the client’s inheritance goes to the bayt màl al-muslimìn (overseen by the imàm). If the manumitter is female, then the walà" is transferred to the woman’s 'aßaba, according to the established order. Failing that, the walà" (and hence the duties of liability and benefits of inheritance) go to the imàm and the bayt al-màl. A woman then is excluded from both inheriting the patronate and inheriting by the patronate except if she freed the client herself, or

patronate in early shì'ite law

159

if her client, after manumission, frees a slave and predeceases her leaving no relatives.57 As ˇùsì states: A woman only inherits through walà" in two circumstances: if she performs the manumission herself and has a client, or if she is the patron to a patron [who predeceases her].58

There is an acceptance in later fiqh that although women form part of the dhawì al-ansàb and hence can inherit wealth, they do not inherit patronate ties, either by their relation to the deceased or via the deceased’s 'aßaba. The reason for this exclusion is obviously linked to the woman’s exclusion from being a member of the 'àqila for a relative culprit. Being a patron brings the risk of similar liability and hence she is also excluded here. The patronate tie of the client is also prone to transference, both in life and after death. The child of a freewoman (for example) has walà" to the patron of his mother. If his father also acquires walà" (through manumission), then the father’s patronate tie takes precedence over that of the mother. The basis for the priority of the father’s walà" is firstly the Qur"ànic verse, “Call them after their fathers” (Q 33:5). This verse refers to adopted sons, but is interpreted by analogy to apply to walà" also. Just as the father’s name takes precedence over that of the mother, the father’s walà" takes precedence over that of the mother. The principle here is also provided with evidence from the akhbàr, such as: Al-'Ays asked the imàm about a man who bought a slave who had children by a free woman. The man then frees the slave. The imàm said, “The walà" of his offspring belong to the one who freed him.”59

Mu˙aqqiq explores the inheritance of a client’s walà" through a series of examples (masà"il ) such as the following: 57 Ibn Idrìs does not appear to make walà" inheritance dependent upon the gender of the manumitter. He also does not use the term 'aßaba common in other works of Imàmì fiqh. However, he does restrict the inheritance of walà" “to one who is closest to the patron on his father’s side—not his mother’s side”. He claims ˇùsì’s support for such a view, though as we have seen, ˇùsì has a number of views, expressed in different works of fiqh, as to how walà" should be inherited. See Ibn Idrìs, Sarà"ir, III, 266. 58 Ibid., IV, 95. 59 Al-Kulaynì, Kàfì, VII, 170; Ibn Bàbùya, Faqìh, III, 79; al-ˇùsì, Tahdhìb, VIII, 250; id., Istibßàr, IV, 21. This report might be said to contradict the report related

160

robert gleave If a slave marries a freedwoman and has a child by her, then the walà" of the offspring goes to the mother’s patron [not the slave’s since he is not free]. If the slave dies and the grandfather of the children [i.e., the father’s father] is freed, then, according to ˇùsì, the walà" of the children is transferred (injarra) to the patron of the grandfather, for he stands in place of the father. This would be the case if the father was alive [but not free]. If the father is freed, then the children’s walà" is transferred again from the patron of the grandfather to the father’s patron, because he is nearer in relation.60

From this, it can be seen that for Mu˙aqqiq, walà" derived from a parent can be transferred between mother and grandfather, but once it reaches the father, it is henceforth immoveable. Of course the legal relevance of this matter is minimal, since the patron’s inheritance and liability for delicts is excluded by the presence of dhawì al-ansàb. If however, the child’s relatives predecease him, and he has no other heir, then his father’s patron would inherit. The complexity of this case, however, palls in comparison to those described by ˇùsì. For example: A slave marries the freedwoman of a man. By her, the slave fathers two girls. The two girls are free because they are the offspring of the freedwoman. Their walà" is to the patron of the mother and [their mother’s] walà" is also to her patron. The two girls buy, together, their father. He is immediately freed because “when a child buys his slave father, the father is immediately free” [i.e., one cannot, de jure, be in a master-slave relationship with one’s father]. The father’s walà" is transferred to his own patron. So his walà" is to his own daughters. When this is established, the walà" [of the daughters, which was previously to the mother’s patron] must be transferred to the patron of their father [since the father’s walà" takes priority]. They are themselves the patron to the father. Hence they have become the patron to themselves. However, it is impossible, ˇùsì states, for one to have walà" to oneself:

by Ibn Bàbùya, from imàm Ja'far, “The freedman is the client, and the son associates with whomever he pleases” (Ibn Bàbùya, Faqìh, III, 135–36). However, this contradiction is easily solved by the altering vowelling (al-mu'taq/freedman could be al-mu'tiq/manumitter) and the ambiguity in the term mawlà (patron or client). 60 Al-Mu˙aqqiq, Sharà"i', IV, 37.

patronate in early shì'ite law

161

Each one of the sisters owes their walà" to the patron of their father. This walà" is divided in half [because their father has two patrons, namely themselves]. Hence each daughter is in a half-walà" relationship with each of her father’s patrons. [The walà" she is due to herself is obliterated, because one cannot have walà" to oneself ] The other half of the walà" goes to her sister.61

Hence the sisters are in a half-walà" relationship to each other, through their father, to whom they are the patron! Of course with regard to inheritance (or indeed liability), most of this calculation is irrelevant, since the presence of a close relative (be it a sister, father or mother) excludes the patron, yet ˇùsì cannot resist computing the various combinations and comparing them with Sunnì fiqh (for which the presence of the relative is not such a definitive bar on the inheriting patron). ˇùsì’s final example demonstrates the tenacity of walà". Though rendered irrelevant temporarily, it lies dormant for a generation only to reappear at some later time: A slave marries a freedwoman and fathers a son by her. The son is free due to the free-status of the mother. His walà" is due to his mother’s patron. The son grows up and buys a slave, whom he frees. The freedman has walà" to the son. Time passes and this freedman buys the father of his patron, whom he then frees. When the father is free, the walà" of the son, which had previously been towards the patron of his mother, is transferred to the patron of his father. This [latter patron] is, in fact, his own client. Each one of the two, son and client, is patron and client to the other. They are both patron (mawlà min a'la) and client (mawlà min asfal ) to each other. [The mother dies] Then the father dies. The son gets the inheritance by kinship (nasab). The freedman dies and has no relative. The son gets the inheritance by walà". If however the son dies [after the father but before the freedman], then his wealth goes to the freedman by walà", via his father. If they both die [simultaneously, one presumes (RG)] and there are no other relatives [of either], then the wealth goes to the patron of the mother.62

The complicated reasoning here can be explained as follows:

61 62

Al-ˇùsì, Mabsù†, IV, 104–105. Ibid., IV, 106–107.

162

robert gleave

Freedman’s inheritance He has no relatives, so his inheritance goes to his patron (the son). However his patron is dead (as they died simultaneously), so therefore it should go to the 'aßaba of the patron/son. However, there are no 'aßabàt, and hence it should go to the patron of the patron, who was originally the patron of the son’s mother. Hence the mother’s patron receives the freedman’s inheritance. The son’s inheritance The son’s inheritance should go to his relatives, but both his parents are dead. Therefore it goes to his patron. His patron is his father’s patron in the first instance [i.e., the freedman] by transferred walà" from the mother’s patron. However the father’s patron/freedman is also dead, and the inheritance belongs to the patron’s 'aßaba. There are no 'aßabàt, and hence it should go to the patron’s patron (which is the son himself, who is also dead). Next it is due to the patron’s patron’s 'aßaba, but they are also dead (i.e., the father of the son). Finally, then it must go to the original patron of the patron’s patron, which was the mother’s patron (mawlà mawlà mawlà). Therefore the son’s inheritance is also delivered to the mother’s patron. I do not believe that ˇùsì expected this case to be taken too seriously, and he probably considered it a humorous example of the complexity of the walà" laws on inheritance. What it does demonstrate, however, is that though walà" may be rendered temporarily irrelevant (such as the mother’s and son’s walà" to her patron), it is not obliterated. It can, if circumstances allow, re-emerge at some later time and provide an unexpected benefit to a patron some time after the original manumission. The manner in which walà" survived the death of the parties and was passed on to their families was complex and I have only given a taster of the intricacies of its operation here. The point that should be stressed however is the tenacity of the walà" link. The Imàmì jurists, like their Sunnì counterparts, were willing to exclude it from consideration in many cases, but never to nullify it. In Imàmì Shì'ism, this gives some idea of the quasi-sacred character of the tie, based as it is on the theological relationship between a believer and his imàm.

patronate in early shì'ite law

163

Of particular legal interest here, as in the discussion on 'àqila, is the acceptance of the ta'ßìb system (albeit in modified form) which is missing from the Imàmì system of wealth inheritance. This incorporation of ta'ßìb means that in Imàmì fiqh the walà" is inherited by the patron’s 'aßaba, or his/her male descendents. There was an underlying tendency for walà" to end up in the hands of men, with women being excluded unless they were the manumitter themselves or the patron of a manumitter who predeceases them with no relatives. In this sense, walà" can be likened to the institution of Imàma, which passed through Fà†ima to 'Alì, and on through 'Alì’s male descendents.

Conclusions The primary patronate duty a believer owes to his imàm was expressed in Imàmì fiqh by the overarching security provided by the imàm. A person who has no (family) heir or liability group is provided with these by the imàm himself. Whilst in other areas of Imàmì law (such as zakàt, khums or Friday prayer)63 the occultation of the imàm presented immediate issues of legal validity and legitimacy. It took some time for the legal implications of the occultation to filter through to the intricacies of walà" law. If the imàm is absent due to his occultation, how is one with no walà" and no relatives (be he both free or sà"iba) provided with the legal protection guaranteed under the law. ˇùsì hardly discusses walà" al-imàma, and Ibn Idrìs appears unconcerned that the imàm is actually in hiding and therefore cannot provide the protection due to the believer with no kin.64 Mu˙aqqiq, once again, shows a more thorough legal understanding and explicitly discusses the question of the absent imàm in his section on walà" al-imàm: If the imàm is absent, then the person’s inheritance is divided between the poor and the indigent (al-fuqarà" wa-l-masàkìn). It is not given to anyone other than the sul†àn al-˙aqq except if there is fear

63 On these duties and their delegation to the fuqahà" during the ghayba, see Calder, “Zakàt in Shì'ì Jurisprudence from the Tenth to the Sixteenth Century A.D.,” BSOAS 64/IV (1981); Calder, “Khums in Imàmì Shì'ì Jurisprudence from the Tenth to the Sixteenth Century A.D.,” BSOAS 65/I (1982); and Sachedina, The Just Ruler in Shi'ite Islam, 177–203. 64 Ibn Idrìs, Sarà"ir, III, 265–66.

164

robert gleave

[for the shì'a’s well-being, i.e., through taqiyya] or if there is force [i.e., if the illegitimate, non-Shì'ì sul†àn takes the inheritance illegally].65 Interestingly, the discussions concerning the liability for the delict of the person with no 'àqila seem untouched by the question of the doctrine of occultation. Whilst the doctrine that the fuqahà" are the delegates of the imàm (nawwàb) during his occultation permitted them to collect and distribute khums and zakàt, lead Friday prayers and (much later in Imàmì fiqh) legitimate jihàd, it does not seem to have been discussed in relation to the payment of liability claims. Here there is no discussion of whether the fuqahà" represent the imàm during the ghayba in terms of diya payment. The fuqahà" were representatives only in the collection of monies. Mu˙aqqiq does describe the course of action in cases where some or all of the 'àqila is existent but absent: If some of the 'àqila are absent, then those present are not responsible for the remainder of the diya.66

The diya payment, he explains is delayed until such time as they appear. The duty of payment does not disappear, but remains as a debt upon them and their heirs. This is because the diya is a debt upon each of the 'àqila which cannot be transferred (ibtadà"an) in Imàmì law, and this would presumably include the debt owed by the imàm. This was one area of law which could await the imàm’s return. The above description of the development of walà" in early Imàmì Shì'ite jurisprudence is aimed to exemplify the manner in which certain distinctive legal doctrines became legal orthodoxy through a process of debate and discussion. Imàmì walà" was subject to the same juristic discussion found in other areas of the law, and here, as elsewhere, opinions were formed out of a dialectic in which the major players were the revelatory material (Qur"àn, Prophetic reports and the akhbàr of the imàms), the opinions of the Sunnìs and the various techniques of exegesis and legal reasoning present also in other fiqh traditions. The major concerns of the Imàmì jurists in 65 Al-Mu˙aqqiq, Sharà"i', IV, 40. MuΩaffar (d. 1322/1904) helpfully explains that it is preferable to give the inheritance to the “general delegate of the imàm” (nà"ib al-imàm al-'àmm) during the occultation. This phase, as is well-known, was code for the fuqahà". 66 Ibn Idrìs, Sarà"ir, IV, 291. The great commentator, Najafì, gives no hint that this doctrine might be affected by the ghayba. Jawàhir, XV, 533–34.

patronate in early shì'ite law

165

developing a coherent theory of the patronate exemplify this dialectic: To what extent can walà" be seen as analogous to kinship ties, and when it is not analogous, what model should be adopted? This is particularly evident in the acceptance of the ta'ßìb principle in terms of delict-liability and inheritance of walà". How might inconvenient elements of revelatory material, particularly the akhbàr of the imàms, be interpreted so as to be in line with the emerging legal orthodoxy? In what ways might the quasi-sacred character of the patronate tie be preserved, bearing in mind that it is an intermediary loyalty which might obscure the religious devotion of the believer to the imàm? The Imàmì answers to such questions lead to a theory of walà" which was, perhaps, more distinctive than other areas of the law, where Imàmì fiqh shows a greater similarity with Sunnì doctrine.

Bibliography Primary Sources al-Ba˙ranì, Muflih al-Íamìrì. Ghàyat al-maràm fì shar˙ sharà"i' al-islàm. 4 vols. Beirut: Dàr al-Kutub, 1420/1999. Ibn Bàbùya, Mu˙ammad b. 'Alì. Man là ya˙∂uruhu al-faqìh. 4 vols. Qum: Mu"assasat al- Nashr al-Islàmì, 1404 AH. id., Al-Muqni' , in the collection Al-Jawàmi' al-fiqhiyya. Qum: Kitàbkhàna-yi Àyatallàh Mar'ashì, 1404 AH. Ibn Idrìs, Mu˙ammad. Sarà"ir al-islàm. 3 vols. Najaf: Mu"assasat al-Nashr al-Islàmì, n.d. al-Kulaynì, Mu˙ammad b. Ya'qùb. Al-Kàfì fì 'ilm al-dìn. 8 vols. Tehran: Dàr alKutub al- Islàmiyya, 1367 AH. al-Mu˙aqqiq, Ja'far b. al-Óasan. Sharà"i' al-islàm. 4 vols. in 2. Najaf: Dàr al-Kutub al-Islàmì, 1389 AH. al-MuΩaffar, Muhammad. Taw∂ì˙ al-kalàm fì shar˙ sharà"i' al-islàm. 10 vols. Baghdàd: Dàr al-Ta'lìm, 1400 AH. al-Najafì, Ja'far. Jawàhir al-kalàm fì shar˙ sharà"i' al-islàm. 15 vols. Beirut: Mu"assasat al- Murta∂à, 1412 AH. al-ˇùsì, Mu˙ammad b. al-Óasan. Al-Mabsù†. 10 vols. Beirut: Dàr al-Kutub, 1412 AH. id., Al-Tahdhìb. 10 vols. Tehran: Dàr al-Kutub al-Islàmiyya, 1364 AH. id., Al-Istibsar. 4 vols. Tehran: Dàr al-Kutub al-Islàmiyya, 1390 AH. id., Al-Nihàya, in the collection al-Jawàmi" al-fiqhiyya. Qum: Kitàbkhàna-yi Àyatallàh Mar'ashì, 1404 AH. Secondary Sources Amir-Moezzi, M. The Divine Guide in Early Shi'ism. New York: SUNY, 1994. Brunschvig, R. art. “'Abd,” EI 2, I, 24–40.

166

robert gleave

Calder, N. “Zakàt in Shì'ì Jurisprudence from the Tenth to the Sixteenth Century A.D.,” BSOAS 64/IV (1981). id., “Khums in Imàmì Shì'ì Jurisprudence from the Tenth to the Sixteenth Century A.D.,” BSOAS 65/I (1982). Corbin, H. En Islam Iranien. 4 vols. Paris: Buchet and Chastel, 1971–72. Crone, P. Roman, provincial and Islamic Law: The Origins of the Islamic Patronate. Cambridge: Cambridge University Press, 1987. id., art. “Mawlà,” EI 2, VI, 874–882. Gleave, R. “Between Hadìth and fiqh: The ‘Canonical’ Imàmì Collections of akhbàr,” Islamic Law and Society 8 (2001). id., “The ‘First Source’ of Islamic Law: Muslim Legal Exegesis of the Qur"àn,” in R. O’Dair and A. Lewis, eds., Law and Religion. Oxford: Oxford University Press, 2001. Hawting, G.R. “An ascetic vow and an unseemly oath? ila and zihar in Muslim Law,” BSOAS 57 (1994). Kimber, R. “The Qur"anic law of inheritance,” Islamic Law and Society 5 (1998). Mitter, U. “Unconditional Manumission in Early Islamic Law: a ˙adìth analysis,” Der Islam (2001). Sachedina, A. The Just Ruler in Shi'ite Islam. Cambridge: Cambridge University Press, 1988.

MAWÀLÌ AND THE PROPHET’S FAMILY: AN EARLY SHÌ'ITE VIEW Patricia Crone

For Professor Madelung on his 70th birthday

The Imàmì work credited to the shadowy Sulaym b. Qays al-Hilàlì (d. c. 90/708f.)1 contains a letter allegedly sent by Mu'àwiya to his governor of Iraq, Ziyàd b. Abìhi, with instructions for Ziyàd to tear up the letter as soon as he had read it and never to divulge its contents to anyone. Sulaym b. Qays, an investigative journalist avant la lettre, managed to copy the letter thanks to his friendship with a Shì'ite secretary in Ziyàd’s service. Needless to say, it was a scoop. It revealed that 'Umar had been deeply hostile to non-Arab Muslims; that Mu'àwiya was continuing his policies against them; that the Umayyads owed their power to 'Umar and his ßà˙ib, i.e., Abù Bakr; and that non-Arab Muslims were natural allies of the Prophet’s family who would destroy the Umayyads as bearers of black banners from Khuràsàn.2 The letter is clearly fictitious. Mu'àwiya never wrote any such letter; nor is it obvious that the reader is being asked to believe that he did, though on this point opinions may differ. On the one hand, the sarcasm is so heavy that one would have thought the author beyond attempts at verisimilitude: he works like a cartoonist by drawing an exaggerated and distorted picture of the practices he dislikes and then having Mu'àwiya recommend them with pride.3 But on 1 Cf. EI 2, s.v., where his existence is questioned (Djebli). I owe my interest in the book to a paper presented by Maria Dakake at the MESA conference in Chicago 1998 and am indebted to Etan Kohlberg, Michael Cook, Samer Traboulsi and the participants in the conference for comments on an earlier draft. 2 Kitàb Sulaym b. Qays al-Hilàlì, ed. M. Bàqir Anßàrì (Qum: Nashr al-Hàdì, 1995), II, 739ff. There are several Najaf printings with different paginations where the letter is easy to find via the table of contents (kitàb mu'àwiya ilà ziyàd b. abìhi fi sha"n shì'at 'alì ). It is also reproduced in al-Majlisì, Bihàr al-anwàr (Beirut: Dàr al-Ri∂à, n.d.), XXXIII, 261ff., from Sulaym’s book. 3 It is the same method which is used in the anti-Ismà'ìlì Kitàb al-balàgh (cf. Stern, “The ‘Book of the Highest Initiation’ and other anti-Ismà'ìlì travesties,” in his Studies in Early Ismà'ìlism ( Jerusalem/Leiden: Magnes Press, 1983).

168

patricia crone

the other hand, the naive indignation expressed by some Shì'ite readers shows that the sarcasm was not always noticed, and the author may well have expected his audience to take his parody at face value.4 Either way, the letter is a clever piece of political satire. It is also of great historical interest. In what follows I give a summary of its contents, a full translation, a discussion of its date, and an analysis of what it tells us.

Summary Sulaym’s sensational story goes as follows. Mu'àwiya wrote to Ziyàd b. Abìhi, his adopted brother and governor of Iraq, telling him to enforce the discriminatory measures against non-Arab Muslims that 'Umar had instituted in matters such as inter-marriage, inheritance, stipends and appointments to positions of authority. Mu'àwiya reminded Ziyàd that 'Umar would actually have preferred to exterminate nonArab Muslims and that he had written to his governor of Baßra, Abù Mùsà al-Ash'arì, telling him to cut off the heads of all those of them who had reached a certain height. But back in those days Ziyàd b. Abìhi had himself been a client (mawlà), or so he had thought, this being before Mu'àwiya had adopted him as his brother, so he had dissuaded Abù Mùsà, his employer at the time, and also warned 'Umar that the measure would probably make the clients join the Prophet’s family, whom 'Umar had likewise maltreated, and that this might enable 'Alì to put an end to 'Umar’s mulk. This was all the more likely to happen, Ziyàd had said, in that he had heard 'Alì predict that the non-Arabs (al-a'àjim) would become better and braver Muslims than the Arabs, and that they would eventually kill the Arabs and take over their spoils. 'Umar had responded that this was precisely why he wanted to get them first: it was his intention to order all his governors to do what he wanted Abù Mùsà to do. But Ziyàd had pointed out that the non-Arabs were too numerous, that 'Alì and his family were too brave, and that their hatred of 'Umar and his companion, i.e., Abù Bakr, was too great, so 'Umar had abandoned his plans. 4 My thanks to Samer Traboulsi for opening my eyes to this possibility; cf. the indignant editorial note in the Najaf edition triggered by the passage translated below as §7. (The anti-Ismà'ìlì book was understood in the same literal vein, even by modern Islamicists, until quite recently.)

MAWÀLÌ

and the prophet’s family

169

Had he gone ahead, Mu'àwiya says, killing non-Arabs would have become sunna just as so many other rulings of 'Umar’s had become sunna even though they were contrary to the Prophet’s practice. The non-Arabs would in that case have been eradicated once and for all. Unfortunately, this had not happened, but Abù Bakr and 'Umar had at least succeeded in keeping the Hàshimites away from the caliphate, and it was the lowly status of their clans which had emboldened the Umayyads to seek the caliphate for themselves. If it had not been for Abù Bakr and 'Umar, the entire nation would have been mawàlì of the Hàshimites. In fact, this might still be the fate ahead, for Ziyàd had heard 'Alì predict, in the reign of 'Uthmàn, that the people coming with black banners from Khuràsàn would be non-Arabs (al-a'àjim) and that they would defeat the Umayyads, deprive them of their power, and kill them under every star.5

Translation6 (1) Abàn7 from Sulaym: Ziyàd b. Sumayya had a Shì'ite secretary 739 who was a friend of mine and who read aloud to me a letter that Mu'àwiya had written to Ziyàd in reply to a letter from Ziyàd to him. (It said:) (2) Now, you have written to me asking me about the Arabs, whom you should honor and whom you should despise, whom you

5 The author uses the term mawlà now in the narrow sense of freedman and now in the broad sense of non-Arab Muslims, but for non-Arab Muslims in general he prefers the term al-a'àjim, presumably because he found their client status demeaning: he carefully distinguishes clients in the sense of freedmen from other non-Arab converts (al-mawàlì wa-man aslama min al-a'àjim) in §§4,9 of the translation below. In later texts 'ajam is a derogative term for unassimilated Iranians (cf. the attestations in Crone, “The 'Abbàsid Abnà" and Sàsànid Cavalrymen,” Journal of the Royal Asiatic Society 8 (1998), 14). 6 The translation renders Anßàrì’s text without reference the variants in the notes, except where immediately relevant. 7 I.e., Abàn b. Abì 'Ayyàsh (Fìrùz), a Baßran mawlà who died in 140 or earlier and who was a companion of 'Alì b. al-Óusayn, Mu˙ammad al-Bàqir and Ja'far al-Íàdiq according to the Imàmìs, but who transmitted from jamà'ì scholars according to the Sunnìs. Both sides characterize him as unreliable, but the Sunnìs know nothing about his Shì'ism or his transmission of Sulaym’s book (Dà"irat al-ma'àrif al-islàmiyya al-kubrà (Tehran: Markaz Dà"irat al-Ma'àrif al-Islàmiyya, 1995), s.v. (I, 45ff.) and the sources cited there. My thanks to Etan Kohlberg for drawing this work to my attention).

170

patricia crone

should favor and whom you should keep at arm’s length, whom you should trust and of whom you should beware. (3) I, brother, know more about the Arabs than anyone else. Look to this tribe of Yemen, honor them in public and despise them in private, for that is how I treat them. I favor their public gatherings (majàlisahum) and give them to understand that they matter more to me than others, but in private my munificence and benevolence goes to others, because so many of them fight me on behalf of this man (sc. 'Alì).8 // And look to Rabì'a b. Nizàr. 740 Honor their nobles and despise their common folk ('àmmatuhum), for their multitudes follow their nobles and chiefs. And look to Mu∂ar. Let them beat one another, for they are crude, arrogant, proud and extremely haughty. If you do that and let them beat one another, they will save you the problem of keeping control of them. Don’t accept any statement of theirs which is not accompanied by action, or any opinion of theirs which falls short of certainty.9 (4) And look to the clients and non-Arabs who convert to Islam (almawàlì wa-man aslama min al-a'àjim). Treat them in accordance with the custom (sunna) of 'Umar b. al-Kha††àb, for it humiliates and disgraces them, namely: Arabs marry (women) from among them but do not give (women) in marriage to them;10 Arabs

8 There is some variation in the different transmission here, and the explanation that so many Yemenìs are 'Alid sympathizers is only found in one manuscript. 9 This paragraph is based on Mu'àwiya’s advice to Ziyàd as known, for example, from al-Haytham b. 'Adì: Ziyàd wrote to Mu'àwiya, asking for advice about the proper way of dealing with the Arabs; Mu'àwiya replied, “Look to this people of Yemen: honor them in public and despise them in secret. And look to this tribe of Rabì'a: honor their nobles and despise their masses (al-safila), for their masses follow the nobles. And as for this tribe of Mu∂ar, they are rough and crude, so put them in charge of one another ( fa-˙mil ba'∂ahum 'alà riqàb ba'∂in). . . .” (alBalàdhurì, Ansàb al-ashràf, ed. M.J. Kister ( Jerusalem: Magnes Press, 1971), IV a, 186; for another version, see al-Madà"inì, ibid., 13, where Mu'àwiya adds that Ziyàd should choose his governors from Mu∂ar). 10 For the Arab reluctance to give their women in marriage to non-Arabs, see Goldziher, Muhammedanische Studien (Halle: Max Niemeyer, 1889–90), I, 127ff. The Arabs placed non-Arab Muslims in the same category as Christians and Jews by marrying their women without giving women marriage to them, as 'Alì is said to have observed in disapproval of the practice (al-Kulaynì, Kàfì, ed. 'A.A. al-Ghaffàrì (Tehran: Dàr al-Kutub al-Islàmiyya, 1342–47 AH), V, 318f.; cf. also below, notes 64–66). The Prophet had declared all Muslims to be equal and it was 'Umar who first distinguished between Arabs and non-Arabs, Qurashìs and non-Qurashìs for purposes of marriage (Abù l-Qàsim al-Kùfì, Kitàb al-istighàtha (n.p. [Maktabat Niniwì l-Óadìtha], n.d.), 44f. (drawn to my attention by Etan Kohlberg); Majlisì, Bihàr,

MAWÀLÌ

and the prophet’s family

171

inherit from them whereas they do not inherit from Arabs;11 you curtail their stipends and rations;12 they go forth in campaigns to repair the road and cut trees;13 none of them acts as prayerleader for Arabs; none of them stands in the front line when Arabs are present, except to complete the line;14 you do not appoint anyone to any Muslim frontier province (thaghr) or garrison city (mißr), nor do you put any of them in charge of the judgeship (qa∂à") or the laws (a˙kàm) of the Muslims.15 (5) This was 'Umar’s custom (sunna) and practice (sìra) regarding them, may God reward him greatly for his services to Mu˙ammad’s community, and to the Umayyads in particular. By my life, if it had not been for what he and his companion (sc. Abù Bakr) did, XXXI, 35f.). The Sunnìs also know 'Umar to have had such views, though they do not describe them as an innovation (cf. below, note 60). 11 Patrons had a title to the estates of their clients, whereas clients did not inherit from their patrons (Crone, Roman, Provincial and Islamic Law (Cambridge: Cambridge University Press, 1987), 36ff.). But patrons were not always Arabs: mawàlì rapidly acquired freedmen and other clients of their own, who acquired freedmen and clients in their turn, and so on. Sulaym presumably held this to be an accidental feature of the institution. 12 The standard pay for an Arab soldier in the Umayyad period was 30 dinars/300 dirhams a year. That for a mawlà is said to have been 15 dinars under Mu'àwiya, 20 under 'Abd al-Malik, 25 under Sulaymàn, and 30 under Hishàm (Ibn 'Abd Rabbih, Al-'Iqd al-farìd, ed. A. Amìn et al., repr. (Cairo: Lajnat al-Ta"lìf wa-l-Tarjama wa-l-Nashr, 1965), IV, 400.2). The Sunnìs insist that 'Umar awarded the same pay to Arabs and mawàlì when he established the dìwàn (al-Balàdhurì, Futù˙ al-buldàn, ed. M.J. de Goeje (Leiden: E.J. Brill, 1866), 458.2; Abù 'Ubayd, Al-Amwàl, ed. M.K. Haràs (Cairo: Maktabat al-Kulliyàt al-Azhariyya, 1969), 335f.). But according to the Imàmìs, it was 'Umar who fixed the pay of mawàlì at 250 rather than 300 dirhams (thus al-Fa∂l b. Shàdhàn, Al-Ì∂à˙ (Beirut: Mu"assat al-A'lamì li-lMa†bù'àt, 1982), 136f., citing the Sunnìs themselves); and it was 'Alì who gave the same to Arabs and non-Arabs, even freed slaves (thus al-Ya'qùbì, Ta"rìkh, ed. M.T. Houtsma (Leiden: E.J. Brill, 1883), II, 213; al-Kulaynì, Kàfì, V, 318f.; VIII, 69; Ibn Abì l-Óadìd, Shar˙ nahj al-balàgha, ed. M.A.-F. Ibràhìm (Cairo: 'Ìsà Óalabì, 1965–67), II, 200f.). Cf. now also P. Crone, ‘The Pay of Client Soldiers’, Der Islam 80 (2003). 13 This particular grievance does not seem to be attested elsewhere, though mawàlì are also deemed worth having as road workers in the Sunnì version of Sulaym’s story (below, notes 55, 56). 14 That Arabs found it demeaning to pray behind mawàlì is well known: when Nàfi' b. Jubayr b. Mu†'im prayed behind a mawlà, he explained that he wanted to humiliate himself (Ibn 'Abd Rabbih, 'Iqd, III, 412f.); Salmàn al-Fàrisì reputedly said, “One can’t lead you Arabs in prayer or marry your women, God favored you over us with Mu˙ammad” (Ibn Qudàma, Al-Mughnì, ed. T.M. al-Zaynì et al. (Cairo: Maktabat al-Qàhira, 1968–70), VII, 33, no. 5126); and so forth. But the detail about not having them in the front line does not seem to be mentioned elsewhere. 15 This is somewhat exaggerated (cf. art. “Mawlà,” EI 2, VI, 878). But it is true that no mawlà was appointed to a coveted province such as Iraq or Khuràsàn in the Umayyad period, and that mawàlì were not acceptable as judges.

172

patricia crone

or for their strength and stubbornness in God’s religion, we and this entire nation would have been the clients of the Hàshimites (la-kunnà wa-jamì' hàdhihi l-umma li-banì hàshim al-mawàlì) and they would have inherited the caliphate, one after the other, in the 741 same way as the families // of Kisrà and Qayßar.16 But God removed it from the Hàshimites at their hands and gave it to B. Taym b. Murra, from where it passed to B. 'Adì b. Ka'b. No two tribes of Quraysh are smaller or more ignoble and despicable than they are, so they made us (Umayyads) desirous of it. We have a better right to it than they or their descendants do because we are wealthy and powerful,17 and we are closer to the Prophet in terms of kinship ( fì l-ra˙im) than they are.18 (6) Then our man (ßà˙ib) 'Uthmàn obtained it by consultation and general consent (bi-shùrà wa-ri∂à min al-'àmma) after consultation among six for three days.19 Those who had obtained it before had done so without shùrà. When our man 'Uthmàn was killed unjustly, we obtained it through him because God has given authority to the wàlì of someone killed unjustly.20 (7) By my life, brother, if 'Umar had fixed the blood-money of a client at half that of an Arab, that would have been more in keeping with piety.21 If I could find a way of doing it and could entertain hope that the common folk (al-'àmma)22 would accept it, then I would do it. But I have recently been at war and am

16 The reader is presumably meant to find it ridiculous that Mu'àwiya should make this comparison given that he introduced the kisrawì and qayßarì mode of succession himself. Dynastic succession was as wrong in the Umayyad case as in that of the Byzantines and Persians because it allowed ordinary families to monopolize power whereas it was right in the Shì'ite case because the Hàshimites were a sacred lineage. 17 Fìnà al-tharwa wa-l-'izz. The Najaf printing has al-ghazw for al-'izz. 18 'Abd Shams and Hàshim were brothers. Taym and 'Adì were collaterals of older ancestors (cf. Caskel, ]amharat an-nasab. Das Genealogische Werk des Hi“àm ibn Mu˙ammad al-Kalbì (Leiden: E.J. Brill, 1966), I, table 4). 19 Compare Akhbàr al-dawla al-'abbàsiyya wa-fìhi akhbàr al-'abbàs, ed. 'A.-'A. al-Dùrì and 'A.-J. al-Mu††alibì (Beirut: Dàr al-ˇalì'a li-l-ˇibà'a wa-l-Nashr, 1971), 51, where Mu'àwiya tells the Hàshimites that the caliphate has passed from one sub-tribe of Quraysh to another bi-ri∂à al-'àmma wa-bi-l-shùrà l-khàßßa. 20 Cf. Q.17:33 (wa-man qutila maΩlùman fa-qad ja'alnà li-walìyihi sul†ànan). This paragraph reproduces genuine Umayyad arguments, except for the Shì'ite point that shùrà was an innovation. 21 In law the blood-money of a client was the same as that for an Arab, though a client’s life was widely felt to be worth less (cf. EI 2, VI, 876b). 22 It also means non-Shì'ites, as one is undoubtedly meant to know.

MAWÀLÌ

and the prophet’s family

173

afraid of splitting people and turning them against me, so you must be content with what 'Umar instituted (sanna) regarding them, for that will humiliate and disgrace them. When this letter of mine reaches you, treat the non-Arabs with contempt and disdain, drive them away and don’t ask for help from any of them, and don’t fulfill any request of theirs. (8) By God, you are the son of Abù Sufyàn and have come from his loins. You don’t share any genealogy with 'Ubayd apart // 742 from Adam.23 You have told me, and I regard you as trustworthy, O brother, that you read 'Umar’s letter to the Ash'arì in Baßra. You were his secretary at the time and he was governor of Baßra, and you were the lowliest of people in his view, and submissive in spirit (dhalìl al-nafs) because you thought that you were a client of Thaqìf. If you had known the certain truth in those days, as you do today, namely that you are a son of Abù Sufyàn, you would have risen in your own estimation and been too proud to serve as secretary to someone falsely claiming to be an Ash'arì (da'ì al-ash'arìyyin).24 You know for sure, and we do too that {Abù Sufyàn’s // grandfather Umayya b. 'Abd Shams 743 went out to Syria on a trading journey and passed by Íaffùriyya, where he bought a slave and his son 'Abdallàh, and that}25 Abù Sufyàn used to follow the example of Umayya b. 'Abd Shams.26 (9) Ibn Abì Mu'ay†27 told me that you told him that you read 'Umar’s

23 Ziyàd counted as the son of 'Ubayd, a mawlà of Thaqìf, until Mu'àwiya claimed him as his brother (cf. the long editorial note to the passage; al-ˇabarì, Ta"rìkh alrusul wa-l-mulùk, ed. M.J. de Goeje et al. (Leiden: E.J. Brill, 1879–1901), II, 1382.4). 24 The irony is rich inasmuch as it was Ziyàd who was a da'ì (as he is explicitly called by an adherent of al-Óusayn in al-ˇabarì, Ta"rìkh, II, 293.18). 25 The sentence in curly brackets, which is missing in the Najaf printing, is a gloss which has been clumsily inserted in the text so that it reads anna abà sufyàn kharaja ma'ahu jadduhu umayya b. 'abd shams, though the events are set before Abù Sufyàn’s time. The translation assumes the intended meaning to have been anna umayya b. 'abd shams jadd abì suyfàn kharaja. 26 The story is normally told slightly differently: Umayya b. 'Abd Shams, the ancestor of the Umayyad dynasty, left Mecca after losing a merit competition with Abù Hàshim and spent ten years in Syria, where his slave girl was impregnated by a Jew from Íaffùriyya and had a son, whom Umayya acknowledged as his own (Crone, Meccan Trade and the Rise of Islam (New Jersey: Princeton University Press, 1987), 102 note 70, 117 note 45). Mu'àwiya similarly acknowledged Ziyàd, the son of a slave or freedman, as his brother. 27 The Jewish child adopted by Umayya in Syria (normally called Dhakwàn rather than 'Abdallàh) was the ancestor of 'Uqba b. Abì Mu'ay†, an archenemy of the Prophet who was executed after the battle of Badr. The reference here is to 'Uqba’s

174

patricia crone

letter aloud to Abù Mùsà al-Ash'arì.28 He ('Umar) had sent him five spans of rope and told him, “Inspect the Baßrans under you. Whoever you find from among the clients and non-Arab converts to Islam (al-mawàlì wa-man aslama min al-a'àjim) who reaches five spans, move him forward and cut off his head.”29 Abù Mùsà consulted you about it, and you told him not to, saying that he should try to make 'Umar reconsider, so he did and you took the letter to 'Umar. You did all that as a partisan of the clients (ta'aßßuban li-l-mawàlì), for at the time you thought that you were one of them and that you were a son of 'Ubayd, and you kept arguing with 'Umar until you had made him change his mind; you made him afraid that there might be a schism ( furqat al-nàs), so he was dissuaded; you asked him, “how can you be sure, given your hostile treatment of this family, that they will not flock to 'Alì so that he can rise up with them and put an end to your power (mulk)?” So he desisted. (10) Brother, I don’t know any son of Abù Sufyàn’s family who brought worse luck to them than you did when you made 'Umar 744 change his mind and told him not to do it. // He informed me that (the argument) with which you made him go back on his decision to kill them was that you had heard 'Alì b. Abì ˇàlib say, “The non-Arabs (al-a'àjim) will beat you in the cause son al-Walìd, 'Uthmàn’s half brother who was governor of Kùfa until he was dismissed for drunkenness (cf. Caskel, ]amhara, II, Register, s.vv.). Al-Walìd b. 'Uqba is excoriated as an 'ilj from Íaffùriyya in Majlisì, Bi˙àr, XIX, 260, 264; XXXXIX, 81. 28 In the previous paragraph (§8) it is Ziyàd who tells Mu'àwiya about how he read 'Umar’s letter to Abù Mùsà; here and in the beginning of the next paragraph (§10) it is Ibn Abì Mu'ay† who tells him; at the end of §10 the informant is once more Ziyàd (cf. below, note 33). One cannot remove the passages with one or the other informant as interpolated without making the story unintelligible, but it is possible that two different versions have been cobbled together in the story as we have it. (I am indebted to Samer Traboulsi for drawing this problem to my attention.) 29 Cf. the catalogue of 'Umar’s misdeeds in Sulaym, Kitàb, II, 682: he sent a rope of five spans to Basra with instructions to cut off the heads of anyone that high. The length is presumably determined, directly or indirectly, by the fact that 'Alì presumed a boy of five spans to have reached puberty (thus al-Zamakhsharì, Al-Kashshàf (Beirut: Dàr al-Kitàb al-'Arabì, n.d. [1947]), III, 254, ad Q.24:59; my thanks to Harald Motzki for this reference). Elsewhere in Sulaym’s book the rope is sent by 'Umar for the measurement of thieving boys in Basra, whose hands were to be cut off if they met the height requirement (Kitàb, II, 683). The Shì'ites also claim that 'Umar expelled every non-Arab (a'jamì) from Medina (ibid., 682) and that he wanted to kill all Zoroastrians until 'Alì persuaded him to accept jizya from them (Abù Óàtim al-Ràzì, A'làm al-nubuwwa, ed. Í. al-Íàwì and Gh. M. A'wànì (Tehran: Imperial Iranian Academy of Philosophy, 1977), 176).

MAWÀLÌ

and the prophet’s family

175

of this religion the second time round as you beat them the first time round”;30 and he said, “God will fill your hands with non-Arabs. They will become lions31 who do not flee, so they will smite your necks and take over your spoils ( fay").”32 So 'Umar said to you, “I have heard that from the Messenger of God, that is what made me write to your employer (ßà˙ib) about killing them. I had decided to write to my governors of all the garrison cities about it.” But you said to 'Umar, “Don’t do it, Commander of the Faithful. You can’t be sure that 'Alì won’t call on them to help him. There are many of them, and you know the bravery of 'Alì and his family and his hatred for you and your companion.” So you dissuaded him. You told me that it was partisan feelings ('aßabiyya) that made you dissuade him, and that it was cowardice which made 'Umar change his mind.33 You (also) told me that you mentioned this to 'Alì in the reign of 'Uthmàn and that he replied, “The people bearing the black banners which will come forth from Khuràsàn, they are the non-Arabs (hum al-a'àjim) and they are the ones who // will 745 wrest power from the Umayyads and kill them under every stone and star.” (11) Brother, if you had not made 'Umar change his mind, it would have become sunna, and God would have eradicated and exterminated them. The caliphs after him would have followed his custom (istannat bihi ) until there would not have remained of them a single hair or nail or anyone at all.34 They are the bane of religion (àfat al-dìn).

30 Also cited in Abù 'Ubayd, Gharìb al-˙adìth (Hyderabad: Ma†ba'at Majlis Dà"irat al-Ma'àrif al-'Uthmàniyya, 1964–67), III, 484; Ibn al-Faqìh, Kitàb al-buldàn, ed. M. J. de Goeje (Leiden: E.J. Brill, 1885), 317f. (both with al-˙amrà" for al-a'àjim). 31 Thus the Najaf printing (usd ), emended to “strong” (ashiddà") by Anßàrì. 32 Also cited in 'Abd al-Razzàq, Al-Mußannaf, ed. Ó.-R. al-A'Ωamì (Beirut: alMajlis al-'Ilmì, 1972), XI, no. 20811 (from al-Óasan). Compare Nu'aym b. Óammàd, Al-Fitan, ed. M. al-Shùrì (Beirut: Dàr al-Kutub al-'Ilmiyya, 1997), no. 658 (ed. S. Zakkàr (Mecca, n.d.), 141.22), where this is what will happen if the believers do not perform the duty of commanding right and forbidding wrong (cited in Cook, Commanding Right and Forbidding Wrong in Islamic Thought (Cambridge: Cambridge University Press, 2000), 36f., with a further reference to al-Muttaqì al-Hindì). 33 Following the second reading in note 52. Note that it is once more Ziyàd who is Mu'àwiya’s informant. 34 Literally “a blower of fire” (nàfikh nàr), cf. Lane, An Arabic-English Lexicon (London: Williams and Norgate, 1863–93), s.v. “nfkh.”

176

patricia crone

(12) How much of what 'Umar instituted in this community is contrary to the sunna of the Prophet, yet people followed him regarding it and adopted it. This would have been like one of those institutions. They include his removal of the maqàm from the place in which the Messenger of God had put it;35 his change of the ßà' and mudd of the Messenger of God, which he enlarged;36 his refusal to allow a person in a state of major ritual pollution to purify himself with sand,37 and many other things of diverse kinds under more than a thousand headings, of which the most important and dearest and most delightful to me is his removal of the caliphate from the Hàshimites, who are the people entitled to it and its font and origin (ma'dinihà), for it is not right for anyone except them, nor will the earth come right except through them.38 When you read this letter of mine, keep its contents secret and tear it up. (13) He said: when Ziyàd39 had read the letter, he flung it on the 746 ground, then he went up to me and said, “Woe unto me // for what I have left and what I have entered! I belonged to

35 'Umar is commonly said to have moved the stone in the Meccan sanctuary known as maqàm Ibràhìm; the Shì'ites say that it was to its Jàhilì location that he moved it (Kister, art. “Maqàm Ibràhìm,” EI 2, VI, 104b–107b). 36 For the disagreement over these measures, see Abù 'Ubayd, Amwàl, 688ff. (drawn to my attention by Avraham Hakim). 'Alì accuses 'Umar of having changed the Prophet’s measures in Sulaym, Kitàb, 677, and explains that this and other changes, including the maqàm, could not be reversed, ibid., 720f.; reproduced in Majlisì, Bihàr, LXXX, 350, and XXXIII, 265; cf. also XXXIV, 168, 174, where the explanation is cited again, but not from Kitàb Sulaym. (I owe the first and the last four references to Avraham Hakim.) 37 Sunnì jurists allow both major and minor pollution to be removed with sand (cf. art. “Tayammum,” EI 2), but 'Umar, Ibn Mas'ùd and Ibràhìm (al-Nakha'ì) are all said to have denied that tayammum could be used for major pollution (al-ˇabarì, Jàmi' al-bayàn fì tafsìr al-qur"àn (Beirut: Dàr al-Fikr, 1988), V, 113, ad Q.4:43; my thanks to Avraham Hakim for references which led to this one). 'Ammàr b. Yàsir reminded 'Umar of the Prophet’s sunna and Abù Mùsà al-Ash'arì reminded Ibn Mas'ùd of Q.5:6 (al-ˇabarì, loc. cit., also in the Sunnì Óadìth collections). But according to Sulaym (Kitàb, II, 680), 'Umar “wrote to all his governors that when someone in a state of major ritual pollution has no water, he may not pray; he cannot do dry ablution with what he finds on the ground. . . .” 38 Mu'àwiya is such a cynic that he praises 'Umar for violating the sunna and opposing the one man who could put the world right: to a modern reader, the parody is now so crude that the author might as well curse Mu'àwiya directly instead of trying to impersonate him. 39 The name seems to have been inserted by mistake. It must be Ziyàd’s secretary who flings the letter to the ground after reading it aloud to Sulaym (cf. §1), who picks it up and copies it.

MAWÀLÌ

and the prophet’s family

177

the party of Mu˙ammad’s family and band (shì'at àl mu˙ammad wa-˙izbihi ), but I have left it and entered the party of Satan and his band, and the party of someone who writes a letter like this! By God, I am like Iblìs who refused to prostrate to Adam out of arrogance and unbelief and envy.” (14) Sulaym said: I had copied the latter before the evening came. When night fell, he40 asked for the letter and tore it up, saying, “Let no one know what was in this letter.” He did not know that I had copied it.

The date of the letter The author of Sulaym’s sensational letter was clearly a Ràfi∂ì (his attitude to Abù Bakr and 'Umar is one of venomous hatred) and he wrote after the 'Abbàsid revolution, in full awareness of black banners from Khuràsàn and the massacres of Umayyads with which they were associated (§10). Yet he presents the revolution as a fulfillment of 'Alì’s hopes, with no sign of disappointment that it had led to the accession of 'Abbàsid rather than 'Alid caliphs. This suggests that he was a Hàshimite Shì'ite. Hàshimite Shì'ism was the belief that the Prophet’s family, defined as the entire Hàshimite clan, not just the 'Alids, had a hereditary right to the caliphate which had been usurped by others.41 When Mu˙ammad died, the knowledge which God had revealed to him passed to his family ('i†ra), not to anyone else, as a missionary on behalf of the Hàshimiyya explained in Khuràsàn in 129/746; therefore Mu˙ammad’s family were the keepers of the Prophet’s legacy and the font and origin of divine knowledge.42 Any Ràfi∂ì might have agreed with this, but the missionary did not equate the Prophet’s family with the 'Alids. It was this acceptance of all descendants of Hàshim as potential imàms in (usually, but not always, ràfi∂ì) preference to everyone else which distinguished Hàshimite Shì'ism from other forms, and it was this form of Shì'ism that the 'Abbàsids initially stood for. “There has not been any caliph among you since 40

Anßàrì identifies the subject as Ziyàd, but again it must be the secretary. Cf. Madelung, “The Hàshimiyyàt of al-Kumayt and Hàshimì Shì'ism,” Studia Islamica 70 (1989). 42 Al-ˇabarì, Ta"rìkh, II, 1961f. (the missionary was Abù Dàwùd). 41

178

patricia crone

the Prophet died, apart from 'Alì b. Abì ˇàlib and this Commander of the Faithful who is behind me,” Dàwùd b. 'Alì famously told the Kùfans in his accession speech on behalf of Abù l-'Abbàs in 750.43 But Hàshimite Shì'ism was a short-lived creed. Since the 'Abbàsids were not going to share the caliphate with the 'Alids, they were ill advised to continue stressing the special status of 'Alì, while the 'Alids themselves had nothing to gain from remaining under the shared Hàshimite umbrella. One soon had to choose between shì'at 'alì and shì'at banì l-'abbàs. Several participants in the revolution drifted from Hàshimite to 'Alid Shì'ism from the 140s/750s onwards,44 and in 145/762 the 'Alids rebelled against the 'Abbàsids under the leadership of Mu˙ammad al-Nafs al-Zakiyya, irrevocably splitting the Hàshimite family into two. Al-Manßùr responded (if his correspondence with the rebel is authentic) by asserting the rights of al-'Abbàs over those of 'Alì; and when the testament of a prominent leader of the revolution who died in 785 declared 'Alì to be the Prophet’s legatee (waßì) and heir to the imamate after him (wàrith al-imàma ba'dahu), al-Mahdì is said to have disapproved.45 He disapproves of raf∂ in favor of 'Alì elsewhere too,46 including in another story about a veteran who died in 785 (presumably a doublet of the first). Here the veteran asks the caliph to find favor with his son, of whom alMahdì says that he “is on the wrong track and at odds with our views and yours: he slanders the two shaykhs Abù Bakr and 'Umar and speaks ill of them,” to which the dying veteran replies that “he is adhering to the stance we stood for and called to when we rebelled.”47

43

Al-ˇabarì, Ta"rìkh, III, 37; al-Mas'ùdì, Murùj al-dhahab, ed. C. Pellat (Beirut: al-Jàmi'a al-Lubnàniyya, 1966–79), IV, §2279 (ed. Paris, VI, 55). Compare al-ˇabarì, Ta"rìkh, III, 33.4: “no caliph has ascended this pulpit of yours since the (death of ) the Prophet apart from the Commander of the Faithful 'Alì b. Abì ˇàlib and the Commander of the Faithful 'Abdallàh b. Mu˙ammad (i.e., Abù l-'Abbàs).” Slightly different versions appear in Ya'qùbì, Ta"rìkh, II, 420, and al-Azdì, Ta"rìkh al-mawßil, ed. 'A. Óabìba (Cairo: Lajnat I˙yà" al-Turàth al-Qawmì, 1968), 124. 44 Cf. Omar, The 'Abbàsid Caliphate, Baghdàd (National Printing and Publishing Co.) 1969, 157f., 182, 205f. 45 Al-ˇabarì, Ta"rìkh, III, 532, on al-Qàsim b. Mujàshi' al-Tamìmì. 46 Cf. the story in which he accuses the Kùfan judge Sharìk b. 'Abdallàh al-Nakha'ì of raf∂ (Wakì', Akhbàr al-qu∂àh, ed. 'A.M. al-Maràghì (Beirut: 'Àlam al-Kutub, n.d. [originally Cairo 1947–50]), III, 155f.; al-Marzubànì, Akhbàr shu'arà" al-shì"a, ed. M.H. al-Amìnì, repr. (Beirut: Shirkat al-Kutubì, 1993), 80; cf. also al-Kha†ìb alBaghdàdì, Ta"rìkh al-baghdàd (Cairo: Ma†ba'at al-Sa'àda, 1349 AH), IX, 292). 47 Al-ˇabarì, Ta"rìkh, III, 537, on Abù 'Awn 'Abd al-Malik b. Yazìd.

MAWÀLÌ

and the prophet’s family

179

By then the stance was no longer viable. Some people continued in government service without renouncing their allegiance to the 'Alids, but they did so by learning to live with divided loyalties, not by remaining loyal to a single, undivided house.48 Precisely what the official alternative to Hàshimite Shì'ism was is anything but clear, given that al-Mahdì is described now as a Ràfi∂ì or at least devotee of 'Alì’s,49 now as an anti-Ràfi∂ì, and now as the sponsor of a new form of raf∂ amounting to rejection of all pre-'Abbàsid caliphs, including 'Alì, as usurpers on the grounds that the 'Abbàsids had enjoyed an exclusive right to the caliphate from the start.50 But even adherents of the new type of raf∂ maintained a favorable attitude to Abù Bakr and 'Umar in public, we are told;51 and Hàrùn (786–809) reputedly went so far as to display a positive interest in 'Uthmàn.52 It was by recognizing 'Uthmàn as legitimate (under al-Mutawakkil?) that the 'Abbàsids repudiated their Shì'ite past altogether.53 The author of our letter rejects Abù Bakr and 'Umar in favor of the Hàshimites, casts the revolution as a fulfillment of predictions by 'Alì, and settles scores with the Umayyads in the triumphant knowledge that they have fallen without betraying any awareness of splits among the Hàshimites who have ousted them. This suggests that he was writing in the immediate aftermath of the revolution. So too does the fact that his work survived as an Imàmì document: like so many others, its author probably drifted to 'Alid Shì'ism when the dust began to settle. All in all, then, he probably wrote before 762, but in any case not after the 780s, when the last adherents of Hàshimite Shì'ism must have been dead or dying.

48

The best-known example is Yaq†ìn and his family. Cf. Khalìfa b. Khayyà†, Ta"rìkh, ed. S. Zakkàr (Damascus: Ma†àbi' Wizàrat al-Thaqàfa wa-l-Siyà˙a wa-l-Irshàd al-Qawmì, 1967–68), 702; Azdì, Mawßil, 238, where he cites the Ràfi∂ì proof text man kuntu mawlàhu fa-'alì mawlàhu against a Khàrijite, possibly meaning no more than that the Khàrijite was at fault for speaking ill of 'Alì. 50 Ps.-Nàshi" in J. van Ess, ed., Frühe mu'tazilitische Häresiographie (Beirut/Wiesbaden: Frank Steiner, 1971), §47; Akhbàr al-dawla al-'abbàsiyya, 165; al-Nawbakhtì, Firaq alshì'a, ed. H. Ritter (Istanbul: Staatsdruckerei, 1931), 43, cf. 42; Sa'd b. 'Abdallàh al-Qummì, Kitàb al-maqàlàt wa-l-firaq, ed. M.J. Mashkùr (Tehran: Óaydarì Press, 1963), 65f. 51 Al-Qummì, Kitàb al-maqàlàt, 65.4 (taken by Nawbakhtì, 42, to refer to another pro-'Abbàsid sect), 66.1 (missing in Nawbakhtì). 52 Al-ˇabarì, Ta"rìkh, III, 749; cited in Zaman, Religion and Politics under the Early 'Abbasids (Leiden: E.J. Brill, 1997), 57f. 53 Zaman, Religion and Politics, 58f., sees it as happening already under Hàrùn. 49

180

patricia crone The caliphs and their policies

In agreement with the Imàmì literature in general,54 our author depicts 'Umar as a staunch protagonist of Arabism and 'Alì as a universalist. His most sensational allegation is that 'Umar would have killed all mawàlì who had reached the height of five spans if it had not been for Ziyàd’s intervention. In a Sunnì version of what is probably the same story, it is Ziyàd who wants to cull them, or who tells of how Mu'àwiya wanted to do so. “I see that these non-Arabs (al-˙amrà") have grown numerous,” Ziyàd here declares, “and I suspect they will attack the Arabs and the government, so I have decided to kill half of them and to leave the other half to maintain the nobles (al-ashràf ) and repair the roads”;55 or Mu'àwiya says, “I see that they are reviling our pious ancestors (al-salaf ) and suspect that they will attack the Arabs and the government, so I think I shall kill part of them and leave part of them to maintain the markets (al-aswàq) and repair the roads.”56 The fact that the non-Arabs revile the salaf shows that they are Shì'ìs here too, but it is characteristic that 'Umar has nothing to do with the plans for their elimination in these versions. To the Shì'ìs, 'Umar was an Arab chauvinist whose policies were continued by the Umayyads; to the Sunnìs, 'Umar was a universalist whose policies the Umayyads had changed. That the Umayyads were Arab chauvinists was taken for granted by both. In historical terms, the Shì'ite view of 'Umar (suitably toned down) is the more plausible, for even the mainstream tradition contains material in which he is a great protagonist of the Arabs.57 It is difficult to see how he could have been anything else. He lived his entire life in tribal Arabia, where practically the only non-Arabs were slaves, freedmen and sundry non-tribesmen under Arab protection, whom it was natural to hold in contempt; and God Himself seemed to have a preference for Arabs, to whom He sent His prophet and 54

Cf. above, notes 10, 12, 29. Al-Ràghib al-Ißfahànì, Mu˙à∂aràt al-udabà" (Beirut: Dàr Maktabat al-Óayyàt, 1961), I, 350, reading aqtulu for aqbalu. 56 Ibn 'Abd Rabbih, 'Iqd, III, 413; translated (slightly differently) in B. Lewis, Islam from the Prophet Muhammad to the Capture of Constantinople (New York/Oxford: University Press, 1987), II, 204 (drawn to my attention by Michael Cook). 57 Cf. Madelung, The Succession to Mu˙ammad (Cambridge: Cambridge University Press, 1997), 74ff.; Kister, “‘The Crowns of this Community . . .’. Some Notes on the Turban in the Muslim Tradition,” Jerusalem Studies in Arabic and Islam 24 (2000), 221f. 55

MAWÀLÌ

and the prophet’s family

181

whom He allowed to conquer the world. Non-Arabs captured in the wars of conquest were of course free to convert and soon did, but they can hardly have started doing so on a significant scale already in 'Umar’s time, since it was only under him that the wars of conquest began in earnest, and it would in any case have been very odd if conversion had sufficed to grant them the same status as their captors: it was after all by fighting against God that they had lost their freedom. Deserters like the Asàwira who joined the Arabs of their own accord were rewarded with high pay and quasi-tribal status, but the vast majority of non-Arab Muslims entered the Muslim community as defeated slaves. It seems unlikely that 'Umar should have given much thought to their position if and when they became free Muslims, let alone that he should have granted them the same status as their captors. Pace the Shì'ites, however, it also seems unlikely that 'Alì’s views should have been significantly different from 'Umar’s. All the first caliphs were Arab tribesmen who had imbibed contempt for nontribesmen as part of growing up and whose pride in their own ethnicity had been reinforced by the religious developments culminating in the conquests; none of them is likely to have had an attitude to non-Arabs significantly different from that of their subjects, and prejudice against mawàlì ran from the top to the bottom of Arab society. If 'Alì had been ahead of his time on this issue, one would have expected him to have been derided for his views, but the sources preserve no recollection of this. Everything suggests that it was only when non-Arabs converted on a significant scale that their status became an object of debate. The Umayyads were cast as protagonists of the Arabs because they presided over a conquest society in which discrimination against native converts was a fact of life, not because they led the way by discriminatory policies and legislation. Sulaym (as we may continue to call our author) is of course right that the regulations of military pay must have been instituted by caliphal decree, probably by Mu'àwiya rather than 'Umar (§4);58 but it is hard to believe that

58 Apart from the Asàwira and other deserters, who became confederates (˙alìfs), not clients, there were no non-Arab soldiers worth speaking of in 'Umar’s time. The captives had to be manumitted in order for the question of their position in Muslim society (including the army) to arise, cf. Crone, Roman, Provincial and Islamic Law, 89ff.

182

patricia crone

Mu'àwiya, or any other early caliph, could have ordered the conquerors to share their fay" with their former captives on an equal basis without causing violent outrage among their subjects. Similarly, Mu'àwiya (rather than 'Umar)59 may be responsible for the one-sided nature of walà", which granted the patron a legal claim to his client’s estate without awarding a corresponding right to the client (§4), but no Arab is on record as having found this unfair. The remaining “legislation” by 'Umar and Mu'àwiya adduced by Sulaym does not seem in fact to be legislation at all. Arabs were indeed reluctant to give their daughters in marriage to non-Arabs, and both 'Umar I and 'Umar II are said to have shared this prejudice, and even to have forbidden such unions.60 But the reports on these two caliphs are a-historical, and it is not clear that any caliph took it upon himself to lay down the law on this issue. In so far as one can tell, it was up to the marriage guardians whether to accept such unions or not. Similarly, there does not seem to have been any formal rules regarding the appointment of clients to positions of authority (§4). In fact, the Umayyads and their governors would appear frequently to have been ahead of their subjects in their willingness to employ clients, much as they may have regretted the necessity.61 Their nonArab appointees were typically greeted with gross abuse from the Arabs of whom they were put in charge. The reader is meant to approve, not to feel outraged, when a Óimßì tradition has Ka'b alA˙bàr declare that, “If I were put in charge of the captives in your hands, I would kill nine out of every ten.” Here the vision of a massacre is meant for people who deplored the impact of non-Arabs on Muslim society rather than those who were outraged by the prejudice against them.62 And even after the 'Abbàsid revolution, people found it perfectly plausible that a massacre at Mosul should have

59

Cf. the preceding note. For 'Umar I, see 'Abd al-Razzàq, Mußannaf, VI, no. 10324 (widely cited elsewhere as well); Ibn Abì Shayba, Al-Mußannaf, ed. M.A. al-Nadwì (Bombay: al-Dàr al-Salafiyya, 1979–83), XII, no. 12522; cf. no. 12516, where he predicts that the Arabs will perish when the children of Persian mothers grow up. For 'Umar II, see Abù 'Ubayd, Gharìb al-˙adìth, II, 219; Ibn ManΩùr, Lisàn al-'Arab (Bùlàq: Ma†ba'at al-Mìriyya, 1301 AH), X, 103f., s.v. “†abi'.” 61 Cf. Crone, “Were the Qays and Yemen of the Umayyad Period Political Parties?” Der Islam 71 (1994), 12ff. 62 Nu'aym b. Óammàd, Fitan, no. 655 (ed. Zakkàr, 140). 60

MAWÀLÌ

and the prophet’s family

183

been triggered by protests at the appointment of a client as governor to that overwhelmingly Arab town.63 'Umar and 'Alì, on the other hand, were built up as universalists by scholars, many of them mawàlì themselves, who disapproved of the attitudes of their contemporaries. As the Umayyads typified the imperfect present in which they lived, so these two caliphs exemplified the ideal as practiced in the past, for it was of course inconceivable that the ideal should be new: if universalism was right now, it had always been right; only practice could have changed. By the midUmayyad period there cannot have been much doubt that Islam was universalist in the sense of directed to all human beings; perhaps there never was. But this is universalism in a loose sense, for it still left the question whether access to Mu˙ammad’s God required membership of Arab society (in which case the Arabs remained a chosen people) or, if the Arabs were merely a subgroup within Muslim society, whether they were to enjoy a privileged status within it (allowing their chosen status to persist to a greater or lesser degree). One has grant Sulaym that the early Shì'ites come across as more genuine universalists than their jamà'ì counterparts, for whereas 'Umar serves as the mouthpiece for a spectrum of views ranging from endorsement of the special position of the Arabs in both Sunnì and Shì'ite literature to outright rejection of it in the Sunnì literature alone, 'Alì seems to be depicted as an enemy of Arab privilege with great regularity in Shì'ite and Sunnì literature alike, suggesting that this was a point on which Shì'ites had agreed since early times. This is also suggested by the fact that the superiority of Arabs for purposes of marriage was rejected in all early Shì'ite law, whether Zaydì,64 Imàmì,65 or Ismà'ìlì,66 though the Zaydìs were later to accept

63 Azdì, Mawßil, 145f.; cf. C. Robinson, Empire and Elites after the Muslim Conquest: The Transformation of Northern Mesopotamia (Cambridge: Cambridge University Press, 2000), chaps. 6–7. 64 See Zayd b. 'Alì (attrib.), Majmù' al-fiqh, ed. E. Griffini (Milan, 1919), 199 (probably c. 750–70), with the commentary by al-Óusayn b. A˙mad al-Siyàghì (d. 1221/1806f ), Kitàb al-raw∂ al-na∂ìr (ˇà"if: Maktabat al-Mu"ayyad, n.d.), IV, 259ff. 65 Al-Kulaynì, Kàfì, 339–46; al-Mufìd, Al-Muqni'a (Qum: Mu"assasat al-Nashr alIslàmì, 1410 AH), 512; al-Hillì, Sharà"i' al-islàm, ed. 'A.M. 'Alì (Najaf: Ma†ba'at alArbàb, 1969), II, 299f.; cf. also above, note 10; Marlow, Hierarchy and Egalitarianism in Islamic Thought (Cambridge: Cambridge University Press, 1997), 33f. 66 Al-Nu'màn b. Mu˙ammad, Da'à"im al-islàm, ed. A. Fay∂ì (Cairo: Dàr al-Ma'àrif, 1951–60), II, nos. 728–31.

184

patricia crone

it.67 This is of some importance inasmuch as it is by refusing to give their women in marriage to other groups, and to accept food from them, that elites turn themselves into a castes. No Arabs to my knowledge are on record as having refused to eat the food of mawàlì, but a great many refused to give their daughters to them, as noted already, and the religious scholars had to decide whether this should be condemned as jàhilì prejudice or on the contrary endorsed as good Islamic practice with reference to the chosen status of the Arabs. Most jamà'ì Muslims endorsed it.68 All Shì'ites by contrast condemned it, clearly including Sulaym.

Mawàlì and Shì'ism Sulaym’s view that there was a natural affinity between mawàlì and Shì'ism, once shared by Western Orientalists, has long been out of favor on the grounds that it was among the Arabs that Shì'ism began and that it was only by historical accident at a late stage in history that Iran became a Shì'ite country. Both points are correct, of course, but Sulaym’s testimony suggests that as far as the Umayyad period is concerned there is something to the old view after all. Clients were converts to the beliefs of foreign invaders who had destroyed their polities and who continued to treat them with appalling arrogance even after they had adopted the invaders’ creed. It is a situation which a modern reader ought to find immediately intelligible, for it is closely paralleled by that created by the European expansion in the nineteenth and early twentieth centuries. Like the Arabs, the Europeans destroyed indigenous polities and converted their victims to their own beliefs (in the form of secular modernity or Christianity, occasionally both) without accepting the converts as full members of their own societies. A native who adopted the religion and/or secular culture of the British did not become a Brit or a full member of the wider Christian/Western society but rather a “Westernized Oriental Gentleman,” or “wog” for short. Similarly, a 67 See for example A˙mad b. Ya˙yà b. al-Murta∂à, Al-Ba˙r al-Zakhkhàr (Beirut: Mu"assasat al-Risàla, 1947), IV, 48f.; al-Qàsim b. Mu˙ammad b. 'Alì, Al-I'tißàm bi˙abl allàh al-matìn (Ían'à": Maktabat al-Yaman al-Kubrà, n.d.), III, 251ff. (my thanks to Bernard Haykel for drawing this to my attention). 68 For the Sunnì rules of marriage equality, see Y. Linant de Bellefond, art. “Kafà"a,” EI 2 and the literature cited there.

MAWÀLÌ

and the prophet’s family

185

native who converted to Islam and adopted the language and culture of the Arabs did not become an Arab or a full member of the wider Muslim society, but rather a mawlà, a term for which “wog” sometimes suggests itself as an effortless translation. In both cases the conquerors lacked a universalist high culture capable of uniting people in a single, educated elite transcending ethnicity (such as emerged after the 'Abbàsid revolution). “As for what you say about the sunna of the Prophet, what do you know about that?” as Salm b. A˙waz al-Tamìmì, a pillar of the Umayyad regime, is supposed to have put it to the Hàshimiyya in Khuràsàn; “we are more likely to understand it than you are, for we are Arabs and descendants of Arabs whereas you are barbarian rabble and worshippers of cats.”69 Like the Westernized elites of modern times, Muslim non-Arabs did not belong fully to either native society or that of the conquerors. Rejoining native society was rarely an option. One way or the other they had to make it in their new society, and they could only do so by severing the link between the truth and the conquerors to whom they owed it. By the mid-Umayyad period, both Khàrijism and Shì'ism offered ways of effecting such a severance. The Khàrijites famously denied that the truth had anything to do with ethnicity by proposing that God’s deputy could be of any descent: the most meritorious person was the most God-fearing (Q. 49:12), not necessarily an Arab or a Qurashì. But there were also ways in which Khàrijism seemed to restate rather than to reject tribal values,70 and Sulaym is probably right that clients were strongly drawn to the Prophet’s family. This is not surprising. For one thing, they could cast the members of that family as fellow-victims of the existing order: 'Umar had maltreated both of them, as Sulaym explains (§9), and as the clients in al-Mukhtàr’s revolt and the adherents of the Hàshimiyya in Khuràsàn agreed: Jàhilì-minded Arabs had excluded both the Prophet’s kinsfolk and his many non-Arab devotees from their rightful position in Muslim society.71 For another thing, the Hàshimites transcended ethnic divisions simply by being kinsmen of the Prophet, whether victimized 69

Akhbàr al-dawla al-'abbàsiyya, 287. Some of them are discussed in P. Crone and F. Zimmermann, The Epistle of Sàlim b. Dhakwàn (Oxford: University Press, 2000), 215ff. 71 Cf. Crone, “The Significance of Wooden Weapons in al-Mukhtàr’s Revolt and the 'Abbàsid Revolution,” in I.R. Netton, ed., Studies in Honour of Clifford Edmund Bosworth (Leiden: E.J. Brill, 2000), I, 176, 179f. 70

186

patricia crone

or not. They formed a sacred lineage so greatly elevated above Arabs and non-Arabs alike that the differences between the two were drained of importance. Sunnìs might consider love of the Arabs to be part of the faith, but to Shì'ites, it was love of the imàms that was mandatory, and it left no room for love of a special people as well. This is the point Sulaym captures so well with the statement that but for the first two caliphs; the entire umma would have been clients of the Hàshimites. Modern Shì'ites sometimes deny that they attach any importance to descent at all: the imàms were the most outstanding people in their own right, regardless of their kinship with the Prophet, they will say. This argument is also found in early Islam, for there were Zaydìs who declared 'Alì to have been the most meritorious man of his time irrespective of his kinship with the Prophet, which was not in itself a source of merit in their view.72 This was a coherent position for a universalist to take, but it is not easy to see how it could be squared with a sharp focus on àl rasùl allàh al-mu†ahharùn, and in practice it was easier to reject ethnic particularism by means of a holy lineage than by simply declaring descent to be irrelevant. Mawàlì and the 'Abbàsid revolution Sulaym adds yet another testimony to the growing mound of evidence that the Hàshimite revolution was once seen as a non-Arab revolt by friends and foes alike. He presents it as a movement in which the clients turned the tables on their captors: 'Alì had foretold that the non-Arabs (al-a'àjim) would one day beat the Arabs in the name of religion as the Arabs had originally beaten them, and that the non-Arabs in question would come from Khuràsàn (§10). Ibn al-Faqìh, who also knew 'Alì’s prediction of a role reversal, similarly held it to have been fulfilled when the Khuràsànìs “beat the Arabs out of anger on behalf of God’s religion and in disapproval of the behavior of the Umayyads.”73 That the roles of conquerors and victims were being reversed was what the Hàshimite soldiers themselves were signaling by using wooden clubs: the slaves were

72 See for example al-Jà˙iΩ, “Maqàlat al-zaydiyya wa-l-ràfi∂a,” in 'A.-S. M. Hàrùn, ed., Rasà"il (Cairo: Maktabat al-Khànjì, 1965–79), IV, 311ff., 317. 73 See the reference given above, note 30.

MAWÀLÌ

and the prophet’s family

187

rebelling as free and rightly guided Muslims against their erstwhile captors, who had once been rightly guided conquerors entitled to their power, but who had now turned into wrongdoers deserving of enslavement themselves.74 When the qà"im came, there would be nothing but the sword between him and the Arabs, as the Shì'ites continued to say after they had stopped regarding the Hàshimite revolution as a fulfillment of their dreams.75 The interpretation of the revolution as a role reversal is also found in an 'Abbàsid development of the “Yemen/Mu∂ar/Rabì 'a/the mawàlì ” trope, though it is played down in another. The first is Ibràhìm al-Imàm’s famous order that Abù Muslim should kill all Arabs in Khuràsàn. “Look to this tribe of Yemen and honor them and settle among them, for God will not complete this matter except through them,” Ibràhìm begins, implicitly assuming Yemenìs to be Shì'ites:76 they had fought with 'Alì against Mu'àwiya, as Sulaym explains (§3). “Look to this tribe of Rabì'a and be suspicious of them,” he continues, here with reference to the idea that the Rabì'a were prone to Khàrijism and fitna because they were angry with God for sending the Prophet to Mu∂ar. “And look to this tribe of Mu∂ar, they are the enemy close by,” he concludes with implicit reference to the fact that most governors of Iraq and Khuràsàn in the later Umayyad period were Mu∂arìs;77 “so kill anyone about whom you have doubts, of whom you are suspicious, and of whose reliability you are not sure. If you are able to leave no speaker of Arabic in Khuràsàn, then do so, and kill any youth you distrust who has reached five spans.”78 The instruction is mentioned again in connection with Ibràhìm al-Imàm’s death, this time as a straightforward order to kill all Arabs/speakers of Arabic.79 In short, Yemenìs were 74 Crone, “Wooden Weapons,” 178f., 180f. For arguments to the contrary, see Elad, “The ethnic composition of the 'Abbàsid Revolution: a reevaluation of some recent research,” Jerusalem Studies in Arabic and Islam 24 (2000), esp. 275ff. 75 Al-Nu'mànì (Ibn Abì Zaynab), Al-Ghayba (Beirut: Mu"assasat al-A'lamì li-lMa†bù'àt, 1983), 154f. 76 Al-ˇabarì, Ta"rìkh, II, 493, 1858; III, 329. 77 Cf. the tables in Crone, “Qays and Yemen,” 11f., 17f., 28–30, with 54f. It is also reflected in al-Madà"inì’s version of the trope, above, note 9. 78 Al-ˇabarì, Ta"rìkh, II, 1937; Azdì, Mawßil, 65 (slightly apocopated, but cf. 107.3); Ibn al-Athìr, al-Kàmil fì l-ta"rìkh (Beirut: Dàr al-Íàdir, 1982), V, 348; cf. Elad, “Ethnic Composition,” 255ff. My thanks to Amikam Elad for reminding me of the five spans here. 79 Al-ˇabarì, Ta"rìkh, II, 1974; III, 25; Azdì, Mawßil, 107.3; Akhbàr al-dawla al"abbàsiyya, 392.3; al-Madà"inì in al-Dhahabì, Ta"rìkh al-islàm (Cairo: Maktabat Dàr al-'Arùba, n.d.), V, 202, and in id., Siyar a'làm al-nubalà" (Beirut: Mu"assasat al-Risàla,

188

patricia crone

to replace Mu∂arìs as the ruler’s favorites, and it was Arabs rather than mawàlì over the height of five spans who were to be massacred. It is noteworthy that the Yemenìs do not seem to count as Arabs in this story.80 Presumably they were seen as Arabs by ethnic origin alone, not in terms of cultural outlook, for Arabs in the disapproving sense were people who attached importance to their Arab descent as a source of special merit in, and knowledge of, Islam.81 Differently put, they were Arabs such as they had been back in tribal Arabia before the rise of Islam rather than Muslims as they ought to be now: drawn from any ethnic origin and united by a shared religion and high culture, in which Arabic provided the common language as English does in the global culture of today.82 “God has removed the pride of the Jàhiliyya and its boasting about ancestors from you,” as a Shì'ite version of the Farewell Oration put it: “Arabic is not the father or mother of anyone, merely a language.”83 All members of the Hàshimiyya, whatever their origin, were Shì'ite universalists committed to the view that ethnic origin was a mere accident of birth of no religious or cultural consequence. The Muslim community had however been dominated by Arab chauvinists such as 'Umar and the Umayyads since the Prophet died, according to Sulaym. It was by eliminating such Arabs that the clients turned the table on their captors. The other version of the trope has Mu˙ammad b. 'Alì instruct his first missionary to “settle among the people of Yemen, conciliate Rabì'a, beware of Mu∂ar . . . and recruit many non-Arabs (wastakthir al-a'àjim), for they are the people of our mission and God will assist it through them.”84 Here a more positive attitude to the angry

1982–85), VI, 59 (where the order is in a letter intercepted by Marwàn II); Ibn 'Abd Rabbih, 'Iqd, IV, 479 (slightly different); al-Dìnawarì, Al-Akhbàr al-†iwàl, ed. W. Guirgass (Leiden: E.J. Brill, 1888), 358 (where it is Abù l-'Abbàs who passes the order to Abù Muslim after Ibràhìm al-Imàm’s death). 80 Highlighted by Sharon, Revolt. The Social and Military Aspects of the 'Abbàsid Revolution ( Jerusalem: Institute of Asian and African Studies, 1990), 53, who sees a contradiction here. 81 As in Salm’s statement quoted above, note 69. 82 For other examples of the same usage, see Crone, “Wooden Weapons,” 179f. 83 Al-Kulaynì, Kàfì, VIII, 246; Qà∂ì Nu'màn, Da'à"im, II, no. 729. 84 Akhbàr al-dawla al-'abbàsiyya, 204.8; reduced to “settle among Yemen and be kind to Mu∂ar” in Madà"inì’s version in al-ˇabarì, Ta"rìkh, II, 1501.8. For another version, put in the mouth of Ibn al-Óanafiyya, see Ibn 'Abd Rabbih, 'Iqd, IV, 476, where the mawàlì are missing.

MAWÀLÌ

and the prophet’s family

189

Rabì'a is adopted, and although non-Arabs are singled out for special importance, there is no order to kill Mu∂ar or the Arabs in general. On the contrary, as Abù Muslim later assured Naßr b. Sayyàr, although it was true that the imàm had ordered him to settle among Yemen, conciliate Rabì'a, beware of Mu∂ar and pay special attention to the a'àjim, it was only the Mu∂arì followers of the Umayyads that he was meant to beware of, not the upright ones; the result was that non-Arabs flocked to the da'wa, closely followed not just by Yemen and Rabì'a, but by some Mu∂arìs of insight as well.85 Here, then, the role reversal is belittled, presumably to reassure Mu∂arì members of the 'Abbàsid elite (who were actually numerous) that there was no need to feel alienated, the constant harping on the theme of Yemenìs and non-Arabs in the 'Abbàsid revolution notwithstanding. Clearly, all the many versions of the “Mu∂ar/Rabì'a/Yemen/ clients” trope give us later interpretations, not factual details.86 Like the predictions attributed to 'Alì, they summarize what they see as the meaning of the revolution, and their pithy formulation suggests that they were used for purposes of propaganda. But the idea that non-Arab Muslims played a major role in the Hàshimite movement cannot be dismissed as mere propaganda intended to assure nonMuslims that there was room for them in the new establishment, for the enemies of the revolution had the same impression. To the Umayyads and their supporters, the revolutionaries were runaway slaves admixed with lowly clients and Arab riffraff,87 longhaired people whose genealogies were their villages and whose slogan was Persian bekosh, bekosh, “kill, kill,”88 native rabble from the back and beyond, uncircumcised pagans who worshipped fire, cats, anything

85 Akhbàr al-dawla al-'abbàsiyya, 284f. Discussed (in a vein unpersuasive to me) by Sharon, Revolt, 58f.; Elad, “Ethnic Composition,” 258f. 86 For a sample (including some cited here), see Crone, “Qays and Yemen,” 55f., with 48f. 87 Nu'aym b. Óammàd, Fitan, no. 565 (ed. S. Zakkàr, 122). For runaway slaves in the revolution, see also Omar, 'Abbàsid Caliphate, 96; add 'Abd al-Óamìd b. Ya˙yà, Rasà"il, ed. I. 'Abbàs (Amman: Dàr al-Shurùq li-l-Nashr wa-l-Tawzì', 1988), 199 (letter no. 8). 88 Nu'aym b. Óammàd, Fitan, nos. 550–52, 558 (ed. Zakkàr, 118, 120); cf. Akhbàr al-dawla al-'abbàsiyya, 206.1, where the fact that their genealogies are their villages is singled out for praise. The participants in the revolution were identified by their villages rather than their tribes in Abù Muslim’s register (al-ˇabarì, Ta"rìkh, II, 1969).

190

patricia crone

except God and Mu˙ammad, and who had come to kill the Arabs along with upright mawàlì and to destroy Islam.89 In a more dispassionate vein Óamza al-Ißfahànì observes that it was the 'ajam of Khuràsàn along with miscellaneous Arabs and Bedouin who transferred power from the Umayyads to the 'Abbàsids.90 His statement makes explicit what everyone knew to be the case, namely that many participants in the revolution, including most of the leaders, were ethnic Arabs, who moved on to a play a dominant role in the caliphate down to the fourth civil war (though the early 'Abbàsid caliphate struck many as non-Arab in character, too).91 It is on the participation of ethnic Arabs that the modern literature has concentrated for the past thirty years or so, with salutary effect.92 But they were a minority, and what was so remarkable about the revolutionaries to contemporaries was in any case their ability to transcend ethnicity: Arabs or Iranians, all were clients of the Hàshimites, as Sulaym would have said. None of this does anything to explain why it was Khuràsànìs who took it upon themselves to enact the role reversal. Sulaym’s cata89 Cf. Naßr b. Sayyàr’s poems (in Dìnawarì, Akhbàr, 360; Ibn A'tham, Kitàb alfutù˙ (Hyderabad: Dà"irat al-Ma'àrif al-'Uthmàniyya, 1968–75), VIII, 161–63; Mas'ùdì, Murùj, IV, 2286 (ed. Paris, VI, 62); Ibn 'Abd Rabbih, 'Iqd, IV, 478f.); 'Abd al-Óamìd b. Ya˙yà’s letters to Marwàn II (in his Rasà"il, nos. 8, 38); and Nu'aym b. Óammàd, Fitan, no. 565 (ed. S. Zakkàr, 122). The charges of cat worship and the like are refuted in Akhbàr al-dawla al-’abbàsiyya, 282, 287, 290ff. Compare the traditions in S. Bashear, Arabs and Others in Early Islam (Princeton: Darwin Press, 1997), 103. Cf. also al-Maqrìzì, Al-Nizà' wa-l-takhàßum (Cairo: Maktabat al-Ahràm, n.d.), 66; Al-Maqrìzì’s Book of Contention and Strife, trans. C.E. Bosworth (Manchester: Journal of Semitic Studies, 1980), 88 (107): the 'Abbàsid took over with the help of the 'ajam ahl Khuràsàn; they overthrew the Umayyads with 'ajam Khuràsàn. Cf. now also S.S. Agha, The Revolution which Toppled the Umayyads: neither Arab nor 'Abbàsid (Leiden: Brill, 2003). 90 Óamza al-Ißfahànì, Ta"rìkh sinì mulùk al-ar∂ wa-l-anbiyà" (Beirut: Maktabat alÓayyàt, 1961), 160 (chap. 10, section 9). 91 Thus an anti-'Abbàsid tradition in Nu'aym b. Óammàd, Fitan, no. 538, where Ibn 'Abbàs tells people to honor the Persians when black banners arrive, “for our regime will be based on them ( fa-inna dawlatanà fìhim)”; cf. also nos. 543–4, 547, on the theme of “woe to the Arabs” and dhill al-'arab after the third civil war (ed. Zakkàr, 116–18). Compare al-Jà˙iΩ, Al-Bayàn wa-l-tabyìn, ed. 'A.-S.M. Hàrùn (Cairo: Maktabat al-Khànjì, 1948–50), III, 366 (their dawla was 'ajamiyya khuràsàniyya); Ibn Óazm in Ibn 'Idhàrì, Kitàb al-bayàn al-mughrib, ed. G.S. Colin and É. Lévi-Provençal (Leiden: E.J. Brill, 1948–51), I, 63f. (similarly); Mas'ùdì, Murùj, V, §3446; (ed. Paris, VIII, 291f.) where the Arabs lose out to al-Manßùr’s clients and slaves rather than to Khuràsànìs. 92 The foremost representative of this line of research is now Elad, whose emphasis on the Arab nature of the revolution sometimes seems a little excessive (cf. his “Ethnic Composition” and his contribution to the present volume).

MAWÀLÌ

and the prophet’s family

191

logue of discriminatory practices is of interest for reflecting the grievances of people who were doing well in Muslim society, not those of villagers and other humble people oppressed by Arab fiscal demands.93 Villagers found it difficult even to gain acceptance as members of Muslim society. In so far as they avoided being deported as illegal immigrants, their standard complaint about the Umayyad army was that they could not get into it, not that they were paid less than Arabs there, and they are not likely to have chafed under their inability to secure government positions or to marry Arab women. By contrast, there is a story of a mawlà, the son of a captive from Sìstàn, who was offended when he was appointed governor of Kirmàn: he had set his eyes on Khuràsàn and attributed his failure to receive the governorship of that province to his mawlawiyya.94 Marwànid society was full of such men who would have made it to the top if they had not been non-Arabs, or who actually did make it to the top, but who still had to endure insults and prejudicial treatment from Arabs less able or educated than themselves. It is mawàlì of this type who were likely to sympathize with the Hàshimiyya. But it is impossible to say more without going into conditions specific to Khuràsàn, and on that there is nothing to be learnt from Sulaym’s letter.

A context for the letter? Abù Muslim was rumored to have killed all the Arabs of Rabì'a, Mu∂ar, Nizàr and Yemen who remained in Khuràsàn after Naßr b. Sayyàr’s flight and Qa˙†aba’s departure for Iraq.95 This is obviously an exaggeration. There is however every reason to believe that Abù Muslim engaged in purges, and the exaggeration was presumably born of shock and outrage at their magnitude. What right did he have to kill all these people? Khuràsànì Arabs who had not openly committed themselves to the Umayyad cause could not automatically be assumed to be supporters of the old regime. His answer presumably 93 Wellhausen, who saw the revolution as a non-Arab revolt triggered by social grievances, had oppressed people in mind (Wellhausen, The Arab Kingdom and its Fall (London: Curzon Press, 1973); (German original, 1902), chaps. 8–9). 94 Al-ˇabarì, Ta"rìkh, II, 1458 (year 104). For his ancestry, see Khalìfa, Ta"rìkh, 168 (year 29). 95 Al-'Uyùn wa-l-˙adà"iq, ed. M.J. de Goeje (Leiden: E.J. Brill, 1869), 193.13.

192

patricia crone

was that if they had not joined the da'wa either, their loyalty to the Hàshimiyya was equally uncertain. It may well have been in justification of his brutal treatment of such people that Ibràhìm al-Imàm was credited with his famous waßiyya regarding the elimination of Arabs from Khuràsàn: Abù Muslim was only following orders; the imàm (who was always right) had instructed him to kill any Mu∂arì “about whom you have doubts, of whom you are suspicious, and of whose reliability you are not sure,” and “to leave no speaker of Arabic in Khuràsàn,” but rather to “kill any youth you distrust who has reached five spans.” Either the formulation was meant to recall 'Umar’s order regarding the extermination of mawàlì or else the story about 'Umar (and Ziyàd or Mu'àwiya’s) order regarding the extermination of mawàlì was brought into circulation in order to justify Abù Muslim’s killing of Arabs. Either way, Sulaym could well have composed his particular account of 'Umar’s order in response to criticism of the murderous nature of the new regime. This would have the merit of explaining why he should have found it worth his while to put together a letter calculated to work its readers into a state of raging fury at a dynasty that had already fallen.

Bibliography Primary Sources 'Abd al-Óamìd b. Ya˙yà. Rasà"il, ed. I. 'Abbàs. Amman: Dàr al-Shurùq li-l-Nashr wa-l- Tawzì') 1988. 'Abd al-Razzàq. Al-Mußannaf, ed. Ó.-R. al-A'Ωamì. Beirut: al-Majlis al-'Ilmì, 1972. Abù Óàtim al-Ràzì. A'làm al-nubuwwa, ed. Í. al-Íàwì and Gh.M. A'wànì. Tehran: Imperial Iranian Academy of Philosophy, 1977. Abù l-Qàsim al-Kùfì. Kitàb al-istighàtha. n.p. [Maktabat Niniwì l-Óadìtha], n.d. Abù 'Ubayd. Al-Amwàl, ed. M. K. Haràs. Cairo: Maktabat al-Kulliyàt al-Azhariyya, 1969. Abù 'Ubayd. Gharìb al-˙adìth. Hyderabad: Ma†ba'at Majlis Dà"irat al-Ma'àrif al'Uthmàniyya, 1964–67. Akhbàr al-dawla al-'abbàsiyya wa-fihi akhbàr al-'abbàs, ed. 'A.'A. al-Dùrì and 'A.J. alMu††alibì. Beirut: Dàr al-Talì'a li-l-ˇibà'a wa-l-Nashr, 1971. al-Azdì. Ta"rìkh al-mawßil, ed. 'A. Óabìba. Cairo: Lajnat I˙yà" al-Turàth al-Qawmì, 1968. al-Balàdhurì. Futù˙ al-buldàn, ed. M.J. de Goeje. Leiden: Brill, 1866. id., Ansàb al-ashràf, IV a, ed. M.J. Kister. Jerusalem: Magnes Press, 1971. Dà"irat al-ma'àrif al-islàmiyya al-kubrà, II. Tehran: Markaz Dà"irat al-Ma'àrif alIslàmiyya, 1995. al-Dhahabì. Siyar a'làm al-nubalà". Beirut: Mu"assasat al-Risàla, 1982–85. id., Ta"rìkh al-islàm. Cairo: Maktabat Dàr al-'Arùba, n.d. al-Dìnawarì, Al-Akhbàr al-†iwàl, ed. W. Guirgass. Leiden: Brill, 1888. al-Fa∂l b. Shàdhàn. Al-Ì∂à˙. Beirut: Mu"assat al-A'lamì li-l-Ma†bù'àt, 1982.

MAWÀLÌ

and the prophet’s family

193

Óamza al-Ißfahànì. Ta"rìkh sinì mulùk al-ar∂ wa-l-anbiyà". Beirut: Maktabat al-Óayyàt, 1961. al-Hillì. Sharà"i' al-islàm, ed. 'A.M. 'Alì. Najaf: Ma†ba'at al-Arbàb, 1969. Ibn 'Abd Rabbih. Al-'Iqd al-farìd, ed. A. Amìn et al., 2nd print. Cairo: Lajnat alTa"lìf wa-l- Tarjama wa-l-Nashr, 1965. Ibn Abì l-Óadìd. Shar˙ nahj al-balàgha, ed. M.A.-F. Ibràhìm. Cairo: 'Ìsà Óalabì, 1965–67. Ibn Abì Shayba. Al-Mußannaf, ed. M.A. al-Nadwì. Bombay: al-Dàr al-Salafiyya, 1979–83. Ibn A'tham. Kitàb al-futù˙. Hyderabad: Ma†ba'at Majlis Dà"irat al-Ma'àrif al'Uthmàniyya, 1968–75. Ibn al-Athìr. Al-Kàmil fì l-ta"rìkh. Beirut: Dàr al-Íàdir, 1982. Ibn al-Faqìh. Kitàb al-buldàn, ed. M.J. de Goeje. Leiden, 1885. Ibn 'Idhàrì. Kitàb al-bayàn al-mughrib, ed. G.S. Colin and E. Lévi-Provençal. Leiden: Brill, 1948–51. Ibn ManΩùr. Lisàn al-'Arab, X. Bùlàq: Ma†ba'at al-Mißriyya, 1301 AH. Ibn al-Murta∂à, A˙mad b. Ya˙yà. Al-Ba˙r al-zakhkhàr. Beirut: Mu"assasat al-Risàla, 1947. Ibn Sa'd. Kitàb al-†abaqàt al-kabìr, ed. E. Sachau et al. Leiden: Brill, 1904–17. Ibn Qudàma. Al-Mughnì, ed. T.M. al-Zaynì et al. Cairo: Maktabat al-Qàhira, 1968–70. al-Jà˙iΩ. Al-Bayàn wa-l-tabyìn, ed. 'A.-S.M. Hàrùn. Cairo: Maktabat al-Khànjì, 1948–50. id., “Maqàlàt al-zaydiyya wa-l-ràfi∂a,” in his Rasà"il, ed. 'A.-S. M. Hàrùn. Cairo: Maktabat al-Khànjì, 1965–79. Khalìfa b. Khayyà†. Ta"rìkh, ed. S. Zakkàr. Damascus: Ma†àbi' Wizàrat al-Thaqàfa wa-l-Siyà˙a wa-l-Irshàd al-Qawmì, 1967–68. al-Kha†ìb al-Baghdàdì. Ta"rìkh baghdàd. Cairo: Ma†ba'at al-Sa'àda, 1349 AH. al-Kulaynì. Al-Kàfì, ed. 'A.A. al-Ghaffàrì. Tehran: Dàr al-Kutub al-Islàmiyya, 1375–77 AH. al-Majlisì. Bihàr al-anwàr. Beirut: Dàr al-Ri∂à, 1983. al-Maqrìzì. Al-Nizà' wa-l-takhàßum fìmà bayna banì umayya wa-banì hàshim. Cairo: Maktabat al-Ahràm, n. d.; al-Maqrìzì’s Book of Contention and Strife, trans. C.E. Bosworth. Manchester: Journal of Semitic Studies, 1980. al-Marzubànì. Akhbàr shu'arà" al-shì'a, ed. M.H. al-Amìnì, 2nd print. Beirut: Shirkat al-Kutubì, 1993. al-Mas'ùdì. Murùj al-dhahab, ed. C. Pellat. Beirut: al-Jàmi'a l-Lubnàniyya, 1966–79. al-Mufìd. Al-Muqni'a. Qum: Mu"assasat al-Nashr al-Islàmì, 1410 AH. Nawbakhtì. Firaq al-shì'a, ed. H. Ritter. Istanbul: Staatsdruckerei, 1931. Nu'aym b. Óammàd. Al-Fitan, ed. M. al-Shùrì. Beirut: Dàr al-Kutub al-'Ilmiyya, 1997. al-Nu'màn b. Mu˙ammad. Da'à"im al-islàm, ed. A. Fay∂ì. Cairo: Dàr al-Ma'àrif, 1951–60. al-Nu'mànì (Ibn Abì Zaynab). Al-Ghayba. Beirut: Mu"assat al-A'lamì li-l-Ma†bù'àt, 1983. Pseudo-Nàshi": see van Ess. al-Qàsim b. Mu˙ammad b. 'Alì. Al-I 'tißàm bi-˙abl allàh al-matìn, vol. 3. Ían'à": Maktabat al-Yaman al-Kubrà, n.d. al-Qummì, Sa'd b. 'Abdallàh. Kitàb al-maqàlàt wa-l-firaq, ed. M.J. Mashkùr. Tehran: Óaydarì Press, 1963. al-Ràghib al-Ißfahànì. Mu˙à∂aràt al-udabà". Beirut: Dàr Maktabat al-Óayyàt, 1961. al-Siyàghì, al-Óusayn b. A˙mad. Kitàb al-raw∂ al-na∂ìr. ˇà"if: Maktabat al-Mu"ayyad, n.d. Sulaym b. Qays (attrib.). Kitàb Sulaym b. Qays al-Hilàlì, ed. M. Bàqir Anßàrì. Qum: Nashr al- Hàdì, 1995.

194

patricia crone

al-ˇabarì. Jàmi' al-bayàn fì tafsìr al-qur"àn. Beirut: Dàr al-Fikr, 1988. id., Ta"rìkh al-rusul wa-l-mulùk, ed. M.J. de Goeje et al. Leiden: Brill, 1879–1901. Al-'Uyùn wa-l-˙adà"iq, ed. M.J. de Goeje. Leiden: Brill, 1869. Wakì'. Akhbàr al-qu∂àh, ed. 'A.M. al-Maràghì. Beirut: 'Àlam al-Kutub, n.d. (originally Cairo, 1947–50). al-Ya'qùbì. Ta"rìkh, ed. M.T. Houtsma. Leiden: Brill, 1883. al-Zamakhsharì. Al-Kashshàf. Beirut: Dàr al-Kitàb al-'Arabì, n.d. [1947]. Zayd b. 'Alì (attrib.). Majmù' al-fiqh, ed. E. Griffini. Milan, 1919. Secondary Sources S.S. Agha, The Revolution which Toppled the Umayyads: neither Arab nor 'Abbàsid (Leiden: Brill, 2003). Bashear, S. Arabs and Others in Early Islam. Princeton: Darwin Press, 1997. Caskel, W. ]amharat an-nasab. Das Genealogische Werk des Hi“àm ibn Mu˙ammad al-Kalbì. Leiden: Brill, 1966. Cook, M. Commanding Right and Forbidding Wrong in Islamic Thought. Cambridge: University Press, 2000. Crone, P. “The 'Abbàsid Abnà" and Sàsànid Cavalrymen,” Journal of the Royal Asiatic Society 8 (1998). id., Meccan Trade and the Rise of Islam. Princeton: University Press, 1987. id., ‘The Pay of Client Soldiers,’ Der Islam 80 (2003). id., Roman, Provincial and Islamic Law. Cambridge: University Press, 1987. id., “The Significance of Wooden Weapons in al-Mukhtàr’s Revolt and the 'Abbàsid Revolution,” in I.R. Netton, ed., Studies in Honour of Clifford Edmund Bosworth. Leiden: Brill, 2000, I. id., “Were the Qays and Yemen of the Umayyad Period Political Parties?” Der Islam 71, 1994. id. and F. Zimmermann. The Epistle of Salim b. Dhakwan. Oxford: University Press, 2000. EI 2: Encyclopaedia of Islam, 2nd ed. Leiden, 1960–. Elad, A. “The ethnic composition of the 'Abbàsid Revolution: a reevaluation of some recent Research,” Jerusalem Studies in Arabic and Islam 24 (2000). van Ess, J. ed. Frühe mu'tazilitische Häresiographie. Beirut/Wiesbaden: Frank Steiner, 1971. Goldziher, I. Muhammedanische Studien. Halle: Max Niemeyer, 1889–90. Kister, M.J. “The Crowns of this Community . . .”. Some Notes on the Turban in the Muslim Tradition,” Jerusalem Studies in Arabic and Islam 24 (2000). Lewis, B. Islam from the Prophet Muhammad to the Capture of Constantinople. New York/ Oxford: University Press, 1987. Madelung, W. “The Hàshimiyyàt of al-Kumayt and Hàshimì Shì'ism,” Studia Islamica 70 (1989). id., The Succession to Mu˙ammad. Cambridge: University Press, 1997. Marlow, L. Hierarchy and Egalitarianism in Islamic Thought. Cambridge: University Press, 1997. Lane, E.W. An Arabic-English Lexicon. London: Williams and Norgate, 1863–93. Omar, F. The 'Abbàsid Caliphate. Baghdàd: National Printing and Publishing Co., 1969. Robinson, C. Empire and Elites after the Muslim Conquest: The Transformation of Northern Mesopotamia. Cambridge: University Press, 2000. Sharon, M. Revolt. The Social and Military Aspects of the 'Abbàsid Revolution. Jerusalem: Institute of Asian and African Studies, 1990. Stern, S.M. “The “Book of the Highest Initiation” and Other Anti-Ismà'ìlì Travesties,” in S.M. Stern, ed., Studies in Early Ismà'ìlism. Jerusalem/Leiden: Magnes Press, 1983. Wellhausen, J. The Arab Kingdom and its Fall. London: Curzon Press, 1973 (German original, 1902). Zaman, M.Q. Religion and Politics under the Early 'Abbàsids. Leiden: Brill, 1997.

MAWÀLÌ AND MUWALLADÙN IN AL-ANDALUS (SECOND/EIGHTH-FOURTH/TENTH CENTURIES) Maribel Fierro

Introduction The onomastic information contained in biographical dictionaries sheds some light on the ethnic landscape of al-Andalus, how it changed through the centuries and its relationship with Andalusì social history. A study I carried out on this topic was based on the onomastic data of three biographical dictionaries of Andalusì scholars.1 Two were written in the fourth/tenth century: Ibn Óàrith alKhushanì (d. 361/971) lived during the Cordoban Umayyad caliphate, as did Ibn al-Fara∂ì (d. 403/1013), who also witnessed the 'Àmirid rise to power and the ensuing collapse of the Cordoban Umayyad caliphate. The third biographical dictionary was written in the sixth/twelfth century by Ibn Bashkuwàl (d. 578/1183), who lived under the Berber dynasties of the Almoravids and the Almohads. In my study, I took into account a point made by L. Molina namely, that scholars who appear in Andalusì biographical dictionaries often can be gathered into families, so that Andalusì biographical dictionaries do not give information only about individuals. In Ibn alFara∂ì’s dictionary, which contains 1650 biographies, 20% can be put in the context of families or lineages, or in other words, those 1650 individuals belong to 1300 families.2 The results all refer to the second through fourth century AH/eighth–tenth century CE and they are as follows.

1

See Fierro, “Árabes, beréberes, muladíes y mawàlì. Algunas reflexiones sobre los datos de los diccionarios biográficos andalusíes,” in M. Marín, ed., EOBA, VII (Madrid, 1995), 41–54. See also Guichard, Al-Andalus. Estructura antropológica de una sociedad islámica en Occidente (Barcelona, 1976), 301–302 (reprint Granada, 1998). 2 See Molina’s articles “Familias andalusíes: los datos del Ta"rìj 'ulamà" al-Andalus de Ibn al-Fara∂ì,” in M.L. Ávila, ed., EOBA, II (Granada, 1989), 19–99; EOBA, III (Granada, 1990), 13–58; and in L. Molina, ed., EOBA, IV (Granada, 1990), 13–40.

196

maribel fierro

Absence of ethnic affiliation Approximately half of the biographies contained in those three dictionaries are characterized by absence of any ethnic nisba, as well as absence of any indication of ethnic affiliation and of ties of clientship. This percentage is higher in the oldest dictionaries (47% and 52%) and lower in the latest (42%), which seems to indicate a tendency to acquire an ethnic nisba as time goes by (see below, Arabic nisbas). When biographees with no nisba have a short nasab (because they are recent converts) they also tend to have no kunya. Most scholars with no nisba were part of the indigenous population of the Iberian Peninsula that converted without establishing walà" ties with Arab tribes. This convert population is called by Andalusì sources musàlima, musàlimat ahl al-dhimma and muwalladùn, terms I shall deal with later. Not all the scholars with no nisba are to be understood in this way. Sometimes a scholar appears in biographical dictionaries lacking a nisba, but if he was a member of a scholarly family that absence could be supplied by the onomastic data provided in the biographies of other members of the same family. Clients of Arab women bore no nisba, as happened in the case of the famous traditionist Baqì b. Makhlad (d. 276/889), who is said to have been the descendant of the mawlà of a woman from Jayyàn ( Jaén).3 In the case of the Berber Nàßi˙ b. Yltìt b. Qa†arì, who is said to have been the slave of the daughter of one al-Thaqafì al-Jazìrì, he did bear the nisba al-Thaqafì.4 Although there is no mention of his manumission having entailed walà", he probably acquired his nisba from a male member of his owner’s family who inherited him from her.5

3 Another case is that of 'Amr b. 'Abd Allàh (d. 273/886), considered by some sources to have been the first qà∂ì of mawlà origin, was the descendant of a client of one of the daughters of 'Abd al-Ra˙màn I. No source gives him any nisba. See on him L. Molina, “An Arab among Muwallads: Mu˙ammad ibn 'Abd al-Salàm al-Khushanì,” in M. Marín, ed., The Formation of al-Andalus. Part 1: History and Society, The Formation of the Classical Islamic World, vol. 46 (Aldershot: Ashgate Variorum, 1998), 125 (English translation of “Un árabe entre muladíes: Mu˙ammad b. 'Abd al-Salàm al-Ju“anì,” Estudios onomástico-biográficos de al-Andalus 4 (1994), 337–51). See on the issue on when the first non-Arab judges were named in al-Andalus notes 49 and 81. 4 See H. de Felipe, Identidad y onomástica de los beréberes de al-Andalus (Madrid, 1997), 187. 5 Women do not inherit clients, see Ibn al-'A††àr (d. 399/1009), Kitàb al-wathà"iq wa-l-sijillàt, ed. P. Chalmeta and F. Corriente (Madrid, 1983), 271 and its Spanish translation, Formulario notarial y judicial andalusí, ed. P. Chalmeta and M. Marugán

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

197

Arabic nisbas The two dictionaries of the fourth/tenth century include approximately 30% biographies with an Arabic nisba (30% and 32%), whereas the dictionary written in the sixth/twelfth century included 48% biographies with an Arabic nisba. That nisba tends to be not so much a traditional Arabic tribal nisba, but an “Islamic” nisba 6 such as “alAnßàrì.”7 The growing predominance of the nisba al-Anßàrì (to which I hope to devote a future study) becomes apparent from the fifth/ eleventh century onwards and seems to be related with the growing importance of an Islamic identity over and against an Arabic identity.8 The use of “religious” nisbas can be traced back to the third/ninth century, as illustrated by the case of the traditionist Thàbit b. Óazm (d. 313/925). His Berber ancestor was a client (walà" 'alàqa)9 of the Banù Zuhra, as were all the Berbers of the Upper Frontier of alAndalus. But Thàbit had problems with his Arab patrons and decided not to continue bearing the nisba al-Zuhrì, adopting instead the nisba “al-'Awfì,” implying that clientship had been established (through conversion) with a descendant of the Prophet’s Companion 'Abd

(Madrid, 2000), 468. Oliver (in her article quoted below, note 75) suggests that the term mawlà might have been applied to individuals who were integrated into families by means of women, without having really established any walà" ties. It would be the case of Fà"iq (or Rà"iq), brother of the jàriya Íub˙, wife of the caliph alÓakam II, who is called in the sources mawlà of al-Óakam II. In this case, his consideration as a mawlà seems to derive solely from the relationship of his sister with the caliph. 6 In the version of this article presented at the Nijmegen Conference “Mawàlì in the early Islamic period” (May 2001), I used the expression “religious nisba” to refer to cases such as al-Anßàrì or al-'Awfì (see below), i.e., cases in which individuals or groups show a tendency to acquire a nisba associated specifically with the Islamic experience or history and not with Arabic tribalism. Objections made by some of the participants, especially by Ella Landau-Tasseron, have made me change it to “Islamic nisba.” The issue deserves more attention and I hope to devote a study to it. 7 Descendants of the famous Prophet’s Companion Sa'd b. 'Ubàda l-Anßàrì are said to have settled in al-Andalus. 8 I indicated this phenomenon in my article “Los mawàlì de 'Abd al-Ra˙màn I,” Al-Qan†ara 20 (1999), 67, note 6. See also the example given by de Felipe, Identidad, 63. On the fa∂à"il of the Anßàr, see Juynboll, “Shu'ba b. al-Óajjàj (d. 160/776) and his position among the traditionists of Baßra,” Le Muséon 111 (1998), 211–12. 9 I understand this term as equivalent of walà" al-muwàlàt. See Fierro, “Los mawàlì de 'Abd al-Ra˙màn I,” 90. I have not found the term in Crone, Roman, provincial and Islamic law. The origins of the Islamic patronate (Cambridge, 1987).

198

maribel fierro

al-Ra˙màn b. 'Awf al-Zuhrì who would have entered the Iberian Peninsula during the time of the conquest.10 The increase in the number of Arabic nisbas recorded by Ibn Bashkuwàl has to be related to the small percentage of mawàlì mentioned in his work, which in many cases corresponds to the fact that former mawàlì were now seen as being Arab tout court. Berber nisbas The number of Berber nisbas is very low, around 1.5% in the dictionaries of the fourth/tenth century, with an increase in the dictionary written in the sixth/twelfth century (2.1%). This small increase seems to reflect the waves of “new” Berbers who settled in the Iberian Peninsula after the second half of the fourth/tenth century. These “new” Berbers did not assimilate into the common “Andalusì” identity as the old Berbers did. Mawàlì The percentage of mawàlì in the two oldest dictionaries is around 8% (8.8% and 7.4%). It decreases dramatically in the one written in the sixth/twelfth century: only 1.7%. As I have mentioned before, this decrease should be related with the increase in the number of Arabic nisbas. There is also an increase in the number of people bearing the nisba al-Umawì (7.3% in Ibn Bashkuwàl, as compared with only 4.6 and 0.6% in the biographical dictionaries written in the fourth/tenth century). Most of the scholars with the nisba alUmawì were not members of the Umayyad family, but clients. The increase in the percentage should be understood in the sense that by the fourth/tenth century, the number of clients of Arab tribes or individuals decreased and there was a parallel increase in the number of Umayyad mawàlì (I shall deal with this process below) that is reflected in those biographical dictionaries written after the fall of the Cordoban Umayyad caliphate.

10 See on this Ibn Khayr, Fahrasa, ed. F. Codera and J. Ribera, 2 vols. (Zaragoza, 1893) (BAH, IX–X), 193; Molina, “Familias,” EOBA, II, 69–70.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

199

The scarce percentage of mawàlì will point to an “Andalusì uniqueness” compared to what happened in the central Islamic lands.11 The first part of this article will be devoted to the characteristics of walà" ties in al-Andalus. The second part will deal with the converts without walà" called in the historical sources muwalladùn. The third part will be devoted to ethnic solidarity in al-Andalus as reflected in scholars’ biographies.

Were there mawàlì in al-Andalus? Before dealing with the Andalusì case, I shall summarize briefly the Sunnì legal doctrine on walà".12 There are two kinds of walà". Patronate arising upon manumission (walà" al-'itàqa or al-'itq)13 implies that the manumitter acquires a title to the freedman’s estate as the latter’s last agnate, that he is obliged to pay blood-money on behalf of the freedman (assuming that the latter has no agnates of his own), that he may act as a marriage guardian to his freedwoman or freedman’s daughter, and according to the Màlikìs he also qualifies for the custody of the freedman’s children. For his part, the freedman does not acquire any title to his manumitter’s state, he does not qualify for membership of the manumitter’s blood-money group (although some Màlikìs do in the absence of all other candidates), he cannot act as marriage guardian to the manumitter’s daughter (although Màlik is said to have allowed it) and he might not be awarded custody of the manumitter’s children (Màlikìs permitted it in the absence of all other candidates). In practice, freedmen were both expected and known to assist their manumitters, be it financially, politically or otherwise. The relationship between manumitter and freedman is as enduring as are kinship ties: it cannot be terminated by the manumitter or the freedman, being inherited by their descendants.

11 The absence of mawàlì is even more characteristic of Morocco, a region which had a very specific pattern of conversion still awaiting a thorough study. But then very few Arabs settled in Morocco in the first centuries, whereas Arabic settlement in al-Andalus was not negligible. See on this settlement (with figures) Guichard, AlAndalus, 442–58 and Chalmeta, Invasión e islamización. La sumisión de Hispania y la formación de al-Andalus (Madrid, 1994), 122–27. 12 I closely follow Crone, Roman, provincial and Islamic law, 36–40. 13 It is also referred in some sources as walà" in'àm/ni'ma, see Fierro, “Árabes,” 53, and Fierro, “Los mawàlì de 'Abd al-Ra˙màn I,” note 52.

200

maribel fierro

In pre-classical law there existed walà" al-islàm, that is, a tie which arose on conversion at the hands of another, as well as walà" almuwàlàt, some other agreement or association between a Muslim and a non-Arab.14 Of the classical legal schools, only Óanafìs have maintained walà" al-muwàlàt (rejecting as do the other schools walà" alislàm), the fundamental requirement of which is that the client must be somebody entirely devoid of blood relations and patrons within the Muslim community. This kind of patronate implies that the patron acquires a title to the freedman’s estate (but not as an agnatic relation), that he is obliged to pay blood-money on behalf of the client, that he may act as a marriage guardian to his client’s daughter. For his part, the client acquires a title to his patron’s estate if the parties have stipulated that succession be mutual, and he assumes responsibility for the payment of the blood-money on behalf of the patron if the parties have stipulated that his obligation be mutual. This tie may be terminated by either party as long as the patron has not paid blood money on the client’s behalf. Why was walà" almuwàlàt rejected (only Óanafìs kept it) and in particular walà" alislàm? Crone explains this rejection as caused by the loss of the tribal roots of Arab society, so that the subjection of non-Arab Muslims to client status ceased to have much point and it rapidly came to be regarded as offensive; at the hands of the 'ulamà" the institution was to be partly reshaped and partly rejected. On the one hand, the tie was reinterpreted as one of kinship rather than dependence, and it is to this interpretation that it owes most of its classical features. And on the other hand the 'ulamà" increasingly ruled that free converts were exempt from it, or, differently put, that they became mawàlì in the sense of kinsmen of the Muslims at large . . . the tendency to interpret walà" as a kinship tie is present in our earliest material: the reshaping of the institution must have been under way by the 720s at the latest. The rejection of walà" over free converts seems to have followed somewhat later, being first attested for lawyers who died in the 760s.15

14

See also Crone, Slaves on horses. The Evolution of the Islamic Polity (Cambridge, 1980), 49. 15 Crone, Roman, provincial and Islamic law, 90. N. Calder has pointed that walà" al-muwàlàt deserves further discussion, see Studies in early Muslim jurisprudence (Oxford, 1993), 205. The Andalusì data on this kind of walà" are sparse and of difficult interpretation. An example is the case of 'Àßim al-'Uryàn al-Thaqafì, called muwàlì of the Umayyad 'Abd al-Ra˙màn I, see my article “Los Banù 'Àßim al-Ôaqafì, antepasados de Ibn al-Zubayr,” Al-Qan†ara 7 (1986), 56.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

201

It is commonly said that one of the differences in the relationship between Arab and non-Arab Muslims in al-Andalus compared to that of the East is that walà" played virtually no role in it. The reasons for it would have been the Berber component in the conquering army, the absence of garrison cities (amßàr) and closely connected to the latter, the way conversion usually took place. The Berbers As Crone reminds us, al-Andalus lacked the purely Arab conquest elite characteristic of the East: many of the conquerors were Berbers, and such ties of walà" as they had with Arab patrons lost all significance when they acquired the status of conquerors themselves.16 H. de Felipe’s study on Berber onomastic data and identity in alAndalus during the Umayyad period shows that not all Berbers of the first “wave” (the so-called “old Berbers,” those who settled in alAndalus during the first period of Muslim rule) fit into the same patterns of Arabization and Islamization. The Berber majority settled in tribal groups in rural areas, mainly in the Marches. Very little, not to say nothing, is known of the more humble groups, as those who appear in the Arabic sources are the elites, especially those who acted as umarà" al-thaghr. The latter are usually mentioned because of their involvement in armed conflict with the Cordoban ruler or his representatives in the frontier areas. Sometimes, this armed conflict assumed a religious outlook characterized by leadership of a charismatic figure associated with Messianic expectations, as was the case of the Berber al-Fà†imì during the emirate of 'Abd al-Ra˙màn I (years 151/768–169/777)17 and of the Mahdì Ibn al-Qi†† (an Umayyad), who rebelled against the Umayyad amìr 'Abd Allàh with a Berber army (year 288/900).18 Some of these Berbers settled in rural areas managed to preserve their tribal identity, reflected in onomastic usages, and later became rulers of ˇà"ifa kingdoms in the fifth/eleventh century.

16

Crone, art. “Mawlà,” EI 2, section on Muslim Spain. See Fierro, “On al-fà†imì and al-fà†imiyyùn,” Jerusalem Studies in Arabic and Islam 20 (1996), 135–36. 18 See Fierro, La heterodoxia en al-Andalus durante el periodo omeya (Madrid: I.H.A.C., 1987), 111–13. 17

202

maribel fierro

Those Berbers who lived in urban settlements, having entered the state administration or become scholars, are differentiated from the Berber rural elites by their onomastics, showing a higher degree of Arabization and Islamization. Berber names appear only in the oldest part of the nasab; there is a high frequency of Arabic or geographical nisbas, with a corresponding very low frequency of Berber nisbas; the kunya element is present. This situation is especially so during the caliphal period, thus indicating that by then ethnic or tribal affiliation with Berber origins was not functional among urban elites. This does not mean that the memory of those origins was lost, as shown by the fact that those Berber urban elites were often employed in missions dealing with Berber elements both in al-Andalus and North Africa,19 as well as acting as “intermediaries” where there was conflict between other ethnic groups.20 Out of 59 Berber lineages reconstructed by de Felipe, information of walà" ties with Arab tribes is given of only five.21 Three Berber lineages have an Arabic nisba, but the sources do not refer to them as mawàlì.22 The term mu˙àlafa or ˙ilf (alliance) is mentioned twice, referring to the Banù Abì 'Ìsà (see below) and to the family of the Banù l-Muß˙afì who bore the nisba al-Qaysì.23 Patronate ties with the Umayyads are recorded in seven cases.24 No mention of walà" nor of an Arabic nisba is given for the majority of cases.25 The sources from which the above mentioned information has been extracted are late sources and although they might preserve

19

See de Felipe, Identidad, 77–79. The Berber Abù Mùsà al-Hawwàrì was named imàm of the mosque of Écija when there were tensions between Arabs and muwallads in the times of 'Abd alRa˙màn II (206/822–238/852), see de Felipe, Identidad, 246. 21 See de Felipe, Identidad, 95–100 (al-'Awfì), 122–25 (Banù l-Faraj, ties with the Banù Makhzùm), 146–57 (Banù Abì 'Ìsà l-Laythì), 249–52 (Banù Abì Zamanìn alMurrì), 187–92 (they bore the nisba al-Thaqafì because an ancestor had been the slave of the daughter of one al-Thaqafì). Of the Berbers of the Upper Frontier is said that they had all established ties of walà" 'alàqa with the Banù Zuhra. See above note 10; and Manzano, La frontera de al-Andalus en época de los omeyas (Madrid: CSIC, 1991), 213. 22 See ibid., 102–103, 105–108, 169–71. See also Fierro, “Árabes,” 46. 23 De Felipe, Identidad, 177–86. We shall see below (cf. note 29) that this term could be understood as the way in which late sources referred to cases of walà" almuwàlàt. 24 De Felipe, Identidad, 108–114, 126–27, 157–61, 163–65, 172–73, 174–75 and 230–38. 25 See for example de Felipe, Identidad, 118–122, 125–26, 132–35, 135–36, 161–63, 186, 196–97, 198–200, 200, 201–220, 225, 225–27, 227–28, 242, 253–67. 20

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

203

old strata of information, they were written in a very specific context, that of the Cordoban Umayyad caliphate, when (real or alleged) walà" ties with the Umayyads grew in number and importance and at the same time walà" ties with Arabic tribes lost relevance. This implies that from those sources we cannot know how important walà" ties were in the first two centuries of Muslim presence in the Iberian Peninsula. Those Berbers who settled in lands rejected by the Arabs, and therefore had no Arabs as neighbors, probably soon lost the memory of any ties they could have had with Arab patrons. The Berbers who settled in the vicinity or among Arabs probably preserved for a longer time the memory of those ties, as happened for example in the Upper Frontier. It was in the second half of the third/ninth century when the Berber scholar Thàbit b. Óazm decided not to use the nisba al-Zuhrì that the Berbers of the Upper Frontier bore and to change it to the Islamic nisba al-'Awfì. An example of preserving the memory of walà" ties with an Arab tribe is provided by the case of the Berber family Banù Abì 'Ìsà who bore the nisba al-Laythì.26 To that family belonged the famous scholar Ya˙yà b. Ya˙yà, credited with the introduction of Màlik’s Muwa††a" in al-Andalus. Only a few biographical dictionaries (none earlier than the fourth/tenth century) record the fact that he was a Maßmùdì, while usually they all record his Arabic nisba, al-Laythì. The historian al-Ràzì (d. 344/955)27 states that Ya˙yà b. Ya˙yà’s ancestor, Wislàs, had converted to Islam in North Africa at the hands of a Yazìd b. 'Àmir al-Laythì (Layth Kinàna, Mu∂ar), became his mawlà and therefore bore the nisba al-Laythì. This story is upheld by a later source against another version implying that there was also manumission, i.e., that Wislàs was a slave ([ Ya˙yà b. Ya˙yà] laythì bi-l-walà" wa-qad ya'taqidu man là 'ilm lahu bi-annahu walà" al-'itàqa walaysa kadhàlika innamà huwa walà" l-islàm aslama Wislàs 'alà yaday Yazìd ).28 The Mafàkhir al-Barbar do not mention walà", but conversion and

26 See Marín, “Una familia de ulemas cordobeses: los Banù Abì 'Ìsà,” Al-Qan†ara 6 (1985), 291–320; and Fierro, “El alfaquí beréber Ya˙yà b. Ya˙yà, ‘el inteligente de al-Andalus’,” in M.L. Ávila and M. Marín, eds., Estudios Onomástico-Biográficos de al-Andalus, vol. 8 (Granada/Madrid, 1997), 269–344. 27 See 'Iyà∂, Tartìb al-madàrik, 8 vols. (Rabat, . . .–1983), III, 379, quoting al-Ràzì al-Istì'àb (Kitàb fì ansàb mashàhìr ahl al-Andalus). From this and other sources it is evident that Ya˙yà b. Ya˙yà’s family relationship with the Arab Banù l-Layth was a matter of perplexity. 28 Al-Tujìbì (d. 730/1329), Barnàmaj, ed. 'A.Ó. Manßùr (Tunis, 1981), 56–57.

204

maribel fierro

alliance (˙ilf ) established between Ya˙yà’s ancestor and the al-Laythì at the hands of whom he had converted.29 Wislàs’ grandson (the father of our scholar) entered al-Andalus with the army of the Berber ˇàriq b. Ziyàd. Ya˙yà b. Kathìr b. Wislàs was therefore a baladì. The baladiyyùn were those Arabs (and their Berber clients) who settled in the Iberian Peninsula after the first wave of conquest. They are differentiated from those Arabs who came later (around the year 123/740) with the Syrian army defeated in North Africa by Berber troops. The latter are called shàmiyyùn or Syrians. Ya˙yà b. Kathìr b. Wislàs and his Arab patrons Banù Layth seem to have settled in the area of Algeciras and Sidonia before the Umayyad period. The first Umayyad emir entrusted Wislàs’ grandson with the governorship there and from then onwards, his descendants became important in the Umayyad entourage, whereas the Banù Layth fell into oblivion (they seem to have shown disaffection to 'Abd al-Ra˙màn I). Ya˙yà b. Ya˙yà was the first member of his family to become a scholar. He traveled to the East and is said to have met Màlik b. Anas in the last year of his life, therefore being able to hear from him the Muwa††a". Although Ya˙yà b. Ya˙yà never accepted any legal position such as judge, he had a huge influence on the nomination of legal positions in Cordoba, being a member of the shùrà, the advisory board that the ruler and the judge had to consult. Ya˙yà b. Ya˙yà and his descendants, known as the Banù Abì 'Ìsà, became influential members of the Cordoban notables and were extremely powerful during the Umayyad period, being connected with the introduction and predominance of the Màlikì school of law. They held important positions and had their own mawàlì. The Umayyads employed the Banù Abì 'Ìsà in missions with Berbers. This urban family preserved ties with their Maßmùda tribe: for example, Ya˙yà b. Ya˙yà sought refuge with them when he had to escape Cordoba after having been involved in the so-called al-Raba∂ revolt.30

29

“Mafàkhir al-barbar,” in M. Ya'la, ed. Tres textos árabes sobre beréberes en el Occidente islámico. Ibn 'Abd al-Óalìm (s. VIII/XIV), Kitàb al-ansàb. Kitàb Mafàjir al-barbar (Anónimo). Abù Bakr Ibn al-'Arabì (m. 543/1149), Kitàb “awàhid al-jilla (Madrid, 1996), 202. Ya˙yà b. Ya˙yà’s ancestor is called Shimlàl: wa-kàna islàm Shimlàl 'alà yad Yazìd b. 'Àmir al-Laythì fa-hum yantamùna ilaihi bi-l-˙ilf. 30 This was the revolt that would have caused the Umayyad amìr 'Abd al-Ra˙màn II realize that the Umayyads, from then onwards, had to count for the government of al-Andalus on the support of the emerging group of scholars. The institution of

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

205

We have here a Berber family connected with the conquest that preserved the memory of its walà" ties with an Arab tribe. Of the disparate versions concerning how those ties were established, the one favored was pointing to conversion, with no manumission involved. The Berber scholars al-'Awfì of the Upper Frontier were also interested in stressing the point that their walà" did not arise from manumission. Ya˙yà b. Ya˙yà was a scholar credited with having introduced Màlikì doctrine in al-Andalus. According to that doctrine, conversion without manumission does not imply walà" ties with the person at whose hands conversion took place. In other words, according to Màlikism, Ya˙yà b. Ya˙yà’s family should have acquired an Arabic nisba only in the case of their ancestor having been a converted manumitted slave. However, these cases should not be understood as attempts at denying the existence of slavery in the genealogy of the family. They are examples of legal practices belonging to a pre-Màlikì period, when patronate ties (with the subsequent acquisition of Arabic nisbas) were established through means other than manumission. Why was there in this case preservation of the memory of those ties? The Banù Abì 'Ìsà, for all their close connection with the Umayyads, are never presented as Umayyad mawàlì. A tentative answer is that Ya˙yà b. Ya˙yà’s family had used their walà" al-islàm ties with the Arab Banù Layth in order to obtain economic privileges when the latter fell in disgrace with 'Abd al-Ra˙màn I. But if that was so, we shall never know for sure. The converted indigenous population (musàlima) of al-Andalus The Arabs and Berbers who conquered the Iberian Peninsula, the baladiyyùn, settled all over the land and became landowners.31 When

the shùrà (advisory council formed by 'ulamà") would have been the outcome of such realization. See on this Marín, “”ùrà et ahl al-“ùrà dans al-Andalus,” Studia Islamica 62 (1985), 25–51; Fierro, “Ya˙yà b. Ya˙yà,” 289–98. 31 See Manzano, “The Settlement and Organisation of the Syrian Junds in alAndalus,” in M. Marín, ed., The Formation of al-Andalus. Part 1: History and Society, The Formation of the Classical Islamic World, vol. 46 (Aldershot: Ashgate Variorum, 1998), 85–114 (English translation of E. Manzano Moreno, “El asentamiento y la organización de los Ëund sirios en al-Andalus,” Al-Qan†ara 14 (1993), 327–59). On the disputed status of the conquered lands, see now Manzano, “Las fuentes árabes sobre la conquista de al-Andalus: una nueva interpretación,” Hispania 49/II (1999), 389–432.

206

maribel fierro

the Syrian junds came later on, an arrangement for them was found by the governor Abù l-Kha††àr al-Kalbì (125/743). He assigned those junds to specific districts: the jund of Damascus was settled at Elvira, the jund of Óimß at Sevilla and Niebla, the jund of Jordan at Rayyo (Málaga), the jund of Qinnasrìn at Jaén, the jund of Palestine at Algeciras and Sidonia, and the jund of Egypt at Beja and Tudmìr. The junds received a third of the taxes paid by the Christians. They most probably were put in charge of collecting them, thus obtaining fiscal control over the territories to which they were respectively assigned and becoming involved in the administration of the lands that they had occupied. After appropriating a third of the payments made by the Christians, the rest had, in principle, to be turned over to the central government. In addition, they were obliged to a tributary payment on the income they retained for their upkeep. The payment was made to the wàlì, representative of the Umayyad administrative apparatus. However, the hold of the central government over the Muslim population of al-Andalus decreased after the governorship of Abù l-Kha††àr, while the influence and power of the Syrians increased. When the Umayyad 'Abd al-Ra˙màn b. Mu'àwiya rose to power in al-Andalus, his policy was directed at countering the political dominance which the junds had acquired. He did it in different ways. The weakening of the junds was followed by the recruitment of Berber contingents who owed allegiance to the emir personally. The jund system was maintained, but with certain changes: The system of stipends in use in the Umayyad period confirms that the Arab military elements only received payment for campaigns they had conducted, and that receipt of such payments was regulated by a rigid hierarchy controlled by the sovereign. The absence of regular wages is explained by the existence of concessions—whether in the form of lands or taxes—which guaranteed the upkeep of these military elements.32

With the Muslims (be they Arabs or Berbers) settled in the land, there were not those garrison cities (amßàr)33 which in the central 32

Manzano, “Settlement,” 114. See Kennedy, Muslim Spain and Portugal. A political history of al-Andalus (London/New York: Longman, 1996), 16–18. For fortified spaces that might have been associated with the junds, see Acién, “Settlement and fortification in Southern al-Andalus: the formation of a land of ˙ußùn,” in M. Marín, ed., The Formation of al-Andalus. Part 1: History and Society, The Formation of the Classical Islamic World, vol. 46 (Aldershot: Ashgate Variorum, 1998), 368 (English translation of “Poblamiento y fortificación 33

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

207

Islamic areas “attracted dhimmì immigrants and caused Muslims to exclude from their ranks all converts without a patron” and therefore in al-Andalus conversion “did not normally involve either hidjra or walà", the converts adopting Islam wherever they happened to be.”34 By the fifth/eleventh century al-Andalus was overwhelmingly Muslim and by the end of the sixth/twelfth century there were no noticeable Christian communities left. R. Bulliet’s well known study on the conversion curve in Islamic lands35 has a section devoted to al-Andalus, according to which the high point of the curve is found there during the fourth/tenth century and prior to the middle of this century half the population of al-Andalus was still Christian. Bulliet’s approach to conversion in al-Andalus, although has been subject to various criticisms not yet presented in a systematic way,36 is generally accepted. As happens for other regions of the Islamic world, “conversion stories” are almost non-existent in Andalusì sources.37 Regarding the reasons for conversion and the ways in which it took place, most of them were the same that supported conversion in other Islamic lands.38 M. de Epalza has stressed one specific reason: the Church’s inability to cater for the officially Christian mass of the population at the moment of the conquest. This

en el sur de al-Andalus: la formación de un país de ˙ußùn,” Actas III Congreso de Arqueología Medieval Española (Madrid, 1989), 137–47). See also al-'Udhrì, Nußùß 'an al-Andalus min Kitàb tarßì 'al-akhbàr wa-tanwì' al-àthàr, ed. 'A. al-Ahwànì (Madrid, 1965), 56 (reference to a place called al-'askar because in it there have been a settlement of Arabs). 34 Crone, art. “Mawlà,” EI 2, section on al-Andalus. 35 Conversion to Islam in the Medieval Period. An Essay in quantitative history (Cambridge/ London, 1979), especially 114–27 (Spain). 36 See de Epalza, “Mozarabs: an Emblematic Christian Minority in Islamic alAndalus,” in M. Marín, ed., The Formation of al-Andalus. Part 1: History and Society, The Formation of the Classical Islamic World, vol. 46 (Aldershot: Ashgate Variorum, 1998), 192 (previously published in S.Kh. Jayyusi, ed., The Legacy of Muslim Spain (Leiden, 1992), 149–70); Penelas, “Some remarks on conversion to Islam in alAndalus,” Al-Qan†ara XXIII, 2002, 193–200. For myself, I have pointed that the onomastic data employed by Bulliet for establishing the time of conversion of Andalusì scholars should be known in order to ascertain their evidence: Fierro, “Árabes,” note 43. 37 See on this issue Bulliet, “Conversion stories in early Islam,” in M. Gervers and R.J. Bikhazi, eds. Conversion and continuity. Indigenous Christian communities in Islamic lands (Toronto, 1990), 123–33; and Morony, “The Age of Conversions: A Reassessment,” in M. Gervers and R.J. Bikhazi, eds., Conversion and continuity. Indigenous Christian communities in Islamic lands (Toronto, 1990), 135–50. 38 I shall come back to this issue below.

208

maribel fierro

inability reveals itself in two crucial areas of religious and political importance: baptism as a means of receiving new members into the Christian communities, and bishoprics as a means of supporting these communities and representing them vis-à-vis the Muslim authorities.39 Andalusì Christianity remained organized in some small urban and semi-urban communities as long as it could maintain a line of bishops. This was mainly the case in what is now Andalucía, where such communities were of symbolic value to the Cordoban political power and were protected accordingly, though there was also a strong impetus to convert to Islam due to social pressure from Muslims and a loss of cultural prestige on the part of the Latin Christendom.40 Conversion would, therefore, have been mainly due not so much to a voluntary passage to the new religion, but to an involuntary inability to continue being a Christian. Converts are referred to as musàlima or musàlimat ahl al-dhimma (the term muwalladùn usually associated with musàlima has a very special meaning, as we shall see later on). This is the term found in both historical sources and in biographical dictionaries.41 It refers to the indigenous Christian population that had been granted the dhimma status and at a certain point converted to Islam. As P. Chalmeta has pointed: “Probably due to the fact that, in historical terms, it is statically more exact to speak of the occupation of al-Andalus than of the conquest, the majority of the population were covered by a pact (mu'àhid ) that was far superior to that of ‘vanquished people’,

39 De Epalza, “Falta de obispos y conversión al Islam de los cristianos de alAndalus,” Al-Qan†ara 25 (1994), 385–400. 40 De Epalza, “Mozarabs,” 193. 41 The nisba al-Islàmì is extremely rare in biographical dictionaries, as shown by Penelas (see above note 36). It appears, however, in shurù† manuals when recording documents on conversion, as well as in historical sources. One such case is the ancestor of Ibn Óafßùn, referred to as Ja'far al-Islàmì, see Fierro, “Four questions in connection with Ibn Óafßùn,” in M. Marín, ed., The Formation of al-Andalus. Part 1: History and Society, The formation of the classical Islamic world, vol. 46 (Aldershot: Ashgate Variorum, 1998), 292 (English translation of “Cuatro preguntas en torno a Ibn Óafßùn,” Al-Qan†ara 16 (1995), 221–57). P. Chalmeta states that in al-Andalus “aslamì was used to designate the ex-Christian convert, whereas the term islàmì was reserved for the former Jew,” see Chalmeta, art. “Muwallad,” EI 2. But this distinction is not supported by the sources. Simonet, Historia de los Mozárabes de España (Madrid, 1897–1903), XVI, states that Christian converts to Islam were called, among other names, “Aslamies” and that the nisba Aslamì is often found in biographical dictionaries. But he confuses the tribal nisba al-Aslamì with al-Islàmì, a mistake that was made by the Andalusìs themselves as shown by a text from Ibn al-'A††àr, Formulario, 633, no. 163.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

209

a circumstance which implies the minor importance of the number of those who ‘submitted’, susceptible of being transformed into mawàlì.”42 Now, according to the settlement patterns of the conquerors described above, it is clear that the Muslim population would have had no interest in promoting the conversion of the dhimmìs because of the financial losses conversion would have meant for them. Each Arab “tribal” group might have been interested in increasing their own numbers,43 but that could be done through means other than conversion44 and that increase, in any case, must have been sought after within certain limits. The Muslim Arabs were military elites who depended for their privileged livelihood on the existence of a mass who, by being nor Muslim nor Arab, were not entitled to the same privileges. The aversion at increasing the number of converts must have been especially felt by the Syrians who, compared to the baladiyyùn, do not seem to have owned land (at least in the early period). After all, the Syrians’ livelihood depended on the Christian rural population in a more dramatic way than elsewhere in the Islamic world, as the Syrians’ stipends for waging war were extremely low. Discouragement on the part of baladiyyùn and Syrians for conversion would have probably meant a low number of walà" ties of the type walà" al-islàm, i.e., those arising from conversion. Walà" ties arising from manumission (walà" al-'itàqa) may have been higher than those arising from conversion as they were more beneficial for patrons

42 Chalmeta, art. “Muwallad,” EI 2, adds that this would be confirmed by the Wathà"iq of Ibn al-'A††àr, which stated that “‘in al-Andalus the walà" of conversion is not considered identical to that of manumission . . .’ The clientage of conversion does not entail personal connections or socio-economic obligations and its members are clients not of an individual but of the whole community. The judicial authorities only register this clientage (of conversion) in the case where the individual comes to reside ‘in a land where it is effective’.” I understand Ibn al-'A††àr’s text (ed., 410; trans., 637, no. 164) as implying that in al-Andalus no consensus has yet been reached in Ibn al-'A††àr’s times (fourth/tenth century) regarding the legal implications of conversion, especially in matters of inheritance. That is the reason that the name of the person “at whose hands” conversion took place is included by Ibn al-'A††àr in the document, because a legal case involving conversion might be judged by a qà∂ì holding the doctrine that walà" al-islàm is similar to walà" al-'itq or a convert might die in a place (balda) where people follow such a doctrine. See also notes 51 and 67. 43 After the arrival of the Syrians there was not any noticeable wave of Arab immigrants from other regions of Islamic world. 44 Children born from Arab tribesmen and local women or slave concubines were Arabs and not muwalladùn.

210

maribel fierro

than the latter from an economic point of view,45 but they would have not affected much the indigenous inhabitants of al-Andalus, because the conquest, as we have seen, did not entail enslavement.46 The relative scarce reflection in biographical and historical sources of cases of walà" al-'itq47 probably suffer from later embarrassment on the part of the family to have had a slave as an ancestor.48 Conversion could not be discouraged in the long run. Umayyad policy led to a weakening of Syrian power and, at the same time, promoted conversion to Islam by offering converts opportunities of social and economic advancement.49 One of the earliest legal texts of Màlikì orientation written in al-Andalus, al-'Utbiyya by al-'Utbì (d. 255/869) (who was an Umayyad mawlà), clearly shows that Andalusì jurists’ were in favor of facilitating the convert the process of conversion to Islam.50 Umayyad support and encouragement of conversion meant that the number of Muslims increased, but that increase did not involve an increase in the number of Arabs within the Andalusì population. This was so because the reception of Medinese

45 That was especially so in the early period, when walà" was regarded as a residue of ownership, not as an agnatic tie, see Fierro, “Los mawàlì de 'Abd alRa˙màn I,” 71 and note 19. 46 Cases of enslavement of the indigenous population were usually associated with rebellions, see Fierro, “Four questions,” 319–22; Kennedy, Muslim Spain, 25. 47 See below. 48 An example of denial of having had a slave as an ancestor in Ibn Óayyàn, Al-Muqtabis al-thànì, f. 129v (Spanish translation, p. 134). For a case that such denial did not take place, see Fierro, “Los mawàlì de 'Abd al-Ra˙màn I,” 79. 49 From the times of the amìr Mu˙ammad (238/852–273/886) onwards, it can be noticed that place was being made to Christian converts in positions of prestige and authority. It was then that non-Arabs were named judges (see above note 3 and below note 81). The Christian secretary Qùmis b. Antunyàn was active during the emirate of Mu˙ammad. Arab and mawàlì notables resented the presence of a Christian in the emiral khidma. Because of their pressure, Qùmis converted in order to keep his position, see Fierro, Heterodoxia, 77–80. He was a contemporary of the so-called Mozarab martyrs, whose movement of resistance as Christians shows the increasing Islamization of the society in which they lived. See on this point the recent studies by Wolf, Christian martyrs in Muslim Spain (Cambridge: Cambridge University Press, 1988) and Coope, The martyrs of Córdoba: community and family conflict in an age of mass conversion (Lincoln/London: University of Nebraska Press, 1995). 50 Their attitude contrasts with that of non-Andalusì jurists, see Fernández Félix, Cuestiones legales del Islam temprano: la 'Utbiyya y el proceso de formación de la sociedad islámica andalusí, Madrid: CSIC, 2003, 436–458; and Fernández Félix and Fierro, “Cristianos y conversos al islam en al-Andalus bajo los Omeyas. Una aproximación al proceso de islamización a través de una fuente legal andalusí del s. III/IX,” Anejos Archivo Español de Arqueología 23 (2000), 417–29, 428.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

211

(Màlikì) doctrine meant that free converts to Islam became Muslims without any need for an Arabic nisba, although it is not easy to establish when this tendency became predominant.51 Màlikism penetrated in al-Andalus during the third/ninth century and became its official doctrine in the fourth/tenth century.52 R. Brunschvig already indicated that Màlikism recognized no legally valid effect (in matters such as, for example, succession or penal solidarity) except in the case of patronage arising from the manumission of slaves.53 Free non-Arabs who converted to Islam were regarded under Màlikism as clients of the Muslim community as a whole, and not of the individual or group before whom they had renounced their former religion. They did not acquire Arabic tribal nisbas,54 something that must have been seen quite appropriate by the Umayyads and their ruling elites, as the loyalty of the converts could be shifted from tribes to the dynasty. The new Muslims and their descendants had no special legal status among the Màlikìs.55 The mawàlì Compared to the central areas of the Islamic world and according to the figures I have given in the introduction, al-Andalus lacked a sizeable number of mawàlì. The percentage of mawàlì in the two oldest biographical dictionaries is around 8% (8.8% in Ibn Óàrith alKhushanì and 7.4% in Ibn al-Fara∂ì). We shall never know the extent of the influence that in registering such low percentage had the fact that those sources were written during the period of the

51 The tendency started in the East around the 760s (Crone, Roman, provincial and Islamic law, 91), but as we shall see (note 67) it does not seem to have been fully implemented in al-Andalus till the end of the third/ninth century. 52 The idea that it became the “official doctrine” of al-Andalus in the third/ninth century should be abandoned. See Fierro, “Ya˙yà b. Ya˙yà,” 317, 337–38; and Fernández Félix, “Biografías de alfaquíes: la generación de al-'Utbì,” EOBA, VIII (1997), 141–75, 152–65. 53 See Lévi-Provençal, Histoire de l’Espagne musulmane, 3 vols. (Paris/Leiden, 1950–53), 117, note 74. See now also Fernández Félix and Fierro, “Cristianos y converses,” 428; and Ibn al-'A††àr, Formulario, 637 (no. 164). 54 Onomastic trends among converts in al-Andalus still await a thorough study. 55 This does not mean that “new Muslims” could not suffer legal discrimination regarding “old Muslims.” See Fernández Félix and Fierro, “Cristianos y converses,” 429.

212

maribel fierro

Cordoban Umayyad caliphate,56 that is, after the disappearance of the conquest society and the loss of influence of the Arab junds. There were mawàlì enlisted in the junds that settled in al-Andalus, but information about mawàlì other than those of the Umayyads is likely to have been mostly lost. Books on mawàlì that may have included early strata of information have not been preserved, such as that of Abù Mu˙ammad 'Abd Allàh b. 'Ubayd Allàh al-Azdì alÓakìm (var. al-Óukayyim) (d. 341/952), entitled Ansàb al-dàkhilìn ilà l-Andalus min al-'arab wa-ghayrihim or Kitàb alladhì dhukira fìhi al-khulafà" wa-man tanàsala minhum bi-l-Andalus wa-min sà"ir Quraysh wa-mawàlìhim wa-ahl al-khidma wa-l-taßarruf lahum wa-mashàhìr al-'arab al-dàkhilìn ilà l-Andalus min al-Mashriq min ghayr Quraysh wa-mawàlìhim wa-mashàhìr qabà"il al-barbar alladhìna ˙talù al-Andalus,57 or that of al-Ràzì (d. 344/ 955), Kitàb a'yàn al-mawàlì bi-l-Andalus. Patronate ties arising upon manumission must have existed for a long time, according to the information provided by the kutub alwathà"iq, where notary documents on legal situations connected with this kind of walà" ties are found until the sixth/twelfth century.58 This presence is evidently due to the fact that the Sunnì classical legal schools preserved in their legal doctrine a section dealing with walà" al-'itq. However, the information provided by biographical dictionaries on cases of walà" al-'itq is very scarce after the fourth/tenth century.59 Clients who were manumitted slaves seem to have had a low social status as Muslims or, at least, their origins could be remembered in order to put them at a disadvantage even when they had reached positions of relevance in society. At the same time, descendants of manumitted slaves could insist that walà" should be regarded as an agnatic tie and not as a residue of ownership and be suc-

56 See on how caliphal legitimacy and propaganda affected historical writing Martínez-Gros, L’idéologie omeyyade. La construction de la légitimité du Califat de Cordoue (X e–XI e siècles) (Madrid, 1992); and Safran, The second Umayyad caliphate. The articulation of caliphal legitimacy in al-Andalus (Cambridge, Mass./London, 2000), 111–83. 57 See M. b. Sharìfa, “Óawla mu"arrikh andalusì majhùl min al-'aßr al-umawì,” Awràq ˝adìda 7–8 (1984–85), 57–68. 58 See de la Puente, “Entre la esclavitud y la libertad: consecuencias legales de la manumisión según el derecho màlikí,” Al-Qan†ara 21 (2000), 339–60, note 69. There might have been an attempt to “revive” the institution of walà" under the Almohads. 59 See Fierro, “Los mawàlì de 'Abd al-Ra˙màn I,” 89 and note 123.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

213

cessful if they had managed in becoming more important than their former masters.60 We have seen the existence in al-Andalus of another kind of patronate not incorporated by the classical legal schools, such as patronate arising upon conversion. Discussions regarding the origin of Ya˙yà b. Ya˙yà’s Arabic nisba are to be understood within this context. His ancestor was said to have converted at the hands of an Arab and from this conversion he bore the Arabic nisba al-Laythì. By later legal standards Ya˙yà b. Ya˙yà’s ancestor could have acquired such nisba only from manumission, and therefore he must have been a slave. Those who denied this possibility (whatever the reasons) were in fact reminding their readers that patronate arising upon conversion had once existed in the Muslim community. When dealing with how Ya˙yà b. Ya˙yà’s family acquired their Arabic nisba, the Mafàkhir al-barbar substituted walà" for ˙ilf 61 and in this way there was no problem in stating that there had been conversion at the hands of a Laythì, because the ensuing tie had not been walà", but merely an alliance. When it was being practiced, walà" al-islàm had problems of its own, as shown in the case of the Banù Qasì. This family, who played a crucial political and military role in the Upper Frontier during the Umayyad period,62 had as its ancestor a Casius described by Ibn Óazm as a “count of the frontier in the times of the Visigoths.” This Casius would have traveled to Damascus where he converted to Islam at the hands of the Umayyad caliph al-Walìd (86/705–96/715), becoming his mawlà and boasting of this relationship.63 There was, however, another version regarding the identity of their patrons by walà" al-islàm. Ibn Óàrith al-Khushanì records the biography of a man called Óassàn b. Yasàr al-Hudhalì, qà∂ì in Saragossa at the time of 'Abd al-Ra˙màn I’s arrival in the Peninsula. It was said that the Banù Qasì had converted at his hands. When 60 This would be the case of the Banù Mùsà l-Ghàfiqì, see Fierro, “Los mawàlì de 'Abd al-Ra˙màn I,” 89, note 124, which would provide also an early example of intermarriage between Arabs and mawàlì. Oliver (art. cit. in note 75), in her note 12, interprets differently Ibn al-Qù†iyya’s text, but her interpretation is not supported by Ibn Óayyàn, Al-Muqtabis al-thànì, f. 125v (Spanish translation, p. 120). 61 See on “alliance” Crone, Roman, provincial and Islamic law, 51–56. 62 See on it Viguera, Aragón musulmán (Zaragoza, 1988), passim; Manzano, La frontera de al-Andalus, index. 63 Ibn Óazm, Jamharat ansàb al-'arab, ed. H.A.S. Hàrùn (Cairo, 1962), 502. Ibn Óazm refers to the Banù Qasì as muwalladùn and not as mawàlì.

214

maribel fierro

this rumor reached Mu˙ammad b. Lubb,64 this member of the family of the Banù Qasì sought Óassàn’s son Muntaßir and killed him for having pretended such a thing.65 Why this reaction? Mu˙ammad b. Lubb’s career shows that his rivalry with other members of his family made him seek the Umayyad emir’s support, remaining loyal to the ruler for a long period of time. Within this context and taking into account that Umayyad elites were considered to be Umayyad mawàlì, Mu˙ammad b. Lubb must have had a vested interest in promoting the version according to which his ancestor had converted at the hands of an Umayyad caliph.66 Therefore, he would have opposed any other version that attached him to an Arab tribe, especially considering that his most serious opponent for the control of Saragossa was an Arab, Mu˙ammad al-Anqar b. 'Abd al-Ra˙màn al-Tujìbì, whose family had received the support of the Umayyads in order to weaken the Banù Qasì’s power.67 This episode might also indicate that Óassàn’s family could have attempted at claiming certain rights arising upon such conversion, perhaps in the field of inheritance as indicated in Ibn al-'A††àr’s text.68 I have already pointed that Arabs must have regarded patronate ties arising upon conversion with mistrust after the conquest of the Iberian Peninsula, as their establishment could promote the erosion of the social basis of the livelihood of the Arab military elites, i.e., promote the decrease of the dhimmì community through their conversion to Islam. When the Umayyad emirs came to power, they had a vested interest in keeping those elites under control by diminishing their power and their numbers. As we have seen, Umayyad policies promoted conversion in the long run, but it did not mean 64 I identify him with Mu˙ammad b. Lubb b. Mùsà b. Mùsà b. Furtùn, d. 285/898. 65 Ibn Óàrith al-Khushanì, Akhbar al-fuqahà" wa-l-mu˙addithìn, ed. M.L. Ávila and L. Molina, Madrid, 1992, no. 74. 66 This version had already been used by his ancestor Mùsà b. Mùsà in a context in which he tried to gain the favor of the Umayyad emir, see Ibn Óayyàn, Al-Muqtabis al-thànì, f. 188r (Spanish translation, p. 321). 67 See on him and his family: Guichard, art. “Tudjìb (Banù),” EI 2; Viguera, Aragón musulmán, passim; Manzano, La frontera de al-Andalus, 337–38. The Tujìbìs were Southern (Qa˙tàn) Arabs. Hudhayl, the tribe of the judge who pretended to have converted the Banù Qasì’s ancestor, belongs to the Northern (Mu∂ar) Arabs. 68 See note 42. If that was the case, it would mean that at the time walà" alislàm was still legally binding. It would also mean that Ibn al-'A††àr in his text might have been referring to the Upper Frontier and therefore that Màlikism had not yet been imposed there.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

215

an increase in the number of Arabs. The reception of classical fiqh and more specifically of the Medinese legal school (later to constitute Màlikism) meant that patronate arising upon conversion was no longer considered to be legally effective69 and therefore converts did not become Arab clients. The establishment of Màlikism (encouraged especially under the caliphate, by which time it became the official doctrine) meant that contractual patronate (walà" al-muwàlàt) also ceased to be considered legally valid. When Thàbit b. Óazm (d. 313/925) decided to change his nisba from al-Zuhrì to al-'Awfì he was entitled to do so according to the old legal doctrine, because his client ties with the Banù Zuhra did not arise upon manumission, but were of a contractual nature (walà" 'alàqa meaning walà" al-muwàlàt)70 and therefore he could terminate those ties. This is an example of how by the second half of the third/ninth century, Màlikì doctrine was not yet predominant and practices derived from former legal doctrines were still being followed. For the same period we have other examples of walà" almuwàlàt still being practiced, those examples showing its use among rebels against the Umayyads.71 The latter had therefore a vested interest in supporting the disappearance of walà" al-muwàlàt, as it meant eliminating a way through which their enemies could establish enduring alliances among themselves, having a legal basis. Against this background, it should not come as a surprise that by the fourth/tenth century the term mawàlì, in plural, is found almost exclusively in connection with the ruling Umayyad family. Historical sources insist on the presence of Umayyad mawàlì among the Muslim armies settled in the Iberian Peninsula.72 Their role was, according to those sources, very important in supporting the cause of the first Umayyad emir and therefore making it possible the later establishment of the Umayyad caliphate. The behavior of those mawàlì towards the fugitive Umayyad gives an idea of the importance attached to walà" ties. References to the clients of 'Abd al-Ra˙màn I are abundant in the sources, not only because of their deeds but also because many of them founded families who played relevant roles in the history of the Umayyad dynasty. I have been able to trace information 69 70 71 72

But see note 67. It might have also been a case of walà" al-islàm, see note 67. See Fierro, “Four questions,” 306. Manzano, “Settlement,” 87 and note 5.

216

maribel fierro

on 19 clients inherited by 'Abd al-Ra˙màn I from his Umayyad family and to 22 clients with whom he himself established walà" ties.73 Of these 22, seven were clients arising upon manumission and two were cases of contractual ties (walà" al-muwàlàt) established with Arabs.74 Other cases could be interpreted in the same sense. No case of walà" al-islàm is recorded, although the reasons for this absence are not clear.75 The “direct” clients families provided scholars and members of the civil and military khidma. On the other hand, the “indirect” clients’ families (many of them belonging to the Syrians) seem to have owned rural properties and in general had a better economic position. In a recent study on the mawàlì in al-Andalus, D. Oliver76 has clearly shown how any analysis of the term mawlà and its plural mawàlì needs to take into account the historical period in which they are used, the region, the social context, the date of redaction of the source and the political alignment of the author, as they underwent changes and acquired nuances according to each context. To translate mawlà as client does not always work: other possibilities are ally, defender, friend, companion, relative, lord. Oliver argues that in the time previous to 'Abd al-Ra˙màn I’s arrival in the Peninsula (92/711–137/755) the so-called Umayyad mawàlì should be understood not in a strictly technical sense, but as being “relatives” of the Umayyad prince. In fact, no distinction is made, when they are referred to, among “real” clients, Umayyad family members and Arabs such as Qurashìs. After the fall of the Eastern Umayyad caliphate, the Banù Umayya would have become a “tribe” in which no distinction was made between Umayyads by blood or by walà". During the same period and within the tribal See Fierro, “Los mawàlì de 'Abd al-Ra˙màn I.” Again, these cases give us examples of Andalusì legal practices before the reception of Màlikism. 75 The manumitted slaves who became clients of the first Umayyad amìr seem to have been already Muslims (none is recorded to have converted to Islam). In the caliphal propaganda directed by 'Abd al-Ra˙màn III to his Berber allies in North Africa there is insistence on the fact that their ancestors converted under the Umayyads, implying that the latter were therefore entitled to the Berbers’ loyalty. See Fierro, “Sobre la adopción del título califal por 'Abd al-Ra˙màn III,” Sharq alAndalus 6 (1989), 33–42. 76 “Sobre el significado de mawlà dentro en la historia omeya de al-Andalus,” alQan†ara 22 (2001), 321–344. 73 74

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

217

milieu, the expression mawàlì was used to refer to those tribal clans or families that were subordinate to the clan or family that had assumed leadership, regardless of the existence of any formal ties. Conversely, if the leader family or clan lost power, there would follow a decrease in the number of their mawàlì, as the latter recovered their former genealogies. This would be reflected in the following cases: among the Syrians who settled in al-Andalus, there is mention of 2000 Umayyads and 8000 Arabs, whereas some years later only 500 Banù Umayya are recorded. According to Oliver, after the fall of the Eastern Umayyad caliphate they recovered their original nisbas. Some became mawàlì of Yùsuf al-Fihrì when he assumed the governorship of al-Andalus.77 After the triumph of 'Abd al-Ra˙màn I, there was an increase in the number of Umayyad mawàlì. This increase happened through different means: an effort was made by the Umayyads to gather in al-Andalus the Umayyads, both by blood and by walà", scattered in different regions, and an effort was also made in establishing new walà" ties. At the same time, the term mawlà does not have always a technical meaning. The establishment of the Umayyad state brought along the creation of hierarchical categories within the ruling “family,” and a distinction was made among close agnate relatives, as well as between old and new clients.78 After 'Abd al-Ra˙màn II’s reign, the term mawàlì was used to refer to an institutionalized social group. This is evident in descriptions of court receptions, with references to alQuraysh wa-l-mawàlì being situated in specific parts of the reception room. When 'Abd al-Ra˙màn III assumed the caliphal title, he tried to situate himself hierarchically well above his own mawàlì, avoiding their familiarity and the insolent behavior they attempted towards the caliph. And he did so by creating a new clientele situated above his own family. At the same time, among his relatives a distinction was made, not according to their genealogy, but according to their position in the khidma.79 The mawàlì during the Umayyad period are It was the case of 'Ìsà b. 'Abd al-Ra˙màn al-Umawì, who served Yùsuf alFihrì and did not show the expected loyalty to 'Abd al-Ra˙màn I. Although he was forgiven for that, he was never put on the same level as the other Umayyad mawàlì by the Umayyad amìr. See Fierro, “Los mawàlì de 'Abd al-Ra˙màn I,” 83. 78 In my own terminology, between indirect and direct clients. 79 A prosopographical study of the buyùtàt of mawlà origin involved in the Umayyad administration can be found in Meouak, Pouvoir souverain, administration centrale et élites politiques dans l’Espagne umayyade (II e–V e/VIII e–X e siécles) (Helsinki, 1999). 77

218

maribel fierro

thus conceived by Oliver as an aristocracy composed by families related to the Umayyads by different ties. Within them there were ranks. In the lowest place there were those who had become members of the Umayyad family by ties of walà", although those who were manumitted slaves never had their origins forgotten. Then followed those who were considered relatives, even if they were descendants of mawàlì of the former Umayyad caliphs in the East.80 In the caliphal period, a new ranking was established, this time putting the Quraysh and the mawàlì quite low in receptions, after the immediate Umayyad family (brothers, sons), the viziers and the court functionaries.81 The reason for this down-grading seems to indicate that the increase in family members forced to adopt a new policy towards them, considering them an otiose aristocracy entitled to certain privileges and welcomed at the court, but regarded as useless in terms of government. Another important development taking place during the Umayyad period was that mawlà and mawàlì came to refer to those who were loyal to the Umayyad state as the opposite to “Arabs,” synonym often of rebels. This is to be connected with the fact that the Umayyads took care not to be referred to as “Arabs” and with the fact that in Umayyad chronicles people in the khidma are not referred to by recording their Arabic tribal nisbas, the only exception being the judges in the early period of the Umayyad emirate.82

Who were the muwalladùn? The relationship between Arab and non-Arab Muslims in Spain differed from that of the east in three major respects. First, walà" played virtually no role in it . . . Secondly, Spain saw armed conflicts between

See on this Fierro, “Los mawàlì de 'Abd al-Ra˙màn I,” 91. See a visualization of the court ceremonial in Barceló, “The Manifest Caliph: Umayyad Ceremony in Córdoba, or the Staging of power,” in M. Marín, ed., The Formation of al-Andalus. Part 1: History and Society, The Formation of the Classical Islamic World, vol. 46 (Aldershot: Ashgate Variorum, 1998), 425–56 (English translation of “El califa patente: el ceremonial omeya de Córdoba o la escenificación del poder,” Estructuras y formas del poder en la historia (Salamanca, 1991), 447–55). 82 The fact that judges were Arab had evidently to do with the Arab ethnicity of Muslims in that early period and with Arab privileges. There is some discrepancy regarding who was the first non-Arab judge to be named. See on this Fierro, “Los mawàlì de 'Abd al-Ra˙màn I,” 76–77. 80 81

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

219

Arabs and indigenous Muslims. Throughout most of the Umayyad period, the mawàlì of the East were ethnically too diverse and socially too dispersed in Arab society to rebel as mawàlì against Arabs, while at the same time non-Arabs who had stayed together had also failed to adopt Islam. Only shortly before and after the 'Abbàsid revolution, when on the one hand whole localities adopted Islam together, while on the other hand government was still identified as Arab, did nonArab Muslims rebel against Arab rule. They did not, however, rebel as mawàlì, but rather as heretics (as in North Africa) or even as nonMuslims, rejecting the Arabs and Islam together (as in both North Africa and Iran). In Spain, where Arabs and muwallads coexisted as distinct groups, such revolts could in principle have erupted any time. In practice, they only came in the third/ninth century, perhaps provoked by the growth of the Umayyad state (Arab and Berber leaders also rebelled, and the upshot was the centralized state of 'Abd alRa˙màn III); and here for once the rebels took action as mawàlì, explicitly invoking the cause of the non-Arab Muslims (da'wat al-muwalladìn wa-l-'ajam . . .) under the leadership of men such as 'Umar b. Óafßùn . . . Being short of traditions of their own, partly because they were natives of provincial Spain and partly because they were Muslims of long standing, they had no alternative to Cordovan Islam. Accordingly, they did not deny the legitimacy of the Cordovan state as heretics; and though Ibn Óafßùn did in the end reject Islam for Christianity, few muwallads followed suit . . .83

This is an accurate account of the situation, although mawàlì and muwalladùn are not the same thing. The rebels mentioned by Crone did not rebel as mawàlì, but as muwalladùn. This term84 (in plural) refers to Arabic-speaking Muslims of Hispanic origin, converted without walà", who, according to D. Oliver, were characterized by the fact that, around the second decade of the third/ninth century, they formed their own tribal groups and exhibited the same patterns of behavior as rebels of Arab and Berber origin.85

83

Crone, art. “Mawlà,” EI 2. It has produced the Spanish form “muladí” (from muwallad ), pl. muladíes. The complex history of this Spanish term has been studied by Oliver Asín, Conferencias y apuntes inéditos, ed. D. Oliver (Madrid, 1996), 58–61. 85 See Oliver, “Una nueva interpretación de ‘árabe’, ‘muladí’ y mawla’ como voces representativas de grupos sociales,” Proyección histórica de España en sus tres culturas: Castilla y Léon, América y el Mediterráneo, Arabe, hebreo e historia de la medicina, vol. 3 (Valladolid: Junta de Castilla y León, 1993), 143–55. 84

220

maribel fierro

The muwalladùn as tribal groups D. Oliver puts emphasis on the fact that a clear distinction must be made among the muwalladùn, the mawàlì and the 'ajam (descendants of the Hispano-Roman and Hispano-Gothic population who had maintained their Christian faith). Muwallad means, as recorded by R. Dozy, “anyone who, without being of Arab origin, is born among the Arabs and has been raised as an Arab.”86 Now, the plural form of muwallad seems to have been restricted to al-Andalus. There is no mention of groups who received the label of muwalladùn before the year 206/821. The references that occur after that time are associated almost exclusively with the following geographical zones: Mérida and adjacent regions, Granada, Sevilla and Jaén. Oliver concludes that “we must situate the Muwallads in Mérida on the one hand and on the other in the military districts where the Syrian Arabs who entered with Balj established themselves (Seville, Granada, Malaga, Jaén, Beja and Cádiz); it was the descendants of the latter group who were the chief instigators of the civil war that erupted at the end of the ninth century and who caused the appearance of tribal groups made up of Hispano-Goths,”87 such as Ibn Óafßùn. In a famous speech, Ibn Óafßùn addressed his followers saying: “Too long already . . . have you borne the yoke of this sultan who seizes your possessions and crushes you with forced tribute. Will you allow yourselves to be trampled underfoot by the Arabs, who regard you as slaves? . . . Do not believe that it is ambition that makes me speak thus; no, I have no other ambition than to avenge you and deliver you from servitude” (†àla mà 'annafa 'alaykum al-sul†àn wa-ntaza'a

86 See Dozy, Supplément aux dictionnaires arabes, 2 vols. (Leiden/Paris, 1967), s.v. musàlima. See also see Chalmeta, art. “Muwallad,” EI 2. 87 Oliver does not mention the muwallads of the Upper Frontier, see Fierro, “Four questions,” 308. There is famous passage by al-'Udhrì where it is said that in Huesca there were not during his time (fifth/eleventh century) any real Arabs descendants of those who conquered the town, only those who took pride in belonging to the lineage of those who converted at the hands of the Arabs at that time: Nußùs, 57 (trans. F. de la Granja, La Marca Superior en la obra de al-'U≈rì (Zaragoza, 1966), 64). This means that Arabic nisbas deriving from walà" al-islàm were preserved by some of the inhabitants of the Upper frontier, thus maintaining an old practice that was not supported by Màlikism. Al-'Udhrì’s remark might have to do with issues of his own times. An Arab Tujìbid dynasty (branch of the Banù Íumàdi˙) ruled Almeria where he lived and it was related to the Tujìbids who had once reigned in the Upper Frontier, being later on substituted by the Banù Hùd, who claimed Judhàmì (Southern Arabs) ascent.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

221

amwàlakum wa-˙ammalakum fawqa †àqatikum wa-adallatkum al-'arab wasta'badatkum wa-innamà urìdù an aqùma bi-tha"rikum wa-ukhrijakum min 'ubùdiyyatikum).88 Oliver interprets the word 'arab as referring to “the men of that race [i.e., ethnic background] who boast of their ancestral traditions and who attempt to bring down the Empire built by the Umayyads.” Oliver’s interpretation is reinforced by other texts, such as the poem by Ibn Jùdì in which he exhorts the Umayyads to abandon the mulk, since the mulk belongs to the abnà" al-'arab; the Arab leaders of Elvira are called amìr al-'arab; and the Yemenì Arab Ibràhìm b. Óajjàj b. 'Umayr al-Lakhmì is called by the sources “king” (malik) of Seville. These Arab chieftains were imitated by clients (mawàlì) of Arab tribes, by Berbers and also by Muwallad leaders as a reaction against the Arab bid for power in al-Andalus. The muwalladùn as feudal lords Ibn Óafßùn’s rebellion has been subject to different interpretations. M. Acién Almansa has provided us with one that deserves careful attention.89 For him, Ibn Óafßùn was the prime embodiment of the struggle by the old Visigothic “rent-lords” against the imposition of Islamic society. The Umayyad emirate (mainly during the third/ninth century) is seen by him as a transition period from which ultimately the “Islamic social formation”90 emerged triumphant in al-Andalus, but only after a long contest between, on the one hand, the tribal groupings within which the bulk of the conquering population (Arabs and Berbers) was structured, and, on the other hand, the feudal social formation which predominated in the final years of Visigothic Hispania. The inheritors of the feudal formation of the Visigothic era were the great Muwallad thuwwàr. These, in Acién’s words, “far from being a bunch of parvenus who appeared out of nowhere, had ancestors who linked them to the former ruling classes,” as it would have been Ibn Óafßùn’s case, as one of his ancestors had belonged

88

See Fierro, “Four questions,” 300. Acién Almansa, Entre el feudalismo y el Islam: 'Umar ibn Óafßùn en los historiadores, en las fuentes y en la historia ( Jaén, 1994); (repr. with a new introduction: Jaén, 1997). 90 See on his definition of the “Islamic social formation,” Acién Almansa, “El papel de la ideología en la caracterización de las formaciones sociales islámicas,” Hispania 200 (1998), 915–68. 89

222

maribel fierro

to the Hispano-Gothic aristocracy. Ibn Óafßùn’s revolt had two main causes, related to the process of the transition to an Islamic society as represented by the caliphate, to wit: 1) a remote cause, the breakup of Visigothic society; 2) a cause that was nearer to hand, the reforms carried out by 'Abd al-Ra˙màn II. With respect to the first, according to Acién a sizeable part of the Visigothic aristocracy had from the beginning managed to hold onto the privileges that they possessed in zones that were primarily rural. Their continued grip on privileges could be interrupted in favour of the conquerors. Nevertheless, the greatest danger which confronted the HispanoGothic aristocracy was not land seizures, but rather mixed marriages (what P. Guichard has referred to as the absorption of indigenous lineages into the rigid patrilineal system of the Arabs). The Muwallad lineages that were active in the fitna were, Acién claims, those who had been able to maintain, on the one hand, their male bloodline and, on the other, their economic base, which was overwhelmingly rural and from which derived their dominant position in society. The “Visigothic-ness” of Ibn Óafßùn would also be shown in the fact that he restored to areas under his control internal norms of conduct affecting, for example, the role of women in a way that was “deliberately different from Islamic norms.” For Acién, thus, the fitna at the end of the emirate must be understood as the violent resolution of the transition to an Islamic society, as represented by the caliphate, in conflict against tribal and feudalizing forms of society. The rebellions of the Muwallads of the second half of the third/ninth century represent the parting shots of the Visigothic feudal society in the face of the steadily more effective imposition of tributary Islamic society. C. Wickham has characterized this Islamic society by the control exercised by “a ‘state-class’ based on a public institution, with political rights to extract surplus from a peasantry that it does not tenurially control.”91 If feudal society is characterized by rent-taking and tributary society by tax raising, the Muwallad rebels of the second half of the third/ninth century tended to behave like feudal lords, in the broad sense that they were “rent-collectors.” Furthermore, Acién states that we should not take

91 “The uniqueness of the East,” Journal of Peasant Studies 12 (1985), 166–96, 170–71.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

223

the portrayal of the fitna in terms of ethnic confrontation seriously, in the way that, for example, the historian Ibn Óayyàn does when writing of 'aßabiyyat al-muwalladìn 'alà l-'arab, 'aßabiyya li-l-muwalladìn wa-l-'ajam wa-n˙iràf an al-'arab, da'wat al-'arab, da'wat al-muwalladìn wal-'ajam 'alà l-'arab, da'wat al-muwalladìn, thawrat al-muwalladìn, or in calling Ibn Óafßùn imàm tilka al-na˙la (namely, al-musàlima, al-muwalladùn, naßàrà l-dhimma) or the words put in the mouth of Ibn Óafßùn when, during his alliance with Ibràhìm b. Khumayr (the Umayyad emir 'Abd Allàh’s general) he writes in secret to another Muwallad rebel, Ibn Mastana, denying his muwàlàt al-'arab and assuring him of his al-da'wa al-muwalladiyya. Acién in fact emphatically rejects the “ethnic” interpretation of the first fitna and later fitnas, on the grounds that their supposed ethnic basis is no more than an attempt to fit events into stereotypical moulds, and a simplification of certain processes whose meaning the chroniclers had not managed to grasp. In answer to the ethnic approach, Acién carries out an analysis of the terminology used to describe the relations between Ibn Óafßùn and his followers, as well as between Ibn Óafßùn and other rebels. His conclusions are that there is: . . . a variety of human groups that is much wider than the ethnic conceptualization that the sources themselves admitted. Among this variety we can find the Berbers of the Lower and Middle Marches, who always appear with their tribal names and are always led by chiefs who receive the title shaykh or muqaddam. Groups of Arabs make up another variant, which I have called a society of lineages, led by one of these lineages, who receive the name aß˙àb . . . A third definite group is that of the hierarchical societies linked by bonds of dependence, among which stand out aß˙àb and quwwàd, men who are nevertheless always subordinate to strong individuals like the Óafßùnìs . . ., who receive specific titles such as sayyid or rabb. And, finally there are the urban societies, based around assemblies and councils, at whose head a varying number of ru"asà" are to be found.

Acién tries to show the complexity of the alliances which were formed and the dissension that existed within each group. Precisely because he does not accept the ethnic framework, Acién interprets in his own fashion the term 'arab which appears in the famous speech by Ibn Óafßùn to the people of the ˙ußùn. According to Acién, 'arab should not be interpreted as meaning the Arabs, but rather as a way of referring to the sul†àn, that is, the Cordoban state and its apparatus of power (the army and Màlikism).

224

maribel fierro

I have elsewhere criticized Acién’s understanding of Muwallad rebels as former Visigoth rent-lords.92 Acién considers that his argument is corroborated by the Visigothic ancestry of the Muwallad rebels, the role of women in the territory under Ibn Óafßùn’s control, the use of specific titles such as sayyid or rabb by the Muwallad rebels, and by their opposition to Umayyad rule. To the first, I have indicated that there is no evidence that the Muwallad rebels has any Visigothic noble origin.93 To the second, that there is no evidence whatsoever that the situation of women under Ibn Óafßùn’s rule was significantly different from the rest of Muslim society.94 To the third, that the so-called title rabb that Ibn Óafßùn would have assumed is in fact a quotation from Qur"àn 79:24.95 To the fourth, that in Ibn Óafßùn’s speech, 'arab should be understood according to Oliver’s interpretation, that is, he was not referring to the Umayyads (never referred in the sources as 'arab), but to the Arab lords who were trying to put an end to Umayyad rule and who stressed Arab superiority over all the rest of the groups. While agreeing with Acién that the Muwallad rebels tended to behave like feudal lords, in the broad sense that they were “rentcollectors,” I differ in seeing them as necessarily descending from the Visigothic nobility, as in order for them to have acted as feudal lords it is neither necessary nor relevant that they should have had Visigothic noblemen as ancestors. As Wickham explains, the feudal mode of production may exist without the tributary mode, but the same is not true of the reverse, since even when the feudal mode does not exist within the tributary mode, feudalizing tendencies will inevitably appear in the latter. The fact that the Muwallad rebels behaved like feudal lords cannot be explained solely by postulating the survival of the former Visigothic nobility, a nobility opposed to the attempt by an incipient “state class” to assume control. Whereas Acién states that the Muwallad rebels were in no sense “a bunch of parvenus who appeared out of nowhere,” nearly all of them appear to be pre92 Fierro, “Four questions,” art. cit. in note 41. Acién has discussed some of my comments in the second edition of his book. 93 The only exception appeared to be Ibn Óafßùn’s genealogy. D. Wasserstein has shed now doubts in it, considering it an invention in one of the many attempts Ibn Óafßùn made for legitimizing his bid for power. See Wasserstein, “Inventing Tradition and constructing Identity: The Genealogy of 'Umar Ibn Óafßùn between Christianity and Islam,” Al-Qan†ara XXIII (2002), 269–98. 94 Fierro, “Four questions,” 295–96. 95 Fierro, “Four questions,” 326.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

225

cisely that, namely, “parvenus” who were hoping to extract rent from the peasant population by means of violence, taking advantage of the weakness of a tributary system that had not yet been fully implemented. On the one hand, Arab tribal leaders were resisting the imposition of Umayyad rule which effectively meant the loss of their privileges and status. On the other hand, Umayyad mawàlì, who benefited from the Umayyad rulers whom they served, were acting like those military leaders described by P. Crone who had under their control a large number of military and administrative positions for which they had to find trustworthy men; and when it came to selecting candidates, their criterion was kinship ties, whether direct (family) or indirect (clients).96 One of such men was the Syrian (of Umayyad mawlà origin) Hàshim b. 'Abd al-'Azìz, an Umayyad general and ˙àjib active during the times of the emir Mu˙ammad (238/ 852–273/886). Positions such as those attained by Hàshim generated intense competition, for they were positions of power, prestige and above all wealth, not only because they entailed a salary, but also because they allowed those who held such positions to siphon off part of the taxes they collected for their own private use. About the time when Ibn Óafßùn’s rebellion occurred (the year 265/878), the governor of the rural district of Rayya was a man who seems to have been a member of Hàshim b. 'Abd al-'Azìz’ family. This governor increased taxes in an exorbitant fashion, which caused the people of the district to flee to the mountains. It is in this context that Ibn Óafßùn’s first “rebellion,” characterized by plunder and pillage, took place. As a result of his defeat by Hàshim in 271/883, he was taken to Cordoba to be integrated into the emiral army. Ibn Óafßùn deserted it (as other Muwallad rebels did) because of the poor treatment he received at the hands of another Umayyad mawlà. During the early period of the Umayyad emirate, the possibilities of gaining positions of responsibility and recompense depended on one’s being either an Arab or a mawlà.97 On the one hand, the emir tended to place his confidence in—and hence entrust the positions of responsibility (consequently those positions that were the most lucrative) to—his clients. On the other hand, the mawàlì were not well disposed toward admitting “competitors” when the emiral recompenses 96 Crone, “Were the Qays and the Yemen Political Parties?,” Der Islam 71 (1994), 1–57, 50–51, 53. 97 See note 49.

226

maribel fierro

were handed out, above all when there were few such recompenses to go round, resulting in more intense competition for those few. The “poor treatment” of Ibn Óafßùn occurred during the rule of the emir Mu˙ammad (238/852–273/886), of whom it is said that he reduced soldiers’ and officers’ wages and cut back the amount of monthly gifts. The second rebellion (this time in the strict sense of the word) of Ibn Óafßùn (that is, when he quit the emir’s army) must be owed to the fact that he, like other converts, was not allowed to compete on an equal footing with Arabs and mawàlì for the highest honors and stipends in this army. As long as the process of conversion/Islamization was confined to a minority of the population, the recompenses offered by the emir were shared out among very few (Arabs, mawàlì, Berber military leaders and some dhimmìs like Qùms). With the increase of conversion/Islamization, the new Muslims began to claim their right to join in the competition for such recompenses. For the non-Arabs, incorporation into the group of Umayyad mawàlì was what permitted access to these recompenses, but in order to prevent competition, the mawàlì must have tended to try to keep their numbers from growing and to set up obstacles in the path of those who were not clients. Where the converts first succeeded in fully assimilating themselves was in the religious offices. In the long run, as it happened in the fourth/tenth century under the Umayyad caliphate, the Umayyad rulers will opt for creating their own elites regardless of their ethnic origin. Contrary to Acién, I think that Ibn Óafßùn was not recreating the “feudal” ties of his (alleged) Visigothic ancestors, but acting as a feudal lord at a time when the Umayyad state was incapable of keeping in check the feudalizing tendencies of the Arab “tribal” leaders. Ibn Óafßùn and other Muwallad leaders imitated the latter and therefore they had their own mawàlì and muwàlùn. According to Oliver, the Muwallads were the section of the converted Hispanic element that had formed their own tribal groups and begun to exhibit the same patterns of behavior as the rebels of Arab and Berber ethnic background. For Oliver, what we have in the Umayyad period is the conflict between two societies, which arranged a succession of truces in an attempt to survive; on the one hand, the tribal society of Arabs and Berbers which, in the third/ninth century, managed to drag the Hispanic community along with it, and on the other hand, the Islamic “state” society. Acién’s view is the exact reverse. For Acién, the Arab lineages of the fitna were

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

227

“closely assimilated in their behavior” to the Muwallad rebels, “heirs of the Hispano-Gothic world who preserved obviously feudalizing or proto-feudal structures and attitudes,” adding that “the contamination” of the former groups by the latter “is revealed by links of mutual assistance” that were established between them, “which shows a relationship based on class and not at all on ethnic differences in the way that the Romantic historiography that has come down to us would have it.” While I may be in disagreement with Acién regarding the survival of Visigothic feudal structures, I disagree no less with Oliver regarding the “tribalization” of the Muwallads. I stand with Crone in believing that the Arab tribes had ceased to function as such and what we are dealing with are factions rather than tribes98 and with T. Glick’s interpretation99 that the Muwallads rebelled for their right to share power and compete for economic and social rewards on equal terms with the Arabs and, in the case of converts, with the old Muslims, i.e., with the mawàlì. The competitive scheme of ethnic stratification is of relevance here. Under this scheme, the ethnic dividing line splits the class system vertically, in such a fashion that in theory members of each group occupy the same position on each side of the line, though in practice one group almost invariably occupies the upper stratum, and this gives rise to resentment and discontent in the other groups, whose expectations of equal status are not seen to be fulfilled. The Muslim Muwallads100 corresponded to this “ethnic” group as distinguished from the “old” Arab Muslims; theoretically and according to the Màlikì legal school, the Muslim Muwallads could participate at all levels of the Arab-Muslim society, but in practice any attempts on the part of the Muwallads to occupy a place in that society ran into obstacles and resistance. One 98 See the case of the Umayyad mawlà 'Abd Allàh b. Mu˙ammad b. Khàlid b. Martinìl, sent by the amìr Mu˙ammad to pacify Beja, a town where there was conflict between Mu∂ar and Yaman. He concluded that there was no way to know who belonged to which tribe: 'Iyà∂, Tartìb, IV, 240. 99 See Glick, “The ethnic systems of premodern Spain,” Comparative Studies in Sociology 1 (1978), 157–71, and Glick and Pi-Sunyer, “Acculturation as an explanatory concept in Spanish history,” Comparative Studies in Society and History 11 (1969), 136–54. See also Marín Guzmán, “Social and ethnic tensions in al-Andalus,” Islamic Studies 32/II (1993), 279–318; and id., “The causes of the revolt of 'Umar ibn Óafßùn in al-Andalus (880–928). A study in Medieval Islamic social history,” Arabica 42/II (1995), 180–221. 100 For this precision (Muslim), see below, p. 228.

228

maribel fierro

of the possible ways out for those who saw their hopes frustrated was to take up arms and, if circumstances proved favorable, comport themselves like “feudal lords.” The process of conversion and the muwalladùn We have seen that the term muwalladùn appears at a specific period of Andalusì history (roughly third/ninth century), after a century of Muslim presence in the Iberian Peninsula, and disappears with the establishment of the Umayyad caliphate. The meaning of muwallad is, as we have seen, “anyone who, without being of Arab origin, is born among the Arabs and has been raised as an Arab.” This definition does not include Islam. Because Arabs in al-Andalus were by definition Muslims, there are many chances that a muwallad was also a Muslim. Oliver reaches this conclusion also from the fact that the sources distinguish between muwalladùn and 'ajam/naßàrà. However, a distinction is also made in the sources between al-muwalladùn and al-musàlima.101 This distinction has been explained as referring respectively to recent converts and old converts, or to converts of Hispanic origin and to Berbers, or between the native-born population who had not been granted dhimma status (i.e., those conquered by force and obliged to convert to Islam) and converted dhimmìs.102 However, muwallad designates someone who was “Arabized” but not necessarily “Islamized.” Given the fact that we find muwalladùn in those places where there were Arab settlements, muwallad could also refer to those Christians who, by having contact with Arabs, had become Arabized without necessarily converting to Islam. These Christians are known in late Christian texts and in Western scholarship as “Mozarabs,” but they are never referred to with that name in Arabic sources, which ignore the term musta'rab to designate the Arabized Christians who lived among them. I therefore think that the Arabic term used in al-Andalus for people who were “Arabized,” independently of their religious status, may have been muwalladùn. With the passing of time, linguistic and cultural assimilation would lead to religious assimilation, and in this fashion the original mean-

101 Wa-ta˙azzabat al-musàlima ma'a l-muwalladìn wa-tamayyazat ilayhim naßàrà l-dhimma: Ibn Óayyàn, Al-Muqtabis, ed. P. Melchor Antuña (Paris, 1937), 51. 102 Fierro, “Four questions,” 308.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

229

ing of muwallad (“Arabized”) was lost, and the term came to be understood as “Muslim convert.” Bulliet has put forward the following questions: “How did Muslims know who among the conquered populace had joined their religion? How did members of the non-Muslim communities know which of their coreligionaries had abandoned the ancestral faith for Islam? And how did the convert know that he or she was, in fact, a Muslim?” and he has shown that the answers to these questions lay more in the area of social conduct than in the realm of belief. The environment in which Ibn Óafßùn lived was not one where there was always a clear dividing line between being and not being a Muslim, for in many cases Islamization was not a radical change but rather a slow process, during part of which time the religious affiliation of individuals remained ambiguous. Examples showing perplexity about whether someone was a Muslim or a Christian are not scarce in the period,103 especially taking into account what has been said above that conversion would have been mainly due not so much to a voluntary passage to the new religion, but to an involuntary inability to continue being a Christian. R. Bulliet104 has shown that the process of conversion to Islam “was largely governed by access to knowledge. People simply could not become Muslims if they didn’t know what Islam was.” There is no evidence of missionary activity, so that information about Islam was obtained in four ways: 1) people conquered by force were offered the possibility of converting, they converted and once they were legally Muslims, they slowly found out afterward what that implied by observing the Muslim Arabs’ behavior and asking them questions 2) people who might have been captured in battle and, as part of the booty, been distributed as slaves. They would have learned about Islam by living among Arab Muslims as slaves and quickly come to realize that conversion would greatly enhance the possibility of manumission

103 See the examples given in Fierro, “Four questions,” 309–10 and 317–18. See also Ibn al-'A††àr, Formulario notarial, 626: after conversion by pronouncing the profession of faith, it might happen that the convert did not perform any prayer. In this case, if the convert reverts to his/her former religion, he/she is not considered an apostate. 104 Islam. The view from the edge (New York, 1994), 40–41.

230

maribel fierro

3) people who lived in proximity to a group of Arab Muslims and observed their behavior, very possibly discovering that imitating them enhanced their acceptance and gained them a livelihood purveying goods and services to them. This possibility would have been limited by the fact that a large proportion of the Arabs in the conquering armies settled in large, widely separated cantonments, the amßàr 4) people who might have encountered some bilingual Muslim capable of explaining to them at least some aspects of the Arabs’ religion. The muwalladùn would correspond to the third case. Bulliet indicates that this case, in the East, would have not been very frequent, as the Muslim Arabs lived separated from the rest of the population in garrison cities, the amßàr. It was the absence of these garrison towns in al-Andalus that would have favored Arabization leading to Islamization.105 Briefly, the muwalladùn were active as a social group only during the third/ninth century (the period when the term is employed in the sources). They were Arabized indigenous inhabitants of the Iberian Peninsula, whose Islamization came about as a result of that Arabization. On its turn, Arabization had been made possible by their living among Arabs and this was the result of the settlement patterns of the Arabs (no amßàr, living on the taxes collected from the indigenous population). The Arabization/Islamization of the muwalladùn did not lead to the establishment of clientship ties with Arab tribes for different reasons. If the muwalladùn were descendants of converts who enjoyed free status, those early converts had not been offered the possibility of walà" al-islàm (the muwalladùn do not bear Arabic nisbas): because Arabs were not interested in the conversion of the dhimmìs, or because walà" al-islàm was not any more effective or because the Umayyads wanted to monopolize the establishment of walà" ties. If the muwalladùn were recent converts, the reception of Medinese/Màlikì law did not acknowledge validity to the former walà" al-islàm, so that conversion stopped entailing clientship ties with Arab tribes. In any case, conversion in al-Andalus seems to have been mainly due not so much to a voluntary passage to the new religion, but to an involuntary inability to continue being a 105

See on this Fernández Félix and Fierro, “Cristianos y conversos.”

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

231

Christian. The muwalladùn tried to find a place as Muslims in a society where to be a Muslim was still closely connected to being an Arab firstly and secondly, to being a mawlà or a Berber (because of the Berbers’ role in the conquest). The tensions between Arabs and the ruling dynasty provided the muwalladùn leaders with a model of behavior (rebellion) once the Umayyad elites made it clear that they were not willing to easily grant the muwalladùn space among themselves. Once the muwalladùn joined the field in the struggles that were to end with the imposition of the Umayyad caliphate, they had different options in order to give an aim to their own fighting and to legitimize their bid for power. All those options were undertaken by Ibn Óafßùn. When his incorporation into the Umayyad khidma failed because of mawàlì opposition, he acknowledged 'Abbàsid and also Fà†imid legitimacy. He also tried to present himself as a new 'Abd al-Ra˙màn I within a general climate of messianic expectations. And as D. Wasserstein has recently shown, he also tried to present himself as a descendant of the Visigoths, an option probably connected to his return to Christianity.106 This last move indicates that the repertoire of Muwallad legitimacies was limited to: 1) moving back to the East, that is, renouncing the Umayyad secession of the Islamic “West” and reestablishing links with the 'Abbàsid caliphate 2) abandoning Sunnism and becoming an “heretic” 3) entering eschatological time 4) moving back to the pre-Islamic past of the Iberian Peninsula. If the first three options involved continuing being a Muslim, the last one made it difficult preserving the convert’s identity.

Andalusì scholars and ethnic solidarity in al-Andalus 107 Together with the scarce number of non-Umayyad mawàlì and the existence of the group of the muwalladùn with its very specific characteristics, the other difference between al-Andalus and the East is 106 If Wasserstein’s findings are followed (see note 92), Chalmeta’s statement (“The political model in no way envisaged the re-establishment of a Visigothic state, but the setting up of an autonomous Islamic unit,” see Chalmeta, art. “Muwallad,” EI 2) should be abandoned. 107 This paragraph merely offers an approach to the topic.

232

maribel fierro

the absence of shu'ùbiyya. It was only in the fifth/eleventh century that a man of letters of Íaqlabì origin, Ibn García, wrote an Andalusì shu'ùbì treatise dealing with non-Arab Muslim superiority over Arabs.108 He borrowed his arguments from Eastern writers. But he had an Andalusì agenda: that of legitimizing the rule of his master, the ruler of the ˇà"ifa kingdom of Denia, who was himself of Íaqlabì origin. Ibn García’s shu'ùbì Risàla gave rise to several refutations in the following two centuries, i.e., during the rule of the Berber dynasties of the Almoravids and Almohads, thus showing that “Arabness” was still an important issue in Andalusì identity, even if only from a rhetorical point of view.109 But under the Almoravids and Almohads it was especially the Berber context that promoted criticism of the Arabs.110 Berbers are known to have opposed Arab supremacy in alAndalus very early (second/eight century). The Berber Óafß b. Maymùn had a dispute with the Arab Ghàlib b. Tammàm defending the superiority of the Maßmùda against the Arabs. The dispute ended with Óafß’s murder at the hands of Ghàlib.111 But what concerns us here are not the Berbers,112 but the muwalladùn. The only case in which there is evidence that they tried to connect themselves with the pre-Islamic past is Ibn Óafßùn’s Visigothic genealogy and his apostasy from Islam. These moves alienated him from the Muslim muwalladùn, thus indicating that Islam was a nondispensable part of their identity. Another crucial part was Arab culture. The muwalladùn did not try to revive Latin-Christian culture. This is something that Christians such as Eulogio and Alvaro, the promoters of the so-called Mozarab movement of martyrdom, did try, but their activity in this sense reached an unforeseen result, that of alienating Christians from their former culture.113 108 The classical study is Monroe, The Shu'ùbiyya in al-Andalus (Berkeley/Los Angeles, 1970). See also Graham, “The meaning of slavery and identity in al-Andalus: the epistle of Ibn Garcia,” Arab Studies Journal/Majallat al-Diràsàt al-'arabiyya 3/I (1995), 68–79. 109 See the shu'ùbì text included by Ibn al-Ashtarkuwì (d. 538/1143) in his alMaqàmàt al-luzùmiyya discussed by Monroe in Al-Qan†ara 22 (2001), 218–19. 110 The Andalusì Màlik b. Wuhayb, who worked as secretary and vizier for the Almoravids, wrote a book entitled Qarà∂at al-dhahab fì dhikr li"àm al-'arab, see alMarràkushì, Kitàb al-Mu'jib, ed. R. Dozy, 2nd ed. (Leiden, 1881); trans. E. Fagnan, Histoire des Almohades (Argel, 1893), 133/161; (ed. Cairo, 1949), 184–85. 111 De Felipe, Identidad, 76, 173. 112 See on Berber shu'ùbism M. Shatzmiller, The Berbers and the Islamic state. The Marìnid Experience in Pre-Protectorate Morocco (Princeton, 2000). 113 See on this issue Wright, “The End of Written Ladino in al-Andalus,” in

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

233

But even if the muwalladùn did not try to revive their Latin-Christian cultural legacy, they could have adopted shu'ùbì arguments, which they did not: But having avoided enslavement and migration, the muwallads had also failed to acquire culture and positions of influence in the society of their conquerors. Where eastern mawàlì had spokesmen among bureaucrats and courtiers, the leaders of the muwallads were country squires more noted for their virtú than for their polish; indeed, the smarts and insults suffered by such rural lords at the court of Cordova played a role in the outbreak of several muwallad revolts. The muwallads thus lacked both the education and the influence required for a literary onslaught on Arab superiority. Instead, however, they were in a position to take up arms . . .114

If the muwalladùn had failed in acquiring classical Arabic culture, that had to do with its late development in al-Andalus, to be dated in the reign of 'Abd al-Ra˙màn II (206/822–238/852). It was then that a secretarial class appeared, within the context of the centralizing efforts of the Umayyad emir and his control over the junds.115 Arab culture till then was limited to poetry. It is therefore not surprising that muwallad solidarity was expressed by their poet 'Abd alRa˙màn al-'Ablì, whereas in the opposite camp there was the poet al-Asadì, the champion of the Arabs. Before the times of 'Abd al-Ra˙màn II, Arab-Muslim superiority was expressed in very crude terms, as shown in the anecdote of the Syrian jundì l-Íumayl b. Óàtim’s reaction to the Qur"ànic verse 3:140: he was convinced that al-nàs in that verse could only mean al-'arab, and when corrected, he did not like the fact that it implied that

M. Fierro and J. Samso, eds., The formation of al-Andalus. Part 2: Language, Religion, Culture and the Sciences, The formation of the Classical Islamic World, vol. 47 (Aldershot: Ashgate Variorum, 1998) 19–36 (English trans. of “La muerte del ladino escrito en al-Andalús,” Evphrosyne. Revista de Filología Clásica, nova série, 22 (1994), 255–68). 114 Crone, art. “Mawlà,” EI 2. 115 The conquerors had settled as property owners and they do not seem to have received pensions. This meant that “there was no need for the elaborate and precocious bureaucracy which had grown up to service the system, nor any need to compile dìwàns or lists of those entitled to pensions. The slow development of bureaucracy in al-Andalus may in turn have been a reason for the slow development of a literary culture: nor until the time of 'Abd al-Ra˙màn II (822–852) did administration and a native literary culture begin to appear”: Kennedy, Muslim Spain, 16. For more nuanced interpretation of the same process, see Soravia, “Entre bureaucratie et littérature: la kitàba et les kuttàb dans l’administration de l’Espagne Umayyade,” Al-Masaq 7 (1994), 165–200.

234

maribel fierro

Arabs were mixed with the slaves and the populace.116 But the Umayyads had to count on that populace, especially when they were trying to weaken Arab power. Because of their needs as rulers of an ethnically diverse country who had put an end to Arab autonomous rule, the Umayyads were more “egalitarian” than the Arab tribes had been. The Muslim tradition of egalitarianism found its way in al-Andalus with the introduction of Arabo-Islamic classical literature in Umayyad court circles.117 The Umayyads in the long run will bring about an Andalusì identity that revolved around Islamic religion and Arab culture, with Arabic ethnicity relegated to a secondary position.118 The main road leading to this common Andalusì identity was the incorporation of converts to the world of Muslim scholarship. By the fourth/tenth century, the 'ulamà" were mostly of convert origin, as I have indicated at the beginning of this article. But this was not yet the case in the third/ninth century when Islamic culture was still being introduced and assimilated in al-Andalus by a very scanty number of Muslim scholars.119 Let’s examine the reflection of this situation in two examples. Against the Arabs: the situation in the Upper Frontier From the very early period of Andalusì history, the Upper Frontier saw the appearance of powerful lineages belonging to the local population, such as 'Amrùs b. Yùsuf, called al-Muwallad, and his descendants, and the Banù Qasì. The Upper Frontier, as compared to the Middle and the Lower, had been settled by Arabs. However, and especially after the arrival of 'Abd al-Ra˙màn I, they had soon to face the rise to power of those local families. By the second half of the third/ninth century, the sources record a surprising high num-

116 Ibn al-Qù†iyya, Ta"rìkh iftità˙ al-Andalus, ed. and trans. P. de Gayangos and J. Ribera, (Madrid, 1926), 40–1/31–2. 117 See on this tradition Marlow, Hierarchy and egalitarianism in Islamic thought, (Cambridge, 1997), with numerous quotations of al-'Iqd al-farìd, composed by the Andalusì Ibn 'Abd Rabbihi al-Marwànì (246/860–328/940). There is still no study of the issue of egalitarianism devoted specifically to al-Andalus. 118 Arabic ethnicity became again important among Andalusìs in opposition against Berber identity, especially after the sixth/twelfth century. 119 Fierro and Marín, “La islamización de las ciudades andalusíes a través de sus ulemas (ss. II/VIII-comienzos s. IV/X),” in P. Cressier and M. García-Arenal, ed., Genèse de la ville islamique en al-Andalus et au Maghreb occidental (Madrid, 1998), 65–98.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

235

ber of attacks against Arabs and Arabic genealogy. Mu˙ammad b. Sulaymàn b. Mu˙ammad b. Talìd al-Ma'àfirì (d. 295/907) was trained as a scholar in Cordoba and in the East, and later became qà∂ì in Huesca and Saragossa. He was of mawlà origin and famous for his solidarity on behalf of the muwalladùn (kàna shadìd al-'aßabiyya li-l-muwalladìn).120 'Abd Allàh b. al-Óasan Ibn al-Sindì (d. 335/947) was named by 'Abd al-Ra˙màn III judge of Huesca, Barbastro and Lérida. This Ibn al-Sindì became quite rich. He is described as keen in having contacts with the powerful and so vain that he did not greet anybody nor answered when greeted. But he was especially known for his hatred of the Arabs and his solidarity for the muwalladùn (kàna shadìd al-'aßabiyya li-l-muwalladìn wa-'aΩìm al-karàhiya li l-'arab), so that he collected proofs of their virtues and of those of the slaves ('abìd ) and proofs of the iniquities of the Arabs. His father was client of the ascetic Ibn al-Mughallis who is credited with having influenced the disappearance of the Arab Banù Salama in the area of Huesca. During the second/eighth century, the Tujìbì Banù Salama ruled tyrannically121 in Huesca with an absolute disdain for the people whom they ruled, as shown by the fact that one of their falcons would have devoured the baby of a local woman without its owner stopping it. They were eventually defeated and exterminated by a muwallad leader, Buhlùl b. Marzùq b. Uskarà (his rebellion against the Banù Salama took place in the year 182/798–99). Before this, the tyrannized subjects addressed themselves to an Arab ascetic called Ibn al-Mughallis, who asked God against the Banù Salama and his prayer was answered with Buhlùl b. Marzùq. Ibn al-Mughallis is described as an Arab himself, but also as a mawlà of the Banù Fihr.122 In both cases we have local scholars named by the Umayyads judges in their own land who were of mawlà origin and whose solidarity went to the converts. We have seen before how the Berber

Ibn Óàrith al-Khushanì, Akhbàr, no. 196; Ibn al-Fara∂ì, Ta"rikh 'ulamà" alAndalus, ed. F. Codera, 2 vols. (Madrid, 1891–92), (BAH, VII–VIII), no. 1147; 'Iyà∂, Tartìb, IV, 472–73. 121 The tyrannical rule of Arab tribes gave rise to a number of legal masà"il preserved in the 'Utbiyya, see Fernández Félix and Fierro, “Cristianos y conversos,” 422. 122 See on him Ávila and Molina, “La Marca Superior de al-Andalus en el siglo VIII: el asceta Ibn al-Mugallis y los Banù Salama,” Homenaje al Profesor J. Ma Fórneas Besteiro, 2 vols. (Granada, 1995), II, 703–709. 120

236

maribel fierro

traditionist Thàbit b. Óazm (d. 313/925) decided to stop bearing the nisba al-Zuhrì deriving from his family’s contractual patrons, adopting instead an “Islamic” nisba. We have also seen how Mu˙ammad b. Lubb denied that his family, the Banù Qasì, had converted at the hands of the Arab Óassàn b. Yasàr al-Hudhalì, qà∂ì in Saragossa at the time of 'Abd al-Ra˙màn I’s arrival in the Peninsula, denial that he brought to the extent of killing al-Hudhalì’s son. In my opinion, the underlying context is that of the rise to power in the Upper Frontier of the Arab family of the Tujìbids which put an end to a long period of predominance of muwallad families such as the Banù 'Amrùs, the Banù Qasì themselves and the Banù Shabrì†. The degree of Arabization must have been quite high in the Upper Frontier as there had been an important settlement of Arabs and it was an important zone in terms of producing local scholars. Rule by local convert families was here put in danger by Umayyad support of an Arabic lineage. Local scholars reacted by remembering the oppression of the times of the conquest, by rejecting Arabic nisbas and by reminding the Umayyads of their clientship. In favor of the Arabs: the situation in Cordoba The second half of the third/ninth century witnessed the tensions that arose in Cordoba between the two contending factions of the ahl al-ra"y and the ahl al-˙adìth. The most important representative of the latter group was the mawlà123 Baqì b. Makhlad, who suffered persecution and a trial. Also persecuted was the traditionist Mu˙ammad b. 'Abd al-Salàm al-Khushanì, who boasted descent from a Companion of the Prophet. Put in prison during that persecution, he asked his jailers in impeccable Arabic, “Where are your latrines to be found?” After being obliged to repeat the question in terms that were more explicit and comprehensible to his guards, these men showed him the area reserved for such needs, next to the patrol road about the prison, in full public view, and there the manifestations of its use had accumulated. Faced with this prospect, al-Khushanì

123 See above note 3, where it is indicated that as client of a woman, Baqì did not bear any Arabic nisba.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

237

opted to refuse all food during the three days he remained in jail in order to avoid the need to visit such an unappealing spot.124 Ibn 'Abd al-Salàm al-Khushanì is remembered for his 'aßabiyya of the Arabs and his scorn for non-Arabs: On one occasion he revealed his hatred of the descendants of local converts and mawàlì when, having received news that a number of such people had died, he exclaimed, “They have been exterminated, and now we are done with that breed.” But his preferred victim, at whom he constantly directed malicious allusions, was Mu˙ammad ibn Wa∂∂à˙, the grandson of one of 'Abd al-Ra˙màn I’s slaves, whose imperfect relationship with the Arabic language served al-Khushanì as the platform from which to display his disdain for Ibn Wa∂∂à˙ in public. Mu˙ammad ibn A˙mad ibn al-Zarràd, one of Ibn Wa∂∂à˙’s staunchest supporters, was once in the presence of al-Khushanì, when the individual responsible for reading the text made an error of pronunciation, provoking the following comment from al-Khushanì: “Where did that come from? Oviedo,125 wouldn’t you say?” Ibn alZarràd, suspecting that al-Khushanì was making an indirect reference to Ibn Wa∂∂à˙, angrily replied: “I do not have to remain in a gathering at which Ibn Wa∂∂à˙ is criticized!” Al-Khushanì snapped, “God be praised! Do you think that that man is superior to everyone else?” This allusion on the part of al-Khushanì is repeated in another anecdote at which Ibn Wa∂∂à˙ plays a part. On hearing of the interpretation that Ibn Wa∂∂à˙ gave to a particular term from a ˙adìth, an interpretation that differed from his own, al-Khushanì commented, “Someone from Oviedo will certainly know how to explain the Prophet’s ˙adìth.”126 Ibn Wa∂∂à˙ was the grandson of a manumitted slave of the first Umayyad emir.127 Both his grandfather and uncle held important positions in the army and in the Umayyad khidma. Ibn Wa∂∂à˙ was the first member of the family to become a scholar. He traveled twice to the East, becoming interested in ˙adìth. Together with Baqì 124

Molina, “An Arab among Muwallads,” note 2. It refers to a town in Northern Spain (in the region now called Asturias), where Muslim rule did not extend. 126 Molina, “An Arab among Muwallads,” 122. 127 See Fierro, “Bazì', mawlà de 'Abd al-Ra˙màn I, y sus descendientes,” AlQan†ara 8 (1987), 99–118. Its document of manumission has been preserved. 125

238

maribel fierro

b. Makhlad, he is credited with having introduced in al-Andalus the 'ilm al-˙adìth. However, contrary to the latter, he was also a Màlikì and obviously tried to find a common path between Màlikì doctrine and ˙adìth. His role during Baqì b. Makhlad’s persecution is to be understood in this sense.128 Al-Khushanì’s specific attacks against him might have derived from that episode during which as scholars they had stood in different factions. As Molina has pointed out, during Baqì b. Makhlad’s persecution it is not possible to find distinctive ethnic affiliations in those factions. But even so, doctrinal affinities did not necessarily involve the disappearance of ethnic solidarity.129 It was precisely in the generation of Ibn Wa∂∂à˙’s students that non-Arabs became majority in the world of Andalusì scholarship, something that must have worried Arab elites whose military role was in the decline and who therefore had to find new “work” areas for their members. Now, it was Ibn Wa∂∂à˙ who became the most important teacher of his generation, followed by Baqì and al-Khushanì. Being descendant of a manumitted slave, that was not small achievement. The following anecdote shows how Ibn Wa∂∂à˙ paid careful attention to the way his social position was acknowledged. The anecdote concerns his contemporary, the judge Mu˙ammad b. Ziyàd al-Lakhmì: He was one of the most sensible, indulgent and polite men. One day, I went with him and with a man from Quraysh to visit one of his brothers who was sick. Ibn Ziyàd asked permission to enter his house and when the slave inquired about our identity, he answered: “Tell your master that Fulàn from Quraysh, the jurist Ibn Wa∂∂à˙ and Mu˙ammad b. Ziyàd are here.” He put himself at the end, after us and he called us by our kunya, whereas he gave his ism. His courtesy was excellent and we thanked him.130

Ibn Wa∂∂à˙ was obviously struck by the fact that the judge had mentioned him before giving his own name: a mawlà was put before an Arab and both the mawlà and the Qurashì were referred to by their kunya, a sign of respect.131 128 See Fierro, “The introduction of ˙adìth in al-Andalus (2nd/8th–3rd/9th centuries),” Der Islam 66 (1989), 68–93. 129 In the same way that religious egalitarianism needed not to imply a rejection of the existing social order, see Marlow, Hierarchy, 45–46. 130 Ibn Óayyàn, Muqtabis, ed. M.'A. Makkì (Beirut, 1393/1973), 72. See also Ibn Óàrith al-Khushanì, Qu∂àt Qur†uba, ed. and trans. J. Ribera (Madrid, 1914), 101–2/ 123–5. 131 See Goldziher, “The use of the kunya as a means of paying respect,” in Muslim Studies, trans. S.M. Stern, 2 vols. (London, 1967–71), I, 242.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

239

Conclusions The non-urban Berber population in al-Andalus during the first centuries of Islamic rule did not preserve the memory of the walà" ties established with Arab tribes, if any. That loss might be attributed to the sources at our disposal, which reflect an Umayyad context (see below), and also to the fact that Berber settlement took often place in regions discarded by the Arabs, so that the Berbers did not have any reason for preserving that memory. As regards the local population, the conquest took place mostly by capitulation, so that there was no general enslavement and therefore walà" al-'itq could be established only with a small section of the Andalusì inhabitants. Clientship arising upon manumission seems to have been remembered, both by the descendants of the manumitted slave (documents of manumission were kept in the family) and by society, as that origin carried with itself a certain stigma. Clientship arising upon conversion seems to have been very scarce. On the one hand, the settlement patterns of the Arabs caused that they must have been not interested in promoting the conversion of those on whom their livelihood depended. On the other hand, walà" al-islàm started to be rejected in the central lands of the Islamic world around the 760s. This means that in those regions where there seems to have been applied (Upper Frontier), it could not last long once it enjoyed Islamic disapproval. The reception and acceptance of Medinese/Màlikì law put an end to it, although it is possible to see that its previous existence left traces in both historical and legal sources. Contractual walà", apart from that of conversion, appears in the sources in the context of armed conflict, when different parties establish an alliance against a common enemy. Umayyad rule was decisive in the ways patronate and clientship ties evolved in al-Andalus. The Umayyads were interested in keeping limited the number of Arabs and in monopolizing the use of walà" for their own interest. By the fourth/tenth, the mawàlì appear as a rank within Umayyad elites. The muwalladùn were active as a social group only during the third/ninth century. They were Arabized indigenous inhabitants of the Iberian Peninsula, whose Islamization came about as a result of that Arabization. On its turn, Arabization had been made possible by their living among Arabs and this was the result of the settlement patterns of the Arabs (no amßàr, living on the taxes collected

240

maribel fierro

from the indigenous population). The Arabization/Islamization of the muwalladùn did not lead to the establishment of clientship ties with Arab tribes for different reasons. If the muwalladùn were descendants of converts who enjoyed free status, those early converts had not been offered the possibility of walà" al-islàm (the muwalladùn do not bear Arabic nisbas): because Arabs were not interested in the conversion of the dhimmìs, or because walà" al-islàm was not any more effective or because the Umayyads wanted to monopolize the establishment of walà" ties. If the muwalladùn were recent converts, the reception of Medinese/Màlikì law did not acknowledge validity to the former walà" al-islàm, so that conversion stopped entailing clientship ties with Arab tribes. In any case, conversion in al-Andalus seems to have been mainly due not so much to a voluntary passage to the new religion, but to an involuntary inability to continue being a Christian. The muwalladùn tried to find a place as Muslims in a society where to be a Muslim was still closely connected to being an Arab firstly and secondly, to being a mawlà or a Berber (because of the Berbers’ role in the conquest). The tensions between Arabs and the ruling dynasty provided the muwalladùn leaders with a model of behavior (rebellion) once the Umayyad elites made it clear that they were not willing to easily grant the muwalladùn space among themselves. Once the muwalladùn joined the field in the struggles that were to end with the imposition of the Umayyad caliphate, they had different options in order to give an aim to their own fighting and to legitimize their bid for power. All those options were undertaken by Ibn Óafßùn. When his incorporation into the Umayyad khidma failed because of mawàlì opposition, he acknowledged 'Abbàsid and also Fà†imid legitimacy. He also tried to present himself as a new 'Abd al-Ra˙màn I within a general climate of messianic expectations. And as D. Wasserstein has recently shown, he also tried to present himself as a descendant of the Visigoths, an option probably connected to his return to Christianity. This last move indicates that the repertoire of Muwallad legitimacies was limited to: 1) moving back to the East, that is, renouncing the Umayyad secession of the Islamic “West” and reestablishing links with the Abbasid caliphate 2) abandoning Sunnism and becoming an “heretic” 3) entering eschatological time 4) moving back to the pre-Islamic past of the Iberian Peninsula.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

241

If the first three options involved continuing being a Muslim, the last one made it difficult preserving the convert’s identity. The late development of Arabic culture in al-Andalus is one factor explaining that the muwalladùn did not produce shu'ùbì writings. Their ethnic solidarity found expression in Arabic poetry. It is in the Upper Frontier where we find a more intense expression of a local reaction against the Arabs in the second half of the third/ninth century. I understand it as indicating the loss of muwallad control in the area once the Umayyads decided to support the rule of a local Arab family. Even if the Umayyads had tried to weaken Arab power, the development of the fitna shows that when Arab and muwalladùn rebels sought their intervention, the Umayyads always opted for the former. The muwalladùn’s defeat has its reflection on the fact that when the Umayyad caliphate disintegrated and the ensuing kingdoms were ruled by dynasties (mulùk al-†awà"if ) which could be characterized by their ethnicity (Arab, Berber, Íaqlabì), there were no Muwallad mulùk al-†awà"if. Equal conditions with Arabs were achieved by mawàlì scholars, who paved the way for the muwallad. Ibn Wa∂∂à˙’s role was crucial in this process. The attacks against him coming from the Arab scholar al-Khushanì indicate that Arab elites had realized that from then onwards they would have to share the field of knowledge with the non-Arabs on an equal footing, and that mastery of the language was not any longer a recipe for success.

Bibliography Primary Sources Ibn al-'A††àr. Kitàb al-wathà"iq wa-l-sijillàt, ed. P. Chalmeta and F. Corriente. Madrid: Academia Matritense del Notariado/Instituto Hispano-árabe de Cultura, 1983; Spanish trans. P. Chalmeta and M. Marugán as Formulario notarial y judicial andalusí. Madrid: Fundación Matritense del Notariado, 2000. Ibn al-Fara∂ì. Ta"rikh 'ulamà" al-Andalus, ed. F. Codera, 2 vols. Madrid: La Guirnalda, 1891–92 (Bibliotheca Arabico-Hispana, VII–VIII). Ibn Óàrith al-Khushanì. Akhbar al-fuqahà" wa-l-mu˙addithìn, ed. M.L. Ávila and L. Molina. Madrid: Consejo Superior de Investigaciones Científicas/Agencia Española de Cooperación Internacional, 1992 (Fuentes Arábico-Hispanas, III). Ibn Óàrith al-Khushanì. Qu∂àt Qur†uba, ed. and trans. J. Ribera. Madrid: Imprenta Ibérica/E. Maestre, 1914. Ibn Óayyàn al-Qur†ubì. Al-Muqtabas min abnà" ahl al-Andalus, ed. M.'A. Makkì. Beirut: Dàr al-Kitàb al-'Arabì, 1393/1973. id., Al-Muqtabis, ed. P. Melchor Antuña. Paris: Librairie Orientaliste Paul Geuthner, 1937.

242

maribel fierro

id., Al-Muqtabis al-thànì, edición facsímil de un manuscrito árabe de la Real Academia de la Historia (Legado Emilio García Gómez) al cuidado de Joaquín Vallvé Bermejo. Madrid: Real Academia de la Historia, 1999; Spanish trans. Ma˙mùd 'Ali Makkì y Federico Corriente as Crónica de los emires Al˙akam I y 'Abdarra˙man II entre los años 796 y 847 [Almuqtabis II–1]. Zaragoza: Instituto de Estudios Islámicos y del Oriente Próximo, 2001. Ibn Óazm. Jamharat ansàb al-'arab, ed. 'Abd al-Salàm Mu˙ammad Hàrùn. Cairo: Dàr al-Ma'àrif, 1962. Ibn Khayr. Fahrasa, ed. F. Codera and J. Ribera, 2 vols. Zaragoza: F. Comas, 1893 (Bibliotheca Arabico-Hispana, IX–X). Ibn al-Qù†iyya. Ta"rìkh iftità˙ al-Andalus, ed. and trans. P. de Gayangos and J. Ribera. Madrid: Tipografía de la Revista de Archivos, 1926. Ibn Sharìfa, M. “Óawla mu"arrikh andalusì majhùl min al-'aßr al-umawì,” Awràq ˝adìda 7–8 (1984–85), 57–68. 'Iyà∂. Tartìb al-madàrik. 8 vols. Rabat: Wizàrat al-Awqàf wa-l-Shu"ùn al-Islàmiyya, . . . –1983. al-Marràkushì. Kitàb al-Mu'jib, ed. R. Dozy, 2nd ed. Leiden, 1881; repr. Amsterdam: Oriental Press, 1968; trans. E. Fagnan as Histoire des Almohades. Argel, 1893; ed. Cairo, 1949. Tres textos árabes sobre beréberes en el Occidente islámico. Ibn 'Abd al-Óalìm (s. VIII/XIV), Kitàb al-ansàb. Kitàb Mafàjir al-barbar (Anónimo). Abù Bakr Ibn al-'Arabì (m. 543/1149), Kitàb “awàhid al-#illa, ed. M. Ya'là. Madrid: Consejo Superior de Investigaciones Científicas/Agencia Española de Cooperación Internacional, 1996 (Fuentes ArábicoHispanas, XX). al-Tujìbì. Barnàmaj, ed. 'A.Ó. Manßùr. Lybia/Tunis: al-Dàr al-Magribiyya li-l-Kitàb, 1981. al-'Udhrì. Nußùß 'an al-Andalus min Kitàb tarßì'al-akhbàr wa-tanwì' al-àthàr, ed. 'A. alAhwànì. Madrid: Ma†ba'at Ma'had al-Diràsàt al-Islàmiyya, 1965; Spanish trans. F. de la Granja as La Marca Superior en la obra de al-'U≈rì. Zaragoza, 1966. Secondary Sources Acién, M. “Settlement and fortification in Southern al-Andalus: the formation of a land of ˙ußùn,” in M. Marín, ed., The Formation of al-Andalus. Part 1: History and Society, The Formation of the Classical Islamic World, 46. Ashgate: Variorum, 1998, 347–76; English trans. of “Poblamiento y fortificación en el sur de alAndalus: la formación de un país de ˙ußùn,” Actas III Congreso de Arqueología Medieval Española. Madrid, 1989, 137–47. id., Entre el feudalismo y el Islam: 'Umar ibn Óafßùn en los historiadores, en las fuentes y en la historia. Jaén: Universidad de Jaén, 1994; repr. with a new introduction, Jaén, 1997. id., “El papel de la ideología en la caracterización de las formaciones sociales islámicas,” Hispania 200 (1998), 915–68. Ávila, M.L. and L. Molina. “La Marca Superior de al-Andalus en el siglo VIII: el asceta Ibn al-Mugallis y los Banù Salama,” Homenaje al Profesor J. Ma Fórneas Besteiro, 2 vols. Granada: Universidad de Granada, 1995, II, 703–709. Barceló, M. “The Manifest Caliph: Umayyad Ceremony in Córdoba, or the Staging of Power,” in M. Marín, ed., The Formation of al-Andalus. Part 1: History and Society, The Formation of the Classical Islamic World, 46. Ashgate: Variorum, 1998, 425–56; English trans. of “El califa patente: el ceremonial omeya de Córdoba o la escenificación del poder,” Estructuras y formas del poder en la historia, Salamanca, 1991. Bulliett, R. Conversion to Islam in the Medieval Period. An Essay in quantitative history. Cambridge/Mass./London: Harvard University Press, 1979.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

243

id., “Conversion stories in early Islam,” in M. Gervers and R.J. Bikhazi, eds., Conversion and continuity. Indigenous Christian communities in Islamic lands, Papers in Mediaeval Studies, 9. Toronto: Pontifical Institute of Mediaeval Studies, 1990, 123–33. id., Islam. The view from the edge. New York: Columbia University Press, 1994. Calder, N. Studies in early Muslim jurisprudence. Oxford: Clarendon Press, 1993 Chalmeta, P. Invasión e islamización. La sumisión de Hispania y la formación de al-Andalus. Madrid: Editorial Mapfre, 1994. id., art. “Muwallad,” EI 2. Coope, J. The martyrs of Córdoba: community and family conflict in an age of mass Conversion. Lincoln/London: University of Nebraska Press, 1995. Crone, P. Roman, provincial and Islamic law. The origins of the Islamic patronate. Cambridge: Cambridge University Press, 1987. id., Slaves on horses. The Evolution of the Islamic Polity. Cambridge: Cambridge University Press, 1980. id., art. “Mawlà,” EI 2, VI, 874–882. id., “Were the Qays and the Yemen Political Parties?” Der Islam 71 (1994), 1–57. Dozy, R. Supplément aux dictionnaires arabes. 2 vols. Leiden/Paris: Brill, 1967. de Epalza, M. “Mozarabs: an Emblematic Christian Minority in Islamic al-Andalus,” in M. Marín, ed., The Formation of al-Andalus. Part 1: History and Society, The Formation of the Classical Islamic World, 46. Ashgate: Variorum, 1998, 183–204 (Previously published in S. Kh. Jayyusi, ed., The Legacy of Muslim Spain. Leiden, 1992, 149–70). id., “Falta de obispos y conversión al Islam de los cristianos de al-Andalus,” AlQan†ara 25 (1994), 385–400. de Felipe, H. Identidad y onomástica de los beréberes de al-Andalus. Madrid: Consejo Superior de Investigaciones Científicas, 1997. Fernández Félix, A. Cuestiones legales del Islam temprano: La 'Utbiyya y el proceso de formación de la sociedad islámica andalusi, Madrid: CSIC, 2003, 436–458. id., “Biografías de alfaquíes: la generación de al-'Utbì,” Estudios Onomástico- Biográficos de al-Andalus. VIII, ed. M.L. Ávila and M. Marín. Granada/Madrid: Consejo Superior de Investigaciones Científicas, 141–175. id., and M. Fierro. “Cristianos y conversos al islam en al-Andalus bajo los Omeyas. Una aproximación al proceso de islamización a través de una fuente legal andalusí del s. III/IX,” Anejos Archivo Español de Arqueología 23 (2000), 417–29. Fierro, M. “Bazì', mawlà de 'Abd al-Ra˙màn I, y sus descendientes,” Al-Qan†ara 8 (1987), 99–118. id., “The introduction of ˙adìth in al-Andalus (2nd/8th–3rd/9th centuries),” Der Islam 66 (1989), 68–93. id., “Árabes, beréberes, muladíes y mawàlì. Algunas reflexiones sobre los datos de los diccionarios biográficos andalusíes,” Estudios Onomástico-Biográficos de al-Andalus. VII, ed. M. Marín. Madrid: Consejo Superior de Investigaciones Científicas, 1995, 41–54. id., “Los mawàlì de 'Abd al-Ra˙màn I,” Al-Qan†ara 20 (1999), 65–97. id., “Los Banù 'Àßim al-ˇaqafì, antepasados de Ibn al-Zubayr,” Al-Qan†ara 7 (1986), 53–84. id., “On al-fà†imì and al-fà†imiyyùn,” Jerusalem Studies in Arabic and Islam 20 (1996), 130–61. id., La heterodoxia en al-Andalus durante el periodo omeya. Madrid: Instituto HispanoArabe de Cultura, 1987. id., “El alfaquí beréber Ya˙yà b. Ya˙yà, 'el inteligente de al-Andalus,” Estudios Onomástico-Biográficos de al-Andalus. VIII, ed. M.L. Ávila and M. Marín. Granada/ Madrid: Consejo Superior de Investigaciones Científicas, 1997, 269–344.

244

maribel fierro

id., “Four questions in connection with Ibn Óafßùn,” in M. Marín, ed., The Formation of al-Andalus. Part 1: History and Society, The formation of the classical Islamic world, 46. Ashgate: Variorum, 1998, 291–328; English trans. of “Cuatro preguntas en torno a Ibn Óafßùn,” Al-Qan†ara 16 (1995), 221–57. id., “Sobre la adopción del título califal por 'Abd al-Ra˙màn III,” Sharq al-Andalus 6 (1989), 33–42. id. and M. Marín. “La islamización de las ciudades andalusíes a través de sus ulemas (ss. II/VIII-comienzos s. IV/X),” in P. Cressier and M. García-Arenal, eds., Genèse de la ville islamique en al-Andalus et au Maghreb occidental. Madrid: Casa de Velázquez/Consejo Superior de Investigaciones Científicas, 1998, 65–98. Glick, Th. “The ethnic systems of premodern Spain,” Comparative Studies in Sociology 1 (1978), 157–71. id. and O. Pi-Sunyer. “Acculturation as an explanatory concept in Spanish history,” Comparative Studies in Society and History 11 (1969), 136–54. Goldziher, I. “The use of the kunya as a means of paying respect,” Muslim Studies, ed. S.M. Stern, trans. C.R. Barber and S.M. Stern, 2 vols. New York: University Press, 1971–77. Graham, C. “The meaning of slavery and identity in al-Andalus: the epistle of Ibn Garcia,” Arab Studies Journal/Majallat al-Diràsàt al-'arabiyya 3/I (1995), 68–79. Guichard, P. Al-Andalus. Estructura antropológica de una sociedad islámica en Occidente. Barcelona: Barral, 1976; repr. Granada: Servicio de Publicaciones de la Universidad, 1998. id., art. “Tudjìb (Banù),” EI 2, X, 582–584. Juynboll, G.H.A. “Shu'ba b. al-Óajjàj (d. 160/776) and his position among the traditionists of Baßra,” Le Muséon 111 (1998), 187–226. Kennedy, H. Muslim Spain and Portugal. A political history of al-Andalus. London/New York: Longman, 1996. Lévi-Provençal, E. Histoire de l’Espagne musulmane. 3 vols. Paris/Leiden, 1950–53. Manzano, E. La frontera de al-Andalus en época de los omeyas. Madrid: Consejo Superior de Investigaciones Científicas, 1991. id., “The Settlement and Organisation of the Syrian Junds in al-Andalus,” in M. Marín, ed., The Formation of al-Andalus. Part 1: History and Society, The Formation of the Classical Islamic World, 46. Ashgate: Variorum, 1998, 85–114; English trans. of “El asentamiento y la organización de los íund sirios en al-Andalus,” AlQan†ara 14 (1993), 327–59. id., “Las fuentes árabes sobre la conquista de al-Andalus: una nueva interpretación,” Hispania 59/II, n. 202 (1999), 389–432. Marín, M. “Una familia de ulemas cordobeses: los Banù Abì 'Ìsà,” Al-Qan†ara 6 (1985), 291–320. id., “”ùrà et ahl al-“ùrà dans al-Andalus,” Studia Islamica 62 (1985), 25–51. Marín Guzmán, R. “Social and ethnic tensions in al-Andalus,” Islamic Studies 32/II (1993), 279–318. id., “The causes of the revolt of 'Umar ibn Óafßùn in al-Andalus (880–928). A study in Medieval Islamic social history,” Arabica 42 (1995), 180–221. Marlow, L. Hierarchy and egalitarianism in Islamic thought. Cambridge, 1997. Martínez-Gros, G. L’idéologie omeyyade. La construction de la légitimité du Califat de Cordoue (X e–XI e siècles). Madrid: Casa de Velázquez, 1992. Meouak, M. Pouvoir souverain, administration centrale et élites politiques dans l’Espagne umayyade (II e–V e/VIII e–X e siécles). Helsinki: Academia Scientiarum Fennica, 1999. Molina, L. “Familias andalusíes: los datos del Ta"rìj 'ulamà" al-Andalus de Ibn alFara∂ì,” EOBA, II, ed. M.L. Àvila. Granada, 1989, 19–99; EOBA, III, ed. M.L. Àvila. Granada, 1990, 13–58; and EOBA, IV, ed. L. Molina. Granada, 1990, 13–40.

MAWÀLÌ

and

MUWALLADÙN

in al-andalus

245

id., “An Arab among Muwallads: Mu˙ammad ibn 'Abd al-Salàm al-Khushanì,” in M. Marín, ed., The Formation of al-Andalus. Part 1: History and Society, The Formation of the Classical Islamic World, 46. Ashgate: Variorum, 1998, 115–28; English trans. of “Un árabe entre muladíes: Mu˙ammad b. 'Abd al-Salàm al-Ju“anì,” Estudios onomástico-biográficos de al-Andalus, VI. Madrid, 1994, 337–51. Monroe, J.T. The Shu'ùbiyya in al-Andalus. Berkeley/Los Angeles: University of California Press, 1970. Morony, M. “The Age of Conversions: A Reassessment,” in M. Gervers and R.J. Bikhazi, eds., Conversion and continuity. Indigenous Christian communities in Islamic lands. Toronto, 1990, 135–50. Oliver, D. “Sobre el significado de mawlà en de la historia omega de al-Andalus,” Al-Qan†ara XXII, 2001, 321–344. id., “Una nueva interpretación de ‘árabe’, ‘muladí’ y ‘mawla’ como voces representativas de grupos socials,” Proyección histórica de España en sus tres culturas: Castilla y Léon, América y el Mediterráneo, Arabe, hebreo e historia de la medicina, 3. Valladolid: Junta de Castilla y León, 1993, 143–55. Oliver Asín, J. Conferencias y apuntes inéditos, ed. D. Oliver. Madrid: Agencia Española de Cooperación Internacional, 1996. Penelas, M. “Some Remarks on convertion to Islam in al-Andalus,” Al-Qan†ara XXIII (2002), 193–200. de la Puente, C. “Entre la esclavitud y la libertad: consecuencias legales de la manumisión según el derecho màlikí,” Al-Qan†ara 21 (2000), 339–60. Safran, J.M. The second Umayyad caliphate. The articulation of caliphal legitimacy in alAndalus. Cambridge/Mass./London, 2000. Shatzmiller, M. The Berbers and the Islamic state. The Marìnid Experience in Pre- Protectorate Morocco. Princeton: Markus Wiener Publisher, 2000. Simonet, F.J. Historia de los Mozárabes de España. Madrid: Real Academia de la Historia, 1897–1903. Viguera, M.J. Aragón musulmán. Zaragoza: Mira, D.L., 1988. Wasserstein, D. “Inventing Tradition and constructing Identity: The Genealogy of 'Umar Ibn Óafßùn between Christianity and Islam,” Al-Qan†ara XXIII (2002), 269–98. Wickham, Ch. “The uniqueness of the East,” Journal of Peasant Studies 12 (1985), 166–96. Wolf, K.B. Christian martyrs in muslim Spain. Cambridge, Cambridge University Press, 1988. Wright, R. “The End of Written Ladino in al-Andalus,” in M. Fierro and J. Samsó, eds., The formation of al-Andalus. Part 2: Language, Religion, Culture and the Sciences, The formation of the Classical Islamic World, 47. Aldershot: Ashgate Variorum, 1998, 19–36; English trans. of “La muerte del ladino escrito en al-Andalús,” Evphrosyne. Revista de Filología Clásica, nóva série, 22 (1994), 255–68.

CONVERSION-BASED PATRONAGE AND ONOMASTIC EVIDENCE IN EARLY ISLAM Richard W. Bulliet

Scholarly opinion has long maintained that in the early Islamic period the word mawlà denoted a non-Arab convert to Islam who entered into a relationship of clientage with an Arab tribe by converting “at the hands of ” an Arab. At a certain point in time, however, generally thought to coincide with the mid eighth-century transition from Umayyad to 'Abbàsid rule, this meaning of the word fell into abeyance. Non-Arab Muslims subsequently were able to make their way in Muslim society without Arab tribal patronage, or whatever else was implied by the relationship of walà". Since conversion to Islam and the status and identity of the convert are central to this scheme of development, it is reasonable to ask how well the scheme fits with other conceptualizations of conversion as a social process. Religious conversion has only slowly gained prominence as a focus of scholarly attention. The pioneering studies of William James and Arthur Darby Nock, among others, took as a model the phenomenon of Protestant conversion in America from the Great Awakening of the seventeenth century down to the surge of evangelical preaching that marked the early decades of the twentieth century. Other scholars contented themselves with nominalist and elitist readings of the available data and accepted with little question the idea that the professed beliefs of rulers and intellectuals, on the one hand, and missionary exploits and the spread of ecclesiastical officialdom, on the other, adequately mirrored religious change in society. Systematic inquiry into the process of that religious change as a broad social phenomenon scarcely arose. In Islamic studies, the hoary myth of conversion by the sword was readily accepted, despite a paucity of evidence, until the idea of differential taxation emerged as an alternative model. In this model, the caliphal promulgation of a new fiscal system in which nonMuslims bore the burden of a special tax ( jizya) prompted non-Arabs to convert from their ancestral faiths as a matter of tax avoidance. Both conversion by the sword and conversion by taxation pressure

onomastic evidence in early islam

247

logically implied that the first century after Mu˙ammad’s death in 632 CE must have been the heyday of conversion since that was when the conquests were most vigorous and when the conquered populations were first subjected to discriminatory taxation. Though conversion to Islam remained on the fringes of scholarly interest, it seemed reasonable that the status of mawàlì (plural of mawlà), in the sense of non-Arab converts to Islam, should be particular to the pre-'Abbàsid period because that was when the bulk of conversion was (tacitly) assumed to have taken place. Change in the meaning of the word after that period, in other words, might plausibly be ascribed to its obsolescence: since there was no one left to convert, to take the most extreme statement of the implied situation, there could be no new mawàlì, and hence no more use for this particular meaning of the word. Some twenty years ago, I and others attempted to draw attention to some of the unspoken assumptions in the history of conversion to Islam and to suggest alternative ways of thinking about the problem. Using different methods and drawing upon previously unexamined bodies of information, I argued that conversion to Islam during the period in question probably took place over a period of three or four centuries, with the greatest numerical increases in the size of the Muslim community occurring after—well after—the Umayyad-'Abbàsid transition in the eighth century. Though the particular methods used to reach these conclusions did not convince all readers, the subsequent tendency in the field, reflected even in the work of scholars who retain doubts as to those particular methods, has been to move toward a distinctly later timetable of conversion. This change in chronological assumptions raised new questions about the institution of conversion-linked patronage and the changing meaning of the word mawlà that at one time denoted it. In the new chronology, non-Arab converts would seem to have been called mawàlì only when they were comparatively few, and the designation would seem to have disappeared as they became more numerous. Moreover, the change in terminology would appear to have taken place spontaneously, that is, without official prompting or sanction, early in the course of a much longer conversion process and could not therefore be taken either as an implicit marker or as a linguistic consequence of the end of the conversion era. Underlying the question of how the change in the meaning of mawlà might have related to the chronology of conversion lies a

248

richard w. bulliet

deeper one: how plausible is the purported institution of conversionlinked patronage, i.e., the status of mawlà as generally understood, in light of other models of conversion as a societal process? We must first acknowledge the particular situation of non-Muslim prisoners of war, who undoubtedly constituted a substantial portion of the early mawàlì population. Deprived of their freedom, and in many instances assigned as slaves to one of their captors, these unfortunates surely saw in conversion a route to bettering their circumstances. Surviving sources may not permit a detailing of conversion under these coercive circumstances, but a relationship of walà" can easily be imagined as a condition of manumission-linked conversion. But unless one assumes that all, or almost all, converts down to the year 750 CE were prisoners of war—a fairly extreme reversal of the traditional models of rapid conversion based on death threats and tax burdens—one must either assume that the word mawlà did not include free converts, i.e., non-prisoners of war who chose, for whatever reason, to embrace the Muslim faith, or attempt to understand what a relationship of walà" may have meant for such “free” converts. In the abstract, a person seeking to become a Muslim under noncoercive circumstances might be imagined frequenting the company of Muslim Arabs and eventually uttering the confession of faith before one of them. That Muslim Arab would then accept the non-Arab free convert as a protected client of his tribe. For this to have been literally true, one must assume: 1) that the prospective convert understood the conversion process, with its attendant client membership in an Arab tribe, despite a probable poor command of Arabic; 2) that Arab Muslims were sufficiently well spaced geographically to be accessible to prospective free converts, the vast majority of whom lived in dispersed farming villages; and 3) that each individual Arab (male) had, and realized that he had, the capacity to confer his tribe’s patronage on a convert. Of course, no one would insist absolutely on these three conditions obtaining in every case of conversion, or on all free conversions taking place in this indivuated manner. But if these conditions had not been present in a substantial proportion of conversion situations prior to 750, the sense of “mawlà” as it comes through in the later historical narratives as a general word for “nonArab convert to Islam,” as opposed to a specific word restricted to prisoners of war and freedmen, would be hard to explain. So how might one imagine people becoming mawàlì through free conversion? Geographically, the narrative sources lead us to believe that the

onomastic evidence in early islam

249

Arab armies of the Ràshidùn and Umayyad periods were largely concentrated in garrisons and that Arab tribes and individuals seldom settled in the countryside or in communities that did not have an active Arab governing presence. If we can use the term “presentation conversion” for conversion “at the hands of ” an Arab Muslim, the standard phrase in the sources, it is evident that presentation conversion must have been much easier for non-Arabs who lived in close proximity to the Arab garrisons than for those who lived remote from an Arab presence. Moreover, since there would be no reason for non-Muslims to circulate information on the process one would have to go through to convert to Islam, it seems safe to assume that the further non-Muslims lived from an Arab Muslim presence, the less the likelihood that they would understand that presentation was a part of the conversion process, or that presentation would normally or necessarily be accompanied by the conferral of client status by an Arab tribe. Given a nearly complete lack of Arabic literacy among non-Arab non-Muslims, the absence of a mutually intelligible language by which the generality of non-Arabs could converse with the Arabs, and the improbability of non-Muslim elites choosing to disseminate information on conversion to their flocks, how can it be supposed that knowledge of a system of presentation conversion and Arab tribal patronage came to be known outside the narrow range of prisoners of war and residents of Arab governing centers? By the same token, how is one to imagine a non-Arab convert operationalizing the status of mawlà in situations where the Arab (or mawlà of an Arab tribe) at whose hands he converted was not present to confirm and enforce that status? For example, if a non-Arab presented himself for conversion before an ordinary tribesman of the tribe of Tamìm and was accepted as a tribal client by that individual and his family, how did other members of the tribe of Tamìm, particularly those living at a distance or in another military camp, or members of any other Arab tribe, for that matter, learn that the convert was now a Tamìmì client? One way would be for the mawlà to adopt the name al-Tamìmì, and perhaps some other distinctively Arab personal name. But if conversion clientage was of concrete benefit, what was to prevent just any non-Arab from adopting an Arab proper name and a tribal surname (nisba)? Or a non-Arab who had converted to Islam but not gone through the process of presentation conversion from entering a town and announcing that

250

richard w. bulliet

he, or his father, was a mawlà of the tribe of Tamìm? In the absence of rosters, how could patronage be controlled? And how could anyone be sure that a Muslim of non-Arab background had ever gone through the process of presentation and conferral of walà"? Complicating this situation are the cases of the prospective nonArab convert approaching a non-Arab Muslim mawlà, i.e., an earlier convert, for the purpose of presentation conversion and gaining thereby, at second hand so to speak, the patronage of that mawlà’s tribe. Logically, given the disproportion between the small, scattered Arab populations and the immeasurably larger conquered populations, this form of conversion should have given rise to a larger and larger population of non-Arab converts whose conversion had been accomplished through presentation before other non-Arab mawàlì (or mawàlì of mawàlì ) rather than through actual contact with Arabs. If patronage had been duly conferred in these instances of conversion (and signaled by use of a tribal nisba), the total numbers of Arab tribesmen, mawàlì included, would have steadily risen, and their actual identity as people of Arab genetic lineage would have become increasingly diluted. Theoretical operational matters of this kind may not be illuminated by the meager and late historical sources available. But they may perhaps be obviated by reformulating the idea of presentation conversion and conversion-linked patronage for free converts. Suppose one were to redefine mawàlì as “that subset of non-Arab converts to Islam who were accepted as clients of an Arab tribe” and dispense with the notions that presentation was required for conversion and that presentation, as opposed to prisoner of war status, normally led to tribal patronage. Most of the awkward circumstances outlined above fade away under this redefinition, but some remain. First, why do historical sources give the impression that (all) non-Arab converts were classified as mawàlì until well into the Umayyad period? Second, how did conversion occur in the absence of presentation and patronage? Third, why did the institution of conversion-linked patronage, even as a subset of overall conversion, disappear by the late eighth century when the generally accepted long chronology of conversion suggests that the conversion process as a whole was still gaining momentum and the most expansive period of conversion was still in the future? Two theoretical models of innovation diffusion may help answer these questions. In my 1979 work on this subject I introduced the

onomastic evidence in early islam

251

first model, which understood innovation diffusion in terms of access to information. Starting from the self-evident premise that one cannot consciously adopt an innovation that one knows nothing about, I sought to outline the parameters that might have affected the dissemination of information about the new religion of Islam among the conquered population: 1) scarcity of non-Arab speakers of Arabic; 2) scarcity of bilingual individuals able to translate information about Islam from Arabic into one of the languages spoken among the conquered population; 3) rarity of literacy; 4) absence of media of mass communication other than those controlled by non-Muslim religious elites, e.g., preaching; 5) residential concentration of Arabs in the aftermath of the conquests as compared with broad geographical dispersal of residence among the conquered population. Since all of these factors would have inhibited the dissemination of information about Islam, and there seemed to be no offsetting factors, such as organized missionary activities, I concluded that conversion must have taken place comparatively slowly. To be sure, instances of mass conversion in which, for example, a tribal or community leader might adopt Islam on behalf of his followers, could not be discounted; but if they were important, they seem to have left few traces in the historical sources. Critics of the empirical research that prompted me to apply a model of information-based innovation diffusion noted that the sources I used, biographical dictionaries emanating from urban milieus of the tenth century and later, could not reasonably be thought of as representing the rural as well as the urban population. Since the most basic object of my study was to argue for a much slower rate of conversion than those previously imagined, this caveat did not do my thesis much damage. No one has ever proposed that rural districts change religious identity faster than cities do so corrections proposed to take account of an under-representation of rural folk in my sources tended to make the conversion process even slower. Nevertheless, it would be desirable to account in a systematic way for a likely rural-urban differential in conversion. The second model of innovation diffusion attempts just that. It proposes that since new things or ideas have to originate in specific geographic locales, it should be possible to map their spread in both space and time. Theorists ask one to imagine a pebble dropped in a body of still water and observe the ripple that radiates from its point of entry. The ripple moves outward in a circle that encompasses

252

richard w. bulliet

a larger and larger area as time progresses. By the same token, they argue, adoption of an innovation radiates outward from the nodal point or points where it first manifests itself. Hence, if one collects the dates of first usage from sampling points at varying distances from the diffusion nodes, one should find that equidistant points first acquire the innovation after approximately the same lapse of time from the date of its first appearance at the nodal point. A series of lines linking similar dates of first adoption outside a node, therefore, should form a series of concentric circles, with due allowance made for particularities of landscape such as rivers or mountains. In modern times, mass communication media may stand in the way of this sort of model being empirically demonstrated. An innovation that first appears in a capital city, for example, might be reported on radio or television and thus become known in all parts of the country at exactly the same moment. Similarly, programmed dissemination, whether by a private corporation or a government agency, might target specific places for reasons unrelated to distance from the diffusion node. Different problems arise in trying to apply the model to pre-modern situations. The most common one is a lack of precise information on dates of first appearance of an innovation. The Renaissance-era invention of the pivoting front axle of a four-wheeled wagon, for example, shows up in drawings and paintings from various parts of Europe. But the preservation of such representations is mostly a matter of chance, and their dating is often problematic. Over time, the innovation became general; but lacking data about nodal points and dates of first appearance, a pattern of diffusion cannot be discerned. For the spread of Islam in the medieval period, problems of data loom large. The dates when mosques were founded or expanded are sometimes given, but they pertain mostly to major cities. Óajj records that might ideally note where pilgrims hailed from are non-existent. The quantifiable data compilations that do exist, namely, the thousands of biographies contained in the biographical dictionaries known generally as “Classes” (ˇabaqàt), do not at first glance appear to be useful sources. Compiled mostly from the tenth century onward, they are devoted primarily to biographies of urban-dwelling Muslim teachers, students, and dignitaries. In most biographies, there is no explicit statement of the subject’s place of birth or family origin. Yet they do have the virtue of recording how individuals were identified, or identified themselves, in successive chronological periods.

onomastic evidence in early islam

253

Identification is a key aspect of human existence. Our contemporary pattern of given name and surname is normally quite sufficient for everyday purposes, though everyone is aware of overlaps and confusions stemming from identical or similar names. On a larger scale, the system is clearly insufficient, however. Social security numbers, passport numbers, and the like take the place of names in most official transactions that require definitive identification. Identification was flexible and could be quite complex in the early Islamic centuries. Individuals might be known by any combination of given name (ism); name derived from real of prospective offspring (kunya); patrilineal genealogy (nasab), most often limited to father’s name and seldom going beyond great-grandfather; one or more relational adjectives (nisba) derived from a profession or locale associated with individuals or their families at some stage of life; and formal or informal honorific or nickname (laqab). The system was so flexible that the same individual might be identified in several ways in different sources, or even in the same source; and the reader seldom knows which form was most often used on an everyday basis. This onomastic flexibility led to changing patterns of identification over time. These patterns are preserved in the biographical dictionaries, though editorial preference normally fixes on the subject’s given name since that is the portion of the name by which the editor alphabetizes the compilation. Yet it is not uncommon for an editor to situate a biography alphabetically by the subject’s given name but to refer to him by a different name, a nisba perhaps, on other occasions in the same book. Recognizing, then, that there is a measure of artificiality in the way editors identify biographical subjects onomastically, it is nevertheless possible to derive useful information from a quantified treatment of the entire assemblage of names. The remainder of this inquiry will be devoted to exploring this possibility in two ways, using as a source the late-tenth century biographical compilation of alÓàkim al-Naysàbùrì (d. 405/1015) known as the Ta"rìkh Naysàbùr. The full text of this work is lost, but an extant manuscript preserves a complete list of the names of the subjects along with occasional scraps of data from specific entries. Like most compilers, al-Óàkim divided his work into chronological classes. The first three classes—ßa˙àba, tàbi'ùn, atbà' al-tàbi'ìn—he explicitly defines according to degrees of remove from the Prophet Mu˙ammad, allowing one to presume that he conceived of at least

richard w. bulliet

254

the initial part of his work as a listing of individuals who transmitted reports about the words and deeds of the Prophet (˙adìth). The fourth class, atbà' al-atbà', he also labels the “fourth century [sic] after the Prophet.” Classes five through seven he simply numbers, changing the word “century” to “class.” The duration of each class, as determinable from the occasional death dates scattered among the surviving biographical scraps, varies widely and is seldom as short as a single generation. For current purposes the dates have been thrown back thirty years as a way of approximating the early stage of the subjects’ careers rather than the time of their deaths. The following table summarizes the structure of the compilation: Table 1: Biographies in al-Óàkim al-Naysàbùrì’s Ta"rìkh Naysàbùr Class

Dates AH

Dates CE

# of Years

# in Class

# per Year

I II III IV V VI VII

0–45 45–80 80–170 170–240 240–284 284–305 305–358

622–665 665–699 699–786 786–854 854–897 897–917 917–969

45 35 90 70 44 21 53

30 73 88 632 521 311 948

7 2 1 9 12 15 18

Before turning to an analysis of naming patterns, one should observe the substantial variability in the number of biographies per class and the steady increase in the number of biographies per year. The latter observation is entirely compatible with a steady growth in the size of the Muslim community over several centuries, but it could as easily be accounted for in other ways. The former observation cautions the analyst to focus more on percentages or averages per class than on absolute numbers. Finally, the transition period from Umayyad to 'Abbàsid rule that appears crucial to the history of patronage falls in the middle of Class III and would do so whether one used death dates or estimated the floruits of individuals by pushing the dates back twenty or forty years instead of thirty, as shown on the table. Class IV, therefore, should be taken as the definitive post-Umayyad class, that is, the class for which the putative early meaning of the word mawlà has become entirely obsolete. For this reason, the following onomastic analysis will concentrate primarily on the first four classes.

onomastic evidence in early islam

255

The most direct way of addressing the patronage question is to look at the role in personal identification played by tribal nisbas. Any member of an Arab tribe could be identified by a tribal nisba, and so could any mawlà of that tribe. If all or most converts were tribal mawàlì, therefore, an increase in the use of tribal nisbas might signal the course of conversion. Table 2 shows the incidence of identifiable tribal nisbas in the first four classes of al-Óàkim’s compilation and the subsequent incidence in classes V–VII of the nisbas represented in those first four classes: Table 2: Tribal Nisbas in al-Óàkim al-Naysàbùrì’s Ta"rìkh Naysàbùr I

II

III

IV

V

VI

VII

1 2 3 2 2 3 1

1 1 3 0 0 1 1

0 1 0 0 1 2 0

0 5 0 0 27 11 1

0 5 0 0 6 5 4

2 2 0 0 2 3 1

0 1 1 0 6 13 2

Azdì Óamdànì Khuzà'ì Kindì Makhzùmì

1 2 1 1 1

1 0 0 0 0

1 0 5 1 0

1 0 1 1 0

0 0 3 0 0

1 0 1 0 0

Class II Total % of Biographies

13 18%

Màzinì Qurashì Qushayrì

1 3 1

0 14 10

0 3 4

1 1 2

0 9 1

Class III Total % of Biographies

10 11% 8 1 1 9 2 4

3 2 0 6 1 0

0 2 1 1 0 0

7 0 0 3 0 0

Nisba 'Adawì Anßàrì Aslamì Ghifàrì Sulamì Tamìmì Thaqafì Class I Total % of Biographies

'Abdì Bàhilì Bajalì Dhuhlì ÓanΩalì Hilàlì

Class

14 47%

richard w. bulliet

256 Table 2 (cont.) Class

IV

V

VI

VII

Óimyarì Numayrì Qaysì

1 1 1

0 1 2

0 0 1

0 2 0

Class IV–VII Totals % of Biographies

103 16%

45 9%

22 7%

47 5%

Nisba

I

II

III

Nothing in these figures supports the notion that presentation conversion resulting in conversion patronage was the norm in the pre'Abbàsid period. Far from growing, the percentage of individuals identified by tribal nisbas declines markedly from the generation of the Prophet down to the mid-ninth century, after which it remains consistently low. The greatest decline falls in the pre-'Abbàsid period. Moreover, only two individuals are identified specifically as mawàlì, one in the first and one in the second class. In both cases they are represented as clients of noteworthy individuals rather than of tribes, reflecting the understanding of the word that is generally associated with the post-Umayyad period. There are many ways, of course, of accounting for the pattern of tribal nisbas. Editorial policy on the part of the compiler seems unlikely since there is no apparent reason for systematically reducing the percentage of tribal nisbas used in the identification of biographees. But it is entirely possible that many tribal clients did not use the nisba of their tribe. It is also possible that since many of the individuals in the first two classes were known historical personages associated with the religious and political foundations of Islam, less need was felt to include a tribal nisba in their nomenclature. Then again, it might be that so few non-Arabs were included in the early classes, consisting as they do of eminent rather than rank-and-file Muslims, that even if they were all tribal clients, their naming patterns would remain invisible in the data. All in all, however, it is evident that a general practice of conversion-based patronage cannot be confirmed through an examination of tribal nisbas. If onomastic identification by tribe, whether for Arab or non-Arab mawlà, has left rapidly diminishing traces during the period when

onomastic evidence in early islam

257

conversion-based patronage is thought to have been the norm for non-Arab converts, what forms of identification were on the increase? To answer this question we turn to a study of geographical nisbas as represented on Table 3: Table 3: Geographical Nisbas in al-Óàkim al-Naysàbùrì’s Ta"rìkh Naysàbùr Categories of Nisbas

Class

I

A) Primordial Urban Centers Mecca Medina Kùfa Baßra Damascus TOTAL % of Biographies

0 0

B) Eastern outside Khuràsàn Balkh Bukhàrà Samarqand Jurjàn TOTAL % of Biographies

0 0

C) Khuràsàn and Nìshàpùr al-Khuràsànì al-Nìsàbùrì TOTAL % of Biographies Total % Geographical Nisbas

0 0

D) Cities of Khuràsàn Marv Herat Sarakhs Nasà TOTAL 0 % of Biographies 0 Total % of Geographical Nisbas

II

III

IV

V

VI

VII

1 0 1 8 0

0 1 1 2 1

0 1 2 7 0

0 4 1 0 0

0 0 1 2 0

0 1 4 8 2

10 14%

7 8%

10 2%

5 1%

3 1%

15 2%

2 0 0 0

7 2 1 2

14 8 3 5

11 2 5 7

5 7 1 2

12 37 1 25

2 3%

12 14%

30 5%

25 5%

15 5%

84 9%

2 2

1 24

1 376

0 318

0 212

0 463

4 5% 22%

25 28%

377 60%

318 61%

212 68%

463 49%

4 4 2 1

30 10 1 2

27 5 2 9

12 4 2 7

18 17 8 6

0 0

12 14% 63%

44 7%

43 8%

25 8%

49 5%

richard w. bulliet

258 Table 3 (cont.) Categories of Nisbas

Class

I

II

E) Western Cities Baghdàd Ißfahàn Rayy TOTAL % of Biographies

0 0

0 0

III

IV

V

VI

VII

9 3 3

4 3 6

1 22 3

27

1

23

13 2%

13 2%

7 2%

72 8%

3 4 9 5 5 2 0

1 2 10 6 9 1 2

1 0 4 2 9 1 1

0 1 14 5 26 6 1

28 4%

31 6%

18 6%

53 6%

10 3 3

24 5 7

15 6 10

45 21 12

16 3% 77%

36 7%

31 10%

79 8%

0 0

F) Nìshàpùr area, lesser towns Abìward Ustuwà Isfarà"in Bayhaq ˇùs Zawzan Juwayn TOTAL % of Biographies

0 0

0 0

0 0

G) Immediate Nìshàpùr area City quarter District name Villages TOTAL 0 % of Biographies 0 Total % of Geographical Nisbas

0 0

0 0

(Categories B, D, E, and F contain representative names only from Class V onward. Totals in Category G represent minimum numbers because of difficulty of identifying village and rural place names.)

This table should be studied chronological class by chronological class. In Class I there are no geographical nisbas. In fact, the only nisbas that occur are tribal, occurring, as we have seen in 47% of all biographies. Individuals without nisbas are identified for the most part by a combination of ism, kunya, and nasab, with one person being listed as a mawlà. Since this was the generation of those who knew the Prophet personally, many of them have a high enough profile in Islamic tradition to be easily identifiable through minimal naming. There have been many Muslims, for example, named 'Abdallàh ibn al-'Abbàs, but there is only one famous 'Abdallàh ibn al-'Abbàs,

onomastic evidence in early islam

259

the first cousin of the Prophet. So long as one knows one is dealing with the generation of the Prophet’s companions, one does not need more of a name to identify him. People in Class II, defined as those who followed the generation of the Prophet’s companions as transmitters of prophetic lore, are only occasionally as well-known as those in Class I. Not surprisingly, therefore, their names convey more specificity. Geographical specification by nisba appears for the first time and inaugurates a tendency that will grow from designating just under a quarter of this class (the cumulative 22% listed for nisba categories A–C as “Total % Geographical Nisbas”) to designating more than three-quarters by Class IV (the cumulative 77% listed for nisba categories A–G as “Total % Geographical Nisbas”). Identification by place is clearly well on its way to supplanting identification by tribe even in the early Umayyad period encompassed by Class II. As might be expected, the locales signaled by the rising use of geographical nisbas are predominantly the early Muslim urban centers that served as centers of caliphal authority: Mecca, Kùfa, and Baßra. The eastern frontier stronghold of Balkh in northern Afghanistan is a surprising addition. Two people are designated as Nìshàpùrìs, and two others are named Khuràsànì. The latter name is a good indicator that there were few enough Muslims in the east to make a broad provincial name, as opposed to a locality within Khuràsàn, a sufficiently distinctive indicator of identity. In Class III the number of people identified by one or more geographical nisbas—the practice of multiple nisbas begins at this time with two examples—grows to 63% of the total with the largest number, 28%, being called Nìshàpùrì. The pattern of locales signified also changes. The primordial Muslim urban centers in Iraq and Arabia lose importance, and eastern cities, both within Khuràsàn and beyond, gain markedly, totaling as many individuals as are called Nìshàpùrì. Class IV, unequivocally in the post-Umayyad period, continues some of the trends of Class III. The share of nisbas pertaining to primordial cities falls to 2%. The percentage of people called Nìshàpùrì soars to 60%, with many of them bearing two nisbas. On the other hand, the share pertaining to other eastern and Khuràsànì cities falls off. Even combined with the newly appearing western cities of Baghdàd, Ißfahàn, and Rayy their total reaches only 14% (nisba cat-

260

richard w. bulliet

egories B, D, E), just half the share of the eastern and Khuràsànì cities (nisba categories B, D) in Class III. The new trends in Class IV are nisbas relating to smaller towns in the immediate hinterland of Nìshàpùr, and to city quarters, rural districts (rustàq), and villages. Both of these newly appearing nisba categories display a good deal of double naming with Nìshàpùr as one of the nisbas, e.g., Nìshàpùrì Juwaynì (from the town of Juwayn near Nìshàpùr) or Nìshàpùrì Quhandizì (from the citadel quarter of the city of Nìshàpùr). The analysis of subsequent classes is more suggestive than definitive because the growing practice of double nisbas, the difficulty of identifying unusual nisbas—presumably most often from villages—and the large size of the classes makes precise quantification more difficult. It seems apparent, nevertheless, that the types of nisba that increase most rapidly are those pertaining to the Nìshàpùr hinterland, both towns and villages, and to the city’s residential quarters. What can be concluded from the data on geographical nisbas? First, place became an important, and eventually a key, component of personal identification starting in the Umayyad period. Secondly, identification by place had virtually supplanted identification by tribe long before the 'Abbàsid seizure of power. And thirdly, the changing array of locales signaled by geographical nisbas is strongly suggestive of a pattern of innovation diffusion, in this case conversion to Islam, extending outward from specific nodes. The sequence of change in frequency of nisba identification over time, as seen from the eastern city of Nìshàpùr, is the following: primordial cities > eastern cities, including Nìshàpùr, with Khuràsàn as a general category > Nìshàpùr itself as the dominant locale with Khuràsànì identification supplanted by specific Khuràsànì cities > increasing incidence of nisbas relating to towns, villages, and rural districts around Nìshàpùr and from the city’s residential quarters. This sequence is precisely what one would expect from a conversion process radiating outward first from the primordial cities of early Islam and subsequently from governing centers established during conquests in the east. Penetration of Islam into rural areas around Nìshàpùr becomes marked only in Class IV covering the period of the late eighth to mid-ninth centuries. These years coincide almost exactly with the most rapid acceleration of growth of the Muslim community in Iran projected on entirely different methodological bases in my work of 1979.

onomastic evidence in early islam

261

The fact that the percentages for the Nìshàpùr hinterland seem to grow still further in classes V through VII, however, strongly suggests that penetration of the countryside was still in fairly early stages during what appeared from my earlier methodology to be the heyday of Iranian conversion. In other words, I am led by the methodology I am employing here to the conclusion that there was, indeed, an urban bias in my earlier work and that the curve of conversion in Iran, taken to include both urban, urbanizing, and rural populations, was probably slower by some fifty to a hundred years than I formerly projected. With respect to the question of conversion-based patronage, a slower rate of conversion, and one that reached only gradually into the countryside from nodes of diffusion determined by the distribution of Arab cantonments and governing centers, makes the notion of presentation conversion before an Arab or a mawlà of an Arab tribe, with a resulting conferring of walà", highly implausible as the primary form of free conversion. To the extent that the practice ever existed, it seems reasonable to conclude that it applied primarily to converts in non-free circumstances and lost prominence as a proportion of all conversion very, very early in the overall conversion process. The onomastic evidence analyzed here, coming from but a single city, can only suggest reasons for the disappearance of conversion patronage. In Nìshàpùr, far from the lands of heaviest Arab settlement, the individuals for whom walà" figured as a major component of personal Muslim identity seem to have been comparatively rare. Non-mawàlì converts must have begun to outnumber mawàlì converts well before the end of the Umayyad period. By early 'Abbàsid times, tribal mawàlì of non-Arab convert origin must have become quite a rare subset of all converts. Indeed, if the dynamic of conversion was truly following the patterns indicated by the information diffusion and nodal diffusion models, becoming a mawlà of an Arab tribe through presentation conversion must have become obsolete well before the advent of the 'Abbàsids. If anyone could become a Muslim, after all, as the data indicate was increasingly the case, why should people any longer seek Arab patronage? The protection that walà" theoretically afforded to converts becomes harder and harder to conceptualize as the Muslim community becomes more ethnically diverse and extends into smaller and more remote communities. Why, then, the prominence of conversion-based patronage in the sources? Given the comparatively late date of the historical narratives

262

richard w. bulliet

that convey this sense of prominence, and the lack of correspondence between those narratives and onomastic data preserved in a biographical dictionary of similar temporal remove from the period in question, I would suggest that the narratives reflect 1) Arab tribal traditions that put strong emphasis on tribal membership, and hence on mawlà status; 2) preferential transmission of historical lore from government/military environments in which non-free converts, i.e., mawàlì originating as prisoners of war, play roles disproportionate to their representation in the overall convert community; and 3) the desire of later families of conversion-based mawàlì origin, or pretending to such an origin, to project upon the past a special status, and early date of conversion, for their ancestors. The last of these explanations could also be indicated by the occasional biographical dictionary entry that preserves at length a putative Arab tribal genealogy, but the very nature of the dictionaries was such as to confine such myth building to specific entries. The bulk of the data available to the compiler reflected a much greater diversity of conversion experience and suggests therefore a more variegated view of the early conversion process and as lessened role for conversion-based patronage in the growth of the Muslim community.

Bibliography al-Óàkim al-Naysàbùrì, Ta"rìkh Naysàbùr, facsimile ed. in Richard N. Frye, ed. The Histories of Nishapur, Cambridge, Mass.: Harvard University Press, 1965. Bulliet, Richard W., Conversion to Islam in the medieval period. An essay in quantitative history, Cambridge, Mass., Harvard University Press, 1979.

THE ECONOMIC STATUS OF THE MAWÀLÌ IN EARLY ISLAM Jamal Juda

Recent studies have devoted much attention to the institution of walà" and to its relationship with the caliphal institution.1 Yet most of these studies have one serious shortcoming: they tend to generalize on the basis of one town or settlement (mißr, pl. amßàr) as if what was observed there was representative for the entire economic situation of all mawàlì in all amßàr. As a direct consequence, these studies do not specify which differences existed between various places in the Islamic empire; for instance, conditions between the two famous Iraqi amßàr of Basra and Kufa differed and in turn these conditions 1 A. von Kremer, Culturgeschichte des Orients unter den Chalifen, 2 vols. (Wien: Wilhelm Braumüller, 1875–1877), II, 136, 145; I. Goldziher, Muhammadanische Studien, 2 vols. (Halle: Max Niemeyer, 1889–90), I, 104; II, 101; G. van Vloten, Recherches sur la domination arabe, le chiitisme, et les croyances messianiques sous le Khalifat des Omayades (Amsterdam: J. Muller 1894), 13, 80; Cl. Cahen, Ta"rìkh al-'Arab wa-l-shu'ùb alislàmiyya mundhu Ωuhùr al-islàm ˙attà bidàyat al-imbirà†ùriyya al-'uthmàniyya, trans. Badr al-Dìn al-Qàsim (Beirut: Dàr al-Óaqìqa lil-ˇibà'a wa-l-Nashr, 1977), 37–40; W.M. Watt, Islam and the integration of society (London: Routledge & Kegan Paul, 1961), 108; R. Levy, The social structure of Islam (Cambridge: Cambridge University Press, 1962), 57; L. Vaglieri, “The patriarchal and Umayyad Caliphates,” in P.M. Holt, A.K.S. Lambton and B. Lewis (eds.), Cambridge History of Islam (Cambridge: Cambridge University Press, 1970), I, 57–104; W. Ende, Arabische Nation und Arabische Geschichte. Die Umayyaden im Urteil Arabischer Autoren den 20. Jahrhunderts (Beirut/Wiesbaden: Frank Steiner Verlag, 1977), 1ff.; P. Crone, The mawàlì in the Umayyad Period (London, Ph.D. Diss., 1973), 92–128; R. Nicholson, A literary history of the Arabs (Cambridge: Cambridge University Press, 1930), 219, 248, 278; Th. Arnold, al-Da'wa al-islàmiyya (Cairo 1947, trans. from the English), 57; U. Haarmann, Geschichte der Arabischen Welt (München: Beckverlag 1987), 95; G. Zaydàn, Ta"rìkh al-tamaddùn al-islàmì, 2 vols. (Beirut: Manshùràt Dàr Maktabat al-Óayàt, 1902, 1906), I, 220; III, 360; M. al-Najjàr, al-Mawàlì fì l-'aßr al-umawì (Cairo 1945), 1–10; M. al-Miqdàd, al-Mawàlì wa-niΩàm al-walà" min al-jàhiliyya ilà awàkhir al-'aßr al-umawi (Damascus: Dàr al-fikr, 1988), 210, 250; Sh. Mu߆afà, Dawlat banì l-'Abbàs (Kuwait 1973), I, 25; F. 'Umar, ˇabì'at al-da'wa al-'abbàsiyya (Beirut: Dàr al-Qalam l-Tibà'a, 1970), 132; A. al-Kha†ìb, al-Óukm al-umawì fì Khurasàn (Beirut: Mu"assasat al-A'lamì/Baghdad: Dàr al-tarbiya, 1975), 14; 'A. al-Dùrì, Muqaddima fì ta"rìkh ßadr al-islàm (Beirut: al-Ma†ba'a alkathùlìkiyya, 1961), 88 and his Muqaddima fì ta"rìkh al-iqtißad al-'arabì (Beirut: Dàr al-Mashriq, 1969), 44; M. Óigàd, Ma"àthir al-shu'ùbiyya fì l-adab al-'arabì (Cairo: Dàr i˙yà" al-kutub al-'Arabiyya, 1961), 163; I. Bay∂ùn, al-Dawla al-umawiyya wa-l-mu'àra∂a, 2 vols. (Beirut: al-Mu"assasa al-Jàmi'iyya lil-Diràsat wa-l-Nashr, 1985), I, 26; H. Atwàn, al-Da'wa al-'abbàsiyya (Amman: Maktabat al-Mu˙taßib, 1984), 17–20.

264

jamal juda

were different from those of Syria and Egypt and so on. Additionally, many scholars have been influenced in their writings by political and/or ideological struggles between Islam and the West or by the current of pan-Arab nationalism or religious ideologies and, unfortunately, some of these biases have found their way in their writings on this topic. At once, most of these studies espouse the view that the Arabs had, during the early period of Islam, deprived the mawàlì of any right to be added to the dìwàn al-'a†à" (alternatively called the dìwàn al-jund ), which, if correct, would have meant that the mawàlì received no annual salary or monthly sustenance or provisions like the Arabs did. It is, moreover, asserted that the mawàlì were overburdened by taxes (kharàj, jizya and 'ushr). These measures— implying no right to any continuous state salary together with overtaxation—naturally would have had a negative effect on the economic conditions of the mawàlì. Taking this line of thinking further, these studies suggest that the mawàlì were consequently forced to join political and religious opposition groups that participated in rebellions ( fitan) against the caliphal state. The purpose of this article is to question the validity of these statements. It does so by scrutinizing our main primary sources to gather information about occupations pursued by mawàlì from the early period of Islam up to the end of the Umayyad period, which roughly corresponds with the first century and a half of Islam, 632–750 CE. A picture is offered of the economic conditions—and their corresponding social standing—of the mawàlì while noting a number of differences in these conditions across the Islamic empire.

Mawàlì in crafts and trade Readers of early primary sources will find it difficult to talk about one unified socioeconomic condition for the mawàlì. In fact, several forms of walà" existed, each offering a different socioeconomic condition. Some of these forms were a continuation of pre-Islamic forms of walà", such as walà" al-'itàqa, walà" al-˙ilf (also known as walà" al'iqd ), walà" al-khidma, walà" al-inqi†à' and walà" al-jiwàr (or al-suknà ). Islam introduced other forms, notably the walà" al-islàm.2 Additionally, allegiance amongst the mawàlì themselves also existed, which can be 2 See also the contributions by Landau-Tasseron, Lecker and Mitter on (dis)continuity with the pre-Islamic period.

the economic status of the

MAWÀLÌ

265

called mawàlì loyalty. As a direct result of these varying forms of walà", varying social conditions for the mawàlì existed. It should also be pointed out that social divisions amongst the mawàlì were not solely determined by their economic condition but rather by their legal and/or social dependence or association with Arabs as well. A sense of the different social and economic conditions for the mawàlì, first, amongst themselves and, secondly, across amßàr can easily be gauged by a listing of the various activities in which the mawàlì were active according to the primary sources: they were merchants, salesmen, brokers, commercial agents, artisans, craftsmen, writers, civil servants, tax collectors, administration officials, workers, rulers, soldiers, military leaders, policemen, private guards, gatekeepers, Qur"àn reciters, scholars, ˙adìth narrators, storytellers, judges, scholars of Islamic law ( fiqh), muftis, educators, teachers, money changers, landowners, goldsmiths, landlords, builders and engineers (more on this below). Various Arabic sources maintain that the majority of those involved in crafts and goods manufactured during the caliphate consisted of mawàlì and slaves.3 In this regard, 'Abdallàh b. 'Amr b. 'Abd al-Qays (d. late first century AH/early eighth century CE) told Óumràn b. Abàn (d. ca. 75/694), the caliph 'Uthmàn b. 'Affàn’s client: “May God increase the number of people like you, so that you would become tailors, tanners and shoemakers.”4 Al-A˙naf b. Qays al-Tamìmì (d. 67/686) is another case in point. When Ziyàd b. Abìhi, Mu'àwiya b. Abì Sufyàn’s governor in Basra, solicited al-A˙naf ’s advice concerning the killing of the mawàlì and slaves of Basra, he wondered, “Do you want the Arabs to be butchers, plasterers ( jaßßàßìn) and doctors (˙ajjàmìn)?”5 The mawàlì excelled in more than one craft or profession.6 3 J. Juda, Die sozialen und wirtschaftlichen Aspekte der Mawàlì in frühislamischer Zeit (Tübingen, Ph.D. Diss., 1983), 105–10; al-Miqdàd, Mawàlì, 211–14. 4 al-Fasawì, Ya'qùb b. Sufyàn. Kitàb al-ma'rifa wa-l-ta"rìkh (ed. Akram Îiyà" al'Umarì, Baghdad: Maktabat al-Muthanna, 1974–76), II, 69; Ibn Qutayba, 'Abdallàh b. Muslim. Kitàb al-ma'àrif (ed. Tharwat 'Ukàsha, Cairo: Dàr al-Ma'àrif, 1960), 439; Ibn 'Abd Rabbihi, A˙mad b. Mu˙ammad. al-'Iqd al-farìd, ed. A˙mad Amìn, A˙mad al-Zayn and Ibràhìm al-Ibyarì (Cairo: Ma†ba'at Lajnat al-Ta"lìf wa-l-Tarjama wal-Nashr, 1940–44), IV, 414; Ibn 'Asàkir, 'Alì b. al-Óasan b. Óibbatallàh. Tahdhìb al-ta"rìkh al-kabìr, ed. 'Abd al-Qàdir Badràn (Damascus: al-Maktaba al-'Arabiyya, 1330–32), VII, 169. 5 Ibn 'Asàkir, Tahdhìb, VII, 15; Mubarrad, Mu˙ammad b. Yazìd, Kitàb al-kàmil fì l-adab, ed. Mu˙ammad Abù l-Fa∂l Ibràhìm and Sayyid Shahàta (Cairo: Ma†ba'at Nah∂at Mißr, n.d.), III, 314; Ibn Abì l-Óadìd, 'Izz al-Dìn Abù Óamìd. Shar˙ nahj al-balàgha, ed. Mu˙ammad Abù l-Fa∂l Ibràhìm, 20 vols. (Cairo: Dàr I˙ya" al-Kutub al-'Arabiyya, 1959), IV, 164. 6 Ibn Qutayba, Ma'àrif, 485.

266

jamal juda

The mawàlì were also involved in commerce, selling and buying locally (within the same mißr) or between various amßàr and even internationally between the Islamic empire and non-Muslim lands.7 In this context, the caliph 'Alì b. Abì ˇàlib (r. 35–40/656–661) once remarked to a Kufan mawlà, when the latter objected to the inequality between mawàlì and Arabs when it came to rewards each group received: “Oh, you mawàlì, these [Arabs] have made you equal to the Jews and the Christians—keep trading, may God bless you. The Prophet Mu˙ammad, peace be upon him, said: ‘Sustenance is ten parts. Nine parts are in trading.’”8 An Umayyad poet expressed this same message in one of his verses: “I have contemplated Iraq’s markets only to find that the mawàlì are the owners of their shops.”9 Furthermore, the involvement of mawàlì in trade and their practice of professions and crafts had drawn al-Jà˙iΩ’s (d. 254/868) attention when he wrote: “The Arabs were neither tradesmen, artisans, physicians, accountants nor cultivators.”10 The mawàlì were able to dominate crafts and trade primarily because they owned slaves and were the patrons of other mawàlì, who acted as craftsmen, salesmen and traders for their mawàlì masters or patrons.11 Eventually some of these even found themselves in a better economic situation than that of their masters or patrons. This phenomenon is apparent from both the quantity and the contents of their correspondence with their masters for emancipation.12

7 S. al-Qawàsmih, al-Tijàra wa-dawlat al-khilàfa fì ßadr al-islàm (Nablus, Mg. Thesis, 1999), 7–20. 8 Kulaynì, Mu˙ammad b. Ya'qùb, Kitàb al-kàfì, ed. 'Alì Akbar al-Ja'farì, 8 vols. (Teheran, 1377–89), V, 318. 9 Jà˙iΩ, 'Amr b. Ba˙r, al-Rasà"il, ed. Mu˙ammad 'Abd al-Salàm Hàrùn, 2 vols. (Cairo: Maktabat al-Khànjì), II, 251. 10 Jà˙iΩ, Rasà"il, I, 69. 11 Ibn Sa'd, Mu˙ammad, Kitàb al-†abaqàt al-kubrà, ed. Eduard Sachau, 9 vols. (Leiden: Brill, 1909), V, 24; Abù 'Ubayda, Ma'mar b. al-Muthannà. Naqà"i∂ Jarìr wa-l-Farazdaq, ed. Anthony Ashly Bevan, 2 vols. (Leiden: E.J. Brill, 1905), II, 350; Wakì', Mu˙ammad b. Khalaf b. Óayyàn, Kitàb akhbàr al-qu∂àt, ed. 'Abd al-'Azìz Mu߆afa al-Maràghì, 3 vols. (Cairo: Ma†ba'at al-Istiqàma, 1947), II, 276; Ißfahànì, 'Alì b. al-Óusayn, Kitàb al-aghànì, 24 vols. (Cairo: Dàr al-Kutub al-Mißriyya, 1927–74), I, 41; Abù Nu'aym, A˙mad b. 'Abdallàh al-Ißfahànì. Óilyat al-awliyà" wa-†abaqàt alaßfiyà", 10 vols. (Beirut: Dàr al-Kitàb al-'Arabì, 1970), I, 90; Sukkarì, Abù Sa'id alÓusayn, Dìwàn Abì l-Aswad al-Du"alì (Beirut, 1974), 127. 12 Ibn Óabìb, Mu˙ammad. Kitàb al-mu˙abbar (Hyderabad: Ma†ba'at Dàr al-Ma'àrif al-'Uthmàniyya, 1942), 342.

the economic status of the

MAWÀLÌ

267

And this can additionally be derived from the fact that there are reports about selling allegiances at very high prices.13 The mawàlì benefited from their relationship with the Arabs and they improved their economic situation. The mawàlì of caliphs, governors, tribal and military leaders were the first beneficiaries. For example, the mawàlì of the caliphs 'Umar b. al-Kha††àb (r. 13–23/ 634–644) and 'Uthmàn b. 'Affàn (r. 23–35/644–656) had the monopoly in the grain trade of Medina during the former’s caliphate.14 The mawlà of one of the caliph Mu'àwiya’s (r. 41–60/661–680) wives, Fàkhita bint QurΩa, asked Mu'àwiya to grant him business concessions in Basra, which the caliph did.15 It is also narrated that Abù Kathìr, a mawlà of the Aslam tribe, enjoyed a good relationship with an Umayyad governor of Egypt, 'Abd al-'Azìz b. Marwàn (d. 85/704), who had exempted this mawlà from paying customs on trade once he entered Egypt as a tradesman.16 Sources also report a connection between trading mawàlì and moneychangers (who were usually also mawàli ) on the one hand, and the bayt al-màl (alternately known as dìwàn al-'a†à", dìwàn al-kharàj and dàr al-rizq) or caliphal treasury on the other. The caliphal treasury also used to borrow money from mawàlì. Moreover, mawàlì were involved in minting coins for the caliphate.17 All anecdotes like the ones just narrated strongly reinforce the impression that the mawàlì enjoyed some form of economic superiority. Some sources even deal with the excessive wealth some mawàlì reportedly had. 'Abdallàh b. Farwa, the caliph 'Uthmàn b. 'Affàn’s mawlà, is one example of such a case since he was considered one of the richest people of Medina.18 Mùsà al-Salmànì, a mawlà of Óa∂ramawt, living in Basra, was one of the wealthiest traders in that city.19 During the reign of the military governor of Iraq, alÓajjàj b. Yùsuf (ca. 41–95/661–714), one of the most well-to-do 13

311.

Ibn Sa'd, ˇabaqàt, VI, 203; Fasawì, Ma'rifa, II, 588; Ibn 'Asàkir, Tahdhìb, VI,

14 Ibn Óamìd, Mu˙ammad b. 'Abìd. Kitàb al-muntakhab min musnad 'Abìd b. Óamìd, ed. Íub˙ì al-Samarrà"ì and Mu˙ammad al-Sa'ìd (Cairo: Maktabat al-Risàla, 1988), 34, 35. 15 Balàdhurì, Ansàb, IV, 207. 16 Balàdhurì, Ansàb, IV, 163; ms. II, 164. 17 Fasawì, Ma'rifa, I, 439–41; al-Qawàsmih, Tijàra, 94–104. 18 Ibn Qutayba, Ma'àrif, 202. 19 Ibn 'Abd Rabbihi, 'Iqd, VI, 96.

268

jamal juda

people in Basra was the mawlà Fayrùz Óusayn; his financial resources reportedly even exceeded those of Basra’s bayt al-màl at the time.20 Al-Madà"inì mentions that a mawlà of the Basran governor 'Abdallàh b. 'Àmir (d. ca. 59/680) had lent the general al-Muhallab b. Abì Íufra (d. ca. 83/703) 300,000 dirhams when he assumed power over the province of Khurasan in 78/697.21 Moreover, Abù Abdrab, a mawlà of the Banì 'Uthra, was one of the wealthiest people of Damascus.22 Also found in the sources is a statement that Dìnàr, a mawlà of the Banì Qu†ay'a in Kufa, had purchased Ibn al-Ash'ath’s maintenance, during his rebellion in the year 82/702 against alÓajjàj and 'Abd al-Malik; al-Óajjàj b. Yùsuf later had him killed.23 The writer Ibn Óabìb reports in his book al-Mu˙abbar about one of the Kufan mawàlì saying: “Abù Dukayn, a mawlà belonging to alJamàliyyìn of Muràd had agreed with his master to give him 100,000 dirham in return for his emancipation. [This] Abù Dukayn [also] sold merchandise to al-'A†à", selling perfumes, slaves and horses and he lent the Muràd tribe 700,000 dirhams until they would receive their salaries (their 'a†à").”24 Inasmuch as economic responsibility was, during this early period, primarily collective due to the still existing tribal system and since the mawàlì were mainly traders, craftsmen and proprietors, having as such economic leverage over Arab patrons, it should come as no surprise that many Arab tribal figures depended on their mawàlì for supporting their families. One of the leaders of the tribe of Qays, for instance, a certain Zufar b. al-Óàrith, is quoted as saying, when one of his mawàlì was killed along with his son during the Marj Ràhi† battle (64/684): “I was sadder for the death of my mawlà than for the death of my son because my mawlà was my source of sustenance for my family.”25 In short, many Arabs at that time were used to borrowing money from their mawàlì whenever they needed it.26 What is more, the Arabs tended to increase their capital by investing it with their mawàlì for trade.27 20

Mubarrad, Kàmil, III, 352–53. ˇabarì, Mu˙ammad b. Jarìr, Ta"rìkh al-rusul wa-l-mulùk, ed. Mu˙ammad Abù l-Fa∂l Ibràhìm, 11 vols. (Cairo: Dàr al-Ma'àrif, 1966–70), VI, 320. 22 Fasawì, Ma'rifa, II, 417. 23 Balàdhurì, Ansàb, ms. II, 1221. 24 Ibn Óabìb, Mu˙abbar, 342. 25 Balàdhurì, Ansàb, V, 157. 26 Ibn Óabìb, Mu˙abbar, 342. 27 Juda, Mawàlì, 113–15. 21

the economic status of the

MAWÀLÌ

269

Mawàlì in administrative functions The early Islamic empire in general continued the administrative systems they encountered in the lands they had conquered. Arab Muslims reached agreements with the administrative officials or with local rulers concluding treaties and truces with them. In the conquered lands, the Muslims introduced changes that were meant to refine the older systems in order to make them more harmonious with the principles of Islam. The dawàwìn al-kharàj (treasures for the land-tax) remained in the hands of the indigenous employees albeit under Arab Muslim supervision. After the indigenous people had actually experienced that Islam upheld their social and economic status, many of them converted to the new religion.28 With the passing of time, many employees in the land-tax treasuries also converted to Islam. Accordingly, most if not all of them became the mawàlì of Arabs and at the same time kept their administrative, social and economic positions. A large number of mawàlì al-'itàqa29 was charged with running land-tax treasuries (al-kharàj dìwàns) for rulers and governors or they were made responsible as tax collectors and clerks in these dìwàns.30 The mawàlì also worked as personal clerks for their patrons or secretaries or as treasurers and officials in charge of their patrons’ money and expenses.31 The mawàlì even worked as clerks while accompanying military campaigns. In this case, the duties of the mawàlì included writing letters and treaties, the distribution of booty and supervision of spending on the army.32 Some mawàlì worked in post 28 Balàdhurì, Futù˙ al-buldàn, ed. De Goeje (Leiden: Brill, 1866), 265; ˇabarì, Ta"rìkh, III, 517–18; Abù Nu'aym, Óilya, I, 81; Ißfahànì, Aghànì, II, 154–55; Ibn Abì l-Óadìd, Shar˙, III, 204; al-Ja˙shiyàrì, Mu˙ammad b. 'Abdùs, Kitàb al-wuzarà" wa-l-kuttàb, ed. Mu߆afa al-Íaqqà (Cairo: Ma†ba'at Mu߆afa al-Bàbì al-Óalabì), 61; J. Juda, al-'Arab wa-l-ar∂ fì l-'Iràq fì ßadr al-islàm (Amman: al-Sharika al-'Arabiyya lilˇibà'a wa-l-Nashr, 1979), 107; al-Miqdàd, al-Mawàlì, 154. 29 Mawàlì al-'itàqa are but one category of mawàlì; these were slaves who were manumitted after which they became the mawàlì of their former masters (see Juda, Mawàlì, 22–25 and Mitter’s contribution in this book). 30 Juda, Mawàlì, 115–20. 31 Khalìfa b. Khayyà†, Ta"rìkh, ed. Suhayl Zakkàr, 2 vols. (Damascus: Wizàrat al-Thaqàfa wa l-Irshàd al-Qawmì, 1967), I, 77; Balàdhurì, Ansàb, II, 191; IV, 101; ms. II, 508, 684; Ibn Qutayba, Ma'àrif, 244; ˇabarì, Ta"rìkh, V, 560; VII, 148; Ißfahànì, Aghànì, III, 290; XVII, 99; Ibn Óajar, al-Ißàba fì tamyìz al-sa˙àba, 8 vols. (Cairo: Ma†ba'at al-Sa'àda/Beirut: Dàr Íadr), 1325, III, 553; Tahdhìb, II, 184; Jahshiyàrì, Wuzarà", 25, 34. 32 Balàdhurì, Futù˙, 394; Ansàb, ms. ii, 43, 194; ˇabarì, Ta"rìkh, iv, 31, 116;

270

jamal juda

offices (al-khatam dìwàns) and most of these also belonged to the category of mawàlì al-'itàqa.33 In this way, the mawàlì dominated in different types of public and private dìwàns. Caliphs, local rulers, leaders and tribal leaders had mawàlì gatekeepers.34 They also had mawàlì and slaves as personal bodyguards.35 Police battalions in big cities consisted mostly of mawàlì.36 This means that the protection of state officials and the maintenance of security inside the amßàr belonged to the responsibilities of mawàlì who would receive their provisions or sustenance and salaries from the state treasury. Mawàlì also worked as papermakers and copyists, especially of the Qur"àn, and this activity constituted a source of income.37 Some mawàlì additionally worked as paid teachers for children in special offices.38 A large number of educated mawàlì were involved in teaching the children of caliphs, leaders and rulers for which they received high wages in return.39 Still other mawàlì composed poetry praising caliphs and emirs in their poetry and at once making this a source of their livelihood.40 Other mawàlì were musicians and singers and taught music to others.41 Mawàlì were also charged with the collection of customs and taxes on trade for the state or region at the border or on trade roads.42 They also worked in post offices,43 as mentioned above, and as controllers in markets to check prices.44 They also assumed responsibility for the caliphal treasury in the city

Sukkarì, Dìwàn, iii, 714; Nìsàbùrì, al-Fa∂l b. Shadhàn al-Azdì, Kitàb al-ì∂à˙, ed. Jalàl al-Dìn al-Óusaynì (Teheran 1972), 91. 33 Khalìfa b. Khayyà†, Ta"rìkh, 412, 427, 468; Balàdhurì, Ansàb, ms. II, 108, 149, 236, 312, 1174; IV, 8; Futù˙, 352; Jahshiyàrì, Wuzarà", 35, 44, 68, 69, 72; Ibn Qutayba, Kitàb 'uyùn al-akhbàr, 4 vols. (Beirut: Dàr al-Kutub al-Mißriyya, Cairo 1952), III, 173; ˇabarì, Ta"rìkh, VI, 544; VII, 211. 34 Juda, Mawàlì, 117. 35 Juda, Mawàlì, 117. 36 Juda, Mawàlì, 117. 37 Ibn Sa'd, ˇabaqàt, VII, 11; Ibn Qutayba, Ma'àrif, 470. 38 Ibn Qutayba, Ma'àrif, 549. 39 Fasawì, Ma'rifa, I, 568; Balàdhurì, Ansàb, III, 100; Ibn 'Asàkir, Tahdhìb, VI, 277; Ibn Óajar, Tahdhìb, I, 317. 40 Balàdhurì, Ansàb, V, 131; Ißfahànì, Aghànì, I, 342. For mawàlì and poetry see van Gelder’s article in this book. 41 Ißfahànì, Aghànì, VIII, 321; Ibn Khaldùn, 'Abd al-Ra˙màn b. Mu˙ammad, alMuqaddima, ed. M. Quatremere, 3 vols. (Paris 1858), II, 360. On mawàlì and music see Kilpatrick’s contribution to this book. 42 Ibn 'Asàkir, Tahdhìb, V, 321; Ibn Óajar, Isàba, I, 491; Tahdhìb, III, 274. 43 Balàdhurì, Ansàb, ms. II, 890. 44 Ibn Sa'd, ˇabaqàt, V, 130.

the economic status of the

MAWÀLÌ

271

in which they were living; others were in charge of their patrons’ houses of finance.45 Summing up, it appears that the bulk of administrative positions in the early caliphal state were in the hands of mawàlì al-'itàqa and mawàlì al-'iqd (contract) or mawàlì al-islàm.46 Simultaneously, the mawàlì were a dominant constituent of police forces in urban centers. All of this evidence underscores the fact that the mawàlì received steady and permanent income provisions and salaries.

Mawàlì as landowners According to the primary sources, mawàlì also became landowners acquiring land through purchase or law grants given to them by the caliph. This was especially true for mawàlì al-'itàqa. Mawàlì al-islàm were originally either prominent or small landowners, just like contract mawàlì were either peasants or very rich. According to some sources, many mawàlì owned villages and country estates in different amßàr.47 In some cases, mawàlì worked as agents for their masters in estates and buildings.48 It is interesting to note here that almost all recent studies (mentioned at the outset of this article) maintain that the Arabs or the caliphal state overburdened the mawàlì with taxes and sometimes even imposed the jizya (head tax meant for nonMuslims) on them, though the vast majority of mawàlì were Muslims: there seems, then, to be much exaggeration about this issue. Islam offered non-Muslims the possibility to embrace the new religion and as such to be exempted from paying jizya. As Muslims they would have to pay alms (zakàt) rather than jizya, which meant that instead

Ibn Sa'd, ˇabaqàt, VII, 178. Unlike the mawàlì al-'itàqa, mawàlì who, as noted above, entered a relationship with their patrons after manumission, the mawàlì al-'iqd and mawàlì al-islàm constitute a patron-client relationship between two free individuals. Usually the mawlà involved had converted to Islam through the mediation of his Arab Muslim patron while the mawàlì al-'iqd converted to Islam by himself (see Juda, Mawàlì, 71–73). 47 Ibn Sa'd, ˇabaqàt, VII, 71, 88; Ibn Qutayba, Ma'àrif, 74, 422; 'Uyùn, II, 35; Balàdhurì, Futù˙, 351, 361, 363, 366–69; Ansàb, ms. II, 1221; ˇabarì, Ta"rìkh, VII, 239; Ibn Óabìb, Mu˙abbar, 345; Ibn Óajar, Ißàba, III, 408; Ibn 'Asàkir, Tahdhìb, IV, 340; Sukkarì, Dìwàn, 127. 48 Balàdhurì, Futù˙, 8, 138, 366; Ansàb, IV, 110; V, 130; Ibn Qutayba, 'Uyùn, I, 214; Mubarrad, Kàmil, IV, 124; ˇabarì, Ta"rìkh, VII, 203; Ißfahànì, Aghànì, XIV, 321. 45

46

272

jamal juda

of paying the state between 50%–80% of their revenues or production, they would only have to pay 10%.49 This offer was of course a strong incentive for non-Muslims to embrace Islam. A considerable number of mawàlì benefited from this offer in the early period of Islam and they improved their economic circumstances after they started to pay zakàt instead of jizya. Yet the spread of Islam had affected the state treasury. The Arabs’ ownership of non-Muslim land (ar∂ al-kharàj ) usually transformed the land automatically into 'ushr land (land-tax for non-Muslims). Consequently, the spread of Islam and the expansion of Arab-Muslim ownership negatively affected the state treasury. These effects began to appear first in Iraq and Greater Syria, after approximately half a century after the Arab-Muslim conquest. This problem, according to the sources, manifested itself critically during the reign of 'Abd al-Malik b. Marwàn (r. 66–86/685–705). To solve this problem, the caliph and his governor al-Óajjàj b. Yùsuf felt compelled to impose land tax ( jizya, jizyat ar∂, kharàj ) on both Arabs and mawàlì for the first time in Iraq and Greater Syria. When 'Umar b. 'Abd al-'Azìz assumed office (r. 99–101/717–719), he pursued this matter further and adopted al-Óajjàj’s measures but only for Iraq and Syria since the problem did not present itself in other regions. He issued a decree stipulating that kharàjiyya land, after 100 AH/718–719 CE, would keep that status regardless of the owner, thus making land-tax the same for Muslims and non-Muslims in the empire, taking away the fiscal advantage for Muslims—at least in Iraq and Syria. This caliph limited the jizya to head tax only and exempted those who had converted to Islam from paying it. The financial crisis appeared in Egypt during the second decade of the reign of al-Walìd b. 'Abd al Malik (86–96/705–714) and in Africa during the rule of Yazìd b. 'Abd al-Malik (101–105/719–723). In Khurasan the financial crisis occurred during the latter part of the caliph Hishàm b. 'Abd al-Malik’s reign (105–125/723–742). The authorities there adopted the same measures initiated by al-Óajjàj and approved by 'Umar b. 'Abd al-'Azìz to save the state treasury.50 These financial measures seem to have had an effect on the status

49 Abù Yùsuf, Ya'qùb b. Ibràhìm, Al-Kharàj (Cairo: al-Ma†ba'a al-Salafiyya wamaktabatuhà, 1392), 44 (on the taxes that al-Îa˙˙àk b. Qays, d. 128/746, had imposed on the people of al-Jazìra al-furàtiyya and Syria). 50 J. Juda, “al-Íul˙ wa-l-'anwa ladà 'ulamà" al-amßàr fì ßadr al-islàm,” Majalat alNajà˙ lil-Ab˙àth, 2/8 (1995), 7–38.

the economic status of the

MAWÀLÌ

273

of both mawàlì and Arabs. As a consequence, some Arab Muslims and mawàlì participated in Ibn al-Ash'ath’s rebellion in Iraq during 'Abd al-Malik’s rule (80–83/699–702); both landowners as well as leaders were amongst them. The same thing happened to the (mawlà) governor Yazìd b. Abì Muslim (d. 102/720–721) in Africa when the Berbers and some Arabs rebelled against him and killed him.51 In Khurasan, later, the mawàlì and some Arabs upheld the 'Abbàsid call and participated in the rebellion that ultimately resulted in the overthrow of Umayyad rule in 132/750. Based on the above, it is unreasonable to accept the conclusions of some modern studies that the Arabs had overburdened the mawàlì with taxes and head tax. The only problem was the state’s imposition of land tax ( jizyat al-'ar∂ ), for the first time, on the properties of mawàlì and Arabs after the financial crisis of the state treasury, which made it unable to make ends meet and to provide provisions and salaries. Therefore, the mawàlì landowners, during the first Islamic century, enjoyed a prosperous life after embracing Islam and concluding agreements of allegiance with the Arab Muslims until the 'Abbàsid caliphs of later came to impose the taxing system more strictly which did lead to a somewhat diminished economic position for the mawàlì.

Mawàlì in the military Primary sources stress that the mawàlì participated alongside Arabs in fighting either as individuals with their Arab patrons’ tribes or as independent battalions that fought side by side with Arab Muslims. In the last quarter of the first Islamic century (694–719 CE), the mawàlì formed the majority of fighters in the Umayyad army and in all amßàr.52 Many leaders, rulers, caliphs and tribal leaders as well as Umayyad family members had special mawàlì fighting forces.53 Sources also indicate that inequality in salaries given to Arabs and 51 Ibn Khallikàn, A˙mad b. Mu˙ammad, Wafayàt al-a'yàn wa-anbà" abnà" al-zamàn, ed. Mu˙ammad Mu˙yì l-Dìn 'Abd al-Óamìd, 6 vols (Cairo: Maktabat al-Nah∂a al-Mißriyya), 1948, VI, 354. 52 Juda, Mawàlì, Part 3 (al-Mawàlì wa-dìwàn al-jund ). Compare Elad’s contribution to this book on the composition of the army under the 'Abbàsid caliph alMa"mùn (r. 198–218/813–833). 53 Juda, Mawàlì, 120–32.

274

jamal juda

mawàlì existed.54 However, other sources state that the mawàlì received the same amount as Arabs (sharaf al-'a†à").55 Some sources specify that many mawàlì reached the honor of receiving 'a†à".56 Concerning the opposition voiced by some Kufan Arabs about the rebel al-Mukhtàr (d. 67/687) giving these 'a†à" salaries to mawàlì and slaves,57 it should be emphasized that this had been a special case for Kufa and that it cannot be generalized.58 It is also worth noting here that mawàlì fighters received booties like their fellow Arab Muslim fighters. These booties were more rewarding financially than normal salaries; one poet even described the mawàlì as booty-inviting wealth.59 Moreover, the participation of mawàlì in conquering new regions and lands meant that the newly conquered areas would bring prosperity to all, including the mawàlì. This applied to the eastern regions, Africa and Andalusia. The mawàlì’s involvement in the army and fighting, side by side, with Arabs, was financially and economically rewarding for them. They received their provisions and salaries. They had their share in booties and money from the lands conquered by them, along with their fellow Arab Muslims. It should also be pointed out that conquest campaigns were used to improve economic activity, particularly commercial enterprises and the increasing demand for luxury items. Inasmuch as traders, craftsmen and artisans were mostly mawàlì, they were the first to benefit from these economic gains that helped improve their economic status.

54 Balàdhurì, Ansàb (Nasab al-Zubayr), 141; Ibn Abì l-Óadìd, Shar˙, II, 200; Kulaynì, Kàfì, V, 69; Ißfahànì, Aghànì, VI, 146; Abù 'Ubayd, Amwàl, 335. 55 Balàdhurì, Futù˙, 458; ˇabarì, Ta"rìkh, IV, 186. 56 Balàdhurì, Futù˙, 265; Ya'qùbì, A˙mad b. Wà∂i˙, Ta"rìkh, ed. M.Th. Houtsma, 2 vols. (Leiden: Brill 1969), II, 176; Ibn 'Asàkir, Tahdhìb, VII, 203; al-Kindì, Mu˙ammad b. Yùsuf, Kitàb al-wulàt wa-l-qu∂àt, ed. Rheuvon Guest (Beirut, 1908), 51. 57 Balàdhurì, Ansàb, V, 300; ˇabarì, Ta"rìkh, II, 559; VI, 28, 43, 44; Jahshiyàrì, Wuzarà", 42; Ya'qùbì, Ta"rìkh, II, 362. 58 The rule of walà" al-'itàqa in Kufa and the failure of al-'itàqa mawàlì to participate in conquering Iraq on the one hand, and al-Mukhtàr’s dependence on many of them in his army on the other, had harmed the indigenous Arabs of the city, especially those who were receiving the 'a†à". This angered the Arabs and compelled them to oppose al-Mukhtàr’s measures. Another version has it that the mawàlì of caliph 'Umar b. 'Abd al-'Azìz had protested, arguing that some 50,000 mawàlì fighters had been fighting in Khurasan without receiving salaries. However, it should be stressed that many of these were volunteers who had sought economic gain, booties and ultimately better salaries. See the main text above that follows. 59 Ißfahànì, Aghànì, XI, 213.

the economic status of the

MAWÀLÌ

275

In closing, let me summarize the main findings of this contribution. We have seen above that the mawàlì, shortly after the conquests, were in charge of business activities, professions and crafts, markets, money changing and administrative jobs within the dìwàns. They were also members of the police force or served as personal bodyguards and in the army. They worked as clerks for their masters but they were owners of estates and villages, too. All this shows that the mawàlì and the non-Muslims were strong productive forces during the early caliphate and that they had benefited greatly from their positions. In other words, their economic situation was not bad, especially when compared with their Arab co-religionists since, at once, the very same sources state that bad economic situations existed for both Arabs and mawàlì alike.

Bibliography Primary Sources Abù Nu'aym, A˙mad b. 'Abdallàh al-Ißfahànì. Óilyat al-awliyà" wa-†abaqàt al-aßfiyà", 10 vols., Beirut: Dàr al-Kitàb al-'Arabì, 1970. Abù 'Ubayd, al-Qàsim b. Sallàm. Kitàb al-amwàl, ed. Mu˙ammad Khalìl al-Harràs, Cairo: Maktabat al-Kulliyya al-Azhariyya, 1968. Abù 'Ubayda, Ma'mar b. al-Muthannà. Naqà"i∂ Jarìr wa-l-Farazdaq, ed. Anthony Ashly Bevan, 2 vols., Leiden: E.J. Brill, 1905. Abù Yùsuf, Ya'qùb b. Ibràhìm. Al-Kharàj, Cairo: al-Ma†ba'a al-Salafiyya wa-maktabatuhà, 1392. al-Balàdhurì, A˙mad b. Ya˙yà b. Jàbir. Ansàb al-ashràf, vol. I, ed. Mu˙ammad Óamìdullàh, Cairo: Dàr al-Ma'àrif, 1959; vol. III, ed. 'Abd al-'Azìz al-Dùrì, Beirut: al-Ma†ba'a al-Kathùlìkiyya, 1978; vol. IV, ed. Schloessinger, Jerusalem: Hebrew University, 1971; vol. V, ed. Goitein, Jerusalem: Hebrew University, 1936. id., Nasab al-Zubayr b. 'Abd al-Mu††alib, ed. Mu˙ammad Bàqir al-Ma˙mùdì, Beirut: Dàr al-Ta'àwun lil-ma†bù'àt, 1977; ms. Istanbul Reisülkütap no. 598. id., Futù˙ al-buldàn, ed. De Goeje, Leiden: Brill, 1866. al-Fasawì, Ya'qùb b. Sufyàn. Kitàb al-ma'rifa wa-l-ta"rìkh, ed. Akram Îiyà" al-'Umarì, Baghdad: Maktabat al-Muthanna, 1974–76. Ibn 'Abd Rabbihi, A˙mad b. Mu˙ammad. al-'Iqd al-farìd, ed. A˙mad Amìn, A˙mad al-Zayn and Ibràhìm al-Ibyarì, Cairo: Ma†ba'at Lajnat al-Ta"lìf wa-l-Tarjama wa-l-Nashr, 1940–44. Ibn Abì l-Óadìd, 'Izz al-Dìn Abù Óamìd. Shar˙ nahj al-balàgha, ed. Mu˙ammad Abù l-Fa∂l Ibràhìm, 20 vols., Cairo: Dàr I˙ya" al-Kutub al-'Arabiyya, 1959. Ibn 'Asàkir, 'Alì b. al-Óasan b. Óibbatallàh. Tahdhìb al-ta"rìkh al-kabìr, ed. 'Abd alQàdir Badràn, Damascus: al-Maktaba al-'Arabiyya, 1330–32, 1349–51. Ibn Óabìb, Mu˙ammad. Kitàb al-mu˙abbar, Hyderabad: Ma†ba'at Dàr al-Ma'àrif al'Uthmàniyya, 1942. Ibn Óajar, A˙mad b. 'Alì. al-Ißàba fì tamyìz al-sa˙àba, 8 vols., Cairo: Ma†ba'at alSa'àda/Beirut: Dàr Íadr, 1325. id., Tahdhìb al-tahdhìb, Beirut: Dàr al-Fikr, 1984.

276

jamal juda

Ibn Óamìd, Mu˙ammad b. 'Abìd. Kitàb al-muntakhab min musnad 'Abìd b. Óamìd, ed. Íub˙ì al-Samarrà"ì and Mu˙ammad al-Sa'ìd, Cairo: Maktabat al-Risàla, 1988. Ibn Khaldùn, 'Abd al-Ra˙màn b. Mu˙ammad. al-Muqaddima, ed. M. Quatremere, 3 vols., Paris 1858. Ibn Khallikàn, A˙mad b. Mu˙ammad. Wafayàt al-a'yàn wa-anbà" abnà" al-zamàn, ed. Mu˙ammad Mu˙yì l-Dìn 'Abd al-Óamìd, 6 vols., Cairo: Maktabat al-Nah∂a alMißriyya, 1948. Ibn Qutayba, 'Abdallàh b. Muslim. Kitàb al-ma'àrif, ed. Tharwat 'Ukàsha, Cairo: Dàr al-Ma'àrif, 1960. id., Kitàb 'uyùn al-akhbàr, Beirut: Dàr al-Kutub al-Mißriyya, Cairo 1952. Ibn Sa'd, Mu˙ammad. Kitàb al-†abaqàt al-kubrà, ed. Eduard Sachau, 9 vols., Leiden: Brill, 1909. al-Ißfahànì, 'Alì b. al-Óusayn. Kitàb al-aghànì, 24 vols., Cairo: Dàr al-Kutub alMißriyya, 1927–74. al-Jà˙iΩ, 'Amr b. Ba˙r. al-Rasà"il, ed. Mu˙ammad 'Abd al-Salàm Hàrùn, 2 vols., Cairo: Maktabat al-Khànjì, 1964–65. al-Jahshiyàrì, Mu˙ammad b. 'Abdùs. Kitàb al-wuzarà" wa-l-kuttàb, ed. Mu߆afa alÍaqqà, Cairo: Ma†ba'at Mu߆afa al-Bàbì al-Óalabì, 1938. Khalìfa b. Khayyà†. Ta"rìkh, ed. Suhayl Zakkàr, 2 vols., Damascus: Wizàrat alThaqàfa wa l-Irshàd al-Qawmì, 1967. al-Kindì, Mu˙ammad b.Yùsuf. Kitàb al-wulàt wa-l-qu∂àt, ed. Rheuvon Guest, Beirut 1908. al-Kulaynì, Mu˙ammad b. Ya'qùb. Kitàb al-kàfì, ed. 'Alì Akbar al-Ja'farì, Teheran 1377–89. al-Mubarrad, Mu˙ammad b. Yazìd. Kitàb al-kàmil fì l-adab, ed. Mu˙ammad Abù l-Fa∂l Ibràhìm and Sayyid Shahàta, Cairo: Ma†ba'at Nah∂at Mißr, n.d. al-Nìsàbùrì, al-Fa∂l b. Shadhàn al-Azdì. Kitàb al-ì∂à˙, ed. Jalàl al-Dìn al-Óusaynì, Teheran 1972. al-Sukkarì, Abù Sa'id al-Óusayn. Dìwàn Abì l-Aswad al-Du"alì, Beirut 1974. id., Shi'r al-Akh†al, Aleppo 1970. al-ˇabarì, Mu˙ammad b. Jarìr. Ta"rìkh al-rusul wa-l-mulùk, ed. Mu˙ammad Abù l-Fa∂l Ibràhìm, 11 vols., Cairo: Dàr al-Ma'àrif, 1966–70. al-Wakì', Mu˙ammad b. Khalaf b. Óayyàn. Kitàb akhbàr al-qu∂àt, ed. 'Abd al-'Azìz Mu߆afa al-Maràghì, 3 vols., Cairo: Ma†ba'at al-Istiqàma, 1947. al-Ya'qùbì, A˙mad b. Wà∂i˙. Ta"rìkh, ed. M.Th. Houtsma, 2 vols., Leiden: Brill, 1969. Secondary Sources Arnold, Th. al-Da'wa al-islàmiyya, Cairo 1947 (translation from the English). Atwàn, H. al-Da'wa al-'abbàsiyya, Amman: Maktabat al-Mu˙taßib, 1984. Bay∂un, I. al-Dawla al-umawiyya wa-l-mu'àra∂a, 2 vols., Beirut: al-Mu"assasa al-Jàmi'iyya lil-Diràsat wa-l-Nashr, 1985. Cahen, Cl. Ta"rìkh al-'Arab wa-l-shu'ùb al-islàmiyya mundhu Ωuhùr al-islàm ˙attà bidàyat al-imbirà†ùriyya al-'uthmàniyya, trans. Badr al-Dìn al-Qàsim, Beirut: Dàr al-Óaqìqa lil-ˇibà'a wa-l-Nashr, 1977. Crone, P. The mawàlì in the Umayyad Period, London, Ph.D. Diss., 1973. al-Dùrì, 'A. Muqaddima fì ta"rìkh ßadr al-islàm, Beirut: al-Ma†ba'a al-kathùlìkiyya, 1961. id., Muqaddima fì ta"rìkh al-iqtißad al-'arabì, Beirut: Dàr al-Mashriq, 1969. Ende, W. Arabische Nation und Arabische Geschichte. Die Umayyaden in Urteil Arabischer Autoren den 20. Jahrhunderts, Beirut/Wiesbaden: Frank Steiner Verlag, 1977. Goldziher, I. Muhammadanische Studien, 2 vols., Halle: Max Niemeyer, 1889–90. Haarmann, U. Geschichte der Arabischen Welt, München: Beckverlag, 1987.

the economic status of the

MAWÀLÌ

277

Óigàb, M. Ma"àthir al-shu'ùbiyya fì l-adab al-'arabì, Cairo: Dàr I˙yà" al-Kutub al'Arabiyya, 1961. Juda, J. al-'Arab wa-l-ar∂ fì l-'Iràq fì ßadr al-islàm, Amman: al-Sharika al-'Arabiyya lilˇibà'a wa-l-Nashr, 1979. id., Die sozialen und wirtschaftlichen Aspekte der Mawàlì in frühislamischer Zeit, Tübingen, Ph.D. Diss., 1983. id., “al-Íul˙ wa-l-'anwa ladà 'ulamà" al-amßàr fì ßadr al-islàm”, in Majalat al-Najà˙ lil-Ab˙àth, 2/8 (1995), 7–38. al-Kha†ìb, A. al-Óukm al-umawì fì Khurasàn, Beirut: Mu"assasat al-A'lamì/Baghdad: Dàr al-Tarbiya, 1975. Kremer, A. von. Culturgeschichte des Orients unter den Chalifen, 2 vols., Wien: Wilhelm Braumüller, 1875–1877. Levy, R. The social structure of Islam, Cambridge: Cambridge University Press, 1962. al-Miqdàd, M. al-Mawàlì wa-niΩàm al-walà" min al-jàhiliyya ilà awàkhir al-'aßr al-umawì, Damascus: Dàr al-Fikr, 1988. Mu߆afa, Sh. Dawlat banì l-'Abbàs, Kuwait 1973. al-Najjàr, M. al-Mawàlì fì l-'aßr al-umawì, Cairo 1945. Nicholson, R. A literary history of the Arabs, Cambridge: Cambridge University Press, 1930. al-Qawàsmih, S. al-Tijàra wa-dawlat al-khilàfa fì ßadr al-islàm, Nablus, Mg. Thesis, 1999. 'Umar, F. ˇabì'at al-da'wa al-'abbàsiyya, Beirut: Dàr al-Qalam l-ˇibà'a 1970. Vaglieri, L. “The patriarchal and Umayyad Caliphates,” in P.M. Holt, A.K.S. Lambton and B. Lewis (eds.). Cambridge History of Islam (Cambridge: Cambridge University Press, 1970. Vloten, G. van. Recherches sur la domination arabe, le chiitisme, et les croyances messianiques sous le Khalifat des Omayades, Amsterdam: J. Muller, 1894. Watt, W.M. Islam and the integration of society, London: Routledge & Kegan Paul, 1961. Zaydàn, G. Ta"rìkh al-tamaddùn al-islàmì, 2 vols., Beirut: Manshùràt Dàr Maktabat al-Óayàt, 1902, 1906.

MAWÀLÌ IN THE COMPOSITION OF AL-MA"MÙN’S ARMY: A NON-ARAB TAKEOVER?1 Amikam Elad

Introduction In this article I wish to present the beginnings of a study on the ethnic character of al-Ma"mùn’s army. In this way I wish to examine, from a different perspective and a later period, an issue I have discussed in previous articles, that is the social-cultural-ethnic character of the early 'Abbàsid caliphate.2 In those articles I argued against the accepted view in research, that from its foundation, the 'Abbàsid caliphate is distinguished by the influence of non-Arab elements. This influence increases rapidly and is obvious already at the time of the caliph al-Mahdì (reigned 158–69/775–85).3 The height of Iranian-Khuràsànì dominance over the Caliphate, according to this view, was the rule of al-Ma"mùn (r. 198–218/813–33). The struggle between al-Amìn and al-Ma"mùn (193–98/809–13) has been viewed by scholars, from the nineteenth century almost to the present day, as a struggle between the Arabs and the non-Arabs, mainly Iranian-Khuràsànìs (see the discussion further below). My main conclusion was that “it is thus necessary to modify greatly the generally accepted view concerning the considerable Iranian 1

Parts of this article are based on my unpublished Ph.D. dissertation “Characteristics of the Development of the 'Abbàsid Army (especially ahl Khuràsàn and al-abnà" Units), With an Emphasis on the Reign of al-Amìn and al-Ma"mùn” (Ph.D. diss., The Hebrew University of Jerusalem, 1987, in Hebrew); I am grateful to Professor Patricia Crone for her detailed criticism and comments on the article, which I found very helpful and stimulating and have used partially in this work. 2 A. Arazi and A. Elad, “L’Épître à l’armée, al-Ma"mùn et la second da'wa,” Studia Islamica 66 (1987), (1st part) 27–70; 67 (1988), (2d pt.) 29–73; A. Elad, “Aspects of Transition from Umayyad to the 'Abbàsid Caliphate,” Jerusalem Studies in Arabic and Islam 19 (1995), 89–132. 3 D. Ayalon, “The Military Reforms of Caliph al-Mu'taßim: Their Background and Consequences,” in idem, Collected studies: Islam and the Abode of War: Military Slaves and Islamic Adversaries (London: Variorum, 1994), I, 2–4, 35–36 and the important information in the addenda; P. Crone, Slaves on Horses: The Evolution of the Islamic Polity (Cambridge: Cambridge University Press, 1980), 68, esp. 74; H. Kennedy, The Early 'Abbàsid Caliphate (London, 1981), 102–103; Elad, “Transition,” 118–19.

MAWÀLÌ

in the composition of al-ma"mùn’s army

279

influence (or even dominance) in various lines of government from the inception of the 'Abbàsid state. It is clear that a large proportion of the answers and solutions given so far to this problem are inadequate, even wrong in many cases and that the direction of research must thus be changed.”4

A summary of my view on the position of the mawàlì in the early 'Abbàsid caliphate It seems that the mawàlì only gradually penetrated and came to influence the caliphate, and at first were certainly limited to less important spheres, mainly in personal service posts in court and in administrative positions. It seems that most of these mawàlì were mawàlì muwàlàt, that is confederates, or allies who had entered into voluntary agreement with their patrons by a contractual agreement, although a certain percentage may have been manumitted slaves.5 Although they were influential, the power base of the regime was the army. The main limitation of these mawàlì was their lack of significant support in the military.6 Evidence of mawàlì units in the early 'Abbàsid army is extremely scarce. If they are mentioned at all, they are small units; none is mentioned among the most important and well-known commanders of the imperial armies of ahl Khuràsàn.7 It is well-known that some of the senior commanders of ahl Khuràsàn and al-abnà" were mawàlì. The decisive majority of them were mawàlì of Arab tribes. “They were totally committed to Arab culture and its values which were deeply rooted within them. They were raised and acted according to the values of the highest levels of Arab society, belonging to the military-administrative elite of the 'Abbàsid

4

A. Elad, “Transition,” 127. It seems impossible at this stage of the research to successfully define the type of walà" mentioned in the sources and to draw detailed and unequivocal conclusions on this topic. 6 J. Lassner, The Shaping of 'Abbàsid rule (New Jersey: Princeton University Press, 1980), 13–16, 89–102; F. Amabe, The Emergence of the 'Abbàsid Autocracy. The 'Abbàsid Army, Khuràsàn and Adharbayjàn (Kyoto: Kyoto University Press, 1994), 81: “Generally speaking, they [that is the mawàlì ] were much less important in military affairs than the Khuràsànis, Baßrans or Syro-Jazìris;” see also the discussion, ibid., 82–84. 7 A. Elad, “Transition,” 119 note 146. 5

280

amikam elad

state. They were loyal to the Arabic ('Abbàsid) dynasty and to the Arabic heritage for which it stood.”8 The negative attitude against the mawàlì did not change during the early 'Abbàsid period. The principles that guided the Umayyads in their relations with the mawàlì, at least for rulers and the upperclass of society, certainly still existed in the period of Hàrùn alRashìd and even later.9 Ample evidence testifies to the central place of Arabic culture, and interest taken in it, throughout the early 'Abbàsid period.10

The period of al-Ma"mùn The discussions in my previous articles do not exhaust the problem. As noted above, in this article I propose to continue following this issue into al-Ma"mùn’s reign. This is because his reign is considered by all scholars to be the peak of the non-Arabs (mainly Iranians) penetration of the caliphate, especially in the army. In another place I have dealt with certain cultural and social aspects of the reign of this caliph. I showed that Arabic poetry was honored, appreciated and admired by the caliph and the dignitaries of the caliphate, but also by the lower classes. Evidence for this caliph’s love of the Arabic language and Arab heritage from the Jàhiliyya is abundant. As in the reign of Hàrùn al-Rashìd, we find a number of excellent poets in this period who composed lengthy poems glorifying and praising the large tribal confederations, Qays and Yaman.11 It is highly unlikely that these poems fell on deaf ears; in other words, the ground was well prepared for them. They responded to the innermost thoughts, and were an echo, of the social reality that absorbed, understood and related to this poetry.

8

Ibid., 119–20. Ibid., 120 note 151; id., 'Abbàsid Army, 80, 86. 10 Ibid., 121; one of the accepted practices commonly used by the mawàlì to integrate into Arab society was the creation of a forged Arab lineage (iddi'à"). This method existed throughout the whole of the Umayyad period and continued to be used well into the 'Abbàsid Caliphate. Even at the time of al-Ma"mùn such a phenomenon was not rare among the poets, singers and senior government secretaries, who mixed in elite social circles, see ibid., 126–27 notes 170–75. 11 A. Elad, “Transition,” 122. 9

MAWÀLÌ

in the composition of al-ma"mùn’s army

281

Al-Ma"mùn certainly was part of this reality, and paid attention to it, to the tribal conflicts and particularly to the poetry that accompanied them. He did not hesitate to intervene, actively and personally, when a qaßìda was aimed against the Nizàr (Mu∂ar).12 This also testifies to the great importance of nasab during this period. AlMa"mùn was visited by many scholars, who at his initiative and encouragement, developed the science of genealogy, as well as the sciences of the Arabic language, rhetoric and ˙adìth.13

Outlines of the development of al-Ma"mùn’s army The first stages before armed conflict Immediately upon his arrival in Marw in 193/809, al-Ma"mùn began to organize and consolidate his army. He wished to gain the support of ahl Khuràsàn for his faction. They joined his forces gradually.14 After the first defeat of al-Amìn’s army before ˇàhir b. al-Óusayn allegiance was sworn to al-Ma"mùn in all regions of Khuràsàn and its immediate vicinity. The traditions emphasize that the entire ahl Khuràsàn joined his faction.15 In another place I have demonstrated that the expression ahl Khuràsàn in these texts, and in the vast majority of other contemporary ones, denotes fighting unites. Moreover, the texts indicate that in Khuràsàn after 190/805 there is identity between the Arabs and ahl Khuràsàn. The expression does not refer to non-Arab residents, local nobility or a separate local Eastern entity.16 The use of the expression ahl Khuràsàn during al-Ma"mùn’s rule in Khuràsàn includes the veteran abna" units, which had come from Iraq with Hàrùn al-Rashìd, and also Arab tribes from Khuràsàn itself.17 Some of them took part in the 'Abbàsid Revolution and in 12

For examples, see ibid., 122–26. Ibid., 125. 14 Ibn A'tham, Kitàb al-futù˙, ed. Mu˙ammad 'Abd al-Mu'ìd Khàn et al. (Hyderabad: Dà"irat al-Ma'àrif al-'Uthmàniyya, 1388–95/1968–75), VIII, 296. 15 Al-Ya'qùbì, Ta"rìkh (Beirut: Dàr Íàdir, 1379/1960), II, 438 ( fa-a'†à jamì' ahl Khuràsàn al-†à'a li-l-Ma"mùn). 16 A. Arazi and A. Elad, “L’Épître à l’armée,” (1st pt.) 58–9; (2nd pt.) 29–32; A. Elad, “The Ethnic Composition of the 'Abbàsid Revolution: A Reevaluation of Some Recent Research,” Jerusalem Studies in Arabic and Islam 24 (2000), esp. 249–75, 283–89, 292–99, 300–310. 17 On the identification of al-abna" with ahl Khuràsàn, and the changing usage of 13

282

amikam elad

the battles against the Umayyads, and the 'Abbàsid nuqabà" and du'àt came from their ranks. Al-Ma"mùn turned to them on his arrival in Khuràsàn. The units of ahl Khuràsàn of al-Ma"mùn’s army were composed of them together with non-Arabs, just as the army of the first revolution had included units—admittedly, only few—of non-Arabs.18 The organization of the army In the month of Sha'bàn 192/June 808, Hàrùn al-Rashìd left Baghdàd for Khuràsàn at the head of a great army, largely commanded by senior officers of the abnà", to suppress the rebellion of Ràfi' b. althis term in relation to those military units, see Arazi and Elad, “L’Épître à l’armée,” (1st pt.) 55–57. 18 Few of the 'Abbàsid propagandists were non-Arabs, see al-ˇabarì, Ta"rìkh alrusul wa-l-mulùk, ed. M.J. de Goeje (Leiden: E.J. Brill, 1879–1901), II, 1993 (mentioned by D.C. Dennett, Marwàn b. Muhammad: The Passing of the Umayyad Caliphate (Ph.D. diss., Harvard University, 1939), 302); al-ˇabarì, Ta"rìkh, II, 1994: 'idda min a'jam al-shì'a (both references from al-ˇabarì are mentioned by Ayalon, “al-Mu'taßim,” 5); Akhbàr al-dawla al-'abbàsiyya wa-fìhi akhbàr al-'Abbàs wa-wuldihi, ed. 'Abd al-'Azìz alDùrì and 'Abd al-Jabbàr al-Mu††alibì (Beirut: Dàr Íàdir, 1971), 285, 325; Lassner, Shaping, 214: the al-Furs and al-Khwàrizmiyya units in al-Manßùr’s army; Crone, Slaves, 572 (quoting J. Lassner, The Topography of Baghdàd in the Early Middle Ages, Text and Studies (Detroit: Wayne State University Press, 1970), 68); Crone, Slaves, 16, mentioning “a proto-mamlùk army in al-Manßùr’s period” according to the chronicle of Dionysius of Tell Ma˙re (ed. Chabot, 84); also we find in the army of the da'wa some noble Iranian commanders; they are called awlàd mulùk Khuràsàn. Three are mentioned by name: 1) Íàli˙ Íà˙ib al-Mußallà, and Shabìb b. Wàj (al-Kha†ìb al-Baghdàdì, Ta"rìkh Baghdàd (Cairo, 1349/1931), XI, 130; Elad, 'Abbàsid Army, 130; and Turàrkhudà. P. Crone, “The 'Abbàsid Abna" and the Sàsànid Cavalrymen,” Journal of the Royal Asiatic Society 8 (1998), 12, 16, 17: Shabìb and Turàrkhudà; p. 17: Íàli˙ Íà˙ib al-Mußallà; Shabìb b. Wàj was one of the commanders of ahl Khuràsàn (al-ˇabarì, Ta"rìkh, II, 1960) and later on we find him in the ˙aras of alManßùr (ibid., III, 111); in 162/778–779, during al-Mahdì’s reign, he headed an expedition against the Khawàrij (ibid., 492); a quarter in Baghdàd was named after him (al-Kha†ìb al-Baghdàdì, Ta"rìkh Baghdàd, I, 84; Turàrkhudà is also mentioned as one of the important commanders of ahl Khuràsàn (al-ˇabarì, Ta"rìkh, III, 93 (Lassner, Shaping, 258 note 29, according to al-ˇabarì); he is mentioned as belonging to the sons of the kings of the non-Arabs of Khuràsàn (wa-kàna min abnà" mulùk a'àjim Khuràsàn, al-ˇabarì, Ta"rìkh, III, 356, quoted by Crone, “Abna",” 16–17). For the opposite view of the “New Revisionists” of the 'Abbàsid Revolution, that is, that the ahl Khuràsàn units were mainly composed of non-Arabs, and the term ahl Khuràsàn, refers mainly to the non-Arab, mainly Iranian population, see Crone, “Abnà",” 11–12, 13, where she does not mention or relate to the specific term, but her conclusions on the subject are clear cut; but see id., “The Significance of Wooden Weapons in al-Mukhtàr’s Revolt and the 'Abbàsid Revolution,” in I.R. Netton, ed., Studies in Honour of Clifford Edmund Bosworth. Vol. I: Hunter of the East: Arabic and Semitic Studies (Leiden: E.J. Brill, 2000), 184–85, where she specifically refers to ahl Khuràsàn; Saleh Said Agha, The Agents and Forces that Toppled the Umayyad Caliphate (Ph.D. diss., University of Toronto, 1993), 364–66, 368–84, 403–405, 415,

MAWÀLÌ

in the composition of al-ma"mùn’s army

283

Layth and to pacify the important province.19 He sent his son alMa"mùn from his camp twenty-three days before his death, with some of the senior commanders to Marw, the capital of Khuràsàn.20 Hàrùn al-Rashìd died on his way to Marw, at ˇùs, at the beginning of 193/March 809.21 According to Hàrùn al-Rashìd’s will, which was drawn up in 186/802 in Mecca and read out to the 'Abbàsid family and the commanders, who even signed it, on al-Rashìd’s death al-Amìn would become caliph and al-Ma"mùn would become governor of Khuràsàn. A paragraph in the will states that the army and commanders who had been with al-Rashìd must pass to al-Ma"mùn’s disposal. A large part of the army and the commanders did indeed remain with him, but part returned to Iraq.22 From now on the relations between the two brothers would deteriorate to a bitter struggle, war, a siege of Baghdàd its conquest and the death of al-Amìn.23 The abnà" units in al-Ma"mùn’s army in Khuràsàn The abna" units with which al-Ma"mùn arrived in Marw are referred to in the sources also as the army of ahl Khuràsàn.24 Of the twelve 419, 499; id., “The Arab Population in Khuràsàn during the Umayyad Period: Some demographic Computations,” Arabica 46 (1999) 211–13; E. Daniel, The Political and Social History of Khuràsàn under 'Abbàsid Rule (Minneapolis/Chicago, 1979), 26, 36, 50–52, 58, 60–61, 86, 189; id., “Arabs, Persians and the Advent of the 'Abbàsids Reconsidered [A Review article of the book of R.M. Guzman, M. Sharon and M. Zakeri],” Journal of the American Oriental Society 117 (1997), 546, 547–548; id., “The Ahl al-Taqàdum and the Problem of the Constituency of the 'Abbàsid Revolution in the Merv Oasis,” Journal of Islamic Studies 7 (1996), 152 note 4; 156ff.; 165 note 43; 179; but cf. Elad, “The Ethnic Composition.” Zakeri’s opinion (pp. 265ff.), that the revolutionary troops mainly consisted of Sàsànid horsemen, belonging to the lower Sàsànid nobility, and hence, their descendants, the abnà", were the sons of these Sàsànid gentry, has been subjected to severe criticism, see for instance, Frye (see Bibliography); Crone, “Abnà",” 1–3, 5, 16. 19 Al-ˇabarì, Ta"rìkh, III, 730–32; F. Gabrieli, “La successione di Hàrùn al-Rashìd e la guerre fra al-Amìn,” Rivista degli Studi Orientali 11 (1926–28), 350–351; Kennedy, The Early 'Abbàsid Caliphate, 123–33. 20 Al-ˇabarì, Ta"rìkh, III, 733–34; al-Ya'qùbì, Ta"rìkh (ed. Beirut), II, 430. 21 Khalìfa b. Khayyà† al-'Ußfurì. Ta"rìkh Khalìfa b. Khayyà†, ed. S. Zakkàr (Damascus, 1968), II, 740; al-Ya'qùbì, Ta"rìkh, II, 430. 22 Gabrieli, “Hàrùn ar-Rashìd,” 346–47 note 2, argues that this close in the will was interpolated later; see also, Kennedy, The Early ‘Abbàsid Caliphate, 124–25; M. Shaban, Islamic History. A New Interpretation. A.D. 600–750 (A.H. 132) (Cambridge: Cambridge University Press, 1971), II, 39–40. 23 The description of the course of events, see Gabrieli, “Hàrùn ar-Rashìd,” 354–95; Kennedy, The Early 'Abbàsid Caliphate, 135–48; Shaban, History, 42–46. 24 Al-ˇabarì, Ta"rìkh, III, 771: al-jund, identified by al-Dìnawarì, Al-Akhbàr al-ˇiwàl, ed. V. Guirgass (Leiden: E.J. Brill, 1888), 392–93 as ahl Khuràsàn: yà ahl Khuràsàn, jaddidù l-bay'a li-imàmikum al-Amìn.

284

amikam elad

commanders of al-Rashìd’s army mentioned in his campaign to ˇùs, ten are well-known,25 of whom seven or eight belonged to al-abna".26 The other two commanders are well-known Arabs, whose families served the Umayyads and proceeded to serve the 'Abbàsids.27 Of the eight commanders mentioned, who according to our definition and concepts assuredly belong to the abnà", seven moved with alMa"mùn to Marw and remained with him there: 1) 2) 3) 4) 5) 6) 7)

'Abdallàh b. Màlik b. al-Haytham. Al-'Abbàs b. al-Musayyab b. al-Zuhayr. Nu'aym b. Khàzim b. Khuzayma. Shabìb b. Óumayd b. Qa˙†aba. Ya˙yà b. Mu'àdh b. Muslim. Al-'Abbàs b. Ja'far b. Mu˙ammad b. al-Ash'ath. Óàtim b. Harthama.28

1) 'Abdallàh b. Màlik [b. al-Haytham al-Khuzà'ì] (al-ˇabarì, Ta"rìkh, III, 734, 769, 772); 2) Ya˙yà b. Mu'àdh [b. Muslim] (ibid., 734, 772); 3) Nu'aym b. Khàzim [b. Khuzayma] (ibid., 734); 4) Shabìb b. Óumayd b. Qa˙taba (ibid., 772); 5) al'Abbàs b. al-Musayyab b. Zuhayr (ibid.); 6) al-'Abbàs b. Ja'far b. Mu˙ammad b. al-Ash'ath (ibid., 734); 7) 'Abd al-Ra˙màn b. Jabala (ibid., 773); 8) Óàtim b. Harthama [b. A'yan] (ibid., 769); 9) al-Sindì b. al-Óarashì (ibid., 734); 10) Asad b. Yazìd b. Mazyad (ibid., 770; 11) al-'Alà", Mawlà Hàrùn al-Rashìd (ibid., 772); 12) Ayyùb b. Sumayr [in charge of al-kitàba] (ibid., 734, 772; al-Jahshiyàrì, Kitàb al-wuzarà" wa-l-kuttàb (Cairo: Mu߆afà l-Bàbì l-Óalabì, 1357/1938), 266; on these commanders, see Crone, Slaves, Index; on Nu'aym b. Khàzim, see al-Jahshiyàrì, Kitàb al-wuzarà", 312; al-ˇabarì, Ta"rìkh, III, 734 [Ibn al-Athìr, Al-Kàmil fì l-ta"rìkh, ed. Carl J. Tornberg (Leiden: E.J. Brill, 1851–76), VI, 212: Óàzim, but see ibid., note 3, ms. R: Khàzim; al-ˇabarì, Ta"rìkh, III, 841: Óàzim, but on p. 1022: Khàzim; see also [Pseudo] al-Jà˙iΩ, Kitàb al-†àj fì Akhlàq al-Mulùk, ed. Fawzì 'A†wà (Beirut, 1970), 59: Khàzim.] 26 Nos. 1, 2, 3, 4, 5, 6, 7, 8 [?]. 27 1) Asad b. Yazìd b. Mazyad (on him and his family, see Crone, Slaves, 169–70); 2) al-Sindì b. Ya˙yà al-Óarashì (on him, see ibid., 145); both are members of wellknown families of commanders, who were in the service of the Umayyads. 28 See al-ˇabarì, Ta"rìkh, III, 734, 769, 772, for evidence for their traveling with al-Ma"mùn to Marw. Evidence for their stay in Marw at al-Ma"mùn’s side: 1) 'Abdallàh b. Màlik b. al-Haytham: al-Jahshiyàrì, Kitàb al-wuzarà", 309; al-Qummì, 'Uyùn Akhbàr al-Ri∂à (Najaf, 1390/1970), II, 151. 2) Al-'Abbàs b. al-Musayyab b. al-Zuhayr: al-ˇabarì, Ta"rìkh, III, 772 [Crone, Slaves, 177–78]. 3) Nu'aym b. Khàzim b. Khuzayma: al-Jahshiyàrì, Kitàb al-wuzarà", 310, 313; al-ˇabarì, Ta"rìkh, III, 841, ll. 15–16. 4) Shabìb b. Óumayd b. Qa˙†aba: see Crone, Slaves, 189, who believes that he went back to Baghdàd, with no evidence; see al-Ya'qùbì, Ta"rìkh (Leiden: E.J. Brill, 1882), II, 574: Shabìb is in charge of the ˙aras of al-Ma"mùn, dismissed and appointed as the governor of al-Qùmìs [Crone, Slaves]; see also, al-Mas'ùdì, Kitàb al-tanbìh wa-l-ishràf, Bibliotheca Geographorum Arabicorum, vol. 8 (Leiden: E.J. Brill, 1894), 352: the ˙àjib of al-Ma"mùn. Shabìb died in 204/820 (Ibn Abì ˇàhir ˇayfùr, [al-Juz" al-sàdis min] Kitàb Baghdàd, ed. H. Keller (Leipzig, 1908), 349). 5) Ya˙yà b. 25

MAWÀLÌ

in the composition of al-ma"mùn’s army

285

To this list of commanders belonging to al-abnà" we may add five additional banawì commanders, in Khuràsàn: 8) Harthama b. A'yan, who was nominated to be ßà˙ib al-˙aras of al-Ma"mùn.29 9) Al-'Abbàs b. 'Abdallàh b. Màlik b. al-Haytham, who was dismissed by al-Ma"mùn from the governorship of al-Rayy.30 10) Mu˙ammad b. al-'Abbàs b. Musayyab b. Zuhayr, who stayed with his father in Marw.31 11) A son of al-'Abbàs b. Ja'far b. Mu˙ammad b. al-Ash'ath, who replaced his father as ßà˙ib al-shur†a of 'Alì al-Ri∂à.32 He may

Mu'àdh b. Muslim: al-ˇabarì, Ta"rìkh, III, 1026, 1027; al-Jahshiyàri, Kitàb al-wuzarà", 316 [Crone, Slaves, 184]; see also, Ibn Óabìb, “Asmà" al-mughtàlìn min al-ashràf fì l-jàhiliyya wa-l-Islàm,” in 'Abd al-Salàm Mu˙ammad Hàrùn, ed., Nawàdir al-makh†ù†àt (Cairo, 1393/1773), VI, 201. 6) Al-'Abbàs b. Ja'far b. Mu˙ammad b. al-Ash'ath: Ibn Óabìb, Asmà" al-mughtàlìn, 7) Óàtim b. Harthama: al-Ya'qùbì, Ta"rìkh, II, 557, but cf. Crone, Slaves, 177. 29 Al-ˇabarì, Ta"rìkh, III, 777; Crone, Slaves, 177; it seems highly probable that he belonged to a (distinguished) family of the al-abnà", that remained in Khuràsàn after the 'Abbàsid revolution; discussion and arguments in this vein, see A. Elad, “The Southern Golan in the Early Muslim Period. The Significance of Two Newly Discovered Milestones of 'Abd al-Malik,” Der Islam 76 (1999), 57–58 notes 114–15. It is possible that another commander, al-Walìd b. al-Musayyab, who may be a relative of al-'Abbàs b. al-Musayyab b. Zuhayr and the well-known commander alZuhayr b. al-Musayyab (see further on in this note), stayed with al-Ma"mùn in Marw as well. (If my assumption is correct, al-Walìd should be added to the biographies of this family in Crone, Slaves, 186f.): see Ibn A'tham, Futù˙, VIII, 274: Hàrùn al-Rashìd reconfirms the conditions of his will in the year 186/Dec. 802; among the sections of the will was the adjoining of “well-known commanders like Harthama b. A'yan, al-Walìd b. al-Musayyab, 'Abdallàh b. Màlik, Ya˙yà b. Mu'àdh, and others to al-Ma"mùn”; see also al-ˇabarì, Ta"rìkh, III, 666–67: the reconfirmation of the conditions of the will, stressing the adjoining of commanders to al-Ma"mùn, but no names of the commanders are given; see the long and exhaustive discussion of Gabrieli, “Hàrùn ar-Rashìd,” 346–47 note 2 (according to al-ˇabarì; Ibn A'tham’s book was published much later); as to al-Zuhayr b. al-Musayyab [b. al-Zuhayr]: a) he was nominated to the governorship of Sìstàn [by Hàrùn al-Rashìd? or maybe by al-Ma"mùn? Hàrùn al-Rashìd died, in February or March, 809, see al-ˇabarì, Ta"rìkh, III, 739; al-Ya'qùbì, Ta"rìkh (ed. Leiden), II, 521 (February); Ibn Khayyà†, Ta"rìkh, II, 740; al-ˇabarì, Ta"rìkh. (March)]; al-Zuhayr b. al-Musayyab arrived at Sìstàn in Shawwàl 193/July–Aug. 809 [that is, a few months after Hàrùn al-Rashìd died] and stayed there until Dhù l-Qa'da 194/August 810, see C.E. Bosworth, Sìstàn under the Arabs from the Islamic Conquest to the Rise of the Íaffàrids (30–250/651–864) (Rome, 1968), 100; Crone, Slaves, 187 (according to the Persian Ta"rìkh Sìstàn). b) In 196/811–812, he is mentioned among the senior commanders of al-Ma"mùn, besieging Baghdàd (Crone, Slaves.). 30 Crone, Slaves, 181–82; nothing more is known about him. 31 Ibn Abì ˇàhir ˇayfùr, Kitàb Baghdàd, 11. 32 Ibn Óabìb, Asmà" al-Mughtàlìn, 201.

286

amikam elad

have been 'Uqba b. al-'Abbàs b. Ja'far, who is mentioned as one of the commanders of Hàrùn al-Rashìd.33 12) A˙mad b. Ya˙yà b. Mu'àdh.34 These commanders, who had joined Hàrùn al-Rashìd’s army with their regiments, remained with al-Ma"mùn in Khuràsàn. It can be safely assumed that they joined al-Ma"mùn with their regiments.35 It would seem that their forces were not small, as these are most of the commanders of al-Rashìd’s army. These commanders filled highly important positions in Marw: they were in charge of al-Ma"mùn’s ˙aras and shur†a, the standard-bearers of his wazìr, al-Fa∂l b. Sahl, in charge of the shur†a and the ˙ijàba of 'Alì al-Ri∂à, al-Ma"mùn’s heir, and also served as governors of Khuràsàn and in other districts.36 It will be recalled that in Baghdàd the abnà" held these key posts of the caliphate, and here in Khuràsàn as well, under al-Ma"mùn’s government, they retain their senior position. It is clear that these senior commanders who joined al-Ma"mùn fulfilled the condition that

33 Al-Azdì, Ta"rìkh al-Mawßil, ed. 'A. Óabìba (Cairo: Dàr al-Ta˙rìr li-l-ˇab' wal-Nashr, 1967), 308 (year 190/805–806): at the head of al-sàqa in al-Rashìd’s army at the Byzantine front; see also al-ˇabarì, Ta"rìkh, III, 709; Crone, Slaves, 185, mentions that he was in opposition to al-Ma"mùn (according to al-Kha†ìb al-Baghdàdì, Ta"rìkh, I, 81), but this source does not tell when and under what circumstances 'Uqba rose against al-Ma"mùn. 34 Assuming that he accompanied his father and stayed with him in Khuràsàn, and together they probably moved to Iraq. The first time he is mentioned in Iraq is in 210 [!] (al-ˇabarì, Ta"rìkh, III, 1075: Rabì' II, 210/July–August 825; on this family, see Crone, Slaves, 184). 35 Al-ˇabarì, Ta"rìkh, III, 769–70; see also al-Jahshiyàrì, Kitàb al-wuzarà", 276, a partial text, mentioning al-Fa∂l b. al-Rabì'’s army; the addressee of the letter of al-Amìn is al-Ma"mùn and not their brother Íàli˙. 36 1) Harthama b. A'yan, ßà˙ib al-Óaras of al-Ma"mùn until 196/812, see Crone, Slaves, 177; 2) Nu'aym b. Khàzim, carried the flag ('alam) of al-Fa∂l b. Sahl, see al-Jahshiyàrì, Kitàb al-wuzarà", 306; al-ˇabarì, Ta"rìkh, III, 841 (year 196); Ibn alAthìr, Al-Kàmil fì l-ta"rìkh (Beirut: Dàr Íàdir-Dàr Beirut, 1385/1965), VI, 257: qalam, instead of 'alam; Óàzim instead of Khàzim; 3) al-'Abbàs b. al-Musayyab b. Zuhayr was ßà˙ib al-shur†a of al-Ma"mùn until the latter entrance to Baghdàd (in 204/819); 4) al-'Abbàs b. Ja'far b. Mu˙ammad b. al-Ash'ath, is in charge of the shur†a of 'Alì al-Ri∂à (in the year 201/816–817), see Ibn Óabìb, Asmà" al-Mughtàlìn, 201; 5) his son ('Uqba?), was the substitute of his father for this position, see ibid.; 6) Ya˙yà b. Mu'àdh b. Muslim was appointed to the Óijàba of 'Alì al-Ri∂à (in the year 201), see ibid.; he was appointed as the governor of al-Jazìra and Armenia in 204/819–820, see Crone, Slaves, 184; 7) al-'Abbàs b. 'Abdallàh b. Màlik b. al-Haytham was for a short time, the governor of al-Rayy, see Crone, Slaves, 181–82; 8) al-Zuhayr b. al-Musayyab b. al-Zuhayr, was the governor of Sìstàn, see note 29 above.

MAWÀLÌ

in the composition of al-ma"mùn’s army

287

the supreme and most distinguished body of the second da'wa was composed of men who had participated in both da'was. It also seems quite likely that they comprised an important part of those called by the sources “the Shì'a of al-Ma"mùn.”37 (On this see the discussion below.) In the first years of al-Ma"mùn’s rule in Khuràsàn there is no evidence for tension or hostility between the commanders of the abnà" and al-Ma"mùn. On the contrary—the senior posts at court, the governorships they held, indicate a preferred status. The efforts at rapprochement with the Arab tribes in Khuràsàn in the attempt to establish a new da'wa certainly strengthened their position as well, as al-Ma"mùn could pride himself before the tribes on the presence of abnà" al-shì'a at his court and on their support of him (see further below). However, an important point should be noted: the fact that none of the commanders of the 'Iràqì abnà" was sent to fight against al-Amìn’s abnà" armies. Even when this was suggested to them they refused.38 It can be assumed that they recoiled at fighting against members of their families and comrades in arms, who had remained in Baghdàd. The sources give no evidence of pressure from al-Ma"mùn on this matter. Al-Ma"mùn surely needed them in Marw itself, to help establish and consolidate his governorship, and certainly to organize the second da'wa. It should be remembered that these experienced commanders and their fathers had served the different caliphs throughout 'Abbàsid rule, holding key positions, and no doubt had much to contribute to al-Ma"mùn. And indeed, we have seen that in Marw they were appointed to important posts. Commanders from the western section of the abnà" were sent to Iraq only after a third [!] army sent from Baghdàd had been defeated by ˇàhir b. al-Óusayn, thus effectively sealing al-Amìn’s fate. Only in 196/812, some three years after arriving in Marw, did al-Ma"mùn send Harthama b. A'yan and Zuhayr b. al-Musayyab to help ˇàhir in the conquest of Iraq.

37 Al-Jahshiyàrì, Kitàb al-wuzarà", 289–90: year 183/808–809; compare Ibn A'tham, Futù˙, VIII, 295–96: wa-shàwara [al-Ma"mùn] al-Fa∂l b. Sahl wa-akhàhu wa-jamì' shì'atihi; al-ˇabarì, Ta"rìkh, III, 778–79 (year 194): wa-hàdhà bayn akhwàlihi wa-shì'atihi; ibid., 780: al-amr mukh†ir wa-laka min shì'atika wa-ahl baytika bi†àna. 38 Al-ˇabarì, Ta"rìkh, III, 825.

288

amikam elad

The mobilization of additional armies and commanders In addition to the army of the abnà"/ahl Khuràsàn mentioned above, al-Ma"mùn began to mobilize other forces. In the first stage, he gathered important commanders and armies stationed in the areas of Khuràsàn remote from Marw.39 This army is not described as a new army. Possibly these were garrisons stations in different parts of Khuràsàn. They may have been veteran forces of ahl Khuràsàn or abnà" sent from Iraq, or local ones, Arabs or perhaps non-Arabs. Together with these commanders, who al-Ma"mùn summons from the four corners of Khuràsàn, their sons and families also arrive at al-Ma"mùn’s court, at his express command.40 One of these senior commanders was Óusayn, ˇàhir’s father, who came to al-Ma"mùn with his son and other commanders who were family members.41 ˇàhir was sent with an army to al-Rayy,42 with which he seems to have fought against the army of 'Alì b. 'Ìsà, al-Amìn’s commander of the abnà" army, in 195/811. Some of his units were non-Arab, of different ethnic backgrounds: Turks, Khwàrizmìs and Bukhàrìs. We do not know what proportion of ˇàhir’s army they comprised. This army is usually described as a small one, between four thousand and five thousand soldiers.43 We do not know when this army was mobilized, whether all or part of ˇàhir’s army was the same as he had fought with in previous years in battles against the Khawàrij at Bùshanj in 180/796–97,44 or against the army of Ràfi' b. al-Layth

39

Al-ˇabarì, Ta"rìkh, III, 816. Al-Jahshiyàrì, Kitàb al-wuzarà", 291. 41 Ibid.; al-ˇabarì, Ta"rìkh, III, 771: Óusayn and his nephew al-Khalìl b. Hishàm; Ibn al-Athìr, Al-Kàmil (ed. Tornberg), VI, 158: Hishàm, the father of al-Khalìl b. Hishàm, one of the close and confidential advisor of al-Ma"mùn in the year 194; al-ˇabarì, Ta"rìkh, III, 799–802; al-Mas'ùdì, Murùj al-dhahab wa-ma'àdin al-jawhar, ed. C. Pellat (Beirut, 1966–1979), IV, 265: A˙mad b. Hishàm, the cousin of ˇàhir b. al-Óusayn; al-ˇabarì, Ta"rìkh, III, 1108; on his important position in Marw, see alˇabarì, Ta"rìkh, III, 841, ll. 15; but cf. al-Jahshiyàri, Kitàb al-wuzarà", 306. 42 Al-ˇabarì, Ta"rìkh, III, 794. 43 But see al-Dìnawarì, al-Akhbàr al-†iwàl, 397: 10,000 soldiers; Bal'amì’s translation of al-ˇabarì (French Translation: Chronique de . . . Tabari, traduit sur la version persane d’Abou-'al, Mohammed Bel'am, trans. H. Zotenberg (Paris, 1867–74), IV, 482): 20,000 soldiers; These and many other sources for information about the number of ˇàhir’s armies are recorded by Daniel, Khuràsàn, 179 and note 94, citing the above quoted sources and others; these numbers are of course extremely exaggerated and are totally unreliable. 44 Ibn al-Athìr, Al-Kàmil (ed. Tornberg), VI, 151. 40

MAWÀLÌ

in the composition of al-ma"mùn’s army

289

between 193/809 and 194/809–810,45 or whether all or part of the army was subordinated to him when he was sent as governor of alRayy to prepare for battle against al-Amìn.46 We hear of the Turks in his army only once more, during the siege of Baghdàd, especially towards its end.47 Almost nothing is known of Harthama b. A'yan’s army, which arrived in Iraq only in 195 or 196. It seems to have been relatively large. One source reports that he reached ˇàhir at Óulwàn at the head of thirty thousand soldiers of the armies of Khuràsàn ( junùd Khuràsàn). It is highly plausible that his army consisted of abnà" as well as of Arab units.48 Al-Ma"mùn drafted into his army and into his service a number of other notables from Khuràsàn, whose families had actively participated in the first 'Abbàsid da'wa. One of the most prominent was (the Arab) Óumayd b. 'Abd al-Óamìd al-ˇà"ì, whose father was one of the 70 du'àt.49 These families preserved their distinctiveness and high status in Khuràsàn. There is of course a difference between them and the abnà" who came from Baghdàd, a different history causing a different group identity. However, it is important to emphasize the historic roots of these families, which are identical to those 45 Joining the army of Harthama b. A'yan: ibid., 209; see also al-ˇabarì, Ta"rìkh, III, 777; in 194/809–810 ˇàhir is mentioned among the commanders of Harthama in the siege of Ràfi' in Samarqand, Fragmenta Historicorum Arabicorum [Kitàb al-'uyùn wa-l-˙adà"iq], ed. M.J. De Goeje (Leiden: E.J. Brill, 1871), 322. 46 Ayalon, “al-Mu'taßim,” 5; Pipes, Slave Soldiers and Islam. The Genesis of a Military System (New Haven/London, 1981), 180. 47 Al-ˇabarì, Ta"rìkh, III, 891; al-Mas'ùdì, Murùj, IV, 292, but cf. al-ˇabarì, Ta"rìkh, III, 915: instead of Turks and Daylamites: ahl al-Ar∂; an exhaustive discussion on ˇàhir’s campaigns and the composition of his armies, see Elad, 'Abbàsid Army, 248, 279ff., 283, 285, 387–407. 48 Thirty thousands soldiers: al-Dìnawarì, al-Akhbàr al-†iwàl, 399. This number is a schematic number, cliché with al-Dìnawarì. It is precisely the number of ahl Khuràsàn units that were in Samarqand and fought at the side of Ràfi' b. al-Layth; some of them joined Harthama’s army, including some commanders who belonged to abnà" al-shì'a of the 'Abbàsid state ( fì 'idda min abnà" al-shì'a). The contingents of Ràfi' b. al-Layth, were composed, then, mainly of units of ahl Khuràsàn, that is to say, an army which certainly had Abnà" units and other Arab units (see al-Dìnawarì, Al-Akhbàr al-†iwàl, 391). Turks and other non-Arabs indeed fought alongside Ràfi', but they had left him before the siege of Harthama began. Thus, it is certainly impossible to state that “al-Ma"mùn’s new dual attitude towards Islam and the Turks caused Ràfi' b. al-Layth and his Turkish supporters to join al-Ma"mùn.” (Ismà'ìl, “Mu'taßim and the Turks,” Bulletin of the School of Oriental and African Studies 29 (1966), 14); the evidence of the Turk’s abandonment of Ràfi' ’s army: Ibn alAthìr, Al-Kàmil (ed. Torenberg), VI, 155. 49 Akhbàr al-'Abbàs, 218, 221; on him and his family, see Crone, Slaves, 174–75.

290

amikam elad

of the abnà" from Iraq, and their ties with the 'Abbàsid caliphate and devotion to it. Their position in Khuràsàn derived from these special ties.50 In Óumayd b. 'Abd al-Óamìd’s army there were Arab commanders and Arab units, part of whom certainly came from Khuràsàn. This is learned from a unique and extraordinary tradition, related by al-Azdì, about the big military expedition of Óumayd’s son Mu˙ammad, against Bàbak.51 The abnà" units after al-Amìn’s death Towards the end of 196/summer 812, al-Amìn’s army began to disintegrate and his commanders, together with their soldiers, began to abandon his camp and move to that of ˇàhir.52 Towards the end of the siege the stream of al-Amìn’s commanders who transferred their allegiance to ˇàhir increased, including the most important of the abnà" officers.53 Together with them, veteran Arab commanders who had served al-Amìn joined al-Ma"mùn’s army.54 The abnà" continued to form an active and vital military force (albeit divided and quarrelsome) during the years 200/815–204/819. Their number was not small;55 they fought against al-Óasan b. Sahl’s army also during the caliphate of Ibràhìm b. al-Mahdì (202/817– 203/818).56 They gradually began to pass over to al-Ma"mùn’s faction.57 This occurred on a large scale during 203, possibly after the death of al-Fa∂l b. Sahl became known.58

Elad, 'Abbàsid Army, 152–79. Al-Azdì, Al-Mawßil, 386–87, 395; al-Ya'qùbì, Ta"rìkh (ed. Leiden), II, 565, 569–70. Abù Sa'ìd, Mu˙ammad b. Yùsuf b. 'Abd al-Ra˙màn al-ˇà"ì is in charge of the Qalb of the army: al-Azdì, Al-Mawßil, 387; al-ˇabarì, Ta"rìkh, III, 1407: alMarwazì, that is, his origin is from Marw; ibid.: mawlà ˇayyi" (Ibn al-Athìr, AlKàmil (ed. Beirut), VI, 412: al-ˇà"ì); Ibn Khayyà†, Ta"rìkh, II, 745–46: Aßram b. 'Abd al-Óamìd, the brother of Óumayd b. 'Abd al-Óamìd, was the governor of Sìstàn, on behalf of al-Rashìd (Crone, Slaves, 175). His son, al-Mahdì (or Sa'dì or Ía'dì) b. Aßram, headed the maymana of the army of Mu˙ammad b. Óumayd against Bàbak (al-Ya'qùbì, Ta"rìkh, 565; Crone, loc. cit.). 52 Al-ˇabarì, Ta"rìkh, III, 867, 871. 53 Ibid., 881, 882–83, 903–904; Elad, 'Abbàsid Army, 289, 304–305. 54 Al-ˇabarì, Ta"rìkh, III, 986, 992, 1003, 1005, 1017–1018, 1020, 1028. 55 Ibid., III, 1007–1008: the exaggerated figure given by the transmitter is 125,000 horsemen and infantry (cited in Crone, “Abnà",” 18). 56 Al-ˇabarì, Ta"rìkh, III, 1016, 1018–19, 1022; Elad, 'Abbàsid Army, 418–21. 57 Al-ˇabarì, Ta"rìkh, III, 1030–32. 58 Ibid., 1032–33. 50

51

MAWÀLÌ

in the composition of al-ma"mùn’s army

291

Al-Ma"mùn entered Baghdàd on Íafar 16, 204/August 12, 820.59 It may be assumed that he arrived from Marw with an army and commanders, but the sources at our disposal do not provide us with any information as to its size and make-up. It seems that he did not arrive from Khuràsàn with a large army. Units of his army that were in Iraq, which had come years before with their commanders; Arab commanders with their units; abnà" commanders with their units—all seem to have fought under one banner. The sources mention only two commanders who came with al-Ma"mùn from Khuràsàn.60 At least four of al-Ma"mùn’s senior commanders who left Khuràsàn for Iraq were descendants of members of the original 'Abbàsid da'wa whose families had remained in Khuràsàn. These are ˇàhir b. al-Óusayn and his family, of whom 'Alì b. Hishàm may be noted in particular;61 Óumayd b. 'Abd al-Óamìd and his son Mu˙ammad, together with other family members; and 'Ujayf b. 'Anbasa.62 Ibn Abì ˇàhir ˇayfùr, Kitàb Baghdàd, 2–3; [al-ˇabarì, Ta"rìkh, III, 1037]. The first is Ya˙yà b. Mu'àdh b. Muslim, see Crone, Slaves, 184; the second (who is mentioned for the first and the last time as a commander) is Dàwùd b. Màsajùr, see al-ˇabarì, Ta"rìkh, III, 1045. 61 The identification of 'Alì b. Hishàm’s family with the ˇàhirid family can be established by the following evidence: 1) al-Khalìl b. Hishàm [b. Muß'ab b. Zurayq (Ruzayq)] is mentioned as the nephew of al-Óusayn b. Muß'ab [b. Zurayq] (alˇabarì, Ta"rìkh, III, 771); 2) al-Khalìl b. Hishàm is also mentioned as the brother of 'Alì b. Hishàm (al-Ißfàhànì, Kitàb al-aghànì (Bùlàq, 1284–85 AH), XVII, 145); 3) A˙mad b. al-Khalìl b. Hishàm is mentioned as the son of the sister [ibn ukhtihimà, sic!, a misprint, read: ibn akhìhimà] of 'Alì and Óusayn, the sons of Hishàm (alYa'qùbì, Ta"rìkh (ed. Leiden), II, 570; [ed. Beirut, II, 467]); 4) a poem lamenting the death of 'Alì b. Hishàm mentions his four brothers: A˙mad, Óusayn, Naßr and al-Khalìl (al-Ißfahànì, Aghànì (ed. Bùlàq), VII, 36 [ed. Dàr al-Kutub, VII, 306]); on 'Alì b. Hishàm, see Elad, 'Abbàsid Army, 227–28, 315–17. 62 On Óumayd b. 'Abd al-Óamìd and his family, see Elad, 'Abbàsid Army, 317–24; Kennedy, The Early 'Abbàsid Caliphate, 165; Crone, Slaves, 175; id., “Abnà",” 6; on 'Ujayf b. 'Anbasa, see Ibn Óazm, Naq† al-'arùs fì tawàrìkh al-khulafà", ed. Shawkì Îayf (Cairo, 1951), 166; id., Jamharat ansàb al-'arab, ed. 'Abd al-Salàm Mu˙ammad Hàrùn (Cairo, 1962), 184: Joining the rebellion of Ràfi' b. al-Layth, together with ˇàhir b. al-Óusayn; al-ˇabarì, Ta"rìkh, III, 732 (year 192/807–808): leaving the camp of Ràfi' b. al-Layth together with a number of abnà" al-Shì'a) (mentioned by Crone, “Abnà",” 4); he was a relative of the 'Abbàsid family; 'Alì b. 'Abdallàh b. al-'Abbàs’ wife, Sa'dà, the mother of Íàli˙ and Sulaymàn the sons of 'Alì, was from the family of 'Ujayf b. 'Anbasa, see Akhbàr al-'Abbàs, 155; al-Mubarrad, AlKitàb al-kàmil, ed. W. Wright (Leipzig, 1864–92), II, 220 (quoted by al-Dùrì, the editor of Akhbàr al-"Abbàs, 155 note 2: min saby al-ßughd min rah† 'Ujayf b. 'Anbasa); see also al-Balàdhurì, Ansàb al-ashràf, ed. 'A.'A. Dùrì (Beirut, 1978), III, 72; the next time we hear about him is twenty years later, as one of the commanders of al-Ma"mùn, see al-ˇabarì, Ta"rìkh, III, 1093; al-Azdì, Al-Mawßil, 368; he became 59

60

292

amikam elad

At first fate did not smile upon the important abnà" commanders in Baghdàd: rather, al-Ma"mùn tried to attack and neutralize their senior position. Al-Ma"mùn was certainly caught on the horns of a difficult dilemma. How should he treat the abnà" commanders and their armies that had fought against him for most of the period he had spent in Khuràsàn, almost until his entry into Baghdàd? The expulsion of his representative in Iraq, al-Óasan b. Sahl, the military struggle against his commanders in Iraq, the nomination of Ibràhìm b. al-Mahdì as anti-caliph, could not but leave residues of bitterness in his heart and mind against the abnà". Even if we accept al-Jà˙iΩ’s claim that the removal of the abnà" from their senior positions was a private initiative on the part of Óumayd b. 'Abd alÓamìd, it was received positively and approved by al-Ma"mùn. However, this initiative failed. Al-Ma"mùn paid the abnà" commanders all payments due to them and returned them to their former positions.63 The reasons for this were no doubt practical. Al-Ma"mùn could not manage without the military power of the abnà" [!] Most of the western provinces of the caliphate were in a state of anarchy and insurrection. Al-Ma"mùn’s primary task was to suppress these rebellions and restore order to the provinces. Abnà" commanders and units were sent to carry out many of these assignments.64 At this stage al-Ma"mùn was unable (and perhaps unwilling) to raise a new army. The organization of a new army, an alternative to the abnà", who held positions of social and economic power and influence in the 'Abbàsid state, would no doubt have met with stiff resistance on their part. Their position was deeply rooted in the ideological struc-

one of the most important commanders of al-Ma"mùn, see for instance, al-ˇabarì, Ta"rìkh, III, 1105, 1107; Ibn Abì ˇàhir ˇayfùr, Kitàb Baghdàd, 267, 269–70 [alˇabarì, Ta"rìkh, III, 1108–1109]; al-Ißfahànì, Aghànì (ed. Bùlàq), VII, 31, 32 [ed. Dàr al-Kutub, VII, 294, 296]; al-Ya'qùbì, Ta"rìkh (ed. Leiden), II, 574: the Óàjib of al-Ma"mùn; after the death of the latter he continued to serve al-Mu'taßim. He was executed by this caliph, being one of the leaders of the rebellion against him, see al-ˇabarì, Ta"rìkh, III, 1256–57, 1266. 63 Al-Jà˙iΩ, “Dhamm akhlàq al-kuttàb,” in 'Abd al-Salàm Mu˙ammad Hàrùn, ed., Rasà"il al-Jà˙iΩ (Cairo, 1384/1965) II, 206–208; M. Sharon, “The Military Reforms of Abù Muslim their Background and Consequences,” in M. Sharon, ed., Studies in Islamic History and Civilization in Honour of Professor David Ayalon (Leiden/Jerusalem: E.J. Brill and Cana Publishing House, 1986), 137–38; Crone, “Abnà",” 9–10. 64 Al-ˇabarì, Ta"rìkh, III, 1044–1045 (year 205/820–821, 'Ìsà b. Mu˙ammad b. Abì Khàlid al-Abnàwì, the senior commander of al-abnà" is sent against Bàbak, and appointed simultaneously as the governor of Armìniya and Àdharbayjàn); al-Ya'qùbì, Ta"rìkh (ed. Leiden), II, 563–64.

MAWÀLÌ

in the composition of al-ma"mùn’s army

293

ture of the caliphate, forming part of the historic legitimization of the existence of 'Abbàsid rule. 'Ìsà b. Mu˙ammad b. Abì Khàlid alAbnàwì is praised as a man who belongs to al-Ma"mùn’s dynasty (min ahl dawlatika) and this position was gained by the position of his forefathers in the 'Abbàsid early dawla: wa-ammà 'Ìsà b. Mu˙ammad b. Abì Khàlid fa-rajul min ahl dawlatika wa-sàbiqatuhum wa-sàbiqat man ma∂à min salafihi, sàbiqatuhum tarji' 'alayka bi-dhàlika);65 abnà" are described as Anßàr al-Dawla of al-Ma"mùn.66 Necessity caused the position of the abnà" during al-Ma"mùn’s reign to continue to be strong, mainly due to their unabated military vigor. According to Ibn Abì ˇàhir ˇayfùr (or his source), they were the best units in the 'Abbàsid army at that time. Ibn Abì ˇàhir ˇayfùr, the contemporary of al-Ma"mùn, records most probably an imaginary discussion between al-Ma"mùn, al-Muta'ßim, and an anonymous commander, in which they disagreed about who was the best army. Al-Ma"mùn was in favor of 'ajam ahl Khuràsàn, al-Mu'taßim was in favor of the Turks and the commander was in favor of the abnà", ending his speech saying: “For it was they who put [the heads of ] the Turks into dogs’ collars, and it was their fathers who were the leaders of the ['Abbàsid] dynasty, and it was they [the abnà"] who undertook the fighting against the caliph [Ma"mùn], then they obeyed him and the caliphate was kept in good order by means of them” (wa-abà"uhum hum alladhìna qàdù al-dawla wa-hum qàmù bi-˙arb Amìr al-Mu"minìn thumma a†à'ùhu fa-istaqàmat al-khilàfa bihim). To solve the dispute Naßr b. Shabath [who rebelled against the 'Abbàsids, captured and was kept in Baghdàd] is brought, his opinion is asked regarding this topic and he highly praises the abnà", as the best of all armies.67 Regarding this text Ayalon remarks that “It is doubtful, of course, whether the discussion described above ever took place. Yet, had the abnà" not constituted a major military force in the 'Abbàsid army, such a story could not have been invented. The military importance of the abnà" is also confirmed by the prominent place which al-Jà˙iΩ accords them in his Manàqib, and far more by their share in the AmìnMa"mùn struggle. ˇayfùr, if I understood him correctly, also states

65

Al-ˇabarì, Ta"rìkh, III, 1068. Ibid., 1070; for further discussion, see Elad, 'Abbàsid Army, 426–37. 67 Ibn Abì ˇàhir ˇayfùr, Kitàb Baghdàd, 143, l. 7–144, l. 14; Ayalon, “al-Mu'taßim,” 30–31; Lassner, Shaping, 281 note 72; Elad, 'Abbàsid Army, 429–30; Crone, “Abnà",” 10. 66

294

amikam elad

that it was they who fought al-Ma"mùn.68 But the same author goes much further in attributing the stability of the caliphate under Ma"mùn to the abnà", who came to terms with him and accepted his rule. The position of the abnà" under Ma"mùn is very vague. Yet, it is quite certain that ˇayfùr erred on the side of exaggeration. The abnà", in all probability, though much shaken, still remained a force to be reckoned with.”69 It seems though, that the description of the course of events in Baghdàd by al-Jà˙iΩ, after al-Ma"mùn’s entrance to the city (in 204/819), corresponds exactly to the course of events described by Ibn Abì ˇàhir ˇayfùr (through the mouth of the commander), about the importance of the abnà" units after al-Ma"mùn has settled in Baghdàd. If the discussion occurred at all, it must have taken place after the year 210, for Naßr b. Shabath was brought to Baghdàd on the 7th of Íafar 210/23 of May 825.70 One of the strongest pieces of evidence for the position of the abnà" in Baghdàd and their being the main military force is al-Mu'taßim’s move to Sàmarrà" with his Mamlùk units. The bitter antagonism and struggle in Baghdàd between the abnà" and al-Mu'taßim’s troops were a decisive reason for his move to the new city.71 There are relatively many reports of commanders who belonged to the most distinguished families of the abnà". A large proportion of them and members of their families held posts as senior commanders, governors and in the highest levels of administration during al-Ma"mùn’s caliphate,72 and even afterwards.73

68

This is also stressed by Crone, loc. cit. Ayalon, “al-Mu'taßim,” 31. 70 Al-ˇabarì, Ta"rìkh, III, 1073; if invented, it is not a sophisticated invention; it is a well-known technique in Arabic literature: the inventors of the tradition bring Naßr b. Shabath, the fierce enemy of the abnà" to speak in their praise. 71 Ayalon, “al-Mu'taßim,” 5, 11, 31; id., “The Muslim City and the Military Aristocracy,” in Proceedings of the Israel Academy of Sciences and Humanities, no. 2 (1968), 314. [Reprinted in id., Studies on the Mamlùks of Egypt. 1250–1517, no. 7 (London: Variorum Reprints, 1977)]; id., “Preliminary Remarks on the Mamlùk Military Institution in Islam,” in V.J. Perry and M.E. Yapp, eds., War, Technology and Society in the Middle East (Oxford, 1975), 53–54. [Reprinted in D. Ayalon, The Mamlùk Military Society, no. 9 (London: Variorum Reprints, 1979]; Ismà'ìl, “al-Mu'taßim,” 3, 5; Crone, Slaves, 261 note 625; Pipes, Slaves, 151; Lassner, Shaping, 281 note 73; but cf. Shaban, History, II, 63. 72 Crone, Slaves, 177, 183–85; id., “Abnà",” 4; Elad, 'Abbàsid Army, 431–36. 73 The family of Óumayd b. 'Abd al-Óamìd: commanders and governors, see 69

MAWÀLÌ

in the composition of al-ma"mùn’s army

295

The appeal to the Arab tribes in Khuràsàn (and their mobilization?) Upon his arrival in Marw, al-Ma"mùn was faced with several urgent problems. A number of kings and princes in Transoxania had rebelled, or were on the verge of rebellion. These areas were ruled by kings of Irano-Turkish origin, who paid tribute to the caliphate.74 It appears that al-Ma"mùn was able to subdue these rebellious regions relatively easily.75 A second and more arduous task facing him was the consolidation of his rule within Khuràsàn itself, especially the establishment of a large and strong army as quickly as possible. We do not have much information about the ways and means by which al-Ma"mùn strengthened his rule in Khuràsàn. The question is whom he wanted to base his rule upon. Whose support was required in order to establish secure rule in Khuràsàn, and among which population groups could he raise forces for his army? We have no evidence that he wanted to base himself upon non-Arab forces, or that he did so. Did he want to base himself upon the Arab tribes living in Khuràsàn? In this context, the question must immediately be asked: Were there Arab tribes in Khuràsàn in 193/809? Did they live in the same centers in which they had lived towards the end of the Umayyad rule? What was their socialcultural-economic character? During the 'Abbàsid period we hear no more of tribal struggles in Khuràsàn. We do not know what happened to those tribesmen and their families who settled in the cities Crone, Slaves, 175; Barmakids: governors and secretaries, see ibid, 177; the family of Harthama b. A'yan: governors, see ibid., 178; the family of Jibrìl b. Ya˙yà alBajalì: heads of the shur†a, see ibid., 180; the family of Khuzayma b. Khàzim: commanders, ibid., 181; the family of Mu'àdh b. Muslim: governors and heads of the ˙aras, ibid., 184; the family of Mu˙ammad b. al-Ash'ath: ßà˙ib al-barìd, ibid., 185; the family of Mu˙ammad b. Abì Khàlid: a commander, see al-ˇabarì, Ta’rìkh, III, 1318–19; evidence for the activities of al-abnà" units after al-Ma"mùn’s reign, see Ayalon, “al-Mu'taßim,” 35–36; Lassner, Shaping, 281 note 73 (relating to abnà" units in 248/862–863, during the reign of al-Musta'ìn according to al-Ya'qùbì, Ta"rìkh, 604; al-ˇabarì, Ta"rìkh, III, 1510–11 (Crone, “Abnà",” 5), and in 256/869–870 during al-Muhtadì’s reign, according to al-Ya'qùbì, Ta"rìkh, 618–19 (Crone, loc. cit.); in both cases al-abnà" units fought with other units). Al-Abnà" units played a decisive military role in the campaigns against Bàbak, but we also find them in other campaigns, during al-Mu'taßim’s reign, see al-ˇabarì, Ta"rìkh, III, 1169, 1173, 1180–81, 1273–74, 1278–79 (year 224 [Shaban, History, II, 67]), 1318–19; al-Ißfahànì, Aghànì (ed. Bùlàq), XVIII, 18, 46 [ed. Dàr al-Kutub, XX, 157]; and see Crone, loc. cit., for additional evidence of units of abnà". 74 Shaban, History, II, 42; Kennedy, The Early 'Abbàsid Caliphate, 138. 75 Al-ˇabarì, Ta"rìkh, III, 816; Gabrieli, “Hàrùn ar-Rashìd,” 358; Kennedy, loc. cit.

296

amikam elad

and the villages surrounding the cities, and what processes they underwent. We have a small number of reports (almost rare ones) of the existence of Arab tribes in Khuràsàn in the 'Abbàsid period, including the time of al-Amìn and al-Ma"mùn. As noted they are few, scattered and unfocussed geographically, and it is hard to draw a picture from them that can reflect, even if only partially, the state of the tribes in Khuràsàn in that time.76 An interesting and most important tradition sheds light regarding the Arab tribes in Khuràsàn (apparently mainly in Marw and its vicinity) in 193/809. The complete tradition is cited by al-ˇabarì from al-Fa∂l b. Sahl, speaking in the first person. It describes the situation in Khuràsàn immediately after Hàrùn al-Rashìd’s death. Al-Fa∂l encourages al-Ma"mùn, by telling him that his condition in Khuràsàn is relatively good and stable and adds: “Tell me, how do you think the people77 were, when they received the news of [the rebellion of ] Ràfi' [b. al-Layth]? He [al-Ma"mùn] said: ‘I saw them78 in a state of great commotion and disturbance.’ I said: ‘All the more so regarding you79 when you are lodged among your maternal uncles (bayn akhwàlika), with the oath of allegiance to you binding them, (see) how [big] will the disturbance of the troops of Baghdàd (ahl Baghdàd ) be. Be patient, and I will guarantee you the caliphate.’ I placed my hand on my chest. He said: ‘Indeed I am going right away to pass the matter [i.e., the rule] into your hands. Accept it and carry it out.’ I said [al-Fa∂l], ‘let me give you sincerely good advice (la-aßduqannaka),80 'Abdallàh b. Màlik, Ya˙yà b. Mu'àdh, and the leading commanders (umarà" al-ru"asà") we have named—if they undertake the matter for you, they will be more advantageous for you than I, with their renowned leadership and because of the strength they possess for fighting. Whoever undertakes the matter, I will be his servant until you attain your desire and decide as you think best about me. So I met them in their lodgings. I reminded them of the oath of allegiance binding them and of their obligation to be loyal (to it). It was as if I had brought them carrion on a plate. One of them said, ‘this is not lawful; go away [!]’ Another of For evidence, see Elad, 'Abbàsid Army, 165–67, 195. Ar. al-nàs; Fishbein, The History of al-ˇabari. Volume XXXI: The War Between Brothers, trans. and ed. M. Fishbein (New York: State University of New York Press, 1992), 16: the troops. 78 Ar. ra"aytuhum. Fishbein, loc. cit.: I think. 79 Ar. fa-kayfa bika. Fishbein, loc. cit.: How will they be regarding you. 80 Fishbein, loc. cit.: “By God, I will speak the truth to you.” 76

77

MAWÀLÌ

in the composition of al-ma"mùn’s army

297

them said, ‘who would interfere between the Commander of the Faithful and his brother?’ So I came and informed [al-Ma"mùn].’ He said: ‘undertake the matter.’ I said [to al-Ma"mùn]: ‘you have read the Qur"àn, heard Óadìth, and you have possessed knowledge of the law (religion?).81 The best advise (opinion) is for you to send for the fuqahà" who are preset [in Marw?]82 summoning them to justice and its performance, and to revive the Sunna. You should sit on felt carpets and answer the cases brought to you (in the framework of the) maΩàlim (institution)’ [which is part of the office of the caliph].83 We did so. We sent to the scholars of law ( fuqahà"). Also we treated the commanders, kings and sons of kings generously. We would say to a member of the tribe of Tamìm, ‘we will put you in (the distinguished) place of Mùsà b. Ka'b’; to a member of Rabì'a [we used to say], ‘we would set you in the place of Abù Dàwùd Khàlid b. Ibràhìm’; and to the Yamanì [we used to say] ‘we will put you in the (distinguished) place of Qa˙†aba [b. Shabìb] and Màlik b. al-Haytham.’84 And we summoned each tribe to (follow the greatness of ) the [twelve] senior propagandists [of the 'Abbàsid da'wa] (who were) their tribal leaders. We gained the hearts of the heads (of the tribes) by telling them things similar to this.85 We reduced Khuràsàn’s kharàj ’s tax by a quarter. This [policy] pleased them and they were glad of it and they said, ‘The son of our sister, and the son of the (paternal) uncle of the Prophet, may God bless him and grant him peace.’”86

81 Ar: tafaqqahta fì l-dìn; Fishbein, loc. cit.: become learned in religion; professor Crone has pointed out to me that this is a Qur"ànic expression (Q. 9:123); note that al-Jahshiyàrì, Kitàb al-wuzarà", 278, l. 16 has: wa-fahimta amr al-dìn. 82 Fishbein, loc. cit.: to the learned men in your entourage. 83 Ar.: wa-tarudda l-maΩàlim; Fishbein, loc. cit.: and relieve injustices. 84 All belonged to the highest body of propagadists (nuqabà"), within the framework of the first da'wa. On them, see Crone, Slaves, p. 186; Akhbàr al-'Abbàs, pp. 215–216 (Mùsà b. Ka'b); Crone, op. cit., p. 181 (Màlik b. Haytham); p. 188; Kennedy, 'Abbàsid Caliphate, pp. 62–63 (Qa˙†aba b. Shabìb); on Abù Dàwùd, Khàlid b. Ibràhìm, see Akhbàr al-'Abbàs, p. 216; Sharon, Black Banners, pp. 193, 221–222; Daniel, Khuràsàn, Index. 85 Fishbein, War Between Brothers, 17, ll. 14–18: “We summoned each tribe to the greatness of their heads. We gained the favor of the heads and said things like this to them.” 86 Al-ˇabarì, Ta"rìkh, III, 773–74; my translation is partially based on Fishbein’s. Partial parallel texts: Ibn al-Athìr, Al-Kàmil (ed. Tornberg), VI, 154–55; al-Maqrìzì, Kitàb al-muqaffà al-kabìr, ed. Mu˙ammad al-Ya'làwì (Beirut: Dar al-Gharb al-Islàmì, 1411/1991), VI, 265 (copies Ibn al-Athìr); al-Jahshiyàrì, Kitàb al-wuzarà", 277–79 (an abridged tradition with some changes, see discussion below); Fragmenta, 321 (an abridged tradition with major changes, see discussion below).

298

amikam elad

This tradition mentions important matters, some of which will be discussed below. (a) The first point is the emphasis on Khuràsàn and its prominence vis-à-vis Iraq. This is not necessarily a contrast. There is an emphasis throughout the 'Abbàsid caliphate on Khuràsàn’s importance to the center of the caliphate, to Iraq, and the many difficulties that appeared in Iraq when the need arose to suppress the great rebellions in Khuràsàn. Here there is no trend of separatism and breaking away from Iraq. On the contrary, al-Fa∂l b. Sahl tells alMa"mùn that his stay in Khuràsàn will increase the confusion and disturbance in the army in Baghdàd. The aim, however, is to gain control of the caliphate. There is no separatist trend at work here (anà u∂minu laka al-khilàfa). Al-Ma"mùn’s situation is better than that of all preceding rebels in Khuràsàn, as he lives, according to al-Fa∂l, among his maternal uncles and they are tied to him by the oath of allegiance they swore to him ( fa anta nàzil fì akhwàlika wa-bay'atuka fì a'nàqihim. See discussion below). (b) The second point is the position and the role of al-Fa∂l b. Sahl. Within this framework I am unable to discuss this in detail. His character, position and role have been accurately and extensively portrayed by Sourdel.87 The aim of the tradition before us is to praise and glorify al-Fa∂l and to emphasize his crucial position and influence over the caliph. It stresses his great modesty and humility. Against his will he is “forced” to accept the running of the caliphate, in a kind of false modesty that is accentuated in a very unsophisticated way in the tradition.88 In this context the tradition highlights the contrast and tension existing between al-Fa∂l b. Sahl and the senior abnà" commanders in al-Ma"mùn’s army. The two commanders mentioned, together with other commanders, came with al-Ma"mùn from ˇùs to Marw. However, the reason for the opposition is unclear. It is unclear from the text what al-Fa∂l was asking from them. His demand was left out of the text, and only their definite answer is given, expressing disgust and abhorrence for his proposal. It is clear that the author 87 D. Sourdel, Le vizirat 'Abbaside de 749 a 936 (Damascus, 1959–60), Index, esp. 196–213; id., art. “al-Fa∂l b. Sahl,” EI 2, s.v. 88 See for instance, al-ˇabarì, Ta"rìkh, III, 774, ll. 1–2 (Ibn al-Athìr, Al-Kàmil (ed. Beirut), VI, 224, ll. 14–15, a different variant).

MAWÀLÌ

in the composition of al-ma"mùn’s army

299

of the tradition believes him to be asking them to take active step against al-Amìn that would cause his downfall and the enthronement of al-Ma"mùn as caliph in his place. Thus he says to al-Ma"mùn in the tradition itself: “I will guarantee you the caliphate.” It is difficult otherwise to understand the commander’s definitely negative answer. Of course, it is possible to claim that here, too, we have “praise of al-Fa∂l,” and an anachronistic viewpoint. It may have seemed obvious to the composer of the tradition that even before the struggle between the brothers began, al-Fa∂l and al-Ma"mùn had already thoughts of deposing al-Amìn and winning the caliphate. This is an interesting point. It would be tempting to think that alMa"mùn was thinking of taking over the caliphate already when he arrived in Marw (if not before) and had no intention of carrying out his father’s will. We have two traditions, which provide faint echoes of such a trend.89 However, they are swallowed up in a sea of traditions that present al-Ma"mùn as attacked, pitiable, peace loving, standing against the unequivocally aggressive policy of his brother al-Amìn to strip him of his heirdom and of the governorship of Khuràsàn. This view is in fact accepted by all scholars who have studied the conflict between al-Amìn and al-Ma"mùn. (d) An additional point is the ideological and religious basis upon which al-Fa∂l wishes to found al-Ma"mùn’s rule in Khuràsàn. AlMa"mùn has the legal basis for ruling—knowledge of the Qur"àn, the ˙adìth and fiqh. He bases his rule on the Sunna according to which justice must reign. There is no hint of non-Sunnì tendencies (certainly not extreme ones) in this tradition. (e) Yet, the most important point (certainly within the framework of this article) is al-Ma"mùn’s activity among the Arab tribes of Khuràsàn: “And we paid respect to the commanders, the kings and the descendants of kings . . .” The translation was made, as noted, according to De Goeje’s edition of al-ˇabarì, with the important filling of a lacuna in the text according to the Egyptian edition of Mu˙ammad Abù l-Fa∂l Ibràhìm.90 Parallel versions may be found in Ibn al-Athìr, al-Maqrìzì, al-Jahshiyàrì and the anonymous Fragmenta (or Kitàb al-'uyùn wa-l-˙adà"iq).

89 Al-ˇabarì, Ta"rìkh, III, 771; a parallel garbled text (with changes) by al-Azdì, Al-Mawßil, 317–18. 90 Also noticed by Fishbein, War Between Brothers, 17 note 78.

300

amikam elad

Al-ˇabarì’s text (III, 774): l. 9) fa-fa'alnà wa-ba'athnà ilà l-fuqahà" wa-akramnà al-quwwàd wa-lmulùk wa-abnà" al-mulùk l. 10) fa-kunnà naqùlu li-l-Tamìmì nuqìmuka maqàm Mùsà b. Ka'b wa-lil-Raba'ì maqàm l. 11) Abì Dàwùd Khàlid b. Ibràhìm wa-li-l-Yamànì nuqìmuka maqàm Qa˙†aba wa-Màlik l. 12) b. al-Haytham; fa-kunnà nad'ù kulla qabìla ilà nuqabà"**ru"ùsihim wa-istamalnà l. 13) al-ru"ùs 91 wa-qulnà mithl dhàlika wa-˙a†a†nà 'an Khuràsàn rub' alkharàj fa-˙asuna l. 14) mawqi' dhàlika minhum wa-surrù bihi wa-qàlù: ibn ukhtinà wa-ibn 'amm al-Nabì (ßal'am) The text of Ibn al-Athìr (ed. Leiden, VI, 155; ed. Beirut, VI, 225): l. 2) . . . fa-fa'ala dhàlika jamì'ahu; wa akramahu al-quwwàd wa-l-mulùk wa-abnà" l. 3) al-mulùk. Wa-kàna yaqùlu li-l-Tamìmì nuqìmuka maqàm Mùsà b. Ka'b wa-li-l-Raba'ì l. 4) nuqìmuka maqàm Abì Dàwùd wa 92 Khàlid b. Ibràhìm wa-li-l-Yamànì nuqìmuka maqàm l. 5) Qa˙†aba wa-Màlik b. al-Haytham; wa kàna hà"ulà" nuqabà" al-dawla al-'Abbàsiyya wa-wa∂a'a l. 6) 'an Khuràsàn rub' al-kharàj fa-˙asuna dhàlika 'inda ahlihà wa-qàlù: ibn ukhtinà l. 7) wa-ibn 'amm nabiyyinà.93 Al-Maqrìzì’s Text (IV, 265): l. 2) fa-fa'ala dhàlika jamì'ahu wa-akrama al-quwwàd wa-l-mulùk wa-abnà" al-mulùk. wa-kàna yaqùlu 91 De Goeje’s ed.: fa-kunnà nad'ù kulla qabìla ilà nuqabà" ru"sà . . . tlnà al-ra"s; the correction is according to the edition of Abù l-Fa∂l Ibràhìm (Cairo: Dàr al-Ma'àrif, 1964), VIII, 372. 92 The letter “waw” is of course a misprint. 93 In line two, Ibn al-Athìr changes the first person to third person; he also skips a complete important sentence (ll. 12–13 in al-ˇabarì): wa-kunnà nad'ù kulla qabìla ilà nuqabà" . . . until: wa-qulnà mithl dhàlika. This sentence most probably, seemed problematic to Ibn al-Athìr, as well. Instead he inserts a sentence that clarifies alˇabarì’s text: all these are the nuqabà" of the 'Abbàsid Dawla. He also “explains” and clarifies who exactly liked this policy of al-Ma"mùn and al-Fa∂l: al-ˇabarì’s fa˙asuna mawqi' dhàlika minhum, is changed to: fa-Óasuna mawqi' dhàlika 'inda ahlihà.

MAWÀLÌ

l. 3) l. 4) l. 5) l. 6)

in the composition of al-ma"mùn’s army

301

li-l-Tamìmì nuqìmuka maqàm Mùsà b. Ka'b wa-li-l-Raba'ì nuqìmuka maqàm Abì Dàwùd Khàlid b. Ibràhìm, wa-li-l-Yamànì nuqìmuka maqàm Qa˙†aba waMàlik b. al-Haytham, wa-kullu hà"ulà" nuqabà" al-dawla al-'abbàsiyya. wa-wa∂a'a 'an ahl Khuràsàn rub' al-kharàj fa ˙asuna dhàlika 'inda ahlihà wa-qàlù: ibn ukhtinà wa-ibn 'amm nabiyyinà, ßallà Allàh 'alayhi wa-sallam.

Al-Jahshiyàrì’s text (278): l. 17) . . . wa-an taq'uda 'alà l. 18) lubùd wa-an tuwàßila al-naΩar fì l-maΩàlim wa-tukrima al-quwwàd wa-l-mulùk wa-abnà" al-mulùk p. 279, l. 1) fa-fa'ala dhàlika wa-kàna yaqùl li-l-Tamìmì: nuqìmuka maqàm Mùsà b. Ka'b l. 2) wa-yaqùlu li-l-Raba'ì nuqìmuka maqàm Abì Dàwùd; wa-yaqùlu li-lYamànì nuqìmuka maqàm l. 3) Qa˙†aba wa-Màlik b. al-Haytham. Wa-˙a††a 'an Khuràsàn rub' alkharàj fa-kànù l. 4) yaqùlùn ibn ukhtinà wa-ibn 'amm Rasùl Allàh.94 Up to now we have seen no significant changes in the parallel versions we have adduced. Al-Maqrìzì seems to have copied Ibn alAthìr, as the places where he abridges al-ˇabarì’s text are identical to those in Ibn al-Athìr. Ibn al-Athìr and al-Jahshiyàrì indeed change the spoken body of the tradition, shortening it slightly, but with the exception of one important and central sentence, their abbreviations are not significant. The text of the Fragmenta (Kitàb al-'Uyùn wa-l-˙adà"iq) is as follows (321): l. 10) . . . wa-anta l. 11) qad qara"ta al-Qur"àn wa-sami'ta al-a˙àdìth wa-tafaqqahta fì l-dìn fa-l-ra"y l. 12) an tab'ath ilà man bi-l-˙a∂ra min al-fuqahà" fa-tad'ùhum ilà l-˙aqq wa-l-'amal

94 In this text, as well as that of Ibn al-Athìr, the sentence about the summoning of the tribes in the names of their nuqabà" to al-Ma"mùn is missing.

302

amikam elad

l. 13) bihi wa-i˙yà" al-Sunna. Thumma al-Ma"mùn jalasa 'alà l-lubùd waradda al-maΩàlim l. 14) wa-akrama al-quwwàd wa-abnà" al-mulùk wa-mannà al-nàs waistamàla qulùb al-ru"asà" wa-˙a††a l. 14) 'an Khuràsàn rub' al-kharàj; fa-˙asuna mawqi' dhàlika wa-surrù bihi wa-qàlat al-Furs l. 15) ibn ukhtinà wa-ibn 'amm al-Nabì ßal'am, al-'àlim, al-'àdil al-zàhid. Lines 10 (from wa-anta) until l. 13 (i˙yà" al-Sunna) parallel al-ˇabarì, when suddenly the author of al-'Uyùn wa-l-˙adà"iq summarizes and epitomizes the tradition before him. Before this (in l. 9), the author has omitted the entire tale of al-Fa∂l’s going to the commanders, their answers and al-Ma"mùn’s subsequent selection of him. After that, he copies lines 10–13 above exactly and then begins to abridge the text, using the third person, leaving out the entire matter of the appeal to the different tribes, using the phrase wa-istamàla qulùb alru"asà". However, from his description it is impossible to know who these ru"asà" are (i.e., from which tribes), as he has omitted the sentences clarifying this. He then mentions the reducing of the kharàj tax by a quarter, suddenly copies a sentence from the original tradition, “And they were pleased by this and rejoiced at it,” and finally adds a text not to be found in any of the parallel sources: “And the Persians said, ‘The son of our sister, the son of the Prophet’s (Í) uncle’.” He also adds epithets to al-Ma"mùn’s name that are not found in the parallel versions: the (religious) scholar, the righteous, the ascetic. Al-ˇabarì, after mentioning the tribes and the tribal chiefs, reports “We reduced Khuràsàn’s kharàj etc. . . .95 The situation pleased them and they rejoiced in it.” According to al-ˇabarì, then, this can certainly be understood as referring to the tribes and tribal chiefs mentioned previously. The other versions change the text and thus also its meaning. Ibn al-Athìr and al-Maqrìzì, who copied him, report that “the situation pleased its [most probably Khuràsàn’s] people”: ahluhà, while the author of the Fragmenta adds “clarifying” that the “Persians” said, “the son of our sister etc.” When the al-'Uyùn wa-l-˙adà"iq’s text is studied closely, it seems that there is a method behind the policy of abridgments and omis95 Al-Maqrìzì is the only source to report “He reduced ahl Khuràsàn’s kharàj,” that is, adds the word ahl.

MAWÀLÌ

in the composition of al-ma"mùn’s army

303

sions adopted by the author. He has consistently deleted any mention of the abnà" commanders, any report of their status, importance and al-Fa∂l b. Sahl’s appeal to them in the knowledge that their status was higher than his own. He copies word for word from alˇabarì’s tradition the noble qualities of al-Ma"mùn and his great knowledge in various fields, which he reports in the first person, as in the original tradition. Suddenly he begins to summarize, in the third person, skipping over the entire section of al-Fa∂l’s appeal to the different tribes, calling them to al-Ma"mùn’s cause and attempting to rely on the nuqabà" who belonged to these tribes. Finally, he “clarifies” and adds that it was the Persians who said “the son of our sister, the (religious) scholar, the righteous, the ascetic.” Possibly the tradent, or the anonymous author of al-'Uyùn wa-l˙adà"iq himself saw in the expression “the son of our sister” and the reduction of the kharàj actions aimed at the Persians, and therefore added the phrase “the Persians.” However, it is possible that this was done consciously and deliberately, indicating an idea that the author wished to convey, that is, the view of al-Ma"mùn’s rule as Persian rule. Perhaps he had shu'ùbì tendencies, which caused him to omit the efforts al-Ma"mùn made to become closer to the Arab tribes and to base his rule in Khuràsàn on them.96

The second da'wa It is impossible to ignore the obvious lines of similarity between this propaganda of al-Ma"mùn among the Arab tribes of Khuràsàn to the original propaganda of the 'Abbàsid da'wa. According to this tradition, al-Ma"mùn approached his commanders (at this stage, most of them are the abnà" commanders who came with him from ˇùs), the kings and the descendants of the kings, who seem to be nonArab rulers of kingdoms and principalities in the Khuràsànì border regions, who paid tribute to al-Ma"mùn, and of course the Arab tribes of Khuràsàn. This is a unique piece of evidence for the existence of Arabs in a tribal framework (the tribes of Yaman, Rabì'a and Tamìm) in Khuràsàn and for al-Ma"mùn’s desire to include them in a new, second da'wa. 96 It seems that the word “Furs” is most uncommon within the context of parallel historical texts.

304

amikam elad

The section of the important text under discussion here has gone unnoticed by most students of the period. What was most obvious to them were the sentences emphasizing al-Ma"mùn’s apparent “Persianness.” Three scholars have referred briefly to the part dealing with the tribes, yet even they stress al-Fa∂l b. Sahl’s words to al-Ma"mùn, that he lives in the midst of his maternal uncles (akhwàl , and especially the phrase ibn ukhtinà).97 This text clearly shows that al-Ma"mùn began to assemble supporters and followers in Khuràsàn from among the Arab tribes. According to the wording of the text before us, it is possible to conclude that al-Ma"mùn was trying to revive and reconstruct the early 'Abbàsid da'wa and to create a second da'wa. This text is something of a “desert island” in the sea of historical chronicles. It is difficult to state as unequivocally as we have done that al-Ma"mùn wished to establish a second da'wa only on the basis of this single text. Luckily for us, the epistles of al-Ma"mùn’s secretary, A˙mad b. Yùsuf, which were preserved by Ibn Abì ˇàhir ˇayfùr in his book al-Manthùr wa-l-manΩùm, confirm this assertion, especially in the letter al-Risàla fì l-khamìs. In another place this epistle has been studied and translated.98 It is sufficient, then, to say that this epistle seems to have

97

Gabrieli, “Hàrùn ar-Rashìd,” 354–55, devote a lengthy discussion to this text, which fits well with the Iranian nature of al-Ma"mùn’s policy and preference, ignoring completely the part about the Arab tribes. Daniel, Khuràsàn, 178: “most interestingly, appeals were made to the various tribal factions in terms designed to recall the principles of the original 'Abbàsid da'wa: The Banù Tamìm were reminded of the example of the naqìb Mùsà b. Ka'b; the Rab[a]'ì of Abù Da"ùd, the Yaman of Qa˙†aba, and so on. The result of all this was that the people of Khuràsàn were delighted with al-Ma"mùn. They thought of him as their “nephew.” [My emphasis, A.E.]; Kennedy, The Early 'Abbàsid Caliphate, 138; Daniel and Kennedy do not analyze nor discuss the contents of the text; Mottahedeh, “The 'Abbàsid Caliphate in Iran,” in R.N. Frye, ed., The Cambridge History of Iran, vol. 4 (Cambridge: Cambridge University Press, 1975), 72: “The Khuràsànians accepted al-Ma"mùn as one of their own and, because his mother was an Iranian, called him ‘son of our sister’”; Amabe, 'Abbàsid Autocracy, 96 (possibly relying on this text), stresses the fact that al-Amìn’s mother was of Arab origin, while al-Ma"mùn, whose mother was the daughter of Khuràsànì lord, possibly Ustàdh Sìs, gained the support of these Khuràsànì lords due to this fact; for Fishbein, War Between Brothers, 16 note 70, the mention of al-Ma"mùn’s akhwàl, “alludes to al-Ma"mùn’s mother, Maràjil, who was of Iranian extraction”; and “son of our sister” “refers to the fact that al-Ma"mùn’s mother, Maràjil, was Persian. . . .” (ibid., 18 note 7). 98 The epistle was published by Íafwat, Jamharat rasà"il al-'Arab fì l-'ußùr al-'arabiyya al-zàhira (Cairo, 1356/1937), III, 386; al-Rifà'ì, 'Aßr al-Ma"mùn (Cairo, 1346/1928), III, 26–37. For analysis, discussion and French translation, see Arazi and Elad, “L’Épître à l’armée.”

MAWÀLÌ

in the composition of al-ma"mùn’s army

305

been composed after al-Amìn’s death (198/813), at a time when alMa"mùn was already the official caliph, but apparently still remained in Khuràsàn. According to this risàla, he was among his supporters from ahl Khuràsàn (bayna anßàrihi min ahl Khuràsàn). These supporters, ahl Khuràsàn, fought with him against al-Amìn, referred to here as al-makhlù', and described, as are his followers, as a traitor, a cheat and mainly as deniers of Islam (murràq al-milla, mukhàlifì ahl al-qibla). Their denial is ridda, that is, abandonment of Islam.99 Who are the ahl Khuràsàn, to whom al-Ma"mùn addresses his epistle, and of whom and whose actions does the epistle speak? AlMa"mùn himself (through his secretary) describes ahl Khuràsàn and their special virtues. Two are important to our discussion: (1) The help their fathers gave in the past to the Prophet’s family and to the ancestors of the caliph [al-Ma"mùn who succeeded him] ( fa mà taqaddama li-aslàfikum min nußrat ahl bayt al-Nabì wa-lqà"imìn bi-mìràthihi min àbà" Amìr al-Mu"minìn). (2) The help to al-Ma"mùn during the second da'wa, which was due to God’s preferring them.100 The highest and most honored position among ahl Khuràsàn was reserved for those who took part in both da'was, and by virtue of this, their position was higher than that of any others. ( fa-in kàna Allàh qad qaddamakum fì l-amrayn jamì'an bi-tafawwuq ˙àlikum 'alà ghayrikum).101 This supreme position is referred to a number of times as the Shì'a of Amìr al-Mu"minìn.102 Among the supporters of the second da'wa belonging to ahl Khuràsàn were important Muslims who did not belong to the first rank of ahl Khuràsàn, that is, those whose ancestors had taken part in the first 'Abbàsid da'wa. However, through having joined the second da'wa they became brothers to this first rank.103 Al-Risàla fì l-Khamìs informs us of a second da'wa organized and managed by al-Ma"mùn in Khuràsàn. His supporters are ahl Khuràsàn. The most important and respected body within it are the members 99

Arazi and Elad, “L’Épître à l’armée,” 42–43. Íafwat, Jamharat, III, 386; al-Rifà'ì, 'Aßr al-Ma"mùn, III, 34; Arazi and Elad, “L’Épître à l’armée,” 49–50. 101 Al-Rifà'ì, 'Aßr al-Ma"mùn, III, 31–32; Íafwat, Jamharat, III, 387; Arazi and Elad, loc. cit. 102 Al-Rifà'ì, 'Aßr al-Ma"mùn, III, 33, 37. 103 Ibid., 31. 100

306

amikam elad

of both da'was. Thus al-ˇabarì’s text and its parallels—about the propaganda among the Arab tribes from whose ranks the senior propagandists of the first da'wa had come—is connected to this important epistle, from which we learn that the Arab tribes responded to al-Ma"mùn’s appeal. His most important and loyal supporters, the senior members of his second da'wa, were Arab tribesmen, members of families and tribes whose ancestors had acted and fought in the first da'wa. The abnà" commanders who came with al-Ma"mùn to Marw presumably also became part of his second da'wa and reached the highest rank in its hierarchy. Ibn Ukht al-Nabì Íal'am The struggle between al-Amìn and al-Ma"mùn has been viewed by scholars from the nineteenth century until almost the present day as a struggle between Arabs and Persians.104 Modern scholars are sometimes more cautious in defining the two blocks as Arabs versus Persians, and create new quite sophisticated but certainly general, unclear and inaccurate definitions. Gabrieli was the first to express reservation about the absolute and unequivocal division of the factions into these two blocs.105 He pointed out (for the first time, as far as I know) that this scheme is too general. Although broadly speaking it is correct that these two blocs existed (Gabrieli contin-

104 A partial list of the scholars who held this view is provided by Samadi, “The Struggle Between the Two Brothers al-Amìn and al-Ma"mùn,” Islamic Culture 22 (1958), 102 note 3; many other scholars may be added to this list, for example, Bartold, Turkestan Down to the Mongol Invasion, trans. W. Bartold and H.A.R. Gibb (London, 1928), 207; Gabrieli, “Hàrùn ar-Rashìd,” 348, 353, 359 and 395–96; id., Al-Ma"mùn e gli 'Alidi (Leipzig, 1929), 29ff.; 'A.'A. al-Dùrì, Al-'Aßr al-'Abbàsì al-Awwal (Baghdàd, 1945), 193 and 199–207; M. Shàkir, Dawlat Banì l-'Abbàs (Kuwayt, 1973), I, 444–45; H. al-Bàshà, Diràsàt fì Ta"rìkh al-Dawla al-'Abbàsiyya (Cairo, 1975), 36–37; al-Yazbekì, Al-Wizàra, nash"atuhà wa-ta†awwuruhà fì l-dawla al-'Abbàsiyya (Mosul, 1976), 106–111, esp. 106–107; N.D. Nicol, Early 'Abbàsid Administration in the Central and Eastern Provinces, 132–218/750–833 A.D. (Ph.D. diss., The University of Washington, 1979), 210–12; I.S. al-Karawì, NiΩàm al-Wizàra fì l-'aßr al-'Abbàsì al-awwal (Kuwayt, 1983), 5–6, 28 and 39–40; N.D. Íaqar, Al-Silm fì l-'alàqàt al-'Abbàsiyya al-Bìzan†iyya fì l-'aßr al-'Abbàsì al-awwal (Mecca, 1406/1985), 36–38; W. al-Khàzin, Al-Óa∂àra al'Abbàsiyya (Beirut, 1984), 19 and 25–26; F. Omar, Al-'Iràq wa-l-Ta˙addì al-Fàrisì (Baghdàd: Dàr al-Shu"ùn al-Thaqafiyya, 1987), 63–64 and 65–66; id., Al-Judhùr alta"rìkhiyya li-l-wizàra al-'Abbàsiyya (diràsa ta˙lìliyya naqdiyya li-qirà"at surdìl 'an al-wuzarà" al-'Abbàsiyyìn wa-ab˙àth ukhrà) (Baghdàd: Dàr al-Shu"ùn al-Thaqafiyya, 1989), 151–66. 105 Gabrieli, “Hàrùn ar-Rashìd,” 348, 353, 359.

MAWÀLÌ

in the composition of al-ma"mùn’s army

307

ues), when scrutinized in detail, there are many contradictions and disagreements with the general scheme. With regard to al-Ma"mùn, Gabrieli believes that the Persian background in which he lived, his advisors, the Persian population’s affection for him, his Persian mother—all these caused his Persian tendency; at the same time, with regard to al-Amìn he is more cautious, claiming that his Arabness is dubious.106 Gabrieli’s conclusions as to al-Amìn’s army is rooted in his understanding of the character of ahl Khuràsàn and the abnà", which he treats as Persian units, and additional units, al-Afàriqa (from Ifrìqiya) and a commander whom he mistakenly (it seems) identified as Persian.107 It may be assumed that if Gabrieli were to agree to the accepted thesis of the mainly Arab character of ahl Khuràsàn and the abnà", he would renounce his reservation about al-Amìn’s army. The reservation of Gabrieli’s caused some scholars to be cautious in defining these blocs. Yet, (as already stated) these definitions too are ill defined and inaccurate. Thus Watt recreates the old Persian and Arab blocs as the autocratic bloc as opposed to the constitutional one. The Barmakids and the class of secretaries of Iranian origin belonged to the autocratic bloc. In the eastern part of the empire this bloc was very powerful, and its supporters had strongly Shì'ite tendencies. The constitutional bloc’s center was with al-Amìn in Baghdàd, Iraq and part of the west. The transmitters of traditions, the Sunnì 'Ulamà" formed an important part of this bloc. Watt is aware that these definitions of the blocs are not necessarily accurate and that this is an initial attempt: The blocs are not homogeneous, and they are composed of different groups with different tendencies.108 This view of Watt has been accepted by Steppat.109 Shaban describes how two blocs developed during the struggle between al-Amìn and al-Ma"mùn. One was centered on the abnà" and the 106

Ibid., 395–96; id., “al-Ma"mùn,” 29ff. Gabrieli, “Hàrùn ar-Rashìd,” 349–350, 352 note 4, 365 (ahl Khuràsàn); 384 (al-Afàriqa); 383 (the Persian commander?). 108 Watt, “The Political Attitude of the Mu'tazilah,” Journal of the Royal Asiatic Society (1963), 43–46; this idea was also raised by Watt in his Islamic Thought (Edinburgh, 1973), 173–79, esp. 175–76. 109 F. Steppat, “From 'Ahd Arda-“ìr to al-Ma"mùn: A Persian Element in the Policy of the Mi˙na,” in Wadàd al-Qà∂ì et al., eds., Studia Arabica and Islamica: Festschrift I˙sàn 'Abbàs on His Sixtieth Birthday (Beirut: American University of Beirut, 1981), 454, who on the basis of the definitions of Watt, further develops the arguments in favor of the pro Iranian tendencies of al-Ma"mùn. 107

308

amikam elad

other on the easterners. Shaban calls the abnà" the reactionary bloc and the others—the progressive bloc.110 Mottahedeh argues that the “Khuràsànians accepted al-Ma"mùn as one of their own and, because his mother was an Iranian, called him ‘son of our sister.’”111 The conquest of Iraq by al-Ma"mùn was, according to Mottahedeh, a Khuràsànì conquest which was accomplished “by an army much more clearly Iranian than the army of Abù Muslim.”112 Crone depicts the antagonism between al-Amìn and al-Ma"mùn, as a struggle between the Khuràsànìs and the abnà": “The Khuràsànìs defeated the Abnà": in 813 as in 750 a caliph and his Sunnì notables were swept away by Persian-speaking troops.”113 Kennedy (as Crone before him),114 emphasizes the successful contacts of al-Ma"mùn “among the native aristocracy of Khuràsàn,”115 the importance and high position given to the dihqàns in Khuràsàn by al-Ma"mùn, and the victory of the latter with armies from Khuràsàn.116 Amabe argues in the same vein, and depicts (following Crone and Kennedy?) the war between the two brothers as “a war between the lords and the abnà" l-dawla” and stresses the meaningful ethnic difference between al-Amìn and al-Ma"mùn.117 The victory of al-Ma"mùn was “a complete victory for the princes and lords of Khuràsàn and central Asia.”118 Scholars have accepted, therefore, some of Gabrieli’s claims, mainly with regard to the definition of al-Ma"mùn’s Persian-ness: his maternal origin, the Iranian background in Khuràsàn, and his Iranian supporters and advisors. One of the central justifications for scholars’ assertion for al-Ma"mùn’s pro-Iranian tendency is the fact of his Iranian origin on his mother’s side. Scholars claim that due to this, 110

Shaban, History, II, 41–42. Mottahedeh, “The 'Abbàsid Caliphate,” 72. 112 Ibid., 73. 113 Crone, Slaves, 76, 77 (and note 602 on p. 257): “the delight in Ma"mùn’s Iranian mother” (following Gabrieli). 114 Crone, Slaves, 75–76. 115 Kennedy, The Early 'Abbàsid Caliphate, 184. 116 Ibid., 187; id., The Prophet and the Age of the Caliphates. The Islamic Near East from the Sixth to the Eleventh Century (London: Longman, 1986), 152: “The fundamental cause of the conflict lay in the policy of al-Fa∂l b. Sahl and the influence he exercised over al-Ma"mùn. Al-Fa∂l had forged the alliance between al-Ma"mùn and the leading Khuràsànìs which had ensured success in the struggle with al-Amìn.” 117 Amabe, 'Abbàsid Autocracy, 96, 99. 118 Ibid., 46. 111

MAWÀLÌ

in the composition of al-ma"mùn’s army

309

he was well received by the Iranian population of Khuràsàn. The most important and crucial evidence for this is the epithets by which al-Ma"mùn was called while in Khuràsàn. These epithets appear in the tradition we have discussed at length, on al-Ma"mùn’s first actions in Khuràsàn, immediately after his father’s death. In this text, alFa∂l b. Sahl reminds al-Ma"mùn that he is in the midst of his maternal uncles (akhwàl ), who have sworn allegiance to him. It is clear to scholars that al-Fa∂l is referring to al-Ma"mùn’s mother, Maràjil, who had been an Iranian slave, on whom the Iranians of Khuràsàn pride themselves. The expression is equivalent to the phrase “the son of our sister” (ibn ukhtinà), also appearing in the above mentioned tradition.119 The first question that arises here is: What oath of allegiance is referred to? This naturally leads to the next question: who are the akhwàl (maternal uncles) who gave the bay'a to al-Ma"mùn? Let us reconstruct the historical context: al-Ma"mùn has just arrived in Khuràsàn, and the tradition is reported immediately after his attempt to draw that part of the army moving with al-Fa∂l b. alRabì' towards Baghdàd into his camp, had failed. He is not a caliph. Al-Ma"mùn will be appointed caliph officially only five years later. If so, to what bay'a is the transmitter referring? A most plausible possibility is that this is the oath of allegiance that was given already in 186/802 in Mecca to al-Amìn as caliph and to al-Ma"mùn as heir apparent. If so, who are those that are addressed as his maternal uncles? In Mecca the members of the 'Abbàsid family swore allegiance to al-Amìn and al-Ma"mùn, as did the senior commanders of the abnà".120 Indeed, the formulae of the appointment of alAmìn and al-Ma"mùn was sent to the provincial governors, who were responsible for the bay'a in their provinces,121 but those who swore allegiance to them in Khuràsàn were the high-ranking officers, the tribal leaders and the local notables. Even if we assume that a

119 For Gabrieli, Mottahedeh, Daniel, Crone, Amabe and Fishbein it is indeed the natural conclusion to be deduced from this text, see above notes 107, 111, 112, 114, 116–18. 120 See notes 22 and 29 above. Those who did not attend the bay'a in Mecca swore allegiance to al-Amìn and al-Ma"mùn in Baghdàd, under the strict supervision of Harthama b. A'yan, who was ßà˙ib al-˙aras of al-Rashìd, see al-ˇabarì, Ta"rìkh, III, 667, ll. 1–4. 121 Al-ˇabarì, Ta"rìkh, III, 663–66 (the text of the epistle to the governors; on p. 666, ll. 6–8: commanding the governor to read the will and make it well-known).

310

amikam elad

few of these notables were non-Arabs nobles, descendants of the kings of local dynasties on the borders of Khuràsàn and Transoxania, it is difficult to view them all as single homogeneous group and to call them the caliph’s maternal uncles.122 There is a possibility that the bay'a referred to is in fact a bay'a to al-Ma"mùn as caliph after al-Amìn’s death (or after the defeat of his armies), but we still do not know the identity of the mysterious akhwàl. The tradition becomes clearer as it continues, when al-Fa∂l b. Sahl approaches the abnà" commanders at al-Ma"mùn’s court and describes this in the following words: “I met them in their lodgings and reminded them of the oath of allegiance that they had sworn and of the loyalty to which they were obliged” ( fa-laqìtuhum fì manàzilihim wa-dhakartuhum al-bay'a allàtì fì a'nàqihim wa-mà yajibu 'alayhim min al-wafà").123 This time the identity of the oath-takers is clear. They are the commanders of alabnà" in al-Ma"mùn’s camp. What bay'a is referred to? It seems to refer to the bay'a to al-Ma"mùn as heir apparent and governor of Khuràsàn. From another tradition we learn that immediately after his father’s death, al-Ma"mùn gathered the army and all the notables (al-jund wa-sà"ir al-wujùh) in Marw, ascended the minbar and said: “O ahl Khuràsàn [i.e., the abnà" army that had come with him from ˇùs], renew your oath of allegiance to your imàm al-Amìn” ( yà ahl Khuràsàn jaddidù al-bay'a li-imàmikum al-Amìn).124 It is highly reasonable that allegiance was sworn in that ceremony to al-Ma"mùn as heir apparent as well. In any case, this is clearly not an oath taken by the Iranian population, or even by the Iranian nobility. It may be assumed that the bay'a referred to in both texts is the same bay'a. It was sworn once by the akhwàl and twice by the commanders of the army of ahl Khuràsàn and the notables. My proposed solution of the problem is to identify the akhwàl with the commanders and notables, who were mainly of Arab origin, or descended from the mawàlì of Arab tribes. It seems to me to be possible to assume that the mysterious aunt (who subsequently becomes a sister) is not the caliph’s mother, but rather refers to the 'Abbàsid maternal Arab origin, either to the mother of caliph Abù l-'Abbàs 122 It is impossible to assume the existence of a group Khuràsànì national identity in this period. 123 Al-ˇabarì, Ta"rìkh, III, 774, ll. 2–3. 124 Al-Dìnawarì, Al-Akhbàr al-†iwàl, 392–93.

MAWÀLÌ

in the composition of al-ma"mùn’s army

311

al-Saffà˙ (reigned 132/750–136/754), who was an Arab of the tribe of al-Óàrith b. Ka'b,125 and members of this tribe were considered his maternal uncles (khu"ùla/akhwàl ).126 Abù l-'Abbàs al-Saffà˙ is termed Ibn al-Óàrithiyya in traditions of apocalyptic nature.127 Some pieces of evidence bear witness to the importance of members of this tribe, from the beginning of the 'Abbàsid Caliphate until the reign of Hàrùn al-Rashìd, emphasizing the close ties between dignitaries from al-Óàrith b. Ka'b and the Caliphs.128 It is possible, though, that the attribution is to the southern origin of the 'Abbàsids through the grandfather of al-Saffà˙ and al-Manßùr, 'Alì b. 'Abdallàh b. al-'Abbàs, whose mother was from Kinda. A fascinating tradition describes how, in the year 143/760–761, the leaders of a sub tribe of Kinda asked the caliph al-Manßùr to pardon Qays b. Walì'a alKindì, who had participated in 'Abdallàh b. 'Alì’s revolt, on the grounds of their being al-Manßùr’s maternal uncles (ahkwàl ). This, of course, does not refer to al-Manßùr’s own mother, who like alMa"mùn’s mother was not an Arab, but rather was of Berber origin. The tradition cited (and not by chance, by Abù Zakariyyà" al-Azdì) clarifies this point by saying that they meant 'Alì b. 'Abdallàh b. al-'Abbàs’ mother, Zur'a (Zuhra) bt. Mishra˙ al-Kindiyya, and continues with a description of how members of Qays b. Walì'a’s family came to 'Alì b. 'Abdallàh’s aid at the battle of al-Óarra.129 125 Al-Zubayrì, Kitàb nasab Quraysh, ed. E. Levi-Provençal (Cairo: Dàr al-Ma'àrif, 1953), 30, ll. 15–19; al-Balàdhurì, Ansàb (ed. Dùrì), III, 82; Ibn Óazm, Jamharat, 95; al-Ißfàhànì, Aghànì (ed. Bùlàq), XVIII, 208. 126 Al-ˇabarì, Ta"rìkh, III, 76; Ibn al-'Adìm, Bughyat al-†alab fì ta"rìkh Óalab, ed. Suhayl Zakkàr (Damascus: Ma†àbi' Dàr al-Ba'th, 1408/1988), VII, 3050; Elad, “Transition,” 101 note 57. 127 Al-Balàdhurì, Ansàb (ed. Dùrì), III, 86–87; Akhbàr al-'Abbàs, 168, 207; M. Sharon, Black Banners from the East. The Establishment of the 'Abbàsid State-Incubation of a Revolt ( Jerusalem/Leiden: The Magnes Press and E.J. Brill, 1983), 142–43. 128 Ibid., 141–43 (their importance in the da'wa); Elad, “Transition,” 101–102 note 57, 125–26. 129 Al-Azdì, Al-Mawßil, 178–79 (a parallel tradition with small changes, ibid., 232–35); al-Fàkihì, Akhbàr Makka, ed. 'Abd al-Malik b. 'Abdallàh b. Dahìsh (Mecca: Maktabat wa-Ma†ba'at al-Nah†a al-Óadìtha, 1407/1987), II, 354; Elad, 'Abbàsid Army, 110–12; but especially, M. Lecker, “Judaism among Kinda and the Ridda of Kinda,” Journal of the American Oriental Society 115 (1995), 645 (quoting Abù l-Baqà", al-Manàqib al-Mazyadiyya and al-Mas'ùdì’s Murùj, III, 268–69); on Zur'a bt. Mishra˙, see Lecker, loc. cit.; id., “Judaism among Kinda and the Ridda of Kinda,” (Hebrew version) Taymà 6 (1998), 17 (quoting Akhbàr al-'Abbàs, 117, 159; Ibn al-Kalbì, Nasab ma'add wa-l-Yaman al-kabìr, ed. Nàjì Óasan (Beirut: 'Àlam al-Kutub-Maktabat alNah∂a l-'Arabiyya, 1408/1988), I, 175; id., Jamharat al-nasab, ed. W. Caskel (Leiden: E.J. Brill, 1966), 32; Ibn Óazm, Jamharat, 19).

312

amikam elad

Mishra˙ b. Ma'dì Karib was one of the four kings of Kinda.130 It is worth mentioning here that Caliph al-Mahdì’s mother, Umm Mùsà bt. Manßùr, was also an Arab from a southern (Óimyar) tribe131 (see the discussion further below). The above-mentioned tradition about Qays b. Walì'a al-Kindì, bears witness to the existence and involvement of high rank Yamanì ashràf at the court of caliph al-Manßùr. It reflects the special relationship of the 'Abbàsid state to the southern tribes, who formed the backbone of the military force of ahl Khuràsàn, which had brought down Umayyad rule.132 A rare and most interesting tradition on this matter (although no doubt exaggerated) also bears evidence of a definite policy of preferring the southern tribes on al-Manßùr’s part.133 These tribes, on their part unequivocally supported the problematic appointment of al-Manßùr’s son, al-Mahdì as heir to the throne134 (to which we will return below). Returning to the akhwàl, again, I wish to raise the possibility that the akhwàl mentioned in the tradition are the senior commanders of the abnà" in al-Ma"mùn’s army, most of whom (with the exception of Nu'aym b. Khàzim al-Tamìmì and al-'Abbàs b. al-Musayyab b. Zuhayr, al-Îabbì) are either members of southern tribes or mawàlì of those tribes, plus one from Rabì'a (Ya˙yà b. Mu'àdh b. Muslim al-Dhuhlì), which in Khuràsàn was considered part of the Yamanì confederation due to the covenant (Óilf ) with al-Azd.135 The section most quoted by scholars and which has become a classic of research on the 'Abbàsid caliphate is the sentence in which al-Ma"mùn is called “the son of our sister and the son of the Prophet’s (paternal) uncle.” First, we should ask who says this (according to al-ˇabarì). The scholars have no doubt that this was said by the non-Arab people of Khuràsàn, who supported al-Ma"mùn due to his Iranian origin on his mother’s side.136

130 Ibn Óazm, Jamharat, 428; al-Zubayrì, Nasab Quraysh, 28–29 (Zur'a); Akhbàr al'Abbàs, 117–18; Lecker, loc. cit. 131 Al-Balàdhurì, Ansàb (ed. Dùrì), 275; Ibn Khayyà†, Ta"rìkh, II, 668; Elad, “Transition,” 112. 132 Elad, “Transition,” 103 note 60. 133 Al-Azdì, Al-Mawßil, 218–22, esp. 222. 134 Al-Azdì, Al-Mawßil, 214. 135 For references, see Elad, “Golan,” 53 notes 87–88. 136 J. Zaydàn, Ta"rìkh al-Tamaddun al-Islàmì (Cairo, 1911), IV, the first page of the introduction; Bartold, Turkestan, 207; Gabrieli, “Hàrùn ar-Rashìd,” 359: al-Ma"mùn

MAWÀLÌ

in the composition of al-ma"mùn’s army

313

This view of al-Ma"mùn and his reign was constructed not only on the basis of the feelings and emotions of modern scholars, who lived under the influence of the national movements of the nineteenth century. This view seems to have basis in reality. Beyond the various Weltanschauungen and ingrained conceptions that infiltrate historical writing (very often unconsciously and unknown to the writer), these views, or part of them (naturally, not in modern nationalistic garb) could be found already among the medieval chroniclers. In the eleventh century, or even earlier, the anonymous author of al'Uyùn wa-l-˙adà"iq could change the sentence “And they said, the son of our sister,” and add “And the Persians (al-Furs) said, the son of our sister.” The scholars have not dealt with the entire text nor with the context in which this sentence appears. Al-ˇabarì describes the honor paid to the commanders, the kings and the sons (descendants) of kings, and then comes to the direct appeal to the Arab tribes, after which we have the quoted sentence “And we reduced Khuràsàn’s kharàj tax by a quarter.” This policy (situation?) pleased them; they rejoiced in him (or it?) (wa-˙asuna mawqi' dhàlika minhum wa-surrù bihi ) and said “the son of our sister etc.” Al-ˇabarì does not say explicitly who calls al-Ma"mùn “the son of our sister.” By reducing the land tax, al-Ma"mùn pleased the landowners, holders of small and large estates. They were the ones who paid the tax directly to the government. The peasants paid their taxes to the estate owners.137 Al-Ma"mùn was not interested in gaining the support of the Iranian population. He wanted to gain that of the notables of Khuràsàn who were Arabs and mawàlì of Arab tribes, of the senior commanders (most of whom belonged to the should have felt secure on the firm ground, on which he stepped, between his loyal supporters in Khuràsàn. For them, continues Gabrieli, he was the son of our sister . . . all the eastern part of the Caliphate stood at the side of the son of Maràjil in his struggle against the plots of Baghdàd; see also ibid., 396; Mottahedeh, 'Abbàsid Caliphate, 72; Daniel, Khuràsàn, 178: “The people of Khuràsàn were delighted with al-Ma"mùn. They thought of him as their “nephew;” Crone, Slaves, 75, 77, 257 (quoting Gabrieli); Kennedy, The Early 'Abbàsid Caliphate, 101, 184, in this vein, although with some reservations; Amabe, 'Abbàsid Autocracy, 96 (possibly relying on this text), stresses the fact that al-Amìn’s mother was of Arab origin, while alMa"mùn, whose mother was the daughter of Khuràsànì lord, possibly Ustàdh Sìs, gained the support of these Khuràsànì lords due to this fact; Fishbein, War Between Two Brothers, 16 note 70, 18 note, 79, arguments in this vein. 137 F. Løkkegaard, Islamic Taxation in the Classical Period, with Special Reference to Circumstances in Iraq (Copenhagen, 1950), 135, 139; Cl. Cahen, art. “Îay'a,” EI 2, 187–88.

314

amikam elad

abnà") and of the (non-Arab) kings and descendants of kings. (It seems that these are not the dihqàns, the members of the lower nobility of Khuràsàn.) It should be remembered that in Khuràsàn (and other regions too),138 the Arab tribesmen held estates. Many Arabs lived in the villages.139 The kharàj tax was levied from them from the earliest periods, during the time of 'Abd al-Malik and towards the end of Umayyad rule,140 and most probably this situation continued under 'Abbàsid rule as well. We remember that 'Alì b. 'Ìsà, who governed Khuràsàn between 180/796–191/807) was accused of tyranny, harsh and unjust conduct against the prominent and leading men of Khuràsàn (ashràf, wujùh, kubarà").141 But according to another tradition, this unjust conduct was directed against the Arabs of Khuràsàn.142 The

138 The case of Banù Ma'qil who had estates in the Jibàl district (which was part of the district of Ißfahàn), see A. Elad, “Two Identical Inscriptions from Jund Filas†ìn from the Reign of the 'Abbàsid Caliph al-Muqtadir,” Journal of the Economic and Social History of the Orient 35 (1992), 317ff. Idrìs and 'Ìsà the sons of Ma'qil were imprisoned by the governor of Iraq, Yùsuf b.'Umar “on account of the kharàj” (bisabab al-kharàj, see al-Balàdhurì, Ansàb (ed. Dùrì), III, 118–19; quoted by Daniel, Khuràsàn, 103 with additional sources; cf. Akhbàr al-'Abbàs, 255: bi-l-kharàj ). 139 M. Shaban, The 'Abbàsid Revolution (Cambridge: Cambridge University Press, 1970), XV, 155–56, 158; Sharon, Black Banners, 65–71; id., Revolt. The Social and Military Aspects of the 'Abbàsid Revolution ( Jerusalem: The Max Schloessinger Memorial Fund, The Hebrew University, 1990), 78, 87; contrary to E. Daniel, “The Ahl alTaqàdum and the Problem of the Constituency of the 'Abbàsid Revolution in the Merv Oasis,” Journal of Islamic Studies 7 (1996), 156ff., who argues that (although some Arabs, were owners of estates in Khuràsàn, see E. Daniel, Khuràsàn, 21–22) the villages mentioned in the sources are not necessarily inhabited by Arabs. “An Arab village in Khuràsàn was one owned by an Arab tribe, which is not necessarily the same thing as being inhabited by Arabs,” Daniel, “Arabs, Persians,” 546; id., “Ahl al-Taqàdum,” 165 note 43 (quoting J. Wellhausen, The Arab Kingdom and its Fall (Beirut, 1963), 528 note 3); see also Daniel, Khuràsàn, 50; Crone, “Abnà",” 13; but is seems that Arabs in Khuràsàn did live in and inhabited the villages around Marw; for a detailed discussion, see Elad, “Ethnic Composition,” 47–53. 140 Evidence from 'Abd al-Malik’s reign: al-ˇabarì, Ta"rìkh, III, 1029, quoted and discussed by Shaban, The 'Abbàsid Revolution, 47, 92; the end of the Umayyad period, the governorship of Naßr b. Sayyàr, Wellhausen, The Arab Kingdom, 480 note 1, 481; Shaban, The 'Abbàsid Revolution, 130; Daniel, Khuràsàn, 195–97; and see also D. Dennett, Al-Jizya wa-l-Islàm, Ar. trans. Fawzì Fahìm Jawdallàh (Beirut/New York, 1960), 191–96. 141 Al-ˇabarì, Ta"rìkh, III, 702–703, quoted and discussed in Arazi and Elad, “L’Épître à l’armée,” 29–31; Daniel, Khuràsàn, 176, 197. 142 Al-Dìnawarì, Al-Akhbàr (ed. Cairo), 391, who attributes the rebellion of Ràfi' b. al-Layth to the oppression of 'Alì b. 'Ìsà; see discussion of the text in Arazi and Elad, “L’Épître à l’armée,” 31; also the translation and the important comprehensive notes of C.E. Bosworth in his translation to al-ˇabarì, The 'Abbàsid Caliphate in Equilibrium (New York: State University of New York, 1989), 250–253 ('Alì b. 'Ìsà), 261 (the revolt of Ràfi' b. al-Layth).

MAWÀLÌ

in the composition of al-ma"mùn’s army

315

reduction in the taxes by al-Ma"mùn made things easier for them as well and pleased them, so they too could say “the son of our sister.” At the same time, it is difficult to see Arab tribesmen or notables who were mawàlì of Arab tribes, such as ˇàhir b. al-Óusayn and other distinguished members of his family, calling al-Ma"mùn “the son of our sister” and meaning by that the son of a Persian/Iranian (slave). Why would they want to praise him by such descent, when it was a curse and an insult to call an Arab ibn al-fàrisiyya?143 It would seem that the Khuràsànì tribesmen, the commanders and the notables, are referring here to the well-known ˙adìth “ibn ukht alqawm minhum (or min anfusihim).”144 This ˙adìth expresses one of the most important and basic values of Arab society. Ibn ukht is identical in meaning to khàl.145 Many examples of the widespread use of maternal affinity during the first two centuries of Islamic rule can be cited.146 143

I. Goldziher, Muslim Studies (London: George Allen and Unwin, 1967), I, 120. A.J. Wensinck, Concordance et Indices de la Tradition Musulmane (Leiden: E.J. Brill, 1933, 1963), a.kh.w. s.v.; 'Abd al-Razzàq, Al-Mußannaf, ed. Óabìb al-Ra˙màn alA'Ωamì (Beirut: al-Maktab al-Islàmì, 1403 AH), II, 63; Abù Ya'là, Musnad Abì Ya'là, ed. Óusayn Salìm Asad (Damascus. Dàr al-Ma"mùn li-l-Turàth, 1404/1984), V, 356, 475; VI, 12; VII, 172; al-ˇabàrànì, Al-Mu'jam al-ßaghìr, ed. Mu˙ammad ShakùrMa˙mùd al-Óajj al-Marìr (Beirut/Amman, al-Maktab al-Islàmì-Dàr 'Ammàr, 1405/1985), 142; id., Al-Mu'jam al-kabìr, ed. Óamdì b. 'Abd al-Majìd al-Salafì (Mosul: Maktabat al-'Ulùm wa-l-Óikam, 1404/1983), II, 136; III, 280, 281; XII, 170, 196; XVII, 12, 118; Ibn Óibbàn, Ía˙ì˙ Ibn Óibbàn, ed. Shu'ayb al-Arnà"ù† (Beirut: Mu"assasat al-Risàla, 1414/1993), X, 353 (min anfusihim); XVI, 258. 145 Lisàn al-'Arab, IV, 409 (kh.w.l., s.v.); Ibn 'Asàkir, Ta"rìkh madinat Dimashq, ed. 'Umar b. Gharàma al-'Amrawì (Beirut: Dàr al-Fikr li-l-ˇibà'a wa-l-Nashr wa-lTawzì', 1415/1995), XXXVII, 440. 146 See for instance, al-Bukhàrì, Ía˙ì˙, ed. D. Mu߆afà Dìb al-Bughà (Beirut: Dàr Ibn Kathìr al-Yamàma, 1407/1987), III, 1291, 1363; Muslim, Ía˙ì˙, ed. Fu"àd 'Abd al-Bàqì ([Cairo]: Dàr I˙yà" al-Kutub al-'Arabiyya, 1375/1955), IV, 2311; Ibn Óibbàn, Ía˙ì˙, XIV, 190; XV, 289; Ibn Abì Shayba, Mußannaf ibn Abì Shayba, ed. Kamàl Yùsuf al-Óùt (Riyà∂: Maktabat al-Rushd, 1409 AH), VII, 344, 444; al-ˇabarì, Ta"rìkh, I, 2437, year 16: the conquest of al-Madà"in; al-Dhahabì, Siyar a'làm alnubalà" (Beirut: Mu"ssasat al-Risàla, 1982–85), III, 161; IV, 36: Yazìd b. Mu'àwiya and his akhwàl from Banù Kalb; al-ˇabarì, Ta"rìkh, II, 893, ll. 10–11, 14–15, year 76: the concept among the Khawàrij; Abù Nu'aym, Óilyat al-awliyà" wa-†abaqàt alaßfiyà" (Cairo, 1351 AH), V, 308, l. 17; 309, l. 3: 'Umar b. 'Abd al-'Azìz approves the renovations in the mosque of Banù 'Adì b. al-Najjàr, akhwàl Rasùl Allàh, in alMadìna; al-ˇabarì, Ta"rìkh, II, 1286, l. 8, year 96: the death of Qutayba b. Muslim and his brothers; one of the brothers is saved due to the khu"ùla; 1338, l. 17, year 99: Banù 'Abs and Sulaymàn b. 'Abd al-Malik (see further in this note); 1669, ll. 12–14, year 121: the rebellion of Zayd b. 'Alì, the governor Yùsuf b. 'Umar does not harm Ayyùb b. Salama due to his maternal relationship to Caliph Hishàm [Ibn al-'Adìm, Bughyat al-†alab, IX, 4036]; al-ˇabarì, Ta"rìkh, III, 76, ll. 6, 12–13, 77, l. 3, year 134; al-Balàdhurì, Futù˙ al-buldàn, ed. Mu˙ammad Ri∂wàn (Beirut: Dàr 144

316

amikam elad

I will conclude the discussion at this point, while returning to caliph al-Manßùr and his son al-Mahdì, to a tradition reflecting the support of the southern tribes for al-Mahdì’s appointment as heir apparent. The tradition, which appears in al-Azdì’s book, is transmitted by al-Madà"inì, who relates that in a public dispute between Ismà'ìl b. 'Abdallàh [b. Yazìd. b. Asad] al-Qasrì (from Bajìla)147 and a member of the 'Abbàsid family, the 'Abbàsid taunted Ismà'ìl and asked him: “When will your southern messiah appear” (matà yaΩharu qa˙†ànukum).148 Ismà'ìl replies: “Indeed he has already appeared, and I expect him to ride on your neck and the necks of those like you tomorrow, for he is al-Mahdì, the heir apparent of the Muslims, the son of the Commander of the Faithful, the son of our sister (ibn ukhtinà), for the Messenger of Allàh (Í ) has said: ‘the tribes’ sister’s son is one of them (ibn ukht al-qawm minhum)’.” According to al-Azdì, this episode took place in 151/768–769, and following it, Ismà'ìl b. 'Abdallàh al-Qasrì was appointed governor of Mosul by al-Manßùr.149 Ismà'ìl b. 'Abdallàh, the Yamanì considered himself related to alManßùr and al-Mahdì. This is possibly due, as noted above, to the mother of 'Alì b. 'Abdallàh b. al-'Abbàs, or perhaps to the mother of Abù l-'Abbàs al-Saffà˙, or to the mother of al-Mahdì himself, Umm Mùsà bt. Manßùr al-Óimyariyya.150

al-Kutub al-'Ilmiyya, 1403 AH), I, 79 (Abù al-'Abbàs al-Saffàh and his maternal uncles from al-Óàrith b. Ka'b); al-Ya'qùbì, Ta"rìkh (ed. Beirut), II, 398 (al-Mahdì and Banù Tanùkh, who seek the Caliph’s grace through their khu"ùla); al-Fàkihì, Akhbàr Makka, III, 321, 347; Ibn 'Asàkir, Ta"rìkh (ed. Beirut), XXXIII, 211; LVI, 456; Ibn al-Athìr, Al-Kàmil (ed. Beirut), IV, 397 (year 76); V, 157 (Banù Sulaym and Mu˙ammad b. 'Abdallàh b. al-Óasan, al-Nafs al-Zakiyya); the well-known families of al-'Abbàs and Khulayd b. Jaz" b. al-Óàrith . . . b. 'Abs. b. Gha†afàn. b. Qays 'Aylàn, were blood-relations to al-Walìd and Sulaymàn b. 'Abd al-Malik. This bloodrelationship through the mother, paved the way for the 'Absìs to the caliph’s court and to high and prominent positions and wealth, see Elad, “Golan,” 59–63. 147 On him, see Crone, Slaves, 102–103; al-ˇabarì, Ta"rìkh, II (Index): one of the ashràf, who took an active part in the political events during the end of the Umayyad period; ibid., 1881ff., 1902: the governor of Kùfa (Crone, loc. cit.); III, 66; 402: in the court of al-Manßùr; Ibn 'Asàkir, Ta"rìkh (ed. Beirut), VIII, 427: one of the ßa˙àba of al-Manßùr; id., 428 and Ibn al-'Adìm, Bughyat al-†alab, IV, 1667: one of the ßa˙àba of al-Manßùr and the governor of Mosul. 148 Within this framework there is no space to discuss the southern messiah, the Qa˙†ànì. 149 Al-Azdì, Al-Mawßil, 214; see also the tradition according to which Banù Tanùkh receives Caliph al-Mahdì with gifts, saying: “we are your maternal cousins (akhwàluka), ho Amìr al-Mu"minìn,” al-Balàdhurì, Futù˙ (ed. Ri∂wàn), I, 79. 150 See note 131 above.

MAWÀLÌ

in the composition of al-ma"mùn’s army

317

Summary: Outlines of the development of al-Ma"mùn’s army A comprehensive and general study of al-Ma"mùn’s army does not exist. Thus even a partial discussion of this subject, like the discussion here, is not easy, as almost every step taken sets up upon virgin soil, that has not been studied. Despite this, students of the period of al-Ma"mùn’s reign are united in the opinion that the major part of his army was composed of non-Arab Khuràsànìs, mainly of Iranian origin.151 Scrutiny of the political and social background in Khuràsàn under al-Ma"mùn’s rule reveals, however, that this description of al-Ma"mùn’s armies and commanders is imprecise and one-dimensional. An examination of this army, its mobilization, consolidation and battles, from al-Ma"mùn’s arrival in Marw in 193/809 until his death in 218/833, provokes interesting conclusions that transform the accepted picture of al-Ma"mùn’s activity in Khuràsàn and of the characteristics of the armies he raised in this province. One of the central arguments of scholars aiming to establish the above thesis is the fact (obvious in their opinion) that al-Ma"mùn’s two main commanders during the battles against al-Amìn’s army and during the siege of Baghdàd, ˇàhir b. al-Óusayn and Harthama b. A'yan, were Iranians, local leaders connected to the dihqàns and mainly to non-Arab notables of Khuràsàn,152 and to local representatives of the Khuràsànìs who joined al-Ma"mùn’s cause, and fought against the abnà" and the western caliphate (on this see below). An important point that may open discussion of this issue is the fact that al-Ma"mùn arrived in Khuràsàn with the most important commanders of the abnà" and their forces. A second point, unknown to research (up to Arazi and Elad’s initial study), is that al-Ma"mùn began to organize a second da'wa (da'wa thàniya) in Khuràsàn, in which he tried to incorporate the Arab tribes of Khuràsàn by using 151 Only a small part of the studies will be mentioned here, Zaydàn, Umayyads and 'Abbàsids . . ., trans. D.S. Margoliouth (Leiden, 1907), 206; Bartold, Turkestan, 207; 'Abd al-Mun'im, Al-'Aßr al-'Abbàsì al-awwal. Al-qarn al-dhahabì fì ta"rìkh al-khulafà" al-'Abbàsiyyìn/al-ta"rìkh al-siyàsì (Cairo, 1973), I, 297–98; Ayalon, “al-Mu'taßim,” 25–26; Shaban, History, II, 43–44; Mottahedeh, “The 'Abbàsid Caliphate,” 73; Amabe, 'Abbàsid Autocracy, 96, 99, 102–103; Crone, Slaves, 76; Kennedy, The Early 'Abbàsid Caliphate, 187; Pipes, Slaves, 180. 152 See the references in the preceding note (except Ayalon), esp. Crone, loc. cit.; Shaban, loc. cit.; and Kennedy, The Early 'Abbàsid Caliphate, 139 (ˇàhir b. al-Óusayn); Pipes, loc. cit.; Nicol, Early 'Abbàsid Administration, 154, 156–58, 189–90.

318

amikam elad

the names of the nuqabà" (the supreme body of the propagandists) of the first da'wa who had belonged to these tribes, as well as the kings and descendants of kings, i.e., the high Iranian nobility, presumably the princes of the non-Arab kingdoms on the borders of Khuràsàn and Transoxania, just as was done in the first da'wa, more than sixty years before. In the second da'wa there was a hierarchy. Its notables were called al-Shì'a or Shì'at Amìr al-Mu"minìn, while Khuràsàn was called dàr al-da'wa. The highest rank within this da'wa were those notables, mainly Arabs, who belonged to both the first and the second da'was. In all probability, al-Ma"mùn’s army included non-Arab units, called 'ajam or 'ajam ahl Khuràsàn, but references to them are very few. We have not found evidence of massive mobilization of nonArab soldiers from Khuràsàn or beyond it, in Transoxania. From a single source we learn that while in Khuràsàn, and even in Iraq, alMa"mùn acted to subdue remote regions of Khuràsàn and beyond, while attempting to mobilize the local nobility, not through force or coercion, but rather through persuasion and enticement. This single source does not permit us to draw such far-reaching conclusions as some scholars have. Other very few pieces of evidence refer to this or that non-Arab commander or of a unit of non-Arab soldiers.153 Two pieces of evidence bear witness to relatively large non-Arab units in al-Ma"mùn’s army. The first is the non-Arab units in ˇàhir b. al-Óusayn’s army, units of Turks, Bukhàrìs and Khwàrizmìs.154 From the second evidence we learn of slave soldiers’ units (Mamlùks), most of whom were purchased by al-Mu'taßim already during his brother’s reign. They numbered between 3,000 to 4,000 soldiers.155 In 213/828 a Mamlùk regiment of 4,000 soldiers is mentioned in al-

153 The single evidence: al-Balàdhurì, Futù˙ al-buldàn, ed. M.J. De Goeje (Leiden, 1866), 430–31; Bartold, Turkestan, 212; Katàbjì, Al-Turk fì mu"allafàt al-Jà˙iΩ wamakànatuhm fì l-ta"rìkh al-Islàmì ˙attà awàsi† al-qarn al-thàlith al-hijrì (Beirut, 1972), 113– 114, 123; Ismà'ìl, “al-Mu'taßim,” 14; Ayalon, “al-Mu'taßim,” 26; Pipes, Slaves, 180. 154 Turks: al-ˇabarì, Ta"rìkh, III, 799; Khwàrizmìs: ibid., 801: at least 700 soldiers [!]; al-Mas'ùdì, Murùj, IV, 263; al-Bukhàriyya: al-ˇabarì, Ta"rìkh, III, 800, 802; the important and unique tradition in al-ˇabarì is related from A˙mad b. Hishàm, who was most probably the cousin of ˇàhir b. al-Óusayn (see Elad, 'Abbàsid Army, 227–28); he was ßà˙ib al-shur†a of ˇàhir’s camp (see al-ˇabarì, Ta"rìkh, III, 799–802; according to al-Mas'ùdì, Murùj, IV, 263–65, he is one of the senior commanders (min wujùh al-quwwàd )). 155 Ayalon, “al-Mu'taßim,” 26; Ismà'ìl, “al-Mu'taßim,” 14; Pipes, Slaves, 146–47; Lassner, Shaping, 113.

MAWÀLÌ

in the composition of al-ma"mùn’s army

319

Mu'taßim’s army in Egypt. This is the first time that a slave military unit in such large scale is mentioned.156 We have a few additional references to some commanders, apparently with non-Arab units, but examination of the names of al-Ma"mùn’s senior commanders who came from Khuràsàn to Baghdàd before, and after him, shows that the vast majority of them were not Iranians; rather, they were mostly Arabs. This is particularly striking in the army of Óumayd b. 'Abd al-Óamìd and his son Mu˙ammad. The view held by past and present scholars,157 that ˇàhir b. al-Óusayn and Harthama b. A'yan, were the eminent representatives of the Iranian nobility of Khuràsàn has no basis in fact. The close ties of ˇàhir’s family to the first 'Abbàsid da'wa is what raised them to prominence in the period before al-Ma"mùn and during his reign.158 The causes for the prominence of the families of ˇàhir, Óumayd, and probably also that of Harthama b. A'yan,159 should also be sought in the historical circumstances in Khuràsàn itself from the time of al-Ma"mùn arrived there in 193/809. The top rank in the second da'wa was composed of those propagandists from ahl Khuràsàn whose ancestors had participated in the first da'wa. The combination of both da'was gave a person the most senior and highest rank in the hierarchy of al-Ma"mùn’s second da'wa in Khuràsàn. ˇàhir and Óumayd, and most probably also Harthama b. A'yan (although evidence for his participation in the first da'wa is equivocal), belonged to the top rank of the second da'wa. After the end of the siege of Baghdàd, and especially immediately before and after al-Ma"mùn’s entry into the city (204/819), the abnà" army that had served al-Amìn, and also other commanders and units, for the most part of Arab origin, became integrated into al-Ma"mùn’s army. This and other pieces of evidence reveal the importance of the veteran units of the 'Abbàsid army, al-abnà", in the army of the caliphate even after al-Amìn’s death. What is the reason for the far-reaching assertion in research regarding the non-Arab make-up of al-Ma"mùn’s army? It seems that the mere fact that al-Ma"mùn 156

167.

Pipes, Slaves, 51; Lassner, loc. cit.; see also Kennedy, The Early 'Abbàsid Caliphate,

157 Shaban, The 'Abbàsid Revolution, 43; Crone, Slaves, 75, 78; Kennedy, The Early 'Abbàsid Caliphate, 138–39; Nicol, Early 'Abbàsid Administration, 154, 156–58, 189–90, 212. 158 Arazi and Elad, “L’Épître à l’armée,” (second part) 62, 66–70; Elad, 'Abbàsid Army, 169–79, 185–86. 159 See Elad, “Golan,” 57–58 note 113, arguments in this vein.

320

amikam elad

lived in Khuràsàn and chose it as the center for his governorship, was sufficient for the scholars to argue in favor of this thesis. It is possible that what helped this view develop, at least in the case of ˇàhir, is the fact that the army that fought with him against 'Alì b. 'Ìsà included non-Arab units. This is, as noted above, a rare evidence for the existence of non-Arab military units in al-Ma"mùn’s army. To date there has been no comprehensive and complete analysis of the reports in the sources about al-Ma"mùn’s armies and their commanders, such as the armies of Harthama b. A'yan, Zuhayr b. al-Musayyab al-Abnàwì, Óumayd b. 'Abd al-Óamìd and his son Mu˙ammad (with their Arab commanders), or the armies of 'Alì b. Hishàm, 'Ujayf b. 'Anbasa (who is described as one of abnà" alshì'a),160 and 'Ìsà b. Yazìd al-Jalùdì (al-Amìn’s commander).161 None of these commanders can be considered as belonging to that new non-Arab Khuràsànì army. It is clear that al-abnà" forces and their commanders, who joined al-Ma"mùn’s army, cannot be included in the new non-Arab Khuràsànì army. The picture, then, is not unequivocal and one-dimensional, as it has been described by the various scholars who have depicted al-Ma"mùn’s army. The reports that the Arab’s political and military power had not completely disappeared in various regions of the caliphate; that al-Ma"mùn was required to take them into consideration and to use them as a military force which he mobilized for battle in the different parts of the caliphate;162 that the tribal make-up of this or that area often dictated the caliph’s policy—all these many reports of Arab commanders and their armies during al-Ma"mùn’s days—cannot hide the clear process of the waning of the Arab military and political power in the 'Abbàsid caliphate, which reached its highest point in the days of al-Mu'taßim.163

On 'Ujayf b. 'Anbasa see note 62. On him, see al-ˇabarì, Ta"rìkh, III, 913, 994, 1044; al-Ya'qùbì, Ta"rìkh (ed. Leiden), II, 561; for his nisba, al-Jalùdì, see al-Sam'ànì, Kitàb al-ansàb, (Hyderabad, 1382/1962), III, 306–309: al-Jalùd, a village in Ifrìqiya; Yàqùt, Mu'jam al-buldàn, (Beirut: Dàr Íàdir, 1955–57), II, 156 (al-Jalùdì, s.v.), quoting Ibn Qutayba’s Adab al-kàtib, attributing 'Ìsà b. Yazìd to the small town (balda) in Ifrìqiya. 162 Al-Azdì, Al-Mawßil, 378, 380–82, 386–87 (year 218/833, the army of Mu˙ammad b. 'Abd al-Óamìd, see discussion in Elad, 'Abbàsid Army, 320–24); al-ˇabarì, Ta"rìkh, III, 1112: year 212/827–828. 163 Ayalon, “al-Mu'taßim,” 21–22; Kennedy, The Early 'Abbàsid Caliphate, 165; Pipes, Slaves, 150. 160 161

MAWÀLÌ

in the composition of al-ma"mùn’s army

321

Bibliography Primary Sources 'Abd al-Mun'im, M. Al-'Aßr al-'Abbàsì al-awwal. Al-qarn al-dhahabì fì ta"rìkh al-khulafà" al-'Abbàsiyyìn/al-ta"rìkh al-siyàsì, vol. 1. Cairo, 1973. 'Abd al-Razzàq. Al-Mußannaf 2, ed. Óabìb al-Ra˙màn al-A'Ωamì. Beirut: al-Maktab al-Islàmì, 1403 AH. Abù Nu'aym, A˙mad b. 'Abdallàh, al-Ißfahànì. Óilyat al-awliyà" wa-†abaqàt al-aßfiyà". Cairo, 1351 AH. Abù Zakariyyà" Yazìd b. Mu˙ammad al-Azdì. Ta"rìkh al-Mawßil, ed. 'A. Óabìba. Cairo: Dàr al-Ta˙rìr li-l-ˇab' wa-l-Nashr, 1967. A˙mad b. 'Alì b. al-Muthannà Abù Ya'là al-Mawßilì al-Tamìmì. Musnad Abì Ya'là, ed. Óusayn Salìm Asad. Damascus: Dàr al-Ma"mùn li-l-Turàth, 1404/1984. Akhbàr al-dawla al-'abbàsiyya wa-fìhi akhbàr al-'Abbàs wa-wuldihi, ed. 'Abd al-'Azìz alDùrì and 'Abd al-Jabbàr al-Mu††alibì. Beirut: Dàr Íàdir, 1971. al-Balàdhurì, A˙mad b. Ya˙yà. Ansàb al-ashràf, ed. 'A.'A. Dùrì, vol. 3. Beirut, 1978. id., Futù˙ al-buldàn, ed. M.J. De Goeje. Leiden, 1866; ed. Ri∂wàn Mu˙ammad Ri∂wàn. Beirut. Dàr al-Kutub al-'Ilmiyya, 1403 AH. Bal'amì see al-ˇabarì. The History of al-ˇabarì. Vol. XXX: The 'Abbàsid Caliphate in Equilibrium, trans. and annot. C.E. Bosworth. New York: State University of New York Press, 1989. al-Bukhàrì, Mu˙ammad b. Ismà'ìl. Al-Ía˙ì˙3, ed. D. Mu߆afà Dìb al-Bughà. Beirut: Dàr Ibn Kathìr-al-Yamàma, 1407/1987. al-Dhahabì, Shams al-Dìn, Mu˙ammad b. A˙mad. Siyar a'làm al-nubalà". Beirut: Mu"ssasat al-Risàla, 1982–1985. al-Dìnawarì, Abù Óanìfa A˙mad b. Dàwùd. Kitàb al-akhbàr al-†iwàl, ed. V. Guirgass. Leiden: Brill, 1888; ed. Jamàl al-Dìn al-Shayyàl. Cairo: Dàr I˙yà" al-Kutub al'Arabiyya, 1960. al-Fàkihì, Mu˙ammad b. Is˙àq. Akhbàr Makka, ed. 'Abd al-Malik b. 'Abdallàh b. Dahìsh. Makka: Maktabat wa-Ma†ba'at al-Nah∂a l-Óadìtha, 1407/1987; repr. Beirut: Dàr Khi∂r, 1414 AH. Fishbein, M. ed. The History of al-ˇabari. Vol. XXXI: The War Between Brothers, trans. and annot. M. Fishbein. New York: State University of New York Press, 1992. Fragmenta Historicorum Arabicorum [Kitàb al-'uyùn wa-l-˙adà"iq], ed. M.J. De Goeje. Leiden: Brill, 1871. Ibn Abì Shayba, Abù Bakr 'Abdallàh b. Mu˙ammad. Mußannaf ibn Abì Shayba, ed. Kamàl Yùsuf al-Óùt. Riyadh: Maktabat al-Rushd, 1409 AH. Ibn Abì ˇàhir ˇayfùr, A˙mad. [al-Juz" al-sàdis min] Kitàb Baghdàd, ed. H. Keller. Leipzig, 1908. Ibn al-'Adìm, Kamàl al-Dìn 'Umar b. A˙mad b. Abì Gharàda. Bughyat al-†alab fì ta"rìkh Óalab, ed. Suhayl Zakkàr. Damascus: Ma†àbi' Dàr al-Ba'th, 1408/1988. Ibn 'Asàkir, Abù l-Qàsim 'Alì b. al-Óasan. Ta"rìkh madìnat Dimashq, ed. 'Umar b. Gharàma l-'Amrawì. Beirut: Dàr al-Fikr li-l-ˇibà'a wa-l-Nashr wa-l-Tawzì ', 1415–/1995–. Ibn A'tham al-Kùfì, Mu˙ammad b. 'Alì. Kitàb al-futù˙. MS. Sarai Ahmet, III, 2956; Hyderabad, 1388–95/1968–75. Ibn al-Athìr, 'Alì b. Mu˙ammad, 'Izz al-Dìn. Al-Kàmil fì l-ta"rìkh, ed. M.T. Houtsma. Leiden, 1863–71; Beirut: Dàr Íàdir and Dàr Beirut, 1385/1965. Ibn Óabib, Abù Ja'far Mu˙ammad. “Asmà" al-mughtàlìn min al-ashràf fì l-jàhiliyya wal- Islàm,” in Nawàdir al-makh†ù†àt 2, ed. 'Abd al-Salàm Mu˙ammad Hàrùn, vol. 6. Cairo, 1393/1773. Ibn Óazm, Abù Mu˙ammad, 'Alì b. A˙mad b. Sa'ìd. Naq† al-'arùs fì tawàrìkh alkhulafà", ed. Shawkì ˇayf. Cairo, 1951.

322

amikam elad

id., Jamharat ansàb al-'arab, ed. 'Abd al-Salàm Mu˙ammad Hàrùn. Cairo, 1962. Ibn Óibbàn al-Tamìmi al-Bustì, Abù Bakr. Ía˙ì˙ Ibn Óibbàn2, ed. Shu'ayb al-Arnà"ù†. Beirut: Mu"assasat al-Risàla, 1414/1993. Ibn al-Kalbì, Hishàm b. Mu˙ammad b. al-Sà"ib. Jamharat al-nasab, ed. W. Caskel, Leiden: Brill, 1966; ed. Nàjì Óasan. Beirut: 'Àlam al-Kutub, 1407/1986. id., Nasab ma'add wa-l-Yaman al-kabìr, ed. Nàjì Óasan. Beirut: 'Àlam al-Kutub and Maktabat al-Nah∂a l-'Arabiyya, 1408/1988. Ibn Khayyà† al-'Ußfurì, Khalìfa. Ta"rìkh Khalìfa b. Khayyà†, ed. S. Zakkàr. Damascus, 1968. al-Ißfàhànì, 'Alì b. Husayn Abù l-Faraj. Kitàb al-aghànì. Bùlàq, 1284–85 AH; Cairo: Dàr al- Kutub al-Mißriyya, 1345–/1927–. al-Jà˙iΩ, Abù 'Uthmàn 'Amr b. Ba˙r. “Dhamm akhlàq al-kuttàb,” in 'Abd al-Salàm Mu˙ammad Hàrùn, ed. Rasà"il al-Jà˙iΩ, vol. 2. Cairo, 1384/1965, 186–209. [Pseudo] al-Jà˙iΩ. Kitàb al-†àj fì Akhlàq al-Mulùk, ed. Fawzì 'A†wà. Beirut, 1970. al-Jahshiyàrì, Mu˙ammad b. 'Abdùs, Abù 'Abdallàh. Kitàb al-wuzarà" wa-l-kuttàb. Cairo. Mu߆afà l-Bàbì l-Óalabì, 1357/1938. al-Kha†ìb al-Baghdàdì, Abù Bakr, A˙mad b. 'Alì, Ta"rìkh Baghdàd. Cairo, 1349/1931. al-Mas'ùdì, Abù l-Óasan, 'Alì b. l-Óasan b. 'Alì. Murùj al-dhahab wa-ma'àdin al-jawhar, ed. C. Pellat. Beirut, 1966–79. id., Kitàb al-tanbìh wa-l-ishràf (BGA, VIII). Leiden, 1894. al-Maqrìzì, Taqiyy al-Dìn, Abù l-'Abbàs, A˙mad b. 'Alì. Kitàb al-muqaffà al-kabìr, ed. Mu˙ammad al-Ya'làwì. Beirut: Dar al-Gharb al-Islàmì, 1411/1991. al-Mubarrad, Abù l-'Abbàs Mu˙ammad b. Yazìd. Al-Kitàb al-kàmil, ed. W. Wright. Leipzig 1864–92. Muslim b. Óajjàj al-Naysàbùrì, Abù l-Óasan. Ía˙ì˙, ed. Fu"àd 'Abd al-Bàqì. [Cairo]: Dàr I˙yà" al-Kutub al-'Arabiyya, 1375/1955. al-Qummì, Mu˙ammad b. 'Alì Abù al-Óusayn b. Mùsà Ibn Bàbùy. 'Uyùn Akhbàr al-Ri∂à. Najaf, 1390/1970. al-Sam'ànì, Abù Sa'ìd, 'Abd al-Karìm b. Mu˙ammad b. Manßùr. Kitàb al-ansàb. Hyderabad, 1382/1962. al-ˇabarànì, Abù l-Qàsim, Sulaymàn b. A˙mad. Al-Mu'jam al-kabìr, ed. Óamdì b. 'Abd al- Majìd al-Salafì. Mosul: Maktabat al-'Ulùm wa-l-Óikam, 1404/1983. id., Al-Mu'jam al-ßaghìr, ed. Mu˙ammad Shakùr-Ma˙mùd al-Óàj al-Marìr. Beirut/ Amman: Maktab al-Islàmì and Dàr 'Ammàr, 1405/1985. al-ˇabarì, Mu˙ammad b. Jarìr. Ta"rìkh al-rusul wa-l-mulùk, ed. M.J. De Goeje. Leiden: Brill, 1879–1901 (New Printing Leiden, 1964). Persian version by Abù 'Alì Mu˙ammad Bal'amì. trans. H. Zotenberg as Chronique de Tabari, traduit sur la version persane d’Abou-'Al, Mohammed Bel'am. Paris, 1867–74. al-Ya'qùbì, A˙mad b. Abì Wà∂i˙. Ta"rìkh, ed. M.Th. Houtsma. Leiden, 1882; Beirut: Dàr Íàdir and Dàr Beirut, 1379/1960. Yàqùt al-Rùmì al-Óamawì, Shihàb al-Dìn, Abù 'Abdallàh. Mu'jam al-buldàn. Beirut: Dàr Íàdir, 1955–57. al-Zubayrì, Muß'ab b. 'Abdallàh. Kitàb nasab Quraysh3, ed. E. Levi-Provençal. Cairo: Dàr al-Ma'àrif, 1953. Secondary Sources Agha Saleh Said. The agents and forces that toppled the Umayyad caliphate. Ph.D. diss., The University of Toronto, 1993. id., “The Arab Population in Khuràsàn during the Umayyad Period: Some demographic Computations,” Arabica 46 (1999), 211–29. Amabe, F. The emergence of the 'Abbàsid autocracy. The 'Abbàsid army, Khuràsàn and Adharbayjàn. Kyoto: Kyoto University Press, 1994.

MAWÀLÌ

in the composition of al-ma"mùn’s army

323

Arazi, A. and A. Elad. “L’Épître a l’armée, al-Ma"mùn et la second da'wa,” Studia Islamica 66 (1987), 27–70 (1st part); 67 (1988), 29–73 (2nd part). Ayalon, D. “The Muslim city and the military aristocracy,” in Proceedings of the Israel Academy of Sciences and Humanities, vol. 2. 1968, 311–329; repr. id., Studies on the Mamlùks of Egypt. 1250–1517. London: Variorum Reprints, 1977, no. 7. id., “The military reforms of Caliph al-Mu'taßim: Their background and consequences,” in Collected studies: Islam and the Abode of War: Military Slaves and Islamic Adversaries. n.p.: Variorum, 1994, no. 1. id., “Preliminary remarks on the Mamlùk military institution in Islam,” in V.J. Perry and M.E. Yapp, eds., War, Technology and Society in the Middle East. Oxford, 1975, 44–58; repr. id.. The Mamlùk Military Society. London: Variorum Reprints, 1979, no. 9. Bartold, W. Turkestan down to the Mongol invation2, trans. W. Bartold and H.A.R. Gibb, H.J.W. Gibb Memorial Series, New Series, 5. London, 1928. al-Bàshà, Óasan. Diràsàt fì ta"rìkh al-dawla al-'Abbàsiyya. Cairo, 1975. Bosworth, C.E. Sìstàn under the Arabs from the Islamic conquest to the rise of the Íaffàrids (30 –250/651–864). Rome, 1968. Crone, P. “The 'Abbàsid abnà" and Sàsànid cavalrymen,” Journal of the Royal Asiatic Society 8 (1998), 1–19. id., Slaves on horses: The evolution of the Islamic polity. Cambridge: Cambridge University Press, 1980. id., “The significance of wooden weapons in al-Mukhtàr’s revolt and the 'Abbàsid revolution,” in I.R. Netton, ed., Studies in honour of Cliford Edmund Bosworth. Vol. I: Hunter of the East: Arabic and Semitic studies. Leiden: Brill, 2000, 174–87. Daniel, E.L. “The Ahl al-Taqàdum and the problem of the constituency of the 'Abbàsid revolution in the Merv oasis,” Journal of Islamic Studies 7 (1996), 150–79. id., “Arabs, Persians and the advent of the 'Abbàsids reconsidered [A review article of the books of R.M. Guzman, M. Sharon and M. Zakeri],” Journal of the American Oriental Society 117 (1997), 542–48. id., The political and social history of Khurasan under 'Abbàsid rule. Minneapolis/Chicago, 1979. Dennett, D.C. Al-Jizya wa-l-Islam, Arabic trans. Fawzì Fahìm Jawdallàh. Beirut/New York, 1960. id., Marwàn b. Mu˙ammad: The passing of the Umayyad caliphate. Ph.D. diss., Harvard University, 1939. al-Dùrì, 'Abd al-'Azìz. Al-'aßr al-'Abbàsì al-awwal. Baghdàd, 1945. Elad, A. Characteristics of the development of the 'Abbàsid army (especially ahl-Khuràsàn and al-Abnà" units) with emphasis on the reign of Al-Amìn and al-Ma"mùn. Ph.D. Diss., The Hebrew University, Jerusalem, 1986 (in Hebrew). id., “The ethnic composition of the 'Abbàsid revolution: A reevaluation of some recent research,” Jerusalem Studies in Arabic and Islam 24 (2000), 246–311. id., “The Southern Golan in the early Muslim period. The significance of two newly discovered milestones of 'Abd al-Malik,” Der Islam 76 (1999), 33–88. id., “Two identical inscriptions from Jund Filas†ìn from the reign of the 'Abbàsid caliph al-Muqtadir,” Journal of the Economic and Social History of the Orient 35 (1992), 301–60. id., “Aspects of transition from the Umayyad to the 'Abbàsid Caliphate,” Jerusalem Studies in Arabic and Islam 19 (1995), 89–132. Frye, R.N. “[A review of ] M. Zakeri, Sàsànid soldiers in early Muslim society. The origins of 'Ayyàràn and Futuwwa,” Jerusalem Studies in Arabic and islam 20 (1996), 263–67. Gabrieli, F. “La successione di Hàrùn ar-Rashìd e la guerra fra al-Amìn e alMa"mùn,” Rivista degli Studi Orientali 11 (1926–28), 341–97.

324

amikam elad

id., Al-Ma"mùn e gli 'Alidi. Leipzig, 1929 (Morgenländische Texte und Forschungen, Harausgegeben von A. Fischer 2. Band. Haft 1). Goldziher, I. Muslim Studies, vol. 1. London: George Allen and Unwin, 1967. Ismà'ìl, O.S.A. “Mu'taßim and the Turks,” Bulletin of the School of Oriental and African Studies 29 (1966), 12–24. al-Karawì, Ibràhìm Sulaymàn. NiΩàm al-wizàra fì l-'aßr al-'Abbàsì al-awwal. Kuwayt, 1983. Katàbjì, Z. Al-Turk fì mu"allafàt al-Jà˙iΩ wa-makànatuhum fì l-ta"rìkh al-Islàmì ˙attà awàsi† al-qarn al-thàlith al-hijrì. Beirut, 1972. Kennedy, H. The early 'Abbàsid caliphate. London, 1981. id., The Prophet and the age of the caliphates. The Islamic Near East from the sixth to the eleventh century. London: Longman, 1986. Al-Khàzin, W. Al-Óa†àra al-'Abbàsiyya. Beirut, 1984. Lassner, J. The topography of Baghdàd in the early Middle Ages, Text and studies. Detroit: Wayne State University Press, 1970. id., The shaping of 'Abbàsid rule. Princeton/New Jersey: Princeton University Press, 1980. Lecker, M. “Judaism among Kinda and the Ridda of Kinda,” Journal of the American Oriental Society 115 (1995), 635–50 (repr. id., Collected Studies, Jews and Arabs in Pre-and Early Islamic Arabia. London: Variorum Reprints Collected Studies, 1988, no. 14; Hebrew version in: Taymà 6 (1998), 1–28). Løkkagaard, F. Islamic taxation in the Classical period, with special reference to circumstances in Iraq. Copenhagen, 1950. Mottahedeh, R. “The 'Abbàsid caliphate in Iran,” in R.N. Frye, ed., The Cambridge History of Iran, IV. Cambridge: Cambridge University Press, 1975, 57–89. Nicol, N.D. Early 'Abbàsid administration in the central and eastern provinces, 132–218/750–833 A.D. Ph.D. diss., The University of Washington, 1979. Omar, F. Al-'Iràq wa-l-ta˙addì al-Fàrisì. Baghdàd: Dàr al-Shu"ùn al-Thaqafiyya, 1987. id., Al-Judhùr al-ta"rìkhiyya li-l-wizàra al-'Abbàsiyya (diràsa ta˙lìliyya naqdiyya li-qirà"at surdìl 'an al-wuzarà" al-'Abbàsiyyìn wa-ab˙àth ukhrà). Baghdàd: Dàr al-Shu"ùn alThaqafiyya, 1989. Pipes, D. Slave soldiers and Islam. The genesis of a military system. New Haven/London, 1981. al-Rifà'ì, A˙mad Farìd. 'Aßr al-Ma"mùn, vol. 3. Cairo, 1346/1928, 26–37. Íafwat, A˙mad Zakì. Jamharat rasà"il al-'Arab fì l-'ußùr al-'arabiyya al-zàhira, vol. 3. Cairo, 1356/1937, 377–97. Íaqar, Nàdiya Óusnì. Al-Silm fì l-'alàqàt al-'Abbàsiyya al-Bìzan†iyya fì l-'aßr al-'Abbàsì al-awwal. Mecca, 1406/1985. Samadi, S.B. “The struggle between the two brothers al-Amìn and al-Ma"mùn,” Islamic Culture 32 (1958), 99–120. Shaban, M.A. Islamic History. A New Interpretation. A.D. 600–750 (A.H. 132), vol. 1. Cambridge: Cambridge University Press, 1971. id., The 'Abbàsid Revolution. Cambridge: Cambridge University Press, 1970. Shàkir, Mus†afà. Dawlat banì l-'Abbàs. Kuwayt, 1973. Sharon, M. “The military reforms of Abù Muslim their background and consequences,” in M. Sharon, ed., Studies in Islamic History and Civilization in Honour of Professor David Ayalon. Leiden/Jerusalem: Cana Publishing House and E.J. Brill, 1986, 105–44. id., Black banners from the East. The establishment of the 'Abbàsid state. Incubation of a revolt. Jerusalem/Leiden: The Magnes Press and E.J. Brill, 1983. id., Revolt. The social and military aspects of the 'Abbàsid revolution. Jerusalem: The Max Schloessinger Memorial Fund, the Hebrew University, 1990. Sourdel, D. Le vizirat 'Abbàside de 749 a 936. Damascus, 1959–60.

MAWÀLÌ

in the composition of al-ma"mùn’s army

325

Steppat, F. “From 'Ahd Arda-ìr to al-Ma"mùn: A Persian element in the policy of the Mi˙na,” in Wadàd al-Qà∂ì et al., eds., Studia Arabica and Islamica: Festschrift I˙sàn 'Abbàs on His Sixtieth Birthday. Beirut. American University of Beirut, 1981, 451–54. Watt, M. The formative period of Islamic thought. Edinburgh, 1973. id., “The political attitude of the mu'tazilah,” Journal of the Royal Asiatic Society (1963), 38–57. Wellhausen, J. The Arab kingdom and its fall. Beirut, 1963. Wensinck, A.J. Concordance et indices de la Tradition Musulmane. Leiden, 1933, 1963. Al-Yazbekì, Tawfìq Sul†àn. Al-Wizàrà, nash"atuhà wa-ta†awwuruhà fì l-dawla al-'Abbàsiyya (132–447). Mosul, 1976. Zakeri, M. Sàsànid soldiers in early Muslim society. The origins of 'Ayyàràn and Futuwwa. Wiesbaden: Otto Harrassowitz, 1995. Zaydàn, J. Umayyads and 'Abbàsids, trans. D.S. Margoliouth, E.J.W. Gibb Memorial Series, 4. Leiden, 1907.

MAWÀLÌ AND MUSIC Hilary Kilpatrick

In the epistle which prefaces his Akhbàr Abì Tammàm, al-Íùlì (c. 260– 335/873–946) makes a fleeting reference to vocal music. He calls it “the science of princes, of which they have almost the monopoly” and goes on to outline what it includes: “the poems set to music, the attribution of the lyrics to their authors, the reason for their composition, [the names of ] those who have composed settings for one or other verse and the mention of the [settings’] melodic and rhythmic modes and genres.”1 Most references to musicians in the pre-Islamic and early Islamic period which have survived treat of those who contributed to the courtly tradition of music which alÍùlì describes or else to its forerunners. About other forms of music, such as work songs, children’s songs, music accompanying the important moments of human life, or music related to religious ceremonies, hardly any information is extant. The major source for the study of musicians2 and musical practice up to the end of the third/ninth century is Abù l-Faraj alIßbahànì’s (284–c. 363/879–972/73) Kitàb al-aghànì. The Aghànì starts out with the aim of providing correct information about the list of the Hundred Songs drawn up by Is˙àq al-Mawßilì (155–235/772–850) for al-Wàthiq (195–232/811–847),3 covering the topics listed by alÍùlì. But it branches out into all kinds of information connected with the poets and composers of the songs, later instances of their performance, occasions for which they were composed and so on. It offers a wealth of information about the musicians working in the

Al-Íùlì, Akhbàr Abì Tammàm, ed. Khalìl Mu˙ammad 'Asàkir, Mu˙ammad 'Abduh 'Azzàm and NaΩìr al-Islàm al-Hindì, repr. 3rd ed. (Beirut: Dàr al-Àfàq al-Jadìda, 1400/1980), 8. 2 The term mughannì, rendered here generally by “musician,” in fact denotes the activities of composing, singing and accompanying oneself on an instrument. 3 Abù l-Faraj 'Alì b. al-Óusayn al-Ißbahànì (d. c. 363/973), Kitàb al-aghànì, 24 vols. (Cairo: Dàr al-Kutub and al-Hay"a al-'Àmma al-Mißriyya li-l-Kitàb, 1927–74), I, 5–6. 1

MAWÀLÌ

and music

327

tradition which developed into 'Abbàsid court music, although as will become clear it is not exhaustive. Thus, for instance, it only names certain musicians in passing but does not accord them individual treatment. Some of these lacunas may be explained by the fact that it is unfinished. Ibn Khurdàdhbih’s (c. 211–300/825–911) Mukhtàr min kitàb almalàhì,4 while a pygmy in comparison to the giant Aghànì, adds some names of other early singers, though some of his information is unreliable or too incomplete to be useful.5 The Fà†imid court musician Ibn al-ˇa˙˙àn’s (d. after 449/1057) Óàwì l-funùn wa-salwat al-ma˙zùn6 also has information about individual performers not found elsewhere in works about the history of singing; by contrast, al-Óasan b. A˙mad al-Kàtib’s (early fifth/eleventh c.) Kamàl adab al-ghinà" only mentions singers sporadically in the course of illustrative anecdotes.7 Some adab works of a general nature, such as Ibn 'Abd Rabbih’s (d. 328/940) Al-'iqd al-farìd, accord space to music, while scattered references to singers and singing may be found in al-ˇabarì’s and other histories. Al-Jà˙iΩ’s (c. 160–255/776–868/69) Risàla fì l-qiyàn, despite its promising title, says nothing about music as such. In what follows, I shall sketch the evolution of music from the Jàhiliyya to the 'Abbàsid era, concentrating on the musicians’ social origins, and then present the authors of books on the theory and practice of music. In conclusion I shall try to estimate the importance of the role of different social groups in the development of Arabic music up to the end of the third/ninth century. It should be pointed out that the term mawlà, when applied to musicians, is not further specified, and rarely do subsequent akhbàr about the person in question shed any light on whether “freedman,” “ally,” or “non-

4 Ibn Khurdàdhbih (d. 300/911), Mukhtàr min kitàb al-lahw wa-l-malàhì, ed. Ighnà†iyùs 'Abduh Khalìfa (Beirut: Dàr al-Mashriq, 1969). 5 For instance, he claims an 'Abd al-Ra˙màn b. 'Awf al-Zuhrì was a composer with the nickname Ibn Abì Qabà˙a (Ibn Khurdàdhbih, Mukhtàr, 52). Abù l-Faraj is generally skeptical about his information; cf. the criticism of Ibn Khurdàdhbih’s claims about the musical activities of the caliphs (Aghànì, IX, 250). 6 Ibn al-ˇa˙˙àn al-Mùsìqì, Óàwì l-funùn wa-salwat al-ma˙zùn, ed. Eckhard Neubauer (Frankfurt am Main: Institute for the History of Arabic-Islamic Science, 1990). 7 Al-Óasan b. A˙mad b. 'Alì al-Kàtib, Kitàb kamàl adab al-ghinà", ed. 'Abd alMalik Gha††às Khashaba (Cairo: al-Hay"a al-Mißriyya al-'Àmma li-al-Kitàb, 1395/1975); id., La perfection des connaissances musicales, trans. Amnon Shiloah (Paris: Paul Geuthner, 1972).

328

hilary kilpatrick

Arab convert to Islam” is meant.8 I cannot do otherwise than follow the sources in their imprecise use of the term. The references to pre-Islamic musical practice are sparse, and no clear picture of it can be formed. But on the basis of the terms used, two different types of music can be distinguished.9 One, referred to as ˙udà", appears to denote a simple type of singing, associated with the journeying of the camel-drivers. The other, designated by naßb, sinàd and hazaj, is more evolved. While naßb was performed by men, and in some accounts appears to be close to ˙udà",10 sinàd and hazaj were performed by women for entertainment, associated with urban centers and found in palaces, rich men’s houses and taverns. The sinàd appears to have been a more ornate, “heavy” style, the hazaj a lighter one, but it may be that later writers have projected back into the pre-Islamic period distinctions they were familiar with from their own time. Ibn al-ˇa˙˙àn gives a short list of singers, both men and women, of the pre-Islamic period and the time of the Prophet, and from the names it appears that the men are free-born Arabs, while the women are usually slave-girls, qaynàt, sometimes only referred to as “the slave-girl of So-and-so.”11 One or two women are, however, described as mawlàh.12 Two other forms of music are known to have existed, even if they are not referred to in discussions of early singing. Both are associated with women. One, which has its own term, is the naw˙ or ritual lament. The other, a much more unusual occurrence, is the battle chant, which appears to have had no special designation; it is only to be inferred from a reference to Hind bint 'Utba and her 8 EI 2, art. “Mawlà” II, 1–3 (P. Crone) makes clear how fluid the term was and traces how it developed. 9 See for the whole subject the discussion in Wright, “Music and Verse,” in A.F.L. Beeston et al., eds., Arabic literature to the end of the Umayyad period, The Cambridge History of Arabic Literature (Cambridge: Cambridge University Press, 1983), 435–41. Wright also suggests other interpretations of the terms here and in the EI 2 article “Mùsì˚ì,” but the basic distinctions remain the same. 10 The term rukbànì is also encountered; it may have been a synonym of ˙udà". Cf. Nàßir al-Dìn al-Asad, Al-Qiyàn wa-l-ghinà" fì l-'aßr al-jàhilì, 3rd ed. (Beirut: Dàr al-Jìl, 1988), 95. 11 Ibn al-ˇa˙˙àn, Óàwì l-funùn, 33–35: 'Alas Dhì Jadan, 'Alqama al-Fa˙l, Jadhìma b. Sa'd, Rabì'a b. Hàrùn, al-Na∂r b. al-Óàrith b. Kalada, Rabà˙ b. al-Mughtarif, Ibn Abì Dubàkil; the two singing slaves of Óudhayfa b. Badr, Hind and Fartanà the singing slaves of Óujr b. al-Óàrith, the singing slaves of 'Abd al-Masì˙ in Najràn, ¸abya and al-Rabàb, 'Abdallàh b. Jud'àn’s Two Crickets and others. 12 Al-Asad, Al-Qiyàn, 87–88: Sàra the mawlàh of 'Amr b. Hàshim b. 'Abd alMu††alib or 'Amr b. Abì Sayfì b. Hàshim b. 'Abd Manàf, and 'Azza the mawlàh of al-Aswad b. al-Mu††alib.

MAWÀLÌ

and music

329

women companions urging the Meccans on at U˙ud with poetry accompanied by tambourines.13 The picture that emerges of preIslamic music is of free men and unfree women performing different genres, with free women’s singing in public14 being confined to performing in the exceptional circumstances of campaigns or else, quite possibly, to lamenting their deceased. The early Islamic conquests, which introduced so much change into social and cultural life in Arabia, affected music profoundly. The flow of wealth to the Óijàz and other urban centers encouraged the Arab aristocracy in the practice of maintaining singing slave-girls, while the foreigners who arrived in the Óijàz as slaves brought their own music with them. The stories of some of the early Umayyad musicians’ innovations being sparked off by the Persian music they heard from the workmen singing as they repaired the Ka'ba during Ibn al-Zubayr’s rule (64/683) may not be true in all their details, but they are perfectly credible as an indication of one of the impulses behind the development of music at the time. And the workers did not only sing, they brought their own instruments, in particular the lute.15 Those Arabs who had taken part in campaigns will also have encountered different kinds of music, so that their artistic horizons will have been widened. This is the context in which the existing style of entertainment music evolved to become al-ghinà" al-mutqan (“perfect,” “precise” or “accurate” singing).16 Several names are associated with the pioneering stage: the Persian Nashì†, a shadowy figure who is simply said to have come to the Óijàz and to have sung there, Sà"ib Khàthir (d. 63/683),17 also a Persian, who is recorded as the first to have

13

Aghànì, XV, 190, quoting Ibn Is˙àq. Genres sung in private, such as lullabies, are never referred to. 15 Ibn Misja˙ was influenced by the singing of the Persians (Aghànì, III, 278) while Ibn Surayj copied their use of the lute as an accompanying instrument (Aghànì, I, 250). 16 Cf. the discussion of the term in Wright, “Music and Verse,” 442. This sketch of music in the Umayyad period is based mainly on his chapter in Arabic literature to the end of the Umayyad period, 441–49. 17 The dates given for singers are drawn either from the relevant Encyclopedia of Islam (2nd ed.) articles, or in the case of some 'Abbàsid ones from Neubauer, Musiker am Hof der frühen 'Abbàsiden (Ph.D. diss., Frankfurt am Main, 1965), which is an invaluable source. Determining exactly when musicians lived is often very difficult; sometimes only the names of their teachers and of those before they performed (and in the case of slaves of their owners) enable them to be situated roughly in time. Where the information is too vague I have not given dates. 14

330

hilary kilpatrick

sung al-ghinà" al-mutqan in Arabic, accompanying himself on the lute, and Ibn Mu˙riz and Ibn Misja˙, both of whom, having familiarized themselves with the singing of Mecca and Medina, are said to have traveled to Iran and then Syria to learn the Persian and Byzantine traditions of music.18 Al-ghinà" al-mutqan evolved out of the combining of elements from the Persian and Byzantine musical systems with the older Arab tradition of entertainment music, as emerges both from the Aghànì and from theoretical writings.19 Whatever elements were introduced from abroad, they had to be compatible both with this tradition and with lyrics in Arabic. The new, “accurate,” singing had a codified system of modes, eight in number, and of rhythmic cycles; these latter were evidently more important for identifying a setting, since often they are given alone.20 The distinction between “heavy” and “light” styles noted for the pre-Islamic entertainment music persisted, ramal and hazaj being used for the latter. The Persian wooden-bellied lute gradually established itself as the main instrument accompanying singing, rather than percussion instruments which seem to have been traditional in the Óijàz.21 At the same time composers took to using two lines of poetry, rather than one, as the basis for their songs.22 Accompanying all these changes was a growing awareness of the subtleties of vocal technique and of skill in playing instruments. It is possible to form an idea of the social background and status of those who took part in this development of Arabic music, even though information about musicians’ lives is sketchy and dating them precisely is usually out of the question. The great majority of Umayyad

18

Aghànì, VIII, 321–22 (Sà"ib Khàthir, with a passing mention of Nashì†); I, 378 (Ibn Mu˙riz); III, 276 (Ibn Misja˙). Here, too, it is unlikely that both men followed the same itinerary. But there is nothing strange in musicians traveling fì †alab alghinà", as the biographies of many European composers show. An example of the traveling music student from the early 'Abbàsid period is Ibràhìm al-Mawßilì, for whom see below. 19 The early development of Arabic musical theory is discussed in detail in Neubauer, “Die acht ‘Wege’ der arabischen Musiklehre und der Oktoechos,” Zeitschrift für Geschichte der arabisch-islamischen Wissenschaften 9 (1994), 373–414; and id., “Alflalìl, die Töne und die musikalischen Metren,” Zeitschrift für Geschichte der arabischislamischen Wissenschaften 10 (1995–96), 255–323. 20 This is the rule in Ibn Khurdàdhbih’s Mukhtàr min kitàb al-malàhì. 21 Al-Walìd b. Yazìd, who was a versatile instrumentalist, accompanied himself either on the lute, on a hand-drum, or walking up and down beating a tambourine “in the Óijàzì fashion” (Aghànì, IX, 274). 22 This is another innovation ascribed to Ibn Mu˙riz (Aghànì, I, 379).

MAWÀLÌ

and music

331

male musicians whose status is mentioned are mawàlì. This applies not only to leading singers, such as Ibn Surayj (d. 96/714 or after), al-Gharì∂ (d. c. 98/716–17), or Ibn 'À"isha (d. before 125/743), but also to lesser performers such as Buday˙ or Ash'ab (1st half second/ eighth c.).23 Furthermore, most were apparently born mawàlì; only Ma'bad (d. 125/743 or 126/744) and Ya˙yà Qayl are described as having started life in slavery.24 They are drawn equally from Mecca and Medina, the exceptions being Ibn Mish'ab, who came from ˇà"if but settled in Mecca and 'Umar al-Wàdì, the founder of a tradition of music in Wàdì l-Qurà who trained in Mecca.25 Many of them are stated to have been of at least partly non-Arab descent, Persian (Sà"ib Khàthir, Ibn Mu˙riz, 'Umar al-Wàdì, Yùnus al-Kàtib), African (Ibn Misja˙, Ma'bad), Turkish (Ibn Surayj) or Berber (al-Gharì∂).26 Occasionally their original occupation is given; al-Gharì∂ was a tailor to start with, Sa'ìd al-Hudhalì a stone-cutter and maker of jars, Ma'bad was in the livestock trade, 'Umar al-Wàdì engaged in irrigation works (muhandis).27 Yùnus al-Kàtib’s (d. c. 147/765), Sinàn alKàtib’s (mid second/eighth c.) and Khalìl al-Mu'allim’s titles speak for themselves; they show that by the end of Umayyad era singing did not only attract men from the lowest points of the social scale. The custom of Arab tribesmen becoming musicians did not die out entirely. A˙mad al-Naßbì (d. 83/702), of the Banù Hamdàn and closely related to A'shà Hamdàn, was a very fine pandore player. Another achievement of his is harder to interpret. He is said to have been the first to sing naßb songs and to have introduced them into al-ghinà".28 Unless another meaning of naßb is intended, it is wrong to attribute the creation of this genre to A˙mad.29 But the second part of the statement is less problematic, if it is understood to indicate that he adapted one of the pre-Islamic styles performed by men

23

Aghànì, I, 248; II, 359; II, 203; XV, 174; XIX, 135. Aghànì, I, 39; III, 110. 25 Aghànì, IV, 321; VII, 85. 26 Aghànì, VIII, 321; I, 378; VII, 85; IV, 398; III, 276; I, 36, 250; II, 359. Here, too, the information is usually imprecise, in particular because only the father’s side of the family is mentioned 27 Aghànì, II, 360; V, 65; I, 39; VII, 85. 28 Aghànì, VI, 63. 29 But cf. Is˙àq al-Mawßilì’s remark that the general style of singing in al-Óìra was somewhere between naßb and hazaj, and closer to naßb (Aghànì, II, 352). In that case A˙mad could have been the first in Iraq to sing proper naßb settings. 24

332

hilary kilpatrick

to the evolving “art music” of the Umayyad era.30 Slightly later than A˙mad al-Naßbì is Óunayn b. Balù' al-Óìrì, a member of the Christian 'Ibàdì (Nestorian) community. Óunayn is said to have been in the business of renting out camels, which suggests he would have been familiar with the primitive camel-drivers’ songs, even though his own style apparently resembled that of his Óijàzì contemporaries.31 Two other tribesmen were Sa'ìd al-Dàrimì and Màlik b. Abì l-Sam˙, of Tamìm and ˇayy respectively, whose families had sought refuge from conflict and drought in the Óijàz. While Sa'ìd is better known as a poet, Màlik became a prominent singer in the later Umayyad period. The most socially eminent musicians were members of Quraysh, the caliphs 'Umar b. 'Abd al-'Azìz (c. 60–101/680–720) and al-Walìd b. Yazìd (90–126/709–744). During his time as governor of the Óijàz (87–93/706–712) 'Umar is said to have interested himself in music to the point of composing songs for his wife.32 Whereas music making appears to have been a private matter indulged in by 'Umar during a short period of his life, it was a consistent trait of al-Walìd, who not only invited many musicians to his court but himself sang in the presence of his courtiers.33 While mawàlì generally supplanted members of Arab tribes as performers in the Umayyad period, the latter played a crucial role as patrons and, on occasion, protectors of musicians. No-one is recorded as regretting that one of his mawàlì was musically gifted; after all, a mawlà who could sing well was not only a pleasure to listen to but might bring in an income from teaching. Some akhbàr set in the Óijàz portray a group of wealthy young men with time on their hands who were never happier than when listening to a singer’s latest composition. A few well-connected members of Quraysh, such as 'Abdallàh b. Ja'far and Ibn Abì 'Atìq, were known for their appre-

30 If this interpretation is correct, it provides the only evidence of musical innovation outside the Óijàz in this period. 31 There is no firm information about Óunayn’s manner of singing, but the fact that one of his settings was included in the Top Hundred as well as the anecdotes which portray his contacts with the Óijàzì musicians suggest that he was working in the same tradition as they were; cf. Aghànì, II, 260, 353–56. 32 Abù l-Faraj has submitted this information to critical examination before including it in the Aghànì; cf. his remarks IX, 250–51. 33 Aghànì, IX, 274–75. Among Arabs Ibn Khurdàdhbih adds not only 'Abd alRa˙man b. 'Awf al-Zuhrì but also 'Abdallàh b. Mu'àwiya al-Bàhilì, who accompanied Qutayba b. Muslim to Khuràsàn and settled in Rayy (Mukhtàr, 26).

MAWÀLÌ

and music

333

ciation of music34 and could be relied on to defend singers from the attacks of philistine governors.35 Against this background Yazìd b. 'Abd al-Malik’s (68–105/688–724) interest in music and particularly in singing-girls appears simply as a further stage in an already wellestablished development. In the Umayyad period an influential section of society was well disposed towards music, and the development of the art owes much to this. Unlike those of men, the social origins of women singers did not change markedly between the pre-Islamic and Islamic periods, as far as can be seen. At most it could be argued that mawlayàt became relatively more important in the Umayyad era, although given the extremely fragmentary character of information about the Jàhiliyya this is little more than speculation. The first to acquire the techniques of al-ghinà" al-mutqan was a mawlàh of the Anßàr, 'Azza alMaylà". But she also represents a link to pre-Islamic entertainment music, since the names of several of her teachers are given. While not an innovator herself, it was she who made the new manner of singing widely popular.36 Like 'Azza, Umm Ja'far, a mawlàh of 'Abdallàh b. Ja'far,37 and Jamìla, a mawlàh of Banù Sulaym, were Medinans. Jamìla, in particular, enjoyed great prestige, as is reflected in the account of the music festival she organized following her pilgrimage, at which all the singers of Mecca and Medina performed. The account, which ignores chronological considerations and is evidently apocryphal, seeks to provide a panorama of musical activity in the Óijàz during the Umayyad period, but the fact that Jamìla

34 'Abdallàh was the nephew of 'Alì b. Abì ˇàlib; he died in or after 80/699–700 (EI 2, s.v. (K. Zetterstéen)). Ibn Abì 'Atìq’s full name was 'Abdallàh b. Mu˙ammad b. 'Abd al-Ra˙màn b. Abì Bakr. He was a contemporary of 'Umar b. Abì Rabì'a and Kuthayyir (EI 2, s.v. (Ch. Pellat)) (i.e., late first/seventh and early second/eighth c.). 35 Cf. Aghànì, VIII, 341–42, where Ibn Abì 'Atìq prevails upon the new governor of Medina, 'Uthmàn b. Óayyàn al-Murrì, not to expel the singers; XVII, 176–77, where 'Abdallàh b. Ja'far persuades a governor to rescind the order that 'Azza l-Maylà" should stop performing. 36 Aghànì, XVII, 162. Among 'Azza’s teachers was Sìrìn, very likely the same who is mentioned as belonging to Óassàn b. Thàbit and singing to her own accompaniment on the mizhar, probably a zither-like instrument (XII, 67). The observation that none of 'Azza’s teachers is mentioned elsewhere in the Aghànì would then be incorrect (Everett K. Rowson, “The Effeminates of Early Medina,” Journal of the American Oriental Society 111 (1991), 679). The lack of reference to these singers may be explained simply by the fact that none of their settings survived; consequently they did not qualify for an article of their own in the Aghànì. 37 Aghànì, VI, 253, as the composer of one of the Top Hundred songs.

334

hilary kilpatrick

is the festival’s director is significant. Jamìla had many pupils, both men and women, which brought her patrons a considerable income.38 Another mawlàh was Shahiyya, like al-Gharì∂ and Ya˙yà Qayl connected to the 'Abalàt.39 From the end of the Umayyad period the name of Khulayda, a Meccan mawlàh of partly African origin, has been preserved. Khulayda’s reply to an offer of marriage from Mu˙ammad b. 'Abdallàh, the caliph 'Uthmàn’s great-grandson, is worth quoting for the glimpse it gives of the standing and character of some women singers of this period. She answered Mu˙ammad’s messenger: “You have related your master’s genealogy and made it clear, so now listen to mine, my dear. My father was not sold according to an Islamic contract or covenant, he lived as a slave and died with his feet in fetters and a chain round his neck. He was a runaway and a thief. My mother bore me out of wedlock and died while she was on the run. So now you know who I am. If your master wants a legal marriage or an affair which is public knowledge, let’s go. I’m his woman.” “He will never embark on anything illicit.” “And he must not be ashamed of anything licit. A secret marriage is out of the question. I shall never go in for that and bring shame on the singing-girls.” Mu˙ammad abandoned the plan, but asked Khulayda if he could just take a look at her to console himself, to which she agreed.40 Among singing slave-girls in the early Umayyad period are ˇawrà" and al-Baghùm, specialists in the ritual lament.41 Somewhat later are Sallàmat al-Qass and ˇabàba, who were trained in Medina before Yazìd b. 'Abd al-Malik bought them and took them to Syria. Sallàma, in particular, evidently enjoyed wide popularity, for the whole city turned out to take leave of her when she set out for Damascus;42

38 Aghànì, VIII, 208–20 for the festival and 186–87 for Jamìla’s career. She refers to her earnings as follows (187): “wa-qad kasabtu li-mawàlìya mà lam yakh†ur lahunna [kadhà] bi-bàl.” The biographies of many later musicians mention Jamìla as one of their teachers. 39 Aghànì, VI, 100, as the composer of one of the Top Hundred songs. The 'Abalàt refers to a clan of 'Abd Shams b. 'Abd Manàf (Ibn Óazm, Jamharat ansàb al-'arab, ed. 'Abd al-Salàm Mu˙ammad Hàrùn, 5th ed. (Cairo: Dàr al-Ma'àrif, 1982), 74). The individuals concerned were Thurayyà, 'Umar b. Abì Rabì'a’s beloved, and her sisters al-Ru∂ayyà, Qurayba and Umm 'Uthmàn (Aghànì, II, 359). 40 Aghànì, XVI, 190, 191–92. 41 They took pity on al-Gharì∂ when Ibn Surayj dismissed him as a pupil; Aghànì, II, 361. 42 Aghànì, VIII, 344 (Sallàma); XV, 122 (Óabàba).

MAWÀLÌ

and music

335

she appears to have been the better musician. Yazìd’s son, al-Walìd, is known to have possessed two singing-girls, Shuhda and al-Nawàr.43 Besides these male and female musicians about whose origins and careers something is known, the Aghànì includes the names of many others. It is likely that the obscure composers of songs from the Top Hundred date from the Umayyad period. In other cases one can sometimes infer from the context when a musician lived. A geographical designation is often attached to men’s names, so that certain conclusions about these otherwise forgotten artists can be drawn. The most striking feature is the number of singers, ten altogether, associated with al-Óìra44 and Kùfa.45 As a nursery of obscure performers Wàdì l-Qurà rates next, with four names,46 while Mecca provides two47 and ˇà"if and Jedda one each.48 Given these figures, and bearing in mind that the Kùfan singers between them furnished seven of the Top Hundred settings, the question arises whether the contribution of Iraqì musicians to music in the Umayyad period has not been underestimated in the sources.49 Another noteworthy feature of these names of unknowns is that they may include the designation of a trade. Qafà the Carpenter, the Jeddan Date-seller: these performers, like some of the mawàlì mentioned above, were of humble origin.50

43

Aghànì, VI, 260; VII, 69. 'Abàdìs, Zayd b. al-ˇulays, Zayd b. Ka'b, Màlik b. ˇumama (Aghànì, II, 352). Their singing is said to have been closer to naßb than to hazaj. 45 'Azzùr (Aghànì, III, 50), Mùsà b. Khàrija (IV, 133), Bàbawayh (IV, 213; VIII, 266; XIII, 125), his brother Sulaymàn (VI, 340), Dukayn b. Yazìd (VI, 159), Mu˙ammad Na'ja (VII, 228). 46 Sulaymàn al-Wàdì, Khulayd b. 'Atìk, Ya'qùb al-Wadì (Aghànì, VI, 280); alAfrak (Ibn Khurdàdhbih, Mukhtàr, 52). 47 Ya˙yà b. Wàßil (Aghànì, IV, 300, 366; V, 259); Fazzàr (VII, 113). 48 Ibn Zurzùr (Aghànì, III, 267; XIII, 199): al-Ra††àb (XIX, 314). 49 In a discussion of the metrical tendencies discernible among pre-Islamic and early Islamic poets, Dimitry Frolov argues that the poets of al-Óìra show preferences different from those of the Bedouin poets, but that this is not reflected in the anthologies made in the 'Abbàsid period, apart from the Aghànì (Frolov, “Metrical dimensions of medieval poetical anthologies: the case of al-Óìra,” in W. Madelung et al., eds., Proceedings of the 17th Congress of the UEAI (St. Petersburg: Thesa, 1997), 80–94). It is possible that a distinct musical tradition in al-Óìra has also been neglected by later sources, with the Aghànì again forming something of an exception. 50 Qafà l-Najjàr (Aghànì, III, 18, 44); al-Ra††àb al-Jiddì (XIX, 314). If the form Marzùq al-Íarràf given in the Aghànì (IV, 366; V, 259) is correct, it would indicate another trade, money changing, but I am inclined to see al-ßarràf as a corruption of al-∂arràb (lutenist). 44

336

hilary kilpatrick

The names of lesser-known women singers are uninformative. At best they include the indication of an owner or patron. Ibn Shammàs, for instance, is described as the mawlà of Khulayda, Rubay˙a and 'Uqayla, who were therefore known as al-Shammàsiyyàt.51 That there were many more singing-girls is suggested by the list of names of participants in Jamìla’s festival as well as those given by Ibn alˇa˙˙àn.52 But unlike the women to whom the Aghànì accords some space, they were presumably just performers and not composers. One group of music-making mawàlì of the Umayyad period remains to be mentioned, the Medinan mukhannathùn or effeminates.53 Their contribution to the development of Arabic music seems to have been less important than is suggested by the fact that Jamìla allotted them the second day of her three-day music festival. Only two of the seven mukhannathùn who performed then, ˇuways and al-Dalàl, appear to have enjoyed real standing as singers. ˇuways, in particular, is credited with having been the first to perform songs in the rhythmic modes and introduce the “light” modes of ramal and hazaj into Arabic music.54 Two other musicians, the Meccans Ibn Surayj and al-Gharì∂, are reported to have behaved in an effeminate fashion,55 but they are not reckoned among this group. Ibn Surayj began his career as a singer of laments, a genre performed by women, and when he quarreled with his pupil al-Gharì∂ the latter’s patrons (mawlayàt) invited him to set their elegies to music in the style of the naw˙.56 Musically speaking, then, both men were close to women. The ambiguity of this situation is reflected in Is˙àq’s reply to a question about the best singer; among the men he names Ibn Mu˙riz and among the women Ibn Surayj57—but whether he is referring to the genres Ibn Surayj excelled at, his voice, his style of singing or his behavior in general cannot be determined. At all events, as Rowson has shown, the mukhannathùn cannot be regarded as representing a transitional phase in the evolution of Arabic music from an art dominated by women to one dominated by men.

51

Aghànì, XVI, 190. Aghànì, VIII, 209; Ibn al-ˇa˙˙àn, Óàwì l-funùn, 36–37. 53 They have been discussed in detail by Rowson in “The Effeminates of Early Medina.” Ibn Khurdàdhbih also names one Meccan mukhannath, Madàr (Mukhtàr, 45). 54 Aghànì, III, 29; IV, 219. 55 Aghànì, I, 249, 274; II, 359–60. 56 Aghànì, I, 254–55; II, 360. 57 Aghànì, I, 252. 52

MAWÀLÌ

and music

337

The early 'Abbàsid period appears to have been one of developing further and refining the discoveries made during the Umayyad period. Few major innovations are mentioned; one such is the màkhùrì rhythmic cycle introduced by Ibràhìm al-Mawßilì (125–88/743–804).58 The popularization of the †unbùr, pandore or long-necked lute, which had two or sometimes three strings,59 is also an 'Abbàsid phenomenon. It is probable that the melodic modes were clearly defined at this time; Is˙àq al-Mawßilì may well have been responsible for the final codification of the system of eight melodic modes and eight rhythmic cycles, as they were used by performers.60 After the establishment of the 'Abbàsid caliphate the centre of musical gravity gradually moved east. While al-Saffà˙ (r. 132–36/ 750–54) was not averse to Óijàzì singing, al-Manßùr (c. 90–158/ 709–775) was a conservative, preferring the uncouth camel-drivers’ ˙udà" to it.61 But members of the 'Abbàsid clan in Baßra, the Banù 'Alì and al-Saffà˙’s son Mu˙ammad, were more open-minded, and several singers frequented them.62 With al-Mahdì’s (c. 127–69/745–85) accession music acquired a firm place at court. Already as a prince al-Mahdì had acquired singing-girls,63 and as caliph he summoned singers from the Óijàz to perform in Baghdàd. From then on until al-Muhtadì (r. 255–56/868–70) there was an uninterrupted series of music-loving caliphs, and high officials and courtiers shared their enthusiasm. The earliest male performers at the 'Abbàsid court show the same social profile as in the Umayyad period; in some cases they were the same people. Most are mawàlì, such as Ibn 'Abbàd al-Kàtib, Yazìd Óawrà" (d. c. 185/801), Siyà† (c. 122–69/739–85), Fulay˙ b. Abì l-'Awrà", Ya˙yà al-Makkì (c. 110–218/728–833) and Ibràhìm 58 Aghànì, V, 231–35, where Ibràhìm ascribes this rhythm to Iblìs’s inspiration. For an explanation of the term, see Sawa, Music Performance Practice in the early 'Abbasid Era 132–320 AH/750 –932 AD (Toronto: Pontifical Institute of Mediaeval Studies, 1989), 52–53. 59 Sawa, Music Performance Practice, 81–83. 60 Wright, art. “Mùsì˚ì,” EI 2. For the details of his system, see Neubauer, “Die acht ‘Wege’,” 384–94. 61 Pseudo-Jà˙iΩ, Le livre de la couronne. Kitàb al-tà[ fì a§làk al-mulùk, trans. Charles Pellat (Paris, Société d’édition “Les belles lettres,” 1954), 60–61; al-Mas'ùdì, Murùj al-dhahab wa-ma'àdin al-jawhar, 7 vols. (Beirut: Publications de l’Université Libanaise, 1966–79), § 2334; Aghànì, XV, 30. 62 Óakam al-Wàdì (Aghànì, VI, 284), 'A†arrad (III, 303), Da˙màn al-Ashqar (VI, 21), Ibràhìm al-Mawßilì (V, 159). 63 They included Maknùna, 'Ulayya’s mother (Aghànì, X, 149).

338

hilary kilpatrick

al-Mawßilì.64 It is noteworthy, however, that while the Arab tribe to which they are affiliated and often the place they come from are named, details of their ethnic background are seldom given. One exception is Ibràhìm al-Mawßilì, whose parents fled from the exactions of a harsh Umayyad governor of Fars to Kùfa.65 A few are Arabs, including the already mentioned Màlik b. Abì l-Sam˙, Mu˙ammad b. al-Ash'ath b. Fajwa al-Kùfì, and Ismà'ìl b. Jàmi' (d. before 188/804).66 Even fewer are slaves by origin. One such is al-Mu'allà b. ˇarìf, and his biography is interesting in another respect too. Of partly Arab and partly Persian origin from Kùfa, he was bought by alManßùr and given to al-Mahdì, who freed him. He had a career in the administration, being in charge of the post and luxury textile production in Khuràsàn and then briefly governor of Baßra and adjoining regions. He studied singing and the lute with Ibràhìm alMawßilì, Ibn Jàmi' and Óakam al-Wàdì (d. before 193/809), and appeared as a performer before Hàrùn al-Rashìd (149–93/766–809).67 He seems to have been the first example in the 'Abbàsid period of a man who, having had a career in public life, was prepared to appear at court as a singer. From the time of al-Rashìd on a new social group engages in singing—the princes. The first and most eminent of the music-making 'Abbàsids was Ibràhìm b. al-Mahdì (162–224/779–839), who emerged as the rival of Is˙àq b. Ibràhìm al-Mawßilì and the leader of one of the two musical factions at court. The picture of Ibràhìm in the Aghànì reflects Abù l-Faraj’s bias in favor of Is˙àq al-Mawßilì, but it is none the less clear that Ibràhìm played a very important part in musical life, having his own style of performance, training young singers and putting forward his own system of terminology for the rhythmic modes.68 Ibràhìm and Is˙àq disagreed on many matters, but one source of tension was precisely the fact that Is˙àq regarded 64

Aghànì, VI, 171; III, 251; V, 152; IV, 359; VI, 173; V, 155. Aghànì, V, 154–55. 66 Aghànì, XV, 55; VI, 289. 67 Aghànì, VI, 239; XVIII, 308; al-ˇabarì, Ta"rìkh al-rusul wa-l-mulùk, ed. M.J. de Goeje (Leiden: Brill, 1879–1901), III, 518, 521. 68 Ibràhìm’s role as a musician is discussed in J.E. Bencheikh, “Les musiciens et la poésie,” Arabica 22 (1975), 123–26. The significance of the changes he introduced in performance practice is set out in Sawa, Music Performance Practice, 187–88. His influence is shown by the fact that in the fourth/tenth century one of the two schools of singing in Baghdàd still claimed him as its founder (Aghànì, X, 70). 65

MAWÀLÌ

and music

339

himself, and was regarded by Ibràhìm, as a professional, obliged to earn his living as a musician, while Ibràhìm considered himself an amateur.69 Two 'Abbàsid caliphs were committed musicians; in the case of one, al-Wàthiq (194–232/810–847), this seems to have been his main claim to fame,70 while the other, al-Mu'ta∂id (c. 245–89/ 860–902), who was a very competent ruler too, earns Abù l-Faraj’s admiration for his skill in composing and grasp of musical theory.71 Other musically gifted princes mentioned in the Aghànì are Abù 'Ìsà b. al-Rashìd (d. 209/824), 'Abdallàh b. Mùsà al-Hàdì, 'Abdallàh b. Mu˙ammad al-Amìn, the prolific composer Abù 'Ìsà b. al-Mutawakkil and 'Abdallàh b. al-Mu'tazz (247–96/861–908); the last two in particular Abù l-Faraj holds in high esteem.72 Where the ruling family showed the way, their courtiers followed. Al-Ma"mùn’s general and later governor of Khuràsàn 'Abdallàh b. ˇàhir (182–230/798–844) was an accomplished composer, although because he was unwilling to be known as a musician he did not set his name to his settings. But when hearing one of his songs, which he had ascribed to Màlik b. Abì l-Sam˙, being performed before the caliph, he admitted that he was the author; evidently the temptation to show that he had fooled all the court musicians with his composition was irresistible.73 The audience at which he gave the show away occurred after he had returned from Egypt, and it may be that having established himself as a commander and administrator he felt he could afford to admit to his musical hobby without it affecting his standing. 'Ubaydallàh (223–300/838–913), 'Abdallàh’s son, inherited his father’s love of music and also his reluctance to talk about his activity as a composer, attributing his settings to his slave-girl Shàjì. He was knowledgeable about music theory, even if Abù l-Faraj dismisses his claim to have composed a melody combining the ten tones.74 Another member of the same family, 'Alì b. alHishàm, who was al-Ma"mùn’s commander of the expedition against 69 Aghànì, X, 142–48 quotes a correspondence between the two men which reflects the points of disagreement between them. 70 Cf. Kennedy, The Prophet and the Age of the Caliphates. The Islamic Near East from the sixth to the eleventh century (London: Longmans, 1986), 168: “No other caliph of the period has left so little trace on the history of his times and it is impossible to form any clear impression of his personality.” 71 Aghànì, IX, 344–45; X, 41–42. 72 Aghànì, X, 201–202, 276. 73 Aghànì, XII, 106, 111–12. 74 Aghànì, VIII, 374–75; IX, 40–41, 43.

340

hilary kilpatrick

Bàbak until his fall and execution in 217/832, not only owned singing-girls and followed Is˙àq al-Mawßilì’s writings on music, but was also an occasional composer.75 Abù Dulaf al-Qàsim b. 'Ìsà al'Ijlì (d. c. 225/840) and his brother Ma'qil were also composers, and seem to have had less hesitation in admitting the fact than the ˇàhirids 'Abdallàh and 'Ubaydallàh, quite possibly because they were associated with al-Wàthiq.76 The number of male singers of slave origin seems to have remained steady as time passed. Ma'bad al-Yaq†ìnì, who was of partly African origin, had his early training in Medina before being brought to Baghdàd. He performed before Hàrùn al-Rashìd, as did Mu˙ammad b. Óamza b. Nußayr the Attendant (waßìf ), who is described as a mawlà of al-Manßùr.77 The career of 'Aqìd, a mawlà of Íàlih b. alRashìd, spanned the period from al-Rashìd to al-Wàthiq.78 Another musician, Yanshù, recounts how he and his companion Mahmùm were purchased by Abù A˙mad b. al-Rashìd, sent to Baghdàd to be trained by Is˙àq, and later introduced to sing before al-Wàthiq.79 Undoubtedly the most successful of the singers of servile origin was Mukhàriq (d. 231/845 or 232/846), the Kùfan butcher’s son, who after attracting 'Àtika bint Shuhda’s attention because of his fine voice received his first musical training from her. He then had several owners and teachers before being acquired by al-Rashìd, who freed him after a particularly successful performance; he later described himself as the mawlà of the song he had sung then.80 As for those musicians who started life neither as slaves, nor as princes, nor belonged to an Arab tribe, their designation as mawàlì becomes less frequent and seems to change its meaning. Mu˙ammad b. al-Óàrith b. Buskhunnar (d. after 232/847) came of a family of the Persian nobility and his father was one of al-Hàdì’s governors; 75 For 'Alì’s career, set out by al-Ma"mùn to explain his execution, see Ibn Abì ˇàhir ˇayfùr, Kitàb Baghdàd, ed. 'Izzat al-'A††àr al-Óusaynì (Cairo: Maktab Nashr al-Thaqàfa l-Islàmiyya, 1368/1949), 146. 'Alì was the owner of Mutayyam alHishàmiyya, the subject of the article in Aghànì, VII, 293–306 and of the younger Badhl (VII, 297–98). The Aghànì quotes Is˙àq’s letter to him about “work in progress” (XVII, 111–12). His setting of 'Abdallàh b. Ja˙sh’s poetry introduces the article on the latter (XIX, 211). 76 Aghànì, VIII, 248, 251–52; XXI, 92. 77 Aghànì, XIV, 116; XV, 356. 78 Aghànì, XVIII, 69; V, 394. 79 Aghànì, V, 293–95. 80 Aghànì, XVIII, 336–38; 340–41.

MAWÀLÌ

and music

341

they are described as mawàlì of al-Manßùr, which Abù l-Faraj understands as a relationship of service, nor emancipation.81 The tendency of the caliph to attract all relationships of walà" to himself can be seen in an exchange between al-Rashìd and Ibràhìm al-Mawßilì, when the caliph asked the musician how he came to be a mawlà of Tamìm. Ibràhìm, whose father died when he was very young, explained that he had been taken care of by members of Tamìm who were his foster-kin, and they had looked after him well. AlRashìd commented: “In that case you must be my mawlà.”82 The assimilation of walà" in all its forms to khidma can be sensed on another occasion. Al-Rashìd suddenly heard that 'Abdallàh b. al'Abbàs al-Rabì'ì (d. c. 247/861), the great-grandson of al-Rabì' b. Yùnus, al-Manßùr’s freed slave and chamberlain, was a budding composer, and ordered his grandfather al-Fa∂l to bring him to court. “Damn you, Fa∂l,” he reproached him, “you have a grandson who can sing and is a good enough musician to compose two songs which Is˙àq and the other singers admire and the singing-girls teach each other, and you haven’t told me about it. It looks as though you thought he was above serving me in this matter.” 'Abdallàh b. al'Abbàs is not introduced as a mawlà at the beginning of the article on him, but everyone must have been aware of his inherited status. Poor Fa∂l could not refuse, although he considered it a social disaster that his grandson should appear at court as a singer.83 The designation of mawlà continued to be used for some musicians, but as time passed and ties of loyalty diversified it became less meaningful. 'Allùya (d. 236/850), who was born during al-Rashìd’s caliphate, was of Soghdian origin, his forefather having been among those taken prisoner by al-Walìd b. 'Uthmàn b. 'Affàn, which made him a mawlà of the Umayyads. But he himself stood in a relationship of walà" al-khidma to the 'Abbàsids.84 No walà", however, is mentioned Aghànì, XII, 48 (wa-a˙sabuhu walà"a khidmatin là walà"a 'itq). Aghànì, V, 155–56 (Fa-mà aràka idhan illà mawlàya). This presumably means that because al-Rashìd, too, had been a generous patron to Ibràhìm and looked after him well he now had the right to his walà". Whereas the bond of walà" between Mu˙ammad b. al-Óàrith’s family and al-Manßùr can be understood as a tie of loyalty between men with political power (the title came to be used as an honorific; cf. Crone, art. cit.), this can scarcely be the case with Ibràhìm al-Mawßilì. 83 Aghànì, XIX, 222–23. 84 Aghànì, XI, 333; IV, 354, where he rashly points out to al-Ma"mùn how unfavorably his own situation as a mawlà of the 'Abbàsids contrasts with that of Ziryàb, originally a mawlà of the 'Abbàsids who has moved to Spain to serve the Umayyads. 81 82

342

hilary kilpatrick

for the Kùfan Sulaym b. Sallàm, whose father was a dà'ì and confident of Abù Muslim, nor for the later 'Abdallàh b. Abì l-'Alà" from Sàmarrà", a pupil of Is˙àq’s, the fuller’s son Sulaymàn b. al-Qassàr (d. c. 255/869) who performed before al-Mu'taßim (c. 179–227/ 794–842) and al-Mu'tazz (231–55/846–69), or 'Umar al-Maydànì and al-Masdùd (d. before 279/892) the butcher’s son, both of whom were from Baghdàd.85 They were not mawàlì, just men of the people. In this period, as under the Umayyads, the status of women singers does not evolve entirely parallel with that of men. The phenomenon of the mawlàh, whether in the sense of “freedwoman” or “attached to an Arab clan” who otherwise lived an independent life almost disappears. An example of such a figure from the beginning of the period is the Medinan 'Àtika bint Shuhda, daughter of al-Walìd b. Yazìd’s singing-girl, who settled in Baßra and is said to have discovered Mukhàriq.86 Much later 'Ubayda, a highly regarded pandore-player whose father was a mawlà of one of 'Abdallàh b. ˇàhir’s companions, picked up singing from a frequent visitor to her father and then pursued her career independently, only turning to her father’s patron for help in a difficult moment.87 Most 'Abbàsid women singers, like many of their colleagues earlier, started their lives as slaves, bought by men who were attracted by their appearance and believed they were gifted enough to justify the investment required for their training. The Medinan Baßbaß, the best singer among Ibn Nufays’s slave-girls, who failed to win alManßùr over to al-ghinà" al-mutqan and Sallàma al-Zarqà" and Rubay˙a, bought from Ibn Ràmìn by Ja'far b. Sulaymàn b. 'Alì and his brother Mu˙ammad respectively, are among the earliest 'Abbàsid examples of this type.88 But two of the most famous singers, 'Arìb and Shàriya, despite being first slaves and later freedwomen, claimed to have been free at birth. 'Arìb maintained that she had been sold by her nurse after the execution of her father, Ja'far b. Khàlid al-Barmakì, while Shàriya had either been stolen from her parents, not recognized by her father, a member of Quraysh, and so relegated to the slave status of her mother, or else sold by her own mother.89 Whether these 85

Aghànì, VI, 164; XXIV, 1; XIV, 112; XXIII, 140; XX, 228. Aghànì, V, 260–61; XVIII, 336. 87 Aghànì, XXII, 208–209. 88 Aghànì, XV, 27, 29–31, 63, 71. 89 Stigelbauer, Die Sängerinnen am Abbasidenhof um die Zeit des Kalifen al-Mutawakkil nach dem Kitàb al-A©ànì des Abù l-Fara[ al-Ißbahànì und anderen Quellen dargestellt (Ph.D. diss., Vienna, 1975), 89–90. 86

MAWÀLÌ

and music

343

claims were true or simply invented to enhance the prestige of the singer in question, they suggest that such things could happen, and so perhaps some of the singing slave-girls had been born free. When al-Mahdì integrated music into court life the best singinggirls gravitated to the capital, ending up after several moves in the palaces of the courtiers and often finally those of the caliphs. In the early 'Abbàsid period some of them came from the Óijàz, receiving their early training in Baßra before going to Baghdàd.90 Danànìr and Badhl were two such musicians. Their careers also illustrate another aspect of the situation of jawàrì at the 'Abbàsid court, the idea that the ruler had a right to the most gifted girls. There is a suggestion of this in the story of al-Rashìd’s repeated visits to Ya˙yà b. Khàlid al-Barmakì to listen to Danànìr; fortunately for Ya˙yà, Zubayda complained to the rest of the family about her husband’s behavior, obliging him to declare that he was only interested in the girl’s singing.91 The idea becomes reality in al-Amìn’s case; having heard about the accomplishments of his cousin Ja'far b. Mùsà al-Hàdì’s slave-girl Badhl, he went to see her and offered to buy her. Ja'far refused either to sell her or to give her away, but al-Amìn got him drunk and went off with Badhl, later compensating him with a large sum of money.92 The ruler could also appropriate the singing-girls of his predecessor; al-Mu'taßim took over those of al-Ma"mùn, and al-Mutawakkil those of al-Wàthiq.93 When a caliph died, those who left the palace might either make advantageous marriages, such as Duqàq and the older Farìda, or refuse all offers and live on the presents they had received and what they earned as performers, as Badhl did.94 Badhl is an example of a singer who after the death of her protector, the caliph, remained an important figure in musical life.95 90 Stigelbauer, Die Sängerinnen, 86–87, notes the importance of the Óijàz as a source of singing-girls. 91 Aghànì, XVIII, 65, 67. 92 Aghànì, XVII, 75–76. This is one of three singers called Badhl in the Aghànì, the other two, Badhl al-Kabìra and Badhl al-Saghìra, ending up in al-Mu'taßim’s palace (VII, 297–98). The only sure way to keep a singing-girl was to have a child with her, and this is how 'Alì b. Hishàm resisted al-Ma"mùn’s demands to buy Mutayyam—which according to some was what prompted al-Ma"mùn to cause his downfall and execution (VII, 296). 93 Aghànì, VII, 298; IV, 115. 94 Aghànì, XII, 282; IV, 113; XVII, 75. 95 Aghànì, XVII, 77–79 mentions her performing at court, discussing with leading musicians, and compiling a book of songs.

344

hilary kilpatrick

Like their male relatives, 'Abbàsid princesses took to singing and composing. The only one whose activities are known in any detail is 'Ulayya bint al-Mahdì (160–210/777–825) who seems to have been at least as good a musician as her brother Ibràhìm.96 Another one mentioned en passant in the Aghànì is Khadìja bint al-Ma"mùn.97 Two daughters of al-Rashìd, Umm Abìhà and Óamda, Umm 'Abdallàh bint 'Ìsà b. 'Alì, Laylà bint 'Alì b. al-Mahdì and Fà†ima bint 'Abdallàh b. Mùsà are also described as musicians.98 In a court as full of musicmakers and music-lovers as the third/ninth century 'Abbàsid one, one may assume that there were more singing princesses, but their names have not been recorded; when men of the standing of 'Abdallàh b. ˇàhir and his son 'Ubaydallàh were reluctant to be known as musicians, it is hardly surprising if royal women, and even more their menfolk, considered that having their names linked to music harmed their reputation. The history of writings on music starts with the song collection. Yùnus al-Kàtib was the first to compile a song book; his example was followed by Is˙àq al-Mawßilì and his son Óammàd, Ya˙yà alMakkì and his son A˙mad (d. 248/862), Óabash, Badhl, Danànìr, 'Amr b. Bàna (d. 278/891) and others. Authors of song books from the latter half of the third/ninth century are Ja˙Ωa al-Barmakì (224– 324/839–936) and Ya˙yà b. 'Alì al-Munajjim (241–300/855–912).99 Apart from these general collections, the repertoire of individual singers was also noted.100 Musicological questions appear first to have been addressed by alKhalìl b. A˙mad (c. 100–175/719–791), author of a treatise on musical metres where he proposes a synthesis between the theory of 'arù∂ 96 Aghànì, X, 162–85; 163 for this appreciation of her. Cf. Neubauer, art. “Ulayya bint al-Mahdì,” EI 2. It is noteworthy that al-Íùlì arranged 'Ulayya’s poetry which she had set to music not according to a conventional system for a dìwàn but according to the rhythmic modes (al-Íùlì, Ash'àr awlàd al-khulafà" wa-akhbàruhum, ed. J. Heyworth Dunne, repr. 3rd ed. (Beirut: Dàr al-Masìra, 1401/1982), 64–76). 97 Aghànì, XVI, 16. 98 Ibn al-ˇa˙˙àn, Óàwì l-funùn, 118. 99 Ibn al-Nadìm, Al-Fihrist, 157–62, mentions Is˙àq al-Mawßilì and his son Óammàd, the Banù Munajjim, Óabash (whom he refers to as Óubaysh al-Íìnì), Abù Óashìsha and Ja˙Ωa. Abù l-Faraj also speaks of the song books of Ya˙yà and A˙mad al-Makkì (the father’s being as unreliable as the son’s was valuable), Badhl and Danànìr (Aghànì, VI, 175; XVI, 311; XVII, 75; XVIII, 65). He draws on them all for performance indications. 100 Neubauer, Musiker, 35–36, gives a list of these books, which he qualifies as aide-mémoire.

MAWÀLÌ

and music

345

and that of ìqà'.101 But the earliest significant contribution is that of Abù Yùsuf al-Kindì (d. c. 252/866). He has left treatises on several subjects, including the analysis of the scale, the structure of musical metrics, composition and the description of the lute. In general he was concerned to reconcile Greek philosophical thought with the practice and theory being evolved by Arab musicians.102 He put forward the idea of correspondences between the four strings of the lute and other tetradic phenomena, such as the seasons, the humors and the elements; this was taken further by later philosophers. His pupil Abù l-'Abbàs al-Sarakhsì (c. 220–286/835–99), al-Mu'ta∂id’s tutor and later nadìm, also wrote on music, though all that has survived of his works are a few quotations and the titles given in the Fihrist. 103 The short text by another nadìm of this period, Ibn Khurdàdhbih, Mukhtàr min kitàb al-lahw wa-l-malàhì, has already been mentioned; apart from giving names of singers, it discusses musical instruments and the place of music among different peoples. Another writer on musical instruments, especially the lute, is the lexicographer al-Mufa∂∂al b. Salama (before 215–90/830–903), author of Kitàb al-'ùd wa-l-malàhì.104 Finally, Ibn Abì l-Dunyà (208–281/823–94) initiated the genre of condemnations of music, instruments and singing-girls with his collection of ˙adìths entitled Dhamm al-malàhì.105 Apart from the compilers of songbooks, these authors are too late for the designation mawàlì to be appropriate; what is important is that except for al-Khalìl they were all attached more or less closely to the court. As far as the early collectors of songs are concerned, they were practicing musicians and, like their fellows, mainly mawàlì. The specifically Arab contribution is represented, very modestly, by al-Khalìl b. A˙mad and significantly by al-Kindì. 101 How many works al-Khalìl devoted to music, and what their titles were, is uncertain. Cf. Neubauer, “Al-flalìl b. A˙mad,” 260–63. 102 For his method, see Neubauer, “Die acht ‘Wege’,” 381–82. Al-Kindì’s musical writings are passed over in silence in the EI 2 article on him. They are listed in Farmer, The Sources of Arabian Music (Leiden: Brill, 1965), 8–10. 103 Rosenthal, art. “al-Sara¶sì, Abù l-'Abbàs A˙mad b. al-ˇayyib,” EI 2. 104 “The Kitàb al-malàhì of Abù ˇàlib al-Mufa∂∂al b. Salama,” ed. and trans. J. Robson, Journal of the Royal Asiatic Society 1938, 231–49. 105 For his life and works, see Weninger, Qanà'a (Genügsamkeit) in der arabischen Literatur anhand des Kitàb al-Qanà'a wa-l-ta'affuf von Ibn Abì d-Dunyà (Berlin: Klaus Schwarz Verlag, 1992), 61–71; for the text “Tracts on listening to music, being ‹amm al-malàhì by Ibn Abì l-Dunyà and Bawàriq al-ilmà' by Majd al-Dìn al-ˇùsì al-Ghazàlì,” ed and trans. James Robson, Royal Asiatic Society (1938). According to Weninger (Qanà'a, 64, 67) this is an abridgement of the original text.

346

hilary kilpatrick

The above sketch will have demonstrated that mawàlì, however the term is understood, made an essential contribution to the development of art music in the Umayyad and early 'Abbàsid period. But it is clear that this development is not due only to the gifts and hard work of the individuals concerned. Two other factors must be mentioned too. First is the abandonment of a sexual division of singing and the involvement of men and women alike in the elaboration and transmission of al-ghinà" al-mutqan.106 When men engaged in performing art music they raised its status and brought it further into the public sphere. According to the sources, they were more innovative and interested in questions of theory and codification. Women, on the other hand, provided continuity with the pre-Islamic tradition. Slave-girls (of unknown origin) had been the main performers of pre-Islamic entertainment music, and despite the emergence of some prominent mawlayàt under the Umayyads and music-making princesses under the 'Abbàsids, most singing-girls continued to be slaves.107 Nonetheless they made a recognized contribution to the preservation and evolution of the musical tradition. Two of them are named as authors of songbooks, and many more as teachers of both men and women. Women, just as much as men, were transmitters of the oral tradition which formed an indispensable complement to the songbooks with their summary noting of performance indications.108 Abù l-Faraj mentions checking information about songs with mughanniyàt al-qußùr and turning to named singing-girls for other information.109 It seems that music is the domain of medieval Arabic culture to which women contributed most, even if they tended to

106

The realities (and images) of gender roles in early Islamic society have been examined recently by Julia Bray in a penetrating article, “Men, women and slaves in Abbasid society” in Leslie Brubaker and Julie M.H. Smith (eds.), Gender in the Early Medieval World. East and West, 300 –900 (Cambridge: Cambridge University Press, 2004) 121–46. 107 The obvious reason for this is that to sing in public was considered incompatible with the honor of a free woman—and her male relatives. There is probably another reason too. Being a good singer was a full-time job, requiring keeping one’s voice in training, practicing the lute regularly and learning or composing new songs. A free woman, expected to run a household and bear and bring up children, would not have had the leisure to devote herself to these pursuits. 108 For this question, cf. Kilpatrick, “The transmission of songs in mediaeval Arabic culture,” in U. Vermeulen and D. De Smet, eds., Philosophy and the Arts in the Islamic World. Proceedings of the 18th Congress of the UEAI, Leuven 1996 (Leuven: Peeters, 1998), 73–82. 109 Aghànì, IX, 61; VI, 175; XIV, 114.

MAWÀLÌ

and music

347

be transmitters rather than innovators, and where their contribution was most readily recognized.110 Al-Óasan b. A˙mad affirms that among the rare musicians combining the four qualities essential to their art, almost all are women. And he quotes a remark, attributed to Is˙àq al-Mawßilì, comparing singing to a book which men compose (ta˙rìr) and women write up (inshà").111 At all events the cooperation between men and women in this field was fruitful. Second, singing would never have acquired the status it did under the early 'Abbàsids without the involvement of the ruling family not only as patrons but also as performers. The caliphs may have derived their ideas about the place of music at court from the imperial Persian ideology, but they took them further. For the Sassanian emperors are not described as singing, only as listening to singers. If Ibràhìm b. al-Mahdì and al-Wàthiq had not been enthusiastic (as well as gifted) composers and singers, music might have remained a kind of spectator sport for most courtiers—or at least an activity they would not admit to. Instead, music making was legitimized through the ruling family’s involvement in it. Arabic music reached the level it did at the 'Abbàsid court thanks to princes and female slaves as well as mawàlì.112

Bibliography Primary sources al-Óasan b. A˙mad b. 'Alì al-Kàtib. Kitàb kamàl adab al-ghinà", ed. 'Abd al-Malik Gha††às Khashaba. Cairo: al-Hay"a al-Mißriyya al-'Àmma li-l-Kitàb, 1395/1975; French trans. Amnon Shiloah, as La perfection des connaissances musicales. Paris: Paul Geuthner, 1972. [Ibn Abì l-Danyà]. Tracts on listening to music, being ‹amm al-malàhì by Ibn Abì l-Dunyà and Bawàriq al-ilmà' by Majd al-Dìn al-ˇùsì al-Ghazàlì, trans. and ed. James Robson. London: Royal Asiatic Society, 1938. Ibn Abì ˇàhir ˇayfùr, A˙mad. Kitàb Baghdàd, ed. 'Izzat al-'A††àr al-Óusaynì. Cairo: Maktab Nashr al-Thaqàfa al-Islàmiyya, 1368/1949.

110

Stigelbauer, Die Sängerinnen, 2. Al-Óasan b. A˙mad, Kamàl adab al-ghinà", 119; trans. 166–67 (where nusuj is corrected to nusakh). 112 The close connection between art music and the court (and the sums lavished on musicians, men and women) also fuelled opposition to this music in some religious circles, as can be seen from some ˙adìths in Ibn Abì l-Dunyà’s Dhamm alMalàhì, where it is assimilated to other kinds of luxury. 111

348

hilary kilpatrick

Ibn Óazm, 'Alì b. A˙mad. Jamharat ansàb al-'arab, ed. 'Abd al-Salàm Mu˙ammad Hàrùn, 5th ed. Cairo: Dàr al-Ma'àrif, 1982. Ibn Khurdàdhbih, Abù l-Qàsim 'Ubaydallàh b. A˙mad. Mukhtàr min kitàb al-lahw wa-l-malàhì, ed. Ighnà†iyùs 'Abduh Khalìfa. Beirut: Dàr al-Mashriq, 1969. Ibn al-ˇa˙˙àn al-Mùsìqì, Abù l-Óusayn Mu˙ammad b. al-Óasan. Óàwì l-funùn wasalwat al-ma˙zùn, ed. Eckhard Neubauer. Frankfurt am Main: Institute for the History of Arabic-Islamic Science, 1990. al-Ißbahànì, Abù l-Faraj 'Alì b. al-Óusayn. Kitàb al-aghànì, 24 vols. Cairo: Dàr alKutub and al-Hay"a al-'Àmma al-Mißriyya li-l-Kitàb, 1927–74. al-Mas'ùdì, 'Alì b. al-Óusayn. Murùj al-dhahab wa-ma'àdin al-jawhar, 7 vols. Beirut: Publications de l’Université Libanaise, 1966–79. [al-Mufa∂∂al b. Salama] ed. and trans. J. Robson. “The Kitàb al-malàhì of Abù ˇàlib al-Mufa∂∂al b. Salama,” Journal of the Royal Asiatic Society (1938), 231–49. Pseudo-Jà˙iΩ. Le livre de la couronne. Kitàb al-tà[ fì a§là˚ al-mulùk, trans. Charles Pellat. Paris: Société d’édition “Les belles lettres,” 1954. al-Íùlì, Abù Bakr Mu˙ammad b. Ya˙yà. Akhbàr Abì Tammàm, ed. Khalìl Mu˙ammad 'Asàkir, Mu˙ammad 'Abduh 'Azzàm and NaΩìr al-Islàm al-Hindì, 3rd ed. Beirut: Dàr al-Àfàq al-Jadìda, 1400/1980. id., Ash'àr awlàd al-khulafà" wa-akhbàruhum, ed. J. Heyworth Dunne, 3rd ed. Beirut: Dàr al-Masìra, 1401/1982. Secondary sources al-Asad, Nàßir al-Dìn. Al-Qiyàn wa-l-ghinà" fì l-'aßr al-jàhilì, 3rd ed. Beirut: Dàr alJìl, 1988. Bencheikh, J.E. “Les musiciens et la poésie,” Arabica 22 (1975), 114–52. Farmer, Henry George. The sources of Arabian music. Repr. Leiden: Brill, 1965. Frolov, Dimitry. “Metrical dimensions of medieval poetical anthologies: the case of al-Óìra,” in W. Madelung et al., eds., Proceedings of the 17th Congress of the UEAI. St. Petersburg: Thesa, 1997, 80–94. Kennedy, Hugh. The Prophet and the age of the caliphates. The Islamic Near East from the sixth to the eleventh century. London: Longmans, 1986. Kilpatrick, Hilary. “The transmission of songs in mediaeval Arabic culture,” in U. Vermeulen and D. de Smet, eds., Philosophy and the arts in the Islamic world. Proceedings of the 18th Congress of the UEAI, Leuven 1996. Leuven: Peeters, 1998, 73–82. Neubauer, Eckhard. “Die acht ‘Wege’ der arabischen Musiklehre und der Oktoechos,” Zeitschrift für Geschichte der arabisch-islamischen Wissenschaften 9 (1994), 373–414. id., “Al-flalìl, die Töne und die musikalischen Metren,” Zeitschrift für Geschichte der arabisch-islamischen Wissenschaften 10 (1995–96), 255–323. id., Musiker am Hof der frühen 'abbàsiden. Ph.D. diss., Franktfurt am Main, 1965. Rowson, Everett K. “The effeminates of early Medina,” Journal of the American Oriental Society 111 (1991), 671–93. Sawa, George Dimitri. Music performance practice in the early 'Abbasid era 132–320 AH/750–932 AD. Toronto: Pontifical Institute of Mediaeval Studies, 1989. Stigelbauer, Michael. Die Sängerinnen am Abbasidenhof um die Zeit des Kalifen al- Mutawakkil nach dem Kitàb al-A©ànì des Abù l-Fara[ al-Ißbahànì und anderen Quellen dargestellt. Ph.D. diss., Vienna, 1975. Weninger, Stefan. Qanà'a (Genügsamkeit) in der arabischen Literatur anhand des Kitàb alQanà'a wa-l-ta'affuf von Ibn Abì d-Dunyà. Berlin: Klaus Schwarz Verlag, 1992. Wright, Owen. “Music and verse,” in A.F.L. Beeston et al., eds., Arabic literature to the end of the Umayyad period, The Cambridge History of Arabic Literature. Cambridge: Cambridge University Press, 1983, 433–59.

MAWÀLÌ AND ARABIC POETRY: SOME OBSERVATIONS Geert Jan van Gelder

The early Arab critics and anthologists used many different and often surprising criteria in their classifications of poets. Here are some titles:1 Man nusiba ilà ummihi min al-shu'arà" (Poets Called After Their Mothers) by al-Madà"inì, similar works by Ibn al-A'ràbì, Ibn Óabìb and al-Sukkarì, Al-Mughtàlùn (Those [Poets] Who Were Killed) by Ibn Óabìb, al-Mu'ammarùn (Long-Lived [Poets]) by Abù Óàtim al-Sijistànì, al-Mu˙ammadùn min al-shu'arà" (Poets Called Mu˙ammad) by al-Qif†ì, Man ismuhu 'Amr min al-shu'arà" (Poets Called 'Amr) by Ibn al-Jarrà˙, similar works by Abù Îam∂am al-Bakrì and al-Jà˙iΩ, Man 'ashiqa min alshu'arà" (Poets Who Were In Love), by a certain al-Jalùdì, and alMu'tadhirìn ([Poets] Who Apologized) by Ibn Abì ˇàhir ˇayfùr. There are books on the poets of a particular tribe or a particular region or country, books on poetesses, on slaves or slave-girls who made poetry, on leading people (mulùk) who were also poets, and on “modern” poets (mu˙dathùn).2 What I have not found is a book, treatise or even a chapter entitled “Poets Who Were Mawàlì ” or “Mawàlì Who were Poets,” or “Persians Who Made Arabic Poetry.” Sezgin3 mentions as a possible candidate a Kitàb al-mawàlì by alJà˙iΩ; this work is lost, but Sezgin is surely right in supposing that it dealt at least partly with poets and poetry. Yet none of the references to it made by al-Jà˙iΩ himself 4 makes it likely that it was specifically about poets. He says that he was criticized because in that work he had given the mawàlì less than their due and credited the Arabs with more than they were entitled to, although he himself had merely wished to put the mawàlì in their proper place, stating 1 All can be found in Fuat Sezgin, Geschichte des arabischen Schrifttums, Bd. II: Poesie bis ca. 430 H. (henceforward GAS, II) (Leiden: Brill, 1975), II, 98–101. 2 On the last category, see GAS, II, 439–40. 3 GAS, II, 97, 99. 4 Al-Jà˙iΩ, Al-Óayawàn, ed. 'Abd al-Salàm Mu˙ammad Hàrùn, 8 vols. (Cairo: Mu߆afà l-Bàbì l-Óalabì, 1965–69), I, 5; al-Jà˙iΩ, Rasà"il, ed. 'Abd al-Salàm Mu˙ammad Hàrùn, 4 vols. (Cairo: Maktabat al-Khànjì, 1964–79), II, 22 (end of al-Nàbita).

350

geert jan van gelder

their virtues and shortcomings. A quotation from this Kitàb al-mawàlì in al-'Iqd al-farìd, at least, discusses Abù Nuwàs as one of the mawàlì. He is described as min aqdar al-nàs 'alà l-shi'r wa-a†ba'ihim fìh, “one of the most capable and most natural of all poets.”5 Abù Nuwàs (d. c. 199/814) may serve as an example that might go some way towards an explanation why the category of “mawàlì poets” is conspicuously absent from traditional criticism. His status as client depended on the fact that his great-grandfather, al-Íabbà˙, presumably a Persian,6 became a mawlà of al-Jarrà˙ b. 'Abdallàh alÓakamì who was governor of Khuràsàn under 'Umar b. 'Abd al'Azìz and who was killed in 112/730 fighting the Khazar Turks in Armenia.7 We can assume that Abù Nuwàs’s grandfather 'Abd alAwwal and his father Hàni", who served in the army of the last Marwànid ruler, had become Arabized; Abù Nuwàs called himself al-Óakamì, after the South Arabian tribe of al-Jarrà˙ b. 'Abdallàh, and he was a strong partisan of the South Arabian cause. From more than one poem it appears that he considered himself to be a “South Arabian,” and he was not the only one. Thus Abù 'Ubayda is reported to have said, “Al-Yaman has taken both serious and jesting poetry: Imra" al-Qays took care of the serious and Abù Nuwàs of the jesting”; and a certain Abù l-Óasan al-ˇùsì said, “The three poets of al-Yaman are Imra" al-Qays, Óassàn, and Abù Nuwàs.”8 Interestingly, the great Abù Nuwàs scholar Ewald Wagner, after having pointed out the background of the poet in as much detail as is available, goes on to say that “from the paternal side he was an Arab, from the maternal side a Persian”;9 he thus literally followed the well-known maxims al-walà" lu˙ma ka-lu˙mat al-nasab “clientship is an interconnection like that of descent” and mawlà l-qawm minhum “a client of a tribe is one of them.”10 5 Ibn 'Abd Rabbih, Al-'Iqd al-farìd, ed. A˙mad Amìn et al., 7 vols. (Cairo: Dàr al-Kutub, 1948–53), VI, 77. 6 Even though none of the sources states this explicitly, I believe; the pre-Islamic name al-Íabbà˙ (attested eight times in Caskel, ]amharat an-nasab: das genealogische Werk des Hi“àm ibn Mu˙ammad al-Kalbì, 2 vols. [Leiden: Brill, 1966]) does not sound like the kind of names normally adopted by Persian converts. 7 Wagner, Abù Nuwàs: Eine Studie zur arabischen Literatur der frühen 'Abbàsidenzeit (Wiesbaden: Harrassowitz, 1965), 16. 8 Abù Nuwàs, Dìwàn, ed. Ewald Wagner, I (Cairo/Wiesbaden: Lajnat al-Ta"lìf and Franz Steiner, 1958), 11, cf. Wagner, Abù Nuwàs, 20. 9 Wagner, Abù Nuwàs, 137: “Von väterlichen Seite war er Araber, von mütterlicher Perser.” 10 Attributed to the Prophet in al-Jà˙iΩ, Al-Nàbita, in his Rasà"il, II, 21, and see Goldziher, Muhammedanische Studien, 2 vols. (Halle, 1889–90), I, 107.

MAWÀLÌ

and arabic poetry

351

Since the paternal side was ultimately decisive, what then was Abù Nuwàs, a Persian or an Arab? Race and genetics are obviously irrelevant, perceived ethnic identity is all, which is rarely “pure.” In some poems Abù Nuwàs fuses the South Arabian cause with pride in his Persian past, which, together with his many stabs at Bedouin culture, earned him the sobriquet of shu'ùbì, undeservedly.11 To the Arab genealogists tribal affiliations were important, as was someone’s status as a mawlà. The early works on poets—I have scanned through Ibn Qutayba’s al-Shi'r wa-l-shu'arà", Ibn al-Mu'tazz’s ˇabaqàt al-shu'arà" al-mu˙dathìn, Ibn al-Jarrà˙’s al-Waraqa, al-Íùlì’s alAwràq, Abù l-Faraj al-Ißfahànì’s al-Aghànì, and al-Marzubànì’s Mu'jam al-shu'arà"12—regularly give information on these matters, or perhaps one should say irregularly. Al-Marzubànì mentions some fifty poets explicitly as mawàlì, but in many cases persons that are obviously Persians or non-Arabs are not given a walà" relationship. Some of these were Christians and therefore no mawàlì,13 but others were Muslims.14 'Abbàsid poets are often not given a tribal affiliation, so that on the basis of al-Marzubànì’s text we cannot say whether they were Arabs or mawàlì. To give a few examples of poets called Mu˙ammad: (p. (p. (p. (p. (p. (p. (p. (p. (p. (p. (p. (p.

354) 358) 359) 360) 366) 369) 389) 395) 397) 402) 403) 424)

Mu˙. b. Umayya b. Abì Umayya (a mawlà, cf. GAS, II, 607) Mu˙. b. 'Alì al-S'ìnì (cf. GAS, II, 601) Abù Shihàb Mu˙. b. Mihrawayh al-Baßrì Abù Muslim al-Khalaq al-Baßrì Mu˙. b. Íabà˙ Mu˙. b. Ma'rùf al-Baghdàdì Abù l-'Abbàs Mu˙. b. al-Qàsim al-Dimashqì Mu˙. b. 'Arùs al-Kàtib al-Shìràzì Abù Na'àma Mu˙. b. al-Daqìqì (GAS, II, 537) Mu˙. b. Ibràhìm al-Jurjànì Mu˙. b. 'Imràn, “min ahl Ißbahàn” Abù Ja'far Mu˙. b. Ya'qùb al-Mithqàl al-Wàsi†ì (GAS, II, 603) Abù l-'Abbàs Mu˙. b. 'Imràn al-Óalabì

11 Wagner, Abù Nuwàs, 133–42, and very thoroughly Albert Arazi, “Abù Nuwàs fut-il “u'ùbite?” Arabica 26 (1979), 1–61. 12 Ibn Sallàm al-Juma˙ì’s ˇabaqàt fu˙ùl al-shu'arà" (ed. Ma˙mùd Mu˙ammad Shàkir (Cairo: Dàr al-Ma'àrif, 1952)) is less relevant here, since it mentions only one mawlà poet (Nußayb), while Ziyàd al-A'jam, of uncertain descent, is included but without any mention of his ethnic and social status. 13 E.g., Abù Mùsà 'Ìsà b. Farrukhshàh al-kàtib, see al-Marzubànì, Mu'jam alshu'arà", ed. 'Abd al-Sattàr Farràj (Cairo: 'Ìsà l-Bàbì l-Óalabì, 1960), 100. 14 E.g., several members of the Banù l-Munajjim (al-Marzubànì, Mu'jam, 141–42, 156, 493–94) or 'Alì b. al-'Abbàs al-Nawbakhtì (ibid., 155–56).

352

geert jan van gelder (p. 429) Abù 'Abdallàh Mu˙. b. A˙mad al-Mufajja' al-Baßrì (GAS, II, 509) (p. 431) Abù Bakr Mu˙. b. Ya˙yà (. . .) b. Íùl

As is clear from this list, tribal nisbas, still dominant in al-Marzubànì’s lexicon of poets, are gradually being superseded by geographical nisbas, or non-Arabic family names, such as in the case of the lastmentioned, who is the well-known Abù Bakr al-Íùlì, one of alMarzubànì’s teachers. There are many poets whose status as either a mawlà or not is dubious, and others known to be mawàlì but their precise affiliation being uncertain. Thus Bashshàr b. Burd is said to be a mawlà of Banù 'Aqìl or Banù Sadùs;15 Mùsà Shahawàt was a mawlà either of Banù Sahm or 'Adì b. Ka'b (to which Sahm belonged) or Banù Taym b. Murra;16 Dàwùd b. Salm was mawlà of Àl Abì Bakr, or of Àl ˇal˙a, or of Banù Taym b. Murra (to whom both Abù Bakr and ˇal˙a belonged);17 'Abdallàh al-Khayyà† was mawlà of Quraysh or of Hudhayl;18 'Alì b. Abì Kathìr was mawlà of Banù Asad or Banù Taym Allàt b. Tha'laba.19 Abù l-'Atàhiya is generally known as a mawlà of 'Anaza, but his son Mu˙ammad claimed that they belonged to 'Anaza itself;20 Yùsuf b. al-Óajjàj al-Íayqal was either a mawlà or perhaps a full member of Thaqìf;21 Mu˙ammad b. Yasìr al-Riyàshì was a mawlà of Banù Riyàsh, but others say that he belonged to them ßalìbatan;22 Thàbit Qu†na belonged to Banù Asad b. al-Óàrith, but others said that he was their mawlà;23 the poet al-Óazìn, according to al-Wàqidì, belonged to Kinàna ßalìbatan, but according to 'Umar b. Shabba he was a mawlà.24 Musàwir al-Warràq belonged to 'Adwàn (Qays b. 'Aylàn); it is also said that he was a mawlà of them.25 Muslim Ibn al-Mu'tazz, ˇabaqàt al-shu'arà", ed. 'Abd al-Sattàr A˙mad Farràj, †ab'a 2 (Cairo: Dàr al-Ma'àrif, 1968), 24, cf. Abù l-Faraj al-Ißfahànì, Al-Aghànì, 24 vols. (Cairo: Dàr al-Kutub and al-Hay"a l-Mißriyya l-'Àmma, 1927–74), III, 136. 16 Aghànì, III, 351. 17 Aghànì, VI, 10. 18 Aghànì, XX, 1. 19 Al-Marzubànì, Mu'jam, 134. 20 Aghànì, IV, 1, 3, 4–5. 21 Aghànì, XXIII, 217. 22 Aghànì, XIV, 17 (the text has, wrongly, ßulbiyyatan). 23 Aghànì, XIV, 263. Judging by his poetry and his career (he became governor of part of Khuràsàn) he was an Arab. 24 Aghànì, XV, 323. 25 Aghànì, XVIII, 149, 150. On the basis of his Bedouin-sounding name (Musàwir occurs eight times in Caskel, ]amharat an-nasab) and that of his father (Sawwàr, six times in Caskel, ]amharat an-nasab) one may conclude that, mawlà or no mawlà, he 15

MAWÀLÌ

and arabic poetry

353

b. al-Walìd is generally considered a mawlà of the Anßàr,26 but Ibn Qutayba says he was min abnà" al-Anßàr.27 Especially interesting is the case of Ziyàd al-A'jam, who died at an advanced age around 100/718. Sezgin says that he was Persian, but things are not that unequivocal. He certainly spoke like a Persian, hence his nickname al-A'jam, and he lived in Khuràsàn for a long time. His father’s name is given as Sulaymàn28 or Sulaym29 or Salmà,30 or Jàbir b. 'Amr b. 'Àmir b. al-Óàrith,31 and that last series sounds very Arabic and Arab. He dressed like a Persian, in a qabà" dìbàj, for which he was punished and derided by al-Muhallab, who said, “Why do you make yourself resemble the unbelievers and polytheists?”;32 which he would have said to an Arab but not to a Persian. This seems to be confirmed by those sources that make him a member of the tribe of 'Abd al-Qays.33 On the other hand, he is often called a mawlà of that tribe;34 he seems to have taken part in the conquest of I߆akhr in 23/643 and to have settled there before he moved to Khuràsàn.35 Someone said that aßlahu wa-mawlidahu wa-mansha"ahu bi-Ißbahàn, his origin, place of birth and where he grew up was Ißfahàn.36 He is called 'ilj, a term usually referring to non-Arabs, in verses by an Arab opponent, alMughìra b. Óabnà",37 who also says to him wa-mà laka aßlun yà Ziyàdu ta'udduhù/wa-mà laka fì l-ar∂i l-'arì∂ati wàlidù.38 He and his daughters are said to be aqlaf, uncircumcised, in lines by Ka'b al-Ashqarì39 and

was probably an Arab, although one occasionally finds non-Arabs with “Bedouin” names. 26 Aghànì, XIX, 31. 27 Ibn Qutayba, Al-Shi'r wa-l-shu'arà", ed. A˙mad Mu˙ammad Shàkir, 2 vols. (Cairo: Dàr al-Ma'àrif, 1966–67), 832. 28 E.g., Aghànì, XV, 380. 29 Ibn Sallàm, ˇabaqàt fu˙ùl al-shu'arà", 551. 30 Ibn Qutayba, Shi'r, 430. 31 Aghànì, XV, 380, cf. Ibn Qutayba, Shi'r, 430. 32 Aghànì, XV, 384. 33 E.g., Ibn Qutayba, Shi'r, 430; cf. Ibn Sallàm, ˇabaqàt, 551. 34 E.g., Aghànì, XV, 380; Abù 'Ubayd al-Bakrì, Dhayl Sim† al-la"àlì, ed. 'Abd al'Azìz al-Maymanì, repr. (Beirut: Dàr al-Kutub al-'Ilmiyya, n.d.), 8. 35 Yàqùt, Mu'jam al-udabà" (Cairo: Dàr al-Ma"mùn, 1936–38), XI, 168; Ibn Shàkir, Fawàt al-Wafayàt, ed. I˙sàn 'Abbàs, 5 vols. (Beirut: Dàr Íàdir, 1973–74), II, 29; alÍafadì, Al-Wàfì bi-l-wafayàt, 25 vols. (Wiesbaden: Franz Steiner, 1962–), XIV, 244. 36 Aghànì, XV, 380. 37 Aghànì, XIII, 92, 96. 38 Aghànì, XIII, 95. 39 Aghànì, XV, 393.

354

geert jan van gelder

al-Mughìra.40 Invective verse is not to be trusted, of course. The editor of his collected verse, Yùsuf Óusayn Bakkàr, collects all the evidence, and although he finds it very difficult to reach a decision41 he thinks it more likely on the whole that Ziyàd was ethnically an Arab.42 There is one thing which, I believe, strongly argues against this, or at least seems to indicate that Ziyàd’s original language was not Arabic. I could imagine that someone who lives for a long time among foreigners forgets some of his mother tongue—and Ziyàd, an esteemed poet, did obviously not forget his Arabic. I cannot imagine, however, that one would lose the ability to pronounce the sounds of one’s original language. And this is precisely what we must believe if Ziyàd was a true Arab, for there are several anecdotes illustrating his lukna or 'ujma: he said da"awtuka instead of da'awtuka, tasna" instead of taßna', shultàn instead of sul†àn, kult and kaws instead of qult and qaws.43 This would only happen to someone whose mother tongue was not pure Arabic. It is not terribly important whether or not Ziyàd was technically an Arab or a Persian. He may have had an Arab father but been brought up by a Persian mother in Persian-language surroundings. For the history of Arabic poetry what matters is precisely this background: Ziyàd is a very early example of a famous poet whose mother tongue was not Arabic, if my theory is correct. His poetry is mostly praised as sound and eloquent ( jazl ) and of pure diction ( faßì˙ alalfàΩ) in spite of his faulty pronunciation,44 even though a syntactical oddity (laysa ghàdin wa-là rà"i˙ù, instead of ghàdiyan and rà"i˙à) is also noted as being the result of fasàd lisànihi bi-Fàris, the corruption of his language in Persia.45 Ziyàd al-A'jam may have been the first Arabic poet whose mother tongue was not Arabic. Another poet who spoke with a strong foreign accent, Abù 'A†à" al-Sindì, lived and died considerably later. He was a mawlà whose father came from Sind; he grew up in Kùfa

40

Aghànì, XIII, 94, 95. Ziyàd al-A'jam, Shi'r, ed. Yùsuf Óusayn Bakkàr (Beirut: Dàr al-Masìra, 1983), 15. 42 Ziyàd al-A'jam, Shi'r, 22. 43 Aghànì, XV, 380; al-Jà˙iΩ, Al-Bayàn wa-l-tabyìn, ed. 'Abd al-Salàm Mu˙ammad Hàrùn, 4 vols. (Cairo: Maktabat al-Khànjì, 1968), I, 71; al-Taw˙ìdì, Al-Baßà"ir wal-dhakhà"ir, ed. Wadàd al-Qà∂ì, 9 vols. (Beirut: Dàr Íàdir, 1988), VI, 188. 44 Aghànì, XV, 380. 45 Ibn Qutayba, Shi'r, 432. 41

MAWÀLÌ

and arabic poetry

355

and died before 158/775 according to some, or after 180/796.46 His father was, according to Abù l-Faraj47 “sindiyyan a'jamiyyan là yufßi˙,” but the son’s Arabic pronunciation was little better, it appears: kànat fìhi 'ujma,48 kàna fì lisàn Abì 'A†à" lukna shadìda wa-luthgha fa-kàna là yufßi˙, even though he was jayyid al-shi'r.49 Again, we are given some examples of his mispronunciations: marhaban hayyàkumu llàh for mar˙aban ˙ayyàkumu llàh, hàja for ˙àja, zuzz for zujj, zaràda for jaràda, azunnu for aΩunnu, say†àn for shay†àn, madahtuhu for mada˙tuhu, hazawtuhu for hajawtuhu, fadalat for fa∂alat.50 In fact, some could hardly understand him at all: kàna là yakàd yufham kalàmuhu,51 and he confessed this in a poem in which he asks for a ràwì to recite his poetry: A'wazatnì l-ruwàtu yà bna Sulaymin wa-abà an yuqìma shi'rì lisànì Wa-ghalà bi-lladhì ujamjimu ßadrì wa-jafànì bi-'ujmatì sul†ànì (. . .) Wa-tamannaytu annanì kuntu bi-l-shi'ri faßì˙an wa-bàna ba'du bayànì52 (. . .) Fa-kfinì mà ya∂ìqu 'anhu ruwàtì bi-faßì˙in min ßàli˙i l-ghilmànì Yufhimu l-nàsa mà aqùlu mina l-shi'ri fa-inna l-bayàna qad a'yànì.53 I need reciters, Ibn Sulaym! My tongue refuses to straighten my verse. My breast boils over with my gibberish and my master treats me harshly for my foreign accent (. . .)54 I wish I were eloquent in my poetry and that my clear speech55 would somehow be evident (. . .) So help me with what my reciters56 are unable to do 46

GAS, II, 471–72. Aghànì, XVII, 327. 48 Ibn Qutayba, Shi'r, 766. 49 Ibn Qutayba, Shi'r, 766; Aghànì, XVII, 327. 50 Ibn Qutayba, Shi'r, 767–68; Aghànì, XVII, 330–32, 338. 51 Aghànì, XVII, 328. 52 Both Aghànì and Ibn Qutayba have banànì, which does not make sense to me. 53 Aghànì, XVII, 328, 337. 54 Or read with Aghànì, XVII, 337: wa-shakànì 'ujmatì shay†ànì, “My demon complains to me about my foreign accent,” a reading that is probably more original and certainly wittier; see Goldziher, Abhandlungen zur arabischen Philologie, 2 vols. (Leiden: Brill, 1896), I, 13–14; and Fück, 'Arabìya: recherches sur l’histoire de la langue et du style arabe, trans. Claude Denizeau (Paris: Marcel Didier, 1955), 30–31. 55 Reading, with Fück ('Arabìya, 31 note 133) bayànì for banànì. 56 One would have expected “my recitation”; perhaps an irregular maßdar of the verb rawà is intended. 47

356

geert jan van gelder by giving me an eloquent and decent boy, Who will make people understand what I say, for eloquence is beyond my powers!

Abù 'A†à" may have grown up in Kùfa,57 but just as in the case of Ziyàd al-A'jam I suspect that his first tongue was Indian, or rather Sindian. In spite of his background and his accent, al-Aßma'ì pronounced him to be faßì˙.58 It is well-known that in Umayyad times many Arabs were strongly prejudiced against the mawàlì, and that this extended to the field of poetry. To be a good poet one had to be an Arab and free, as expressed in the line by al-Farazdaq, referring to his rival Nußayb b. Rabà˙: Wa-khayru l-shi'ri akramuhù rijàlan wa-sharru l-shi'ri mà qàla l-'abìdù.59

There is a somewhat scabrous anecdote in which an unknown Bedouin says to an equally anonymous Persian poet (shà'ir min abnà" al-Furs): “Poetry belongs to the Arabs; everyone of you who makes poetry only does so because one of us has mounted his mother!,” whereupon the Persian answers, “Likewise, whenever one of you does not make poetry, that is because one of us has mounted his mother.”60 Jokes may not be reliable as reports, but the sentiments expressed in them are often real. Ibn Abì Du"àd (d. 240/854) said, “There is no Arab who is unable to make poetry, which is a natural trait of them, whether they make little or much of it.”61 Al-Óajjàj, mawàlìhater, “wished to remove them from the place of eloquence and good breeding (maw∂i' al-faßà˙a wa-l-àdàb) and to mix them with the villagers and Nabataeans, and said, ‘The mawàlì are louts ('ulùj ); they have been brought from the villages, so they’d better go to their villages’.”62 A Bedouin once saw Bashshàr b. Burd, and when he was 57

Aghànì, XVII, 328. Abù Óàtim al-Sijistànì/al-Aßma'ì, Fu˙ùlat al-shu'arà", ed. Mu˙ammad 'Abd alQàdir A˙mad (Cairo: al-Nah∂a l-Mißriyya, 1991), 123. 59 Ibn Qutayba, Shi'r, 411; al-Sharìf al-Murta∂à, Al-Amàlì (Ghurar al-fawà"id), ed. Mu˙ammad Abù l-Fa∂l Ibràhìm, 2 vols. (Cairo: Dàr I˙yà" al-Kutub al-'Arabiyya, 1954), I, 62. 60 Al-Óußrì, Zahr al-àdàb, ed. Zakì Mubàrak and Mu˙ammad Mu˙yì l-Dìn 'Abd al-Óamìd, 4 vols. repr. (Beirut: Dàr al-Jìl, 1972), 682. 61 Aghànì, XXIII, 56. 62 Al-Mubarrad, Al-Kàmil, ed. W. Wright, 2 vols. (Leipzig: F.A. Brockhaus, 1874–92), 286; cf. Ibn 'Abd Rabbih, 'Iqd, III, 416, quoting from al-Jà˙iΩ’s Kitàb almawàlì wa-l-'arab. 58

MAWÀLÌ

and arabic poetry

357

told that the latter was a poet, he asked, “Is he a mawlà or an Arab?”—“A mawlà,” they answered.—“What have mawàlì to do with poetry! (wa-mà li-l-mawàlì wa-l-shi'r),” exclaimed the Bedouin, whereupon Bashshàr, having asked his patron’s permission, responded with a suitably anti-Bedouin poem.63 Even Shu'ùbìs conceded that poetry was the Arabs’ contribution to civilization: The Arabs never produced any craft and left no trace of philosophy, except the poetry from the past (illà mà kàna min al-shi'r), but the nonArabs (al-'Ajam) have now come to share this with them. For the Byzantines have wonderful poems, with correct metre and prosody.64

Note that the only 'ajam mentioned here are Byzantines, who made Greek poetry, not the Persians and other mawàlì who made Arabic poetry. In any case, the opinion that Arabic poetry belonged to the Arabs alone was soon to disappear. When al-Jà˙iΩ contrasts the Arabs, who used poetry for recording their glorious deeds, with the 'ajam, who did this with architecture, he is talking about the preIslamic Arabs (Kànat al-'Arabu fì jàhiliyyatihà . . .).65 In his own days, poetry is a matter of language, not of ethnicity: “The excellence of poetry is restricted to the Arabs and those who speak Arabic ( fa∂ìlat al-shi'r maqßùra 'alà l-'Arab wa-'alà man takallama bi-lisàn al-'Arab).”66 With his al-Bayàn wa-l-tabyìn al-Jà˙iΩ attempts to prove the superiority of the eloquence and poetry of the Arabs in general; but here too he is speaking of the past, and he mentions non-Arabs such as Bashshàr, Abù l-'Atàhiya and Abàn al-Là˙iqì as gifted “natural” poets (al-ma†bù'ùn 'alà l-shi'r min al-muwalladìn).67 When he says that “badì' (novel or original style) is restricted to the Arabs” (al-badì' maqßùr 'alà l-'Arab), he seems to be referring to speakers of Arabic, not to ethnic Arabs, for he goes on to say that “al-Rà'ì has much badì' in his poetry, and Bashshàr has beautiful badì'.”68 Ibn Qutayba, 63 Aghànì, III, 166–67; cf. Bashshàr, Dìwàn, ed. Mu˙ammad al-ˇàhir b. 'Àshùr, 4 vols. Tunis/Algiers: al-Sharika l-Tùnisiyya and al-Sharika l-Wa†aniyya, 1976), III, 207–11. 64 Ibn 'Abd Rabbih, 'Iqd, III, 405; cf. also the ten categories of adab listed by the Persian al-Óasan b. Sahl, of which poetry, together with genealogy and ayyàm al-nàs, is said to be Arabic (al-Óußrì, Zahr, 196). 65 Al-Jà˙iΩ, Óayawàn, I, 72. 66 Al-Jà˙iΩ, Óayawàn, I, 74–75. 67 Al-Jà˙iΩ, Bayàn, I, 50–51. 68 Al-Jà˙iΩ, Bayàn, I, 55–56; see also I, 49.

358

geert jan van gelder

like al-Jà˙iΩ an opponent of the Shu'ùbìs, said that “God has not restricted knowledge, poetry or eloquence to one particular time, nor given it to one specific people (qawm); rather, He has distributed it so that it is shared by His servants in all ages.”69 Ibn Rashìq, in his encyclopedia of poetry and poetics, devoted a couple of sections to genealogy,70 in which he provides explanations of tribal terms and names, almost all of them Arabs. At the end, a brief note states that there are three kinds of mawàlì: mawlà l-yamìn al-mu˙àlif, mawlà l-dàr wa-l-mujàwir, and mawlà l-nasab ibn al-'amm wal-qaràba. However, nothing is said about the role of Arabs or mawàlì in the history of poetry. The sections are intended only as useful information for poets and critics. That no monographs on mawàlì poets were written, or that books on poets never segregate between mawàlì and Arabs, is not merely because the status of some individuals was uncertain, of course. One could imagine that a more important reason was the sensitive nature of the issue, since by writing such a work one could not help being drawn into the shu'ùbiyya controversy. Any such work could not have failed to show clearly the great contribution of mawàlì to the history of Arabic poetry. Yet some authors did not flinch from venting their anti-Arab partisanship in other fields. Thus the lack of interest in the matter remains to some extent curious and unexplained. The main reason why the mawàlì poets are not dealt with as a group is precisely, I believe, that they did not form a group, being an extremely heterogeneous collection of Persians, “Nabataeans” and others, many of mixed origins, with various linguistic backgrounds, without any kind of group solidarity. I know of only one case where one may perhaps detect such a feeling of solidarity: Nußayb b. Rabà˙, a mawlà of mixed Arab and African descent, and who preferred to remain a mawlà when asked to be incorporated into the clan of his patrons,71 and Ismà'ìl b. Yasàr, a mawlà of Persian origin (and a brother of Mùsà Shahawàt), professed admiration for each other’s verse;72 but no reason is given explicitly. Normally, if mawàlì had any loyalty, it 69

Ibn Qutayba, Shi'r, 63. Ibn Rashìq, Al-'Umda, ed. Mu˙ammad Mu˙yì l-Dìn 'Abd al-Óamìd, 2 vols. repr. (Beirut: Dàr al-Jìl, 1972), II, 190–98. 71 Aghànì, I, 336. This was perhaps for economic rather than ideological reasons: “qad 'alimtu annakum turìdùna bi-dhàlika màlì.” See also Goldziher, Muhammedanische Studien, I, 136. 72 Ibn Sallàm, ˇabaqàt, 348. 70

MAWÀLÌ

and arabic poetry

359

was to the Arab tribe or clan whose clients they were, or else to their own ethnic background, as in the case of the poets with shu'ùbì leanings such as the same Ismà'ìl b. Yasàr and Bashshàr b. Burd. In a patriarchal society, one’s ethnic identity was normally determined by that of one’s father, even though the mother’s background often played a secondary role in determining one’s status.73 The poet Ibn Mayyàda (d. c. 136/754) belonged to Dhubyàn; exceptionally, he took as much pride in his Persian ancestry through his mother (after whom he is called) as in his Arabness.74 The topic “mawàlì and Arabic poetry” gives rise to various questions. Firstly, one may ask what the mawàlì poets have in common, or how they differ from Arab poets. From the above it should follow that there are many different kinds of mawàlì poets, with various linguistic, ethnic and geographical backgrounds, and that there is not much point in the question if it is put in such general terms. Another question would be to ask what the contribution has been of the mawàlì to Arabic poetry. Gregor Schoeler published an essay (originally an inaugural lecture) in 1984 entitled “Ein Wendepunkt in der Geschichte der arabischen Literatur.”75 This “turning point” concerns the rise of mu˙dath verse with Bashshàr and others. Schoeler gives a survey of traditional explanations: either “historical,” by positing a survival of the literary traditions of late antiquity, or “purely literary (innerliterarisch),” as a reaction to the thematic conservatism of Arabic poetry. Finding the former explanation improbable and the latter insufficient, Schoeler’s answer is that the “modern” style came to prominence precisely when the rising class of mawàlì penetrates into Arabic literature and takes over the lead (“Der ‘neue’ Stil tritt genau dann in Erscheinung, als die aufsteigende Schicht der Klienten in die arabische Literatur hineindrängt und in ihr die Führung übernimmt”).76 I should like to discuss Schoeler’s thesis and question it to some extent, even though in my experience this is a hazardous venture 73 See e.g., Goldziher, Muhammedanische Studien, I, 121–27; Juda, Die sozialen und wirtschaftlichen Aspekte der Mawàlì in frühislamischer Zeit (Ph.D. diss., Tübingen, 1983), viii–ix, 27. 74 Aghànì, II, 261, 266–67. 75 Schoeler, “Ein Wendepunkt in der Geschichte der arabischen Literatur,” Saeculum 35 (1984), 293–305. 76 Schoeler, “Ein Wendepunkt,” 304.

360

geert jan van gelder

because he is usually right. In general, I find his formulation too general, which is perhaps not unrelated to the fact that his article was an inaugural lecture, in which one expects broad outlines for a broader public, rather than subtly qualified hypotheses. It is obvious, on the one hand, that there is a link between the rise of the so-called modern (mu˙dath) poetry and the presence and growing influence of the mawàlì on politics, society and culture. On the other hand, not all mu˙dath poets were mawàlì, nor were all mawàlì poets modern. Let us mention some names. By general consensus, the Persian mawlà Bashshàr (d. 167/783) was the first great “modern” poet, “father of the mu˙dathùn”; interestingly, al-Aßma'ì calls him “the seal of poets,” as if he were the last great early poet.77 The greatest “modern” poet, Abù Nuwàs, was a mawlà of Persian descent, it seems. Muslim b. al-Walìd, who is credited with the conscious introduction of the badì' style,78 was probably a mawlà from Kùfa, whose ethnic background is unknown. He died in 208/823, forty years after Bashshàr. Abù l-'Atàhiya likewise was a mawlà who grew up in Kùfa and died two or three years after Muslim b. al-Walìd. These four poets were all of them important mu˙dathùn, arguably the leading poets among the early moderns, and all of them mawàlì. But the important stylistic and thematical innovations in mu˙dath poetry did not begin with these and other mawàlì. A significant contribution was made by Arab contemporaries and even predecessors. Let us again mention some names, even though it may be slightly perverse, in a book devoted to the role of the mawàlì, to focus on Arabs and stress their importance, as if one were following in the footsteps of non-Arab but anti-Shu'ùbite writers such as al-Jà˙iΩ and Ibn Qutayba. Ibn Harma (d. 176/792), of Quraysh, is often called the last of the old poets (al-Aßma'ì: khutima l-shi'r bi-Ibn Harma,79 or khutima l-shu'arà" bi-Ibn Harma;80 he is in “the rearguard of the poets,” sàqat

77

Aghànì, III, 143, 148, 150. E.g., Ibn al-Mu'tazz, ˇabaqàt, 235; Aghànì, XIX, 31; cf. GAS, II, 528; art. “Muslim b. al-Walìd,” EI 2, VII; Heinrichs, “Muslim b. al-Walìd und badì',” in W. Heinrichs and G. Schoeler, eds., Festschrift Ewald Wagner zum 65. Geburtstag, II: Studien zur arabischen Dichtung (Beirut/Stuttgart: Franz Steiner, 1994), II, 211–45. 79 Ibn al-Mu'tazz, ˇabaqàt, 20. 80 Aghànì, IV, 373, 396. 78

MAWÀLÌ

and arabic poetry

361

al-shu'arà",81 and the last poet whose verse may be quoted as linguistic evidence for pure Arabic).82 He is also called a modern poet.83 His verse may not demonstrate much that could be called “modern,” but there is at least one startling novelty: he is said to have composed a poem of some forty lines using only unpunctuated letters, twelve lines of which are preserved.84 It begins in the old style, with a†làl and nasìb: A-rasmu Sawdata 85 ma˙lun dàrisu l-†alalì/mu'a††alun raddahù l-a˙wàlu ka-l-˙ulalì. One could imagine that the absence of dots is symbolical for the absent beloved, and that the word mu'a††al in line 1 is a further pun, although I doubt if the poet intended these things. In any case, this is a very early example of a visual lipogram, a typical mu˙dath technique that became popular much later (see e.g., al-Óarìrì’s maqàmas al-Samarqandiyya, al-Maràghiyya and al-Óalabiyya). Ibn Dàwùd al-Ißbahànì (d. 297/909) quotes a few anonymous examples in his anthology al-Zahra;86 it is regularly included in lists of badì' only centuries later, with al-Zanjànì (d. 660/1262) and others.87 Abù l-Faraj al-Ißfahànì can hardly believe that Ibn Harma’s poem is authentic: “I have not found it in Ibn Harma’s [collected] verse, nor did I think that anyone preceded Ruzayn al-'Arù∂ì in this.”88 81

Ibn Qutayba, Shi'r, 753. 'Abd al-Qàdir al-Baghdàdì, Khizànat al-adab, ed. 'Abd al-Salàm Mu˙ammad Hàrùn, 13 vols. (Cairo: Dàr al-Kàtib al-'Arabì and Maktabat al-Khànjì, 1967–86), I, 425, etc. 83 Mu˙dath (Ibn al-Nadìm, Al-Fihrist, ed. Gustav Flügel et al., 2 vols. (Leipzig, 1871–72), 159), muwallad (al-Jà˙iΩ, Bayàn, I, 51), in both cases together with Bashshàr. Another, earlier “borderline” poet is Ibn Mayyàda (d. c. 136/754), who counts as one of the last “classical” poets used for linguistic evidence by the philologists (GAS, II, 442; Aghànì, II, 269) and who also appears in Ibn al-Mu'tazz’s work on mu˙dath poets (Ibn al-Mu'tazz, ˇabaqàt, 105–109). 84 Aghànì, IV, 378–79. 85 It will be noted that the tà" marbù†a counts as an undotted hà". Similarly, the final yà" that serves as alif maqßùra, as in ra"à and a'là (lines 2 and 11), is also considered undotted. 86 Ibn Dàwùd, Al-Zahra, ed. Ibràhìm al-Sàmarrà"ì, 2 vols. (al-Zarqà" ( Jordan): Maktabat al-Manàr, 1985), 782–83. 87 Al-Zanjànì, Mi'yàr al-nuΩΩàr fì 'ulùm al-ash'àr, ed. Mu˙ammad 'Alì Rizq alKhafàjì, 2 vols. (Cairo: Dàr al-Ma'àrif, 1991), II, 130–31; Íafì al-Dìn al-Óillì, Shar˙ al-Kàfiya al-badì'iyya, ed. Nasìb Nashàwì (Damascus: Majma' al-Lugha l-'Arabiyya, 1982), 276–77; Ibn Óijja, Khizànat al-adab (Cairo, 1291 AH), 536–38; cf. 'Abd alGhanì al-Nàbulusì, Nafa˙àt al-azhàr (Bùlàq, 1299 AH), 254–57; Cachia, The Arch Rhetorician, or The Schemer’s Skimmer: A Handbook of Late Arabic badi' (Wiesbaden: Harrassowitz, 1998), 39–40. In Persian it appears at an earlier stage, see Rashìd al-Dìn Wa†wà†, Óadàyiq al-si˙r fì daqàyiq al-shi'r, ed. 'Abbàs Iqbàl (Tehran: Kitàbkhànayi Kàwa, 1308 AH Sh.), 64–66. 88 Aghànì, IV, 378. On this Ruzayn, or rather Razìn, see al-Jà˙iΩ, Óayawàn, VII, 82

362

geert jan van gelder

Ibn Harma’s contemporary al-Khalìl Ibn A˙mad, of al-Azd (d. 175/ 791), great scholar and minor poet,89 experimented with metres in true mu˙dath fashion: lines that only employ long syllables (Laysa l-mar"u l-˙àmì anfan/mithla l-mar"i l-ßatmi l-rà∂ì ), or the foot fa'ilun (Su"ilù fa-abaw fa-laqad bakhilù/wa-la-bi"sa la-'amruka mà fa'alù).90 He made punning rhymes (three lines rhyming in al-ghurùb in three different senses).91 He was the first to produce an Arabic pangram, a line containing all the letters of the alphabet that makes rather more sense than the English quick brown fox that jumps over the lazy dog (ßif khalqa khawdin ka-mithli l-shamsi idh bazaghat/ya˙Ωà l-∂ajì'u bihà najlà"a mi'†àrì ).92 These experiments by Ibn Harma and al-Khalìl are trivial, perhaps. More important is the innovative character of the poetry of the bad caliph and good poet al-Walìd b. Yazìd, Arab from father’s and mother’s side, whose violent early death in 744/127 prevented him from reaching the Abbasid period and, therefore perhaps, from being ranked among the true mu˙dathùn.93 That he sounds “modern” is a matter of themes (wine, mujùn) as much as form (short metres) and diction (no Bedouin gharìb). The greatest mu˙dath poet, the mawlà Abù Nuwàs, most famous on account of his bacchic poetry, had several important precursors in this genre who were Arabs, among them Abù l-Hindì, who died after 132/750, Wàliba b. al-Óubàb (d. c. 170/786), al-'Ujayr alSalùlì (contemp. with reign of Hishàm, 105–125/724–43), and even earlier poets such as Abù Jilda al-Yashkurì (d. c. 85/704), al-Uqayshir (d. 80/699), al-Óàritha b. Badr (d. c. 64/684) and, earliest of all,

217; Ibn al-Mu'tazz, ˇabaqàt, 295; Aghànì, VI, 160; al-Tha'àlibì, Thimàr al-qulùb fì l-mu∂àf wa-l-mansùb, ed. Mu˙ammad Abù l-Fa∂l Ibràhìm (Cairo: Dàr al-Ma'àrif, 1985), 387. 89 See Weipert, “Al-flalìl Ibn A˙mad—A Poet?” Oriens 35 (1996), 65–104. 90 Al-Marzubànì, Nùr al-qabas al-mukhtaßar min al-Muqtabas fì akhbàr al-nu˙àt wa-ludabà" wa-l-shu'arà" wa-l-'ulamà", ed. Rudolf Sellheim (Wiesbaden: Franz Steiner, 1964), 60. 91 Al-Marzubànì, Nùr al-qabas, 59. 92 Al-Marzubànì, Nùr al-qabas, 59; more sources in Weipert, “Al-flalìl,” 73 (note 21). 93 For the appreciation of his poetry, see especially Renate Jacobi, “Zur ˝azalpoesie des Walìd ibn Yazìd,” in Wolfhart Heinrichs and Gregor Schoeler, eds., Festschrift Ewald Wagner zum 65. Geburtstag, II: Studien zur arabischen Dichtung (Beirut/Stuttgart: Franz Steiner, 1994), II, 145–61; eadem, “Theme and variation in Umayyad ghazal poetry,” Journal of Arabic Literature 32 (1992), 109–19; eadem, art. “al-Walìd Ibn Yazìd,” in J.S. Meisami and P. Starkey, eds., Encyclopedia of Arabic Literature (London, 1998), 803.

MAWÀLÌ

and arabic poetry

363

Abù Mi˙jan al-Thaqafì (d. after 16/637). The early history of wine poetry has been studied in some detail recently by scholars such as Bencheikh and Kennedy.94 Therefore I should like to focus briefly on a few others. 'Ammàr b. 'Amr b. 'Abd al-Akbar did not only make some khamriyyàt 95 but was an early poet of nonsense poetry and light verse. He was an Arab of the tribe of Hamdàn,96 known as 'Ammàr Dhù Kinàz,97 or Dhù Kubàr,98 who lived in Kùfa in the first quarter of the eighth century.99 His poetry was admired by alWalìd b. Yazìd; a book about him was composed by Mu˙ammad b. 'Abdallàh al-Óazanbal.100 Although he lived well before the period of the recognized “moderns,” his verse shows many modern characteristics. One poem, admired by al-Walìd b. Yazìd, of which twenty-four lines are preserved,101 is composed in a short, sixteensyllable khafìf metre. It has an unusual rhyme in -dhà and uses a number of Persian-derived words (mujanbadh: “vault-shaped,” from gunbad; jihbidh/jahbadh “clever,” from kihbud/kahbad, Pahlavi gàhbed “banker”; hirbidh “Magian priest,” from hirbad/hèrbed ). Moreover, it is a jesting and partly obscene poem, with an element of nonsense (Ashtahì minki minki minki makànan mujanbadhà . . . “I lust for that vaulted spot of yours of yours of yours”). In other poems he again uses short metres, as in his priapic rà"iyya.102 He parodies the a†làl theme in a qaßìda made for Khàlid al-Qasrì, by replacing the deserted abodes with his ragged clothes and empty house: Akhlaqat ray†atì wa-awdà l-qamìßù/wa-izàrì wa-l-ba†nu †àwin khamìßù Wa-khalà manzilì fa-là shay"a fìhì . . .

94

Kennedy, The Wine Song in Classical Arabic Poetry: Abù Nuwàs and the Literary Tradition (Oxford: Oxford University Press, 1997); Bencheikh, art. “‡amriyya,” EI 2. 95 Schoeler, “'Abbàsid Belles-Lettres,” in J. Ashtiany et al., eds. (Cambridge: Cambridge University Press, 1990), 292; and Kennedy, Wine Song, 81. 96 Aghànì, XXIV, 220: Hamdàniyyun ßalìbatan. 97 Thus GAS, II, 341. 98 Thus in Aghànì, XXIV, 220ff. The name is a bit of a puzzle; Ibn al-Kalbì/ Caskel, II, 167, mentions a Hamdànì called 'Ammàr b. 'Ubayd b. Zayd b. 'Umayr b. Dhì Kibàr, or 'Ammàr Dhù Kubàr, or Dhù Kubàr b. Sayf b. 'Amr b. Sab' b. al-Sabì'. 99 Aghànì, XXIV, 219–35; VII, 56–57, 67; 'Alì b. Abì l-Faraj al-Baßrì, Al-Óamàsa al-Baßriyya, ed. Mukhtàr al-Dìn A˙mad, 2 vols. (Hyderabad: Dà"irat al-Ma'àrif al'Uthmàniyya, 1964), II, 313; GAS, II, 341. 100 Aghànì, XXIV, 220, 223, 231; on the author cf. Ibn al-Nadìm, Fihrist, 68, 73. 101 In fragments, Aghànì, XXIV, 219, 221–22, 232, 235. 102 20 lines, Aghànì, XXIV, 227–28.

364

geert jan van gelder My thin clothes are threadbare, my shirt has gone to rags, like my loincloth; my belly is starved and hollow. My house is empty, nothing left in it . . .103

Another early poet of light verse is al-Óakam b. 'Abdal, of Asad,104 a much-feared satirist who grew up and lived in Kùfa and died c. 100/718.105 His poems on the cats and mice (both kinds represented as speaking) and other creatures in his house,106 or on himself (lame) and Abù 'Ulayya (blind)107 have a remarkably “modern” ring; and he, too, did not shun jesting and obscene verse.108 Musàwir al-Warràq, a poet and mu˙addith who lived in the middle of the second/eighth century and who was very probably an Arab,109 was a pioneer with his long banquet poem quoted by Ibn 'Abd Rabbih.110 Ibn Abì l-Zawà"id Sulaymàn b. Ya˙yà, of the tribe of Hawàzin, imàm in Medina, straddled the “Two Dynasties” (mukha∂ram aldawlatayn) and made ghazal, obscene verse and wine poetry. He employed rare rhymes (-ìthù, -àdhà) and boasted about it: “Now listen to this dhàl! Bring me a poet who made a rhyme like that!” (Hàdhihi l-dhàlu fa-sma'ùhà wa-hàtù/shà'iran qàla fì l-rawiyyi 'alà dhà! ).111 Many poems by 'Ammàr, al-Óakam b. 'Abdal, and the other Umayyad poets I have mentioned would be difficult to distinguish from “true” mu˙dath poems, and one must conclude that the transition from “early” to “modern” was neither so clear-cut nor so sudden as is suggested, for instance, by Schoeler’s use of the word “precisely” ( genau), when he speaks of the coinciding rise of mawàlì influence and of mu˙dath poetry. One suspects that the usual equation in traditional literary criticism of “old” with pre-Islamic, early Islamic and Umayyad, and of “modern” with 'Abbàsid, has something to do with ideology and politics.

103

Aghànì, XXIV, 229. Caskel, ]amharat an-nasab, Tab. 55. 105 Aghànì, II, 402–26; al-Jà˙iΩ, Óayawàn, I, 249–53; V, 297–300; more references in GAS, II, 331 and Pellat, EI 2, s.v. 106 Al-Jà˙iΩ, Óayawàn, V, 297–99, and see 300. 107 Aghànì, II, 405–406. 108 Aghànì, II, 409–10. 109 See above, note 25, and G.J. van Gelder, “Musàwir al-Warràq and the beginnings of Arabic gastronomic poetry,” Journal of Semitic Studies 36 (1991), 309–27. 110 Ibn 'Abd Rabbih, 'Iqd, VI, 295–97. 111 Aghànì, XIV, 126. On the poet, see Aghànì, XIV, 120–30; GAS, II, 449. 104

MAWÀLÌ

and arabic poetry

365

Possibly the clearest influence of Persian mawàlì on poetry is the occurrence of Persian words. Abù Nuwàs used many of them.112 Far from being the prerogative of the mawàlì, however, Persian words are already found in al-A'shà’s verse in pre-Islamic times; as shown above, 'Ammàr Dhù Kinàz/Kubàr also liked to use them. As alJà˙iΩ says, “It happens that a Bedouin (a'ràbì ) inserts some Persian words into his verse, for the fun of it.”113 The earliest, or almost earliest, (New) Persian poetry on record is a doggerel by the early Umayyad Arab poet Yazìd b. Mufarrigh al-Óimyarì (d. 69/688) from Baßra: Àb-ast nabìdh-ast 'ußàrat-i zabìb-ast Sumayya rùspìd-ast, “It’s water, it’s date wine, it’s grape juice! It’s Sumayya, the whore!”114 In short, the rise of mu˙dath poetry was not due exclusively to the mawàlì, nor did it coincide more or less precisely with the dynastic change from Umayyad to 'Abbàsid times. “Modern” poetry is the natural result of poets living in the multicultural, multilingual urban centers, Kùfa and Baßra. Whether they were mawlà or Arab is relatively immaterial. In this article I have concentrated mostly on the notional status of being a mawlà or an Arab, and I have not spoken on the implications for the poetry made in the context of the patron-client relationship. Patronage in a broader sense (not to be confused with the institutional patronate) undoubtedly affects the production of poetry and its nature. It has little to do with the mawàlì in the sense of “non-Arab Muslims,” for both the client and the patron may be, and often were, Arabs. Or indeed non-Arabs: the Barmakids spring to mind, who as patrons of poets are said to have created a dìwàn al-shi'r for the distribution of rewards.115 Ya˙yà b. Khàlid al-Barmakì put Abàn al-Là˙iqì in charge; poets had to submit their poems to

112

Wagner, Abù Nuwàs, 190–95, 213–14. Al-Jà˙iΩ, Bayàn, I, 141, with some examples by al-'Umànì and others. 114 Al-Jà˙iΩ, Bayàn, I, 143; Aghànì, XVIII, 264; Ibn Qutayba, Shi'r, 361; alBaghdàdì, Khizàna, VI, 46; cf. Fück, 'Arabìya, 14; and especially Rempis, “Die altesten Dichtungen in Neupersisch,” Zeitschrift der Deutschen Morgenländischen Gesellschaft 101 (1951), 220–40 (see 222–23, with a longer version taken from a Persian source). Most seem to think that rwsbyd/rwsby/rwsfyd/rwsfy reflects New Persian rùspì, Pahlavi ròspìg, “whore, courtesan.” On Ibn Mufarrigh, see GAS, II, 324–26; Pellat, EI 2, s.v.; and Pellat, “Le poète Ibn Mufarri© et son œuvre,” in Mélanges Louis Massignon (Damascus: Institut Français de Damas, 1956), III, 195–232. 115 Sourdel, art. “Baràmika,” EI 2, 1035b. 113

366

geert jan van gelder

him first.116 Al-Jahshiyàrì tells that al-Rashìd told poets and orators to make panegyrics on al-Fa∂l b. Ya˙yà al-Barmakì; a rather minor poet, A˙mad b. Sayyàr al-Jurjànì, was ordered to sift the incoming poems and pay the rewards they deserved.117 The medieval critics did not often discuss the effect on poetry of the social and economic dependence of a poet upon his patron. In the innumerable anecdotes in which a poet is rewarded with large sums, this serves to emphasize his excellence and importance. Yet many critics speak of the mercenary sides of poetic practice in order to demonstrate the degeneration of poetry, or how it has become inferior to prose. Abù 'Amr b. al-'Alà" (d. 159/776) says that among the reasons that the status of poet is inferior to that of the orator (al-kha†ìb) is the fact that “they have made poetry a means of earning (maksaba), connecting themselves with the rabble, and attacking people’s reputations.”118 'Abd al-Karìm al-Nahshalì blames the modern poets (al-mu˙dathùn) for perverting the allegedly true function of poetry by turning it into a means of earning a living.119 His pupil Ibn Rashìq, however, in the chapter al-Takassub bi-l-shi'r wa-l-anafa minhu, “Earning by means of poetry, and disdaining this” in his al'Umda,120 does not connect this with either the rise of the mu˙dathùn or the mawàlì. The Arabs themselves were the first: although the Arabs used not to earn anything with their poems, payment for poetry began already in pre-Islamic times, with al-Nàbigha alDhubyànì, who became wealthy through his panegyrics for the 116 Al-Íùlì, Al-Awràq: Akhbàr al-shu'arà" al-mu˙dathìn, ed. J. Heyworth-Dunne (London/Cairo: Luzac and Ma†ba'at al-Íàwì, 1934), 33. 117 Al-Jahshiyàrì, Al-Wuzarà" wa-l-kuttàb, ed. Mu߆afà l-Saqqà et al. (Cairo: Mu߆afà l-Bàbì l-Óalabì, 1980), 191–92. A˙mad b. Sayyàr was sacked when it appeared he had excluded Abù Nuwàs at the instigation of his rivals, including Abàn al-Là˙iqì, Muslim b. al-Walìd and Ashja' al-Sulamì (ibid.). See also Wagner, Abù Nuwàs, 64 note 3 (where A˙mad b. Sayyàr’s nisba appears as al-Jarjarà"ì). 118 Al-Jà˙iΩ, Bayàn, I, 241, with some additions also in al-Marzùqì, Shar˙ Óamàsat Abì Tammàm, ed. A˙mad Amìn and 'Abd al-Salàm Hàrùn, 4 vols. repr. (Beirut: Dàr al-Jìl, 1991), 16–17 (he contrasts the poets not with orators but “the eloquent,” al-bulaghà", obviously including prose writers); and cf. Ibn Rashìq, Al-'Umda, I, 82–83. 119 'Abd al-Karìm al-Nahshalì, Al-Mumti' fì ßan'at al-shi'r, ed. 'Abbàs 'Abd al-Sàtir (Beirut: Dàr al-Kutub al-'Ilmiyya, 1983), 141 ( ja'alùhu muktasaban). 120 Ibn Rashìq, 'Umda, I, 80–86; on takassub, see e.g., al-Khayyà†, Al-Takassub bil-shi'r (Beirut: Dàr al-Àdàb, 1970); Mu߆afà Badr Zayd, Al-Takassub bi-l-shi'r (Cairo: al-Ma†ba'a l-Salafiyya, 1348 AH), (mentioned in A˙mad A˙mad Badawì, Usus alnaqd al-adabì 'inda l-'Arab (Cairo: Nah∂at Mißr, 1960), 682); and Darwìsh al-Jundì, ¸àhirat al-takassub wa-atharuhà fì l-shi'r al-'arabì wa-naqdihi (Cairo, 1970). The two lastmentioned titles were not accessible to me.

MAWÀLÌ

and arabic poetry

367

Lakhmid and Ghassànid kings, with Zuhayr, who received “a little ( yasìran)” from Harim b. Sinàn, and al-A'shà, who is said to have “made poetry a trade ( ja'ala l-shi'r matjaran),” traveled widely and was rewarded by the Persian emperor. It comes as no surprise that Ibn Rashìq condemns the phenomenon; he obviously mentions the preIslamic exceptions partly as an excuse for the widespread occurrence of literary patronage in Islamic times.121 The topic is worth more study; one of the points this article has intended to make, is that this literary patronage has, in principle, little to do with the patronage between Arabs and mawàlì.

Bibliography Primary Sources 'Abd al-Ghanì al-Nàbulusì. Nafa˙àt al-azhàr. Bùlàq, 1299 AH. See also Secondary Sources, s.v. Cachia. Abù Óàtim al-Sijistànì/al-Aßma'ì. Fu˙ùlat al-shu'arà", ed. Mu˙ammad 'Abd al-Qàdir A˙mad. Cairo: al-Nah∂a l-Mißriyya, 1991. Abù Nuwàs. Dìwàn, ed. Ewald Wagner, vol. 1. Cairo/Wiesbaden: Lajnat al-Ta"lìf and Franz Steiner, 1958. Abù 'Ubayd al-Bakrì. Dhayl Sim† al-la"àlì, ed. 'Abd al-'Azìz al-Maymanì. repr. Beirut: Dàr al-Kutub al-'Ilmiyya, n.d. al-Aßma'ì. Fu˙ùlat al-shu'arà". See Abù Óàtim al-Sijistànì al-Baghdàdì, 'Abd al-Qàdir. Khizànat al-adab, ed. 'Abd al-Salàm Mu˙ammad Hàrùn, 13 vols. Cairo: Dàr al-Kàtib al-'Arabì and Maktabat al-Khànjì, 1967–86. Bashshàr Ibn Burd. Dìwàn, ed. Mu˙ammad al-ˇàhir b. 'Àshùr, 4 vols. Tunis/Algiers: al-Sharika l-Tùnisiyya and al-Sharika l-Wa†aniyya, 1976. al-Baßrì, 'Alì b. Abì l-Faraj. Al-Óamàsa al-Baßriyya, ed. Mukhtàr al-Dìn A˙mad, 2 vols. Hyderabad: Dà"irat al-Ma'àrif al-'Uthmàniyya, 1964. al-Óillì, Íafì al-Dìn. Shar˙ al-Kàfiya al-badì'iyya, ed. Nasìb Nashàwì. Damascus: Majma' al-Lugha l-'Arabiyya, 1982. al-Óußrì, Ibràhìm b. 'Alì. Zahr al-àdàb, ed. Zakì Mubàrak and Mu˙ammad Mu˙yì l-Dìn 'Abd al-Óamìd, 4 vols. repr. Beirut: Dàr al-Jìl, 1972. Ibn 'Abd Rabbih. Al-'Iqd al-farìd, ed. A˙mad Amìn et al., 7 vols. Cairo: Dàr alKutub, 1948–53; repr. Beirut: Dàr al-Kitàb al-'Arabì, 1983. Ibn Dàwùd al-Ißbahànì. Al-Zahra, ed. Ibràhìm al-Sàmarrà"ì, 2 vols. al-Zarqà" ( Jordan): Maktabat al-Manàr, 1985. Ibn Óijja al-Óamawì. Khizànat al-adab. Cairo, 1291 AH. Ibn al-Jarrà˙, Mu˙ammad b. Dàwùd. Al-Waraqa, ed. 'Abd al-Wahhàb 'Azzàm and 'Abd al-Sattàr Farràj, †ab'a 2. Cairo: Dàr al-Ma'àrif, [1953] (date of preface).

121 On the early 'Abbàsid period, see e.g., Bencheikh, Poétique arabe: Essai sur les voies d’une création (Paris: Anthropos, 1975), 27–32 (“Evolution du mécénat et fonction de la poésie”) and his “Le cénacle poétique du calife al-Mutawakkil (m. 247): Contribution à l’analyse des instances de légitimation socio-littéraires,” Bulletin d’Etudes Orientales 29 (1977), 33–52.

368

geert jan van gelder

Ibn al-Kalbì, Hishàm. Jamharat al-nasab. See Secondary Sources, s.v. Caskel. Ibn al-Mùtazz. ˇabaqàt al-shu'arà", ed. 'Abd al-Sattàr A˙mad Farràj, †ab'a 2. Cairo: Dàr al-Ma'àrif, 1968. Ibn al-Nadìm. Al-Fihrist, ed. Gustav Flügel et al., 2 vols. Leipzig, 1871–72; repr. Beirut: Maktabat Khayyà†, n.d. Ibn Qutayba. Al-Shi'r wa-l-shu'arà", ed. A˙mad Mu˙ammad Shàkir, 2 vols. Cairo: Dàr al-Ma'àrif, 1966. Ibn Rashìq. Al-'Umda, ed. Mu˙ammad Mu˙yì l-Dìn 'Abd al-Óamìd, 2 vols. repr. Beirut: Dàr al-Jìl, 1972. Ibn Sallàm al-Juma˙ì. ˇabaqàt fu˙ùl al-shu'arà", ed. Ma˙mùd Mu˙ammad Shàkir. Cairo: Dàr al-Ma'àrif, 1952. Ibn Shàkir al-Kutubì. Fawàt al-Wafayàt, ed. I˙sàn 'Abbàs, 5 vols. Beirut: Dàr Íàdir, 1973–74. al-Ißfahànì, Abù l-Faraj. Al-Aghànì. 24 vols. Cairo: Dàr al-Kutub and al-Hay"a l-Mißriyya l-'Àmma, 1927–74. al-Jà˙iΩ. Al-Bayàn wa-l-tabyìn, ed. 'Abd al-Salàm Mu˙ammad Hàrùn, 4 vols. Cairo: Maktabat al-Khànjì, 1968. id., Al-Óayawàn, ed. 'Abd al-Salàm Mu˙ammad Hàrùn, 8 vols. Cairo: Mu߆afà l-Bàbì l-Óalabì, 1965–69. id., Rasà"il, ed. 'Abd al-Salàm Mu˙ammad Hàrùn, 4 vols. Cairo: Maktabat alKhànjì, 1964–79. al-Jahshiyàrì. Al-Wuzarà" wa-l-kuttàb, ed. Mu߆afà l-Saqqà et al., Cairo: Mu߆afà l-Bàbì l-Óalabì, 1980. al-Marzubànì. Mu'jam al-shu'arà", ed. 'Abd al-Sattàr Farràj, Cairo: 'Ìsà l-Bàbì l-Óalabì, 1960. id., Nùr al-qabas al-mukhtaßar min al-Muqtabas fì akhbàr al-nu˙àt wa-l-udabà" wa-l-shu'arà" wa-l-'ulamà", ed. Rudolf Sellheim, Wiesbaden: Franz Steiner, 1964 (BTS, 23a). al-Marzùqì. Shar˙ Óamàsat Abì Tammàm, ed. A˙mad Amìn and 'Abd al-Salàm Hàrùn, 4 vols. repr. Beirut: Dàr al-Jìl, 1991. al-Mubarrad. Al-Kàmil, ed. W. Wright, 2 vols. Leipzig: F.A. Brockhaus, 1874–92. al-Nahshalì, 'Abd al-Karìm. Al-Mumti' fì ßan'at al-shi'r, ed. 'Abbàs 'Abd al-Sàtir. Beirut: Dàr al-Kutub al-'Ilmiyya, 1983. Rashìd al-Dìn Wa†wà†. Óadàyiq al-si˙r fì daqàyiq al-shi'r, ed. 'Abbàs Iqbàl. Tehran: Kitàbkhàna-yi Kàwa, 1308 HSh. al-Íafadì. Al-Wàfì bi-l-wafayàt. 29 vols. to date, Wiesbaden: Franz Steiner, 1962–. al-Sharìf al-Murta∂à. Al-Amàlì (Ghurar al-fawà"id), ed. Mu˙ammad Abù l-Fa∂l Ibràhìm. 2 vols. Cairo: Dàr I˙yà" al-Kutub al-'Arabiyya, 1954. al-Íùlì. Al-Awràq: Akhbàr al-shu'arà" al-mu˙dathìn, ed. J. Heyworth-Dunne. London/Cairo: Luzac and Ma†ba'at al-Íàwì, 1934. al-Taw˙ìdì, Abù Óayyàn. Al-Baßà"ir wa-l-dhakhà"ir, ed. Wadàd al-Qà∂ì, 9 vols. Beirut: Dàr Íàdir, 1988. al-Tha'àlibì. Thimàr al-qulùb fì l-mu∂àf wa-l-mansùb, ed. Mu˙ammad Abù l-Fa∂l Ibràhìm. Cairo: Dàr al-Ma'àrif, 1985. Yàqùt. Mu'jam al-udabà". Cairo: Dàr al-Ma"mùn, 1936–38; repr. Beirut: Dàr alTuràth al-'Arabì, n.d. al-Zanjànì, 'Abd al-Wahhàb b. Ibràhìm. Mi'yàr al-nuΩΩàr fì 'ulùm al-ash'àr, ed. Mu˙ammad 'Alì Rizq al-Khafàjì, 2 vols. Cairo: Dàr al-Ma'àrif, 1991. Ziyàd al-A'jam. Shi'r, ed. Yùsuf Óusayn Bakkàr, Beirut: Dàr al-Masìra, 1983. Secondary Sources Arazi, Albert. “Abù Nuwàs fut-il “u'ùbite?” Arabica 26 (1979), 1–61. Ashtiany, Julia et al. eds. 'Abbàsid Belles-Lettres. Cambridge: Cambridge University Press, 1990.

MAWÀLÌ

and arabic poetry

369

Badawì, A˙mad A˙mad. Usus al-naqd al-adabì 'inda l-'arab. Cairo: Nah∂at Mißr, 1960. Bencheikh, Jamal Eddine. “Le cénacle poétique du calife al-Mutawakkil (m. 247): Contribution à l’analyse des instances de légitimation socio-littéraires,” Bulletin d’Etudes Orientales 29 (1977), 33–52. id., art. “‡amriyya,” EI 2, IV, 998–1009. id., Poétique arabe: Essai sur les voies d’une creation. Paris: Anthropos, 1975. Cachia, Pierre. The Arch Rhetorician, or The Schemer’s Skimmer: A handbook of late Arabic badì'. Wiesbaden: Harrassowitz, 1998. Caskel, Werner. ]amharat an-nasab: das genealogische Werk des Hi“àm ibn Mu˙ammad alKalbì. 2 vols. Leiden: Brill, 1966. The Encyclopaedia of Islam, New Edition, 12 vols. to date. Leiden: Brill, 1960–. Fück, Johann. 'Arabìya: recherches sur l’histoire de la langue et du style arabe, trans. Claude Denizeau. Paris: Marcel Didier, 1955. Goldziher, Ignaz. Abhandlungen zur arabischen Philologie, 2 Bde. Leiden: Brill, 1896. id., Muhammedanische Studien. 2 vols. Halle, 1889–90; repr. Hildesheim: Georg Olms, 1971. Heinrichs, Wolfhart. “Muslim b. al-Walìd und badì',” in Wolfhart Heinrichs and Gregor Schoeler, eds., Festschrift Ewald Wagner zum 65. Geburtstag, II: Studien zur arabischen Dichtung. Beirut/Stuttgart: Franz Steiner, 1994 (BTS, 54), 211–45. Jacobi, Renate. “Theme and variation in Umayyad ghazal poetry,” Journal of Arabic Literature 32 (1992), 109–19. eadem, art. “al-Walìd Ibn Yazìd,” in Julie Scott Meisami and Paul Starkey, eds., Encyclopedia of Arabic Literature, 2 vols. London: Routledge, 1998, 803. eadem, “Zur ˝azalpoesie des Walìd ibn Yazìd,” in Wolfhart Heinrichs and Gregor Schoeler, eds., Festschrift Ewald Wagner zum 65. Geburtstag, II: Studien zur arabischen Dichtung. Beirut/Stuttgart: Franz Steiner, 1994 (BTS, 54), 145–61. Juda, Jamal. Die sozialen und wirtschaftlichen Aspekte der Mawàlì in frühislamischer Zeit. Ph.D. diss., Tübingen, 1983. al-Jundì, Darwìsh. ¸àhirat al-takassub wa-atharuhà fì l-shi'r al-'arabì wa-naqdihi. Cairo: Nah∂at Mißr, 1970. Kennedy, Philip F. The Wine Song in Classical Arabic poetry: Abù Nuwàs and the literary tradition. Oxford: Oxford University Press, 1997. al-Khayyà†, Jalàl. Al-Takassub bi-l-shi'r. Beirut: Dàr al-Àdàb, 1970. Pellat, Charles. “Le poète Ibn Mufarri© et son oeuvre,” in Mélanges Louis Massignon. Damascus: Institut Français de Damas, 1956, III, 195–232. Rempis, Christian. “Die altesten Dichtungen in Neupersisch,” Zeitschrift der Deutschen Morgenländischen Gesellschaft 101 (1951), 220–40. Schoeler, Gregor. “Ein Wendepunkt in der Geschichte der arabischen Literatur,” Saeculum 35 (1984), 295–305. Sezgin, Fuat. Geschichte des arabischen Schrifttums, Bd. II: Poesie bis ca. 430 H. Leiden: Brill, 1975. Van Gelder, Geert Jan. “Musàwir al-Warràq and the beginnings of Arabic gastronomic poetry,” Journal of Semitic Studies 36 (1991), 309–27. Wagner, Ewald. Abù Nuwàs: Eine Studie zur arabischen Literatur der frühen 'abbàsidenzeit. Wiesbaden: Harrassowitz, 1965. Weipert, Reinhard. “al-flalìl Ibn A˙mad—A poet?” Oriens 35 (1996), 65–104. Zayd, Mu߆afà Badr. Al-Takassub bi-l-shi'r. Cairo: al-Ma†ba'a l-Salafiyya, 1348 AH.

THE MAWÀLÌ AND EARLY ARABIC HISTORIOGRAPHY Lawrence I. Conrad

In memory of Suliman Bashear And in spite of this the proud heart feels not terror, nor turns to run, and it is his own courage that kills him. Il., XII, 45–46.

In the year 12/633 Arab forces advanced into the Iraqi village of 'Ayn al-Tamr and seized a number of captives whom they reduced to slavery. One of these unfortunates was a young man by the name of Yasàr b. Khiyàr, probably a Christian Arab,1 who was carried off to Medina and sold to the family of the Qurashì Qays b. Makhrama b. al-Mu††alib.2 Yasàr subsequently converted to Islam and was set free, whereupon he associated himself to an Arab tribe as a mawlà seeking a secure place within the structure of Islamic society.3 He had three sons, one of whom, Is˙àq, married the daughter of another mawlà and in 80/699 had a son named Mu˙ammad.4 His non-

1 See Sayf b. 'Umar (d. c. 180/796) in al-ˇabarì (d. 310/923), Ta"rìkh al-rusul wa-l-mulùk, ed. M.J. de Goeje et al. (Leiden: E.J. Brill, 1879–1901), I, 2064:1–10, where another version of the story tells of how 40 Christian boys studying the Bible in a church locked themselves in when the Muslim attackers occupied the town; this group, taken captive when the door was beaten down, included individuals who subsequently converted to Islam and rose to eminence, and many fathers or grandfathers of such figures. 2 On the name and genealogy, see Caskel, ]amharat an-Nasab. Das genealogische Werk des Hi“àm ibn Mu˙ammad al-Kalbì (Leiden: E.J. Brill, 1966), I, Plate no. 15; II, 461. 3 On the mawàlì in this sense a fundamental study is Jamal Juda, Die sozialen und wirtschaftlichen Aspekte der Mawàlì in frühislamischer Zeit (Tübingen: Universität Tübingen, 1983). Cf. also the important article “Mawlà” by Patricia Crone in EI 2, VI, 874b–882b; the rich materials collected in Suliman Bashear, Arabs and Others in Early Islam (Princeton: Darwin Press, 1997); and Crone’s Roman, Provincial and Islamic Law: the Origins of the Islamic Patronate (Cambridge: Cambridge University Press, 1987). 4 Al-Balàdhurì (d. 279/892), Futù˙ al-buldàn, ed. M.J. de Goeje (Leiden: E.J. Brill, 1866), 247:9–10; al-ˇabarì, Ta"rìkh, I, 2122:2–12; al-Mizzì (d. 742/1341), Tahdhìb al-kamàl fì ma'rifat al-rijàl, ed. Bashshàr 'Awwàd Ma'rùf (Beirut: Mu"assasat al-Risàla, 1406–13/1985–92), VII, 33ult–34:3, 35:15–16; Fück, Mu˙ammad ibn Is˙àq (Frankfurt am Main, 1926), 27 note 2.

the

MAWÀLÌ

and early arabic historiography

371

Muslim heritage seems not to have troubled or hindered the lad at all, if we may judge from the fact that he easily joined in with the other boys in harassing and insulting the aged tradent Anas b. Màlik as he walked about Medina.5 All this would in and of itself have been distinctly unremarkable were it not for the fact that this grandson of the Christian enslaved at 'Ayn al-Tamr was none other than the Mu˙ammad b. Is˙àq (d. 151/761) of sìra fame. This is precisely what engaged the early tradents about the subject, and it shows that already in his own time Ibn Is˙àq must have been a renowned scholar. This was likewise the case shortly after his death, when Sayf b. 'Umar gave his own account of the capture of 'Ayn al-Tamr and in reviewing the captives pointedly omitted the name of Ibn Is˙àq’s grandfather,6 almost certainly out of a disinclination to acknowledge the eminence of his predecessor from Medina. And it is naturally what engages our attention as well, since Ibn Is˙àq’s vision of the life of Mu˙ammad entirely dominates the field as we know it. The extant recensions of his Sìra are indispensable to us now,7 and in other sources his influence is omnipresent: transmission from no fewer than 61 of his students is attested in extant sources,8 and one can only guess as how much information on his circle has been lost. This was not, of course, the only instance in which a scholar of mawlà background quickly emerged in a family of converts and rose to particular prominence. A perusal of the first volume of Sezgin’s Geschichte des arabischen Schrifttums will confirm this, and indeed, the phenomenon only seems surprising within the framework of traditional modern scholarship’s tendency to overstress the tensions inherent in relations between Arabs and others in an early Islamic context, as if there ought to have been no difference between a fully assimilated urbanized mawlà of convert background, such as Ibn Is˙àq, and the unconverted peasant mawàlì of the agrarian hinterlands.9 5 Al-Mizzì, Tahdhìb al-kamàl, VII, 35:17–19, as related by Ibn Is˙àq’s student, Salama b. al-Fa∂l (d. 191/807). 6 Al-ˇabarì, Ta"rìkh, I, 2064:1–10. 7 On these, see note 27 below. 8 See Mu†à' al-ˇaràbìshì, Ruwàt Mu˙ammad ibn Is˙àq ibn Yasàr fì l-maghàzì wa-lsiyar wa-sà"ir al-marwiyàt (Beirut/Damascus: Dàr al-Fikr al-Mu'àßir, 1414/1994), for a summary version of which see also his “Ruwàt al-maghàzì wa-l-siyar 'an Mu˙ammad ibn Is˙àq,” Majallat al-majma' al-'ilmì al-'arabì bi-Dimashq 56 (1981), 533–609. 9 An important illustration of the problem is Philip K. Hitti’s influential History of the Arabs (London: Macmillan, 1937), with at least nine subsequent editions into

372

lawrence i. conrad

The phenomenon of rapid assimilation will undoubtedly be noted by other papers in this volume, and in some respects it invites comparison to the way in which genealogical shifts could result in the relocation of an individual, or indeed an entire tribe, from one kinship identity to another. As Ibn Khaldùn (d. 784/1382) described it: It is clear that a person of a certain lineage may happen to fall within the genealogy of another group, whether by reason of taking up residence with them,10 alliance, clientage, or due to flight from his own people because of some crime he committed. He thus comes to be included in the genealogy of this other group, and is counted among them in matters where eschewing enemies, seeking vengeance, bearing the cost of blood money and other such affairs are concerned. . . . Then over a long span of time his original lineage may be forgotten; those who know it will have perished, and thus in most cases it fades away.11

The social and other implications of a shift from one tribe to another could of course be far exceeded by those involved in assumption of mawlà status by a non-Muslim who in many cases would not even have been a speaker of Arabic.12 Still, the case of Ibn Is˙àq finds many parallels, and not only in early and later medieval Islamic

the 1970s, all repeating the same errors. Much of this carries over into works such as M.A. Shaban, The 'Abbàsid Revolution (Cambridge: Cambridge University Press, 1970); id., Islamic History A.D. 600–750 (A.H. 132): a New Interpretation (Cambridge: Cambridge University Press, 1971), though Shaban nevertheless repeatedly stresses assimilationist tendencies. Cf., on the other hand, the more nuanced approaches in Ignaz Goldziher, Muslim Studies, ed. and trans. S.M. Stern and C.R. Barber (London: George Allen and Unwin, 1967–71), I, 98–154; and Bashear, Arabs and Others. Both of these works reflect problematics of their own, however: Goldziher in a lingering tendency from his early career to draw sharp distinctions between supposed ethnic types, and Bashear in an uncertainty as to the extent to which literary debates reflect genuine and general tensions in society. 10 For bi-nuzù"in read bi-nuzùlin, since what is at issue here is not affection but close physical proximity. See, for example, Abù l-Faraj al-Ißfahànì (d. 356/957), Kitàb al-aghànì, ed. A˙mad Zakì al-'Adawì et al. (Cairo: Dàr al-Kutub al-Mißriyya, 1345–94/1927–74), XX, 404:6–405:2, where the rajaz poet Abù Nukhayla (d. 147/ 764) buys a house in the territory of the Banù Óimmàn in order to “correct his lineage.” That is, he wished to be acknowledged as a member of this tribe. Cases of this kind could be multiplied at length. 11 Ibn Khaldùn, Al-Muqaddima, ed. Etienne Quatremère (Paris: Imprimerie nationale, 1858), I, 238:11–18. 12 As seems to be acknowledged, for example, in a report in which al-Zuhrì (d. 124/742) says: “Whoever wants to know about the maghàzì must consult this mawlà of Qays b. Makhrama.” See al-Bukhàrì (d. 256/870), Al-Ta"rìkh al-kabìr, ed. Mu˙ammad 'Abd al-Mu'ìd Khàn (Hyderabad: Dà"irat al-Ma'àrif al-'Uthmàniyya, 1384/1964), I.1, 40:5–6.

the

MAWÀLÌ

and early arabic historiography

373

times. In 1875, for example, the rector of al-Azhar who granted the young Ignaz Goldziher leave to study there was Shaykh Mu˙ammad al-'Abbàsì, who was the grandson of a Copt who in turn, like Ibn Is˙àq’s grandfather, had converted to Islam at a young age. Again the rise of the new Muslim family to scholarly eminence within Islamic circles was very rapid. This grandfather was already an important Shàfi'ì teacher, and his son, Mu˙ammad Amìn al-Mahdì, was a leading Óanafì intellectual and the muftì of Egypt for a time.13 Here it did not take even one generation for a convert family to produce a leading figure in Islamic scholarship. It is not the concern of this study to consider in any further detail the phenomenon sketched out above. Here I will rather seek to explore a proposition implicit in paradigms of this sort. In our sources the mawàlì who achieved eminence in early Islamic scholarship are quite naturally seen in terms of their contributions as Muslims. That is, the Arabic sources present them in terms of the criteria applied to all participants in Islamic culture: who were their (Muslim) teachers and students, what were the main points of their lives (as Muslims), in what centres of (Islamic) culture were they active, how reliably did they pass on what they received of (Islamic) tradition? In some cases this emphasis can become quite radically pronounced. Abù Nu'aym (d. 430/1038), for example, promotes many tradents and compilers of historical material to the status of “Íùfìs” extolled for their humility and piety,14 and in the work of al-Yàfi'ì (d. 767/1366) it is very often difficult to dig an individual’s actual name out from under the array of pious and laudatory epithets bestowed upon him by the compiler.15 In no case known to me can the information in the early Islamic tradition be augmented from Christian sources in Greek or Syriac, not only because eastern Christian historiography is very limited for this period,16 but also because converts are so 13 See Conrad, “The Near East Study Tour Diary of Ignaz Goldziher,” Journal of the Royal Asiatic Society, 1990, 122–23. 14 Abù Nu'aym al-Ißfahànì, Óilyat al-awliyà" wa-†abaqàt al-aßfiyà" (Cairo: Maktabat al-Khànjì, 1354/1935), in eight volumes. 15 Al-Yàfi'ì, Mir"àt al-janàn wa-'ibrat al-yaqΩàn (Hyderabad: Dà"irat al-Ma'àrif al'Uthmàniyya, 1337–39 AH), in four volumes. 16 See Conrad, “Syriac Perspectives on Bilàd al-Shàm During the 'Abbàsid Period,” in Mu˙ammad 'Adnàn al-Bakhìt and Robert Schick, eds., Bilàd al-Shàm During the 'Abbàsid Period (132 A.H./750 A.D.-451 A.H./1059 A.D.) (Amman: University of Jordan, 1412/1991), 1–44. On the more general dearth in Byzantium in particular, see Haldon, Byzantium in the Seventh Century: the Transformation of a Culture (Cambridge:

374

lawrence i. conrad

often seen off in the sources as faceless renegades and targets of general communal ire.17 But it stands to reason that as mawàlì, in the sense of assimilated converts, assumed roles in early Islamic scholarship they would have brought with them a variety of literary perspectives, techniques and motifs current in their own traditions of learning, culture and folklore.18 Two factors weigh in heavily in favor of this proposition. First, in Umayyad and early 'Abbàsid times the population of Syria and Iraq, regions of primary importance in the development of Islamic tradition, was still overwhelmingly Christian. According to the conclusions of Richard Bulliet in his study of conversion patterns, Muslim representation in either region should not have been more than about three or four percent of the total population.19 Much the same is true of Egypt, for which we have the evidence of contemporary papyri.20 So early Islamic culture in all its forms represented the Cambridge University Press, 1990), 425–27; also Kennedy, Greek Rhetoric under Christian Emperors (Princeton: Princeton University Press, 1983), 265–75, on the meagre representation of Greek rhetoric from the sixth century until John of Sardis in the early ninth, and the longer-term discontinuities discussed in Cyril Mango, “Discontinuity with the Classical Past in Byzantium,” in Margaret Mullett and Roger Scott, eds., Byzantium and the Classical Tradition (Birmingham: Centre for Byzantine Studies, University of Birmingham, 1981), 48–57. An account of the collapse of secular culture everywhere in the Near East except the north of Syria, beginning already before the rise of Islam, is given in Lawrence I. Conrad, “Varietas Syriaca: Secular and Scientific Culture in the Christian Communities of Syria after the Arab Conquest,” in G.J. Reinink and A.C. Klugkist, eds., After Bardaisan: Studies on Continuity and Change in Syriac Christianity in Honour of Han J.W. Drijvers (Leuven: Peeters, 1999), 85–105. 17 Cf. al-ˇabarì, Ta"rìkh, I, 2031:15–2032:7. 18 The argument of what follows here stands in sharp contradiction to the views on the role of the mawàlì in early Islamic literary tradition expressed in G.H.A. Juynboll, “On the Origins of Arabic Prose: Reflections on Authenticity,” in Juynboll, ed., Studies on the First Century of Islamic Society (Carbondale/Edwardsville: Southern Illinois University Press, 1982), 161–75. This study proposes that the mawàlì “must have lacked” the Arabs’ predilection for ahistorical storytelling (165) and, where attitudes toward the past were involved, “were not so much concerned with the hyper-romantic view” of the “purely Arab tribesmen” (167). Juynboll thus proposes that the Arab-Islamic sources can be divided into two different categories based on these distinctions: on the one hand, we have the storytellers’ (i.e., the Arabs’) approach to history, and on the other stands that of the mawàlì (168). Hardly any evidence for this argument is brought to bear, and certainly none from the rich late antique literary heritage, which was, of course, the heritage of the mawàlì themselves. Juynboll is of course well aware that one might expect exceptions to such a broad characterisation (168), but in these pages it will become clear that even in general terms the paradigm must be rejected. 19 Bulliet, Conversion to Islam in the Medieval Period: an Essay in Quantitative History (Cambridge/Mass.: Harvard University Press, 1979), 81, 109. 20 See Lapidus, “The Conversion of Egypt to Islam,” Israel Oriental Studies 2 (1972),

the

MAWÀLÌ

and early arabic historiography

375

agenda of a small minority that was pursued in the midst of the cultural activity of a vast majority with other goals and priorities. One would hardly expect that Islamic literary culture could have developed in isolation from this environment, and there is overwhelming evidence against such a notion. There are simply too many cases where influence can easily be detected, and it has long been known that there are many motifs and arguments—not just in the sìra tradition, but also in Qur"ànic exegesis, history and belles lettres—that obviously represent borrowings from the cultural traditions of late antique Christianity and Judaism.21 There are, however, also numerous points that demonstrate influence in the opposite direction as well. When we see a Byzantine historian such as Theophanes (d. 818) discussing ideas and perspectives that we know to be fictions originating in emergent Arab-Islamic tradition, e.g., a paradigm for the 248–62; Bulliet, Conversion to Islam, 92–103; Frantz-Murphy, “Conversion in Early Islamic Egypt,” in Yousef Ragheb, ed., Documents de l’Islam médiéval (Paris: CNRS, 1991), 11–17. 21 The literature on this influence is enormous. See, for example, von M≥ik, “Die Gideon-Saul-Legende und die Überlieferung der Schlacht bei Badr. Ein Beitrag zur ältesten Geschichte des Islàm,” Wiener Zeitschrift für die Kunde des Morgenlandes 29 (1915), 371–83; Josef Horovitz, “Biblische Nachwirkungen in der Sira,” Der Islam 12 (1922), 184–89; id., Koranische Untersuchungen (Berlin: Walter de Gruyter, 1926), 109–11; Speyer, Die biblischen Erzählungen im Qoran (Gräfenhainchen: Schultze, 1931); Katsh, Judaism in Islam: Biblical and Talmudic Backgrounds of the Koran and its Commentaries (New York: Bloch, 1954); Wensinck, “L’influence juive sur les origines du culte musulmane,” trans. Georges Bousquet in Revue africaine 98 (1954), 84–112; Goitein, Jews and Arabs: Their Contacts through the Ages (New York: Schocken Books, 1955), 46–61; Lecomte, “Les citations de l’Ancien et du Nouveau Testament dans l’oeuvre d’Ibn Qutayba,” Arabica 5 (1958), 34–46; E. Rosenthal, Judaism and Islam (London: Yoseloff, 1961), 10–20; F. Rosenthal, “The Influence of Biblical Tradition on Muslim Historiography,” in Bernard Lewis and P.M. Holt, eds., Historians of the Middle East (London: Oxford University Press, 1962), 35–45; Schwarzbaum, Biblical and ExtraBiblical Legends in Islamic Folk-Literature (Walldorf-Hessen: Verlag für Orientkunde Dr. H. Voradran, 1982); Schoeler, “Mündliche Thora und ÓadìΔ: Überlieferung, Schreibverbot, Redaktion,” Der Islam 66 (1989), 213–51; Sachedina, “Early Muslim Traditionists and their Familiarity with Jewish Sources,” in William M. Brinner and Steven D. Ricks, eds., Studies in Islamic and Judaic Traditions II (Atlanta: Scholars Press, 1989), 49–59; Karoui, Die Rezeption der Bibel in der frühislamischen Literatur (Heidelberg: Seminar für Sprachen und Kulturen des Vorderen Orients, 1997); Schöller, Exegetisches Denken und Prophetenbiographie. Eine quellenkritische Analyse der Sìra-Überlieferung zu Mu˙ammads Konflict mit den Juden (Wiesbaden: Otto Harrassowitz, 1998), 206–14; Hary, Hayes, and Astren, eds., Judaism and Islam: Boundaries, Communication and Interaction—Essays in Honor of William M. Brinner (Leiden: E.J. Brill, 2000), 139–256. Of particular importance now are two fundamental works by Uri Rubin: The Eye of the Beholder: the Life of Mu˙ammad as Viewed by the Early Muslims, a Textual Analysis (Princeton: Darwin Press, 1995), and Between Bible and Qur"àn: the Children of Israel and the Islamic SelfImage (Princeton: Darwin Press, 1999).

376

lawrence i. conrad

overall organisation and division of the ancient tribes of Arabia that was in fact invented in the Umayyad period,22 then the Greek historian’s ultimate source cannot be other than this emergent ArabIslamic tradition.23 Second, and closely related to this point, there are enormous differences between the cultural activities we can detect in pre-Islamic Arabia and those that increasingly come to our attention in early Islamic society. Perhaps unintentionally, Josef Horovitz laid the groundwork for a perspective in favor of some broad sense of continuity in a seminal study in which he argued that the early Arab-Islamic accounts of Mu˙ammad’s campaigns continue the pre-Islamic tradition of the ayyàm al-'arab, the “battle days” narratives of the ancient Arab tribes.24 But in reality the ayyàm al-'arab tales are really very different,25 and in any case no pre-Islamic Arabic prose survives to 22 See Goldziher, Muslim Studies, I, 45–97, 125–36, 164–90; Caetani, Annali dell’Islam (Milan: Ulrico Hoepli, 1905–24), I, 58–90; Caskel, Die Bedeutung der Beduinen in der Geschichte der Araber (Köln/Opladen: Westdeutscher Verlag, 1953), 13–18; id., ]amharat an-Nasab, I, 19–47; Henninger, “La società bédouine ancienne,” in Francesco Gabrieli, ed., L’Antica società beduina (Rome: Università di Roma, 1959), 69–93; id., “Altarabische Genealogie,” Anthropos 61 (1966), 852–70. The dynamics of such creations are neatly summarised in the comment by Ibn Khaldùn cited above; for the modern situation, see Musil, Manners and Customs of the Rwala Bedouins (New York: American Geographical Society, 1928), 46–50; Jabbur, The Bedouins and the Desert: Aspects of Nomadic Life in the Arab East, trans. Lawrence I. Conrad, ed. Suhayl S. Jabbur and Lawrence I. Conrad (Albany: State University of New York Press, 1995), 262–72. 23 Conrad, “Theophanes and the Arabic Historical Tradition: Some Indications of Intercultural Transmission,” Byzantinische Forschungen 15 (1990), 1–44. Theophanes’ source was a Greek recension of the Syriac chronicle of Theophilus of Edessa, on whom more will be said below. 24 Horovitz, “Die poetischen Einlagen der Sìra,” Islamica 2 (1926), 310. It is not clear to me that Horovitz would have agreed with some of the arguments derived from this passage in later scholarship, since his aim was solely to explain the presence of poetry in the sìra and not to account for the origins of historical writing among the Arabs. On the ayyàm lore, see Caskel, “Aijàm al-'arab. Studien zur altarabischen Epik,” Islamica 3, Ergänzungsheft (1930), 1–99; Meyer, Der historische Gehalt der Aiyàm al-'Arab (Wiesbaden: Otto Harrassowitz, 1970); Faruqi, Early Muslim Historiography: a Study of Early Transmitters of Arab History from the Rise of Islam up to the End of the Umayyad Period (612–750 A.D.) (New Delhi: Idarah-i Adabiyat-i Delli, 1979), 37–42; Duri, The Rise of Historical Writing among the Arabs, ed. and trans. Lawrence I. Conrad (Princeton: Princeton University Press, 1983), 16–20; and the collection of 'Àdil Jàsim al-Bayàtì, Kitàb ayyàm al-'arab qabla l-Islàm (Beirut: 'Àlam al-Kutub, 1407/1987). 25 Cf. Conrad, “The Conquest of Arwàd: a Source-Critical Study in the Historiography of the Early Medieval Near East,” in Averil Cameron and Lawrence I. Conrad, eds., The Byzantine and Early Islamic Near East, I: Problems in the Literary Source Material (Papers of the First Workshop on Late Antiquity and Early Islam (Princeton: Darwin Press, 1992), 387–88; and the counterargument of Fred Donner in his

the

MAWÀLÌ

and early arabic historiography

377

offer a basis for comparison, except in recensions that do not predate the second/eighth century and which therefore have been subject to revisions in content, style and dialect over an extended period,26 not to mention recasting under the influence of the “political and social currents” prevailing in early Islamic times.27 As a genre, the ayyàm al-'arab cannot be traced any earlier than the Baßran compiler Abù 'Ubayda (d. 209/824). That is, in the form in which we have them today they are of lesser demonstrable antiquity than either the sìra tradition, for which we have recensions of the materials taught and circulated by Ibn Is˙àq,28 or the futù˙ tradition, which is clearly attested from the mid-second century AH on by the career of Abù Mikhnaf (d. 157/774),29 by fragments of the Kitàb al-ridda wa-l-futù˙ wa-Kitàb al-jamal wa-masìr 'À"isha wa-'Alì by Sayf b. 'Umar (d. ca. 180/796),30 and by extant recensions of the late second-century works of al-Azdì (wr. ca. 190/805)31 and Ibn A'tham al-Kùfì (wr.

Narratives of Islamic Origins: the Beginning of Islamic Historical Writing (Princeton: Darwin Press, 1998), 179–80. 26 See the cautious assessment of these issues in Alan Jones, “The Prose Literature of Pre-Islamic Arabia,” in J.R. Smart, ed., Tradition and Modernity in Arabic Language and Literature (Richmond, Surrey: Curzon, 1996), 229–41; id., “The Qur"àn in the Light of Earlier Arabic Prose,” in Alan Jones, ed., University Lectures in Islamic Studies, I (London: Altajir World of Islam Trust, 1997), 67–83. This crucial difficulty is ignored in any effort to describe pre-Islamic Arabic prose on the basis of these later recensions. See, for example, Hàshim Íàli˙ Mannà', Al-Nathr fì l-'aßr al-jàhilì (Beirut: Dàr al-Fikr al-'Arabì, 1993). 27 Duri, Rise of Historical Writing, 19. Gerald R. Hawting goes even further and argues in his recent The Idea of Idolatry and the Emergence of Islam: From Polemic to History (Cambridge: Cambridge University Press, 1999) that accounts of pre-Islamic religion betray a monotheistic agenda in their origins. 28 Ibn Is˙àq, Kitàb al-siyar wa-l-maghàzì, recension of Yùnus b. Bukayr, ed. Suhayl Zakkàr (Damascus: Dàr al-Fikr, 1398/1978); another version, Sìra, ed. Mu˙ammad Óamìd Allàh (Rabat: Ma'had al-Diràsàt wa-l-Ab˙àth li-l-Ta'rìb, 1976); Ibn Hishàm, Sìrat rasùl Allàh, ed. Ferdinand Wüstenfeld (Göttingen: Dieterische UniversitätsBuchhandlung, 1858–60). Cf. also Alfred Guillaume, New Light on the Life of Mu˙ammad (Manchester: Manchester University Press, 1960). 29 See al-Nadìm (wr. 378/989), Kitàb al-fihrist, ed. Ri∂à-Tajaddud (Tehran: Ma†ba'at Dànishgàh, 1393/1973), 105:18, for Abù Mikhnaf ’s futù˙ works on the conquests in Syria, Iraq, and Khuràsàn; also the detailed research in Ursula Sezgin, Abù Mi¢naf. Ein Beitrag zur Historiographie der umaiyadischen Zeit (Leiden: E.J. Brill, 1971), 42–47, 52–53, 63, 102, 114. 30 Sayf b. 'Umar, Kitàb al-ridda wa-l-futù˙ wa-Kitàb al-jamal wa-masìr 'À"isha wa-'Alì, ed. Qàsim al-Samarrà"ì (Leiden: Smitskamp Oriental Antiquarium, 1995). 31 There are three recensions of this work: Futù˙ al-Shàm, ed. William Nassau Lees (Calcutta: Baptist Mission Press, 1854); Ta"rìkh futù˙ al-Shàm, ed. 'Abd alMun'im 'Abd Allàh 'Àmir (Cairo: Mu"assasat Sijill al-'Arab, 1970); and the recension of al-ˇalamankì (d. 429/1037) in anon., Al-Durr al-nafìs fì uns al-Ωà'in wa-l-jalìs, Real

378

lawrence i. conrad

204/819).32 The assumption of continuity is itself precarious, since as Fred Donner has recently demonstrated, the earliest Muslim community did not envisage their affairs as history in the making, but rather was concerned with issues of personal and communal piety in preparation for an imminently expected Day of Judgment.33 The ayyàm of course do not exhaust the material handed down in Islamic times as reflecting ancient Arabian culture: there is an enormous corpus of poetry, as well as proverbs, wise adages and popular folk tales. But if one accepts that all this in its extant form can tell us a great deal about culture in pre-Islamic Arabia (and I am personally much inclined to this viewpoint), then—quite apart from the fact that none of this is history or historical—what we see is material that moralises and upholds the traditional tribal values that society holds dear in any case. This lore does not envisage disputes with a rival moral system, much less the need to argue in favor of its own perspectives, and it does not seek (or need, for that matter) proofs for its values and ideals in any vision of the past as continuous history. The integrity of its position does not depend on any ability to deploy sophisticated dialectical skills or subtle rhetorical techniques and arguments based on historical examples. Those who defy the established norms are not engaged in disputes: they shame their kin and will quickly be brought into line by them or Academia de la Historia (Madrid), Ms. Gayangos no. 18. On this text see Conrad, “Al-Azdì’s History of the Arab Conquests in Bilàd al-Shàm: Some Historiographical Considerations,” in Mu˙ammad 'Adnàn al-Bakhìt, ed., Proceedings of the Second Symposium on the History of Bilàd al-Shàm During the Early Islamic Period Up to 40 A.H./640 A.D., English and French papers (Amman: University of Jordan, 1987), I, 28–62; LandauTasseron, “New Data on an Old Manuscript: an Andalusian Version of the Works Entitled Futù˙ al-Shàm,” Al-Qan†ara 21 (2000), 361–80. Cf. further Mourad, “On Early Islamic Historiography: Abù Ismà'ìl al-Azdì and his Futù˙ al-Shàm,” Journal of the American Oriental Society 120 (2000), 577–93; Conrad, art. “Al-Azdì” in Julie Scott Meisami and Paul Starkey, eds., Encyclopedia of Arabic Literature (London: Routledge, 1998), I, 117b–118a. 32 Ibn A'tham al-Kùfì, Kitàb al-futù˙, ed. Mu˙ammad 'Abd al-Mu'ìd Khàn (Hyderabad: Dà"irat al-Ma'àrif al-'Uthmàniyya, 1388–95/1968–75). This nearly complete version of the work represents a later recension, but a part of the earliest version survives in a manuscript erroneously ascribed to al-Wàqidì (d. 207/823) and published as such by Mu˙ammad Óamìd Allàh, Kitàb al-ridda wa-nubdha min futù˙ al'Iràq (Paris: Editions Tougui, 1409/1989). Cf. Muranyi, “Ein neuer Bericht über die Wahl des ersten Kalifen Abù Bakr,” Arabica 25 (1978), 233–60. For a survey of the structure, recensions and date of Ibn A'tham’s history see Conrad, art. “Ibn A'tham al-Kùfì,” in Meisami and Starkey, eds., Encyclopedia of Arabic Literature, I, 318b–319a. 33 Donner, Narratives of Islamic Origins, 64–97.

the

MAWÀLÌ

and early arabic historiography

379

else will be expelled from their protecting domain and excluded from the security provided by customary law.34 But a sophisticated sense of history, large-scale recourse to subtle literary motifs and techniques, and specific argument against rival value systems are of course extremely common in late antique society, so when we see all this emerging at a rapid pace among Muslims in early Islamic times, not least of all in historiography, one of our first questions must be: Where is this coming from? This specific issue will engage our attention at the conclusion of this study, and in the interim I will seek to establish that the influence of ideas introduced by non-Arab converts and their descendents should not be seen as an intrusion via points of detail into an Islamic historical tradition already wellgrounded in the literary culture of sixth and seventh-century Arabia, but rather as a fundamental and decisive contribution by the community of the vast majority to the newly emerging literary tradition—including historiography—of the small ruling elite in its midst. It must be clear from the outset that in speaking of indigenous influence on the Arab-Islamic historical tradition, brought to bear by mawàlì entering the ranks of Muslim scholarship, the scope of our inquiry is not limited to manifest borrowings of specific points or arguments for use in accounts with an obvious utility for purposes of religious polemics and apologetics. To take one example, Sayf tells of the Byzantine commander “Georgius” who meets Khàlid b. al-Walìd and asks him how new converts can be equal to the Arabs in religious merit when the latter have preceded them in adherence to Islam. Khàlid replies that the Arabs of course should have accepted Islam, because they had known the Prophet and had seen him receiving the revelations of the Qur"àn; Georgius’ conversion, on the other hand, would be of far greater merit precisely because he had not witnessed these events, and so must be accepting Islam on the basis of faith alone. The general thereupon converts to Islam, performs two rak'as, and then rushes forth and fights his former compatriots with Khàlid from sunrise to sunset before he is finally killed in battle.35 This exchange is obviously borrowed straight from the New Testament, where the doubting Thomas is 34 See, for example, Schacht, An Introduction to Islamic Law (Oxford: Clarendon Press, 1964), 6–9; and on poetry as reflective of pre-Islamic attitudes, Reinert, Das Recht in der altarabischen Poesie (Köln: Universität Köln, 1963). 35 Al-ˇabarì, Ta"rìkh, I, 2097:16–2099:2.

380

lawrence i. conrad

admonished by the risen Christ for refusing to believe until he has seen the wounds in Jesus’ hands and feet and side.36 The larger import of the story would immediately have been appreciated by mawàlì of Christian origin, and it was probably among them (i.e., not with Sayf ) that it originated. But this sort of piecemeal borrowing for specific purposes still allows us to conceive of the early Arab-Islamic tradition as a more or less entirely Arab-Muslim affair, reflecting ideas and cultural sensibilities originating within an Arab cultural elite or continued by them from an Arab Arabian heritage. It is precisely this that I would like to call into question, and I will argue here that the influence of the indigenous culture of the late antique Near East was probably pervasive: its role was not to provide a pre-existing tradition with scattered ideas here and there in the field of apologetics, but rather to participate in the shaping of the tradition itself and to contribute major paradigms and techniques as well as a host of details. To establish in a comprehensive way the means and channels for the influence of the literary tradition of classical and late antiquity on early Arabic historical writing would be the task of a substantial monograph based on a full review of the relevant Graeco-Roman and early Islamic literature. This would show in detail how this occurred. But here my concern is rather to demonstrate only that it occurred, and I will limit myself to a narrow range of illustrative indications drawn from one of the liveliest subfields of antique historical writing, the tradition of retellings of the story of the Trojan War. For comparative purposes this tradition is particularly useful. Like the early Arabic tradition it involves a great deal of popular storytelling. It is also extremely argumentative and seeks to sustain and promote the truth value of a self-serving vision of the past; just as the Trojan War tradition is engulfed by questions of whether Greeks or non-Greeks are the true heirs to the classical heritage and whether Troy deserved to be destroyed, the early Arabic tradition is concerned to argue such highly charged issues as the truth of Islam as the quintessential monotheistic dispensation (especially as opposed to the claims to the same status by Christianity) and the relative status in the new order enjoyed by various ethnic, regional and social groups. Finally, the Trojan War tradition is one that can

36

John 20:29; I Peter 1:8; Hebrews 1:1.

the

MAWÀLÌ

and early arabic historiography

381

be reliably traced in both professional and popular circles from the canonization of the Iliad in antiquity right through to its presentation in the history of Theophilus of Edessa in north Syria in the mid-eighth century. A review of the relevant literature is itself interesting and revealing. The texts of this tradition that I have used here are the following: The ILIAD and the ODYSSEY as they have come down to us in a written form attributed to the poet Homer.37 The possibility that the two works were composed by different poets, as well as the argument that neither epic can be the product of a single creative imagination, need not detain us, since what we are concerned with is the form in which the works were known in classical antiquity and subsequent times, when it was taken for granted that Homer, whoever he was, had written both poems. Neither tells the story of the Trojan War, of course, but both presume a corpus of well-known lore on the subject, and Homer’s work does lay the foundations, however unintentionally, for a more pointed and polemical approach to the subject. Homer uses not the Achaeans but the Trojans to explore issues of suffering and the horrors of war, and even the gods lament the impending destruction of Priam’s city.38 At the same time, however, the poet is not unattached to the Achaean side. Far more Trojans than Achaeans fall in the Iliad,39 and Zeus states that Achilles is a greater warrior than all the Trojans combined and could take Troy by himself.40 Achaeans defeated in combat and facing death never plead for their lives, as Trojans often do, and Homer has Zeus justify the ruin of Troy “because in violent assembly they pass decrees that are crooked, and drive righteousness from among them and care nothing for what the gods think.”41 So in Homer’s work there was a marked issue orientation that encouraged later exploration of various themes, already, for example, in Greek tragedy. Homer would 37 In this paper I have used the masterful translations by Richmond Lattimore of both the Iliad (Chicago: University of Chicago Press, 1951) and the Odyssey (New York: Harper Collins, 1965), with occasional reference to the Oxford Classical Texts edition, David B. Monro and Thomas W. Allen, eds., Homeri opera (Oxford: Clarendon Press, 1912–20), vols. 1–4. 38 Il., IV.44–49. 39 Yes, someone has counted: see Richard Janko in G.S. Kirk, gen. ed., The Iliad: a Commentary (Cambridge: Cambridge University Press, 1985–93), IV, 272. 40 Il., XX.26–27. 41 Ibid., XVI.387–88.

382

lawrence i. conrad

not have found it possible to draw sharp political or ethnic distinctions between Achaeans and Trojans: all spoke Greek, they worshipped the same gods, and their customs were—if not identical—at least mutually intelligible. By the era of Euripides, however, times had changed, as we can see in a passage sung by the chorus in one of his plays in lament for the destruction of Troy: Woe for our babes, for our fathers hoary! Woe for our country mid smoke and smoulder Crashing to ruin, and all her glory Spear-spoiled—and an alien land shall behold her Bond who was free; for that Asia’s shoulder Is bound under Europe’s yoke, and I dwell, An exile from home, in a dungeon of hell.42

Homer would have found it rather curious to contrast Achaeans and Trojans in terms of Asian and European peoples, but in many ways the tragedian is simply fleshing out old Homeric problems in terms of the priorities and sensibilities of a later time. The Epic Cycle.43 This was a collection of poems assembled about 500 BCE to fill in events and details not provided by Homer in the Iliad and Odyssey, using as its source the stories of the Trojan War already known in Homer’s time. The Cycle now survives only in the form of quotations scattered through classical literature and a summary by Photius (wr. 855) in his Bibliotheca, which in turn is based on an earlier summary by Proclus (d. 485) in his Chrestomathia.44 42

Euripides, Hecuba, 477–83. The poems of the Epic Cycle are cited here from vol. 5 of Munro and Allen, Homeri opera, following their numbering of the fragmenta. 44 See Welcker, Der epische Cyclus; oder Die homerischen Dichter, 2nd ed. (Bonn: E. Weber, 1849–65); T.W. Allen, “The Epic Cycle,” Classical Quarterly 2 (1908), 64–74, 81–88; Severyns, Le cycle épique dans l’école d’Aristarque (Paris: E. Champion, 1928); Griffin, “The Epic Cycle and the Uniqueness of Homer,” Journal of Hellenic Studies 97 (1977), 39–53; Malcolm Davies, The Epic Cycle (Bristol: Bristol Classical Press, 1988), esp. 1–12; id., “The Date of the Epic Cycle,” Glotta 67 (1989), 89–100. The so-called neo-analytical approach of critics primarily associated with the work of Wolfgang Kullmann, however, regards the poems of the Epic Cycle as the sources of Homer. See, for example, Kakrides, Homeric Researches (Lund: C.W.K. Gleerup, 1949); Schadewaldt, Von Homers Welt und Werk. Aufsätze zur homerischen Frage, 2nd ed. (Stuttgart: K.F. Koehler Verlag, 1951), 155–202; Kullmann, Die Quellen der Ilias (Troischer Sagenkreis) (Wiesbaden: Franz Steiner Verlag, 1960), esp. 18–57, 204–357. Cf. also Kullmann’s subsequent clarifications in several articles: “Zur Methode der Neoanalyse in der Homerforschung,” Wiener Studien, Neue Folge, 15 (1981), 5–42; “Oral Poetry Theory and Neoanalysis in Homeric Research,” Greek, Roman and 43

the

MAWÀLÌ

and early arabic historiography

383

It is in the Epic Cycle that one finds accounts of such famous episodes as the battle with the Ethiopians and the Amazons and the ruse of the wooden horse. An interpretation of the Trojan War cannot be reconstructed from the extant fragments, but these are nevertheless useful for the many common literary motifs they deploy. The AENEID of Virgil (d. 19 BCE). Book II of Virgil’s epic of Roman imperial destiny is devoted to the Trojan War. Dido, queen of Carthage, welcomes Aeneas and his throng of refugees from the destruction of Troy and asks him to “tell us, my guest, from the first beginning the treachery of the Greeks, the sad fate of your people, and your own wanderings.”45 The son of Priam responds with a tale that stresses the cruelty, barbarousness and duplicity of the Greeks, and holds the Trojans blameless for any responsibility for the outbreak of the war. The latter can pride themselves in the fact that they broke Greek power for ten years, and there is no glory for the Greeks in the slaughter of a sleeping city. The war and the fall of Troy were brought about by the gods, who had other plans and another destiny in mind for Aeneas.46 Book II of the Aeneid thus uses the story of the Trojan War to emblazon into Roman imperial mythology sharp judgmental distinctions between Greeks and Trojans/Romans and to justify the latter at the extreme expense of the former. Virgil’s own “Iliad” appears subsequently, with distinct parallels adduced between the paradigms of Greek and Roman history.47 Given the renown of Virgil’s epic, this practically guaranteed that the subject would become the centre of a continuing debate. The EPHEMERIS BELLI TROIANI attributed to Dictys Cretensis, written about a century after the Aeneid.48 In the prologue to this book we Byzantine Studies 25 (1984), 307–23; “Ergebnisse der motivgeschichtlichen Forschung in Homer (Neoanalyse),” in Joachim Latacz, ed., Zweihundert Jahre Homer-Forschung. Rückblick und Ausblick (Stuttgart/Leipzig: B.G. Teubner, 1991), 425–55. For overviews of the neo-analytical school see Leinieks, “The Iliad and the Epic Cycle,” Classical Outlook 52 (1975), 62–64; Clark, “Neoanalysis: a Bibliographical Review,” Classical World 79 (1985–86), 379–94; Edwards, “Neo-Analysis and Beyond,” Classical Antiquity 9 (1990), 311–25. 45 Aeneid, I.753–55. 46 Aeneid, II.13, 601–604, 777–78, III.1–5. Cf. Beye, Ancient Epic Poetry: Homer, Apollonius, Virgil (Ithaca/London: Cornell University Press, 1993), 235–36. 47 Anderson, “Vergil’s Second Iliad,” Transactions and Proceedings of the American Philological Society 88 (1957), 17–30. 48 See Collilieux, Etude sur Dictys de Crête et Darès de Phrygie (Grenoble: Xavier

384

lawrence i. conrad

are told that Dictys was a Cretan companion of an eminent Greek, whose name, Idomeneus, can of course be found in the Iliad.49 This Cretan warrior kept a diary of his experiences “in Phoenician characters” (litteris Punicis), and the work was buried in a lead box with him when he died. An earthquake (or just the collapse of the tomb) in 66–67 CE resulted in the discovery of the manuscript by some shepherds, and the diary was translated on the order of the emperor Nero.50 The appeal to an ancient book discovered in a tomb is a well-worn topos among forgers of texts of this kind;51 the Greek version referred to in the prologue is obviously the original form of the work and ranks among the many allegedly pre-Homeric texts that emerged in Hellenistic times and after.52 The Ephemeris opposes the Aenead and argues in favour of the Greeks on the superior authority of an eyewitness.53 The alterity of the Trojans is stressed by firstperson narration that, inter alia, repeatedly calls the Trojans “barbarians”54 and by assertion that the conflict was not about the return of Helen or redemption of the honour of Menelaus, but about who will rise to dominion—the Greeks (i.e., the bearers of civilisation) or the Trojans (the champions of barbarism).55 The Iliad ’s sympaDrevet, 1886), 61–87; Griffin, Dares and Dictys: an Introduction to the Study of Medieval Versions of the Story of Troy (Baltimore: J.H. Furst, 1907), 1–3, 7–120; Allen, Homer: the Origins and the Transmission (Oxford: Clarendon Press, 1924), 130–76; Venini, “Ditti Cretese e Omero,” Istituto Lombardo di scienze e lettere, Classe di lettere e scienze, morali e politiche, Memorie 37 (1981), 161–98; Merkle, Die Ephemeris belli Troiani des Diktys von Kreta (Frankfurt am Main: Peter Lang, 1989); id., “Troiani belli verior textus. Die Trojaberichte des Dictys und Dares,” in Horst Brunner, ed., Die deutsche Trojaliteratur des Mittelalters und der frühen Neuzeit (Wiesbaden: Ludwig Reichert Verlag, 1990), 494–507. 49 Il., II.645–50. 50 Cf. ps.-Dictys, Ephemeris Epistula/Prologus, Latin translation ed. Werner Eisenhut, 2nd ed. (Leipzig: B.G. Teubner, 1973), 1:1–3:14. 51 Collilieux, Etude, 61–65; Wolfgang Speyer, Die literarische Fälschung im heidnischen und christlichen Altertum. Ein Versuch ihrer Deutung (Munich: C.H. Beck, 1971), 67–68. 52 Discussion of this phenomenon in Speyer, Literarische Fälschung, 13–106; Kindstrand, Homer in der Zweiten Sophistik (Stockholm: Almqvist und Wiksell, 1973), 221–29; Gianotti, “Le metamorfosi di Omero. Il romanze di Troia della specializzazione delle scholae ad un publico di non specialisti,” Sigma 12 (1979), 15–32; Anderson, Philostratus: Biography and Belles Lettres in the Third Century A.D. (London: Croom Helm, 1986), 241–57; Merkle, Ephemeris, 16–21, 33–82, 249–59. 53 Cf. the detailed discussion in Merkle, Ephemeris, 142–99. 54 From context it is clear that the word is not meant in the sense of the classical Greek barbaroi, “foreigners,” but rather has definite negative connotations. Cf., for example, Ephemeris, III.10, 23; V.5, 11 (ed. Eisenhut, 67:5–8, 77:9–12, 106:15, 112:2–6). 55 Ibid., III.23 (ed. Eisenhut, 77:7–12).

the

MAWÀLÌ

and early arabic historiography

385

thy for the Trojans is rejected, and the sack of the city is attributed to the self-serving and unreasonable behaviour of the Trojans themselves. The Roman apologetic agenda of the Aenead is also dismissed. Troy was not destroyed, but rather only sacked after it was betrayed by its own notables, including Aeneas, who left Troy rejected by his own people and not as a defiant hero. Oration 11 of Dio Chrysostom. A prominent benefactor of his native city of Prusa and a scholar keenly dedicated to the Hellenistic heritage,56 Dio wrote this piece about 71–80 CE.57 Pretending that he is actually speaking from Troy, he first argues that Homer is essentially anti-Trojan, and further, a liar: his vitae show that he was a beggar, and everyone knows that all beggars are incorrigible liars.58 Dio himself will set the record straight by refuting Homer with verses from the Iliad and Odyssey themselves and by correcting Homer’s account with information Chrysostom has received from an elderly priest in Egypt, where all past history is recorded in the temples.59 Helen was never abducted, the Greeks never captured Troy, Hector was not killed by Achilles or anyone else, Priam also survived, and Aeneas sets out westward as a conqueror rather than as a fugitive. What Dio’s oration achieves is the replacement of Virgil’s Phoenixlike vision of the origins of Roman greatness with a paradigm of continuous and uninterrupted glory, quite in keeping with Dio’s dedication to the Hellenistic heritage. That this is at the expense of the Greeks is of course still rigorously maintained, and he seems also to have in mind for Prusa a role as Troy’s successor as a light of civic glory in Asia Minor. The DE EXIDIO TROIAE HISTORIA attributed to a certain Dares Phrygius. Here again we have a supposed eyewitness to the conflict, this time

56 See von Arnim, Leben und Werke des Dio von Prusa (Berlin: Weidmannsche Buchhandlung, 1898); Desideri, Dione di Prusa, un intellectuale greco nell’impero romano (Messina/Florence: G. D’Anna, 1978); Jones, The Roman World of Dio Chrysostom (Cambridge/Mass.: Harvard University Press, 1978). 57 See von Arnim, Dio von Prusa, 166–71, 181–204; Desideri, Dione di Prusa, 431–34; Jones, The Roman World of Dio Chrysostom, 17–18, 45, 134; Seeck, “Dion Chrysostomos als Homerkritiker (or. 11),” Rheinisches Museum für Philologie, Neue Folge, 133 (1990), 97–107. 58 Dio Chrysostom, Orationes, XI.15. 59 Ibid., XI.7, 38.

386

lawrence i. conrad

a Phrygian ally of the Trojans who participated in the entire war.60 An epistle attached to the beginning of the work describes the discovery of the autograph Greek manuscript of the text in Athens and the preparation of a precise Latin translation.61 As with the topos of the opened tomb in ps.-Dictys, the discovery of a rare and wondrous book in a library in the ps.-Dares is yet another motif used to explain why an allegedly ancient work has come to be known only now, nearly a millennium after its composition.62 As was the case with “Dictys,” one is hardly surprised to find the source for the name “Dares” in Homer.63 Whether or not a Greek original ever existed is open to debate, but it seems clear that the book is late and probably originated in the fifth century CE.64 What survives of this is an even later Latin recension executed to a poor critical standard.65 No time is wasted in getting to the point of the text or demonstrating the importance of the motif of the eyewitness, for the reader is invited to consider who has the better account of the war: Dares, who was a contemporary observer to the fighting, or Homer, who was not born until long after the end of the war and whom the Athenian scholars deemed insane for portraying gods in conflict with human beings.66 The anti-Greek agenda of this work is thus evident 60 Collilieux, Etude, 88–101; Griffin, Dares and Dictys, 4–5, 9–17; Schissel von Fleschenberg, Dares-Studien (Halle-an-der-Salle: Max Niemeyer, 1908); Merkle, “Troiani belli verior textus,” 507–22; Beschorner, Untersuchungen zu Dares Phrygius (Tübingen: Gunter Narr Verlag, 1992), 193–270. The name Dares is itself Phrygian; see Hans von Kamptz, Homerischen Personennamen. Sprachwissenschaftliche und historische Klassifikation (Göttingen: Vandenhoeck und Ruprecht, 1982), 338–39. 61 Ps.-Dares Phrygius, De excidio Troiae historia Prologus, 12; ed. Ferdinand Meister (Leipzig: B.G. Teubner, 1873), 1:1–17, 14:9–12. The surviving text is in fact written in very bad Latin. Cf. also Schissel von Fleschenberg, Dares-Studien, 84–96, 157–60; Beschorner, Untersuchungen, 107–108, on the comments in ch. 12. 62 Speyer, Literarische Fälschung, 68–70. 63 Il., V.9, where Dares is a “blameless and bountiful” Trojan priest with two sons who fought in the war. Cf. G.S. Kirk in Kirk, gen. ed., Commentary, II, 54. 64 See Merkle, Ephemeris, 263–83; Beschorner, Untersuchungen, 254–63. Cf. Schissel von Fleschenberg, Dares-Studien, 128–33, 160–69, arguing for a Greek original in the first century CE and dating the Latin version to 510–30 CE. 65 Schetter, “Dares und Dracontius über die Vorgeschichte des trojanischen Krieges,” Hermes 115 (1987), 211–31; id., “Beobachtungen zum Dares Latinus,” Hermes 116 (1988), 94–109. Cf. also Bradley, “Troy Revisited,” Hermes 119 (1991), 232–46, which argues against assertions for the well-planned structure and content of the Latin text advanced in Lumiansky, “Dares’ Historia and Dictys’ Ephemeris: a Critical Comment,” in E. Bagby Atwood and Archibald A. Hill, eds., Studies in Language, Literature and Culture of the Middle Ages and After (Austin: University of Texas at Austin, 1969), 200–209. 66 Ps.-Dares, Historia Prologus (ed. Meister, 1:10–16).

the

MAWÀLÌ

and early arabic historiography

387

from the first page. The war is blamed on the Greeks, whose kidnapping of Priam’s daughter Hesione led to Paris’ tit-for-tat kidnapping of Helen with a trade of victims in mind.67 The Greek response, however, was the siege of Troy, which ps.-Dares presents largely in terms of repeated negotiations and truces. Eventually the city is taken, largely through the treachery of self-serving notables and conniving women. Basically, the ps.-Dares text is a reply to the ps.-Dictys, a pro-Trojan retelling of the story of the war with clear reference to a divide between Europe and Greek-speaking lands to the east. The POSTHOMERICA by Quintus Smyrnaeus, an obscure fourth-century poet who apparently lived in Asia Minor.68 He claims to have been inspired by the Muses while tending his sheep on a mountainside near Smyrna,69 but Smyrna was one of the main claimants to the honour of being the birthplace of Homer, and our poet may simply be trying to associate himself with the legacy of his predecessor. His poem on the Trojan War is in fourteen books, each on one of the themes of the lore on the Trojan War, which Quintus knows primarily through his reading of ps.-Dictys and the great classical tragedians—Aeschylus, Sophocles and Euripides.70 There is no proem, and the text rather follows on from the last line of the Iliad. The poet shuns the prevailing trend of viewing the war in either proTrojan or pro-Greek terms and attributes the conflict and the ruin of Troy to fate and the inscrutable will of the gods. His interest is rather in the human dimension of the subject, as one can see, for example, in his moving account of how the mortally wounded Paris staggers back to Troy and seeks out his wife Oenone (a figure unknown to either Homer or the Cyclic poets) to beg her to save

67

Ibid., 7–11 (ed. Meister, 9:1–14:8). Glover, Life and Letters in the 4th Century (Cambridge: Cambridge University Press, 1901), 77–101; Paschal, A Study of Quintus of Smyrna (Chicago: University of Chicago Press, 1904); Knight, Vergil’s Troy: Essays on the Second Book of the Aeneid (Oxford: Basil Blackwell, 1932), 71–104; Vian, Recherches sur les Posthomerica de Quintus de Smyrne (Paris: C. Klincksieck, 1959). 69 Quintus Smyrnaeus, Posthomerica, XII.308–13. 70 See Vian, Recherches—ps.-Dictys: 101–108; Aeschylus: 26–28, 75, 82–84; Sophocles: 41–51, 56 note 1, 67, 78; Euripides: 51, 53–54, 74, 75, 76–78, 94–95, 97–98, 100, 109; also the often-criticised but still useful remarks in Anton Baumstark, “Die zweite Achilleus-Trilogie des Aeschylos,” Philologus 55 (1896), 277–306. Quintus seems not to have direct access to the poems of the Epic Cycle (Vian, Recherches, 87–94). 68

388

lawrence i. conrad

his life. Filled with jealousy and hate after his philandering with Helen, however, she drives him away to die alone, only to realise later that she still loves him and regrets her cruel words to him. Overwhelmed with grief and remorse she joins him on his funeral pyre, and the ashes of the two are buried together.71 The CHRONICLE of Theophilus of Edessa. The author (d. 785) was a Maronite scholar from north Syria, and his Syriac history was a chronicle, divided into short books (mèmrè ), beginning with Creation and ending with the 'Abbàsid revolution in 750. This history is lost, but it was so frequently used by later authors writing in Greek, Syriac and Arabic that its presence and contents are often easy to establish;72 Theophilus’ account of the Trojan War survives in the anonymous Chronicle of 1234 under the rubric “Concerning the Wars against Troy and its Destruction,” and the chronicler of 1234 describes his source as Books 43 to 51 of a Chronographia that he (or more likely, a later copyist) attributes to none other than Homer himself.73 Basing himself on his own Syriac translations of the Greek poems of the Epic Cycle, supplemented with information available from other lore, Theophilus offers a virulently anti-Greek account of the war in which he stresses Greek cruelty, ferocity and barbarousness. Theophilus is known to have used such ancient accounts and motifs in order to comment upon the current situation of eastern Christians under Islamic rule,74 and this appears to be another case in point. As I argue elsewhere, his version of events sees a parallel between Troy and the fate of the eastern Christian community. The fate of both is undeserved, but both are nevertheless doomed: Troy to be destroyed by the Greeks, eastern Christendom to be destroyed by the expansion of Islam.75 One point of importance that already emerges from this review of the Trojan War tradition as it developed from classical antiquity 71

Posthomerica, X.277–329, 411–31, 459–89. See Conrad, “The Conquest of Arwàd,” 322–40; id., “The Arabs and the Colossus,” Journal of the Royal Asiatic Society, Third Series, 6 (1996), 166–75; Hoyland, Seeing Islam as Others Saw It: a Survey and Evaluation of Christian, Jewish and Zoroastrian Writings on Early Islam (Princeton: Darwin Press, 1997), 400–409, 631–71. 73 Chronicon ad annum Christi 1234 pertinens, ed. J.-B. Chabot (Paris: L. Durbecq, 1916–20; CSCO 81–82, Scr. syri 36–37), I, 66:8–78:24. 74 See Conrad, “The Arabs and the Colossus,” 181–87. 75 Conrad, Muhammadanea Edessensis: the Rise of Islam in Eastern Christian Historiography under the Early 'Abbàsids (Princeton: Darwin Press, forthcoming). 72

the

MAWÀLÌ

and early arabic historiography

389

in the Aegean basin through to early Islamic Syria is a pattern of continuity. The subject was one of enormous interest throughout antiquity, and apart from books such as those discussed above there are archaeological finds, papyri and endless references in other literary works to attest to this engagement. The Iliad in particular was a favored school text, and from second century CE Egypt there survives, for example, a schoolboy’s wooden writing board bearing a paraphrase and glossary of Il., IV.349–73.76 In general it can be said that the Graeco-Roman world never tired of talking about Homer:77 indeed, no less an authority than Plato engages in paraphrase, allusion and pastiche with Homeric passages.78 That something of this trend carried over into Islamic culture is well attested.79 While it of course cannot be said that interest in Homerica and the Trojan War in Islam in any way rivalled that of classical times and late antiquity, it is nevertheless clear that the literary motifs and techniques present in the written texts and oral discussions of this tradition would have been well-known and widely circulated among eastern Christians in early Islamic times.

76 Ashmolean Museum (Oxford), Greek Inscr. 3017. The text in question relates the near death of Paris at the hands of the wronged Menelaus on the battlefield before Troy. The reference to a child repeating Homeric hexameters in Aristophanes (Pax, 1269–87) should thus come as no surprise; the playwright’s apparently flippant attitude, by the way, strikes me as a comic convention and not a genuine critical attitude. 77 See Patzig, “Achills tragisches Schicksal bei Dictys und den Byzantinern,” Byzantinische Zeitschrift 25 (1925), 1–18, 273–91; Browning, “Homer in Byzantium,” Viator 6 (1975), 15–33; id., “The Byzantines and Homer,” in Robert Lamberton and John J. Keaney, eds., Homer’s Ancient Readers: the Hermeneutics of Greek Epic’s Earliest Exegetes (Princeton: Princeton University Press, 1992), 134–48; id., “A FourteenthCentury Prose Version of the Odyssey,” in Anthony Cutler and Simon Franklin, eds., Homo Byzantinus: Papers in Honor of Alexander Kazhdan (Washington, DC: Dumbarton Oaks, 1992; Dumbarton Oaks Papers 46), 27–36; Paulo Cesaretti, Allegoristi di Omero a Bisanzio: ricerche ermeneutiche (11–12 secolo) (Milan: Guerini, 1991), on three writers of the later Byzantine period; Sophia Georgacopoulou, “L’Achilléide de Stace et l’Achilléide anonyme byzantine,” Classica et mediaevalia 95 (1994), 251–86. 78 Lebarbe, L’Homère de Platon (Liège: Faculté de philosophie et lettres, 1949), 340–83. 79 Kraemer, “Arabische Homerverse,” Zeitschrift der Deutschen Morgenländischen Gesellschaft 106 (1956), 259–316; 107 (1957), 511–18; Strohmaier, “Homer in Bagdad,” Byzantinoslavica 41 (1980), 196–200; Khoury, “Die arabischen literarischen Übersetzungen aus dem Griechischen, unter besonderer Berüucksichtigung der Ilias von Homer,” in Piotr O. Scholz and Reinhard Stempel, eds., Nubia et Oriens Christianus. Festschrift für C. Detlef G. Müller zum 60. Geburtstag (Köln: Verlag Jürgen Dinter, 1987), 163–80.

390

lawrence i. conrad

One caveat must be recognised from the outset. What is of interest to us here are the motifs and paradigms that belong to the domain of literature, and not the recording of repetitive patterns that reflect the affairs of real life. There are, for instance, martial heroic motifs normally regarded by Arabists as typical of the early Arabic tradition that are also important in ancient Greek epic, but which reflect the genuine everyday concerns of the two societies. Horses were a prized possession among the Arabs, for example, and an ability to make off with the enemy’s animals is prized in Arabic tradition as a source not only of wealth but also of glory.80 In modern times this keen interest is still well attested.81 In the Iliad—and indeed, in folklore more generally82—one finds much the same. Diomedes comments that if only he could take the two horses that Zeus had given to Aeneas “we should win ourselves excellent glory.”83 The doomed Dolon tries to entice Odysseus to raid the camp of Priam’s Thracian allies, whose horses are the finest and largest he has ever seen, white as snow and fleet as the wind.84 The famous description of the shield of Achilles tells of an image in which shepherds are ambushed and killed and their livestock driven away.85 Achilles plunders the livestock of Aeneas on Mt. Ida,86 and boasts that the Trojans have never driven off his own cattle or horses.87 But these are motifs

80 Eisenstein, “Überlegungen zu einer Darstellung der Rolle des Pferdes in der arabisch-islamischen Kulturgeschichte,” in Arne A. Ambros and Markus Köhlbach, eds., Festschrift für Hans Hirsch zum 65. Geburtstag (Vienna: Institut für Orientalistik, 1996), 107–17; Macdonald, “Hunting, Fighting and Raiding: the Horse in PreIslamic Arabia,” in David Alexander, ed., Furùsiyya: the Horse in the Art of the Near East (Riyadh: King Abdulaziz Public Library, 1996), 73–83. 81 See, for example, Musil, Manners and Customs, 510–11, 539; Jabbur, The Bedouins and the Desert, 154–79, 348–55. 82 My attention was drawn to this point by a typically apt comment in Davies, The Epic Cycle, 42. 83 Il., V.273. 84 Ibid., X.433–37. 85 Ibid., XVIII.523–29. Cf. Mark W. Edwards in Kirk, gen. ed., Commentary, V, 220; Reinhardt, Die Ilias und ihr Dichter, ed. Uvo Hölscher (Göttingen: Vandenhoeck und Ruprecht, 1961), 401–11; Edwards, Homer: Poet of the Iliad (Baltimore/London: Johns Hopkins University Press, 1987), 278–85; Becker, The Shield of Achilles and the Poetics of Ekphrasis (London: Rowman and Littlefield, 1995), 114–24. 86 Il., XX.89–92. 87 Ibid., I.154–55. Cf. Eustathius (d. c. 1194), Eustathii commentarii ad Homeri Iliadem pertinentes, ed. Marchinus van der Valk (Leiden: E.J. Brill, 1971–87), 1203.8; AhlbergCornell, Myth and Epos in Early Greek Art: Representation and Interpretation ( Jonsered: Paul Åströms Förlag, 1992), 53; LIMC, I.1, 95–96 Achilleus nos. 389–90, an uncommon motif.

the

MAWÀLÌ

and early arabic historiography

391

of real life, not of literature, and they appear in literature as an expression of a continuing interest in them in the two societies in question. The fact that they appear in both traditions thus reflects no influence or pattern of borrowing, but rather reflects the continuing historical importance of the horse in the Near East. The ways in which the early Arab-Islamic tradition draws on late antique historical lore can, however, readily be detected in many other perspectives, literary devices and specific motifs that appear in both traditions. As these parallels accumulate, it becomes impossible to dismiss the similarities as mere coincidence or attribute them to general historical sensibility. In most cases, the examples cited could be multiplied considerably. To take the general perspectives first, we may refer to the issue of topoi.88 In Arab-Islamic studies this word will long be associated with the name of Albrecht Noth, who demonstrated that the battle scenes in the early Arabic futù˙ tradition largely consist of motifs that tradents could and did transfer from one account to another in order to provide details appropriate to description of a battle. The key to understanding Noth’s argument on topoi is the notion of transferability. A motif is not a topos because it is necessarily false everywhere it appears, but rather because it has broken loose from its original historical moorings—assuming that it ever had any such grounding—and has become a motif that any tradent could take up into his account to tell a good battle story.89 So to claim that something is topological is not to assert that it is false, but rather to ask how one is to separate the historical instances—if such there be— from the invented ones that the historical ones have inspired. If Noth is right, then the immediate result is the loss of huge swathes of the futù˙ tradition, at least as concerns narration of historical fact about the events in question (as opposed to later perceptions of these events).

88 Cf. Conrad, “Historical Evidence and the Archaeology of Early Islam,” in Samir Seikaly, Ramzi Baalbaki, and Peter Dodd, eds., Quest for Understanding: Arabic and Islamic Studies in Memory of Malcolm H. Kerr (Beirut: American University of Beirut, 1991), 264–68. 89 Noth, “Ißfa˙àn–Nihàwand. Eine quellenkritische Studie zur frühislamischen Historiographie,” Zeitschrift der Deutschen Morgenländischen Gesellschaft 118 (1968), 286–93; id., The Early Arabic Historical Tradition: a Source-Critical Study, 2nd ed. in collaboration with Lawrence I. Conrad, trans. Michael Bonner (Princeton: Darwin Press, 1994), 109–10, 144–45.

392

lawrence i. conrad

This prospect has been unnerving to some historians,90 but there can hardly be any doubt that Noth’s argument is correct.91 The topos is not, however, a peculiarly Arab device, but rather can be traced back through Near Eastern storytelling for more than a millennium. Already in 1930 Milman Parry showed that Homer’s use of repetitive phrases was unparalleled in classical literature. About 20% of the Iliad and Odyssey consist of repeated phrases, whole lines are repeated elsewhere, and there are about 26,000 repeated phrases in the 27,793 verses of the two epics.92 In 1968 Bernard Fenik published an enormously important study in which he showed that Homeric battle scenes are heavily topological.93 Thus battles and duels follow a limited number of specific paradigms and models and do not really represent different accounts. This is most obvious where duels between two opponents are concerned: A hurls a spear at B but misses, B throws at A but misses, A rushes at B and kills him. But it also occurs at a very general level. Thus the aristeia of both Diomedes in Book V of the Iliad and of Patroclus in Book XVI 90 See, for example, Khalidi, Arabic Historical Thought in the Classical Period (Cambridge: Cambridge University Press, 1994), 16 note 13, asking why Noth did not explain “why a concept originally developed by Erich Auerbach and E.R. Curtius for the aesthetic appraisal of medieval Latin literature is also relevant to the assessment of the facticity of early-Islamic historiography.” The implication of such an observation is of course that there is no such relevance, and its aim—evident elsewhere in the book—is polemical: in this way an opposing point of view can be dismissed as illegitimate without actually arguing against it. But that much said, the reply is firstly, and most obviously, that historical writing anywhere is a form of literature and is therefore open to discussion in terms of literary structures and conventions. One may also note that while the roles of Auerbach (Mimesis. Dargestellte Wirklichkeit in der abendländischen Literatur (Bern: A. Francke, 1946) and subsequent editions and translations) and Curtius (Europäische Literatur und lateinisches Mittelalters (Bern: A. Francke, 1948); also subsequent editions and translations) were seminal in promoting the understanding of topoi, they did not “originally develop” the concept. Quite apart from the classical tradition, where one already finds it in the pioneering scholarship of Milman Parry beginning with his L’épithète traditionnelle dans Homère: essai sur un problème de style homérique (Paris: Société d’éditions “Les Belles Lettres,” 1928), if not earlier, it also appears about the same time in European Orientalist circles with Werner Caskel’s study of the “battle days” lore of the Arab tribes; see his “Aijàm al-'arab,” 47–54. 91 See, for example, Robinson, “The Study of Islamic Historiography: a Progress Report,” Journal of the Royal Asiatic Society, Third Series, 7 (1997), 225–27. 92 Parry, “Studies in the Epic Technique of Oral Verse-Making,” Harvard Studies in Classical Philology 41 (1930), 90–117. 93 Fenik, Typical Battle Scenes in the Iliad: Studies in the Narrative Techniques of Homeric Battle Description (Wiesbaden: Franz Steiner Verlag, 1968). Fenik never uses the word “topoi,” but of course it is precisely these, along with what Noth called “schemata,” that he is discussing.

the

MAWÀLÌ

and early arabic historiography

393

begin during a battle and follow the same basic template: an initial kill clears the way for other heroes to take turns, a simile sums up, there is a rout, then each wreaks havoc until he encounters some obstacle.94 At one point an elaborate scene is set forth to describe a battle,95 but Fenik shows that this no sooner ends than it is immediately repeated on a much larger scale,96 with additional details and motifs being used to expand what is essentially the same story.97 The genius of Homer creates an exceptionally rich range of such materials, but they are also to be found in later texts of the tradition. The ps.-Dares, for example, tells his story of the Trojan War in a stiffly stereotyped fashion. His overarching schema is one of repeated negotiations and truces—eleven in all, lasting a total of seven years. This segmentation of the text serves as a means to move from one stage of the story to another; long truces are followed by stereotyped accounts of fighting and “great slaughter.”98 Numerous affinities between the two traditions in this area could be cited but one will serve to illustrate. The obvious Greek caricature of a duel between two warriors, with the contest reduced to a simple exchange of two blows followed by the kill, finds its Arab parallel in single-combat Arab battle scenes in the futù˙, where so many duels consist of one blow by each side expressed by a stock phrase (ikhtalafà ∂arbatayn), followed by the kill.99 Such cases are obviously topoi and entirely fantasy, not only in light of the stereotyped way in which they are expressed, but because there do survive accounts of what may well have been genuine duels. In one case, for example, we read of a great Persian champion who throws out a challenge to single combat. A mediocre Arab warrior hesitates to respond for fear of the mockery of his comrades, but finally steps forward because no one else will. In the duel he is quickly overpowered and thrown down by his rival, but just as the fatal blow is about to fall the Persian is pulled over by his horse, whose reins

94 Ibid., 9–10, 191–92. Cf. also the further discussion of the typology of the aristeia in Tilman Krischer, Formule Konventionen der homerischen Epik (Munich: C.H. Beck, 1971), 23–36. 95 Il., XIII.39–168. 96 Ibid., XIII.169–539. 97 Fenik, Typical Battle Scenes, 119. 98 Ps.-Dares, Historia, 16–38 (ed. Meister, 20:23–47:2). 99 E.g., al-ˇabarì, Ta"rìkh, I, 2024:13, 2101:6. Cf. also I, 2033:14–15 (killed with first blow), 2309ult-2310:10 (one blow each, both fatal).

394

lawrence i. conrad

have become tangled in his armour. The Arab is thus able to jump up, recover his sword, and kill the Persian.100 Another interesting parallel arises when we note that the Greek tradition on the Trojan War makes the abduction/elopement of Helen the cause of the conflict, as if the entire Aegean world of antiquity would have been engulfed by what amounted to a personal quarrel between Menelaus and Paris.101 This willingness to assign the causes of great events to discrete single incidents is well attested in the Arab-Islamic tradition, as in the explanation for the Arab invasion of Visigothic Spain, where, interestingly enough, we find a motif of a similar kind. The Roman governor Julian sends his daughter to be educated at the court of the Visigoths, but while there she is violated by the king. The furious Julian therefore encourages the Muslims under ˇàriq b. Ziyàd to invade Spain, since he can think of no more appropriate punishment for the king than to have the tribal Arabs set upon him.102 While there are differences in detail, the basic motif is identical: challenges to personal honour through violation of the sanctity of one’s womenfolk leads to the dispatch of a mighty host and large-scale conflict between the protagonists. Both traditions resolve anomalies in earlier stages of the development of an account by introducing explanations at some later point. What is happening here is that as society changes and develops, its perspectives on the past also change and it puts new or different questions to this past, and of course arrives at answers that meet its needs. We find similarities between the Greek and Arab-Islamic traditions in the general deployment of the technique, and with the same dramatic results as an old perspective is entirely turned on its head in favor of the opposite conclusion. In the classical Greek tradition, for example, and certainly in all of classical Homerica, Achilles is the archetypical Greek hero: no enemy can defeat or better him, and though he displays some of the usual failings of a great hero,

100

Sayf in al-ˇabarì, Ta"rìkh, I, 2323:9–2324:8. The case for this is secured by the explanation—unknown (or at least unattested) in Homer and the poets of the Epic Cycle—that before selecting Helen’s husband her father obliged all the suitor kings and princes to swear that should any one of the disappointed contestants abduct her, the others would unite to secure the father’s revenge. See Hesiod, Catalogi sive eoearum, lxviii.88–99; Isocrates, Helenea 39–41. Even Herodotus (I.4) finds no problem in accepting that the war was “due to a single Lacedaimonian woman.” 102 Ibn 'Abd al-Óakam (d. 257/871), Futù˙ Mißr wa-akhbàruhà, ed. Charles C. Torrey (New Haven: Yale University Press, 1922), 205:9–206:13. 101

the

MAWÀLÌ

and early arabic historiography

395

such as pride, anger and often poor judgment, no poet or playwright allows any stain of dishonour to touch him.103 But by the time of the ps.-Dictys the tale of the war had become embroiled in bitter quarrels over the merits and prerogatives of Greeks and non-Greeks. One of the many victims of this shift is the mighty Achilles, who becomes so infatuated and obsessed with Priam’s daughter Polyxena that he resolves to betray the entire Greek expedition to the Trojans in exchange for her hand. His friend Agamemnon and eventually almost all of the Greek warriors become suspicious of him as a traitor. In the end he is himself betrayed and murdered in a most unheroic way, with his arms pinned to his sides by one assailant while another runs him through with a sword. The Greek commanders lament the loss of such a hero, but the warriors do not grieve and refuse to perform the customary rites of mourning for a man whom they believe had sought to betray them; in their eyes, Achilles had met an ignominious or even dishonourable death.104 A similar problem often arises in the early Arab-Islamic tradition, where—even in materials collected in the same history, ˙adìth collection, or Qur"àn commentary—an event or point of discussion can be interpreted or explained in various irreconcilable ways. Quite often it is impossible to suggest how this can be sorted out, but in many cases plausible scenarios can at least be proposed. The caliph 'Umar b. al-Kha††àb (r. 13–23/634–44), for example, is said to have set out for Syria from Medina, but somewhere along the way he was warned that the plague was raging in Syria and so decided to turn back. This perfectly ordinary report shows an important leader taking a supremely sensible precaution and probably describes more or less what actually happened. But the account could not be allowed to stand. As a kerygmatic view of the early days of Islam became more prominent, 'Umar’s march became a journey in the cause of God, and thus a mission from which he should not have allowed himself to be diverted. There was also the problem that a tradition 103 This is especially clear in ancient iconography; cf. the recently completed Lexicon iconographicum mythologiae classicae (Zürich/Munich: Artemis, 1981–99), I.1, q.v. “Achilleus.” 104 Ps.-Dictys, Ephemeris, III.2–3, 27; IV.10–11, 13 (ed. Eisenhut, 61:5–62:26, 80pu–81:5, 88pu–90:17, 91:16–17). Cf. Patzig, “Achills tragisches Schicksal,” 1–18, 273–91; Milazzo, “Achille e Polissena in Ditti Cretese: un romanzo nel romanzo?,” Le forme e la storia 5 (1984), 3–24; King, “Achilles Amator,” Viator 16 (1985), 21–64, esp. 31–49; id., Achilles: Paradigms of the War Hero from Homer to the Middle Ages (Berkeley/Los Angeles: University of California Press, 1987), 184–201; Merkle, Ephemeris, 206–22.

396

lawrence i. conrad

of the Prophet had come into circulation saying that there was no such thing as contagion ('adwà),105 in which case 'Umar had no reason to flee: only those would die whose demise was willed by God, marking participation in a dispute among Muslims over divine foreordainment as opposed to human free will.106 So gradually the tradition was recast and elaborated into a much longer story associated with the renowned Medinan traditionist and compiler al-Zuhrì (d. 124/742). The fact that 'Umar had turned back could not be denied, but new details were added to justify his action, a tradition acknowledging contagion was thus turned into one denying it, and the worthies of the archaic age were made to express opinions on foreordainment that in fact belong to a much later time.107 We also find this sort of thing at the level of smaller details, and in ways where similar solutions are found. For example, the identification of Achilles as a great hero falls afoul of the fact that he did not rank among the suitors for the hand of Helen. If he was so great, how could this have been so? Hesiod resolves this difficulty: Achilles was too young at the time, otherwise he would have sought Helen for himself and certainly would have prevailed.108 In early Islamic tradition a similar problem arises over 'Abd Allàh b. 'Umar: if he was such a great Companion of the Prophet, why did he not participate in the battles of Badr or U˙ud? The answer is that he did volunteer, but he was still sexually immature at the time and was deemed too young to fight.109 105

On this see Conrad, “A Ninth-Century Muslim Scholar’s Discussion of Contagion,” in Lawrence I. Conrad and Dominik Wujastyk, eds., Contagion: Perspectives from Pre-Modern Societies (Aldershot: Ashgate, 2000), 163–77. 106 On the early Islamic doctrine and related disputes about qadar, or divine foreordainment, see Cook, Early Muslim Dogma: a Source-Critical Study (Cambridge: Cambridge University Press, 1981), 105–52; van Ess, Anfänge muslimischer Theologie. Zwei antiqadaritische Traktate aus dem ersten Jahrhundert der Hi[ra (Wiesbaden: Franz Steiner Verlag, 1977); id., Theologie und Gesellschaft im 2. und 3. Jahrhundert Hidschra. Eine Geschichte des religiösen Denkens im frühen Islam (Berlin: Walter de Gruyter, 1991–95), I, 24–25, 72–135, 229–33; II, 50–121, 202–206, 643–55, 668–700. 107 Conrad, “'Umar at Sargh: the Evolution of an Umayyad Tradition on Flight from the Plague,” in Stefan Leder, ed., Story-Telling in the Framework of Non-Fictional Arabic Literature (Wiesbaden: Otto Harrassowitz Verlag, 1998), 489–528. 108 Catalogi sive eoearum, LXVIII.100–106. The ps.-Dictys, Ephemeris, I.14 (ed. Eisenhut, 14:2), seems to pick up on this point. 109 Al-Bukhàrì, Al-Jàmi' al-ßa˙ì˙, ed. Ludolf Krehl and T.W. Juynboll (Leiden: E.J. Brill, 1862–1908), III, 55:3–6, 93:10–12 Maghàzì, nos. 6, 29. Similar reports are very common.

the

MAWÀLÌ

and early arabic historiography

397

We may also note that both traditions make use of the motif of the eyewitness as a means of asserting the truth-value of an account. Both the ps-Dictys and the ps.-Dares trawl through the Iliad and discover obscure personalities who are then turned into “authors” providing first-hand accounts of the war. They are both late arrivals in terms of the development of the tradition, we may note, and clearly the point is to overthrow the superior intellectual or literary authority of a great figure of the past by contradicting him with the testimony of someone who, for all his inferior credentials on other fronts, enjoys the decisive advantage of reporting what he has actually witnessed and heard on the scene. That is, the appeal to the authority of the eyewitness is as much as to admit the lateness of the account in question, since it tacitly concedes that there already exists an earlier and highly authoritative stage of discussion against which there is now a need to deploy a counterargument buttressed by the weighty authority of first-hand knowledge. The same occurs in early Islamic historiography and ˙adìth, where chains of authorities extending right back to the event in question are often late developments. Historical accounts related on first-hand authority in the first person are in most cases obvious fabrications. The point is almost always to argue that while there may be conflicting versions of a story in circulation, or while there may be weighty arguments against the account in question, this one given on the authority of an eyewitness ought to be accepted because it is reported by one who actually saw what had happened. That is, and precisely as we find in the late antique tradition, the argument presupposes either 1) the existence of a body of earlier material whose considerable authority must be overthrown in favor of this latest contribution to the discussion, or 2) sharp dispute over the accuracy of the “historical facts” in question. For example, in early Islamic accounts of the alleged letter of Mu˙ammad to Heraclius calling upon the emperor to convert to Islam, the Byzantine ruler interviews Abù Sufyàn in order to gain information about the Prophet and his spiritual message. The story is pro-Umayyad, since it places Abù Sufyàn, the father of Mu'àwiya, the first Umayyad caliph (r. 41–60/661–80), in a favorable light as a great tribal notable of sterling descent and makes much of his knowledge of and closeness to the Prophet. In all likelihood it was sharply disputed by anti-Umayyad elements who preferred to stress that the progenitor of the early Umayyad house

398

lawrence i. conrad

had been an early opponent of Mu˙ammad.110 It thus does not surprise us to see accounts of Abù Sufyàn with Heraclius moving from third-person narratives in which his story is told, to versions in which he speaks in the first person, and even more strikingly, to first-person versions in which the transmitter of the tale assures us that he has the account straight from the mouth of Abù Sufyàn and that we have it straight from his.111 In some cases the lateness of the eyewitness account is obvious: in the materials on 'Umar’s journey to Syria discussed above, for example, an allegedly eyewitness account is confected by simply appropriating a report presented in the third person, rewriting it in the first, and putting it in the mouth of an eminent Companion of the Prophet as an eyewitness. But when we see that the recasting has neglected to change one of the verbs from third to first person, what is really going on is instantly betrayed to us.112 In Homer lists of names and exercises in name-giving are very often crucial to the task of carrying the story forward. In the Iliad,113 for example, Zeus is overcome with desire for his wife Hera and seeks to tempt her back into his bed by assuring her that her charms are far superior to those of his other conquests, which, while Hera smoulders with fury, he obliviously proceeds to retail in a list that includes sublists of the children born by these women.114 Elsewhere, as in lists of those slain by a particular warrior, Homer’s enumeration of names comprises a pause prolonging the suspense or allowing the auditor to reflect on a particular situation. But in Homer one can also detect a trend toward indulgence in lavish onomatomania—specification of names primarily for its own sake and to display knowledge of (often pointless) detail.115 Euryalos, for example, pursues and kills the Trojans Aisepus and Pedasus.116 Who are these 110

See EI 2, I, 151a–b. See, for example, A˙mad b. Óanbal (d. 242/855), Musnad (Cairo: al-Ma†ba'a al-Maymuniyya, 1313 AH), I, 262:2–263:15; al-Bukhàrì, Ía˙ì˙, I, 7:1–8:13 Bad" alwa˙y, no. 6; II, 232ult–235:11 Jihàd, no. 102; III, 214:4–216:8 Tafsìr/Àl 'Imràn (3), no. 4; Muslim b. al-Óajjàj (d. 261/875), Ía˙ì˙ (Cairo: al-Ma†ba'a al-Maymuniyya, 1290 AH), Ía˙ì˙, II, 59:12–60:21 Jihàd, no. 73; al-ˇabarì, Ta"rìkh, I, 1561:10–1565:13; Aghànì, VI, 345:8–349:14. 112 Conrad, “'Umar at Sargh,” 512–13. 113 Il., XIV.312–28. Cf. also the discussion in Eustathius, 988.41–59. 114 Cf. Bowra, Tradition and Design in the Iliad, 2nd ed. (Oxford: Clarendon Press, 1950), 223; Beye, Ancient Epic Poetry, 95–98; Edwards, Homer, 247–50; Janko, Commentary, IV, 201–205. This tactic is discussed in connection with another episode in E.T. Owen, The Story of the Iliad (London: Bristol Classical Press, 1989), 52–53. 115 Bowra, Tradition and Design, 69–81; Edwards, Homer, 120–22. 116 Il., VI.20–28. 111

the

MAWÀLÌ

and early arabic historiography

399

warriors? Elsewhere Aisepus is the name of a river at the foot of Mt. Ida,117 and Pedasus is a town on a certain river,118 or a town somewhere else entirely,119 or one of the horses of Achilles.120 Warriors who fall victim to the military prowess of great heroes are especially likely to be nothing more than random names invented from terms appropriate to the heroic archetype. Thus Ajax kills Pandocus (“the welcomer of everyone”),121 and the leader of the Trojan contingent that seeks to torch the Achaean ships is named Pyraichmes (“armed with fire”).122 In the famous story of the wooden horse in Troy, Helen circles it three times and calls out, imitating the voices of the wives of the Achaean warriors within; the one who comes closest to responding and thus exposing the Achaean trick is a man named Anticlus (“he who responds”), an individual nowhere named in the Iliad.123 Often one finds that the names are quite transferable and that their owners possess remarkable powers of revival. A certain Greek killed by Hector and Ares,124 for example, pops up subsequently as a companion of a Trojan commander125 and is finally killed again—this time as a Trojan.126 The obscure Lycian warrior Halius killed by Odysseus in the Iliad127 is the son of the Phaiacian king in the Odyssey.128 The Trojan warrior Melanippus is killed three times;129 each time the name occurs at the end of the verse, showing that we have to do here not with three different men, but with a random name chosen to fill out the end of the line in a way that will suit the metrical requirements of Greek hexameter.130 Warriors who come forward to accept one challenge are the same ones who volunteer for something else, often in the same or nearly the same order.131 At one point Homer comes close to conceding that one would have to be a god to possess the knowledge required for production of a genuine 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131

Ibid., II.824–25. Ibid., VI.34–35. Ibid., IX.152–53. Ibid., XVI.152–54. Ibid., XI.490. Ibid., XVI.287. Od., IV.274–89. Il., V.703–705. Ibid., XII.139. Ibid., XII.187–94. Ibid., V.674–78. Od., VIII.118–19. Il., VIII.276; XV.576; XVI.695. Edwards, Commentary, V, 264. Il., VII.161–68; VIII.262–65; XXIII.288–90.

400

lawrence i. conrad

list of those involved in various battles.132 Clearly the ancient aoidoi had to hand an onomastic pool into which they could dip for random names of insignificant people and places, selected according to the metrical requirements of the line for which they were needed.133 Later texts of Homerica exult in this sort of thing, a prime example being ps.-Dares, who “knows” and provides so many names that his text at times degenerates into a series of lists—categories of all kinds filled with names. He gives the reader, for example, the names of all the children of Priam, the gates of Troy, the sons and friends of Priam whom he summoned to a certain council of war (but excluding, we are told, all the sons born to him by his concubines), the advisors sent with Paris on his embassy to Sparta, all the places that allied themselves with Troy (including the names of the leaders who personally came to Priam’s aid),134 and all those killed by a particular warrior during a particular battle.135 In the fable book of Hyginus, we may note, some chapters contain no continuous text but rather only lists: 36 suitors of Helen,136 the supposed 55 sons and daughters of Priam,137 and specification of who fought whom, those who killed princes, Achaeans who killed Trojans (with sublists of the victims), and Trojans who killed Achaeans.138 This same love of listing and name-giving, which evoked considerable enthusiasm among Greek and Latin writers,139 is likewise very 132 Ibid., XII.176. Interestingly, this is to my knowledge the only place in the Iliad where Homer refers to himself in the first person singular. Cf. Hainsworth in Kirk, gen. ed., Commentary, III, 337. 133 Cf. Edwards, Homer, 81; Kirk, Commentary, II, 126–27, 130–31, 158–59. 134 The Epic Cycle may also have contained something like this list of allies. See the summary comment in Cypria apud Proclus, Chrestomathia (ed. Allen, 105:17–18); also ps.-Dictys, Ephemeris, II.35 (ed. Eisenhut, 47:4–48:6), where these allies are listed in the order in which they rushed out the gates for battle. 135 E.g., ps.-Dares, Historia, 4, 6, 9, 18, 21 (ed. Meister, 6:12–16, 8:3–6, 11:8–16, 22:12–23:13, 26:17–22). Cf. also Schissel von Fleschenberg, Dares-Studien, 115–28. 136 Oddly enough, this is far fewer than the 108 (plus retainers) already specified in Od., XVI.245–53. Yet another list (31 suitors) is given in Apollodorus, Bibliotheca, III.x.8–9. 137 This list is inspired by Il., XXIV.495, where Priam states: “Fifty were my sons, when the sons of the Achaeans came here.” Cf. also Il., VI.242–50, for the 50 sleeping chambers set aside for the sons of Priam in his palace and the twelve for his wedded daughters. One notes again, of course, the prominent role of the topological number 50. Cf. Germain, Homère et la mystique des nombres (Paris: Presses universitaires de France, 1954), 19; Nicholas Richardson in Kirk, gen. ed., Commentary, VI, 325–26. The number “twelve” is also a motif, and will be discussed below. 138 Hyginus, Fabulae, lxxxi, xc, cxii–cxv. 139 Cf. Matz, Ancient World Lists and Numbers: Numerical Phrases and Rosters in the Greco-Roman Civilizations (London: McFarland, 1995).

the

MAWÀLÌ

and early arabic historiography

401

prominent in the early Arab-Islamic historical tradition, where one encounters similarly extravagant lists and gratuitous specification of names for persons, places, horses, donkeys, swords, elephants, catapults, dogs, and so forth.140 Mu˙ammad b. Óabìb (d. 245/860) is notorious for a work that consists of little more than lists of names,141 early works on horses tend to be lists of famous horses and their pedigrees, organised by name,142 and a book with a title like Kitàb alsilà˙, “Book of Weapons,” is likely to be a work listing the nicknames various warriors had given to their swords, lances, daggers, etc.143 As in late antiquity, then, the exercise of name-giving and listing in early Islamic culture is often its own justification. But the phenomenon also illustrates the frequently fine line between life and literature. The Arabs clearly gave names to animals and implements of use to them; these names appear in poetry, and it was in order to elucidate this verse that works on these names were compiled. But the joy with which naming and listing were pursued in Arabic literature goes far beyond this, and in this sense they continue a pattern that had already been typical of Near Eastern and Mediterranean literature for centuries. There are also many places where one can detect parallels at a more specific level. Some of these have already been discussed above in connection with larger paradigms, and here a few more examples can be cited. Elsewhere I have studied the use of the number four and its multiples by ten (especially 40) in the Arabic literary tradition and have come to the conclusion that these figures usually do not represent actual numeration, but rather express general notions of magnitude, often with some specific symbolic message in mind.144 More than a millennium earlier this sort of usage can already be seen, for example,

140 Cf. the discussions in Noth/Conrad, Early Arabic Historical Tradition, 96–104, 126–29. 141 Mu˙ammad b. Óabìb (d. 245/860), Kitàb al-mu˙abbar, ed. Ilse Lichtenstädter (Hyderabad: Dà"irat al-Ma'àrif al-'Uthmàniyya, 1361/1942). 142 See Levi della Vida, ed., Les “livres des chevaux” de Hi“àm ibn al-Kalbì et Mu˙ammad ibn al-A'ràbì (Leiden: E.J. Brill, 1928). 143 Unfortunately I have not been able to identify this text, as this paper was written without recourse to my personal library. 144 Conrad, “Abraha and Mu˙ammad: Some Observations Apropos of Chronology and Literary Topoi in the Early Arabic Historical Tradition,” Bulletin of the School of Oriental and African Studies 50 (1987), 225–40; further in id., “The Conquest of Arwàd,” 354–58, responding to some important reservations put to me by Michael Lecker.

402

lawrence i. conrad

in the Iliadic “Catalogue of Ships,”145 where “40,” frequently given as the number of warships that came to fight for the Achaeans from this or that location, is a round number conveying a vague sense of magnitude. Of the 29 Achaean contingents described in the “Catalogue,” in fact, 22 are counted in terms of “round” numbers, making use of most of the multiples of ten up to 100. There are other numbers that likewise appear simply to be formulaic, with no specific symbolism behind them beyond the image of a full complement of something. Thus, for example, one frequently finds Homer resorting to the motif of “twelve”:146 e.g., twelve heifers are sacrificed to Athena;147 Agamemnon offers Achilles twelve horses;148 Diomedes kills twelve Thracians;149 Neleus has twelve sons;150 twelve Trojans are speared by Ajax, one after another, as they rush forward to set fire to the Achaean ships;151 twelve Trojans fall dead in their chariots, shocked by the war cry of Achilles;152 twelve Trojans are captured and killed by Achilles and laid on the funeral pyre of Patroclus;153 the North Wind assumes the form of a stallion and mates with mares that bear him twelve colts;154 a great tripod offered as a prize in the funeral games for Patroclus is worth twelve oxen.155 “Twelve” is also important in the Arab-Islamic tradition, to which it passed from antiquity through a specifically monotheistic prism. In the Old Testament one finds twelve tribes of Israel, and when the Levites are excluded Joseph’s sons Ephraim and Manassah become “tribes” in order to keep the twelve complete.156 In the New Testament 145 Il., II.484–759. On this famous passage, see Allen, The Homeric Catalogue of Ships (Oxford: Clarendon Press, 1921); Kullmann, Die Quellen der Ilias, 63–118, 157–68; Giovannini, Etude historique sur les origines du Catalogue des vaisseaux (Bern: Francke, 1969); Simpson and Lazenby, The Catalogue of the Ships in Homer’s Iliad (Oxford: Clarendon Press, 1970); Kirk, Commentary, I, 168–240; Visser, Homers Katalog der Schiffe (Stuttgart: B.G. Teubner, 1997). 146 On “twelve,” see also Waltz, “L’exagération numérique dans l’Iliade et l’Odyssée,” Revue des études homeriques 3 (1933), 22, 25–27, 36; Schultze, “Die Zahlensysteme der homerischen Griechen,” Archeion 19 (1937), 186–87; Germain, Homère et la mystique des nombres, 17–18, 35–36, 47, 54–56. 147 Il., VI.92–93. 148 Ibid., IX.123. 149 Ibid., X.485–88. 150 Ibid., XI.691. 151 Ibid., XV.742–46. 152 Ibid., XVIII.230–31. 153 Ibid., XVIII.336–37; XXI.26–28; XXIII.175–76. 154 Ibid., XX.222–25. 155 Ibid., XXIII.702–703. 156 Genesis 35:22–26; Matthew 14:20, 26:53.

the

MAWÀLÌ

and early arabic historiography

403

Jesus has twelve disciples, and after the suicide of Judas the remaining eleven choose Matthius as a replacement, again maintaining the twelve.157 In the Qur"àn the Biblical connection is already evident in the account of how Moses strikes twelve springs out of the rock,158 which seems to be taken up from the Old Testament reference to the twelve springs of Elim.159 In subsequent Islamic tradition we find that Mu˙ammad used to call out “God is great” twelve times on feast days.160 At the First Pledge of al-'Aqaba twelve accept Islam,161 and at the Second Pledge of al-'Aqaba Mu˙ammad requests the appointment of twelve delegates (nuqabà");162 likewise, the 'Abbàsid activists in southern Syria appoint twelve nuqabà" to attend to their affairs in Merv and elsewhere in Khuràsàn.163 In such cases we can see an imitation of the Christian notion of a complete complement of twelve apostles.164 An Arab commander whose religious sincerity is doubted redeems himself by killing twelve Persians.165 In ˙adìth all this reverberates in traditions about the first twelve caliphs, and among the Shì'a it of course appears in the doctrine of twelve Imàms.166 The number 12,000 is enormously important in apocalyptic, where it is used to describe the size of a wide variety of groups involved in apocalyptic events.167 In connection with this a tradition emerged to assert that a Muslim army of 12,000 would not be defeated by an enemy,168 and this has an impact on subsequent futù˙-accounts (though not as pronounced as one might expect). 157

Acts 1:12–26. Sùrat al-Baqara (2), v. 60. 159 Exodus 15:27. 160 Ibn Óanbal, Musnad, II, 180:18. 161 Ibn Hishàm, I.1, 288:3–289:13. 162 Ibid., I.1, 297:4–299:10. 163 Al-ˇabarì, Ta"rìkh, II, 1358:9–10; Akhbàr al-dawla al-'abbàsiyya, ed. 'Abd al'Azìz al-Dùrì and 'Abd al-Jabbàr al-Mu††alibì (Beirut: Dàr al-Jìl, 1391/1971), 216:1. 164 Islamic tradition held that every prophet has his apostles (˙awàrìyùn) and companions (ßa˙àba). See, for example, al-Bukhàrì, Ía˙ì˙, II, 212:13–19, 247:1–5; Jihàd, nos. 40–41, 135; Ibn Óanbal, Musnad, I, 89:17–23, 102:7–9, 103:29–31, 458:9–13, 461:29–462:2. When the 'Abbàsids claim that their early activists had a similar number of representatives that is as much as to claim religious sanction for their mission. 165 Sayf in al-ˇabarì, Ta"rìkh, I, 2264:17–18. 166 See Kohlberg, “From Imàmiyya to Ithnà'ashariyya,” Bulletin of the School of Oriental and African Studies 39 (1976), 521–34; Rubin, “Apocalypse and Authority in Islamic Tradition: the Emergence of the Twelve Leaders,” Al-Qantara 18 (1997), 11–41. 167 See David Cook, Studies in Islamic Apocalyptic (Princeton: Darwin Press, 2002), index. 168 Discussion and references in Conrad, “The Conquest of Arwàd,” 355. 158

404

lawrence i. conrad

For example, various figures are given for the number of Muslims setting out on the campaign that will culminate in the battle of alQàdisiyya; to a total of 9000 enumerated from various tribes, another 3000 from Asad are added to reach a total of 12,000.169 Where the enemy is numbered at 120,000 (i.e., ten times a full Muslim army) the message is that no matter how numerous the foe is he will still lose.170 As these examples illustrate, there is a message here beyond the element of mere exaggeration;171 in the monotheistic traditions the creation, maintenance or alteration of a “twelve” invited the audience to consider the hand of God at work in the events in question.172 In the Iliad both the Achaeans and the Trojans frequently use a exhorting speech to urge their men to fight on bravely. No one would doubt that here we have to do with a literary motif. Apart from the fact that the epic is obviously a work of fiction, the speeches pursue a limited number of themes that Homer can shift from one scene to the next as suits his purposes for that passage, its metrical requirements, and how long Homer wishes to tarry at this point in the text.173 In the early Arabic futù˙ tradition a variety of speeches are quoted to depict religious convictions or argue for the spiritual superiority of Islam. There may of course be some slender historical basis for such accounts; generals usually recognise the importance of morale in the ranks, and the commanders of the early Arab armies are certainly likely to have encouraged their men in some such way as they closed for battle. But the speeches quoted in extant texts cannot have been the original ones; the way in which they are built using the same stockpile of standard motifs, transferable from one speech or account to another as required, shows that they have broken loose from such “original” moorings as they may once have had and now function—by definition—as topoi.174 Supernatural powers are routinely attributed to great figures in both traditions, and often powers of a very similar kind. In the Epic

169

Sayf in al-ˇabarì, Ta"rìkh, I, 2221:4–2222:15. Cf. also I, 2279:4–5. Ibid., I, 2250:14–15, 2258:6, 2264:1–2, 2294:9–10. 171 Cf. Wellhausen, “Prolegomena zur ältesten Geschichte des Islams,” in his Skizzen und Vorarbeiten (Berlin: Georg Reimer, 1884–99), VI, 71, 77. 172 Where the enemy is concerned, however, 12,000 seems simply to mean “a large force.” See Sayf in al-ˇabarì, Ta"rìkh, I, 2347:8–10, where the Roman army before al-Yarmùk includes 12,000 Armenians and 12,000 Arab confederates. 173 See Latacz, Kampfparänese, Kampfdarstellung und Kampfwirklichkeit in der Ilias, bei Kallinos und Tyrtaios (Munich: C.H. Beck, 1977), 246–50. 174 See Noth/Conrad, Early Arabic Historical Tradition, 87–96. 170

the

MAWÀLÌ

and early arabic historiography

405

Cycle a specially gifted individual can from a mountaintop see across the whole Peloponnesus and detect his foes.175 Similarly, in the ArabIslamic tradition Mu˙ammad is able to follow the course of the battle of Mu"ta in Syria when the ground is raised for him in Medina—i.e., hundreds of kilometres away.176 In the sìra tradition, in fact, Mu˙ammad is endowed with a wide range of miraculous gifts and abilities, and years ago Horovitz was already able to demonstrate that such materials were especially popular among the immediate ancestors of Ibn Is˙àq, whose Christian background would have disposed them favorably toward a Prophet figure with the same miraculous powers attributed to Jesus in the Gospels.177 The military climax of the Iliad comes on the 25th day of the period covered by the text; over more than seven books of his epic Homer describes the tremendous battle that raged through this fateful day.178 And the poet leaves us in no doubt that this is the day when all will be decided. Zeus, for example, warns Hera of the coming duel between Achilles and Hector when what is to be will be,179 and Hector warns the assembled Trojans that the morrow will reveal whether he or Achilles will prevail.180 The Trojans in their vast camp on the plain await the fateful break of dawn,181 while in the Achaean camp Odysseus urges Achilles to set aside his anger and rejoin them, lest disaster overwhelm them.182 The day itself opens with Zeus sending down “the wearisome goddess of Hate, holding in her hands the portent of battle.”183 The same sort of motif also appears in the early Arabic historical tradition, in which a variety of ways are found to express the idea that a certain battle will be decisive and decide the further course of history. That this emphasis represents a secondary accretion and a topos of literary tradition is clear from the fact that so many battles—in the Arab-Islamic tradition at least eight can be counted—vie for such honours in different texts.184 175

Cypria, fr. xi (ed. Allen, 122:5–11). Ibn Sa'd (d. 230/844), Kitàb al-†abaqàt al-kabìr, ed. Eduard Sachau et al. (Leiden: E.J. Brill, 1904–40), II.1, 93:24–26. 177 See Horovitz, “Zur Mu˙ammadlegende,” Der Islam 5 (1914), 41–53. Cf. also the important account in Tor Andrae, Die Person Muhammeds in Lehre and Glauben seiner Gemeinde (Stockholm: P.A. Norstedt, 1918), 26–91. 178 Il., XI.1–XVIII.242. 179 Ibid., VIII.470–83. 180 Ibid., VIII.529–41. 181 Ibid., VIII.553–65. 182 Ibid., IX.225–51. 183 Ibid., XI.2–4. 184 Noth/Conrad, Early Arabic Historical Tradition, 129–32. 176

406

lawrence i. conrad

In a pattern closely related to that of onomatomania one finds that in the Graeco-Latin tradition of Homerica a certain text or group of texts will bear details that are entirely unknown to earlier writers or compilers. Thus, for example, ps.-Dares “knows” the physical appearance and aspects of personal character of the Greek and Trojan heroes in excruciating detail, down to the birthmark between Helen’s eyebrows and Hector’s slight lisp,185 though earlier in the tradition this sort of thing appears to be unknown. He can also tell us, based on the figures he had previously recorded in his “diary” (acta diurna),186 that the war lasted ten years, six months and twelve days, that 866,000 Greeks and 676,000 Trojans fell in the fighting, that Aeneas escaped with 22 ships and 3400 Trojan followers, and that another 2500 escaped with Antenor and 1200 with Andromache.187 That is, a work apparently becomes “better” for its ability to relate details that earlier authors or compilers had “missed.” A similar feature can readily be seen in the early Arabic historical tradition, and again with precise parallels. After he relates the death of the caliph Abù Bakr in 13/634, al-ˇabarì gives an account of the caliph’s appearance: slender, fair-skinned, thin beard, stooped, with a gaunt face and sunken eyes, protruding forehead, and slender legs.188 In other words, the first caliph appears to have no use for the pleasures of this world. Al-ˇabarì’s list of authorities for this account, which is matched by many others in his history and in works by other scholars describing many personalities of the early days of Islam, includes the compiler, jurist, and traditionist al-Wàqidì (d. 207/823). In his narrations one routinely finds very rich details 185 Ps.-Dares, Historia, 12–13 (ed. Meister, 14:9–17:10). On the background to this tradition, see Evans, “Descriptions of Personal Appearance in Roman History and Biography,” Harvard Studies in Classical Philology 46 (1935), 43–84. Similar portraits were taken up by Byzantine historians from the original Greek text of the ps.-Dictys; see Malalas, Chronographia, V.1, 11, 12, 13–40 (trans. Elizabeth Jeffreys, Michael Jeffreys, and Roger Scott, The Chronicle of John Malalas (Melbourne: Australian Association for Byzantine Studies, 1986), 45, 50, 51, 51–56); Schissel von Fleschenberg, Dares-Studien, 9–84; Griffin, Dares and Dictys, 29, 33. In the extant Latin version, however, these no longer survive; cf. Luciano de Biasi, “I ritratti dei personaggi in Darete Frigio: raffronto con i testi fisiognominici,” Koinonia 3 (1979), 53–112; Beschorner, Untersuchungen, 108–27. 186 On the significance of this term, see Gianotti, “Le metamorfosi di Omero. Il romanze di Troia della specializzazione delle scholae ad un publico di non specialisti,” Sigma 12 (1979), 23; Schetter, “Dares und Dracontius,” 107–109; Beschorner, Untersuchungen, 191–92. 187 Ps.-Dares, Historia, 44 (ed. Meister, 52:3–14). 188 Al-ˇabarì, Ta"rìkh, I, 2132:11–2133:4.

the

MAWÀLÌ

and early arabic historiography

407

of the most precise sort, and it is significant that it so often proves that his predecessors were clearly ignorant of these details. This problem arises everywhere in his work: in his accounts of pre-Islamic Arabia,189 of the career of Mu˙ammad,190 and of the conquests.191 Here too, to produce something with details given by no other author makes for a “better” book. There is a valuable cluster of cases in which the central motif in both traditions is very similar, but put to different uses. Three of these are worthy of special attention here by way of illustration. In the Arabic tradition one often finds the motif of the bedouin, a simple rather uncouth individual portrayed as a guileless rustic who is befuddled by a simple problem, fails to see a basic point, or naively behaves in a totally unacceptable way, for example urinating in the mosque. This motif is especially common in ˙adìth, where the prominent religious figure in the story redeems the situation or corrects the bedouin, thereby providing a didactic lesson for the broader benefit of the public circles who will be exposed to this material.192 In the texts of Homerica the guileless rustic is also in evidence, if not for such purposes. The Ephemeris of the ps.-Dictys introduces itself with the claim that the lead box containing Dictys’ Phoenician-language diary was discovered in the ruins of his tomb by some shepherds, who naively retrieved it and handed it over to the authorities. The aim here is clearly to assert the utterly innocent circumstances of the text’s sudden emergence from obscurity: illiterate shepherds could not have forged the work or even read it.193 A similar argument appears at the beginning of the Posthomerica of Quintus, who claims that while tending his sheep on a mountainside near Smyrna he was inspired by the Muses to compose his poem. That is, this work seeks to gain public assent to its version of the Trojan War on the grounds that its author was a simple herdsman without a partisan or polemical agenda to pursue.194 189 Useful assessments of this phenomenon may be found in Patricia Crone, Meccan Trade and the Rise of Islam (Oxford: Blackwell’s, 1987), 223–30. 190 Cook, Muhammad (Oxford: Oxford University Press, 1983), 63–67; LandauTasseron, “Processes of Redaction: the Case of the Tamìmite Delegation to the Prophet Mu˙ammad,” Bulletin of the School of Oriental and African Studies 49 (1986), 253–70. 191 Conrad, “The Conquest of Arwàd,” 373–84. 192 On this motif, see Stetter, Topoi und Schemata im ÓadìΔ (Tübingen: Orientalisches Seminar der Universität Tübingen, 1965), 28–30. 193 Above, note 50. 194 Above, note 69.

408

lawrence i. conrad

The fact is so well-known that it may now simply be stated that the early Arabic historical tradition displays a profound fascination with awà"il, or reports of famous “firsts.”195 These accounts are usually very brief, or even parenthetical, and pursue often trivial issues for which no answer could possibly have been known: who was the first person to say “hello” in Arabic or to light a fire on Muzdalifa, who was the first Arab woman to fall in love with another woman?196 The same interest in “firsts” likewise appears in the Greek tradition. Achilles’ mother Thetis commands her son not to be the first Achaean to alight on Trojan soil, since this warrior is doomed to die.197 The first to do so is of course identified and not surprisingly is named Protesilaus, “the first to leap [ashore]”; equally unsurprising is the fact that he is immediately killed by Hector.198 We are also told that the first Achaean to descend from the wooden horse in Troy was a man named Echion, who jumped and was killed in the fall.199 Homer further tells us that Antilochus was the first to kill an important Trojan200 and asks Patroclus in the second person whom it was that he killed first.201 The difference between these and the Arabic cases is that several of the Greek examples are heavily symbolic, in that they advance the motif of an initiator (however obscure) whose sacrifice of his own life ensures the success of an enterprise.202 The subtle veiling of the point thus reminds the audience to pay attention to the details, so as to catch clues as to what will happen next. The shift to a more superficial deployment of the motif suggests that in many cases the deeper significance of ancient forms was no longer understood. But the opposite could also occur. As we have seen, the number “twelve” appears to have been a general motif for a “full” 195 See Juynboll, Muslim Tradition: Studies in Chronology, Provenance and Authorship of Early Óadìth (Cambridge: Cambridge University Press, 1983), 10–23, for the ˙adìth literature; and Noth/Conrad, Early Arabic Historical Tradition, 104–108, on the historical tradition. 196 Valuable collections of this material are al-Nabìl (d. 287/900), Kitàb al-awà"il, ed. Ma˙mùd Mu˙ammad Ma˙mùd Óasan Naßßàr (Beirut: Dàr al-Jìl, 1991); Abù Hilàl al-'Askarì (d. 395/1005), Kitàb al-awà"il, ed. Mu˙ammad al-Mißrì and Walìd Qaßßàb (Damascus: Wizàrat al-Thaqàfa wa-l-Irshàd al-Qawmì, 1975). 197 Apollodorus, Epitome, III.29–30; scholion to Lycophron, Alexandra 246, ed. Eduard Scheer (Berlin: Weidmann, 1881–1908), II, 110:19–32. 198 Cypria apud Proclus, Chrestomathia (ed. Allen, 105:1); fr. 17 (ed. Allen, 123:19–23). 199 Apollodorus, Epitome, V.20. 200 Il., IV.457. 201 Ibid., XVI.692. 202 Cf. Davies, The Epic Cycle, 47.

the

MAWÀLÌ

and early arabic historiography

409

complement of something, but in the monotheistic dispensations of Judaism, Christianity and Islam took on a more complex meaning. Finally, the trend we see in Homerica for playful tricks with names that invite the reader/auditor to pay close attention is evident in the Arabic futù˙ tradition, but often not in the same way. One especially obvious example of this occurs when names have to be found for the commanders of the left and right flanks of Persian armies. Sometimes the creator of the account, practically inviting his audience to catch him out, simply takes one skeletal form for a “Persian” name and points and vowels it in different ways to produce a pair of absolutely invented commanders: Kharukbadh and Karukbadh in one case,203 Tìrùyah and Bandùyah in another.204 My purpose in pointing to these similarities is not to assert that in early Islamic times Muslim tradents were generally familiar with Homer, or with the Greek, Latin and Syriac literature of subsequent times on the Trojan War. Indeed, the cases of use of the number “twelve” and assignment of miraculous powers to a focal personality illustrate a phenomenon that was probably rather common: what we find in Homerica and what we find in the Islamic tradition reflect different “branches” of development from a common ancient origin. So what the above comparisons show is simply that many of the literary techniques, devices and motifs that one finds in the ArabIslamic tradition had already been in circulation for many centuries, and comprised tools in a repertoire of storytelling well-known throughout the region, not only among formally trained scholars but at the popular level as well. To some extent the Arabs probably already knew about these through centuries of contacts and interactions with the neighboring peoples of Egypt, Syria and Iraq. But there are several factors dictating that it is not within Arabia that we should be looking for the contact point for the late antique influences that can readily be seen in early Arab-Islamic historiography. To be sure, no one of these factors is decisive in and of itself, but taken together they comprise an argument for cultural environment that suffices to direct our attention in one direction as opposed to another. A telling illustration of this arises in the cases where the transmission history of an account shows remarkable discontinuities. As 203 204

Sayf in al-ˇabarì, Ta"rìkh, I, 2117:2. Sayf in ibid., I, 2169:11.

410

lawrence i. conrad

we have seen above, the argumentative value of eyewitness testimony was well-known in antiquity and is well attested in the debates waged in literary venues over important social and cultural issues. If the authority of such a witness had been well-known in the cultural milieu of pre-Islamic Arabia, or indeed, appreciated at all, then we would expect to see it deployed in the earliest stages of arguments in Arab-Islamic historical sources. But this is not the case. When appeal to the alleged eyewitness appears it is always at a late stage in the elaboration of an account, with the shift to first-person sometimes marked by nothing subtler than a recasting of the verbs in an already existing account. In terms of specific chronology this stage is mid-Umayyad to early 'Abbàsid: i.e., it appears at a time when mawàlì were assimilating to Arab-Islamic society in large numbers in Syria and Iraq and are repeatedly attested by name as tradents in literary tradition on all fronts. We may also pose a crucial question implicit in the modern scholarly work on topoi and schemata that has contributed so much to our understanding of the evolution of early Islamic literary tradition. Let us assume that pre-Islamic Arabia had been possessed of a narrative prose tradition that by the time of the Prophet had already assimilated whatever it found useful from other traditions, and then gradually assumed—in some course of continuous evolution—an Islamic identity and sense of historical purpose as Arab society itself developed and adopted such sensibilities in the course of the first centuries of Islamic history. In such a case we would expect to detect, at least in general terms, this pattern of continuity in early Islamic historiography. But this is not what we see. Instead, one finds enormous trends of discontinuity. Noth, for example, has shown that entire accounts prove to be inventions created by organising stereotyped motifs and paradigms in different ways. In some cases the discontinuity and isolation from historical reality is nothing short of astonishing. Here we may refer to the case of the conquest of the island of Arwàd off the Syrian coast in 29/650.205 The island can easily be seen from the mainland, and to this day the Syrian shabàb swim out to it to impress their girlfriends. Its very small size is also immediately evident to anyone who sees it. Early Islamic tradition, however, claims that Arwàd was “a huge island”206 and that it is 205 206

Studied at length in my “The Conquest of Arwàd,” 317–401. Ibn A'tham, Kitàb al-futù˙, II, 145:13. Cf. Conrad, “Conquest of Arwàd,” 352–53.

the

MAWÀLÌ

and early arabic historiography

411

located “near Constantinople.”207 The account of its conquest in the Arabic sources is constructed almost entirely of topoi and stock motifs that have little to do with the events in question. This sort of problem can only be interpreted as argument over a void.208 That is, in earliest Islamic times no coherent account was being transmitted about the conquest of this place, and all that was passed on was knowledge that Arwàd was an island and that it had been taken while Mu'àwiya was governor of Syria. Detailed accounts of the conquest only appear long after the event, when the Arabs found themselves in the midst of a culture in which historiography was a venerable and sophisticated part of literary tradition. No genuinely historical details were known, so an account was assembled from stock motifs reflecting later perceptions of what had usually “happened” when places were taken during the conquests. One could perhaps argue that the lack of genuine information about the capture of Arwàd reflects the unimportance of this event, but this is not an isolated case and scholars of differing methodological tendencies have discovered similar instances elsewhere in the futù˙ tradition: Noth for Ißfahàn and Nihàwand,209 for example, and Donner for Buwayb.210 There is also the fact that in many cases the specific sources for motifs, facts and paradigms in early Arab-Islamic historiography can be found in the literature familiar to the Christian and Jewish communities of late antiquity.211 The Old and New Testaments of course provided both passages for quotation and models for emulation, but that is only part of the story. The sources in use also include exegetical, legal and didactic works, and overall it is impossible to avoid the conclusion that the Islamic works reflecting these sources must have arisen in a cultural milieu in which Jewish and Christian literature was widely known and broadly discussed and debated.212 The

207 Al-Wàqidì in al-ˇabarì, Ta"rìkh, II, 163:14–15. Cf. Conrad, “Conquest of Arwàd,” 380–83. 208 Further discussion of the phenomenon in Conrad, “Conquest of Arwàd,” 387–91. 209 Noth, “Ißfahàn–Nihàwand,” 274–96. 210 Donner, The Early Islamic Conquests (Princeton: Princeton University Press, 1981), 198–200. 211 Cf. note 21 above. 212 This was a key observation informing Patricia Crone and Michael Cook’s Hagarism: the Making of the Islamic World (Cambridge: Cambridge University Press, 1977), but has tended to become lost in the controversy surrounding this work.

412

lawrence i. conrad

Qur"àn of course makes frequent reference to Jews and Christians,213 so could this milieu have been pre-Islamic Arabia? Were this the case, then one would expect to see evidence of such activity and awareness in our Arabic sources for pre-Islamic Arabia. After all, we do possess detailed accounts of, for example, the history of the Jews of Yathrib and other settlements in the peninsula,214 and there is no dearth of Arabic verse handed down as the work of Jewish215 and Christian poets.216 But of such an atmosphere we in fact find hardly any trace at all, and when Jewish and Christian poets display their values and beliefs to us what matters to them is clearly their identity as Arabs, not their religious affiliation with either Christianity or Judaism. The Jewish poet al-Samaw"al, for example, was renowned for being true to his word—an extremely important Arab tribal virtue expressed in the proverb “more faithful than al-Samaw"al.”217 When obliged to see his son slaughtered before his 213

See, for example, Schedl, Muhammad und Jesus. Die christologisch-relevanten Texte des Korans (Vienna: Herder, 1978). 214 See Horovitz, “Judaeo-Arabic Relations in Pre-Islamic Times,” Islamic Culture 3 (1929), 161–99; Torrey, The Jewish Foundation of Islam (New York: Jewish Institute of Religion Press, 1933), 1–27; Nallino, Raccolta di scritti editi e inediti, ed. Maria Nallino (Rome: Istituto per l’Oriente, 1939–48), III, 87–121; Lichtenstädter, “Some References to Jews in Pre-Islamic Arabic Literature,” Proceedings of the American Academy for Jewish Research 10 (1940), 185–94; Goitein, Jews and Arabs, 46–61; Franz Altheim, Entwicklungshilfe im Altertum. Die grosse Reiche und ihre Nachbarn (Reinbeck: Rowohlt, 1962), 53–80; Gil, “The Origin of the Jews of Yathrib,” Jerusalem Studies in Arabic and Islam 4 (1984), 203–24; Newby, “The Sìrah as a Source for Early Arabian Jewish History,” Jerusalem Studies in Arabic and Islam 7 (1986), 121–38; id., A History of the Jews of Arabia: From Ancient Times to their Eclipse under Islam (Columbia, S.C.: University of South Carolina Press, 1988), 14–77; Lecker, Muslims, Jews and Pagans: Studies on Early Islamic Medina (Leiden: E.J. Brill, 1995); id., Jews and Arabs in Pre- and Early Islamic Arabia (Aldershot: Variorum, 1998); Schöller, Exegetisches Denken und Prophetenbiographie, 171–214. Important issues are raised in the studies of M.J. Kister, for which see his Studies in Jàhiliyya and Early Islam (London: Variorum, 1980); Society and Religion from Jàhiliyya to Islam (Aldershot: Ashgate, 1990); Concepts and Ideas at the Dawn of Islam (Aldershot: Ashgate, 1997). 215 See Nöldeke, “Die Gedichte der Juden in Arabien,” in his Beiträge zur Kenntniss der Poesie der alten Araber (Hannover: C. Rümpler, 1864), 52–86; Delitzsch, Jüdischarabische Poesien aus vormuhammedischer Zeit (Leipzig: Dörffling und Franke, 1874); Steinschneider, Die arabische Literatur der Juden (Frankfurt: J. Kauffmann, 1902), 3–8; GAS, II, 249–50, 250–51, 296–97, 297. 216 Still of use is the problematic work (because of how it identifies “Christian” poets) of Louis Cheikho, Shu'arà" al-naßrànìya fì l-jàhilìya (Beirut: Ma†ba'at al-Àbà" al-Mursalìn al-Yasù'ìyìn, 1890–91). Cf. also GAS, II, 109–315, where Christian and Jewish poets are identified as they arise. 217 See al-Maydànì (d. 518/1124), Majma' al-amthàl, ed. Mu˙ammad Mu˙yì l-Dìn 'Abd al-Óamìd (Cairo: Ma†ba'at al-Sunna al-Mu˙ammadiyya, 1374/1955), II, 374a:6–375a:16.

the

MAWÀLÌ

and early arabic historiography

413

eyes rather than betray a pledge, he refers in a famous poem to the importance of upholding the pristine reputation his father had left to him: i.e., he speaks entirely in terms of traditional Arab values without the slightest reference to any sense of Jewish identity.218 We find the same with the Christian poet 'Adì b. Zayd from al-Óìra in Iraq. Arab-Islamic tradition tells us that he was unjustly seized and cast into prison, and from his cell composed poems complaining of his predicament, appealing for fair treatment and calling upon his kinsmen for aid. In these poems his pleas are again along traditional Arab lines: it dishonours his kin for them to allow him to languish in prison, his captor is related to him by marriage and owes him better, his good deeds and loyalty have been repaid with cruelty and injustice, etc. 'Adì occasionally mentions the name of God (but also the “Lord of Mecca”) and betrays that he knows what a monk and a cloister are, but nowhere does he invoke Christian values: in his verse we encounter, for example, no Christian sense of suffering, no appeal to Christ for strength or assistance.219 In other words, our Arabic sources—should we choose to rely on them for pre-Islamic Arabia—display to us Christians and Jews who regard themselves primarily as Arabs and who, even in times of great personal distress, do not fall back on the values of their faith or the solace it may offer. This hardly suggests spiritual communities steeped in and engaged with the literature of their faiths. So again, evidence for the thesis we wish to test is lacking, and again, we are led to suspect that we are looking in the wrong place and the wrong era. Finally, there are instances where the case for an Arabian milieu already steeped in such influences becomes untenable at a very specific level. An illuminating example is Mu˙ammad’s letter of Heraclius, in which, inter alia, the Prophet states that if the emperor refuses to convert or submit to Islam, “then the sin of the arìsìyìn will be held against you.” The term arìsìyìn has puzzled scholars for more than

218 Al-Samaw"al, Dìwàn, ed. Wa∂ì˙ al-Íamad (Beirut: Dàr Íàdir, 1416/1996), 96:6–99:10. 219 See the accounts in al-Ya'qùbì (d. 284/897), Ta"rìkh, ed. M.T. Houtsma (Leiden: E.J. Brill, 1883), I, 241:13–246:7; al-ˇabarì, Ta"rìkh, I, 1016:10–1029:3; Aghànì, II, 105:8–120:5; also Nöldeke, Geschichte der Perser und Araber zur Zeit der Sasaniden (Leiden: E.J. Brill, 1879), 312–32; Horovitz, “'Adì ibn Zayd: the Poet of al-Óìra,” Islamic Culture 4 (1930), 44–49. The poems may also be found scattered through 'Adì’s Dìwàn, ed. Mu˙ammad Jabbàr al-Mu'aybid (Baghdàd: Sharikat Dàr al-Jumhuriyya li-l-Nashr wa-l-ˇab', 1385/1965).

414

lawrence i. conrad

a century, but as I seek to demonstrate elsewhere,220 it is simply an approximation of arìsìn, the plural of the Palestinian Aramaic arìs, a “sharecropper,”221 the allusion of the passage being to the New Testament Parable of the Husbandmen in a context of inter-faith disputes with Christians.222 It can also be shown, on the one hand, that the “Heraclius and Mu˙ammad” traditions revolve around al-Zuhrì, and on the other, that his source for the Parable of the Husbandmen is not the Syriac New Testament itself, where the words arìs and arìsìn never appear, but rather a lectionary in Palestinian Aramaic with the section headings in Arabic, all written in Syriac script. This is not the only place where this lectionary proves to be the source used by early Muslim tradents,223 so the question is simply this: Where have they gained access to such a text? One could of course force or assume a case for Medina, but this is a slippery sphere: one can hold it or drop it, but not do anything with it or discuss it since it rests upon no evidence. Surely the more reasonable conclusion would be that al-Zuhrì was exposed to this work, or more likely to extracts made from it in Arabic, during one of his long sojourns in Syria.224 That this work was available to him in any form at all clearly reflects the kind of social context we find in Umayyad Syria: an increasing pace of conversion to Islam, the presence in Muslim ranks of a new generation of scholars of recent convert background, and religious discussions and disputes that encouraged an interest in the content of other religious traditions. 220 Conrad, “Heraclius in Early Islamic Kerygma,” in Gerrit J. Reinink and Bernard H. Stolte, eds., The Reign of Heraclius (610 –644): Crisis and Confrontation (Leuven: Peeters, 2002), 127–30. 221 The similarity between the Arabic and the Aramaic forms was already noted in Hubert Grimme, Mohammed (Münster: Aschendorff, 1892–95), I, 125 note 4, who admitted, however, that the sense behind such an observation remained unclear to him. 222 Matthew 21:33–41. 223 Baumstark, “Eine altarabisches Evangelübersetzung aus dem Christlich-Palästinensischen,” Zeitschrift für Semitistik 8 (1932), 201–209; Guillaume, “The Version of the Gospels used in Medina ca. A.D. 700,” Al-Andalus 15 (1950), 289–96; Schacht, “Une citation de l’Evangile de St. Jean dans la Sìra d’Ibn Is˙àq,” Al-Andalus 16 (1951), 489–90; Griffith, “The Gospel in Arabic: an Inquiry into its Appearance in the First Abbasid Century,” Oriens Christianus 69 (1985), 137–43. 224 See, for example, Faruqi, Early Muslim Historiography, 235–61; Óusayn 'A†wàn, Riwàyat al-shàmìyìn li-l-maghàzì wa-l-siyar (Beirut: Dàr al-Jìl, 1986), 69–150; KellnerHeinkele, “Rußàfa in den arabischen Quellen,” in Dorothée Sack, ed., Rusafa IV. Die Grosse Moschee von Resafa—Rußàfat Hishàm (Mainz: von Zabern, 1986), 133–54; Lecker, “Biographical Notes on Ibn Shihàb al-Zuhrì,” Journal of Semitic Studies 41 (1996), 21–63.

the

MAWÀLÌ

and early arabic historiography

415

For the milieu for the influences traced out above, then, we should be looking beyond Arabia north to the heartlands of the Near East itself, and to a time several generations after the life of the Prophet when, in various parts of the Umayyad realm, increasing numbers of non-Arabs were converting to Islam and affiliating themselves to Arab tribes in order to attain a secure position within the new social order of Islam. These mawàlì, shifting from the vast non-Arab and non-Muslim majority to the circle of the ruling elite, brought with them their own traditions of narration and literary expression. These ideas and sensibilities played an enormous role in shaping the ArabIslamic historical literature so familiar to us today, or rather in perpetuating literary and cultural traditions that had already flourished for centuries, and giving them a new dynamic sense of direction and purpose. Thinking in terms of the nature of the creative impulse that propels one society but not another to greatness as a world civilisation, it may perhaps be said that late antiquity was but a catalyst to the creative impulse implicit in the social and cultural conditions that came together under the auspices of early Islam. Certainly the combination was explosive in the most positive sense of the term, and led to the establishment of a cultural-literary tradition of an energy and ambition that would have been impossible to envisage earlier.

Bibliography Primary Sources Abù l-Faraj al-Ißfahànì, 'Alì b. al-Óusayn. Kitàb al-aghànì, ed. A˙mad Zakì al-'Adawì et al., 24 vols. Cairo, 1345–94/1927–74. Abù Nu'aym al-Ißfahànì, A˙mad b. 'Abd Allàh. Óilyat al-awliyà" wa-†abaqàt al-aßfiyà". 10 vols. Cairo: Maktabat al-Khànjì, 1354/1935. Abù 'Ubayd al-Qàsim b. Sallàm al-Juma˙ì. Kitàb al-silà˙, ed. Óàtim Íàli˙ Îamìn. Beirut: Mu"assasat al-Risàla, 1985. 'Adì b. Zayd. Dìwàn, ed. Mu˙ammad Jabbàr al-Mu'aybid. Baghdàd: Sharikat Dàr al-Jumhuriyya, 1385/1965. Akhbàr al-dawla al-'abbàsiyya, ed. 'Abd al-'Azìz al-Dùrì and 'Abd al-Jabbàr al-Mu††alibì. Beirut: Dàr al-ˇalì'a, 1391/1971. Apollodorus. Bibliotheca, trans. Robin Hard as The Library of Greek Mythology. Oxford: Oxford University Press, 1997. id., Epitome, ed. Richard Wagner. Leipzig: Hirzel, 1891. Aristophanes. Pax, ed. and comm. J. van Leeuwen. Leiden: A.W. Sijhoff, 1906. al-'Askarì, Abù Hilàl al-Hasan b. 'Abd Allàh. Kitàb al-awà"il, ed. Mu˙ammad alMißrì and Walìd Qaßßàb, 2 vols. Damascus: Wizàrat al-Thaqàfa wa-l-Irshàd alQawmì, 1975.

416

lawrence i. conrad

al-Azdì, Abù Ismà'ìl Mu˙ammad b. 'Abd Allàh. Futù˙ al-Shàm, ed. William Nassau Lees. Calcutta: Baptist Mission Press, 1854; Ta"rìkh futù˙ al-Shàm, ed. 'Abd alMun'im 'Abd Allàh 'Àmir. Cairo: Mu"assasat Sijill al-'Arab, 1970. al-Balàdhurì, Abù l-Óasan A˙mad b. Ya˙yà b. Jàbir. Futù˙ al-buldàn, ed. M.J. de Goeje. Leiden: E.J. Brill, 1866. al-Bukhàrì, Abù 'Abd Allàh Mu˙ammad b. Ismà'ìl. Al-Jàmi' al-ßa˙ì˙, ed. Ludolf Krehl and Th.W. Juynboll, 4 vols. Leiden: E.J. Brill, 1862–1908. id., Al-Ta"rìkh al-kabìr, ed. Mu˙ammad 'Abd al-Mu'ìd Khàn, 8 parts in 4 vols. Hyderabad: Dà"irat al-Ma'àrif al-'Uthmàniyya, 1360–64/1940–44. Caskel, Werner, ]amharat an-nasab. Das genealogische Werk des Hi“àm ibn Mu˙ammad alKalbì. 2 vols. Leiden: E.J. Brill, 1966. Cheikho, Louis, ed. Shu'arà" al-naßràniyya fì l-jàhiliyya. 2 vols. in 6 parts. Beirut: Ma†ba'at al-Àbà" al-Mursalìn al-Yasù'ìyìn, 1890–91. Chronicon ad annum Christi 1234 pertinens, ed. J.-B. Chabot. 2 vols. CSCO 81–81, Scr. syri 36–37. Paris: L. Durbecq, 1916–20. Cypria, see Epic Cycle. ps.-Dares Phrygius. De excidio Troiae historia, ed. Ferdinand Meister. Leipzig: B.G. Teubner, 1873. ps.-Dictys Cretensis. Ephemeris belli troiani, Latin translation ed. Werner Eisenhut. 2nd ed. Leipzig: B.G. Teubner, 1973. Al-Durr al-nafìs fì uns al-Ωà'in wa-l-jalìs. Real Academia de la Historia (Madrid), Ms. Gayangos no. 18. Epic Cycle. Fragments ed. Thomas W. Allen in Homeri opera, V. Oxford: Clarendon Press, 1912. Euripides. Hecuba, ed. and verse trans. Arthur S. Way in the Loeb Euripides. Cambridge/Mass.: Harvard University Press, 1912, I; ed. and prose trans. David Kovacs in the Loeb Euripides. Cambridge/Mass.: Harvard University Press, 1994–98, II, 391–519. Eustathius of Thessalonica. Eustathii commentarii ad Homeri Iliadem pertinentes, ed. Marchinus van der Valk, 4 vols. Leiden: E.J. Brill, 1971–87. Guardini, Maria Lia, ed. and comm. Il mito di Elena: Euripide e Isocrate. Treviso: Editioni Canova, 1987. Herodotus. Historiae, ed. Haiim B. Rosen, 2 vols. Leipzig: B.G. Teubner, 1987–97. Hesiod. Catalogi sive eoearum, ed. and trans. Hugh G. Evelyn-White as The Catalogues of Women and Eoiae in the Loeb Hesiod, rev. ed. London: William Heinemann, 1936, 155–219. Homer. The Iliad of Homer, trans. Richmond Lattimore. Chicago: University of Chicago Press, 1951. id., The Odyssey of Homer, trans. Richmond Lattimore. New York: Harper Collins, 1965. Hyginus. Fabulae, ed. Peter K. Marshall. Leipzig: B.G. Teubner, 1993. Ibn 'Abd al-Óakam, Abù l-Qàsim 'Abd al-Ra˙màn b. 'Abd Allàh. Futù˙ Mißr waakhbàruhà, ed. Charles C. Torrey. New Haven: Yale University Press, 1922. Ibn A'tham al-Kùfì, Abù Mu˙ammad A˙mad. Kitàb al-futù˙, ed. Mu˙ammad 'Abd al-Mu'ìd Khàn et al., 8 vols. Hyderabad: Dà"irat al-Ma'àrif al-'Uthmàniyya, 1388–95/1968–75. id., (falsely ascribed to al-Wàqidì). Kitàb al-ridda wa-nubdha min futù˙ al-'Iràq, ed. Mu˙ammad Óamìd Allàh. Paris: Editions Tougui, 1409/1989. Ibn Óabìb al-Baghdàdì, Abù Ja'far Mu˙ammad. Kitàb al-mu˙abbar, ed. Ilse Lichtenstädter. Hyderabad: Dà"irat al-Ma'àrif al-'Uthmàniyya, 1361/1942. Ibn Óanbal, Abù 'Abd Allàh A˙mad b. Mu˙ammad. Musnad. 6 vols. Cairo: alMa†ba'a al- Maymuniyya, 1311 AH. Ibn Hishàm al-Ma'àfirì, Abù Mu˙ammad 'Abd al-Malik. Sìrat Rasùl Allàh, ed. Ferdinand Wüstenfeld, 2 vols. Göttingen, 1858–60.

the

MAWÀLÌ

and early arabic historiography

417

Ibn Is˙àq, Abù 'Abd Allàh Mu˙ammad. Sìra, recension of Yùnus b. Bukayr, ed. Mu˙ammad Óamìd Allàh. Rabat: Ma'had al-Diràsàt wa-l-Ab˙àth li-l-Ta'rìb, 1976. id., Kitàb al-siyar wa-l-maghàzì, recension of Yùnus b. Bukayr, ed. Suhayl Zakkàr. Damascus: Dàr al-Fikr, 1398/1978. Ibn Khaldùn, Abù Zayd 'Abd al-Ra˙màn b. Mu˙ammad. Al-Muqaddima, ed. Etienne Quatremère, 3 vols. Paris: Imprimerie Nationale, 1858. Ibn Sa'd, Abù 'Abd Allàh Mu˙ammad. Kitàb al-†abaqàt al-kabìr, ed. Eduard Sachau et al., 9 vols. Leiden: E.J. Brill, 1904–40. Levi della Vida, Georgio, ed. Les “livres des chevaux” de Hi“àm ibn al-Kalbì et Mu˙ammad ibn al-A'ràbì. Leiden: E.J. Brill, 1928. Lycophron. Alexandra, text and scholia ed. Eduard Scheer, 2 vols. Berlin: Wiedemann, 1881–1908. Malalas, John. Chronographia, trans. Elizabeth Jeffreys, Michael Jeffreys and Roger Scott as The Chronicle of John Malalas. Melbourne: Australian Association for Byzantine Studies, 1986. al-Maydànì, Abù l-Fa∂l A˙mad b. Mu˙ammad. Majma' al-amthàl, ed. Mu˙ammad Mu˙yì l-Dìn 'Abd al-Óamìd, 2 vols. Cairo: Ma†ba'at al-Sunna al-Mu˙ammadiyya, 1374/1955. al-Mizzì, Abù l-Óajjàj Yùsuf b. 'Abd al-Ra˙màn. Tahdhìb al-kamàl fì ma'rifat al-rijàl, ed. Bashshàr 'Awwàd Ma'rùf, 35 vols. Beirut, 1403–13/1983–92. Muslim b. al-Óajjàj al-Qurashì, Abù l-Óusayn. Ía˙ì˙ Muslim, 2 vols. Cairo: alMa†ba'a al- Maymuniyya, 1290 AH. al-Nabìl, Abù Bakr A˙mad b. 'Amr. Kitàb al-awà"il, ed. Ma˙mùd Mu˙ammad Ma˙mùd Óasan Naßßàr. Beirut: Dàr al-Jìl, 1991. al-Nadìm, Abù Ya'qùb Mu˙ammad b. Is˙àq. Kitàb al-fihrist, ed. Ri∂à-Tajaddud. Tehran: Ma†ba'at Dànishgàh, 1393/1973. Quintus Smyrnaeus. Posthomerica, ed. Albert Zimmermann. Leipzig: B.G. Teubner, 1891. al-Samaw"al. Dìwàn, ed. Wa∂ì˙ al-Íamas. Beirut: Dàr Íàdir, 1416/1996. Sayf b. 'Umar al-Tamìmì. Kitàb al-ridda wa-l-futù˙ wa-Kitàb al-jamal wa-masìr 'À"isha wa- 'Alì, ed. Qasim al-Samarrai. Leiden: Smitskamp Oriental Antiquarium, 1995. al-ˇabarì, Abù Ja'far Mu˙ammad b. Jarìr. Ta"rìkh al-rusul wa-l-mulùk, ed. M.J. de Goeje et al., 15 vols. Leiden: E.J. Brill, 1879–1901. Virgil. Aeneid, trans. H. Rushton Fairclough and rev. G.P. Goold in the Loeb Virgil, 2 vols. New York: Harvard University Press, 1999–2000. al-Yàfi'ì, Abù Mu˙ammad 'Abd Allàh b. As'ad. Mir"àt al-janàn wa-'ibrat al-yaqΩàn, 4 vols. Hyderabad, 1337–39 AH. al-Ya'qùbì, A˙mad b. Abì Ya'qùb b. Wà∂i˙. Ta"rìkh, ed. M.T. Houtsma, 2 vols. Leiden: E.J. Brill, 1883. Secondary Sources Ahlberg-Cornell, Gudrun. Myth and epos in early Greek art: Representation and interpretation. Jonsered: Paul Åströms Förlag, 1992. Allen, Thomas W. “The epic cycle,” Classical Quarterly 2 (1908), 64–74, 81–88. id., Homer: The origins and the transmission. Oxford: Clarendon Press, 1924. id., The Homeric catalogue of ships. Oxford: Clarendon Press, 1921. Altheim, Franz. Entwicklungshilfe im Altertum. Die grosse Reiche und ihre Nachbarn. Reinbeck: Rowohlt, 1962. Anderson, Graham. Philostratus: Biography and Belles Lettres in the third century A.D. London: Croom Helm, 1986. Anderson, William S. “Vergil’s second Iliad,” Transactions and Proceedings of the American Philological Society 88 (1957), 17–30.

418

lawrence i. conrad

Andrae, Tor. Die Person Muhammeds in Lehre and Glauben seiner Gemeinde. Stockholm: P.A. Norstedt, 1918. von Arnim, Hans. Leben und Werke des Dio von Prusa. Berlin: Weidmannsche Buchhandlung, 1898. 'A†wàn, Óusayn. Riwàyat al-shàmìyìn li-l-maghàzì wa-l-siyar. Beirut: Dàr al-Jìl, 1986. Auerbach, Erich. Mimesis. Dargestellte Wirklichkeit in der abendländischen Literatur. Bern: A. Francke, 1946. Bashear, Suliman. Arabs and others in early Islam. Princeton: Darwin Press, 1997. Baumstark, Anton. “Eine altarabisches Evangelübersetzung aus dem ChristlichPalästinensischen,” Zeitschrift für Semitistik 8 (1932), 201–209. id., “Die zweite Achilleus-Trilogie des Aeschylos,” Philologus 55 (1896), 277–306. al-Bayàtì, 'Àdil Jàsim. Kitàb ayyàm al-'arab qabla l-Islàm. Beirut: 'Àlam al-Kutub, 1407/1987. Becker, Andrew Sprague. The shield of Achilles and the Poetics of Ekphrasis. London: Rowman and Littlefield, 1995. Beschorner, Andreas. Untersuchungen zu Dares Phrygius. Tübingen: Gunter Narr Verlag, 1992. Beye, Charles Rowan. Ancient epic poetry: Homer, Apollonius, Virgil. Ithaca/London: Cornell University Press, 1993. de Biasi, Luciano. “I ritratti dei personaggi in Darete Frigio: raffronto con i testi fisiognominici,” Koinonia 3 (1979), 53–112. Bowra, C.M. Tradition and design in the Iliad, 2nd ed. Oxford: Clarendon Press, 1950. Bradley, Dennis R. “Troy revisited,” Hermes 119 (1991), 232–46. Browning, Robert. “The Byzantines and Homer,” in Robert Lamberton and John J. Keaney, eds., Homer’s ancient readers: The hermeneutics of Greek epic’s earliest exegetes. Princeton: Princeton University Press, 1992, 134–48. id., “A fourteenth-century prose version of the Odyssey,” in Anthony Cutler and Simon Franklin, eds., Homo Byzantinus: Papers in Honor of Alexander Kazhdan. Washington, DC: Dumbarton Oaks, 1992; Dumbarton Oaks Papers 46, 27–36. id., “Homer in Byzantium,” Viator 6 (1975), 15–33. Bulliet, Richard W. Conversion to Islam in the medieval period: An essay in quantitative history. Cambridge/Mass.: Harvard University Press, 1979. Caetani, Leone. Annali dell’Islam. 10 vols. Milan: Ulrico Hoepli, 1905–24. Caskel, Werner. “Aijàm al-'arab. Studien zur altarabischen Epik,” Islamica 3, Ergänzungsheft (1930), 1–99. id., Die Bedeutung der Beduinen in der Geschichte der Araber. Köln/Opladen: Westdeutscher Verlag, 1953. Cesaretti, Paulo. Allegoristi di Omero a Bisanzio: ricerche ermeneutiche (11–12 secolo). Milan: Guerini, 1991. Clark, Mark Edward. “Neoanalysis: A bibliographical review,” Classical World 79 (1985–86), 379–94. Collilieux, Eugène. Etude sur Dictys de Crête et Darès de Phrygie. Grenoble: Xavier Drevet, 1886. Conrad, Lawrence I. “Abraha and Mu˙ammad: Some observations apropos of chronology and literary topoi in the early Arabic historical tradition,” Bulletin of the School of Oriental and African Studies 50 (1987), 225–40. id., “The Arabs and the Colossus,” Journal of the Royal Asiatic Society, Third Series, 6 (1996), 165–87. id., “Al-Azdì’s History of the Arab conquests in Bilàd al-Shàm: Some historiographical considerations,” in Mu˙ammad 'Adnàn al-Bakhìt, ed., Proceedings of the Second Symposium on the History of Bilàd al-Shàm during the early Islamic period up to 40 A.H./640 A.D., English and French papers. Amman: University of Jordan, 1987, I, 28–62.

the

MAWÀLÌ

and early arabic historiography

419

id., “The conquest of Arwàd: A source-critical study in the historiography of the early medieval Near East,” in Averil Cameron and Lawrence I. Conrad, eds., The Byzantine and early Islamic Near East, I: Problems in the literary source material. Papers of the First Workshop on Late Antiquity and Early Islam. Princeton: Darwin Press, 1992, 317–401. id., “Heraclius in early Islamic Kerygma,” in Gerrit J. Reinink and Bernard H. Stolte, eds., The reign of Heraclius (610–644): Crisis and confrontation. Leuven: Peeters, 2002, 113–56. id., “Historical evidence and the archaeology of early Islam,” in Samir Seikaly, Ramzi Baalbaki, and Peter Dodd, eds., Quest for understanding: Arabic and Islamic Studies in memory of Malcolm H. Kerr. Beirut: American University of Beirut, 1991, 263–82. id., Muhammadanea Edessensis: The rise of Islam in eastern Christian historiography under the early 'Abbàsids. Princeton: Darwin Press, forthcoming. id., “The Near East study tour diary of Ignaz Goldziher,” Journal of the Royal Asiatic Society (1990), 105–26. id., “A ninth-century Muslim scholar’s discussion of contagion,” in Lawrence I. Conrad and Dominik Wujastyk, eds., Contagion: Perspectives from pre-modern societies. Aldershot: Ashgate, 2000, 163–77. id., “Syriac perspectives on Bilàd al-Shàm during the 'Abbàsid Period,” in Mu˙ammad 'Adnàn al-Bakhìt and Robert Schick, eds., Bilàd al-Shàm during the 'Abbàsid period (132 A.H./750 A.D.-451 A.H./1059 A.D.). Amman: University of Jordan, 1412/1991, 1–44. id., “Theophanes and the Arabic historical tradition: Some indications of intercultural transmission,” Byzantinische Forschungen 15 (1990), 1–44. id., “'Umar at Sargh: the evolution of an Umayyad tradition on flight from the plague,” in Stefan Leder, ed., Story-telling in the framework of non-fictional Arabic literature. Wiesbaden: Otto Harrassowitz Verlag, 1998, 489–528. id., “Varietas Syriaca: Secular and scientific culture in the Christian communities of Syria after the Arab conquest,” in G.J. Reinink and A.C. Klugkist, eds., After Bardaisan: Studies on continuity and change in Syriac Christianity in Honour of Han J.W. Drijvers. Leuven: Peeters, 1999, 85–105. Cook, David, Studies in Islamic apocalyptic. Princeton: Darwin Press, 2002. Cook, Michael. Early Muslim dogma: A source-critical study. Cambridge: Cambridge University Press, 1981. id., Muhammad. Oxford: Oxford University Press, 1983. Crone, Patricia. Meccan trade and the rise of Islam. Oxford: Blackwell’s, 1987. id., Roman, Provincial and Islamic law: The origins of the Islamic patronate. Cambridge: Cambridge University Press, 1987. id. and Michael Cook. Hagarism: The making of the Islamic world. Cambridge: Cambridge University Press, 1977. Curtius, Ernst Robert. Europäische Literatur und lateinisches Mittelalters. Bern: A. Francke, 1948. Davies, Malcolm. “The date of the epic cycle,” Glotta 67 (1989), 89–100. id., The epic cycle. Bristol: Bristol Classical Press, 1988. Delitzsch, Franz. Jüdisch-arabische Poesien aus vormuhammedischer Zeit. Leipzig: Dörffling und Franke, 1874. Desideri, Paolo. Dione di Prusa, un intellectuale greco nell’impero romano. Messina/Florence: G. D’Anna, 1978. Donner, Fred M. The early Islamic conquests. Princeton: Princeton University Press, 1981. id., Narratives of Islamic origins: The beginning of Islamic historical writing. Princeton: Darwin Press, 1998.

420

lawrence i. conrad

Duri, A.A. The rise of historical writing among the Arabs, ed. and trans. Lawrence I. Conrad. Princeton: Princeton University Press, 1983. Edwards, Mark W. Homer: Poet of the Iliad. Baltimore/London: Johns Hopkins University Press, 1987. id., “Neo-analysis and beyond,” Classical Antiquity 9 (1990), 311–25. Eisenstein, Herbert. “Überlegungen zu einer Darstellung der Rolle des Pferdes in der arabisch-islamischen Kulturgeschichte,” in Arne A. Ambros and Markus Köhlbach, eds., Festschrift für Hans Hirsch zum 65. Geburtstag. Vienna: Institut für Orientalistik, 1996, 107–17. van Ess, Josef. Anfänge muslimischer Theologie. Zwei antiqadaritische Traktate aus dem ersten Jahrhundert der Hi[ra. Wiesbaden: Franz Steiner Verlag, 1977. id., Theologie und Gesellschaft im 2. und 3. Jahrhundert Hidschra. Eine Geschichte des religiösen Denkens im frühen Islam, 6 vols. Berlin: Walter de Gruyter, 1991–95. Evans, Elizabeth Cornelia. “Descriptions of personal appearance in Roman history and biography,” Harvard Studies in Classical Philology 46 (1935), 43–84. Faruqi, Nisar Ahmed. Early Muslim historiography: A study of early transmitters of Arab history from the rise of Islam up to the end of the Umayyad period (612–750 A.D.). New Delhi: Idarah-i Adabiyat-i Delli, 1979. Fenik, Bernard. Typical battle scenes in the Iliad: Studies in the narrative techniques of Homeric battle description. Hermes Einzelschrift no. 21. Wiesbaden: Franz Steiner Verlag, 1968. von Fleschenberg, Otmar Schissel. Dares-Studien. Halle-an-der-Salle: Max Niemeyer, 1908. Frantz-Murphy, Gladys. “Conversion in Early Islamic Egypt,” in Yousef Ragheb, ed., Documents de l’Islam médiéval. Paris: CNRS, 1991, 11–17. Fück, Johann. Mu˙ammad ibn Is˙àq. Literarhistorische Untersuchungen. Frankfurt am Main, 1926. GAS = Fuat Sezgin. Geschichte des arabischen Schrifttums, 12 volumes to date. Leiden: E.J. Brill, and Frankfurt: Institut für Geschichte der arabisch-Islamischen Wissenschaft, 1967—proceeding. Georgacopoulou, Sophia. “L’Achilléide de Stace et l’Achilléide anonyme byzantine,” Classica et mediaevalia 95 (1994), 251–86. Germain, Gabriel. Homère et la mystique des nombres. Paris: Presses universitaires de France, 1954. Gianotti, Gian Franco. “Le metamorfosi di Omero. Il romanze di Troia della specializzazione delle scholae ad un publico di non specialisti,” Sigma 12 (1979), 15–32. Gil, Moshe. “The origin of the Jews of Yathrib,” Jerusalem Studies in Arabic and Islam 4 (1984), 203–24. Giovannini, A. Etude historique sur les origines du Catalogue des vaisseaux. Bern: Francke, 1969. Glover, Terrot Reaveley. Life and letters in the 4th century. Cambridge: Cambridge University Press, 1901. Goitein, S.D. Jews and Arabs: Their contacts through the ages. New York: Schocken Books, 1955. Goldziher, Ignaz. Muslim Studies, ed. and trans. S.M. Stern and C.R. Barber, 2 vols. London: George Allen and Unwin, 1967–71. Griffin, Jasper. “The epic cycle and the uniqueness of Homer,” Journal of Hellenic Studies 97 (1977), 39–53. Griffin, Nathaniel Edward. Dares and Dictys: An introduction to the study of medieval versions of the story of Troy. Baltimore: J.H. Furst, 1907. Griffith, Sidney H. “The Gospel in Arabic: An inquiry into its appearance in the first Abbasid century,” Oriens Christianus 69 (1985), 137–43.

the

MAWÀLÌ

and early arabic historiography

421

Grimme, Hubert. Mohammed. 2 vols. Münster: Aschendorff, 1892–95. Guillaume, Alfred. New light on the life of Mu˙ammad. Manchester: Manchester University Press, 1960. id., “The version of the Gospels used in Medina ca. A.D. 700,” Al-Andalus 15 (1950), 289–96. Haldon, J.F. Byzantium in the seventh century: The transformation of a culture. Cambridge: Cambridge University Press, 1990. Hary, Benjamin H., John L. Hayes, and Fred Astren, eds., Judaism and Islam: Boundaries, communication and interaction—Essays in Honor of William M. Brinner. Leiden: E.J. Brill, 2000. Hawting, Gerald R. The idea of idolatry and the emergence of Islam: From polemic to history. Cambridge: Cambridge University Press, 1999. Henninger, Joseph. “Altarabische Genealogie,” Anthropos 61 (1966), 852–70. id., “La società bédouine ancienne,” in Francesco Gabrieli, ed., L’Antica società beduina. Rome: Università di Roma, 1959, 69–93. Hitti, Philip K. History of the Arabs. London: Macmillan, 1937, and various subsequent editions. Horovitz, Josef. “'Adì ibn Zayd: The poet of al-Óìra,” Islamic Culture 4 (1930), 31–69. id., “Biblische Nachwirkungen in der Sira,” Der Islam 12 (1922), 184–89. id., “Judaeo-Arabic relations in pre-Islamic times,” Islamic Culture 3 (1929), 161–99. id., Koranische Untersuchungen. Berlin: Walter de Gruyter, 1926. id., “Die poetischen Einlagen der Sìra,” Islamica 2 (1926), 308–12. id., “Zur Mu˙ammadlegende,” Der Islam 5 (1914), 41–53. Hoyland, Robert G. Seeing Islam as others saw it: A survey and evaluation of Christian, Jewish and Zoroastrian writings on early Islam. Princeton: Darwin Press, 1997. Jabbur, Jibrail. The Bedouins and the desert: Aspects of nomadic life in the Arab East, trans. Lawrence I. Conrad, ed. Suhayl S. Jabbur and Lawrence I. Conrad. Albany: State University of New York Press, 1995. Jones, Alan. “The prose literature of pre-Islamic Arabia,” in J.R. Smart, ed., Tradition and modernity in Arabic language and literature. Richmond, Surrey: Curzon, 1996, 229–41. id., “The Qur"àn in the light of earlier Arabic prose,” in Alan Jones, ed., University Lectures in Islamic Studies. London: Altajir World of Islam Trust, 1997, I, 67–83. Jones, C.P. The Roman world of Dio Chrysostom. Cambridge/Mass.: Harvard University Press, 1978. Juda, Jamal. Die sozialen und wirtschaftlichen Aspekte der Mawàlì in frühislamischer Zeit. Tübingen: Universität Tübingen, 1983. Juynboll, G.H.A. Muslim Tradition: Studies in chronology, provenance and authorship of early Óadìth. Cambridge: Cambridge University Press, 1983. Kakrides, Johannes Th. Homeric Researches. Lund: C.W.K. Gleerup, 1949. von Kamptz, Hans. Homerischen Personennamen. Sprachwissenschaftliche und historische Klassifikation. Göttingen: Vandenhoeck und Ruprecht, 1982. Karoui, Said. Die Rezeption der Bibel in der frühislamischen Literatur. Heidelberg: Seminar für Sprachen und Kulturen des Vorderen Orients, 1997. Katsh, Abraham I. Judaism in Islam: Biblical and Talmudic backgrounds of the Koran and its commentaries. New York: Bloch, 1954. Kellner-Heinkele, Barbara. “Rußàfa in den arabischen Quellen,” in Dorothée Sack, ed., Rusafa IV. Die Grosse Moschee von Resafa—Rußàfat Hishàm. Mainz: von Zabern, 1986, 133–54. Kennedy, George A. Greek rhetoric under Christian emperors. Princeton: Princeton University Press, 1983. Khalidi, Tarif. Arabic historical thought in the Classical Period. Cambridge: Cambridge University Press, 1994.

422

lawrence i. conrad

Khoury, Raif Georges. “Die arabischen literarischen Übersetzungen aus dem Griechischen, unter besonderer Berücksichtigung der Ilias von Homer,” in Piotr O. Scholz and Reinhard Stempel, eds., Nubia et Oriens Christianus. Festschrift für C. Detlef G. Müller zum 60. Geburtstag. Köln: Verlag Jürgen Dinter, 1987, 163–80. Kindstrand, Jan Fredrik. Homer in der Zweiten Sophistik. Stockholm: Almqvist und Wiksell, 1973. King, Katherine Callen. “Achilles Amator,” Viator 16 (1985), 21–64. id., Achilles: Paradigms of the War Hero from Homer to the Middle Ages. Berkeley/Los Angeles: University of California Press, 1987. Kirk, G.S., gen. ed. The Iliad: A Commentary. 6 vols. Cambridge: Cambridge University Press, 1985–93. Vols. 1–2, ed. G.S. Kirk; vol. 3, ed. Bryan Hainsworth; vol. 4, ed. Richard Janko; vol. 5, ed. Mark W. Edwards; vol. 6, ed. Nicholas Richardson. Kister, M.J. Concepts and ideas at the dawn of Islam. Aldershot: Ashgate, 1997. id., Society and religion from Jàhiliyya to Islam. Aldershot: Ashgate, 1990. id., Studies in Jàhiliyya and early Islam. London: Variorum, 1980. Knight, W.F. Jackson. Vergil’s Troy: Essays on the second book of the Aeneid. Oxford: Basil Blackwell, 1932. Kohlberg, Etan , “From Imàmiyya to Ithnà'ashariyya,” Bulletin of the School of Oriental and African Studies 39 (1976), 521–34. Kraemer, Jörg. “Arabische Homerverse,” Zeitschrift der Deutschen Morgenländischen Gesellschaft 106 (1956), 259–316; 107 (1957), 511–18. Krischer, Tilman. Formule Konventionen der homerischen Epik. Munich: C.H. Beck, 1971. Kullmann, Wolfgang. “Ergebnisse der motivgeschichtlichen Forschung in Homer (Neoanalyse),” in Joachim Latacz, ed., Zweihundert Jahre Homer-Forschung. Rückblick und Ausblick. Stuttgart/Leipzig: B.G. Teubner, 1991, 425–55. id., “Oral poetry theory and neoanalysis in Homeric research,” Greek, Roman and Byzantine Studies 25 (1984), 307–23. id., Die Quellen der Ilias (Troischer Sagenkreis). Hermes Einzelschrift no. 14. Wiesbaden: Franz Steiner Verlag, 1960. id., “Zur Methode der Neoanalyse in der Homerforschung,” Wiener Studien, Neue Folge, 15 (1981), 5–42. Landau-Tasseron, Ella. “New data on an old manuscript: An Andalusian version of the works entitled Futù˙ al-Shàm,” Al-Qan†ara 21 (2000), 361–80. id., “Processes of redaction: The case of the Tamìmite delegation to the Prophet Mu˙ammad,” Bulletin of the School of Oriental and African Studies 49 (1986), 253–70. Lapidus, Ira. “The Conversion of Egypt to Islam,” Israel Oriental Studies 2 (1972), 248–62. Latacz, Joachim. Kampfparänese, Kampfdarstellung und Kampfwirklichkeit in der Ilias, bei Kallinos und Tyrtaios. Zetemata no. 66. Munich: C.H. Beck 1977. Lebarbe, Jules. L’Homère de Platon. Liège: Faculté de philosophie et lettres, 1949. Lecker, Michael. “Biographical notes on Ibn Shihàb al-Zuhrì,” Journal of Semitic Studies 41 (1996), 21–63. id., Jews and Arabs in pre- and early Islamic Arabia. Aldershot: Variorum, 1998. id., Muslims, Jews and Pagans: Studies on early Islamic Medina. Leiden: E.J. Brill, 1995. Lecomte, Gérard. “Les citations de l’Ancien et du Nouveau Testament dans l’oeuvre d’Ibn Qutayba,” Arabica 5 (1958), 34–46. Leinieks, Valdis. “The Iliad and the epic cycle,” Classical Outlook 52 (1975), 62–64. Lichtenstädter, Ilse. “Some references to Jews in pre-Islamic Arabic literature,” Proceedings of the American Academy for Jewish Research 10 (1940), 185–94. LIMC = Lexicon iconographicum mythologiae classicae. 8 vols. Zurich/Munich: Artemis, 1981–99. Lumiansky, R.M. “Dares’ Historia and Dictys’ Ephemeris: A critical comment,” in E. Bagby Atwood and Archibald A. Hill, eds., Studies in language, literature and culture of the Middle Ages and after. Austin: University of Texas at Austin, 1969, 200–209.

the

MAWÀLÌ

and early arabic historiography

423

Macdonald, M.C.A. “Hunting, fighting and raiding: The horse in pre-Islamic Arabia,” in David Alexander, ed., Furùsiyya: The horse in the art of the Near East (Riyadh: King Abdulaziz Public Library, 1996), 73–83. Mango, Cyril. “Discontinuity with the Classical Past in Byzantium,” in Margaret Mullett and Roger Scott, eds., Byzantium and the Classical Tradition (Birmingham: Centre for Byzantine Studies, University of Birmingham, 1981), 48–57. Mannà', Hàshim Íàli˙. Al-Nathr fì l-'aßr al-jàhilì. Beirut: Dàr al-fikr al-'arabì, 1993. Matz, David. Ancient World Lists and Numbers: Numerical Phrases and Rosters in the GrecoRoman Civilizations. London: McFarland, 1995. Meisami, Julie Scott, and Paul Starkey, eds.. Encyclopedia of Arabic Literature. 2 vols. London: Routledge, 1998. Merkle, Stefan. Die Ephemeris belli Troiani des Diktys von Kreta. Frankfurt am Main: Peter Lang, 1989. id., “Troiani belli verior textus. Die Trojaberichte des Dictys und Dares,” in Horst Brunner, ed., Die deutsche Trojaliteratur des Mittelalters und der frühen Neuzeit (Wiesbaden: Ludwig Reichert Verlag, 1990), 494–507. Meyer, Egbert. Der historische Gehalt der Aiyàm al-'Arab Wiesbaden: Otto Harrassowitz, 1970. Milazzo, Antonio M. “Achille e Polissena in Ditti Cretese: un romanzo nel romanzo?,” Le forme e la storia 5 (1984), 3–24. Mourad, Suleiman. “On early Islamic historiography: Abù Ismà'ìl al-Azdì and his Futù˙ al-Shàm,” Journal of the American Oriental Society 120 (2000), 577–93. Muranyi, Miklos. “Ein neuer Bericht über die Wahl des ersten Kalifen Abù Bakr,” Arabica 25 (1978), 233–60. Musil, Alois. Manners and customs of the Rwala Bedouins. New York: American Geographical Society, 1928. von M≥ik, Hans. “Die Gideon-Saul-Legende und die Überlieferung der Schlacht bei Badr. Ein Beitrag zur ältesten Geschichte des Islàm,” Wiener Zeitschrift für die Kunde des Morgenlandes 29 (1915), 371–83. Nallino, Carlo Alfonso. Raccolta di scritti editi e inediti, ed. Maria Nallino. 6 vols. Rome: Istituto per l’Oriente, 1939–48. Newby, Gordon Darnell. A history of the Jews of Arabia: From ancient times to their eclipse under Islam. Columbia, S.C.: University of South Carolina Press, 1988. id., “The Sìrah as a source for early Arabian Jewish history,” Jerusalem Studies in Arabic and Islam 7 (1986), 121–38. Nöldeke, Theodor. “Die Gedichte der Juden in Arabien,” in his Beiträge zur Kenntniss der Poesie der alten Araber (Hannover: C. Rümpler, 1864), 52–86. id., Geschichte der Perser und Araber zur Zeit der Sasaniden. Leiden: E.J. Brill, 1879. Noth, Albrecht. The early Arabic historical tradition: A source-critical study, 2nd ed. in collaboration with Lawrence I. Conrad, trans. Michael Bonner (Princeton: Darwin Press, 1994. id., “Ißfa˙àn—Nihàwand. Eine quellenkritische Studie zur frühislamischen Historiographie,” Zeitschrift der Deutschen Morgenländischen Gesellschaft 118 (1968), 274–96. Owen, E.T. The story of the Iliad. London: Bristol Classical Press, 1989. Parry, Milman. L’épithète traditionnelle dans Homère: essai sur un problème de style homérique. Paris: Société d’éditions “Les Belles Lettres,” 1928. id., “Studies in the epic technique of oral verse-making,” Harvard Studies in Classical Philology 41 (1930), 90–117. Paschal, George Washington. A Study of Quintus of Smyrna. Chicago: University of Chicago Press, 1904. Patzig, Edwin. “Achills tragisches Schicksal bei Dictys und den Byzantinern,” Byzantinische Zeitschrift 25 (1925), 1–18, 273–91. Reinert, Werner. Das Recht in der altarabischen Poesie. Köln: Universität Köln, 1963. Reinhardt, Karl. Die Ilias und ihr Dichter. Ed. Uvo Hölscher. Göttingen: Vandenhoeck und Ruprecht, 1961.

424

lawrence i. conrad

Robinson, Chase F. “The study of Islamic historiography: A progress report,” Journal of the Royal Asiatic Society, Third Series, 7 (1997), 199–227. Rosenthal, Erwin I.J. Judaism and Islam. London: Yoseloff, 1961. Rosenthal, Franz. “The Influence of Biblical tradition on Muslim historiography,” in Bernard Lewis and P.M. Holt, eds., Historians of the Middle East (London: Oxford University Press, 1962), 35–45. Rubin, Uri. Between Bible and Qur"àn: The Children of Israel and the Islamic self-image. Princeton: Darwin Press, 1999. id., The Eye of the Beholder: The life of Mu˙ammad as viewed by the early Muslims, a textual analysis. Princeton: Darwin Press, 1995. id., “Apocalypse and authority in Islamic tradition: The emergence of the Twelve Leaders,” Al-Qantara 18 (1997), 11–41. Sachedina, Abdulaziz. “Early Muslim traditionists and their familiarity with Jewish sources,” in William M. Brinner and Steven D. Ricks, eds., Studies in Islamic and Judaic Traditions II (Atlanta: Scholars Press, 1989), 49–59. Schacht, Joseph. “Une citation de l’Evangile de St. Jean dans la Sìra d’Ibn Is˙àq,” Al-Andalus 16 (1951), 489–90. id., An introduction to Islamic law. Oxford: Clarendon Press, 1964. Schadewaldt, Wolfgang. Von Homers Welt und Werk. Aufsätze zur homerischen Frage. 2nd ed. Stuttgart: K.F. Koehler Verlag, 1951. Schedl, Claus. Muhammad und Jesus. Die christologisch-relevanten Texte des Korans. Vienna: Herder, 1978. Schetter, Willy. “Beobachtungen zum Dares Latinus,” Hermes 116 (1988), 94–109. id., “Dares und Dracontius über die Vorgeschichte des trojanischen Krieges,” Hermes 115 (1987), 211–31. Schöller, Marco. Exegetisches Denken und Prophetenbiographie. Eine quellenkritische Analyse der Sìra-Überlieferung zu Mu˙ammads Konflict mit den Juden. Wiesbaden: Otto Harrassowitz, 1998. Schultze, Ernst. “Die Zahlensysteme der homerischen Griechen,” Archeion 19 (1937), 179–91. Schwarzbaum, Haim. Biblical and extra-Biblical legends in Islamic folk-literature. WalldorfHessen: Verlag für Orientkunde Dr. H. Voradran, 1982. Seeck, Gustav Adolf. “Dion Chrysostomos als Homerkritiker (or. 11),” Rheinisches Museum für Philologie, Neue Folge, 133 (1990), 97–107. Severyns, Albert. Le cycle épique dans l’école d’Aristarque. Paris: E. Champion, 1928. Sezgin, Ursula. Abù Mi¢naf. Ein Beitrag zur Historiographie der umaiyadischen Zeit. Leiden: E.J. Brill, 1971. Shaban, M.A. The 'Abbàsid Revolution. Cambridge: Cambridge University Press, 1970. id., Islamic History A.D. 600–750 (A.H. 132): A new interpretation. Cambridge: Cambridge University Press, 1971. Simpson, R. Hope, and J.F. Lazenby. The catalogue of the ships in Homer’s Iliad. Oxford: Clarendon Press, 1970. Speyer, Heinrich. Die biblischen Erzählungen im Qoran. Gräfenhainchen: Schultze, 1931. Speyer, Wolfgang. Die literarische Fälschung im heidnischen und christlichen Altertum. Ein Versuch ihrer Deutung. Munich: Beck, 1971. Steinschneider, Moritz. Die arabische Literatur der Juden. Frankfurt: J. Kauffmann, 1902. Stetter, E. Topoi und Schemata im ÓadìΔ. Tübingen: Orientalisches Seminar der Universität Tübingen, 1965. Strohmeier, Gotthard. “Homer in Bagdad,” Byzantinoslavica 41 (1980), 196–200. al-ˇaràbìshì, Mu†à'. “Ruwàt al-maghàzì wa-l-siyar 'an Mu˙ammad ibn Is˙àq,” Majallat al-majma' al-'ilmì al-'arabì bi-Dimashq 56 (1981), 533–609. id., Ruwat Mu˙ammad ibn Is˙aq ibn Yasàr fì l-maghàzì wa-l-siyar wa-sà"ir al-marwiyàt. Beirut and Damascus: Dàr al-fikr al-mu'àßir, 1414/1994.

the

MAWÀLÌ

and early arabic historiography

425

Torrey, Charles C. The Jewish foundation of Islam. New York: Jewish Institute of Religion Press, 1933. Venini, Paolo. “Ditti Cretese e Omero,” Istituto Lombardo di scienze e lettere, Classe di lettere e scienze, morali e politiche, Memorie 37 (1981), 161–98. Vian, Francis. Recherches sur les Posthomerica de Quintus de Smyrne. Paris: C. Klincksieck, 1959. Visser, Edzard. Homers Katalog der Schiffe. Stuttgart: B.G. Teubner, 1997. Waltz, Pierre. “L’exagération numérique dans l’Iliade et l’Odyssée,” Revue des études homeriques 3 (1933), 3–38. Welcher, Friedrich Gottlieb. Der epische Cyclus; oder Die homerischen Dichter, 2nd ed. 2 vols. Bonn: E. Weber, 1849–65. Wellhausen, Julius. “Prolegomena zur ältesten Geschichte des Islams,” in his Skizzen und Vorarbeiten (Berlin: Georg Reimer, 1884–99), VI, 1–160. Wensinck, A.J. “L’influence juive sur les origines du culte musulmane,” trans. Georges Bousquet in Revue africaine 98 (1954), 84–112.

THE CONTRIBUTION OF MAWÀLÌ AND THE ARABIC LINGUISTIC TRADITION Monique Bernards

Introduction The field of Arabic linguistics traditionally belongs to the Islamic religious sciences. Indeed, it is obligatory for every 'àlim to have a basic knowledge of Arabic grammar in order to properly interpret the Qur"àn, ˙adìth and fiqh. The Arabic linguistic tradition is anchored in the care needed to preserve the language of the Holy Book, Arabic, God’s chosen language. Qur"ànic linguistic studies started early with the collection of the text but linguistics proper is like fiqh a rather late development in the Islamic sciences, at least at the group level.1 Central to the study of the history of Arabic linguistics are two issues. (1) Possible external influences on the emergence and early development of the Arabic grammatical paradigm and (2) the existence of two competing schools of grammar, the so-called Basra/Kufa dichotomy. Most studies of these two key issues are based on linguistic data; these studies revolve around trying to find similarities in grammatical theory between Greek and Arabic grammar or searching for differences in method and technical vocabulary between Basrans and Kufans.2

1 The first real full-fledged linguistic work is, as far as we now know, the famous Kitàb written by the Persian grammarian Sìbawayh but this relatively early work remains an exception. 2 Regarding the possibility of Greek influence on Arabic grammar, first suggested by A. Merx at the end of the nineteenth century, see amongst others H. Fleisch, “Esquisse d’un historique de la grammaire Arabe,” Arabica 4 (1957), 1–22; later more elaborately studied by C. Versteegh, Greek elements in Arabic linguistic thinking (Leiden: E.J. Brill, 1977). M. Carter, “Les origines de la grammaire Arabe,” Revue des Etudes Islamiques 40 (1972), 69–97, on the other hand sees Arabic grammar as an inherent Islamic science that developed in close relation with Islamic law. See C. Versteegh, Arabic grammar and Qur"ànic exegesis in early Islam (Leiden: E.J. Brill, 1993), 20–36 for an overview of the divergent theses on the subject. For a summary of the different ways the Basra/Kufa dichotomy is discussed see M. Bernards, Changing traditions. Al-Mubarrad’s refutation of Sìbawayh and the subsequent reception of the Kitàb (Leiden: Brill, 1997), 93–95.

MAWÀLÌ

and the arabic linguistic tradition

427

This article takes a different approach than the ones just mentioned. Its angle is based on social rather than theoretical data and it focuses on the Arabic linguists as a social group of the first four centuries of Islam. More specifically, it centers on the role of the non-Arab linguists (mostly mawàlì ) versus that of the Arab linguists in the early development of the Arabic linguistic tradition.3 Several aspects of profiles of non-Arab and Arab linguists will be highlighted in what follows. After a discussion of the way I identified and selected the linguists for this study with a general overview of the ethnic background of this entire group, the origin of the Arabic linguistic tradition will be discussed based on a study of non-Arab versus Arab activity up to 200 AH/815–6 CE. Subsequently, a study of the formation of academic circles provides insight into the non-Arab contribution to the various centers of grammatical learning, which emerged during the four hundred years of Islam with special attention for the alleged Basran/Kufan dichotomy. Academic standing will then be discussed through a study of non-Arab versus Arab activity in specific fields of endeavor—both in and outside their own field of Arabic linguistics—together with the extent of their teacher and pupil networks. Finally, an examination of the occupations linguists had alongside their academic involvement will shed light on possible differences between non-Arabs and Arabs in society—at least within the confines of the group of 'ulamà". A summary of the findings closes the article.

Selection of linguists Inasmuch as the period covered by the Ulama Project4—the first four centuries of Islam—includes the inception, evolution and crystallization of Arabic linguistics as a religious science, it was not selfevident who exactly was to be considered a linguist. The main fields of linguistics are na˙w, “grammar,” and lugha, “lexicography,” and linguists are generally referred to as na˙wiyyùn or lughawiyyùn.5 A number 3 For the purposes of the article, a strict distinction must be made between Arabic linguistics and Arab/non-Arab linguists. The former (Arabic) refers to the language, the latter to the Arabs as a people. 4 The project was carried out by John Nawas and myself from 1994 to 2000, financed by a grant through the Dutch government. See the article by John Nawas for more details. 5 Lughawiyyùn were the “field workers,” as it were, who observed and heard from

428

monique bernards

of linguists were, however, active in both fields, and many were additionally involved in one or the other linguistic sub-specialization like ßarf, “morpho-phonology,” or gharìb, “foreign and/or strange vocabulary.” In identifying linguists to be entered in the database, I have relied on the specialized biographical dictionaries of linguists—a genre that dates back to the middle of the third/ninth century when the number of Arabic linguists and the bulk of linguistic material had expanded considerably. Since one criterion for inclusion in the database was that an individual died before or in the Islamic year 400/1009–10 CE, biographical dictionaries composed before that year were used first.6 Data on all linguists mentioned in these sources were entered and functioned as the basic collection. Later biographical works were scrutinized entry-by-entry to see which scholar had died before 400 AH.7 Not only explicit years of death were taken into account; whenever clear evidence was encountered that a scholar had died within the period covered by our research design, that individual was entered into the database. If any doubt remained, the scholar at hand was not entered into the database. Another way through which many “native speakers” and who gathered the material that was subsequently “translated” into rules and theories by na˙wiyyùn (cf. A. Hadj-Salah, “Lugha,” art. EI 2, V, 803–6). 6 These dictionaries are al-Tirmidhì, Abù Óàmid A˙mad b. Mu˙ammad (d. ca. 250/864), Maràtib al-na˙wiyyìn, ed. Hàshim al-Ta''àn, al-Mawrid 3/2 (1974), 137–144; Abù l-ˇayyib al-Lughawì, 'Abd al-Wà˙id b. 'Alì (d. 351/962), Maràtib al-na˙wiyyìn, ed. Mu˙ammad Abù l-Fa∂l Ibràhìm (Cairo: Maktabat Nah∂a, 1955); al-Sìràfì, Abù Sa'ìd al-Óasan b. 'Abdallàh (d. 368/979), Akhbàr al-na˙wiyyìna l-baßriyyìn, ed. Fritz Krenkow (Paris: Paul Geuthner/Beirut: Imprimerie Catholique, 1936); al-Zubaydì, Abù Bakr Mu˙ammad b. al-Óasan (d. 379/989), ˇabaqàt al-na˙wiyyìna wa-l-lughawiyyìn, ed. Mu˙ammad Abù l-Fa∂l Ibràhìm (Cairo: Dàr al-Ma'àrif, 1973); and Ibn alNadìm, Abù l-Faraj Mu˙ammad b. Is˙àq (d. ca 385/995), Kitàb al-fihrist (Cairo: alMa†ba'a l-Ra˙màniyya, 1929; specific chapter on grammar); al-Tanùkhì, Abù l-Ma˙àsin al-Mufa∂∂al b. Mu˙ammad (d. 442/10150), Tàrìkh 'ulamà" al-na˙wiyyìna min al-baßriyyìna wa-l-kùfiyyìna wa-ghayrihim, ed. 'Abd al-Fattà˙ Mu˙ammad al-Óulw (Riyadh: Dàr al-Hilàl, 1981), was used in this initial phase as well because it is in reversed chronological order and as such linguists from before 400 AH are easily identified. 7 These later works are Ibn al-Anbàrì, Abù l-Barakàt 'Abd al-Ra˙màn b. Mu˙ammad (d. 577/1181), Nuzhat al-allibà" fì †abaqàt al-udabà", ed. 'A†ìya 'Àmir (Stockholm: Almqvist and Wiksell, 1963); al-Qif†ì, Jamàl al-Dìn Abù l-Óasan 'Alì b. Yùsuf (d. 646/1248), Inbàh al-ruwàt 'alà anbàh al-nu˙àt, ed. Mu˙ammad Abù l-Fa∂l Ibràhìm, 4 vols. (Cairo: Dàr al-Fikr/Beirut: Mu"assasat al-Kutub, 1986); al-Fìrùzàbàdì, Majid al-Dìn Mu˙ammad b. Ya'qùb (d. 817/1447), al-Bulgha fì taràjim a"immat al-na˙w wal-lugha, ed. Mu˙ammad al-Maßrì (Kuwait: Manshùràt Markaz al-Makh†ù†àt wa-lTùràth, 1987); and al-Suyù†ì, Jalàl al-Dìn 'Abd al-Ra˙màn (d. 911/1505), Bughyat al-wu'àt fì †abaqàt al-lughawiyyìna wa-l-nu˙àt, 2 vols. (Cairo: Dàr al-Fikr, 1979).

MAWÀLÌ

and the arabic linguistic tradition

429

scholars were included in the collection was via networks of teachers and pupils. Listings of teachers and pupils are not often restricted to the field of linguistics, however. To make sure that the listed individuals were actually involved in linguistic activities, we checked their names in the index of al-Suyù†ì’s Bughya and/or al-Qif†ì’s Inbàh— our two largest sources in this area and as such the most likely to provide reliable results.8 This twofold process of selecting linguists of the first four Islamic centuries has resulted in a database that totals 695 individuals.9 We sought data for all linguists in more than eighty extant classical Arabic biographical dictionaries, ranging from the earliest (Ibn Sa'd) to the late compilation by Ibn al-'Imàd. In all, approximately 100 variables could be entered per individual. To facilitate the process of entering information in the database, we classified the variables into five main categories: (1) Onomastics; (2) Vital Statistics; (3) Occupation and Intellectual Specializations; (4) Geography, and (5) Miscellaneous.10 The variable of “ethnic background” of an individual scholar (classified under Vital Statistics) is, of course, central to this article. It should be noted that it was our general policy to enter only information that was explicitly mentioned in the sources. We noted if someone was of Arab descent when this was made explicit (as, for instance, min banì fulàn, min anfusihim), a mawlà, or a nonArab without an explicitly mentioned mawlà-status. When a mawlà, ethnic information on his patron could be entered; information on the non-Arab ethnicity of both the scholar at hand and his patron and/or family was stored as well. Because this article focuses on the role of “non-Arabs” vis-à-vis Arabs in the emergence and development of the Arabic linguistic tradition, in what follows no distinction is made between mawàlì and non-Arabs with no explicit mawlà-status.

8 During the data-entering phase, a few linguists were found by sheer accident in one of the other sources used but in the end it appeared that all were listed in at least one of the specialized linguistic biographical dictionaries. 9 This database includes an additional nine linguists who entered through a teacher’s network but who themselves died after 400/1009–10. Moreover, some doubt still remains about the exact identity of some thirteen individuals listed by the biographers. 10 The database has a “flexible” part for entering specific information on networks of teachers and pupils as well as lines of transmission of linguistic works.

monique bernards

430

General overview Table 1 displays the ethnic distribution of the entire group of 695 linguists for the first four Islamic centuries; table 1a includes all individuals, table 1b shows only Arab and non-Arab linguists. Table 1a: General distribution of Arabic linguists (Arab, non-Arab, and Unknown Ethnicity; N = 695) Arab Non-Arab Unknown

61 118 516

9% 17% 74%

Total

695

100%

Table 1b: General distribution of Arab and non-Arab linguists (N = 179) Arab Non-Arab

61 118

34% 66%

Total

179

100%

The largest contribution—quantitatively speaking of course—to the field of Arabic linguistics is that of scholars whose ethnic background is “unknown,” i.e., the sources do not give us explicit information on their ethnicity (nearly three-quarters of the entire group). The remaining group of 179 linguists for whom ethnicity is explicitly given, consists of 61 Arab linguists and 118 of non-Arab descent or a 34%–66% distribution. Table 2 presents the ethnic background of the entire group again but now across time and for those who have an explicit death year. The period of 400 years was divided into five eighty-year generations since eighty lunar years is the approximate average lifespan of the group of linguists whose birth and death dates are given by the sources.11 11 In the grammarian database, the age of 150 individuals (some 20%), irrespective of ethnicity, is either explicitly mentioned in the sources or can be calculated on the basis of given birth and death years. These 150 individuals have an average age of 78; 22 is the minimum age and 112 the maximum age mentioned by the sources.

MAWÀLÌ

and the arabic linguistic tradition

431

Table 2: Distribution of ethnicity over five 80-year generations (N = 471) Generation

Arab

–80 AH 81–160 161–240 241–320 321–400

1 5 15 12 9

100% 25% 21% 9% 4%

Total

42

9%

Non-Arab

Unknown

Total

14 34 28 26

70% 47% 20% 11%

1 24 102 200

1 5% 33% 72% 85%

100% 20 73 142 235

100% 100% 100% 100%

102

22%

327

69%

471

100%

The first eighty-year period has only one “linguist,” the legendary founder of the field of linguistics, Abù l-Aswad al-Du"alì, who was an Arab. In order to do more justice to the general picture, I left out the first eighty years in figure 1 that graphs the exact same percentages of table 2. Figure 1: Distribution of ethnicity over five 80-year generations (N = 470) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 81–160

161–240 Arab

241–320

Non-Arab

321–400

Unknown

Figure 1 clearly makes visual that non-Arabs were the immediate leaders in the field of linguistics. Until 240 AH (the end of the third period) their contribution is evidently larger than that of the other two groups, the Arabs and the linguists whose ethnic background is

monique bernards

432

not known. Excluding Abù l-Aswad, the share of the Arabs over time did not exceed 25% whereas that of the non-Arabs commences with 70% and decreases over time to 11%. In particular, the fourth period (240–320 AH) witnesses a clear decline in the role of Arabs and non-Arabs in linguistics. The field is gradually carried forth by the increasing group who carried no ethnic attribute, the unknowns as we have called them—illustrating a large-scale islamization of society, a process that apparently reaches its apex by the end of the fifth period, towards the year 400/1009–10.

Origin of linguistic tradition To investigate more closely the role of non-Arabs in the origin of Arabic linguistics, we will now focus our attention on the first two Islamic centuries that, for purposes of analysis, I have divided into forty-year periods. In total, forty-nine linguists were active in this period and the distribution of their ethnicity over time is shown in table 3. Table 3: Distribution of ethnicity over four forty-year generations up to 200 AH (N = 49) Generation 41–80 81–120 121–160 161–200 Total

Arab

Non-Arab

1 100% 2 40% 3 20% 7 25% 13

27%

Unknown

Total 100% 100% 100% 100% 100%

3 11 16

60% 73% 57%

1 5

7% 18%

1 5 15 28

30

61%

6

12%

49

There are no linguists at all who died before 41/660; the one Arab linguist who died between 41/660–80/700 is Abù l-Aswad al-Du"alì. Figure 2 graphs these percentages, again leaving out Abù l-Aswad al-Du"alì’s period for the sake of clarity. The graph displays the period prior to the emergence of “Muslim scholars” whose ethnic background is not made explicit (they become the majority after 200 AH). That the majority of linguists in this period of emergence and earliest development of the Arabic linguistic tradition consists of non-Arabs, amounting to 60–73%, is even more

MAWÀLÌ

and the arabic linguistic tradition

433

Figure 2: Distribution of ethnicity over four forty-year generations up to 200 AH (N = 48) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 121–160

81–160 Arab

Non-Arab

161–200 Unknown

obvious in figure 2 than it was in figure 1. The initial Arab contribution of 40% (in the period 81–120) decreases to 20–25% within forty years and after that, scholars without explicit ethnic background start to really manifest. The issue of non-Arab contribution to, and possible foreign influence on, the development of Arabic linguistics still arouses lively debates amongst students of the history of the Arabic grammatical tradition. The most straightforward indicator of foreign influence is the percentage of non-Arabs versus Arabs active in the field. In the foregoing we have seen that the percentage of non-Arabs is high indeed and that they have taken the lead from the very beginning of the linguistic tradition. This is partly explained by the fact that Arabic linguistics as a science emerged relatively late—before 200 AH only forty-nine scholars were active in the field and this group witnessed a nine-fold increase in the next two centuries. Compared to other fields of Islamic sciences, however, non-Arab contribution to linguistics is substantially higher and this aspect still awaits satisfying clarification. Elsewhere I have argued that a more telling way of determining the role of non-Arabs in linguistics and the possible influence they

monique bernards

434

exerted on the development of the tradition is by examining generational distance.12 I reasoned as follows. Compilers of biographical dictionaries listed the names of scholars together with the names of their fathers, grandfathers and so on, going back presumably as far as they were able to find information. This listing of names of an individual usually started at the moment in time when either the first family member or his father had converted to Islam. Scholars with up to two generations in their names are relative newcomers and these first and second-generation converts still carried some intellectual baggage from their original culture such as to possibly influence their scholarly output. Latecomers—those whose genealogical listing exceeds two generations—on the other hand are presumed to have become integrated and in no position to reflect elements of a culture that lies far behind them. I intend to follow the same track here but now with the entire group. Relevant information is presented in the following three tables. Table 4a includes all 471 linguists with an explicit death year of in or before 400 AH; table 4b is limited to the forty-nine linguists who were active in the first period up to 200 AH while table 4c contains information about the remaining 422 linguists of the later period up to 400 AH. Table 4a: Number of generations in names of linguists for the entire period (N = 471) Number of Generations

Arab

Non-Arab

Up to two Three or more

9 33

Total

42 100%

21% 79%

57 45

56% 44%

102 100%

Unknown

Total

199 128

265 206

61% 39%

327 100%

56% 44%

471 100%

12 M. Bernards, “Social data as indicators for cultural diffusion: Arabic grammar as a case study,” in S. Leder, H. Kilpatrick, B. Martel-Thoumian and H. Schönig, eds., Studies in Arabic and Islam. Proceedings of the 19th Congress, Union Européenne des Arabisants et Islamisants, Halle 1998 (Leuven: Uitgeverij Peeters, 2002), 357–66.

MAWÀLÌ

and the arabic linguistic tradition

435

Table 4b: Number of generations in names of linguists up to 200 AH (N = 49) Number of Generations Up to two Three or more Total

Arab

Non-Arab

Unknown

Total

31% 69%

26 4

87% 13%

5 1

83% 17%

35 14

71% 29%

13 100%

30

100%

6

100%

49

100%

4 9

Table 4c: Number of generations in names of linguists 200–400 AH (N = 422) Number of Generations

Arab

Non-Arab

Unknown

Total

Up to two Three or more

5 24

17% 83%

31 41

43% 57%

194 127

230 192

Total

29 100%

72

100%

60% 40%

321 100%

55% 46%

422 100%

Tables 4a and 4c show more or less the same trend. As expected, Arabs clearly have more generations listed in their names than nonArabs—an approximate 20% of the Arabs go back less than three generations whereas for the non-Arabs this percentage lies around 50%. If we consider table 4b in the light of this observation, we see that in the early period only 13% of the non-Arabs had more than three generations listed in their genealogies. With the passage of time, however, they became an integral part of Islamic society. This can be read from the increase for the percentage of non-Arabs with three or more generations listed. Like the scholars whose ethnic background is not known to us, the percentage was considerably higher in the later period, rising from 13% to 57% for the non-Arabs and from 17% to 40% for the unknowns, thereby coming close to the 45% of the overall distribution reflected in table 4a. My preliminary conclusion of the original study of generational distance as an indicator for foreign influence calls for some modification here. That tentative conclusion was based on the finding that linguists tended to have longer genealogical listings in their name than other 'ulamà" in general—for whom the overall distribution is 41%

436

monique bernards

up to two generations and 59% with three or more.13 Now that we have a more complete picture of the linguists’ distribution, it appears that 56% has up to two generations listed in the name and 44% three or more. Moreover, compared with the overall distribution (41/59% of the 'ulamà" in general) just given, the 87% non-Arabs active as linguists in the first two centuries and having up to two generations in their name is exceptionally high. If we take a generational distance of two or less generations as a sound gauge to determine the possibility of foreign influence, we are bound to conclude that non-Arabs did exert considerable (and possibly foreign) influence on the early development of Arabic linguistics. Before we end our discussion of the non-Arab contribution to the origin and early development of Arabic linguistics, let us examine more closely the ethnic background of twelve individuals who have been exceptionally important to that tradition in that the sources accredit them with an awwalu man and as such can be seen as pioneers in the field of linguistics.14 Not surprisingly, the first pioneer is the Arab Abù l-Aswad al-Du"alì and the six awà"il that are ascribed to him together account for the invention of every main aspect of Arabic grammar. Additionally, Abù l-Aswad sparks off the beginnings of Arabic linguistic activities in Basra, which later became one of two competing centers of grammatical learning, the other being Kufa (or so we are told). Of the remaining eleven pioneers, three are Arabs, seven non-Arabs and the ethnic background of one is not known—thus truthfully reflecting the general distribution of ethnicity in the first two Islamic centuries as we saw earlier when analyzing table 3. This not only underscores the important role non-Arabs have played in the process, but illustrates as well that, with the sole exception of Abù l-Aswad, the biographers of linguists have not intentionally tried to give the linguistic tradition an Arab constitution.

Centers of learning As mentioned earlier, central to the study of the Arabic linguistic tradition is the issue of the existence of two distinct and allegedly 13

Cf. Bernards, “Social data.” Only those who died before 200 AH have been taken into account in what follows. 14

MAWÀLÌ

and the arabic linguistic tradition

437

rival schools of grammar—the school of Basra and the school of Kufa. The Basran tradition reportedly goes back a long while, even as far back as Abù l-Aswad al-Du"alì, but it owes its real fame to great scholars like Sìbawayh, al-Aßma'ì, and al-Akhfash (al-Awsa†), to name but a few of the early Arabic linguists affiliated with that school. At the same time, the Kufan school harbored equally famous scholars like al-Ru"àsì, al-Kisà"ì and al-Farrà". Additionally, some argue that linguistic activities did not start in Basra at all; that center would have been preceded by a Óijàzì school of linguistics.15 At any rate, according to the tradition, these separate schools remained in existence until Baghdad became the center of learning and scholarly activities moved to the capital in the course of the third/ninth century. Whether the distinction between the Basran and the Kufan tradition is marked by methodological or social differences is subject to debate since the beginning of the twentieth century.16 On the basis of my previous research I have concluded that social aspects—living and working in the same area, having the same academic lineage, sharing a network of contacts—rather than methodological aspects define “school” (at least as far as linguistics is concerned).17 That is the main reason why I prefer to speak of “centers of learning” in the following discussion about the contribution of non-Arabs versus Arabs in the formation of academic circles in various geographic areas. The most wide-ranging geographical variable, covering most of a scholar’s adult social and intellectual activities is min ahl so-and-so, meaning that a linguist who is min ahl al-baßra, for instance, is more likely to have Basra as a place of teaching and/or study than as his place of origin. To investigate the emergence and development of the various centers of linguistic learning, I have therefore analyzed

15 Cf. R. Talmon, Eighth-century Iraqi grammar. A critical exploration of pre-flalìlian Arabic linguistics (Winona Lake, Indiana, 2003 [Harvard Semitic Studies 53]). 16 More specifically, the debate started with G. Weil’s introduction to his edition of Ibn al-Anbàrì’s Masà"il al-khilàf, Die grammatischen Streitfragen der Basrer und Kufer (Leiden: E.J. Brill, 1913), and is still going on. Some students of the Arabic linguistic tradition hold the view that the Basra/Kufa dichotomy reflects “historical reality,” while others consider it the product of “back-projection” or retrojection. 17 The term used for “school” in linguistics is madhhab, but its use seems to have developed differently in the context of legal studies. Cf. Bernards, Changing Traditions, 93–97; id., “The delusion of identification: The term madhhab in Arabic grammatical tradition,” in H.L.J. Vanstiphout et al., eds., All those nations . . . Cultural encounters within and with the Near East (Groningen: Styx, 1999, 13–20).

monique bernards

438

the information on geographic affiliation that was entered in the database as min ahl. Inasmuch as the original information entered was how it was found in the sources, some scholars have a city as min ahl while others are affiliated with an area or province (min ahl al-baßra versus min ahl al-'iràq, for instance). Table 5 contains the data based on a re-categorization on the province level, which enables comparisons. Table 5a: General overview of centers of learning of linguists (N=695) No affiliation Iraq Andalus Ifrìqiyà Mißr Khuràsàn Jibàl Shàm

229 190 180 31 22 13 10 5

32.9% 27.4% 25.9% 4.5% 3.2% 1.9% 1.4% 0.7%

Total

695

100%

Óijàz Jazìra Ahwàz Mosul Transoxania Yaman Adhrabijàn

4 4 3 1 1 1 1

0.6% 0.6% 0.4% 0.1% 0.1% 0.1% 0.1%

As one can read from table 5, one-third of our 695 linguists have no explicit geographic affiliation. Iraq is the largest center for those linguists who do, immediately followed by Andalus—together accounting for 370 linguists (53%). All other centers are markedly smaller, some having had only one linguist in the four-hundred years covered by this study. A general overview of the emergence and historical growth of these linguistic centers was achieved when I broke down the information of table 5a into five periods of each eighty years—the five generations we have used before—and reorganized it in such a way that the oldest centers come first. For the sake of clarity, the group of 229 linguists who had no affiliation was left out, but, it is noted in passing, they started to blossom with only one in the second period then gradually increased from 11 in the third, to 45 and 101 in the fourth and fifth period respectively.

MAWÀLÌ

and the arabic linguistic tradition

439

Table 5b: General overview of centers of learning over five generations (N = 313) Center

–80

81–160

161–240

241–320

321–400

Total

Iraq Óijàz Jazìra Yaman Andalus Mißr Ifrìqiyà Khuràsàn Jibàl Shàm Transoxania Ahwàz Mosul

1

14 3 1 1

48 1

42

28

1 1

133 4 3 1 119 17 11 11 6 5 1 1 1

Total

1

134

313

2 7 2 1 1 1 1

19

62

31 7 5 4 3 2 1

97

81 8 5 6 2 2

Again, it is immediately clear that Iraq and Andalus are by far the biggest centers. Now, however, we can see that Andalus came to preponderance at a much later stage in time than Iraq. For the first three generations, up to 240 AH, the field of linguistics is dominated by Iraqi centers of learning. But after that period the growth of the Iraqi contribution first stagnates then wanes and linguistic activity spreads to new centers of learning. The table clearly shows the exponential growth of the Andalusian centers and, to a lesser degree, though nonetheless obvious, an increase of linguistic activity for the smaller centers of Mißr, Ifrìqiyà, Khuràsàn, Jibàl and Shàm. Concentrating now on the non-Arab/Arab contribution to the formation of academic linguistic circles, we can leave out centers that are either too small (Óijàz, Jazìra, Ahwàz, Mosul, Transoxania, Yaman, Adhrabijàn) and/or have been populated mainly by scholars whose ethnicity was not explicitly mentioned by the sources (Ifrìqiyà, Mißr, Khuràsàn, Jibàl, Shàm).18 The two provinces that had

18 Note, however, at this point that Óijàz does not seem to have “made it” as a center of linguistics. Of the linguists of Ifrìqiyà, 94% belong to the group whose ethnicity is not known; for Mißr this percentage is 91%, Khuràsàn 77% and for both Jibàl and Shàm it is 100%. Since we are interested in non-Arab contributions, I have left these out.

monique bernards

440

the most linguistic activity during our period, Iraq and Andalus, remain. To see if ethnic background has played a role in the emergence of the two famous schools of Basra and Kufa, Iraq was divided into “Basra,” “Kufa” and “Iraq” (the last category primarily referring to Baghdad).19 Table 6 displays the Arab and non-Arab contribution to these centers across five generations. Table 6: Distribution of ethnicity over five generations in the three main centers of learning (N = 227) Generation Center

Arab

Non-Arab

Unknown

Up to 80

Basra

1 100%

81–160

Basra Kufa

3 1

60% 11%

2 8

40% 89%

161–240

Basra Kufa Andalus

6 5

24% 24%

15 10 3

60% 48% 43%

4 6 4

241–320

Basra Kufa Andalus

3 2 4

16% 15% 13%

5 2 7

26% 15% 23%

321–400

Basra Kufa Andalus

1 2

25% 2%

1 3 16

9% 75% 20%

Total 1

100%

5 9

100% 100%

16% 28% 57%

25 21 7

100% 100% 100%

11 9 20

58% 70% 65%

19 13 31

100% 100% 100%

10 4 63

91% 100% 78%

11

100%

81

100%

The Arab of the first generation active as a linguist on his own in Basra is, and by now this comes as no surprise, Abù l-Aswad alDu"alì. More telling, however, is the information on the second generation when the distribution of Arabs and non-Arabs in Basra is still dominated by the first group with a 60% Arab representation (the reader will recall the overall distribution of 35% Arab and 65% non-Arab). Kufa, on the other hand, starts off with a bigger grouping—with nine linguists the Kufan school is at that time almost twice as big as its Basran counterpart—that almost exclusively consists of non-Arabs. Note that in this second generation all linguists have an

19 A similar attempt was made for Andalus but without much success because only one big center could be identified, Qur†uba, the others being more or less evenly distributed across the peninsula.

MAWÀLÌ

and the arabic linguistic tradition

441

explicit ethnic background mentioned by the sources and there is as yet no activity in Andalus. The third period witnesses the emergence of Andalus that is dominated by non-Arabs and linguists without any explicit ethnicity. At this same period in time, both Basra and Kufa are represented by non-Arabs and here, too, we see the beginning of the rise of those linguist 'ulamà" who have no explicit ethnic background. Basra as a center of linguistic learning has become slightly larger than Kufa. The Arab/non-Arab representation (without taking the “unknowns” into account) is, in this third period, 28%/72% for Basra and 33%/67% for Kufa—slowly coming to the familiar overall distribution of 35/65%. Subsequently, we observe a gradual decline of Basran activity and a dramatic fall in Kufan activity in the fourth and fifth periods, while the field of linguistics in Andalus almost triples from thirty-one to eighty-one representatives of whom almost 80% have no known ethnic background. With regard to the Arab/non-Arab distribution in Andalus we may conclude that, unlike Basra and Kufa, Andalus did not have any Arab or non-Arab dominance at any time. Although the non-Arabs in Andalus have outnumbered the Arabs from the beginning of linguistic studies in that area ( just like they did in Arabic linguistics in general), the Andalusian center of linguistic learning is mainly based on those whose ethnic identity is not given by the sources.

Academic standing As was stated at the beginning of this article, every 'àlim had to have at least basic knowledge of Arabic grammar to be able to properly interpret the Qur"àn, ˙adìth and fiqh. Arabic linguistics was the most important auxiliary to the Islamic religious sciences. Such a statement could imply an inferior position of linguists vis-à-vis other Islamic religious scholars who saw their own knowledge of Arabic grammar as a necessary tool. Abundant anecdotes, however, tell us of the opposite, presenting eminent linguists as all-round Islamic scholars who were invited to the caliphal court to settle matters of discussion about variegated subjects. Moreover, linguists were the educators of princes, teaching them to read the Qur"àn, instructing them about

442

monique bernards

traditions, enriching them with knowledge on poetry and adab and, additionally, teaching them to know when to speak or be silent and how to properly conduct a conversation.20 Our next focus of attention concerns the academic position of linguists and especially the standing of non-Arab scholars versus Arab 'ulamà". To determine academic standing, I scrutinized the linguists’ scope of interest in scholarly activities, their contribution to specific fields of endeavor both within and outside the field of linguistics as well as the extent of their networks of teachers and pupils. Due to our manner of selecting linguists to be entered in the database (their presence in one of the specialized biographical dictionaries or an explicit mention of linguistic activity) ensured that all scholars had in one way or another at some point in life been involved in linguistics. Alongside linguistics, we additionally entered explicit information on any other intellectual activity. Taken together, these activities provide us with an idea of the span of that person’s intellectual interest; this information is displayed in table 7a. Table 7a: Overview of number of fields of endeavor for entire group N Missing

657 38

Mean Median

4 3

One field Two fields Three fields

Minimum Maximum

1 15

108 111 135

As one can read from table 7a, thirty-eight scholars do not have an explicitly mentioned field of endeavor—we can only infer their involvement in linguistics from their inclusion in one of the specialized biographical dictionaries. For this group, the sources do not tell us exactly what their contribution to linguistics was. Of the 108 scholars who have but one field of endeavor (about 15% of the entire

20

All of these elements have been taken from an anecdote narrating Hàrùn alRashìd’s instructions to the linguist al-A˙mar at the occasion of the latter’s appointment as al-Amìn’s educator (al-Mas'ùdì, Abù l-Óasan 'Alì b. al-Óusayn, Murùj al-dhahab wa-ma'àdin al-jawhar, ed. Yùsuf As'ad Dàghir, 4 vols., Beirut: Dàr al-Andalus, 1983, III, 351).

MAWÀLÌ

and the arabic linguistic tradition

443

group of 695 linguists), seventy were active in either na˙w (forty) or lugha (thirty) (not shown in the table). What the table does show is that linguists could be active in up to fifteen fields of knowledge. The average for the entire group, however, is four fields of intellectual endeavor while half of the group partook in three fields or less (the median). I have used the median of three fields of endeavor to divide the group of linguists into two groups in order to scrutinize more closely the group whose ethnic background is known. Table 7b presents the distribution of Arab and non-Arab linguists over these two groups: those who were interested in up to three fields on the one hand and those who were involved in four or more fields on the other. Table 7b: Number of fields of endeavor of Arab and non-Arab linguists (N = 178)* Arab

Non-Arab

Total

Up to three Four or more

22 38

34% 34%

43 75

66% 66%

65 113

100% 100%

Total

60

34%

118

66%

178

100%

* One Arab has no known interest mentioned explicitly in the sources.

The majority of both Arabs and non-Arabs have a higher average than the entire group (113 out of 178 or 63% have four or more scholarly interests). In other words, no difference whatsoever exists on this point between Arab and non-Arab linguists and this is reflected by the 34/66% distribution within all three rows of table 7b. An examination of non-Arab versus Arab linguists’ contribution to specific fields of endeavor is our next step. Because we entered data exactly as we encountered them in the sources, we ended up with a list of fifty-four different fields of knowledge, ranging from the very broad and general, e.g., adab, to extremely specific and less widespread topics like ashriba (“drinks,” in particular beverages that are forbidden). Before being able to properly analyze the distribution of Arabs and non-Arabs over these intellectual endeavors, a recategorization of the data was called for. I categorized the fifty-four fields into the following ten general fields: lugha, na˙w, ˙adìth, tafsìr, lisàn (including 'arabiyya and faßà˙à, balàgha, khi†àba, tarassul and jadal ), fiqh (including farà"i∂ ), qirà"a (and Qur"àn), tafsìr (and kalàm), and, finally, adab (and shi'r).

monique bernards

444

Table 9 presents the results of this re-categorization for Arabs and non-Arabs. In examining the table, one has to keep in mind that most of the Arab and non-Arab linguists were involved in more than three different fields, hence the total number of 603 in the column before last of the table.21 Table 9a: Distribution of Arab and non-Arab linguists over fields of endeavor na˙w lugha adab lisàn ˙adìth qirà"a fiqh tafsìr

Arab 37 33 40 26 38 16 16 4

34% 35% 39% 37% 36% 28% 32% 29%

Non-Arab 72 66% 61 65% 63 61% 44 63% 67 64% 42 72% 34 68% 10 71%

Total 109 94 103 70 105 58 50 14

100% 100% 100% 100% 100% 100% 100% 100%

total

210

35%

393

603

100%

65%

The table indicates that as far as the group of linguists under scrutiny here (i.e., those whose ethnic background is known) is concerned, adab and ˙adìth were the most popular fields besides linguistics and these are followed by linguistics-related subjects categorized under lisàn. Qirà"a and, more surprisingly (since anecdotal information suggests the opposite) fiqh are no favorites while tafsìr has a score of just fourteen (out of 603, amounting to a mere 2%). With special regard to Arab versus non-Arab scholarly activity, the 35/65% distribution is abundantly present in this table as well, although a few small deviations do manifest themselves: non-Arabs seem to have had slightly more interest in the study of qirà"a and tafsìr than one would expect based on the overall distribution of Arabs and non-Arabs.22 To determine the extent of their networks and to see whether Arabs and non-Arabs differ in this respect, data are presented in tables 8a, b, c.

21 Stated differently, the number of 603 is the total number of specializations rather than the number of individuals involved. 22 A percentage of 82% of the non-Arab linguists writing books on qirà"a lends support to this finding.

MAWÀLÌ

and the arabic linguistic tradition

445

Table 8a: Overview of number of teachers and pupils for entire group Teachers N Missing Mean Median Minimum Maximum

Pupils

493 202 6 3 1 122

338 357 7 3 1 98

Table 8b: Number of teachers of Arab and non-Arab linguists (N = 150) Arab

Non-Arab

Total

Up to three Four or more

13 32

24% 34%

42 63

76% 66%

55 95

100% 100%

Total

45

30%

105

70%

150

100%

Table 8c: Number of pupils of Arab and non-Arab linguists (N = 125) Arab

Non-Arab

Total

Up to three Four or more

20 24

39% 32%

31 50

61% 68%

51 74

100% 100%

Total

44

35%

81

65%

125

100%

The range of number of teachers for the entire group goes as high as 122, the maximum number of pupils is with 98 slightly less (table 8a). The entire group has an average of six and seven for both the number of teachers and the number of pupils, while the median is three—and that is why the cut-off point of three has been used to divide the group into two (tables 8b and 8c). As with number of fields of endeavor, the majority of Arab and non-Arab linguists have a higher average (than the entire group) with four or more teachers and pupils. Again, no major difference between Arabs and non-Arabs exists regarding the overall distribution for the two groups. However, in the group with up to three teachers, we can note the 76% for the non-Arabs (table 8b, middle column, first row), a percentage that is somewhat larger than the expected 65% encountered in the overall distribution of the entire group.

446

monique bernards

At this stage we can conclude that Arab and non-Arab linguists did not differ in their scholarly activities nor did they in academic standing. We have seen that the two groups are comparable as far as their scope of interests, their contribution to specific fields of endeavor as well as the extent of their networks is concerned. Somewhat more non-Arabs than one would expect (given the overall group distribution) tended to have less than four teachers. The same holds for involvement in the fields of qirà"a and tafsìr where non-Arabs deviated from the overall expectation as well. In absolute figures (and somewhat splitting hairs), however, the difference is four scholars in four-hundred years (forty-two instead of the expected thirty-eight) for qirà"a and one (ten instead of nine) for tafsìr.

Social position Before discussing the findings such that the role of non-Arabs in the emergence and development of the Arabic linguistic tradition can be properly assessed, I would like to investigate one remaining topic: the social position of linguists on the basis of the data we gathered concerning their occupations. Inasmuch as the period of investigation is the first four centuries of Islam, this study covers the period prior to the institutionalization of the Islamic religious sciences in the form of the madrasa. Teaching took place in mosques and private homes and during the pre-madrasa period an established curriculum still did not exist.23 The all-round character of the Arabic linguist, if I may call it so, suggests that many linguists found jobs in education—those, for instance, invited to the court probably having been the very eminent—and in administration. Of the entire group of 695 linguists, 286 had one or more occupations. The following table lists the most popular in order of frequency. For this listing, I have re-categorized the occupations that were the most popular among the linguists. 23 According to C. Versteegh, “A sociological view of the Arab grammatical tradition. Grammarians and their professions,” in Paul Wexler, Alexander Borg and Sasson Somekh, Studia linguistica et orientalia memoriae Haim Blanc dedicata (Wiesbaden: Otto Harrassowitz, 1989, 289–302; a study based on professions mentioned in Suyu†ì’s Bughya), Arabic grammar was never institutionalized in the form of a madrasa or comparable institution and therefore “most grammarians were obliged to seek other sources of income” (p. 295).

MAWÀLÌ

and the arabic linguistic tradition

447

Mu"addib, mutaßaddir, mu'allim, kàtib, and warràq are straightforward, faqìh includes both qà∂ì and muftì, government service ranges from 'alà l-shur†a or al-kharàj to 'alà l-sùq or al-barìd, merchant is tàjir of silk, books, wood, etcetera, and function in mosque is either 'alà l-ßalàt or imàm al-jàmi'; wàlì/wazìr is wàlì or wazìr.24 Table 10a: Overview of the most popular occupations of linguists Mu"addib Faqìh Mutaßaddir Mu'allim Kàtib

97 53 51 32 22

Government Service Merchant Function in mosque Warràq Wàlì/Wazìr

18 14 13 9 8

Table 10a informs us that linguists were indeed involved in jobs revolving around education (mu"addib, mutaßaddir, and mu'allim together score 180), but administration and/or governmental service is less popular than I had initially expected. Let us see if there is a difference between Arabs and non-Arabs in their choices of occupation (table 10b). Table 10b: Distribution of Arab and non-Arab linguists over occupations Arab Mu"addib Faqìh Mutaßaddir Mu'allim Kàtib Government Service Merchant Function in mosque Warràq Wàlì/Wazìr Total

Non-Arab

Total

83% 65% 57% 73% 80% 67% 83% 100% 100% 50%

29 20 7 15 5 3 6 5 1 2

100% 100% 100% 100% 100% 100% 100% 100% 100% 100%

75%

93

100%

5 7 3 4 1 1 1

17% 35% 43% 27% 20% 33% 17%

1

50%

24 13 4 11 4 2 5 5 1 1

23

25%

70

In light of the overall group distribution of 35/65% distribution, used here again as for the purpose of comparison, we observe that 24 Among the less favorite occupations (not listed or included in one of the above categories) we find fallà˙, landowner, mustamli, grammarian (one), poet, sultan, and qàßß.

448

monique bernards

Arabs and non-Arabs did not differ as far as their representation in juridical occupations or in government service is concerned. There are, however, considerably more non-Arabs in the group of mu"addibs, mu'allims and merchants, and in the group of mosque functionaries and warràqs we have not encountered one single Arab. Less obvious, though still clearly present, is an overrepresentation of Arabs in the group of mutaßaddirs and wazìrs.25 Summary and discussion of the findings Two main themes capture the attention of the study of the Arabic linguistic tradition: (1) the issue of possible external or foreign influence on the emergence and early development of the science of Arabic grammar and (2) an alleged existence of two distinct and rival grammatical schools, the so-called Basra/Kufa dichotomy. In this article I dealt with these two issues on the basis of social rather than linguistic data, focusing on the linguists of Arabic as a social group. More specifically, the article centers on the contribution of the nonArab linguists (mostly mawàlì ) versus that of the Arab linguists to the early development of the Arabic linguistic tradition. All data collected pertain to the first four centuries of Islam. This period covers the inception, evolution and crystallization of Arabic linguistics as a science and ends before the institutionalization of the Islamic religious sciences in the form of the madrasa. In identifying linguists from that period, I have relied on the specialized biographical dictionaries of linguists—a specific category that dates back to the middle of the third/ninth century when the number of linguists had increased such that scholars in the field started to compose their own Who’s Who’s. In all, 695 individuals were identified who were, in one way or another, involved in Arabic linguistics in this four-century period. Information on these scholars was sought in more than eighty classical Arabic biographical dictionaries and transferred to a computer database under five main rubrics: (1) Onomastics; (2) Vital Statistics; (3) Occupations and Intellectual Specializations; (4) Geography, and 25 It should be kept in mind, though, that within the entire group of linguists who carried out the occupations listed in table 10 the largest representation is of linguists whose ethnic background is not explicitly mentioned by the sources.

MAWÀLÌ

and the arabic linguistic tradition

449

(5) Miscellaneous. Central for my research was, of course, the variable “ethnic background” of an individual scholar. It should be reiterated here that only information explicitly mentioned in the sources was entered into the database. That is, we considered someone to be of Arab descent only when this was made explicit as in, for instance, min banì fulàn, min anfusihim. And, similarly, only when it was explicitly mentioned that someone was a mawlà or a non-Arab without the mawlà-status, this information was entered exactly as such. Inasmuch as my article focuses on the role of “non-Arabs” vis-à-vis Arabs in the emergence and development of the Arabic linguistic tradition, I made no distinction between mawàlì and non-Arabs without explicit mawlà-status but lumped them together in a category “Non-Arabs.” I investigated and analyzed several aspects of the group of linguists from different perspectives. After this summary, I shall now restrict myself to a short discussion of some of the highlights. First, addressing the issue of possible foreign influence on the early development of the Arabic grammatical paradigm, we have seen that non-Arabs almost immediately took the lead in the field of linguistics. That is, if we exclude the legendary Abù l-Aswad al-Du"alì, who in one of the many versions of the anecdote narrating his creation of Arabic grammar exclaims, “Oh those mawàlì who covet Islam and convert to it. They become our brothers, so shouldn’t we teach them the language?!” Abù l-Aswad reportedly wrote a chapter on al-fà'il wa-l-maf 'ùl and left further developments to others.26 Somewhat paradoxically, “these others” were mostly non-Arabs. We have seen that until the middle of the third/ninth century, the non-Arab contribution to the field exceeds that of the Arabs. Afterwards, activities are taken over by the group whose ethnic background is not mentioned by the sources—the “Muslim linguists” who, in general terms, are responsible for the largest contribution to Arabic linguistics. The non-Arab majority in this particular field may partly be explained by the fact that Arabic linguistics as a science emerged relatively late, when non-Arab neophytes started to outnumber Arab Muslims in all segments of Islamic society. At once, however, we should note that the pious desire to grasp the language that God 26 Cf. Ibn ManΩùr, Abù l-Fa∂l Mu˙ammad b. Mukarram, Mukhtaßar tàrìkh dimashq li-Ibn 'Asàkir, eds. Rù˙iyyat al-Khàs and Mu˙ammad Mu†ì ' al-ÓàfiΩ, 29 vols. (Damascus: Dàr al-Fikr, 1984–91), XI, 226.

450

monique bernards

himself had chosen to reveal the message of Islam must have played a role as well. And by “grasping” I mean much more than just knowing and understanding the language. What I am referring to is the desire of non-Arabs to be on a par with Arabs in their involvement regarding matters of religion. We have also seen that many non-Arab linguists were relatively recent converts to Islam. This was inferred from the listing of generations in names of individual scholars. Compilers of biographical dictionaries listed the names of scholars together with the names of their fathers, grandfathers and so on, going back presumably as far as they were able to find information. This holds especially for the genealogy of Arab linguists. For non-Arab linguists, this listing of names of an individual usually started at the moment in time when either the first family member or that person’s father had converted to Islam. In the above, I considered scholars with up to two generations in their names to be relative newcomers who may still have carried some intellectual baggage from their original culture. As such, it may have been possible that this “original culture” influenced in one way or another their scholarly output. On the other hand, latecomers (defined as those whose genealogical listing exceeds two generations) are presumed to have become integrated and in no position to reflect elements of a culture that clearly laid behind them. On the basis of “generational distance” as a gauge, the possibility of non-Arabs exerting foreign influence on the early development of Arabic linguistics has to be taken seriously into account. However, due to a lack of more detailed information in the sources, a thorough investigation could not be carried out about the exact geographic origin and upbringing of early non-Arab linguists who were able to introduce some of their background training and theoretical skills in the field of Arabic linguistics. This, in turn, only leaves us with the “possibility” of this influence. Secondly, we have the issue of the two so-called schools—the Basra/Kufa dichotomy. We have seen that the Basran tradition reportedly goes back as far as the Arab Abù l-Aswad al-Du"alì, but it owes its real fame to great scholars like Sìbawayh (a mawlà of Persian descent), al-Aßma'ì (an Arab), and the mawlà al-Akhfash alAwsa†, to name but a few of the early Arabic linguists affiliated with this school. At the same time, the Kufan school harbored equally

MAWÀLÌ

and the arabic linguistic tradition

451

famous scholars like al-Ru"àsì (mawlà), al-Kisa"ì (mawlà) and al-Farrà" (mawlà). Whether the distinction between the Basran and the Kufan traditions is marked by methodological or social differences has been the subject of debate since the beginning of the twentieth century. In my view, social aspects—living and working in the same area, having the same academic lineage, sharing a network of contacts— rather than methodological aspects define a “school.” This is why I preferred to speak in this article of “centers of learning.” Intriguing was the finding that the Basran center of linguistic learning had started as an Arab bulwark, whereas Kufa was initially populated by non-Arabs. The more so since the “traditional” distinction between the two schools maintains that Kufan linguists were the stern defenders of linguistic usage based on actual Bedouin Speech and transmitted material (samà' ), whereas the Basrans are described as champions of qiyàs and logical reasoning and, as such, founders of what later became the canonical model of Arabic grammar. So what we have are Kufan non-Arab linguists devoting their life to the preservation of Islam’s Arab heritage—an intriguing thought indeed. However, this distinction is “traditional” and it may well be the result of back-projection or retrojection. Moreover, no data are to be found in my investigation to empirically explain this finding. Hence, I would like to close this article with a few speculative thoughts. As mentioned in the article, every 'àlim had to have at least a basic knowledge of Arabic grammar to be able to properly interpret the Qur"àn, ˙adìth and fiqh. Arabic linguistics was not only the most important auxiliary to Islamic religious science but linguists of Arabic were the educators of princes, teaching them to read the Qur"àn, instructing them about traditions, enriching them with knowledge on poetry and adab and, additionally, teaching them to know when to speak or be silent and how to properly conduct a conversation. Amongst the linguists of Arabic we find many non-Arabs who had converted to Islam. Their practical sense had most probably led them to learn Arabic and some may have had a linguistic training of sorts to help enhance the development of an Arabic grammatical paradigm. But it may have been their pious devotion to Islam that made them dive into the material and protect Islam’s Arab heritage from decay. Seen from this angle, it is mainly thanks to the non-Arabs

452

monique bernards

that the field of Arabic linguistics is, despite its speculative character, traditionally reckoned as one of the Islamic religious sciences.

Bibliography Primary Sources Abù l-ˇayyib al-Lughawì, 'Abd al-Wà˙id b. 'Alì, Maràtib al-na˙wiyyìn, ed. Mu˙ammad Abù l-Fa∂l Ibràhìm, Cairo: Maktabat Nah∂a, 1955. al-Fìrùzàbàdì, Majid al-Dìn Mu˙ammad b. Ya'qùb, al-Bulgha fì taràjim a"immat alna˙w wa-l-lugha, ed. Mu˙ammad al-Maßrì, Kuwait: Manshùràt Markaz al-Makh†ù†àt wa-l-Turàth, 1987. Ibn al-Anbàrì, Abù l-Barakàt 'Abd al-Ra˙màn b. Mu˙ammad, Nuzhat al-allibà" fì †abaqàt al-"udabà", ed. 'A†ìya 'Àmir, Stockholm: Almqvist and Wiksell, 1963. Ibn ManΩùr, Abù l-Fa∂l Mu˙ammad b. Mukarram, Mukhtaßar tàrìkh dimashq li-Ibn 'Asàkir, eds. Rù˙iyyat al-Khàs and Mu˙ammad Mu†ì' al-ÓàfiΩ, 29 vols., Damascus: Dàr al-Fikr, 1984–91. Ibn al-Nadìm, Abù l-Faraj Mu˙ammad b. Is˙àq, Kitàb al-fihrist, Cairo: al-Ma†ba'a l-Ra˙màniyya, 1929. al-Mas'ùdì, Abù l-Óasan 'Alì b. al-Óusayn, Murùj al-dhahab wa-ma'àdin al-jawhar, ed. Yùsuf As'ad Dàghir, 4 vols., Beirut: Dàr al-Andalus, 1983. al-Qif†ì, Jamàl al-Dìn Abù l-Óasan 'Alì b. Yùsuf, Inbàh al-ruwàt 'alà anbàh al-nu˙àt, ed. Mu˙ammad Abù l-Fa∂l Ibràhìm, 4 vols., Cairo: Dàr al-Fikr/Beirut: Mu"assasat al-Kutub, 1986. Sìbawayh, Abù Bishr 'Amr b. 'Uthmàn, al-Kitàb, ed. H. Derenbourg, 2 vols., Paris: Imprimerie Nationale, 1881–89; repr. Hildesheim: G. Olms, 1970. al-Sìràfì, Abù Sa'ìd al-Óasan b. 'Abdallàh, Akhbàr al-na˙wiyyìna l-baßriyyìn, ed. Fritz Krenkow, Paris: Paul Geuthner/Beirut: Imprimerie Catholique, 1936. al-Suyù†ì, Jalàl al-Dìn 'Abd al-Ra˙màn, Bughyat al-wu'àt fì †abaqàt al-lughawiyyìna wal-nu˙àt, 2 vols., Cairo: Dàr al-Fikr, 1979. al-Tanùkhì, Abù l-Ma˙àsin al-Mufa∂∂al b. Mu˙ammad, Tàrìkh 'ulamà" al-na˙wiyyìna min al-baßriyyìna wa-l-kùfiyyìna wa-ghayrihim, ed. 'Abd al-Fattà˙ Mu˙ammad al-Óulw, Riyadh: Dàr al-Hilàl, 1981. al-Tirmidhì, Abù Óàmid A˙mad b. Mu˙ammad, Maràtib al-na˙wiyyìn, ed. Hàshim al-Ta''àn, al-Mawrid 3/2 (1974), 137–144. al-Zubaydì, Abù Bakr Mu˙ammad b. al-Óasan, ˇabaqàt al-na˙wiyyìna wa-l-lughawiyyìn, ed. Mu˙ammad Abù l-Fa∂l Ibràhìm, Cairo: Dàr al-Ma'àrif, 1973. Secondary Sources Bernards, M. Changing traditions. Al-Mubarrad’s refutation of Sìbawayh and the subsequent reception of the Kitàb, Leiden: Brill, 1997. id., “The delusion of identification: The term madhhab in Arabic grammatical tradition,” in H.L.J. Vanstiphout et al., eds., All those nations . . . Cultural encounters within and with the Near East, Groningen: Styx, 1999, 13–20. id., “Social data as indicators for cultural diffusion: Arabic grammar as a case study,” in S. Leder, H. Kilpatrick, B. Martel-Thoumian and H. Schönig, eds., Studies in Arabic and Islam. Proceedings of the 19th Congress, Union Européenne des Arabisants et Islamisants, Halle 1998, Leuven: Uitgeverij Peeters, 2002, 357–66. Carter, M. “Les origines de la grammaire Arabe,” Revue des Etudes Islamiques 40 (1972), 69–97.

MAWÀLÌ

and the arabic linguistic tradition

453

Fleisch, H. “Esquisse d’un historique de la grammaire Arabe,” Arabica 4 (1957), 1–22. Hadj-Salah, A. “Lugha,” in EI 2, V, 803–6. Talmon, R. Eighth-century Iraqi grammar. A critical exploration of pre-flalìlian Arabic linguistics, Winona Lake, Indiana, 2003 [Harvard Semitic Studies 53]. Versteegh, C. Greek elements in Arabic linguistic thinking, Leiden: E.J. Brill, 1977. id., “A sociological view of the Arab grammatical tradition. Grammarians and their professions,” in Paul Wexler, Alexander Borg and Sasson Somekh, Studia linguistica et orientalia memoriae Haim Blanc dedicata, Wiesbaden: Otto Harrassowitz, 1989. id., Arabic grammar and Qur"ànic exegesis in early Islam, Leiden: E.J. Brill, 1993. G. Weil, Die grammatischen Streitfragen der Basrer und Kufer, Leiden: E.J. Brill, 1913.

A PROFILE OF THE MAWÀLÌ 'ULAMÀ" John A. Nawas This article presents a profile of the mawàlì 'ulamà" from the beginning of Islam to the year 240 AH/854–5 CE. The mawàlì will be studied under four rubrics: demography; geography; distinctive attributes and practices; and, intellectual indices. We will start off, however, by describing the general framework and data-gathering procedures used in our Ulama Project—of which the mawàlì are a part—and outlining the rationale for restricting the presentation to the first two and a half centuries.

Framework and procedures The main goal of the Ulama Project was the study of the evolution of 'ilm—as encapsulated in the Islamic religious sciences and those who carried these endeavors out, the 'ulamà"—from its inception until the fifth Islamic century. The Islamic sciences studied comprise ˙adìth, qirà"a, tafsìr, fiqh and na˙w. The sources used for obtaining the information needed were classical Arabic biographical dictionaries covering these five sciences. The material gleaned was systematically transformed into computerized data.1 With commitment to the spirit of the social science approach, our point of departure was entire groups rather than selected individuals who are by definition unique. To allow generalization, the sample drawn was random and the data gathered were coded to form

1 The project was carried out by Monique Bernards and myself from 1994 to 2000, financed by a grant through the Dutch government. Both of us worked in close unison, differentiated only by the track chosen as focus. Monique Bernards addressed the grammarians (practically all grammarians of the first four centuries of Islam, totaling 695 individuals) while I devoted my attention to an overview of all five branches of 'ilm just named. Inasmuch as the biographical dictionaries list many thousands of people, the tactic here was to use a representative probability sample that included more than one thousand 'ulamà" (see below for more information on the composition of the sample).

a profile of the

MAWÀLÌ 'ULAMÀ"

455

a part of the computer program that we wrote for the project. So cast, the data permit us to answer many specific questions without the encroachment of subjectivism and to make countless comparisons across groups, geographic locations and time periods. Our data set encompasses some hundred variables distilled from a much larger pool we had started with then discarded in light of the pilot studies we had conducted. We have used more than eighty biographical dictionaries to draw on.2 In transferring information from the biographical dictionaries to the computer database we strictly adhered to the rule of reading the compiler’s text, not reading into it. In other words, we relied on what the compiler said in explicit terms, not on inferences or guesses as to what he was possibly saying. Especially for this article, it is important to stress that an 'àlim was entered as a mawlà (or an Arab) if, and only if, we are so told explicitly by the compilers. The Ulama Project sample for the five sciences comprises 1,003 'ulamà".3 How many of these, per time period, were mawàlì, Arabs and “unknowns” (i.e., were not explicitly designated as a mawlà or an Arab) is graphed in figure 1. Each of the five periods in the figure represents eighty hijrì years—which coincides with the average life span for the entire group during the first four centuries of Islam.4 2 The biographical dictionaries used range from the very earliest (Ibn Sa'd, Khalìfa b. Khayyà†) to the late work by Ibn al-'Imàd (d. 1089/1678) and include all major †abaqàt works that have been published. In some cases, information about a particular individual is found in only one or two sources, while for others we found ourselves consulting twenty or more. On the average, five to six biographical dictionaries covered all that there is about any particular 'àlim. As a rule, we discontinued the search once we failed to gain any new information about the person after examining in succession entries from three different dictionaries. 3 The actual data set comprises 1,049 but 46 of these 'ulamà" were, by sheer chance, encountered twice. For the entire aggregated group, the duplicates were limited to one occurrence to guard against contamination—hence the figure of 1,003. 4 Pellat, Gilbert, Bulliet and }entürk, though using different approaches, also report an average of close to eighty lunar years for the life span of the 'ulamà". Charles Pellat, “Quelques chiffres sur la vie moyenne d’une catégorie de Musulmans,” in Pierre Salmon, ed., Mélanges d’islamologie: Volume dédié à la mémoire de Armand Abel (Leiden: E.J. Brill, 1974, 233–246); Joan E. Gilbert, “Institutionalization of Muslim scholarship and professionalization of the ‘'ulamà"’ in Medieval Damascus,” Studia Islamica 52 (1980), 105–34; Richard W. Bulliet, “The age structure of medieval Islamic education,” Studia Islamica 57 (1983), 105–17; Recep }entürk, Narrative social structure. Anatomy of the Hadith transmission network, CE 610–1505 (Ph.D. Diss., Columbia University, 1997).

john a. nawas

456

Figure 1: Percentage of Mawàlì, Arabs, and “Unknowns” across the Five Periods (N = 1,003) 100% 87%

90%

85%

80% 69%

70% 60%

49%

50%

42%

40%

36%

33%

30% 18%

20% 10%

4%

22%

22%

9%

9%

10%

5%

0% Up to 80 AH (N = 55)

161–240 AH (N = 267)

81–160 AH (N = 177) Mawlà

Arab

241–320 AH (N = 276)

321–400 AH (N = 228)

Unknown

In the first period (up to 80 AH), the mawàlì constituted but a small portion of society (4%), especially when compared with their Arab counterparts. In the second period, the mawàlì still do not make up the largest part of society but their increase is noticeable indeed and this trend continues up to the third period when they constitute about one-third of the total. After 240/854–5, however, not the mawàlì and Arabs together but the “Unknowns” form the majority of the 'ulamà", probably indicating a degree of islamization that rendered such a designation as mawlà and Arab meaningless.5 Inasmuch as we are interested in a profile of the mawàlì 'ulamà" when this designation did appear to carry meaning, the attention will focus on those who died in or before the year 240/854–5. Although the concern in this article is a description of the 157 mawàlì we have for this period by profiling them, I will—in the interest of 5 To be sure, the term mawlà was still invoked in a legal sense as late as the fifteenth and sixteenth centuries CE though not with regard to 'ulamà". Under pressure from the European colonial powers, the Ottomans used the term to help facilitate the manumission of slaves which, they hoped, would ultimately lead to the gradual abolition of slavery as demanded by some western countries (Halil (nalcık, “Servile labor in the Ottoman Empire,” in his Studies in Ottoman social and economic history, London: Variorum Reprints, 1985, article VII).

a profile of the

MAWÀLÌ 'ULAMÀ"

457

sharpening the focus—compare them now and then with their 194 Arab counterparts who spanned the same period.

Demographic data Demographic data include averaged years of birth, death and life span; former religion; ethnic background of the mawlà as well as his patron; tribal affiliation; and, finally, occupation. In what follows, we will be speaking of group averages. As expected (and we have just seen), the mawàlì entered Islamic society at a later time than did the Arabs; their average years of birth and death were for our group respectively 97 AH and 174 AH—compared with 55 AH and 123 AH for the Arabs. The average life span of a mawlà was 76 which was almost identical with that of their Arab counterparts—an average that is somewhat lower than the 80 years which holds for the entire data set as reported above. Compilers of the biographical dictionaries did not seem to have been interested in the pre-Islamic religion of 'ulamà" or their families. This information was given on a mere 4% of the mawàlì 'ulamà", six of the 157. The combined figures for previous religion of the 'àlim and his family reveals that three were Jews, two Christians and the sixth mawlà an ex-Magian—much too small a sample to allow generalization, let alone one in terms of percentages. A bit more information is available on the ethnic background of the mawàlì. We have information for fifteen of the 157 (close to 10%). Of these, eight were Persian (Fàrisì ). The remaining seven individuals came from nearly as many ethnic groupings, two belonging to the remote eastern provinces, Farghàna (Transoxania) and alRukhkhaj (Sijistàn), while the others were Byzantine (Rùmì ), Armenian, Ethiopian and Kurdish. One mawlà was an Arab, the mawlà of another Arab—a phenomenon that occurred only at the very beginning of Islam. This becomes apparent as we look into the distribution of the ethnicity of the patrons of the mawàlì, a subject to which we now turn. Information on the ethnicity of the patrons (in terms of Arab or mawlà) is found for some 40% of the mawàlì (63 out of 157). One of these is the Arab of whom we have just spoken; this man died in 110/728–9. It is noteworthy that this person had “inherited” the

458

john a. nawas

mawlà status from an ancestor who reportedly had become a mawlà during the days of the Prophet.6 Except for this one Arab who himself was a mawlà, the other mawàlì by far—almost 90%—had Arabs as patrons (55 out of 157) while the remaining 10% had mawàlì as patrons (seven out of 157). Knowing when it became a general trend for a mawlà to become himself a patron is important since it would reflect the timing of an upward movement in the social mobility of the mawàlì. The earliest year of death encountered of the mawàlì who had other mawàlì as their patrons is 145/762–3. The years of death of the second and third such person were 153/770, 170/786–7 and the remaining four cases occurred in years of death beyond 200/815–6. Considering these dates on the one hand and their proximity to the year 132/749–50 in which the 'Abbàsid takeover took place, it could well indeed be true as some have suggested that the 'Abbàsid takeover was in part a reaction to the inequality of nonArab Muslims to Arab Muslims and a means for improving the social status of the former.7 Much more data are available with regard to the tribal affiliation of the mawàlì. As stated earlier, all data were recorded as we encountered them in the sources. Afterwards, using Caskel’s ]amharat anNasab des Ibn al-Kalbì as a reference point, we re-categorized all these into 32 main tribes. The distribution of the mawàlì across these main tribes is presented in table 1; the data have been organized starting with highest frequency of occurrence.

The individual involved is Abàn b. Íàli˙ b. 'Umayr b. 'Ubayd, Abù Bakr alQurashì who, in the words of Ibn 'Asàkir (Abù l-Qàsim 'Alì b. Abì Mu˙ammad, Ta"rìkh Madìnat Dimashq, ed. 'Umar b. Gharàma al-'Amrawì, 80 vols., Beirut: Dàr al-Fikr, 1995, VI, 141), was their mawlà who by origin was an Arab taken prisoner (mawlàhum, aßlu˙u min al-'arab wa-asabahu siba"). However, it was not Abàn himself who had been taken prisoner but one of his ancestors, an event that had taken place during the Prophet’s lifetime. Apparently, after manumission this ancestor had become a mawlà. This instance, then, is an illustration of the “inheritance” of the status of mawlà from one generation to another. 7 P. Crone (Roman, provincial and Islamic law. The origins of the Islamic patronate, Cambridge: Cambridge University Press, 1987, 36) asserts “[t]hroughout the Umayyad period (661–750) all non-Arabs who wished to join the ranks of the conquerors had to find an Arab (or, as the Arab character of Muslim society receded, Muslim) patron, who, upon their declaration of conversion of allegiance, obtained very much the same rights and duties vis-à-vis them as the manumitter vis-à-vis his former slave.” 6

a profile of the

MAWÀLÌ 'ULAMÀ"

459

Table 1: Percentage Distribution of Tribes (N = 149) Quraysh Tamìm Azd Aws/Khazraj Gha†afàn Unknown Asad Khuzà'a Bàhila 'Àmir b. Sa'sa'a 'Abdalqays Kinàna Madh˙ij Thaqìf Ash'ar Kinda Rabì'a Sulaym ˇayy" Bakr b. Wà"il Óimyar Hudhayl Kalb Nahd Total

36.2 8.7 8.7 7.4 4.7 4.7 3.4 2.7 2.7 2.7 2.0 2.0 2.0 2.0 1.3 1.3 1.3 1.3 1.3 0.7 0.7 0.7 0.7 0.7 100

We should first note that it appears to have been important for the chroniclers to indicate tribal affiliation as attested to by the fact that they have provided information on 149 of the 157 mawàlì under study. A scrutiny of the data presented in table 1 shows that Quraysh had the largest share of mawàlì. Other tribes who counted mawàlì amongst their ranks were also well known, the Tamìm and Azd having each about 9% of all mawàlì. The Anßàr (Aws and Khazraj) boasted some 7% of the mawàlì. These tribes account for some 60% of the entire group; the remaining 40% are more or less evenly distributed amongst the other main tribal units as one can see in the table (the category “Unknown” are those tribes that we were not able to identify properly through Caskel’s work). With the categorization of the tribes into 32, the distribution of the Arab 'ulamà" did not differ from that of the mawàlì 'ulamà" (data not shown).

john a. nawas

460

Table 2: Percentage Distribution of Mawàlì and Arabs in Northern and Southern tribes

Mawàlì (N = 149) Arabs (N = 193)

Northern

Southern

Unknown

68.5 53.4

26.9 42.5

4.7 4.2

This, however, is not the case once we re-categorize the tribes (again following Caskel) into the much larger units of Northern versus Southern tribes. The relevant data are found in table 2 where the data for the mawàlì are presented together with those for the Arabs. More than two-thirds of the mawàlì are affiliated with Northern tribes while a little more than one-quarter of them have a Southern tribal affiliation. This distribution is statistically different from that of the Arab 'ulamà" who are more or less evenly divided albeit with a slight edge tilted toward the Northern tribes. I can think of no reasonable explanation for these differences. The last variable listed under the rubric of demographic data concerns occupational pursuits. The results are presented in table 3. Table 3: Percentage Distribution of Occupations for Mawàlì and Arabs Commerce Education Administration Military Legal/Religious

Mawàlì (N = 53) Arabs (N = 70)

18.9 4.3

17.0 1.4

26.4 38.6

3.8 8.6

34.0 47.1

Of the five categories that I have classified in larger groupings of occupation, the mawàlì were primarily active in commerce and education: about 36% of them had an occupation in one of these two domains versus a mere 6% of the Arabs. The Arabs found employ in administration (the government) and legal/religious offices ( judge or mufti, for instance); more than 85% of the Arabs versus 60% of the mawàlì were active in these two spheres. All in all, however, the most compelling difference between mawàlì and Arab 'ulamà" pertains to the preponderance of the mawàlì in commerce and education.

a profile of the

MAWÀLÌ 'ULAMÀ"

461

Geographical data Six different geographical variables have been pursued: areas or places of origin, residence, occupation, teaching, death and min ahl.8 For purposes of analysis, all place names encountered were re-categorized into larger units—province. We adopted as reference point the administrative division of the Islamic empire as it stood during its territorial apogee, that is the empire during the reign of the 'Abbàsid caliph Hàrùn al-Rashìd (d. 193/809). The data have been cast into percentages and displayed in the form of maps. Map 1 shows the percentage distribution of the provinces from where the mawàlì (N = 42) originated. The darker the shade of grey in the map, the larger the percentage of mawàlì who came from that province. Except for the central province of Iraq, where one-third of the mawàlì originally came from, the rest have originated in eastern provinces of the Islamic empire. Some 17% of the mawàlì originated in Khuràsàn and another 12% from Fars. All other provinces listed contribute 10% or less. These findings are entirely consistent with the data on ethnicity that I have described earlier. Quite striking is the difference between the mawàlì and Arabs with regard to Yemen where only 2.5% of the former originated (Map 1) compared with 35% of the latter who originated there (data not displayed). All five other geographical variables have to do with a savant’s adult life. A scrutiny of maps 2 through 6 shows that they do tell the same broad story: now the concentration (the darker areas) is found in the central provinces of the empire rather than the eastern ones. Map 2 displays the percentage distribution of the mawàlì with regard to the provinces where they lived (N = 53). More than half of the mawàlì had their residence in Iraq. This province is followed by the one where the Umayyads had their capital, Shàm. Some 9% of the mawàlì lived there in the course of the period under study as did the same percentage in Óijàz. Then follows the province of Mißr with almost an 8% share.

8 What is being presented here is limited to these six variables, the criterion being used is a sample size of no less than thirty mawàlì. The geographical variables that did not meet this numerical criterion and thus left out are places of birth, growing up, study, where fought, visit and burial, “being there” (the reason why the 'àlim was in that particular location is unclear).

john a. nawas

Map 1. Origin

462

MAWÀLÌ 'ULAMÀ"

463

Map 2. Residence

a profile of the

Map 3. Occupation

464 john a. nawas

MAWÀLÌ 'ULAMÀ"

465

Map 4. Teaching

a profile of the

john a. nawas

Map 5. Death

466

MAWÀLÌ 'ULAMÀ"

467

Map 6. Min Ahl

a profile of the

468

john a. nawas

Approximately the same pattern of distribution is visible in Map 3 on the provinces of occupation (N = 33), which is logical since by and large an occupation is also practiced in one’s place of residence. Here, too, the largest province where mawàlì found employment was Iraq (but now only one-third), followed by Óijàz and then Mißr. Shàm does not occupy the same rank it did with regard to their places of residence. The very same holds, not unexpectedly, for places of teaching (Map 4, N = 48): more than 50% of the mawàlì did so in Iraq, followed by Óijàz (19%) and Mißr again (13%). And yet again, also in line with expectation, for provinces where the mawàlì died (Map 5, N = 78). The largest number of mawàlì died in Iraq followed by Óijàz and, what we have in the meantime become accustomed to, Mißr. The variable min ahl tells the same story and the data on this aspect are displayed in the last map (N = 133). In Map 6 we see a similar concentration of mawàlì in the central provinces, the top three again Iraq, Óijàz and Mißr in this order. But this geographic variable and the first one presented on origin are the only ones in which the mawàlì differed from the Arab 'ulamà" (data not displayed). The differences between them revolve primarily around two provinces: Óijàz and Mißr. For the mawàlì, about one-quarter were min ahl alÓijàz compared with 40% of Arab 'ulamà". An opposite trend is observed in Mißr where 13% of the mawàlì and 5% of Arabs were min ahl that province.

Distinctive attributes and practices This rubric covers eight different aspects of social and personal life: ideological position; awwalu man; classificatory labels (e.g., ßà˙ib sunna); mention in one or more of the canonical Sunnite ˙adìth collections; noted garments; dyeing of the beard; and particular physical features as well as religious observances. Reference to Arab 'ulamà" will only be made for those variables where statistically significant differences exist.

a profile of the

MAWÀLÌ 'ULAMÀ"

469

Table 4: Percentage Distribution of Ideologies (N = 33) Shì'ite Murjì"ite Qadarì Mu'tazilì Màlikì Khàrijì Óanafì Shàfi'ì Óanbalì Jahmì Sunnì 'Uthmànì

27.3 18.2 15.2 9.1 9.1 3.0 3.0 3.0 3.0 3.0 3.0 3.0

The first intellectual variable to be considered has to do with ideologies. Relevant information, presented in table 4, is available on 33 of the 157 mawàlì (one out of five). Of our 'ulamà" who received an ideological ascription, more than a quarter was Shì'ites. The large proportions of Murjì"ites and Qadarites (and Mu'tazilites) come as no surprise given the issues debated at the time. All other remaining ideologies are evenly distributed; none of the four Sunnite madhàhib could have acquired many followers because they were just starting.9 The observation that the Màlikites are found three times more often probably has to due with it being one of the earliest. There were seven instances of awwalu man . . . that fell into three categories. Four of the seven were initiators of a specific field of learning, another introduced a book (to a province or particular region) and the two remaining awà"il were too detailed to allow categorization. Table 5: Percentage Distribution of Classifications (N = 38) Tàbi' Íà˙ib sunna al-Abdàl Majhùl Mu'ammar Ía˙àbì

50.0 13.2 7.9 7.9 2.6 2.6

9 Cf. Monique Bernards and John Nawas, “The geographic distribution of Muslim jurists during the first four centuries AH,” ILS 10 (2003), 168–181.

john a. nawas

470 Table 5 (cont.)

Ahl al-Dìn Íà˙ib Qur"àn Laysa Íà˙ib ˙adìth Íà˙ib na˙w wa-'arabiyya Shu'ùbì al-Qurrà" al-'ashra

2.6 2.6 2.6 2.6 2.6 2.6

The next category covers what can be called “labeling” such as ßa˙àbì, tàbi', or a phrase starting with min. . . . Table 5 presents these classifications that were found for some 25% of the mawàlì (38 out of the 157). The largest category by far is that of the tàbi'ùn since half fall into it. This is followed by the vague label “ßà˙ib sunna”10 while, except for the categories “al-abdàl ”11 and “majhùl,” all others are evenly distributed and quite few in number. In comparing mawàlì with Arabs a significant difference did come to light: the percentage of Arab 'ulamà" who were referred to as ßa˙àba was inordinately higher than for the mawàlì. Inasmuch as the ßa˙àba were almost by definition Arabs, I excluded this category from the statistical analysis that compared the groups on the other eleven labels and found no difference.12 Table 6: Listing in One or More of the Six Sunnite Canonical Hadith Collections in Percentages (N = 134) Six One Two Five Four Three

10

38.8 18.7 17.2 10.4 8.2 6.7

Cf. G.H.A. Juynboll, “Some new ideas on the development of sunna as a technical term in early Islam,” JSAI 10 (1987), 97–118 and his “An excursus on the ahl as-sunna in connection with Van Ess, Theologie und Gesellschaft, vol. IV,” Der Islam 75 (1998), 318–30. 11 Abdàl: One of the degrees in Íùfì hierarchy, unknown to the masses (I. Goldziher/H.J. Kissling, “Abdàl,” art. EI 2, I, 95). 12 Approximately 98% of the ßa˙àba were Arabs, a mere 2% mawàlì. For the †àbi'ùn the distribution is about 70% Arabs and 30% mawàlì.

a profile of the

MAWÀLÌ 'ULAMÀ"

471

Thanks to the large compilations by al-Mizzì and Ibn Óajar, we are able to find out in how many of the six Sunnite canonical collections an individual is listed in the isnàd. The findings on this variable are presented in table 6. The lion’s share of mawàlì found their way into the isnàds of the canonical collections in either all six (almost 40%) or in just one of the six (almost 20%). There is nothing strange or even unexpected about this distribution, the more so since it holds for the Arab 'ulamà" as well (data not presented). The biographical dictionaries make occasional mention of noteworthy clothing worn by an individual. Also on this aspect the data are very sparse; for less than one-half percent of the mawàlì is information to be found (seven out of 157) and we need not present these data in tabular form. Three of the seven were noted for wearing black clothing, the remaining four individuals were evenly distributed over the four categories of qalansuwa, silk and woolen clothes and white garments. There were just as many mawàlì 'ulamà" who dyed their beards as those who did not, six individuals in each case. When compared with their Arab counterparts, however, a significant difference emerged; nearly five times as many (23) Arab 'ulamà" dyed their beards versus five who did not. This is a curious finding for which I have no explanation. Juynboll has suggested that Arabs dyed their beards in imitation of the custom encountered in the territories they had conquered.13 This is an intriguing idea but I do not have the size of data to pursue it on either the dimension of time or geography which would be directly relevant. It is not clear to me, too, if the 'ulamà" as a group were amongst those segments of society who were prone to imitate the conquered. Most puzzling indeed is: why a conqueror imitates the conquered while, if anything, common sense suggests the opposite. Table 7: Percentage Distribution of Noted Physical Feature (N = 34) Blind Forgetfulness/Senility Handsome Kahl A'war

29.4 20.6 8.8 5.9 5.9

13 G.H.A. Juynboll, “Dyeing the hair and beard in early Islam. A Hadith-analytical study,” Arabica 33 (1986), 49–75.

john a. nawas

472 Table 7 (cont.) Ashqar Small face Very tall A˙maq Short Pitch-black A'raj Black A˙wal

5.9 2.9 2.9 2.9 2.9 2.9 2.9 2.9 2.9

Table 7 lists the physical features for the mawàlì. Over one-third of the features had to do with the organs of vision (blind, one-eyed, cross-eyed), which is fitting in describing men of learning in a preindustrial society. The table makes clear, too, that we are dealing with an older elite population14 since the highest ranking characteristics have to do with aging (blindness, senility, kahl; over 50% were so described).15 Table 8: Noted Religious Duties (counts) Reading Quran Íalàt Óajj Íawm Good deed

6 3 3 1 1

The final item pertains to unusual or atypical religious observances in the sense that these were carried out beyond the call of duty. This, as shown in table 8, was noted for fourteen mawàlì (about 9%). The most noticeable feature of the table is that six of the fourteen 14 The 'ulamà" who had made a name for themselves and hence were included in biographical dictionaries constituted as such an elite with seniority. Cf. the point made by J. van Ess when discussing why Ibn Óanbal’s only other colleague who refused to submit to the 'Abbàsid caliph al-Ma"mùn’s doctrinal demands during the mi˙na, Mu˙ammad b. Nù˙ al-'Ijlì, had received so little attention despite his heroic stand: he simply died too young to have earned mention as an 'àlim by the compilers of biographical dictionaries (Theologie und Gesellschaft, III, 458). 15 The Arab 'ulamà" did differ significantly on physical features but the difference is due to one qualification that Arab 'ulamà" received while none of the mawàlì did: six of the 45 Arabs on whom information was available were noted for being “samìn” (13%), obese. It remains open to debate if the quality of corpulence implied in early and classical Islam a positive or negative attribute.

a profile of the

MAWÀLÌ 'ULAMÀ"

473

mawàlì were famed for reading the Qur"àn. This finding is probably meaningful and it gains more weight as we compare it with Qur"àn reading amongst the Arab 'ulamà" where only one of the 21 Arabs were noted for this same observance (data not shown).

Intellectual attributes To qualify as an 'àlim an individual must have intellectual attributes. Our data set allows the assessment of these attributes on the basis of six variables: practice of at least one religious science; any additional intellectual pursuit; committing knowledge to writing (explicit yes/no); topics of knowledge put to writing; numbers of both teachers and pupils. Nothing will be said about the comparisons with Arab 'ulamà" until the very end of this section. Table 9: Percentage Distribution of 'Ilm (N = 157) Óadìth Qirà"a Fiqh Tafsìr Na˙w

70.1 10.8 10.8 7.6 0.6

As stated earlier, the point of departure in our research design is the five Islamic religious sciences of ˙adìth, qirà"a, tafsìr, fiqh and na˙w. Table 9 presents the percentage distribution for the mawàlì across these sciences. The vast majority were active in ˙adìth—more than two-thirds. This is followed by both qirà"a and fiqh, each encompassing about 10%. Finally, some 8% were active in tafsìr and less than 1% had been involved in na˙w. The preponderance of ˙adìth comes as no surprise since this endeavor was a foundation for the development of the other religious sciences. The frequency of involvement in fiqh and qirà"a is identical. It is easier to interpret the prevalence of the former than qirà"a especially as we inspect table 10 where data are displayed for a second intellectual pursuit.16 16 It should be recalled that we saw that the mawàlì differed significantly in unusual religious observances from the Arabs for reading the Qur"àn. Inasmuch as qirà"a stood out amongst the pursuits of the mawàlì 'ulamà", this result may be construed as confirmation of their special interest in the Holy Book.

474

john a. nawas

Table 10: Percentage Distribution across an Intellectual Endeavor other than the one that identified the 'Alim (N = 83) Fiqh Óadìth Qirà"a Na˙w Shi'r Tafsìr Kalàm Qißaß Gharìb Sunna Qur"àn Mùsìqà Comp. Religion

32.5 31.3 10.8 7.2 4.8 2.4 2.4 2.4 1.2 1.2 1.2 1.2 1.2

We see in this table that fiqh occupies the top rank that is practically the same for ˙adìth, which is probably indicative of the close affinity between these two disciplines. Qirà"a as an extra discipline is also high in the table—as it was in the previous one. A comparison of participation in na˙w in both tables strongly suggests that the pursuit of this disciple reflects a secondary rather than a primary interest in our period. Poetry (close to 5%) is the first intellectual endeavor we encounter beyond the scope of the Islamic religious sciences; indeed, it assumes a higher rank in frequency than does tafsìr that is a religious science. Additionally, the size of the table speaks for itself and, all in all, 47% of the mawàlì had no other intellectual interest than the one religious science they were engaged in. To gain more insight into the overall intellectual pursuits of the mawàlì beyond those that relate to the religious sciences, I re-categorized the data of table 10 into three larger groupings, fields that have to do with the “sacred,” language and philosophy, and history. This re-categorization is depicted in table 11. Table 11: Percentage Distribution for Larger Categories of Second Intellectual Endeavor (N = 83) Sacred Language/Philosophy History

79.5 18.1 2.4

a profile of the

MAWÀLÌ 'ULAMÀ"

475

Not surprisingly the largest percentage (close to 80%) is in intellectual endeavors that have to do with the sacred. Almost one-fifth of the mawàlì, however, were additionally occupied with subjects that had to do with language and/or philosophy, followed by a smaller percentage (2.4%) in history. Table 12: Percentage Distribution of Knowledge put or not put to Writing (N = 64) Yes No

93.8 6.3

Table 13: Percentage Distribution of Topics put to Writing (N = 55) Óadìth Tafsìr Qur"àn Fiqh Gharìb Na˙w Maghàzì Kalàm Shi'r Qirà"a Adab Tàrìkh Comp. Religion

49.1 12.7 9.1 5.5 3.6 3.6 3.6 3.6 1.8 1.8 1.8 1.8 1.8

Our data entry scheme allowed us to record who had committed learning to writing. In addition, the biographers report if someone’s ˙adìth is not put to writing.17 Table 12, then, shows the percentage of mawàlì who committed any knowledge to writing and those whose ˙adìth was not written down. A little over 6% of the mawàlì fall in this latter category. The topics on which knowledge was written down appear in table 13. Almost half the mawàlì who did write did so about ˙adìth, followed by tafsìr and Qur"àn (each some 10%). Inasmuch as fiqh was a prominent endeavor it is surprising that only about 5% of the topics written down pertained to this field. This is probably because we are dealing with the period preceding 240/854–5 when the Sunnite schools of law were just starting to develop. 17 Cf. M. Cook, “The opponents of the writing of Tradition in early Islam,” Arabica 44 (1997), 437–530.

476

john a. nawas

We turn now to the next variable. We counted for each individual the longest listing of teachers and pupils encountered for that person. Once we started initial analyses we were surprised to learn of the extraordinarily high correlation between the number of teachers and pupils; the correlation coefficient was 0.71 and statistically highly significant, which is quite respectable even in fields more developed than our own.18 At any rate, this correlation means that if someone had many teachers he is very likely to have many pupils as well.19 The median number of teachers and pupils for the mawàlì was respectively 16 and 22.5.20 These figures hold irrespective of the Islamic science involved.21 We have come to the end of this part on the intellectual attributes of the mawàlì. I ran tests to find out if significant differences existed on any of these variables between mawàlì and Arabs. Not one single intellectual attribute appears to distinguish between the two groups. Summary and discussion of the findings I will now summarize and discuss the findings reported above. It should be stressed that these findings pertain to the entire group of mawàlì 'ulamà" who lived before the year 240/854–5. This period was selected because after it, compilers of the biographical dictionaries scarcely made mention of whether an 'àlim was a mawlà or Arab. The mawàlì 'ulamà" this article covered numbered 157 individuals. And now to what I have found. Very little interest is shown by the compilers in the previous religion of the mawàlì 'ulamà" or in their ethnic background. The information on previous religions was too scanty to allow any sensible statement. With regard to ethnic background we can say little more 18 Inasmuch as both variables (number of teachers and pupils) did not meet the normality assumption, Spearman’s rho was used as correlation measure, p < 0.01. 19 A high correlation between two variables often indicates that these variables are measuring one and the same phenomenon, in our case the phenomenon of “transmission of knowledge.” 20 The median rather than the mean is used because our data are skewed and in such instances the median is a more reliable measure than the mean. 21 Statistical tests were run to see if there are significant differences in the number of teachers and pupils the mawàlì had in the five various Islamic religious sciences. No such difference emerged.

a profile of the

MAWÀLÌ 'ULAMÀ"

477

than that there seems to be some evidence that the larger part of the mawàlì 'ulamà" consisted of Persians. This makes sense; the findings on geographic distribution indicate that the mawàlì primarily hailed from the eastern regions of the Islamic empire. Much more data are available on tribal affiliations. Some 60% of the mawàlì 'ulamà" were affiliated with the leading tribes of Quraysh, Tamìm, Azd and the Anßàr (the tribes of Aws and Khazraj). In terms of Northern versus Southern tribes, a clear majority (two-thirds) were affiliated with the Northern ones. We also examined the patrons of the mawàlì 'ulamà"; only about 10% of the patrons were themselves mawàlì, all others Arabs. As we scrutinized the data from the perspective of time, a very interesting finding came to the fore. The earliest year of death of a mawlà who had another mawlà as patron was 145/762–3 which is only a decade removed from the 'Abbàsid revolution of 132/749–50. This event may have served as a springboard for the upward movement of the mawàlì on the social ladder which made it possible for them to become patrons as well. Moving on to their occupations, one-third of the group of mawàlì 'ulamà" was engaged in the sectors of commerce and education, while the rest have performed a scattered range of occupations within the spheres of administration and legal or religious activities. Turning our attention now to distinctive attributes and practices, we saw that about a quarter of the mawàlì 'ulamà" who were reported to have had an ideological position were Shì'ites, followed by Murjì"ites, Qadarites and Mu'tazilites. Considering the period under study this distribution is not unexpected. Other classifications encountered were even less surprising: of those who received a classificatory label the vast majority consisted of tàbi'ùn. The lion’s share of mawàlì (close to 60%) are listed in either one or otherwise all six Sunnite canonical ˙adìth collections. We noted that about half the mawàlì dyed their beards while the other half did not. The listed physical features for the mawàlì 'ulamà" confirm that we are dealing with an older population, a statement derived from the fact that most of these physical features had to do with traits that usually manifest themselves as one grows older—blindness and senility. The first intellectual variable dealt with in this article pertains to the distribution of the 'ulamà" across the five Islamic religious sciences. Some two-thirds of the mawàlì 'ulamà" was active in ˙adìth, followed by qirà"a, fiqh, then tafsìr and, finally, grammar. Almost half

478

john a. nawas

the mawàlì confined their intellectual interests to only one of these five religious sciences. Of those whose interests were wide, the fields of ˙adìth and fiqh went together and took the lead, followed by qirà"a. We noted that fiqh and especially grammar were not studied singly in this period but usually in combination with some other field of intellectual endeavor. Beyond the five religious sciences which we then categorized as “the sacred,” we saw, quite interestingly, that close to one-fifth of the mawàlì 'ulamà" additionally occupied themselves in a subject that had to do with language and/or philosophy. Of the topics the mawàlì 'ulamà" committed to writing, again ˙adìth leads the way, followed by tafsìr and Qur"àn. We explained the small percentage of mawàlì 'ulamà" who wrote on fiqh (a mere 5%) by the fact that the four Sunnite madhàhib were just starting to blossom at the end of our period and that fiqh in this period was studied in conjunction with some other religious science. After making a point about the exceptionally high correlation between the number of teachers and pupils an 'àlim had, we listed the median for each: the median number of teachers for the mawàlì 'ulamà" was 16 while the median number of students was 22.5. We noted that these figures hold across disciplines. To sharpen our focus, comparisons were made between mawàlì and Arab 'ulamà" of the same time span studied. The overall picture of the comparisons shows the two groups to be far more homogeneous than different. In light of the contention that the mawàlì 'ulamà" had an edge over their Arab counterparts in furthering learning, it was surprising that the two groups displayed no differences on any of the variables within the intellectual sphere—at least within the confines of the data we were able to distill from the biographical dictionaries. We did, however, find some differences between the mawàlì and Arab 'ulamà". These centered either on outward appearance on the one hand, or geographic and migratory drifts and occupational pursuits on the other. Among the first of these, distinctly more Arab than mawàlì 'ulamà" were in the habit of dyeing their beards and were described as obese. Leaving possible genetic factors for obesity aside, one can speculate that the Arab 'ulamà" were financially better off to afford an excessive calorie intake and/or culture-bound or otherwise, a large belly may have been viewed as an index of social standing—perhaps as a dyed beard was. It is a pity that the number of 'ulamà" of either group is not large enough to allow break-

a profile of the

MAWÀLÌ 'ULAMÀ"

479

downs in smaller time segments to see if the mawàlì 'ulamà" have gradually begun, through the process of acculturation, to emulate those who were in the majority at the time—the Arab 'ulamà"—in putting on more weight or acquiring the habit of dyeing their beards. The second difference between mawàlì and Arab 'ulamà" which reflects migratory drift and means of livelihood pertains to the distinct over-representation of the mawàlì 'ulamà" in Northern tribes and, secondly, the relatively high degree of them described in the sources as min ahl Mißr. A thought that comes to mind is that these waves of migrants, as in our time, tended to follow in the footsteps of earlier ones and settle where they did. The same lines of reasoning apply to occupational pursuits. A newcomer is more likely than not to engage in occupations that are available and in which other ingroup members appear to find a livelihood in pursuing. Probably, as time went by—but our data are too limited for breakdowns to unveil time-related trends—the mawàlì 'ulamà" increasingly took on a larger share of judgeships and other key administrative positions that earlier on were more likely to be occupied by Arab 'ulamà". All these speculations and reflections do not, however, change the fact that we did not find evidence to suggest that the mawàlì 'ulamà" were any less or any more influential in advancing the cause of Islamic learning. Indeed, all the findings taken together illustrate a very successful historic case of cultural assimilation, mawàlì and Arabs having together evolved into the intellectual group of 'ulamà", the ultimate bearers of religious authority in Islam.

Bibliography

Primary Sources Ibn 'Asàkir, Abù l-Qàsim 'Alì b. Abì Mu˙ammad, Ta"rìkh Madìnat Dimashq, ed. 'Umar b. Gharàma al-'Amrawì, 80 vols., Beirut: Dàr al-Fikr, 1995. Ibn Óajar al-'Asqalànì, Tahdhìb al-tahdhìb, 12 vols., Beirut: Dàr Íàdir, 1968. al-Mizzì, Jamàl al-Dìn Abù l-Óajjàj Yùsuf al-Mizzì, Tahdhìb al-kamàl fì asmà" al-rijàl, ed. 'Awwàd Ma'rùf, 35 vols., Beirut: Mu"assasat al-Risàla, 1992. Secondary Sources Bernards, Monique and John Nawas, “The geographic distribution of Muslim jurists during the first four centuries AH,” Islamic Law and Society 10 (2003), 168–181. Bulliet, Richard W., “The age structure of Medieval Islamic education,” Studia Islamica 57 (1983), 105–17.

480

john a. nawas

Caskel, Werner, ]amharat an-Nasab. Das genealogische Werk des Hishàm ibn Mu˙ammad al-Kalbì , 2 vols., Leiden: E.J. Brill, 1966. Cook, Michael, “The opponents of the writing of Tradition in early Islam,” Arabica 44 (1997), 437–530. Crone, Patricia, Roman, provincial and Islamic law. The origins of the Islamic patronate, Cambridge: Cambridge University Press, 1987. Ess, Josef van, Theologie und Gesellschaft im 2. und 3. Jahrhundert Hidschra. Eine Geschichte des religiösen Denkens im frühen Islam, 6 vols., Berlin/New York: Walter de Gruyter, 1991–1997. Gilbert, Joan E., “Institutionalization of Muslim scholarship and professionalization of the ‘'ulamà"’ in Medieval Damascus,” Studia Islamica 52 (1980), 105–34; Goldziher, Ignaz/H.J. Kissling, “Abdàl,” in EI 2, I, 95. (nalcık, Halil, “Servile labor in the Ottoman Empire,” in id., Studies in Ottoman social and economic history, London: Variorum Reprints, 1985, article VII. Juynboll, G.H.A., “Some new ideas on the development of sunna as a technical term in early Islam,” Jerusalem Studies in Arabic and Islam 10 (1987), 97–118. id., “An excursus on the ahl as-sunna in connection with Van Ess, Theologie und Gesellschaft, vol. IV,” Der Islam 75 (1998), 318–30. id., “Dyeing the hair and beard in early Islam. A Hadith-analytical study,” Arabica 33 (1986), 49–75. Pellat, Charles, “Quelques chiffres sur la vie moyenne d’une catégorie de Musulmans,” in Pierre Salmon, ed., Mélanges d’islamologie: Volume dédié à la mémoire de Armand Abel, Leiden: E.J. Brill, 1974, 233–246. }entürk, Recep, Narrative social structure. Anatomy of the Hadith transmission network, CE 610–1505, Ph.D. Diss., Columbia University, 1997.

LIST OF CONTRIBUTORS

Monique Bernards is Lecturer in Arabic at the University of Groningen. Her publications deal with intellectual and social history, including her Changing Traditions. Al-Mubarrad’s Refutation of Sìbawayh and the Subsequent Reception of the Kitàb (1997). Richard Bulliet is Professor of History at Columbia University. His works touching on early Islamic matters include The Patricians of Nishapur (1972), Conversion to Islam in the Medieval Period (1979), Islam: The View from the Edge (1993), and The Case for Islamo-Christian Civilization (2004). He has also written The Camel and the Wheel (1975) and Hunters, Herders, and Hamburgers (2005). He has edited The Columbia History of the Twentieth Century (1998) and co-authored The Earth and Its Peoples: A Global History (3rd ed. 2004). Lawrence Conrad is University Professor and Professor of the History and Culture of the Middle East at the Asia-Africa Institute, University of Hamburg. One of his recent publications is (editor) History of Historiography in Early Islamic Times: Studies & Perspectives (Studies in Late Antiquity & Early Islam, No 9, 2006). Patricia Crone studied Islamic History at the School of Oriental and African Studies, London, and taught for many years in Oxford and Cambridge. She is now Professor of Islamic History at the Institute for Advanced Study, Princeton. Among her books are Meccan Trade and the Rise of Islam (1987), Pre-industrial Societies: Anatomy of the Pre-Modern World (1989) and Medieval Islamic Political Thought (2004). Amikam Elad is Associate Professor of Arabic and Middle Eastern Studies at the Hebrew University of Jerusalem. His general field of interest is the medieval history of the Middle East, in particular the history of the early 'Abbàsid Caliphate and the history of Syria and Palestine. His publications include Medieval Jerusalem and Islamic Worship. Holy Places, Ceremonies, Pilgrimage (1995). Maribel Fierro is Associate Professor at the CSIC (Madrid). Her research deals mainly with the history of al-Andalus, Islamic law and religious practices in classical Islam. Among her recent publications:

482

list of contributors

Abd al-Rahman III, the first Cordoban caliph (2005) and Los almohades: Problemas y perspectivas (ed. P. Cressier, M. Fierro and P. Guichard, 2005). Robert Gleave joined the University of Bristol in 1995 and is currently Reader in Islamic Studies in the Department of Theology and Religious Studies. He specializes in Shì'ism generally and the development of Shì'ì law in particular. His books include Inevitable Doubt: Two Theories of Shi'i Jurisprudence (2000). Jamal Juda studied at the Eberhard Karls Universität, Tübingen and is now Associate Professor of History, An-Najah National University, Nablus. His books include al-'Arab wa l-ar∂ fì l-'Iràq fì ßadr al-islàm (1979) and Die Sozialen und Wirtschaftlichen Aspekte der Mawàlì in Frühislamischer Zeit (1983). Hilary Kilpatrick studied Arabic at Oxford. She has taught at universities in the United Kingdom, the Netherlands and Switzerland, and now works as an independent scholar in Lausanne. Her numerous publications on modern and classical Arabic literature include a study of the Kitàb al-Aghànì. Making the Great Book of Songs (2003). She has also co-edited Religious Perspectives in Modern Muslim and Jewish Literatures (2005). Ella Landau-Tasseron is Professor at the Hebrew University of Jerusalem. She studies Islamic history, tradition, institutions and historiography. She also participated in the ˇabarì translation project initiated by Columbia University, where she produced the last volume of the series entitled Biographies of the Prophet’s Companions and Their Successors (1998). Michael Lecker is Professor of Arabic at the Hebrew University of Jerusalem. He has extensively published on pre- and early Islamic history, the biography of Mu˙ammad and the Jews of Arabia. A recent publication is The “Constitution of Medina”: Mu˙ammad’s First Legal Document (2004). Ulrike Mitter was an assistant at the University of Nijmegen where she received her PhD in 1999. From 1999 to 2001 she was a parttime Lecturer at the University of Hamburg and since 2002 director of DAAD (German Academic Exchange Service) Information Centre in Baku/Azerbaijan. She has published on Muslim Spain, ˙adìth and the development of Islamic law.

list of contributors

483

John Nawas is Professor of Arabic and Islamic Studies at the Catholic University Leuven. He has written on the socio-political history of the early 'Abbàsid period and on the 'ulamà" as a social group. Currently he is one of the four executive editors of the Encyclopaedia of Islam. Third Edition. Geert Jan van Gelder was Lecturer in Arabic at the University of Groningen for many years, before becoming Laudian Professor of Arabic at the University of Oxford in 1998. He has published widely on classical Arabic literature, including Close Relationships: Incest and Inbreeding in Classical Arabic literature (2005).

INDEX

al-'Abalàt, Tribe, 334 Abàn b. Abì 'Ayyàsh, Traditionist, 169 Abàn b. Íàli˙ b. 'Umayr b. 'Ubayd, Abù Bakr al-Qurashì, Scholar, 458 n. 6 Abàn al-Là˙iqì, Poet, 357, 365, 366 n. 117 al-'Abbàs b. 'Abdallàh b. Màlik b. al-Haytham, Commander, 285, 286 n. 36 al-'Abbàs b. Ja'far b. Mu˙ammad b. al-Ash'ath, Commander, 284, 286 n. 36 – Son of, Commander and ßà˙ib al-shur†a of 'Alì al-Ri∂à, 285 al-'Abbàs b. al-Musayyab b. al-Zuhayr, Commander, 284, 286 n. 36, 312 'Abbàsid revolution, 177–8, 182, 186–92, 273, 281–2, 285 n. 29, 289, 291, 297, 303–4, 318–9, 458, 477 – Shì'ì view of, 187 – see also Da'wa 'Abd, Tribe, 51 'Abd al-Awwal, Grandfather of Abù Nuwàs, 350 'Abd al-'Azìz b. Marwàn, Governor, 267 'Abd al-'Azìz b. 'Umar, Caliph, 82, 90, 119 'Abd al-Dàr, Tribe, 29 'Abd al-Karìm b. Abì l-Mukhàriq, Traditionist, 91 'Abd al-Karìm al-Nahshalì, Scholar, 366 'Abd al-Malik b. Marwàn, Caliph, 171 n. 12, 268, 272–3, 314 'Abd Manàf, Tribe, 7 'Abd al-Mu††alib, Ancestor of prophet Mu˙ammad, 7, 56 n. 11 'Abd al-Qays, Tribe, 353 'Abd al-Ra˙màn al-'Ablì, Poet, 233 'Abd al-Ra˙màn b. 'Abdallàh b. Mas'ùd, Ally of Zuhra, 15 'Abd al-Ra˙màn b. 'Abd al-Ra˙màn b. 'Abdallàh b. 'Àmir al-Óa∂ramì, Ally of Óarb b. Umayya, 15–6

'Abd al-Ra˙màn b. 'Awf al-Zuhrì, Ibn Abì Qabà˙a, Companion, 28 n. 105, 197–8, 327 n. 5, 332 n. 33 'Abd al-Ra˙màn b. Jabala, Commander, 284 n. 25 'Abd al-Ra˙màn b. Say˙àn, Poet, 15 n. 45 'Abd al-Ra˙màn b. 'Uthmàn b. 'Ubaydallàh, Nephew of ˇal˙a b. 'Ubaydallàh, 17 'Abd al-Ra˙màn I, Emir, 196, 201, 204–6, 215–7, 234, 236, 240 'Abd al-Ra˙màn II, Emir, 204 n. 30, 217, 222, 233 'Abd al-Ra˙màn III, Caliph, 216 n. 75, 217, 219, 235 'Abd al-Razzàq al-Ían'ànì, Traditionist, 21, 86 n. 64 'Abd Shams, Tribe, 172 n. 18, 334 n. 39 'Abd al-'Uzzà, Tribe, 32 n. 125 al-Abdàl, 470 'Abdallàh, Umayyad emir, 201 'Abdallàh b. al-'Abbàs, See Ibn 'Abbàs 'Abdallàh b. 'Abbàs al-Rabì'ì, Musician, 341 'Abdallàh b. 'Abd al-Óakam, Scholar, 75 n. 25 'Abdallàh b. Abì l-'Alà", Musician, 342 'Abdallàh b. Abì Is˙àq, Linguist, 26 n. 95 'Abdallàh b. 'Adì b. al-Óamrà", Ally of Zuhra, 16 'Abdallàh b. 'Alì, 'Abbàsid rebel, 311 'Abdallàh b. 'Àmir, Governor, 268 'Abdallàh b. 'Amr b. 'Abd al-Qays, Traditionist, 265 'Abdallàh b. 'Atìk, Raid of, 61 n. 33 'Abdallàh b. Farwa, Mawlà of Caliph 'Uthmàn, 267 'Abdallàh b. al-Óasan Ibn al-Sindì, Qà∂ì, 235 'Abdallàh b. Ja'far, Musician, 332–3 'Abdallàh b. Ja˙sh, Poet, 340 n. 75 'Abdallàh b. Jud'àn, Qurashì nobleman, 28

486

index

'Abdallàh b. Màlik b. al-Haytham, Commander, 284, 296 'Abdallàh b. Mu'àwiya al-Bàhilì, Musician, 332 n. 33 'Abdallàh b. Mu˙ammad al-Amìn, Musican and 'Abbàsid prince, 339 'Abdallàh b. Mu˙ammad b. Khàlid b. Martinìl, Umayyad mawlà, 227 n. 98 'Abdallàh b. Mùsà al-Hàdì, Musician and 'Abbàsid prince, 339 'Abdallàh b. al-Mu'tazz, Musician and 'Abbàsid prince, 339 'Abdallàh b. Shaddàd, Traditionist, 96–7 'Abdallàh b. ˇàhir, Musician and governor, 339–40, 342, 344 'Abdallàh b. 'Umar, Companion, 396 'Abdallàh b. al-Zubayr, see Ibn al-Zubayr 'Abdallàh al-Khayyà†, Poet, 352 Abnà", 279, 281–7, 289, 306, 309, 314 – And ahl Khuràsàn, 281 n. 17, 283–7, 288, 289 n. 48, 307, 310, 317, 319 – And al-Amìn, 287 – And al-Ma"mùn, 287 – al-Anßàr, 353 – Anßàr al-dawla, 293 – al-Dawla, 308 – Distinguished families of, 294, 294–5 n. 73 – al-Furs, 356 – In Baghdad, 289–92, 294 – al-Shì'a, 289 n. 48, 291 n. 62, 320 – Social status of, 292–3 – See also Ahl Khuràsàn Abrogation, 4, 9 n. 24, 20–1, 27 n. 99 'Abs, Tribe, 315–6 n. 146, Abstinence, Sexual, 142 n. 16 Abù l-'Abbàs al-Saffà˙, Ibn al-Óàrithiyya, Caliph, 178, 310–1, 316, 337 Abù l-'Abbàs al-Sarakhsì, Musician, 345 Abù Abdrab, Mawlà of 'Uthra, 268 Abù A˙mad b. al-Rashìd, Musician, 340 Abù 'Amr b. al-'Alà", Scholar, 366 Abù l-Aswad al-Du"alì, Scholar, 431–2, 436–7, 440, 449–50 Abù 'A†à" al-Sindì, Poet, 354–6 Abù l-'Atàhiya, Poet, 352, 357, 360

Abù Bakr, Caliph, 32 n. 122, 59, 167–9, 171, 177–9, 406 Abù Dukayn, Mawlà of al-Jamàliyyìn, 268 Abù Dulaf al-Qàsim b. 'Ìsà al-'Ijlì, Musician, 340 Abù Óanìfa, Scholar, 23, 93 n. 97, 97, 106 Abù Óasan al-ˇùsì, 350 Abù Óashìsha, Musician, 344 n. 99 Abù l-Hindì, Poet, 362 Abù Óudhayfa, Companion, And family, 26 n. 94 Abù 'Ìsà, Tribe, 202–3, 205 Abù 'Ìsà b. al-Mutawakkil, Musician and 'Abbàsid prince, 339 Abù 'Ìsà b. al-Rashìd, Musician and 'Abbàsid prince, 339 Abù Jilda al-Yashkurì, Poet, 362 Abù Jubayla, 52 n. 5 Abù l-Julayd, Father of 'A†à", 12 n. 33 Abù Kathìr, Mawlà of Aslam, 267 Abù l-Kha††àr al-Kalbì, Governor, 206 Abù Mi˙jan al-Thaqafì, Poet, 363 Abù Mikhnaf, Scholar, 377 Abù Mùsà al-Ash'arì, Governor, 168, 174, 176 n. 37 Abù Mùsà al-Hawwàrì, Berber, 202 n. 20 Abù Muslim, Commander, 187, 189, 191–2, 308, 342 Abù Nu'aym, Scholar, 373 Abù Nukhayla, Poet, 372 n. 10 Abù Nuwàs, Poet, 362, 365, 366 n. 117 – Ethnic background of, 350–1 Abù Qilàba, Successor, 87 Abù Sufyàn, Father of Mu'àwiya, 8, 29 n. 108, 32, 173, 397–8 Abù Tijràt, And sister, 29 Abù 'Ubayda, Scholar, 350, 377 Abù 'Ulayya, Poet, 364 Acién Almansa, M., 221, 222–4, 226–8 Adab, 442–4, 451 – Ten categories of, 357 n. 64 Adam, 173, 177 'Adì, Tribe, 172, 315 n. 146, 352 'Adì b. Zayd, Poet, 413 'Adìd, 87 n. 71 – And ˙alìf, 39–41 – Definition of, 24–33 – Descent group, 21 – See also Descent groups; Dhawù

index 'Adnàn, Tribe, 10 Adoption, 159 'Adwàn, Tribe, 352 al-Afàriqa, Military unit, 307 Afghanistan, 259 al-Afrak, Musician, 335 n. 46 Agnatic tie, 73, 120 – See also Kubr Agnatic title, Of manumitter to freedman’s property, 79 – See also Inheritance; Dhawù; Walà" Ahl al-ar∂, 86, 89, 289 n. 47 Ahl al-balad, 154, 204–5, 209 Ahl al-bayt, 19–20, 154, 167–8 – See also Prophet Mu˙ammad Ahl al-dìwàn, 154 – See also Dìwàn Ahl al-˙adìth, In al-Andalus, 236 – See also Óadìth Ahl Khuràsàn, 279, 281–3, 312 – 'Ajam, 293, 318 – Definition of, 281 – Social status of, 305–6 – See also Abnà" and Ahl Khuràsàn; Anßàr ahl Khuràsàn Ahl al-ra"y, In al-Andalus, 236 A˙mad b. 'Abdàn, Mawlà of 'Abd al-Ra˙màn b. 'Awf, 28 n. 105 A˙mad b. Hishàm, Commander, 288 n. 41 A˙mad b. Sayyàr al-Jurjànì, Poet, 366 A˙mad b. Ya˙yà al-Makkì, Musician, 344 A˙mad b. Ya˙yà b. Mu'àdh, Commander, 286 A˙mad b. Yùsuf, Secretary, 304 A˙mad al-Naßbì, Musician, 331–2 al-A˙mar, Linguist, 442 n. 20 al-A˙naf b. Qays al-Tamìmì, Tamìmì nobleman, 265 Ahwaz, Place, 439 'À"isha, Wife of prophet Mu˙ammad, 85 'Ajam, a'àjim, 174, 175, 186, 188–90, 219–20, 223, 228 – Byzantines, 357 Ajìr, Hireling, 26 n. 94 Akh, 50 – See also Khuwwa; Mu"àkhàt Akhbàr, 141, 145–6, 153, 159, 164–5 – About music, 332 – Imàmì and Prophetic ˙adìth, 140

487

– Rejection of, 146 – Relationship with fiqh, 140 al-Akhfash al-Awsa†, Scholar, 437, 450 Akhwàl, Agnates of mother, 22 – Identification of, 296, 298, 304, 309–16 – See also Agnatic tie; Dhawù; Inheritance; Walà" Àl Hàni", Family, 29 al-'Alà", Mawlà of Hàrùn al-Rashìd, Commander, 284 n. 25 Algeciras, Place, 206 'Alì, Tribe, 337 'Alì b. 'Abdallàh, Commander, 311 'Alì b. Abì Kathìr, Poet, 352 'Alì b. Abì ˇàlib, Caliph, 7, 12, 88–9, 113–4, 143, 153, 168–70, 174–5, 176 n. 36, 177–81, 183, 186–7, 189, 266, 333 n. 34 – As universalist, 180, 183 'Alì b. Hishàm, Commander, 320, 343 n. 92 – And family, 291 and n. 61 'Alì b. al-Hishàm, Musician and commander, 339–40 'Alì b. al-Óusayn, Imam, 169 n. 7 'Alì b. 'Ìsà, Commander, 288, 314 n. 142, 320 'Alì al-Ri∂à, Imam, 285–6 'Alids and Umayyads, Brawl between, 11 'Àliya, Upper Medina, 51–2, 55–6 – See also Medina Alliance(s), 24, 55 n. 10, 59, 372 – Anti-Fà†imid, 10 n. 27 – Between Rabì' and Azd, 10 – Choice of, 13, 16 – Difference with brotherhood, 9 – False claims to, 14–5 – For raids, 18 – Guest, 2 – Óibàl, 29 – Host, 2, 5 n.9, 8, 11–12, 14, 16–8, 33 – Inaugurated in Islamic times, 37–8 – Islamic, 10–8, 15–6 – Islamic objections to, 18–23 – Legal consequences of, 20 – Legal problems of, 13 – Loyalty to, 12 n. 35 – Military and political, 1, 3, 5 n. 9 and 12, 10, 14, 18, 59 – Naming of, 15

488

index

– Non-valid, 11 – Perpetuating pre-Islamic, 34–7 – Pre-Islamic, 1, 2–4, 6, 8–9, 10–1, 14, 16, 17, 20–1, 72 – Prohibition of, 2–5, 10, 12, 14, 33 – Prophet’s principles of, 6 – Without mutual inheritance, 5 n. 11 Alliance ˙adìth, 3–6, 7 n. 18, 9–11, 13–6, 18, 19 n. 62, 20–3, 33 – Óanafì view of, 3 – See also Óilf 'Allùya, Musician, 341 Almoravids and Almohads, Berber dynasties, 195, 232 Altheim, F., 54 Amabe, F., 279 n. 6, 304 n. 97, 308, 313 n. 136 al-Amìn, Caliph, 278, 281, 283, 286 n. 35, 288–90, 304 n. 97, 305, 309–10, 319–20, 343, 442 n. 20 – “Arabness” of, 307 – See also al-Ma"mùn 'Àmir, Contemporary of Óarb b. Umayya, 15 'Àmir b. Màlik, Abù Barà", Leader of 'Àmir b. Ía'ßa'a, 60 'Àmir b. Ía'ßa'a, Tribe, 59–60 'Ammàr, Mawlà of Quraysh, 32 'Ammàr b. 'Amr b. 'Abd al-Akbar, Poet, 363–5 'Ammàr b. Yàsir, Companion, 28 n. 105, 176 n. 37 'Àqila, Blood-money group, 102, 114, 117, 149–50, 152, 154, 163–4 – Óanafì view of, versus Imàmì view, 154 – Women’s exclusion from, 159 – See also 'Aßaba; Blood-money 'Amr b. 'Abdallàh, Mawlà and qà∂ì, 196 n. 3 'Amr b. al-'Àß, Commander, 87 'Amr b. Bàna, Musician, 344 'Amr b. Shu'ayb, Successor, 87 'Amrùs, Tribe, 236 'Amrùs b. Yùsuf al-Muwallad, And family, 234 Amßàr, Garrisons, 249, 265–6, 270–1, 273, 288 – Conditions in, 263 – In al-Andalus, 201, 206, 230, 239 Anas b. Màlik, Traditionist, 371 'Anaza, Tribe, 352

'Anaza al-Sulamì, Mawlà, 28 n. 105 al-Andalus, 274, 438–41 – Arabization and Islamization in, 201–2 Anßàr, 6–7, 9, 19–20, 21 n. 72, 22, 29–30, 51, 54, 55 n. 10, 56–8, 65, 333, 353, 459, 477 – Ahl Khuràsàn, 305 – As mawàlì of Jews, 30 – Genealogy of, 57 – Historiography of, 51, 54–6, 58, 65 – Tribal informants of, 51 – See also Aws and Khazraj Apocalypse, In Islam, 403 Apostates, Safe conduct for, 10 al-'Aqaba, Place, 29 – Pledge of, 6, 30, 58, 403 'Aqìd, Musician, 340 'Aqìl, Tribe, 352 Aqlaf, Uncircumcised, 353 al-Aqrà b. Óàbis, Tamìmì warrior, 61 n. 33 'Arab, Term for Cordovan state, 223 Arab(s) – And Arabic poetry, 233, 360–4 – And music, 328 – Attacks against, in al-Andalus, 234–5 – Ethnic background, 221, 224 – Forged lineages (iddi'à"), 280 n. 10 – Identity, 412 – In al-Andalus, 226 – In the army, 284 – Killing of, 188, 190, 192 – Military elites in al-Andalus, 214 – Musicians, 331–2 – Occupations of, compared with mawàlì, 265 – Patrons of musicians, 332 – Pre-Islamic, 357 – Residential concentration of, 251 – Social status of, Sunnì and Shì'ì view, 183–5, 187 – Status of, compared with mawàlì, 171 – Tribal virtues, 412–3 – With mawlà status, 457–8 Arab(s) and non-Arab(s), 27, 33, 169–71, 218–9, 278, 306–9, 353 – Imàmì and Sunnì view of, 180–1 – Inequality between, 458 – Social status of, 189, 191

index Arab 'ulamà" – And qirà"a, 473 – See also Mawlà and Arab 'ulamà" al-'Aràba, Tribe, 11, 12 n. 33 Arabia, 59, 329 – Old rules, 74 – Pre-Islamic, 31, 62, 64, 72, 80 n. 38, 412–3 – Pre-Islamic wars, 57 – Relationship between settled and nomads, 62–4, 66 – Tribal structure of, 188 Arabic culture in al-Andalus, 233 Arabic grammar, As Islamic science, 426–7, 441 Arabic language, 187–8, 192, 280–1, 354, 357, 361, 372, 414 – Command of, 236–7, 248–9, 251 – Corruption of, 354–6 Arabic linguistic tradition – Basra-Kufa dichotomy, 426–7, 436–7, 440–1, 448, 450–1 – Centers of learning, 436–41 – Foreign influence on, 426, 433–4, 436, 448–50 – Óijàzì school, 437 – Late development of, 426, 433, 449 – Origin of, 432–6 Arabic linguistics – As Islamic science, 448, 451–2 – Na˙w and lugha, 427 Arabic linguists – Academic status of, 441–6 – Assimilation of, in Islamic society, 435 – Conversion to Islam, 450 – Differences between Arabs and non-Arabs, 446 – Ethnic background, 449 – Ethnic distribution of, 430–3 – Generational distance, 434–6 – Listing of intellectual endeavors, 443–4 – Networks of teachers and pupils, 429 n. 9 and n. 10, 444–5 – Number of intellectual endeavors, 442–3 – Occupations of, 446–8 – Social status of, 446–8 Arabic literature, 26, 280–1 – In al-Andalus, 234, 241 – See also Arabic poetry

489

Arabic pangram, 362 Arabic poetry – And Persian poetry, 365 – A†làl, 363 – Bedouin, 365 – Conservatism of, 359 – Modern style of, 359–60 – Qaßìda, 363 – Use of foreign words, 363, 365 – See also Mu˙dath; Poets Arab-Islamic tradition, Parallels with Greek tradition, 391–402 – Awà"il, 408–9 – Causation of great events by discrete single incidents, 394 – Duels between two warriors, 393 – Exhorting speech, 404 – Eyewitness accounts, 397 – Listing and name-giving, 398–401 – Military climax, 405 – Naiveté, 407 – Narrative techniques, 394–6 – Onomatomania, 406–7 – Playful tricks with names, 409 – Supernatural powers, 404–5 – Topoi, 391–3 – Use of numbers, 401–4 Arabization and Islamization – In al-Andalus, 228, 230, 236, 239–40 – See also Assimilation process; Islamization; Islamization process Aramaic language, Palestinian, 414 'Arìb, Musician, 342 Arìsìyìn, Arìs, pl. Arìsìn, Sharecroppers, 413–4 Armenia, Place, 286 n. 36, 292 n. 64, 350, 457 'Arràm al-Sulamì, Scholar, 63 Arwàd, Conquest of, 410–1 'Aßaba, Blood-money group, 86, 88 n. 74, 95, 99 n. 122, 103, 151–3, 155–8, 162 – Shì'ì view of, 154 – Sunnì view of, 154, 157 – See also 'Àqila; Blood-money; Ta'ßìb 'Aßabiyya, 174–5, 223, 237 Asad, Tribe, 352, 364, 404 Asad b. Yazìd b. Mazyad, Commander, 284 n. 25 and n. 27 al-Asadì, Poet, 233 Asàwira, Persian soldiers, 16–7, 181

490

index

Asbàb, Relatives, 153 Aßba˙, Tribe, 28 al-A'shà, Poet, 365, 367 A'shà Hamdàn, Musician, 331 Ash'ab, Musician, 331 Aß˙àb – al-Óadìth, 74 n. 20 – Of Arab lineage, in al-Andalus, 223 – al-Ra"y, 74 and n. 20 Ash'arites, 173 Ashja', Tribe, 6, 59 Ashja' al-Sulamì, Poet, 366 n. 117 'Àßim b. Sulaymàn al-A˙wal, Administrator, 10 n. 26 Aslam, Tribe, 59, 267 al-Aßma'ì, Scholar, 356, 360, 437, 450 Aßram b. 'Abd al-Óamìd, Governor, 290 n. 51 Assimilation process, 234, 371–2, 374, 479 – In al-Andalus, 228 – See also Arabization and Islamization; Islamization 'A†à", Salary, 268, 274 – See also Soldiers al-'A†à", 268 'A†à" b. Abì l-Julayd, 11 'A†à" b. Abì Rabà˙, Traditionist, 80 n. 41, 86 n. 64, 87, 99, 103, 105, 106 n. 155, 110 Atbà' al-atbà', 254 Atbà' al-tàbi'ìn, 253 'Àtika bt. Shuhda, Musician, 340, 342 'A††àf b. Khàlid, 28 n. 105 Awà"il, 468–9 – In Arabic linguistics, 436 – In Arab-Islamic tradition, 408–9 Awlàd mulùk Khuràsàn, Iranian commanders, 282 n. 18 Aws and Khazraj, 50, 52–3, 57 – Banù Qaylah, 53 – Client status of, 58 n. 26, 59 – See also Anßàr Ayalon, D., 293 Ayatollah Khomeini, 134 'Ayn al-Tamr, Place, 370–1 Ayyàm al-'arab, 376–8 Ayyùb b. Salama, Relative of Caliph Hishàm, 315 n. 146 Ayyùb b. Sumayr, Commander, 284 n. 25 Azd, Tribe, 10, 58 n. 26, 312, 459, 477

al-Azdì, Abù Zakariyyà", Scholar, 311, 377 Azerbaijan, Place, 292 n. 64, 439 al-Azhar, 373 'Azza al-Maylà", Musician, 333 'Azzùr, Musician, 335 n. 45 Bàbak, Rebel, 290, 292 n. 64, 295 n. 73, 340 Bàbawayh, Musician, 335 n. 45 Badhl, Musician, 340 n. 75, 343–4 Badì', Novel or original style, 357, 360–1 Badr, Battle of, 54–5, 173 n. 27, 396 Baghdad, Place, 259, 282–3, 285 n. 29, 286–7, 293, 307, 309, 313 n. 136, 319, 337, 340, 342–3, 437 – Siege of, 289, 317, 319 al-Baghùm, Musician, 334 Bahz, Tribe, 11, 12 n. 33 Bajìla, Place, 316 Balì, Tribe, 7, 63 Balkh, Place, 259 Baqì b. Makhlad, Traditionist, 196, 236–8 al-Barà" b. Mu'rùr, Companion, 58 Barbastro, Place, 235 Barìra tradition, 78, 84–5, 94 n. 104, 96 n. 110, 117 – Analysis of, 124–31 Barmakids, 307, 365 Baßbaß, Musician, 342 Bashear, S., 372 n. 9 Bashshàr b. Burd, Poet, 352, 356–7, 359–60 Basra, Place, 10, 87, 89–91, 168, 173–4, 259, 263, 265, 267–8, 337–8, 342–3, 365, 377, 426, 436–7, 440–1 Ba†n, Clan, 52 Bayt al-màl, 99 n. 122, 158, 267–8 Bedouin(s), 53 n. 6, 407 – Speech, 451 Beja, Place, 206, 220, 227 n. 98 Bencheikh, J., 363 Benjamin, Brother of Joseph, 61 Berbers, 198, 201, 219, 273, 311 – As urban elite, 202 – In al-Andalus, 221, 223, 226, 231 – In North Africa, 216 n. 75 – Walà" ties with Arabs, 203 Bergsträsser, G., 79 Bilàl, Companion, 32 Bint Óamza, 96–7 Biographical dictionaries, 262, 455, 457

index – Family ties in, 195–6 – On linguists, 428–9 – Use of, 252–3 Bi"r Ma'ùna, Battle of, 60 n. 29 Bishr b. Sufyàn al-Ka'bì, Tax collector, 59 Black banners, 167, 169, 177, 190 n. 91 Blood-money, 2, 10 n. 26, 12 n. 33, 21–3, 72, 79, 82, 85–6, 90–2, 100–6, 108, 109 n. 172, 113–4, 117–8, 120–2, 138, 140–1, 152, 154–5, 164, 172, 372 – See also 'Àqila; 'Aßaba; Sà"iba Blood revenge, 2, 6, 30, 72 Bray, J., 346 n. 106 Brockopp, J., 73, 74 n. 20 and n. 21 Brunschvig, R., 115 n. 208, 211 Bu'àth, Battle of, 54, 57 Buday˙, Musician, 331 Buhlùl b. Marzùq b. Uskarà, Muwallad leader, 235 Bujayr b. Abì Bujayr, Mawlà, 28 n. 105 Bukhara, Place, 288, 318 Bulliet, R., 207, 229, 374, 455 n. 4 Bùshanj, Battle of, 288 Bußrà, Place, 55 Byzantine Empire, 172 n. 16, 457 – Front, 286 n. 33 – Institutions, 79 Byzantine(s), 'Ajam, 357 Caesar, 172 Captives, Exchange of, 8 Casius, Visigoth, 213 Caskel, W., 392 n. 90, 458–9 Cats – And mice, 364 – Worshippers of, 185, 189 Chalmeta, P., 208, 209 n. 42 Children of Israel, 57 – See also Jew(s) Christian(s), 184, 223, 228, 332, 351, 370, 380, 457 – In al-Andalus, 206–9, 210 n. 49, 219, 228, 232 – Values, 413 – See also Jew(s) and Christian(s) Civil war, 190 and n. 91 Clan, 72 – See also Ba†n; Qawm Client, See Mawlà, mawàlì Common link, 10 n. 26, 13 n. 39, 23

491

n. 82, 77, 85, 88, 89 n. 80, 90 n. 82, 90, 93, 94 n. 103, 96 n. 110, 101 n. 129, 107, 109 n. 172 – Dive, 76 – Inverted partial, 75–6 – Partial, 75–6, 89 n. 79, 93 n. 93 – Real, 76 – Seeming, 76 – Spider, 76 – Theory of, 75 Companions, See Ía˙àbì Constantinople, Place, 411 Conversion, 82, 83 n. 51, 90, 209, 226 – After capture, 181 – After manumission, 248 – And inheritance, 209 n. 42 – As a societal process, 248 – Conditions for, 248 – Curve of, in al-Andalus, 207 – Curve of, in Iran, 261 – Due to social pressure, 208 – Encouragement of, 210, 214 – In al-Andalus, 230 – Individual, 248 – Involuntary, 230, 240, 246 – Involuntary, in al-Andalus, 229 – Màlikì view of, 210–1 – Mass, 251 – Of slaves, 370 – Process of, 250, 260 – Protestant, in America, 246 – Rural-urban differentiation, 251–2, 260–1 – Termination of process, 247 – Voluntary, 181 – With or without walà", 91 – Without manumission, 205 – Without walà", 90, 100, 196, 207, 219, 250, 261 Convert(s), 108 n. 168 – Arab, 82, 86 n. 63, 101 n. 132 – Assimilation of, in al-Andalus, 226 – Non-Arab, 82, 85, 86 n. 63, 92 n. 92, 112, 174 – Not being mawlà, 99 – Social status of, 373 Cordova, Place, 233, 235–6, 440 n. 19 Crone, P., 26 n. 96, 27 n. 99, 30 n. 115, 31 n. 116, 52, 72–3, 74 n. 21, 79, 80 n. 38, 83 n. 51, 92 n. 92, 97 n. 114, 106 n. 158, 109 n. 174, 115 n. 208, 143 n. 20, 200–1, 219, 225, 282 n. 18, 284 n. 28, 285 n. 29, 286 n. 33,

492

index

297 n. 81, 308, 328 n. 8, 341 n. 82, 458 n. 7 – DAEP rule, 95 n. 107 Cádiz, Place, 220 al-Dalàl, Mukhannath musician, 336 Damascus, Place, 268, 334 Îàmin, See Ta∂ammun Îamra b. Tha'laba, Mawlà, 28 n. 105 Danànìr, Musician, 343–4 Daniel, E., 288 n. 43, 304 n. 97, 313 n. 136, 314 n. 139 Dàr al-da'wa, 318 Dàr al-rizq, 267 Îariyya, Place, 64 Da'wa, Second, of al-Ma"mùn, 287, 289, 291, 297, 303–5, 317, 319 – See also 'Abbàsid Revolution Dàwùd b. 'Alì, Commander, 178 Dàwùd b. Màsajùr, Commander, 291 n. 60 Dàwùd b. Salm, Poet, 352 Daylam, Place, 289 n. 47 Denia, ˇà"ifa kingdom, 232 Descent groups, Definition of, 1 n. 1 Dhawù – l-ansàb, 152–3, 156, 159–60 – l-ar˙àm, 95–6, 99 – l-quràba, 19, 153 – l-qurbà, 19 – See also Agnatic tie; Inheritance; Walà" Dhimmì(s), 209, 214, 223, 228, 230, 240 – As patron, 85 – Social status of, 208 – See also Jew(s); Christian(s); Jews and Christians Dhubyàn, Tribe, 66, 359 Dicta, 75 n. 25 Dihqàn, 308, 314 Dìnàr, Mawlà, 268 Dìwàn, 14, 24, 87, 101, 102 n. 136, 103, 171 n. 12, 233 n. 115 – al-'A†à", 267 – al-'A†à"/al-jund, 264 – al-Kharàj, 267, 269 – al-Khatam, 270 – al-Shi'r, 344 n. 96, 365 Diya, See Blood-money Donner, F., 378, 411 Dukayn b.Yazìd, Musician, 335 n. 45 Îumayra, 7 Duqàq, Musician, 343

Egalitarianism, 238 n. 129 – In al-Andalus, 234 Egypt, Place, 264, 267, 272, 339, 373, 409 Elad, A., 190 n. 92 Elvira, Place, 206, 221 Epalza, M. de, 207 Ess, J. van, 472 n. 14 Ethiopia, Place, 457 Ethnicity, 181, 188, 357 – Arab, 190, 234 – Determination of identity, 359 – Khàrijì view of, 185 – Of musicians, 338, 340–1 – Of ˇàhir b. al-Óusayn’s army, 288 – Shì'ì view of, 186 – See also Identity European expansion, 184 Fadak, Place, 51, 60–1 – Jews of, 59 – People of, 60 al-Fa∂l b. al-Rabì', Vizier 286, 290, 303, 341 al-Fa∂l b. Sahl, Vizier 296, 298, 304, 308 n. 116, 309–10 – Glorification of, 298–9 al-Fa∂l b. Ya˙yà al-Barmakì, Administrator, 366 – See also Barmakids Fàkhita bt. QurΩa, Wife of Caliph Mu'àwiya, 267 Falaturi, A., 81 n. 47 Fallà˙, 447 n. 24 Faqìh, Fuqahà", 82 – Representing Imam, 164 – See also Fiqh al-Farazdaq, Poet, 26 n. 95, 356 Farewell oration, Shì'ì view of, 188 Farewell pilgrimage, 4 Farghana, Place, 457 Farìda, Musician, 343 al-Farrà", Scholar, 437, 451 Fars, Place, 338, 461 Fà†ima bt. 'Abdallàh b. Mùsà, Musician and 'Abbàsid princess, 344 al-Fà†imì, Berber rebel, 201 Fay", 182 Fayrùz Óusayn, Mawlà, 268 Fazàra, Tribe, 65 Fazzàr, Musician, 335 n. 47 Felipe, H. de, 201–2 Fenik, B., 392

index Feudalism, In al-Andalus, 221–8 Fihr, Tribe, 235 Financial crises, 272–3 Fiqh, 75, 79, 82, 92 n. 92, 98, 101 n. 132, 140–7, 151–3, 156, 159, 426, 441, 444, 451, 454, 473–5, 477–8 – Foreign influence on, 80 n. 39 – Imàmì, 134–6, 140 – Sunnì, 135 – Ußùl, 142 Fishbein, M., 304 n. 97 Fitna, 187, 222–3, 226, 241 – See also Civil war al-Fi†yawn, Jewish king, 52, 54, 57 Foundling(s), 85–6, 92, 100, 108, 120, 152 – Maintenance of, 93 – Walà" over, 120 Friday prayer, 164 Frolov, D., 335 n. 49 Fulay˙ b. Abì l-'Awrà", Musician, 337 Furs, 303 n. 96, 313 Futù˙ tradition, 377, 391, 393, 403–4, 411 Gabrieli, F., 304 n. 97, 306–8, 308, 313 n. 136 Genealogy, See Descent groups; Nasab Georgius, Byzantine commander, 379 Ghàlib b. Tammàm, Arab, 232 Gharìb, 428 al-Gharì∂, Musician, 331, 334, 336 Ghassanid kings, 367 Gha†afàn, Tribe, 59, 61 n. 33, 65–6, 316 n. 146 al-Ghazàla, Tribe, 12 n. 33 Ghifàr, Tribe, 59 al-Ghinà" al-mutqan, Novel style of singing, 329–30, 333, 342, 346 Gilbert, J., 455 n. 4 Gleave, R., 109 n. 170 Glick, T., 227 Goldziher, I., 4 n. 6, 74 n. 20, 372 n. 9, 373 Granada, Place, 220 Greek grammar, 426 Greeks, And non-Greeks, 380 Grimme, H., 414 n. 221 Guichard, P., 222 Óabash, Musician, 344 al-Hàdì, Caliph, 340 Óadìth, A˙àdìth, 70, 72 n. 12, 74, 76,

493

78, 80–3, 87 n. 70, 92, 98, 103, 109–10, 112, 116, 119–21, 254, 281, 395, 403, 407, 426, 441, 444, 451, 454, 468, 471, 473–5, 477–8 – Analysis of, 73, 75 – And legal practice, 122–3 – Authenticity of, 6, 13, 74, 82 n. 50, 89 n. 79, 89 n. 80, 90, 93, 96, 101, 110, 115 – Canonical, 76, 84 – Condemning music, 345, 347 n. 112 – Contraction to, 18 – Fabrication of, 76–7, 89, 96 n. 110, 118 n. 215, 397 – Legal, 70, 74 n. 20 – Non-Prophetic, 76 – Pre-canonical, 76, 83, 86 – Prophetic, 2–3, 6–8, 74 n. 20, 82, 89 n. 79, 90, 92, 96, 97 n. 112, 101 n. 129, 108, 109 n. 173, 109 n. 175, 111, 118, 135, 164, 237 – See also 'Ilm al-˙adìth Hadramawt, Place, 267 Óafß b. Maymùn, Berber, 232 Hagarism, 411 n. 212 Hajar, Prostitutes of, 51 al-Óajjàj b. Yùsuf, Governor, 267–8, 272 – As hater of mawàlì, 356 al-Óakam, Traditionist, 93 al-Óakam II, Caliph, 197 n. 5 al-Óakam b. 'Abdal, Poet, 364 Óakam al-Wàdì, Musician, 338 al-Óàkim al-Naysàbùrì, Scholar, 253 Óalìf, Óulafà", 72, 81, 181 n. 58 – Confusion with mawlà, 1 – Definition of, 24–33 – See also Óilf Hallaq, W., 30 n. 115, 31 n. 116, 72, 73 n. 16, 80 Hamadan, Place, 88 n. 75 Óamda bt. al-Rashìd, Musician and 'Abbàsid princess, 344 Hamdàn, Tribe, 331, 363 Óammàd, Traditionist, 91 Óammàd b. Is˙àq al-Mawßilì, Musician, 344 al-Óamrà", Non-Arabs, 180 Óamrà" al-daylam, Persian soldiers, 16–7 Óamza al-Ißfahànì, Scholar, 190 Hand-drum, Musical instrument, 330 n. 21 Hàni", Father of Abù Nuwàs, 350

494

index

Óanìfa, Tribe, 62 Óaras, 286 Óarb b. Umayya, Father of Abù Sufyàn, 15, 16 n. 45 Harim b. Sinàn, Poet, 367 al-Óarìrì, Poet, 361 al-Óàrith b. 'Awf, Leader of Murra, 65 al-Óàrith b. Óißn al-Óarshà, Leader of Kalb, 60 al-Óàrith b. Ka'b, Tribe, 311, 316 n. 146 al-Óàritha b. Badr, Poet, 362 al-Óarra, Battle of, 311 Harthama b. A'yan, Commander, 285, 286 n. 36, 287, 289, 317, 319–20 Hàrùn al-Rashìd, Caliph, 179, 280–4, 285 n. 29, 286, 290 n. 51, 296, 311, 338, 340–1, 343, 366, 442 n. 20, 461 al-Óasan al-Baßrì, Scholar, 98 n. 117 – View on walà" of foundling(s), 94 n. 103 al-Óasan b. A˙mad al-Kàtib, Scholar, 327, 347 al-Óasan b. Sahl, Commander, 290, 292 – List of ten categories of adab, 357 n. 64 Hàshim, Ancestor of prophet Mu˙ammad, 7 Hàshim, Tribe, 19, 169, 172, 176–9, 185, 188, 190 Hàshim b. 'Abd al-'Azìz, Commander, 225 Óàshiya, Entourage, 26 Óassàn b. Thàbit, Poet, 333 n. 36, 350 Óassàn b. Yasàr al-Hudhalì, Qà∂ì, 213–4, 236 Óà†ib b. Abì Balta'a, Companion, 14 Óàtim b. Harthama, Commander, 284 Hawàzin, Tribe, 364 al-Haytham b. 'Alì, Scholar, 170 n. 9 Hazaj, Singing by women, 328, 331 n. 29, 335 n. 44, 336 al-Óazìn, Poet, 352 Hejaz, 53 n. 6, 70 n. 2, 85, 110, 113, 115, 117–9, 259, 329–30, 332–3, 337, 343, 439, 468 – Road of, 64 Heraclius, Byzantine ruler, 397–8, 413–4 Óijàba, 286

Hijra – Emigration, 11–2, 51, 54 – Settlements, 207 Óilf, Alliance, 9, 17, 72, 204, 213, 312 – Mu˙àlafa, 202 – Oath/alliance, 26 – Political, 81 n. 44 – Pre-Islamic, 5, 8, 80–2, 121 Óilf al-fu∂ùl, Meritorious alliance 5, 11 – See also Alliance(s) Hilwan, Place, 289 Óimmàm, Tribe, 372 n. 10 Hind bt. 'Utba, Mother of Caliph Mu'àwiya, 328 al-Óìra, Place, 62, 331 n. 29, 335, 413 – King of, 51 Hishàm, Caliph, 29, 171 n. 12, 315 n. 146, 272, 362 Hishàm, Commander, Father of al-Khalìl b. Hishàm, 288 n. 41 Hishàm b. Hubayra, Qà∂ì, 123 n. 220 Hispano-Goths, 220, 222, 227 Hispano-Romans, 220 Historiography – Christian, 373 – Discontinuity in transmission history, 409–10 – See also Anßàr; Music Hitti, Ph., 371–2 n. 9 Óiyal, Legal devices, 123 n. 219 Homicide, Unintentional, 142, 150 Horovitz, J., 376, 405 Hùd, Tribe, 220 n. 87 Óudà", Singing, 328, 337 Óudaybiyya, Treaty of, 7, 12 Hudhayl, Tribe, 12, 214 n. 67, 352 Huesca, Place, 220 n. 87, 235 Óumayd b. 'Abd al-Óamìd al-ˇà"ì, Commander, 289–90, 292, 319–20 – And family, 291 and n. 62 Óumràn b. Abàn, Mawlà of Caliph 'Uthmàn, 265 Óunayn, Place, 57–8 Óunayn b. Balù' al-Óìrì, Musician, 332 Óusayka, Place, 55–6 – Battle of, 55 – Jews of, 56 al-Óusayn, Father of ˇàhir b. al-Óusayn, 288 Iblìs, Musicial inspiration by, 337 n. 58 – See also Satan

index Ibn 'Abbàd al-Kàtib, Musician, 337 Ibn 'Abbàs, Scholar, 12–3, 89, 190 n. 91, 258 Ibn 'Abd Rabbih, Scholar, 327, 364 Ibn Abì 'Atìq, Musician, 332, 333 n. 34 and n. 35 Ibn Abì Du"àd, Qà∂ì, 356 Ibn Abì l-Dunyà, Scholar, 345 Ibn Abì Mu'ay†, Jewish child adopted by Umayya, 173, 174 n. 28 Ibn Abì ˇàhir Tayfùr, Scholar, 293–4, 304 Ibn Abì l-Zawà"id Sulaymàn b. Ya˙yà, Poet, 364 Ibn 'À"isha, Musician, 331 Ibn 'Asàkir, Scholar, 28, 458 n. 6 Ibn al-Ash'ath, Rebel, 268, 273 Ibn A'tham al-Kùfì, Scholar, 377 Ibn Bàbùya, Scholar, 145, 148 Ibn Dàwùd al-Ißbahànì, Scholar, 361 Ibn al-Faqìh, Scholar, 186 Ibn García, Scholar, 232 Ibn Óabìb, Mu˙ammad, Scholar, 12 n. 34, 14–5, 29, 401 Ibn Óafßùn, Rebel, 220–6, 229, 231–2, 240 Ibn Óajar, Scholar, 28, 471 Ibn Óanbal, Scholar, 23 n. 83, 472 n. 14 Ibn Harma, Poet, 360–2 Ibn Óayyàn, Scholar, 223 Ibn Óazm, Scholar, 8, 18, 20, 22 Ibn al-'Imàd, Scholar, 429, 455 Ibn Is˙àq, Scholar, 27, 57, 372, 377 – Ancestors of, 405 – Ethnic background of, 370–1 Ibn Jàmi', Musician, 338 Ibn Jùdì, Poet, 221 Ibn Jurayj, Traditionist, 80 n. 41, 82 n. 48, 86 n. 64, 103 n. 143 Ibn al-Kalbì, Scholar, 62 Ibn Kathìr, Scholar, 5 Ibn Khaldùn, Scholar, 372 Ibn Khurdàdhbih, Scholar, 327, 330 n. 20, 345 Ibn Mas'ùd, Traditionist, 15, 32, 86 n. 64, 87–8, 99, 110–1, 176 Ibn Mayyàda, Poet, 359, 361 n. 83 Ibn Mish'ab, Musician, 331 Ibn Misja˙, Musician, 329 n. 15, 330–1 Ibn al-Mughallis, Ascetic, 235 Ibn Mu˙riz, Musician, 330–1, 336 Ibn Nufays, 342

495

Ibn Qutayba, Scholar, 353, 357–8, 360 Ibn Ràmìn, 342 Ibn Rashìq, Scholar, 358, 366–7 Ibn Sa'd, Scholar, 429, 455 n. 2 Ibn Shaddàd, Traditionist, 98 Ibn Shammàs, Mawlà/Patron, 336 Ibn Surayj, Musician, 329 n. 15, 331, 334 n. 41, 336 Ibn al-ˇa˙˙àn, Scholar, 327–8, 336 Ibn Ukht al-Nabì, “Son of our sister”, 304, 306, 308–16 Ibn Wa∂∂à˙, Scholar, 241 Ibn al-Zubayr, 'Abdallàh, Anti-Caliph, 57, 114–5, 329 Ibn Zurzùr, Musician, 335 n. 48 Ibràhìm b. Óajjàj b. 'Umayr alLakhmì, Arab King of Seville, 221 Ibràhìm b. Khumayr, General, 223 Ibràhìm b. al-Mahdì, Musican and 'Abbàsid prince, 290, 292, 338, 344, 347 Ibràhìm al-Imàm, 187, 192 Ibràhìm al-Mawßilì, Scholar, 330 n. 18, 337–9, 341 Ibràhìm al-Nakha'ì, Traditionist, 89, 97, 176 n. 37 Identification – By name, 253, 258–9 – By place, 259–60 – By tribe, 256, 260 Identity – Andalusì, 197, 198, 231, 232, 234, 241 – Berber, 234 n. 118 – Islamic versus Arabic, 197, 374–5 – Jewish, 413 – Personal, 252–3, 256–7, 260–1 – Tribal, in al-Andalus, 201 – See also Ethnicity Ifrìqiyà, Place, 272–4, 320 n. 161, 439, 307 Ikhtilàf, 91, 118 'Ilhiz, Eaters of, 65 'Ilj, 'ulùj, Non-Arab(s), 353, 356 Illiteracy, 249, 251 'Ilm, Development of, 454 'Ilm al-˙adìth, Introduction of, in alAndalus, 238 Imàm, 70 n. 2, 135, 155, 158, 163 – Occultation of, 163–4 – Patronate of, 138 – Relationship with community, 134 – Relationship with individual, 135

496

index

– Relationship with Shì'a, 134–5 – Zaydì, 10 n. 27 Imam al-Bàqir, 135–6, 141, 145, 169 n. 7 Imam Ja'far, 143, 145, 148, 153, 160 n. 59, 169 n. 7 Imra" al-Qays, Poet, 350 Inheritance, 2, 72, 86, 90, 136–7, 140, 147–50, 152, 171 – Agnates and cognates, 90, 95–99, 153 – By client, 82, 100–1 – By contractual patron, 98 – By foundling, 100 n. 126 – By freedman, 102, 162 – By manumitter, 95–6, 98, 116 – By nasab, 161 – By patron, 79, 84, 95–6, 111 n. 190, 119 – By women, 100 n. 126, 115 – Exclusion of patron, 138, 153 – Exclusion of women, 113 – In case of no heirs, 86, 88 – Islamic laws of, 18 – Màlikì view on, 99 n. 122 – Of Arab and non-Arab strangers, 100 – Of caliphate, 172, 177, 179 – Of walà" by women, 137 n. 10 – Qur"ànic heirs, 95 n. 107, 96, 97 n. 114 – Qur"ànic rules, 115, 153 – Right to, 71–2, 71 n. 6, 81, 101 – Right to, after conversion, 214 – Through alliance, 3, 5 n. 9, 20, 22–3 – Through alliance, Sunnì/Óanafì view of, 21, 23 – Through walà", 27, 161 – See also Agnatic tie; Dhawù; Walà" Innovation diffusion, 260 – Access to information, 251 – Models, 250–2 – Spread in space and time, 251 Iran, Place, 330 – As Shì'ì country, 184 – Muslim community in, 260 Iraq, Place, 70 n. 2, 83, 85, 87, 98, 110, 113, 115, 117, 119, 167–8, 187, 191, 259, 263, 266, 272–3, 274 n. 58, 281, 283, 286 n. 34, 287, 290–2, 307, 318, 331 n. 29, 335, 374, 377 n. 29, 409–10, 413, 438–9, 461, 468

'Ìsà b. 'Abd al-Ra˙màn al-Umawì, Mawlà, 217 n. 77 'Ìsà b. Mu˙ammad b. Abì Khàlid al-Abnawì, Commander, 292 n. 64, 293 'Ìsà b. Yazìd al-Jalùdì, Commander, 320 al-Ißbahànì, Abù l-Faraj, Scholar, 326, 332 n. 32, 338–9, 341, 346, 355, 361, 372 n. 10 Isfahan, Place, 314 n. 138, 259, 353, 411 Is˙àq b. Ibràhìm al-Mawßilì, Musician, 326, 336–8, 340–2, 344, 347 Islam, And monotheism, 380, 409 Islamic religious sciences, 454 Islamic society, Emergence of, in al-Andalus, 221–2 Islamization process, 432 – In al-Andalus, 226, 229 – See also Arabization and Islamization; Assimilation Ism, asmà", 253, 258 – Bedouin, 352–3 n. 25 – Patterns of, 254 Ismà'ìl b. 'Abdallàh b. Yazìd b. Asad, Governor, 316 Ismà'ìl b. Jàmi', Musician, 338 Ismà'ìl b. Yasàr, Poet, 358–9 Isnàd, Asànìd, 3 n. 3–4, 70, 75–6, 83, 85, 89 n. 79, 89 n. 80, 90 n. 81, 93, 95–6, 100 n. 126, 106, 108, 110 n. 180, 114 n. 203, 115, 118 n. 215, 135, 471 – See also Common link; Matn Isnàd-cum-matn analysis, 13, 76–8 Istakhar, Place, 353 Istißràkh, Support, 11 n. 32 Itàwa, Tribute, 58, 65–6 I'tizà", Support, 11 n. 32 Ja'àla, Payment, 60–3, 66 Jaén, Place, 196, 206, 220 Ja'far b. Khàlid al-Barmakì, Vizier, 342 Ja'far b. Mùsà al-Hàdì, 'Abbàsid prince, 343 Ja'far b. Sulaymàn b. 'Alì, 342 Ja'far al-Íàdiq, See Imam Ja'far Jàhiliyya, And Islam, 7, 11–2, 18, 20–2, 27–8, 30, 32–3, 184, 188, 327–9, 333, 364, 376, 378, 410 al-JàhiΩ, Scholar, 266, 292–4, 327, 349, 357–8, 360, 365, 366 n. 118

index al-Jahshiyàrì, Scholar, 366 Ja˙Ωa al-Barmakì, Musician, 344 al-Jalùd, Place, 320 n. 161 James, W., 246 Jamìla, Musician, 333–4, 336 Jàriya b. Óumayl, Companion, 7 Jarr al-walà", 107, 119 – See also Ta∂ammun;Walà" al-Jarrà˙ b. 'Abdallàh al-Óakamì, Governor, 350 Jarrar, M., 70 n.2 al-Jaßßàß, Scholar, 3, 4 n. 5 Jawàrì, Slave-girls, at 'Abbàsid court, 343 Jayyàn, See Jaén al-Jazìra, Place, 286 n. 36, 439 Jesus, 380, 405 Jew(s), 29–30, 50–5, 57–60, 62–4, 66, 173 n. 26 and n. 27, 457 Jewish tribes, 50–1, 53 n. 6, 54–6 Jews and Christians, 266, 408–9, 411–3 – Borrowings from, 375 – Category of, 170 n. 10 – Literature, 411–2 al-Jibàl, Place, 314 n. 138, 439 Jidda, Place, 335 Jihàd, 5, 164 Jiwàr, Protection, 14 Jizya, Poll-tax, 174 n. 29, 246, 264, 271–3 – See also Taxation Joseph, Story of, 61 Ju'ayl, Tribe, 7 Juda, J., 5 n. 12, 6 n. 14, 9 n. 24, 12 n. 33, 12 n. 35, 25 n. 92, 28 n. 103, 71–2, 80, 81 n. 47, 92 n. 92, 98 n. 116, 102 n. 136, 104 n. 145, 109 n. 175, 113 n. 194 Judhàm, Tribe, 220 n. 87 Ju'fì, Tribe, 12 Juhayna, Tribe, 59 Junds, 233 – In al-Andalus, 206, 212 Junùd Khuràsàn, 289 – See also Ahl Khuràsàn Juwayn, Place, 260 Juynboll, G., 74 n. 20, 75–6, 374 n. 18, 471 Ka'b, Tribe, 24, 59 Ka'b al-A˙bàr, Traditionist, 182 Ka'b al-Ashqarì, Poet, 353

497

Ka'b b. Màlik al-Anßàrì, Tax collector, 55, 59 Ka'ba, 329 Kaffàra, Penitential recompense, 143, 149 Kahl, Middle-aged, 472 Kalb, Tribe, 50, 59–60, 61 n. 32, 62, 290 n. 51, 315 n. 146 Kayla, Annual grant, 60, 62–3, 66 Kennedy, H., 233 n. 115, 308, 339 n. 70 Kennedy, Ph., 363 Khabar, See Akhbàr Khadìja bt. al-Ma"mùn, Musician and 'Abbàsid princess, 344 Khàlid b. Ibràhìm, Abù Dàwùd, Commander, 297, 304 n. 97 Khalidi, T., 392 n. 90 Khalìfa b. Khayyà†, Scholar, 455 n. 2 al-Khalìl b. A˙mad, Scholar, 344–5, 362 al-Khalìl b. Hishàm, Commander, 288 n. 41 Khalìl al-Mu'allim, Musician, 331 Khamriyyàt, Wine poetry, 362–4 Khandaq, Battle of, 65 Kharàj, Land tax, 264, 272, 297 – See also Taxation al-Kha†ìb, al-Baghdàdì, Scholar, 26 n. 95 Khaybar, Place, 61 n. 33, 64 – Conquest of, 65 n. 43 – Jews of, 65 Khazraj, Tribe, 54 n. 8 and n. 9, 55 – King of, 51 – See also Aws and Khazraj; Anßàr Khàlid b. al-Walìd, Commander, 379 Khàlid al-Qasrì, Governor, 363 Khàrijites, 179 n. 49, 187, 288, 315 n. 146 – And Shì'ism, 185 Khidma, Service, 216–8, 231, 237, 240 Khosraw, 62, 172 Khudra/Abjar, Tribe, 32 n. 125 Khulayd b. 'Atìk, Musician, 335 n. 46 Khulayda, Musician, 334 Khums, 163–4 – See also Taxation Khurasan, Place, 167, 169, 175, 177, 185–7, 190–2, 259, 268, 272–3, 274 n. 58, 332 n. 33, 338–9, 350, 353, 377 n. 29, 403, 439, 461 – Arabs and non-Arabs of, 303, 319

498

index

– Arabs of, 281, 290–1, 295–7, 304, 313–4, 317–8 – Banawì commanders in, 285 – Garrisons in, 288 – Junùd Khuràsàn, 289 – al-Ma"mùn, governor of, 282–3, 286–90 – Non-Arabs of, 278–9, 295, 310, 312, 317–8, 320 – Prominence of, vis-à-vis Iraq, 298 – Social status of commanders in, 317, 319 – See also Ahl Khuràsàn al-Khusanì, Scholar, 241 Khuwwa, 50, 59, 61 n. 33 – See also Akh; Mu"àkhàt Khuzà'a, Tribe, 7, 16, 59 Khwarizm, Place, 288, 318 Kinàna, Tribe, 352 Kinda, Tribe, 29, 311 al-Kindì, Abù Yùsuf, Scholar, 345 Kinship tie, 120, 135–6, 140, 146–7, 225 – Change of identity, 372 – Fictitious, 81 n. 44 – See also Agnatic tie; Dhawù; Inheritance; Nasab; Walà" Kirmàn, Place, 191 Kirsà, See Khosraw al-Kisà"ì, Scholar, 437, 451 Kister, M., 15 n. 42, 60 Kitàba, Contract of emancipation, 73 n. 16, 106–7, 143 Kremer, A. von, 71–2, 79 Kubr, Next agnatic relative, 114–9, 122, 157 n. 56 – Doctrine, 113 – See also Agnatic tie; Dhawù; Inheritance; Nasab; Walà" Kufa, Place, 10 n. 26, 13 n. 39, 89–90, 174 n. 27, 259, 263, 266, 268, 274, 316 n. 147, 335, 338, 340, 342, 354, 356, 360, 363–5, 426, 436–7, 440–1 Kunya, kunan, 196, 202, 238, 253, 258 Kurds, 457 Kuthayyir 'Azza b. 'Abd al-Ra˙màn, Poet, 333 n. 34 Labor contract, 73 Lakhmid kings, 367 Landau-Tasseron, E., 81 n. 45, 109 n. 173, 197 n. 6 Laqab, alqàb, 253

Laqì†, See Foundling(s) Law – Commercial, 123 n. 219 – Customary, 71, 79, 80 n. 39, 115 n. 208, 123 n. 219 – Endowments, 134 – In al-Andalus before Màlikism, 215–6 – Islamic, 71, 73 n. 17, 74 n. 21, 74–75, 143 n. 20 – Màlikì school of, 204–5, 215, 238–40 – Marriage, 134 – Non-Arab, 123 n. 219 – Of War, 123 n. 219 – Penal, 123 n. 219 – Pre-classical, 78, 143 n. 20 – Pre-Islamic Arabic, 79–80 – Roman, 71 n. 6, 73 n. 17, 79–80 – Schools of, 475 – Shi'ì Inheritance, 153 – See also Fiqh Laylà bt. 'Alì b. al-Mahdì, Musician and 'Abbàsid princess, 344 Layth, Tribe, 203–5 Lérida, Place, 235 Lisàn, 444 Lugha, 443 Lughawì, lughawiyyùn, 247–8 Lute, Musical instrument, 329–30, 338, 345, 346 n. 107 Ma'bad al-Yaq†ìnì, Musician, 331, 340 al-Mada"in, Place, 10 n. 26, 62 – Conquest of, 315 n. 146 Madhhab, madhàhib, School(s), 437, 451, 469, 478 – See also Fiqh; Law Ma∂mùn, See Ta∂ammun Madrasa, 446, 448 Maghàzì, 76 n. 29, 372 n. 12 Magians, 457 al-Mahdì, Caliph, 278, 282 n. 18, 312, 316, 178–9, 337–8, 343 al-Mahdì b. Aßram b. 'Abd al-Óamìd, Commander, 290 n. 51 Mahdì Ibn al-Qi††, Umayyad rebel, 201 Mahmùm, Musician, 340 Majhùl, 470 Màkhùrì, Musical rhythm, 337 Makhzùm, Tribe, 29 n. 108, 32 n. 122 Maknùna, Musician, mother of 'Ulayya, 337 n. 63

index Màl-doctrine, 114–5 Malaga, Place, 220 Malik, mulk, In al-Andalus, 221 Màlik b. Abì l-Sam˙, Musician, 332, 338–9 Màlik b. al-'Ajlàn, Leader of Aws and Khazraj, 53–4 Màlik b. Anas, Scholar, 17–8, 23 n. 83, 27, 75 n. 25, 77 n. 32, 106, 204 – Social status of family, 17–8, 39 Màlik b. al-Haytham, Propagandist, 297 Màlik b. 'Iyà∂, Mawlà of 'Umar, 60 n. 30 Màlik b. ˇumama, Musician, 335 n. 44 Ma'mar b. Ràshid, Traditionist, 86 n. 64 Mamlùk, Slave-soldier, 282 n. 18, 294, 318–9 al-Ma"mùn, Caliph, 339, 341 n. 84, 343, 472 n. 14 – And al-Amìn, struggle between, 278, 283, 290, 293, 298–9, 306, 308, 317 – And scholarship, 281 – Bay'a to, 309–10 – Conquest of Iraq, 308 – Conversation with al-Fa∂l b. Sahl, 296–7, 299–303 – Discussion with al-Mu'taßim, 293 – Entering Baghdad, 291–2, 294 – Governor of Khurasan, 281, 319 – His “Persianness”, 304, 307–8, 312 – His shì'a, 287, 305, 318 – See also Abnà"; Da'wa; Maràjil; Khurasan al-Manßùr, Caliph, 282 n. 18, 311–2, 316 n. 147, 340–1, 178, 190 n. 91, 337–8, 342 Manumission, 70–1, 85, 121–2, 136–7, 142, 146–7, 150–1, 155, 181 n. 58, 196, 229 – By pre-Islamic Arabs, 79–80 – Conditional, 73 n. 16, 102 n. 133 – Gratuitous, 141 – Motivation for, 144 – Obligatory, 141, 143–4, 150 – Relationship between manumitter and freedman, 72–3 – Testamentary, 71 – Unconditional, 79, 85 – View of schools of law, 70

499

– Voluntary, 147 – See also Walà" Maqàm Ibràhìm, 176 Ma'qil, Tribe, 314 n. 138 Ma'qil b. 'Ìsà al-'Ijlì, Musician, 340 Maràjil, al-Ma"mùn’s mother, 309, 313 n. 136 Marj Ràhi†, Battle of, 268 Marriage – Mixed, 168, 170, 171 n. 14, 182, 184, 191, 222, 334 – Sunnì rules of, 184 n. 68 – See also Law Marw, Place, 281, 283–8, 290 n. 51, 291, 295–6, 298–9, 310, 314 n. 139, 317, 403 al-Marzubànì, Scholar, 351–2 Marzùq al-Íarràf/al-Îarràb, Musician, 335 n. 50 al-Masdùd, Musician, 342 Maßmùda, Tribe, 203–4 Mass media, 251–2 Mas'ùd b. Rukhayla al-Ashya'ì, Tax collector, 59 Matn, Mutùn, 70 n. 2, 75–7, 83, 87 n. 69 and n. 71, 87, 89 n. 79, 90 n. 82, 97 n. 112, 98 n. 118, 99 n. 125, 100 n. 126, 106, 108 n. 168, 109, 111, 114 – See also Isnàd Mawàlì/Muwàlùn, 226 Mawlà, mawàlì, 26 n. 95, 86 n. 63, 89, 92 n. 92, 100, 104 n. 143, 113–5, 117, 198, 203, 209, 221, 227, 235, 236, 258, 372 – Administrators, 269–71 – And Arabs, Social status of, 81, 218, 256–7, 266, 273–4, 349–50 – And ˙alìf, 25–6, 28–9, 41–4 – And international trade, 266 – And Shì'ism, 184–6 – Andalusì uniqueness of, 199 – al-'Aqd, 25, 27, 33, 81 n. 45 – Arab, 80, 86 n. 64, 119 – al-Asfal/ al-A'là, 82, 149, 161 – Assimilation process, 415, 279–80, 410 – bi-l-Óilf, 17, 25, 27–9, 33, 81 n. 45 – Borrowing money from, 267–8 – By oath, 21 n. 72 – Contempt by, 217 – Contribution of, to Arabic poetry, 359

500

index

– Conversion of, 85 – Definition of, 160 n. 59, 169 n. 5, 24–33, 216–8, 246–8, 327–8, 429, 455 – Economic status, 216 – Equality with patron, 82 – Ethnicity of patrons, 457–8 – Governors, 171, 183, 191 – In Arabic sources, 373 – In crafts and trade, 264–8 – In junds, in al-Andalus, 212 – In military service, 112 n. 194, 273–5 – In Morocco, 199 n. 11 – Influence of, on society, 360 – l-Islàm, 25 – l-'Itàqa, 25 – Judges, 171, 479 – Killing of, 265 – Kinsmen, 50 – Kitàb al-mawàlì, 349–50 – Lack of, in al-Andalus, 211 – Landowners, 271–3 – Legal status of, 121 – Length of, 168, 174, 180, 188, 192 – Listing of occupations of, 265 – Loyalty of, 265, 358–9 – Manumitted slaves, 81, 112, 279 – Marriage with Arab, 213 n. 60 – Musicians, 331 – al-Muwàlàt, 25 – Non-Arab, 80, 112, 119 n. 216 – Non-Umayyad, in al-Andalus, 231 – Of a woman, 196, 236 n. 123 – Of Banì Abù 'Ìsà, 204 – Of caliphs, 267 – Of God and the Prophet, 31–2 – Of individuals, 256 – Of mawlà, 171 n. 11, 250, 458, 477 – Owning slaves, 266 – Patrons, 266 – Perceived as being Arabs, 198 – Poets, common denominator, 359 – Poets, lack of group solidarity, 358 – Praying behind, 171 n. 14 – Prejudice against, 237, 356 – l-Qabìla, 25 – Rebellions of, 219, 264 – Rejection of change, 108 – Rights and duties of, 72, 83–116 – Sale of, 85, 111 – Social mobility of, 458

– Social status of, 16, 20, 25, 30, 81 n. 47, 112, 121, 122 n. 218, 181–2, 184, 212, 218, 238, 246–7, 262, 264–5, 269, 278–80, 351, 371 – Soldiers, 171 – Status of, before Islam, 52 – Tribal, 255 – Umayyad, 198, 205, 212, 214–7, 230 – Umayyad, in al-Andalus, 225–6, 239–40 – Urbanized versus non-urban, 371 – Wealthy, 267–8, 274 – Without walà", 197 n. 5 – al-Yahùd, 53 – al-Yamìn, 28, 33, 81 n. 45 – See also Walà" Mawlà and Arab 'ulamà" – Comparisons between, 478–9 – Dyeing beards, 471 – Geographical distribution, 478–9 – Occupations, 478–9 – Percentage distribution of occupations, 460 – Percentage distribution in Southern and Northern tribes, 460 – Physical features of, 478–9 – Social status of, 478–9 Mawlà 'ulamà" – Average life span of, 477 – Ethnic background of, 476–7 – Geographical distribution of, 461–8 – Ideologies of, 477 – In secular sciences, 478 – Lifespan of, 472 – Networks of teachers and pupils, 476, 478 – Occupations of, 477 – Percentage distribution of classifications, 469–70 – Percentage distribution of ideologies, 469 – Percentage distribution of physical features, 471–2 – Percentage distribution of tribal affiliations, 458–60 – Percentage distribution of, in ˙adìth collections, 470–1 – Percentage distribution of, in 'ilm, 473–5

index – Percentage distribution of, in knowledge put to writing, 475 – Percentage distribution of, in secular versus sacred sciences, 474–5 – Physical features of, 477 – Previous religion, 476 – Religious duties of, 472 – Tribal affiliations of, 477 – See also 'Ulamà" Mawlàh, 328 Maxim(s), 78, 111, 115, 117 Maymùna, Wife of Prophet, 110 Measurements, Íà' and mudd, 176 Mecca, Place, 5, 7, 11, 32, 79, 81 n. 45, 87, 101, 115 n. 206, 173 n. 26, 259, 283, 309, 330, 331, 333–6, 413 Medina, Place, 11, 13 n. 39, 17, 31–2, 51–3, 55–60, 61 n. 33, 62, 64, 66, 79, 81 n. 45, 87, 96–7, 115, 118, 174 n. 29, 267, 315 n. 146, 330–1, 333–4, 336, 340, 342, 364, 370–1, 395–6, 405, 414 – Constitution of, 29–30, 54, 109, 109 n. 174 – Jews of, 30, 61 n. 31 – Pre-Islamic, 51, 54, 58–9 – Siege of, 65 – See also Yathrib Mérida, Place, 220 Messianic expectations, 201, 231 Metrics, 'Arù∂ and ìqà', 335 n. 49, 344–5, 363 Migrational waves, 478–9 Mi˙na, 472 n. 14 Military commanders, Social status of, 286–7, 289–90 Military families, 288–90 Min ahl, 437–8, 461, 468, 479 al-Miqdàd, Companion, 32 Mishra˙ b. Ma'dì Karib, King of Kinda, 312 Mißr, Place, 439, 468 Missionary activities, 229, 246, 251 Mizhar, Musical instrument, 333 n. 36 al-Mizzì, Scholar, 471 Molina, L., 195, 238 Mosul, Place, 182, 316, 439 Mottahedeh, R., 308 Motzki, H., 10 n. 26, 13, 76, 114 n. 204 Mozarabs, 228, 232 Mu"addib, 447–8

501

Mu'àhid, Pact, 208 Mu"àkhàt, Brotherhood, 9, 22 – See also Akh; Ukhuwwa al-Mu'allà b. ˇarìf, Musician, 338 Mu'allim, 447–8 Mu'àwiya, Caliph, 29, 73, 167–9, 170 n. 9, 171 n. 12, 172 n. 16 and n. 19, 173 n. 23, 174, 176 n. 38, 180–2, 187, 192, 265, 267, 397, 411 Mu'ayqìb b. Abì Fà†ma al-Dawsì, Ally/Mawla, 32 Mudabbar, See Tadbìr Mu∂ar, Tribe, 170, 187–9, 191–2, 203, 214 n. 67, 227 n. 98, 281 al-Mufa∂∂al b. Salama, Scholar, 345 al-Mufìd, Scholar, 156 Mughannì, Definition of, 326 n. 2 Mughanniyàt al-qußùr, 346 al-Mughìra b. Óabnà", Poet, 353–4 al-Mughìra b. Shu'ba, Companion, 32 Muhàjir, muhàjirùn, 7, 9, 19, 21 n. 72, 22, 29 al-Muhallab, 353 al-Muhallab b. Abì Íufra, Commander, 268 Mu˙ammad, See Prophet Mu˙ammad Mu˙ammad, Andalusian emir, 225–6, 227 n. 98 Mu˙ammad al-'Abbàsì, Sheikh al-Azhar, 373 Mu˙ammad Amìn al-Mahdì, Mufti of Egypt, 373 Mu˙ammad al-Anqar b. 'Abd al-Ra˙màn al-Tujìbì, Arab, 214 Mu˙ammad al-Bàqir, See Imam al-Bàqir Mu˙ammad b. al-'Abbàs b. Musayyab b. Zuhayr, Commander, 285 Mu˙ammad b. 'Abd al-Salàm al-Khushanì, Traditionist, 236–8 Mu˙ammad b. 'Abdallàh, Great-grandson of Caliph 'Uthmàn, 334 Mu˙ammad b. 'Abdallàh b. al-Óasan al-Nafs al-Zakiyya, Rebel, 178, 316 n. 146 Mu˙ammad b. 'Abdallàh al-Óazanbal, Scholar, 363 Mu˙ammad b. Abì l-'Atàhiya, Son of poet, 352 Mu˙ammad b. A˙mad b. al-Zarràd, 237 Mu˙ammad b. 'Alì, 188 Mu˙ammad b. al-Ash'ath b. Fajwa al-Kùfì, Musician, 338

502

index

Mu˙ammad b. Óamza b. Nußayr, Musician, 340 Mu˙ammad b. al-Óàrith b. Buskhunnar, Musician, 340 – And family, 341 n. 82 Mu˙ammad b. Óumayd b. 'Abd al-Óamìd, Commander, 290, 319–20 Mu˙ammad b. Idrìs, Scholar, 144, 150, 159 n. 57, 163 Mu˙ammad b. Lubb, 214, 236 Mu˙ammad b. Nù˙ al-'Ijlì, Scholar, 472 n. 14 Mu˙ammad b. Sulaymàn b. 'Alì, 342 Mu˙ammad b. Sulaymàn b. Talìd al-Mà'àfirì, Qà∂ì, 235 Mu˙ammad b. Wa∂∂à˙, 237–8 Mu˙ammad b. Yasìr al-Riyàshì, Mawla (?), 352 Mu˙ammad b. Ziyàd al-Lakhmì, Qà∂ì, 238 Mu˙ammad Óasan al-Najafì, Scholar, 147 Mu˙ammad Na'ja, Musician, 335 n. 45 Mu˙ammad, Son of Caliph al-Saffà˙, 337 al-Muhaqqiq, al-Óillì, Scholar, 151 , 152, 154, 156, 158–9, 160, 163 Mu˙dath, Novel style of poetry, 359, 362, 364–6 – Visual lipogram, 361 – See also Arabic poetry; Poets al-Muhtadì, Caliph, 295 n. 73, 337 Mukàtab, Slave with contract to buy freedom, 110 n. 183, 112–3, 116 – Selling of, 85 – See also Kitàba; Manumission Mukhannathùn, Effeminates and musicians, 336 Mukhàriq, Musician, 340, 342 al-Mukhtàr, Rebel, 185, 274 Muladí, Muladíes, 219 n. 84 al-Munajjam, Tribe, 344 n. 99 Muràd, Tribe, 268 Murji"a, 469, 477 Murra, Tribe of, 65 Mùsà b. Ka'b, Commander, 297, 304 n. 97 Mùsà b. Khàrija, Musician, 335 n. 45 Mùsà al-Ghàfiqì, Tribe, 213 n. 60 Mùsà al-Salmànì, Mawlà, 267 Mùsà Shahawàt, Poet, 352, 358 Muß'ab b. al-Zubayr, Commander, 57

Musàlima, Andalusi converts, 196, 205, 208, 223, 228 – Ahl al-dhimma, 196, 208 Musàwir al-Warràq, Poet and traditionist, 352, 364 al-Muß˙fì, Tribe, 202 Music – And Greek philosophy, 345 – Arabic medieval sources, 327 – Byzantine, 330 – Contribution to, by men and women, 346–7 – Festivals, 333–4, 336 – Historiography, 344–5 – Instruments, 329 – Persian, 329–30 – Pre-Islamic, 330–1, 346 – Scales, 345 – Schools of Singing, 338 n. 68 – Styles, 326, 328, 330, 336 – See also Top Hundred Songs Musician(s) – Occupations of, 331–2, 335, 338 – Social status of, 327, 330–1, 333–5, 337, 339–42, 344, 346–7 Muslim b. al-Walìd, Poet, 352–3, 360, 366 n. 117 Mustamlì, Clerk, 447 n. 24 Musta'rab, Mozarab, 228 Mu"ta, Battle of, 405 al-Mu'ta∂id, Caliph and musician, 339, 345 Mu'taq/mu'tiq, Freedman, 115, 160 n. 59 Mutaßaddir, 447–8 al-Mu'taßim, Caliph, 292 n. 62, 294, 295 n. 73, 318–20, 342–3 al-Mutawakkil, Caliph, 179, 343 Mutayyam al-Hishàmiyya, Musician, 340 n. 75, 343 n. 92 Mu'tazila, 469, 477 al-Mu'tazz, Caliph, 342 al-Mu††alib, Tribe, 19 Muwallad, muwalladùn, Modern poet, 361 n. 83 Muwalladùn, 196, 208, 209 n. 42, 213 n. 63, 218–9, 221–3, 239–40 – Analysis of term, 228 – Compared with mawlà and 'ajam, 220 – Màlikì view on, 227 – Muslim, 228–9, 232

index – Rebellions of, 222, 224–7, 231, 233, 240 – Social status of, 222, 227, 233 – Support for, 235 Muzayna, Tribe of, 59 Muzdalifa, Place, 408 Nabataeans, 92 n. 92 , 356 al-Nàbigha, al-Dhubyànì, Poet, 366 Nadhr, Oath, 149 – Breaking of, 142 Na∂ìr, Tribe, 51, 54–5 Nàfi' b. Jubayr b. Mu†'im, Arab, 171 n. 14 al-Nafs al-Zakiyya, See Mu˙ammad b. 'Abdallàh b. al-Óasan Na˙w, 443, 454, 473–4, 477 Na˙wì, na˙wiyyùn, 427–8 al-Najafì, Scholar, 164 n. 66 Najd, Place, 60 – Road of, 64 al-Namir b. Qàsit, Tribe?, 28 Naqìb, nuqabà", 'Abbàsid, 282, 297 n. 84, 303, 304 n. 97, 318, 403 Nasab, 113 n. 196, 202, 253, 258, 458 – Doctrine, 118 – Importance of, 281 – Kinship tie, 134, 140 – Majhùl 152 – Sale of, 111 – See also Descent groups; Kinship tie Naßb, Singing, 328, 331, 335 n. 44 Nashì†, Persian singer, 329 Nàßi˙ b. Yltìt b. Qa†arì, Berber, 196 Naßr b. Sayyàr, Governor, 189, 191 Naßr b. Shabath, Rebel, 293–4 al-Nawàr, Musician, 335 Naw˙, Singing by women, 328, 336 New Testament, 379, 402–3, 411, 414 Niebla, Place, 206 Nihawand, Place, 411 Ni'ma, Favor, 85, 100, 115 Nisba, nisab 2, 253 – Arabic, 1, 197–8, 202, 211, 213, 217–8, 220 n. 87, 230, 236, 240, 249 – Aslàmì versus Islàmì in al-Andalus, 208 n. 41 – Berber, 198, 202 – Ethnic affiliation, 196 – Geographical, 202, 257–60, 352 – Islamic/religious, 197, 236

503

– Tribal, 197, 208 n. 41, 211, 218, 250, 255–6, 352 – Uncommon, 260 – Walà", 196 Nishapur, Place, 259–1 Nizàr, Tribe, 191, 281 Nock, A.D., 246 Non-Arab(s) – A'àjim, 168–9 – In the army, 280, 282 – Muslims, 167–70, 173, 175, 415 – Rebellion, 219–20 – Social status of, 173, 185 – See also Arabs and non-Arabs; 'Ajam Noth, A., 74 n. 20, 391–2, 410–11 Nu'aym b. Khàzim b. Khuzayma al-Tamìmì, Commander, 284, 286 n. 36, 312 Nu∂ayb b. Rabà˙, Poet, 356, 358 al-Nu'màn b. Bashìr al-Anßàrì, Daughter of, 57 n. 14 al-Nu'màn b. al-Mundhir, King of al-Óìra, 61 n. 32 Oath, See Nadhr Obesity, 472 n. 15, 478 Old Testament, 402–3, 411 Oliver, D., 197 n. 5, 216, 218–21, 224, 226–8 Oman, Place, 58 n. 26 Oviedo, Place, 237 Pandore, Musical instrument, Longnecked lute, 331, 337, 342 Paramonè, 73 n. 16 Parry, M., 392 Patron(s) – Change of, 92, 107, 109–10, 123 n. 220 – Choice of, 88 – Female, 96, 102–3, 156, 158–9 – Of poets, 365–7 – Rejection of change, 108 – Rights and duties of, 83–116, 144 – Status of, 96 – See also Walà" Patronate – Classical Roman, 79 – Late-Roman, 73, 79 – Roman and Islamic, 71 – See also Walà" Pellat, Ch., 455 n. 4

504

index

Persians, 172 n. 16, 182 n. 60, 302–3, 317, 351, 357, 403, 457, 477 – See also 'Ajam; Asàwira; Óamrà" Pipes, D., 92 n. 92, 112 n. 194 Poets – Called Mu˙ammad, 351–2 – Classification of, 349 – Payment of, 365–7 – Social status of, 366 Powers, D., 115 n. 209 Presentation conversion, 249–50, 256, 261 Prisoners of war, 248, 250, 261 Prophet Mu˙ammad, 2, 4, 5 n. 12, 6–9, 11, 13, 19–20, 23, 29–32, 51, 54, 56–9, 61 n. 33, 63–5, 74, 77–80, 82, 84 n. 58, 85, 86 n. 63, 96–7, 100, 109 n. 172 and n. 174, 110, 118–9, 123, 135–6, 146, 170 n. 10, 171 n. 14, 175–8, 180, 187–8, 190, 247, 253–4, 256, 258, 266, 328, 350 n. 10, 371, 379, 396, 397–8, 403, 405, 407, 413–4, 458 – And family, 168, 185–6, 305 – Kinship to, 172 – Violating religious principles, 8–9 – See also Ahl al-bayt Qabà" dìbàj, Persian clothing, 353 Qadar, Divine fore-ordainment, 396 n. 106 Qadariyya, 469, 477 al-Qàdisiyya, Battle of, 24 n. 88, 404 Qafà l-Najjàr, Musician, 335 Qa˙ba, Prostitute, 62 n. 36 Qa˙†aba b. Shabìb, Commander, 191, 297, 304 n. 97 Qa˙†àn, Tribe, 10, 214 n. 67 Qa˙†ànì, Southern messiah, 316 Qalansuwa, 471 Qasàma, Oath, 13, 23 – See also Nadhr Qasì, Tribe, 213–4, 234, 236 Qàßß, 447 n. 24 Qawm, People, tribe, 52, 80 n. 40, 87 n. 71, 101, 121 Qayna, Slave-girl, 328 Qays, Tribe, 268, 280, 316 n. 146, 352 Qays b. Makhrama b. al-Mu††alib, Arab, 370, 372 n. 12 Qays b. Walì'a al-Kindì, Rebel, 311–2 Qayßar, See Caesar al-Qif†ì, Scholar, 429

Qirà"a, 444–4, 473–4, 477–8 Qiyàs, 142, 451 Qu∂à'a, Tribe, 24, 62–3 Quellenlage, Sources, 74–6 Qùmis b. Antunyàn, Secretary, 210 n. 49 Quràba, Family member of slave, 138 – See also Dhawù; Inheritance Qur"àn – 3:140, 233 – 4:33, 12, 20 – 4:43, 176 n. 37 – 4:92, 141 – 5:1, 4, 27 n. 99 – 5:2, 5 – 5:6, 176 n. 37 – 5:89, 142 – 8:75, 21 – 9:29, 53 n. 5 – 17:33, 172 n. 20 – 24:59, 174 n. 29 – 33:5, 107, 159 – 33:6, 21 – 42:23, 58 n. 21 – 58:3, 142 – 79:24, 224 – 106, 12 Qurayba, Sister of Thurayya, 334 n. 39 Quraysh, Tribe, 5, 7, 12, 14–5, 17, 20, 24, 28–9, 31–2, 54, 170 n. 10, 172, 185, 238, 332, 342, 352, 360, 370, 458 n. 6, 459, 477 QurayΩa, Tribe, 51, 54–5 Qußayy, Ancestor of prphet Mu˙ammad, 56 n. 11 Qu†ay'a, Tribe, 268 Qutayba b. Muslim, Commander, 315 n. 146, 332 n. 33 Quwwàd, In al-Andalus, 223 al-Raba∂ revolt, 204 Rabb/Sayyid, Title in al-Andalus, 224 Rabì' b. 'Àßim al-'Uqaylì, Chieftain, 12 Rabì' b. Nizàr, Tribe, 10, 170, 189 al-Rabì' b. Yùnus, Chamberlain, 341 Rabì'a, Tribe, 12, 187–8, 191, 297, 303, 304 n. 97, 312 Radd, Màlikì view on, 96 and n. 109 Ràfi' b. al-Layth, Rebel, 282–3, 288, 291 n. 62, 296, 314 n. 142 Ràfi∂iyya, 177–9 al-Rà'ì, Poet, 357

index Ramal, Musical term, 330, 336 Ràshidùn, 12–4, 19 Ra††àb al-Jiddì, Musician, 335 Raw˙ b. Zinbà' al-Judhàmì, Chieftain, 57 Rayy, Place, 259, 285, 286 n. 36, 288–9, 332 n. 33 Rayyo (Malaga), Place, 206 Reinert, W., 80 n. 39, 81 n. 45, 81 n. 44 Ridda, Wars, 10, 14, 61 n. 33, 305 al-Risàla, fì l-khamìs, 304–5 Riyàsh, Tribe, 352 Rowson, E., 336 al-Ru"àsì, Scholar, 437, 451 Rubay˙a, Musician, 342 Ru∂ayyà, Sister of Thurayyà, 334 n. 39 Rukbànì, Singing, 328 n. 10 al-Rukhkhaj, Place, 457 Rùstàq, Rural districts, 260 Ruzayn al-'Arù∂ì, Poet, 361 al-Íabbà˙, Great-grandfather of Abù Nuwàs, 350 Sabrì†, Tribe, 236 Sa'd, Tribe, 62 Sa'd b. 'Ubàda l-Anßàrì, Chieftain, 58, 197 n. 7 Ía'da, Place, 10 n. 27 Íadaqa, Alms, 63 n. 40, 105 n. 154 Sadùs, Tribe, 352 al-Saffà˙, See Abù l-'Abbàs al-Íaffà˙ Íaffùriyya, Place, 173 Íafiyya tradition, 103, 113–4 Sàfila, Lower Medina, 56 – See also Medina; Yathrib Ía˙àbì, ßa˙àba, Companion(s), 54, 60 n. 29, 74, 77–8, 82, 89 n. 79, 93 n. 94, 96, 110, 113, 117–20, 122–3, 253, 259, 470 – Ethnic distribution of, 470 n. 12 Íà˙ib sunna, 470 Ía˙ìfa, 109 n. 174 Sahm, Tribe, 352 Sà"ib Khàthir, Musician, 329, 331 Sà"iba, sawà"ib, 102, 104–6, 110 n. 182–3, 112, 117, 135, 137, 141–4, 146, 148–52 – Blood-money and inheritance, 111 n. 188 – See also Blood-money; Inheritance; Tasyìb

505

Sa'ìd b. Abì Hind, And family, 32 n. 125 Sa'ìd b. Manßùr, Traditionist, 86 n. 64, 87 n. 69 Sa'ìd al-Dàrimì, Musician, 332 Sa'ìd al-Hudhalì, Musician, 331 al-Salaf, 180 Salama, Tribe, 235 Íalìbatan, 352 Íàli˙ b. al-Rashìd, 'Abbàsid prince, 286 n. 35, 340 Íàli˙ Íà˙ib al-Mußallà, Commander, 282 n. 18 Salima, Clan of Khazraj, 55–6 Sallàm b. Abì l-Óuqayq, Abù Ràfi', Raid against, 61 n. 33 Sallàma al-Zarqà", Musician, 342 Sallàmat al-Qass, Musician, 334 Salm b. A˙waz al-Tamìmì, Commander, 185 Salmàn al-Fàrisì, Companion, 171 n. 14 Samà', 451 Samarqand, Siege of, 289 n. 45 and n. 48 – See also Ràfi' b. al-Layth Samarra, Place, 294, 342 al-Samaw"al, Poet, 412 al-Samhùdì, Scholar, 55 Samura b. Jundab, Companion, 32 n. 125 Saragossa, Place, 213–4, 235–6 al-Sarakhsì, Scholar, 85 n. 62, 86 n. 63, 98, 101 n. 132, 112 n. 191, 116 Íarf, Morpho-phonology, 428 Sassanians, 51, 283 n. 18 – And music, 347 Satan, Iblìs, 177 Sayf b. 'Umar, Scholar, 371, 377, 379–80 Say˙àn, Companion, 14 Sayyid/rabb, Title, in al-Andalus, 223 Schacht, J., 73 n. 18, 74–5, 78 n. 34, 82 n. 50, 95, 96 n. 110, 100 n. 126 Schoeler, G., 76, 359, 364 }entürk, R., 455 n. 4 Sergeant, R.B., 53 Seville, Place, 206, 220–1 Sezgin, F., 349, 353 Shaban, M., 307–8, 372 n. 9 al-Sha'bì, Traditionist, 98 n. 117 Shabìb b. Óumayd b. Qa˙†aba, Commander, 284

506

index

Shabìb b. Wàj, Commander, 282 n. 18 al-Shàfi'ì, Scholar, 20, 22, 103 n. 141, 106 Shahiyya, Musician, 334 Shàjì, Musician, 339 Shàm, Place, 439, 461, 468 Shàmiyyùn, See Syrians al-Shammàsiyyàt, Khulayda, Rubay˙a and 'Uqayla: Musicians, 336 al-Sharìd, Tribe, 62 Shàriya, Musician, 342 Shayba b. 'Uthmàn, Arab, 29 Shaykh/Muqaddam, Title in al-Andalus, 223 Shì'a, 307, 469, 477 – 'Alids, 178–9 – Hàshimites, 177, 178 n. 9 – Imams, 403 Shuhda, Musician, 335 Shùrà, Consultation, 172 – In al-Andalus, 204 – Shì'ì view of, 172 n. 20 Shura˙bìl b. Óasana, Companion, 13–4 Shuray˙, Traditionist, 107 n. 162, 114 Shur†a, 286, 318 n. 154 Shurù†, Legal documents, 123 n. 219 Shu'ùbiyya, 233, 241, 303, 351, 357–60 – In al-Andalus, 232 Sìbawayh, Scholar, 437, 450 Sidonia, Place, 206 Sijistàn, Place, 457 Sinàd, Singing by women, 328 Sinàn al-Kàtìb, Musician, 331 Sind, Place, 354 al-Sindì b. al-Óarashì, Commander, 284 n. 25 and n. 27 Sìra, 371, 375, 405 – Literature, 70 n. 2 – Of caliph 'Umar, 171 Sìrìn, Musician, 333 n. 36 Sistàn, Place, 191, 285 n. 29, 286 n. 36, 290 n. 51 Siyà†, Musician, 337 Slave(s), Arab, 84 n. 58 – Military, 112 n. 194 Soldiers, Payment of, 171, 181, 191 Sourdel, D., 298 Steppat, F., 307 Stiehl, R., 54 Stranger(s), 100, 121 – Arab, 86, 88, 99, 101, 103, 104, 108 n. 167, 112, 120 , 122

– Arab and non-Arab, 80 n. 41 – Non-Arab, 86, 88–9, 99, 103, 108 n. 167, 112, 119, 120, 122 – See also Ajam; Asàwira; Berbers; Byzantines; Óamrà"; 'Ilj; Kurds; Mawlà; Non-Arabs; Persians; Turks; Walà" Íubay˙, Mawlà, 32 Succession, Dynastic, Shì'ì view on, 172 n. 16 Successors, See Tàbi' Sufis, 373 Sufyàn b. 'Uyayna, Traditionist, 81 n. 42 Íuhayb b. Sinàn, Arab mawlà, 28–9 Suicide, Unintentional, 142 Sulaym, Tribe, 11, 12 n. 33, 62–4, 315 n. 146, 333 Sulaym b. Qays al-Hilàlì, Follower of 'Alì, 167–9, 176 n. 39, 177, 181–6, 190–2 Sulaym b. Sallàm, Musician, 342 Sulaymàn, Caliph, 171 n. 12 Sulaymàn, Musician, brother of Bàbawayh, 335 n. 45 Sulaymàn b. 'Abd al-Malik, Eunuch of Caliph, 62 Sulaymàn b. al-Qassàr, Musician, 342 Sulaymàn al-Wàdì, Musician, 335 n. 46 al-Íùlì, Abù Bakr Mu˙ammad b. Ya˙yà, Scholar, 326, 344 n. 96, 352 Íumàdi˙, Tribe, 220 n. 87 al-Íumayl b. Óàtim, Commander, 233 Sunna, 297, 299 – Of caliph 'Umar, 169–71, 173, 175 – Prophetic, 176, 185 al-Suwàriqiyya, Place, 63–5 al-Suyù†ì, Scholar, 429 Syria, Place, 70 n. 2, 83, 119, 173, 264, 272, 330, 334, 374, 377 n. 29, 381, 395, 398, 403, 405, 409–11, 414 Syrians, In al-Andalus, 204, 206, 209–10, 216–7, 220 ˇabàba, Musician, 334 ˇabaqàt, Classes, generations, 252–4 – See also Biographical dictionaries al-ˇabarì, Scholar, 9 n. 24, 11 n. 32, 20 n. 70, 296, 312–3, 327, 406 Tàbi', Tàbi'ùn, Successors, 17, 25, 74, 77–8, 82, 87, 96, 97 n. 113, 105–6,

index 113–4, 117–9, 122, 253, 259, 470, 477 – Ethnic distribution of, 470 n. 12 Ta∂ammun al-jarìra, Contractual walà", 149–152 Tadbìr, Form of manumission, 147 Tafsìr, 395, 444, 454, 473–4, 477–8 ˇàhir b. al-Óusayn, Commander, 281, 287, 289, 290, 317–20 – And family, 291, 315 ˇà"if, Place, 8, 13 n. 39, 31, 331, 335 ˇà"ifa Kingdoms, 201, 241 ˇalab al-ghinà", 330 n. 18 ˇal˙a b. 'Ubaydallàh, Companion, 17 Tambourine, Musical instrument, 329, 330 n. 21 Tamìm, Tribe, 59, 249–50, 297, 303, 304 n. 97, 332, 341, 459, 477 Tanùkh, Tribe, 316 n. 146 and n. 149 ˇàriq b. Ziyàd, Commander, 204, 394 Ta'ßìb, Agnacy, 163, 165 – Reappearance of, 156–7 – See also 'Aßaba Tasyìb, 83 n. 53, 85, 102, 104, 111 n. 190, 116–7, 143, 145, 149–51 – Acceptance of, 118 – Prohibition of, 118 – Rejection of, 105–6 – Shì'ì view of, 104 n. 147 – See also Sà"iba Ta†awwu', Gratuitous or voluntary manumission, 144 ˇawrà", Musician, 334 Taxation, 59, 66, 123 n. 219, 191, 206, 222, 225, 230, 239, 246–8, 264, 269, 271, 273, 313–5 – And conversion to Islam, 271–2 – See also Jizya; Kharàj; Khums; 'Ushr Tayammum, 176 Taym, Tribe, 17–18, 28, 172, 352 ˇayy, Tribe, 290 n. 51, 332 Thàbit b. Óazm, Berber scholar, 197, 203, 215, 236 Thàbit Qu†na, Poet, 352 Thaqìf, Tribe, 7, 55 n. 10, 173, 352 Theophanes, Byzantine historian, 375 Theophilus of Edessa, 381 Thomas, Apostle, 379–80 Thurayyà, Beloved of 'Umar b. Abì Rabì'a, 334 n. 39 Thuwwàr, 221 Top Hundred Songs, List of, by Is˙àq al-Mawßilì, 326, 332 n. 31, 334 n. 39, 335

507

– See also Music Topoi, 410–1 Tradition(s), See Óadìth Transoxania, Place, 295, 318, 439, 457 – Non-Arabs of, 310 Tribal affiliation, See Nasab Tribal structure, 14–5, 18–9, 268, 351, 375–6 – Principle of personal responsibility, 22 – Values, 378–9 – See also Nasab Trojan war tradition, 380–9 – Aeneid, 383 – Chronicle of Theophilus of Edessa, 388–9 – De exidio troiae historia, 385–7 – Ephemeris belli troiani, 383–5 – Epic cycle, 382–3 – Iliad and Odyssey, 381–2 – Motives and paradigms, 390–1 – Oration 11, 385 – Posthomerica, 387–8 Tubba', King of Óimyar, 57 Tudmìr, Place, 206 Tujìb, Tribe, 220 n. 87, 235–6 Turàrkhudà, Commander, 282 n. 18 Turks, 288, 289, 293, 318, 350 ˇùs, Place, 283–4, 298, 303, 310 al-ˇùsì, Mu˙ammad b. Óasan, Scholar, 140, 144–7, 149, 152–7, 159–61, 163 Tyan, E., 72 ˇu'ma, Portion of annual crop, 61 n. 33, 62–3, 97 ˇunbùr, Musical instrument, 337 ˇuways, Mukhannath musician, 336 'Ubayd, Father of Ziyàd b. Abìhi, 173 'Ubayda, Musician, 342 'Ubaydallàh b. 'Abdallàh b. ˇàhir, Musician, 339–40, 344 'Udhra, Tribe of, 62–3 U˙ud – Battle of, 329, 396 – Mountain, 6 'Ujayf b. 'Anbasa, Commander, 320 – And family, 261 and n. 62 al-'Ujayr al-Salùlì, Poet, 362 Ukl, Portion of annual crop, 62–3 'Ulamà", 200, 234, 307, 426–7, 441–2, 451 – Average life span, 455, 457 – Generational distance, 435–6

508

index

– Percentage distribution of ethnicity over time, 456 – Pre-Islamic religious background of, 457 – Ulama Project, 427, 454–5 – See also Mawlà and Arab 'ulamà"; Mawlà 'ulamà" 'Ulayya bt. al-Mahdì, Musician and 'Abbàsid princess, 344 'Ulfà, Share of harvest, 62 'Umar b. 'Abd al-'Azìz, Caliph, 13, 19, 21 n. 72, 27, 98 n. 118, 182, 272, 274 n. 58, 315 n. 146, 332, 350 'Umar b. Abì Rabì'a, Poet, 333 n. 34, 334 n. 39 'Umar b. Óafßùn, Rebel, 219 'Umar b. al-Kha††àb, Caliph, 12–3, 21, 24, 60 n. 30, 63, 79–80, 86–7, 89 n. 79, 93–4, 97 n. 112, 99–101, 103, 105, 113, 167–70, 172, 174–83, 182, 185, 188, 192, 267, 395–6, 398 'Umar b. Shabba, Scholar, 352 'Umar al-Maydànì, Musician, 342 'Umar al-Wà∂, Musician, 331 'Umàra b. al-Walìd b. al-Mughìra, Arab, 29 n. 108 Umarà" al-thaghr, 201 Umayya, Tribe, 173, 216–7 Umm 'Abdallàh bt. 'Ìsà b. 'Alì, Musician and 'Abbàsid princess, 344 Umm Abìha bt. al-Rashìd, Musician and 'Abbàsid princess, 344 Umm Ja'far, Musician, 333 Umm Mùsà bt. Manßùr, Mother of al-Mahdì, 312, 316 Umm 'Uthmàn, Mother of caliph al-Mahdì, 334 n. 39 Universalism – And Shì'ism, 188 – In Islam, 180, 183, 186 – Lack of, 185 Upper Frontier, al-Andalus, 213, 214 n. 68, 239, 241 – Arabs, 234, 236 – Berbers, 197, 202 n. 21, 203, 205 – Muwalladùn, 220 n. 87 'Uqayl, Tribe, 7–8, 12 al-Uqayshir, Poet, 362 'Uqba b. al-'Abbàs b. Ja'far, Commander, 286 'Uqba b. Abì Mu'ay†, Archenemy of prophet Mu˙ammad, 173 n. 27

'Urwa b. Mas'ùd, Chieftain, 31 'Urwa b. al-Zubayr, Traditionist, 110 'Ushr, Tithe, 264, 272 – See also Taxation al-'Utbì, Andalusian jurist, 210 'Uthmàn b. 'Affàn, Caliph, 11–2, 29, 172, 175, 179, 265, 267 'Uthmàn b. Óayyàn al-Murrì, Governor, 333 n. 35 'Uthmàn b. 'Ubaydallàh al-Taymì, Arab, 17 'Uthra, Tribe, 268 'Uyayna b. Óißn, Chieftain, 61 n. 33, 65–6 Versteegh, C., 446 n. 23 Visigoth(s), 213, 221–2, 224, 226–7, 231, 240 Visigothic Spain, Arab invasion of, 394 Wàdì l-Qurà, Place, 51, 62–3, 331, 335 – Jews of, 59 Wagner, E., 350 Wàjibàt, Obligatory reasons, 142 Walà", 14, 70, 73–4, 78, 81, 88 n. 74, 90, 97, 107, 134, 144, 148 – After conversion, 25, 73, 80, 92 n. 92, 99, 119, 197, 213–4, 239, 246, 250, 256–7, 261–2 (See also Walà" without conversion) – After death, 138, 155 – After manumission, 71, 135, 138, 141, 152, 155, 159, 211–2, 215–6 – Agnatic tie, 117, 212 Fictitious, 152 Equal to nasab, 140, 146–7, 150, 156 – Alienation of, 79, 111 – Andalusì uniqueness of, 201 – Conditional, 149 – Different forms of, 264 – Disappearance of institution, 250, 261 – Dissolution of, 73, 85, 104, 106, 116, 122, 143, 215 – Early development of, 116–120 – Early regulations of, 83–116 – Encouragement of, 217 – Half-walà" relationship, 161 – In Imàmì akhbàr, 140–7 – Individual, 72, 79–80, 119, 121 – Inheritance of, 73, 79, 113–5, 117, 122, 156–9, 165, 457–8

index By women, 116 Complexity of laws, 162 Imàmì and Sunnì view, 157, 161 – Insignificance of, 218 – (In)voluntary, 92, 108, 119, 248 – Islamic, 73, 79 – Kinship tie, 111, 117–8, 165, 200 – Legal stipulations, 70, 83 – Likeness to Imàma, 163 – Loyalty, 84 – Non-contractual, 136 – Of Arab and non-Arab strangers, 80, 100, 120 – Of children, 107 – Of foundling(s), 80, 86, 93–5, 119 – Of free person(s), 84, 92 – One-sided, 182 – Origin, 50–1, 70–83 – Patron-client relationship, 81, 148–51 – Pre-classical, 73 – Pre-Islamic, 50, 71–2, 81, 121 – Property, 118, 122 Gift of, 73, 110–1, 117–8, 122, 123 n. 220 Sale of, 73, 81 n. 42, 84, 110–1, 113, 117–8, 122, 267 Transfer of, 31 n. 118, 83, 91, 104, 159–60, 162 – Rejection of, 94 – Right to, 84, 85 – (Semi-)collective, 72, 80 – Shì'ì view, 107, 116, 153 Imàmì, 103 n. 142, 116, 136, 139–40, 147–9, 152, 156, 162, 164 – Sunnì view, 137–8, 140, 153, 155, 162, 164, 152 Óanafì, 112 n. 191, 138, 148, 200 Màlikì, 94, 118, 121, 199, 220 n. 87, 230 Summary of, 199–201 – Tawàlà/tawallà, 109 n. 170–1, 148–51 – Theological importance, 134–5, 144, 162, 165 – With Umayyads, 198, 202–3 – Without conversion, 80, 82, 89 n. 79, 216 n. 75 Walà" 'alàqa, 202 n. 21 – Same as walà" al-muwàlàt, 197 and n. 9, 215 Walà" al-'aqìda, 25 n. 92

509

Walà" al-dàr wa-l-mujàwir, 358 Walà" al-˙ilf, 26, 72, 264 Walà" al-imàma, 163 Walà" al-inqi†à', 264 Walà" al-islàm, 82–4, 83 n. 51, 86 n. 63, 87 n. 69, 91–2, 98–9, 104, 108, 112, 113 n. 194, 119–21, 200, 205, 209, 213, 214 n. 68, 215 n. 70, 216, 220 n. 87, 230, 239–40, 264, 271 – Acceptance of, 89–90 – Origin of, 83 – Rejection of, 90, 92, 119, 200 Walà" al-'itq/al-'itàqa, Servile patronate, 70–1, 72 n. 14, 78–80, 84, 94 n. 104, 98, 101, 104, 107, 109–10, 112, 113 n. 194, 115–6, 118, 120–1, 209–10, 239, 264, 269–71 – Origin of, 83 – Sale and gift, 116 – Sunnì view, 212 Walà" al-jiwàr, 81 n. 47, 264 Walà" al-khidma, 264, 341 Walà" al-muwàlàt, Contractual patronate, 27, 70–2, 73 n. 18, 78, 80–2, 84–6, 98, 101–4, 107, 109, 112, 113 n. 194, 115–6, 118–9, 120–1, 135, 138, 146, 148–52, 155, 200, 202 n. 23, 215–6, 239, 279 – Acceptance of, 88–9 – Approval of, 87 – Conditional, 136 – Equal to ˙alìf, 23 n. 83 – For free individual, 152 – Prohibition of, 85 – Rejection of, 99, 200 – Same as walà" al-'itq, 116 – Transition to walà" al-islàm, 100 – With and without conversion, 84 – With conversion, 120 – Without conversion, 83–4, 86, 87 n. 69, 88, 92, 99, 101, 104 n. 145, 107, 108, 112, 118–20 Walà" al-nasab ibn al-'amm wa-l-qaràba, 358 Walà" al-suknà, 264 Walà" al-tibà'a, Same as walà" almuwàlàt, 72 Walà" al-yamìn al-mu˙àlif, 358 – See also Agnatic tie; Dhawù; Inheritance; Patronate; Tasyìb Walàya, See Wilàya Walì, 134 Wàlì, Umayyad administrator, 206

510

index

Wàliba b. al-Óubàb, Poet, 362 al-Walìd b. 'Abd al-Malik, Caliph, 272 al-Walìd b. 'Abd al-Ra˙màn, Arab mawlà (?), 28 al-Walìd b. al-Musayyab, Commander, 285 n. 29 al-Walìd b. 'Uqba, Companion, 174 n. 27 al-Walìd b. 'Uthmàn b. 'Affàn, Umayyad prince, 341 al-Walìd b. Yazìd, Caliph and poet, 213, 332, 335, 342, 362–3 Wansbrough, J., 74 n. 21 al-Wàqidì, Scholar, 51, 352, 406–7 Warràq, 448 Wasserstein, D., 224 n. 93, 231, 240 al-Wàthiq, Caliph and musician, 326, 339–40, 343, 347 Watt, W.M., 53, 307 Wazìr, 448 Weil, G., 437 n. 16 Wellhausen, J., 51, 191 n. 93 Wensinck, A. J., 53 Wickham, C., 222, 224 Wilàya, 134 – al-faqìh, 134 Wislàs, Ancestor of Ya˙yà b. Ya˙yà, 203–4 WOG, Westernized Oriental Gentleman, 184–5 Women, 26, 31 – As musicians and social status of, 334 Wright, O., 328 n. 9 al-Yàfi'ì, Scholar, 373 Ya˙yà b. 'Alì al-Munajjim, Musician, 344 Ya˙yà b. Kathìr b. Wislàs, Baladì, 204 Ya˙yà b. Khàlid al-Barmakì, Vizier, 343, 365 – See also Barmakids Ya˙yà b. Mu'àdh b. Muslim, Commander, 284, 285 n. 29, 286 n. 36, 291 n. 60, 296, 312 Ya˙yà b. Wàßil, Musician, 335 n. 47 Ya˙yà b. Ya˙yà, Scholar, 203–5, 213 Ya˙yà al-Makkì, Musician, 337, 344 Ya˙yà Qayl, Musician, 331, 334 Ya'là b. Munya, Companion, 14 Yamàma, Place, 62 Yaman, Tribe, 170, 187, 188–9, 191, 227 n. 98, 280, 297, 303, 304

n. 97, 312, 350 Yanshù, Musician, 340 Yaq†ìn, And family, 179 n. 48 Ya'qùb al-Wadì, Musician, 335 n. 46 al-Ya'qùbì, Scholar, 53 n. 5, 54 Yasàr b. Khiyàr, Slave, 370 Yathrib, Place, 50–3, 59, 412 – See also Àliya; Medina; Sàfila Yawm al-Íafqa, 62 Yazìd b. 'Abd al-Malik, Caliph, 272, 333–4 Yazìd b. Abì Muslim, Governor, 273 Yazìd b. Mu'àwiya, Caliph, 315 n. 146 Yazìd b. Mufarrigh al-Óimyarì, Poet, 365 Yazìd Óawrà", Musician, 337 Yemen, Place, 10 n. 27, 17, 62, 88 n. 75, 115 n. 208, 439, 461 – Kings of, 53 n. 5 Yùnus al-Kàtib, Musician, 331, 344 Yùsuf b. al-Óajjàj al-Íayqal, Poet, 352 Yùsuf b. Óusayn Bakkàr, Editor, 354 Yùsuf b. 'Umar, Governor, 314 n. 138, 315 n. 146 Yùsuf al-Fihrì, Governor, 217 Zakàt, 163–4, 271–2 Zakeri, M., 283 n. 18 al-Zanjànì, Scholar, 361 Zayd b. 'Alì, Rebellion of, 315 n. 146 Zayd b. Óàritha, Mawlà and adoptive son of prophet Mu˙ammad, 32 Zayd b. Ka'b, Musician, 335 n. 44 Zayd b. al-ˇulays, Musician, 335 n. 44 ¸ihàr, Vow, 136, 142, 145, 150 – In divorce, 145 Ziryàb, Musician, 341 n. 84 Ziyàd al-A'jam, Poet, 353–4, 356 Ziyàd b. Abìhi, Administrator, 167–9, 170 n. 9, 173 n. 23 and n. 24, 174 n. 28, 176, 180, 192, 265 Ziyàd b. Sumayya, See Ziyàd b. Abìhi, 169 Zoroastrians, 174 n. 29 Zubayda, Wife of Hàrùn al-Rashìd, 343 al-Zubayr b. al-'Awwàm, Companion, 113–5 al-Zubayr b. Bakkàr, Scholar, 57 Zufar b. al-Óàrith, Chieftain, 268 Zuhayr, Poet, 367

index Zuhayr b. Abì Sulmà, Poet, 80 n. 39 Zuhayr b. al-Musayyab, Commander, 285 n. 29, 286 n. 36, 287, 320 Zuhra, Tribe, 15–6, 28 n. 105, 197, 202 n. 21, 215 Zuhra b. Óawiyya al-Tamìmì, Commander, 24 n. 88

511

al-Zuhrì, Ibn Shihàb, Traditionist, 27, 64, 87, 93, 95, 97 n. 112, 99, 106 n. 155, 123, 372 n. 12, 396, 414 Zur'a bt. Mishra˙ al-Kindiyya, Mother of 'Alì b. 'Abdallàh b. al-'Abbàs, 311, 316

ISLAMIC HISTORY AND CIVILIZATION STUDIES AND TEXTS 1. Lev, Y. State and Society in Fatimid Egypt. 1991. ISBN 90 04 09344 3. 2. Crecelius, D. and #Abd al-Wahhab Bakr, trans. Al-Damurdashi’s Chronicle of Egypt, 1688-1755. Al-Durra al Musana Akhbar al-Kinana. 1991. ISBN 90 04 09408 3 3. Donzel, E. van (ed.). An Arabian Princess Between Two Worlds. Memoirs, Letters Home, Sequels to the Memoirs, Syrian Customs and Usages, by Sayyida Salme/ Emily Ruete. 1993. ISBN 90 04 09615 9 4. Shatzmiller, M. Labour in the Medieval Islamic World. 1994. ISBN 90 04 09896 8 5. Morray, D. An Ayyubid Notable and His World. Ibn al-#AdÊm and Aleppo as Portrayed in His Biographical Dictionary of People Associated with the City. 1994. ISBN 90 04 09956 5 6. Heidemann, S. Das Aleppiner Kalifat (A.D. 1261). Vom Ende des Kalifates in Bagdad über Aleppo zu den Restaurationen in Kairo. 1994. ISBN 90 04 10031 8 7. Behrens-Abouseif, D. Egypt’s Adjustment to Ottoman Rule. Institutions, Waqf and Architecture in Cairo (16th and 17th Centuries). 1994. ISBN 90 04 09927 1 8. Elad, A. Medieval Jerusalem and Islamic Worship. Holy Places, Ceremonies, Pilgrimage. 1995. ISBN 90 04 10010 5 9. Clayer, N. Mystiques, État et Société. Les Halvetis dans l’aire balkanique de la fin du XVe siècle à nos jours. ISBN 90 04 10090 3 10. Levanoni, A. A Turning Point in Mamluk History. The Third Reign of al-N§sÊr MuÈammad ibn Qal§wån (1310-1341). 1995. ISBN 90 04 10182 9 11. Essid, Y. A Critique of the Origins of Islamic Economic Thought. 1995. ISBN 90 04 10079 2 12. Holt, P.M. Early Mamluk Diplomacy (1260-1290). Treaties of Baybars and Qal§wån with Christian Rulers. 1995. ISBN 90 04 10246 9 13. Lecker, M. Muslims, Jews and Pagans. Studies on Early Islamic Medina. 1995. ISBN 90 04 10247 7 14. Rabbat, N.O. The Citadel of Cairo. A New Interpretation of Royal Mamluk Architecture. 1995. ISBN 90 04 10124 1 15. Lee, J.L. The ‘Ancient Supremacy’. Bukhara, Afghanistan and the Battle for Balkh, 1731-1901. 1996. ISBN 90 04 10399 6 16. Zaman, M.Q. Religion and Politics under the Early #Abbasids. The Emergence of the Proto-SunnÊ Elite. 1997. ISBN 90 04 10678 2 17. Sato, T. State and Rural Society in Medieval Islam. Sultans, Muqta#s and Fallahun. 1997. ISBN 90 04 10649 9 18. Dadoyan, S.B. The Fatimid Armenians. Cultural and Political Interaction in the Near East. 1997. ISBN 90 04 10816 5 19. Malik, J. Islamische Gelehrtenkultur in Nordindien. Entwicklungsgeschichte und Tendenzen am Beispiel von Lucknow. 1997. ISBN 90 04 10703 7 20. Mélikoff, I. Hadji Bektach: un mythe et ses avatars. Genèse et évolution du soufisme populaire en Turquie. 1998. ISBN 90 04 10954 4 21. Guo, L. Early Mamluk Syrian Historiography. Al-YånÊnÊ’s Dhayl Miŗt al"zam§n. 2 vols. 1998. ISBN (set) 90 04 10818 1 22. Taylor, C.S. In the Vicinity of the Righteous. Ziy§ra and the Veneration of Muslim Saints in Late Medieval Egypt. 1999. ISBN 90 04 11046 1 23. Madelung, W. and P.E. Walker. An Ismaili Heresiography. The “B§b al-shayã§n” from Abu Tamm§m’s Kita-b al-shajara. 1998. ISBN 90 04 11072 0 24. Amitai-Preiss, R. and D.O. Morgan (eds.). The Mongol Empire and its Legacy. 1999. ISBN 90 04 11048 8

25. Giladi, A. Infants, Parents and Wet Nurses. Medieval Islamic Views on Breastfeeding and Their Social Implications. 1999. ISBN 90 04 11223 5 26. Holt, P.M. The Sudan of the Three Niles. The Funj Chronicle 910-1288/ 1504-1871. 1999. ISBN 90 04 11256 1 27. Hunwick, J. Timbuktu and the Songhay Empire. Al-Saådi’s Ta"rÊkh al-såd§n down to 1613 and other Contemporary Documents. 1999. ISBN 90 04 11207 3 28. Munis, S.M.M. and M.R.M. Agahi. Firdaws al-iqb§l. History of Khorezm. Translated from Chagatay and annotated by Y. Bregel. 1999. ISBN 90 04 011365 7 29. Jong, F. de and B. Radtke. Islamic Mysticism Contested. Thirteen centuries of controversies and polemics. 1999. ISBN 90 04 11300 2 30. Meier, F. Essays on Islamic Piety and Mysticism. Translated by J. O’Kane, with editorial assistance of B. Radtke. 1999. ISBN 90 04 10865 3 31. B. Radtke, J.O’Kane, K.S. Vikør & R.S. O’Fahey. The Exoteric AÈmad Ibn IdrÊs. A Sufi’s Critique of the Madh§hib and the Wahh§bÊs. 2000. ISBN 90 04 11375 4 32. H. Motzki (ed.). The Biography of MuÈammad. The Issue of the Sources. 2000. ISBN 90 04 11513 7 33. Flood, F.B. The Great Mosque of Damascus. Studies on the Makings of an Umayyad Visual Culture. 2000. ISBN 90 04 11638 9 34. Weismann, I. Taste of Modernity. Sufism, Salafiyya, and Arabism in Late Ottoman Damascus. 2001. ISBN 90 04 11908 6 35. Frank, A.J. Muslim Religious Institutions in Imperial Russia. The Islamic World of Novouzensk District and the Kazakh Inner Horde, 1780-1910. 2001. ISBN 90 04 11975 2 36. Afsaruddin, A. Excellence and Precedence. Medieval Islamic Discourse on Legitimate Leadership. 2002. ISBN 90 04 12043 2 37. Anetshofer, H. and H.T. Karateke (eds.). Traktat über die Derwischmützen (RÊs§le-Ê T§ciyye) des MüstaqÊm-Z§de Süleym§n Sa#deddÊn (st. 1788). 2001. ISBN 90 04 12048 3 38. Dickinson, E. The Development of Early Sunnite \adÊth Criticism. The Taqdima of Ibn AbÊ \§tim al-R§zÊ (240/854-327/938). 2001. ISBN 90 04 11805 5 39. Fuess, A. Verbranntes Ufer. Auswirkungen mamlukischer Seepolitik auf Beirut und die syro-palästinensische Küste (1250-1517). 2001. ISBN 90 04 12108 0 40. Heidemann, S. Die Renaissance der Städte in Nordsyrien und Nordmesopotamien. Städtische Entwicklung und wirtschaftliche Bedingungen in ar-Raqqa und \arr§n von der beduinischen Vorherrschaft bis zu den Seldschuken. 2002. ISBN 90 04 12274 5 41. Motzki, H. The Origins of Islamic Jurisprudence. Meccan Fiqh before the Classical Schools. 2002. ISBN 90 04 12131 5 42. Heck, P.L. The Construction of Knowledge in Islamic Civilization. Qud§ma b. Ja#far and his Kit§b al-Khar§j wa-ßin§#at al-kit§ba. 2002. ISBN 90 04 12340 7 43. Afl§kÊ, S.D.A. The Feats of Knowers of God (Man§qeb al-#§refÊn). Translated from the Persian by John O’Kane. 2002. ISBN 90 04 12132 3 44. Donohue, J.J. The Buwayhid Dynasty in Iraq 334H./945 to 403H./1012. Shaping Institutions for the Future. 2003. ISBN 90 04 12860 3 45. Robinson, C.F. (ed.). Texts, Documents and Artefacts. Islamic Studies in Honour of D.S. Richards. 2003. ISBN 90 04 12864 6 46. Newman, A.J. (ed.). Society and Culture in the Early Modern Middle East. Studies on Iran in the Safavid Period. 2003. ISBN 90 04 12774 7

47. Jabali, F. The Companions of the Prophet. A Study of Geographical Distribution and Political Alignments. 2003. ISBN 90 04 12923 5 48. Hadi, A. Islam and State in Sumatra. A Study of Seventeenth-Century Aceh. 2004. ISBN 90 04 12982 0 49. Berg, H. (ed.) Method and Theory in the Study of Islamic Origins. 2003. ISBN 90 04 12602 3 50. Agha, S.S. The Revolution which Toppled the Umayyads. Neither Arab nor #Abb§sid. 2003. ISBN 90 04 12994 4 51. Lucas, S.C. Constructive Critics, \adÊth Literature, and the Articulation of SunnÊ Islam. The Legacy of the Generation of Ibn Sa#d, Ibn Ma#Ên, and Ibn \anbal. 2004. ISBN 90 04 13319 4 52. Guo, L. Commerce, Culture, and Community in a Red Sea Port in the Thirteenth Century. The Arabic Documents from Quseir. 2004. ISBN 90 04 13747 5 53. Shoshan, B. Poetics of Islamic Historiography. Deconstructing •abarÊ’s History. 2004. ISBN 90 04 13793 9 54. Shalem, A. The Oliphant. Islamic Objects in Historical Context. 2004. ISBN 90 04 13794 7 55. Sijpesteijn, P. and L. Sundelin (eds.). Papyrology and the History of Early Islamic Egypt. 2004. ISBN 90 04 13886 2 56. Reynolds, G.S. A Muslim Theologian in the Sectarian Milieu. #Abd al-Jabb§r and the Critique of Christian Origins. 2004. ISBN 90 04 13961 3 57. Qutbuddin, T. Al-Mu#ayyad al-ShÊr§zÊ and Fatimid Da"wa Poetry. A Case of Commitment in Classical Arabic Literature. 2005. ISBN 90 04 14103 0 58. Günther, S. (ed.). Ideas, Images, and Methods of Portrayal. Insights into Classical Arabic Literature and Islam. 2005. ISBN 90 04 14325 4 59. Papan-Matin, F. The Unveiling of Secrets Kashf al-Asr§r. The Visionary Autobiography of Råzbih§n al-BaqlÊ (1128-1209 A.D.). With the assistance of M. Fishbein. 2005. ISBN 90 04 14408 0 60. Behrens-Abouseif, D. and S. Vernoit (eds.). Islamic Art in the 19th Century. Tradition, Innovation, and Eclecticism. 2005. ISBN 90 04 14442 0 61. Bernards, M. and J. Nawas (eds.). Patronate and Patronage in Early and Classical Islam. 2005. ISBN 90 04 14480 3 62. Ritter, M. Moscheen und Madrasabauten in Iran 1785-1848. Architektur zwischen Rückgriff und Neuerung. 2006. ISBN 90 02 14481 1