Concubines and Courtesans: Women and Slavery in Islamic History 0190622180, 9780190622183

Concubines and Courtesans contains sixteen essays that consider, from a variety of viewpoints, enslaved and freed women

826 91 10MB

English Pages 368 [369] Year 2017

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Cover
Concubines and Courtesans
Copyright
Contents
List of Contributors
Introduction: Producing Songs and Sons
1. Statistical Approaches to the Rise of Concubinage in Islam
2. Abbasid Courtesans and the Question of Social Mobility
3. A Jariya’s Prospects in Abbasid Baghdad
4. Visibility and Performance: Courtesans in the Early Islamicate Courts (661–​950 ce)
5. The Qiyan of al-​Andalus
6. The Ethnic Origins of Female Slaves in al-​Andalus
7. The Mothers of the Caliph’s Sons: Women as Spoils of War during the Early Almohad Period
8. Concubines on the Road: Ibn Battuta’s Slave Women
9. Slaves in Name Only: Free Women as Royal Concubines in Late Timurid Iran and Central Asia
10. A Queen Mother and the Ottoman Imperial Harem: Rabia Gülnuş Emetullah Valide Sultan (1640–​1715)
11. Hagar and Mariya: Early Islamic Models of Slave Motherhood
12. Between History and Hagiography: The Mothers of the Imams in Imami Historical Memory
13. Are Houris Heavenly Concubines?
14. Educated Slave Women and Gift Exchange in Abbasid Culture
15. Remembering the Umm al-​Walad: Ibn Kathir’s Treatise on the Sale of the Concubine
Epilogue: Avenues to Social Mobility Available to Courtesans and Concubines
Acknowledgements
Index
Recommend Papers

Concubines and Courtesans: Women and Slavery in Islamic History
 0190622180, 9780190622183

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Concubines and Courtesans

i   j

Concubines and Courtesans Women and Slavery in Islamic History

Edited by Matthew S. Gordon and Kathryn A. Hain

i   j

1

1 Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and certain other countries. Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America. © Oxford University Press 2017 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by license, or under terms agreed with the appropriate reproduction rights organization. Inquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above. You must not circulate this work in any other form and you must impose this same condition on any acquirer. Library of Congress Cataloging-​in-​Publication Data Names: Gordon, Matthew, editor. | Hain, Kathryn A., editor. Title: Concubines and courtesans : women and slavery in Islamic history / edited by Matthew S. Gordon and Kathryn A. Hain. Description: New York, NY : Oxford University Press, 2017. | Includes index. Identifiers: LCCN 2016045077 | ISBN 9780190622183 (hardback : alk. paper) Subjects: LCSH: Courtesans—Islamic countries—History. | Slavery—Islamic countries—History. Classification: LCC HQ256.5.A5 C66 2017 | DDC 306.3/6209767—dc23 LC record available at https://lccn.loc.gov/2016045077 9 8 7 6 5 4 3 2 1 Printed by Sheridan Books, Inc., United States of America

i

Contents List of Contributors  vii Introduction: Producing Songs and Sons  1 Matthew S. Gordon

1. Statistical Approaches to the Rise of Concubinage in Islam  11 Majied Robinson



2. Abbasid Courtesans and the Question of Social Mobility  27 Matthew S. Gordon



3. A Jariya’s Prospects in Abbasid Baghdad  52 Pernilla Myrne



4. Visibility and Performance: Courtesans in the Early Islamicate Courts (661–​950 ce)  75 Lisa Nielson



5. The Qiyan of al-​Andalus  100 Dwight F. Reynolds



6. The Ethnic Origins of Female Slaves in al-​Andalus  124 Cristina de la Puente



7. The Mothers of the Caliph’s Sons: Women as Spoils of War during the Early Almohad Period  143 Heather J. Empey



8. Concubines on the Road: Ibn Battuta’s Slave Women  163 Marina A. Tolmacheva

v



9. Slaves in Name Only: Free Women as Royal Concubines in Late Timurid Iran and Central Asia  190 Usman Hamid



10. A Queen Mother and the Ottoman Imperial Harem: Rabia Gülnuş Emetullah Valide Sultan (1640–​1715)  207 Betül İpşirli Argıt



11. Hagar and Mariya: Early Islamic Models of Slave Motherhood  225 Elizabeth Urban



12. Between History and Hagiography: The Mothers of the Imams in Imami Historical Memory  244 Michael Dann



13. Are Houris Heavenly Concubines?  266 Nerina Rustomji



14. Educated Slave Women and Gift Exchange in Abbasid Culture  278 Jocelyn Sharlet



15. Remembering the Umm al-​Walad: Ibn Kathir’s Treatise on the Sale of the Concubine  297 Younus Y. Mirza Epilogue: Avenues to Social Mobility Available to Courtesans and Concubines  324 Kathryn A. Hain Acknowledgements  341 Index 343

List of Contributors

i Betül İpşirli Argıt is an associate professor at Marmara University, where she teaches Ottoman history. She completed her PhD dissertation at Bogazici University in 2009. She is the author of Rabia Gülnuş Emetullah Valide Sultan (1640–​1715), a book project supported by a postdoctoral TUBITAK fellowship. Her main areas of research are early modern Ottoman history, history of women in the Ottoman Empire, material culture, and history of the Ottoman imperial court. Michael Dann is an assistant professor in the Department of Religion at the University of Illinois, Urbana–​Champaign. His research interests center on sectarian boundaries, hadith literature, and historiography in classical and modern Islamic thought. He is currently preparing a book manuscript on the reception of Shi‘ite hadith narrators in Sunni literature. Cristina de la Puente is a scientific researcher in the Department of Jewish and Islamic Studies at the Instituto de Lenguas y Culturas del Mediterráneo and is vice president for scientific and technical research at the Spanish National Council. Her publications include Avenzoar, Averroes, Ibn al-​Jatib: médicos de al-​Andalus, perfumes, ungüentos y jarabes (2003); El banquete de las palabras: la alimentación de los textos árabes (2005); and Judaísmo e islam, with Paloma Díaz-​Más (2007). Her research fields include Islamic law in the Western Muslim world, Islamic theology, the transmission of Muslim traditions and popular religiosity, and the history of al-​Andalus.

vii

viii i

List of Contributors

Heather J. Empey is currently a librarian at the NATO Defense College in Rome, Italy. Her forthcoming PhD dissertation is titled “The Almohad Murshidah: Genealogy and Transmission of a Muslim Creed,” and her research interests include medieval Muslim historiography, Islamic theology, and the history of warfare. Matthew S. Gordon is a professor of Middle East and Islamic history at Miami University (Oxford, Ohio). His publications include The Breaking of a Thousand Swords: A History of the Turkish Military of Samarra (2000) and The Rise of Islam (2005), and a series of articles on gender and slavery in early Islamic society. He is coeditor of the Yaqubi Translation Project and, with Antoine Borrut, an editor of the online journal al-​Usur al-​Wusta. Kathryn A.  Hain earned her PhD from the University of Utah (2016). Her research interests include the social history of power machinations within the harem, and the slave trade of European concubines and eunuchs in the Middle East, India, and China. Usman Hamid is a PhD candidate in the Department of Near and Middle Eastern Civilizations at the University of Toronto. His doctoral dissertation explores the intersections of rituals, relics, memory, and sacred space in the context of north Indian Muslims’ pilgrimage to the Hijaz during the Mughal period. In addition to concubinage, his other research on slavery focuses on the history of eunuchs in sixteenth-​century North India. Younus Y. Mirza is an assistant professor of Islamic Studies at Allegheny College. His recent publications include “Was Ibn Kathir the Spokesperson for Ibn Taymiyya” in the Journal of Qur’anic Studies and “Ishmael as Abraham’s Sacrifice: Ibn Taymiyya and Ibn Kathir on the Intended Victim” in Islam Christian–​Muslim Relations (2013). His research focuses on Islamic movements, marriage and sexuality in Islam, Quranic exegesis, and hadith. Pernilla  Myrne is an associate professor in classical Arabic literature at University of Gothenburg, Sweden. Her research interests include the representation of women in Abbasid literature and premodern Arabic erotic literature. Among her publications are “Discussing ghayra in ʿAbbāsid Literature: Jealousy as a Manly Virtue or Sign of Mutual Affection” (2014) and “Who Was Hubba al-​Madiniyya?” (2015). She is currently working on a monograph on medieval Arabic–​Islamic representations of women’s sexuality. Lisa Nielson is the Anisfield-​Wolf SAGES Fellow at Case Western Reserve University. She is a historical musicologist whose research interests include the intersections of gender, slavery, and music in the early Islamicate courts. Among her publications are “Gender and the Politics of Music in the Early Islamic Courts” in Early Music History (2012). Dwight F. Reynolds is a professor of Arabic language and literature in the Department of Religious Studies at the University of California, Santa Barbara. He is the editor and



List of Contributors

j ix

coauthor of The Cambridge Companion to Modern Arab Culture (2015) and Interpreting the Self: Autobiography in the Arabic Literary Tradition (2001), as well as author of Arab Folklore: A Handbook (2007) and numerous articles on music in medieval Muslim Spain. Majied Robinson received his PhD in 2014 and is a Leverhulme Early Career Research Fellow at Edinburgh University. His primary research interests are in the Arab genealogical literary tradition, and the application of statistical and computer science methodologies to early Islamic history. Nerina Rustomji is an associate professor of history at St. John’s University (Queens, New York). She is the author of The Garden and the Fire: Heaven and Hell in Islamic Culture (2013) and the forthcoming Beauty of the Houri: Heavenly Virgins, Earthly Jihad, and the Feminine Models of Islam. Jocelyn Sharlet is an associate professor of comparative literature at the University of California, Davis. Her publications include “Chaste Lovers, Umayyad Rulers and Abbasid Writers” in Courts and Performance in the Pre-​modern Middle East (in press); “Arabic Praise Poems” in Oxford Bibliographies in Islamic Studies (in press); and Patronage and Poetry in the Islamic World: Social Mobility and Status in the Medieval Middle East and Central Asia (2011), which received an honorable mention of the British–​Kuwait Friendship Society Book Prize (2012). She works on classical Arabic poetry and prose literature. Marina A. Tolmacheva is a professor of history at Washington State University and president emerita of the American University of Kuwait. Her research interests range from Islamic geography and travel to historiography of the Islamic periphery and medieval women’s history. Elizabeth Urban is an assistant professor of the Islamic world at West Chester University of Pennsylvania. Her publications include “Abu Bakra: Mawla of the Prophet or Polemical Tool?” (2012) and “The Foundations of Islamic Society as Expressed by the Qur’anic Term Mawla?” (2013). Her research treats the intersection of slavery, gender, family structures, and political power during the early Islamic empire, which she explores in her forthcoming monograph Conquered Populations in Early Islam: Non-​Arabs, Slaves, and the Sons of Slave Mothers.

Concubines and Courtesans

Introduction Producing Songs and Sons

Matthew S. Gordon

i  The study of slavery in medieval Islamic history has progressed slowly in comparison with the work carried out by historians of ancient Greek and Roman society, early medieval Europe, the premodern Atlantic world, and the antebellum United States.1 But, there are clear indications of new interest in the topic across Arabic, Middle Eastern, and Islamic studies.2 The essays contained in this volume, Concubines and Courtesans: Women and Slavery in Islamic History, reflect such interest and underscore the need to move matters forward at an accelerated pace. The backdrop to the study of Islamic-​era slavery is, of course, the history of early, medieval, and premodern Islamic society. Judith Tucker describes it as a religious and cultural sphere, shaped continually through time from its origins in the seventh-​century Arab/​ Islamic conquests, “a geographic zone within which an educated person might feel culturally at home even as he or she traveled vast distances.”3 Our contributors range precisely over these distances, covering nearly a millennium of Islamic history. The chapters connect the early, formative period (seventh–tenth centuries ce) to the later Ottoman, Safavid, and Mughal eras (sixteenth–​eighteenth centuries), and regions extending from Islamic Spain and North Africa across Egypt and the Levant to Iran and Central Asia. The shared aim is a reconstruction of the lives, careers, and representations of women across this same expanse of time, social organization, and political drama.4 Interest in a gendered approach to Islamic history, society, and religion has sunk deep roots in Middle Eastern and Islamic studies. The effort is not simply to get at the fullness of these topics but also to underscore their centrality to a firm grasp on Islamic history. To borrow phrasing from Penelope Corfield, “Herstory” offers not only a way to update history but 1

2 i

Matthew S. Gordon

also the opportunity to move Islamic and Near Eastern historiography further along “a distinctive and innovatory pathway.”5 The aim of the research showcased here is specific: the study of enslaved and freed women. In her widely cited book, Marriage and Slavery in Early Islam, Kecia Ali comments on the ubiquitous presence of the enslaved in Muslim society: “both life and law were saturated with slaves and slavery.”6 The present essays bear out this very point. The sources on which they rely, literary and documentary alike, point to an extensive trade in girls and young women throughout the Islamic period. Three of the features of this trade are its deep historic roots, connecting Late Antiquity with the long expanse of Islamic history proper; its transregional character, with enslaved women arriving from a range of ethnic and social backgrounds; and, finally, the particular market it served. The likelihood is that a majority of enslaved women were employed in urban households in which they provided a variety of forms of labor, and in which concubinage appears to have had a notable presence. A common view is that even nonelite households employed female slaves, thus the ubiquity to which Ali refers. The essays examine the many connections of feminine slavery with social and political networking, cultural production (music, belles-​lettres, poetry, and dance), sexuality, Islamic family law, and religious tradition making. They cast light on the manner in which medieval Islamic society approached enslaved and freed persons, and the multivalent intersection of slavery and women’s history across centuries. They turn a much-​needed gendered lens on Islamic-​era slave praxis. The elite urban household—​a locus of dynamic social, political, legal, and imaginative relationships—​takes on special significance in this context.7 The concern is with women of slave and freed origin active in these households and, thus, in a complex of overlapping social, religious, and cultural interaction. Arabic, Persian, Turkish, and other traditions of written sources say much of patterns of enslavement, the structures and dynamics of slave commerce, the demands on slave and freed labor, and the social integration of slaves and freed persons. Islamic-​era scholarly culture was remarkably productive, and the medieval library in all languages of the Islamic realm is rich in references to slavery and enslaved persons. The Quran evinces a well-​rooted practice of feminine slavery in early Arab culture, using as it does a variety of terms for enslaved and freed persons, a vocabulary then enlarged over many generations of subsequent Quranic exegesis.8 Works of Arabic, Persian, and Turkish poetry, from a span of periods, serve in this regard, as do collections of adab, an elastic term used for all manner of Arabic and other bodies of Islamic-​era prose writing. Chronicles, biographical dictionaries, and works of geography and political thought also contain much evidence, as do medieval Muslim legal and religious works. And we are fortunate to have an expanding body of edited and translated documents, the majority produced originally on papyrus in medieval Egypt.9 The documentary evidence often works well with the references provided by the medieval “literary” sources. But, the evidence has its limitations. It is scattered across a range of source types, so that discussion of any given topic of Islamic-​era slavery requires much patient spadework.

Introduction

j3

The sources also rely on a range of terms. The vocabulary for all forms of servitude in Islamic society is considerable, and each such term must be contextualized carefully. In addition, the written sources deal almost exclusively with elite sectors of urban society, telling us much of political and cultural brokers but little of the governed and owned. With regard to enslaved and freed women, the sources privilege certain groups—​notably, imperial concubines and elite courtesans. Our essays discuss the lives and careers of these women in detail, drawing on rich evidence. But these same women were the product of a broad, variegated slave trade. The concubines and courtesans arrived in Islamic society alongside far greater numbers of individuals about whom we read too little. Such is the context in which the more prominent careers played out: the women who occur in the medieval and later sources succeeded, so to speak, whereas the great majority of their lesser-​known counterparts did not. Of the countless latter women, only tenuous evidence remains, but it is some of the most valuable in the documentary record. Another point concerns what is typically referred to as the domestic sphere, and it relates, in part, to the physical organization of medieval Near Eastern and Muslim households. Typically, the private/​family space was kept distinct from that used by adult males in their many interactions with society at large. The relationship of the different domains, understood as both social and physical spaces, is the subject of considerable discussion in modern scholarship.10 But, as this suggests, it also relates to the dynamics of family and gender relations—​a topic of wide complexity in medieval Islamic society, as in any other. Emerging clearly is an intricate and consistent pattern of household integration of the enslaved (male and female alike). Such integration laid the groundwork, in turn, for a variety of opportunity for social mobility. As Kristina Richardson has put it, one has to account for “the many nuanced master–​slave relationships in the medieval Islamicate world, where dominion of master over slave was not fixed and absolute.”11 A proper account of female slavery in medieval and premodern Islamic society thus can work only when careful attention is paid to webs of domestic relations as the context in which the specific lives and careers played out.12 But here lies yet a further challenge: the Arabic sources, like those in other Islamic languages, typically say little about familial matters. A vexing question for historians of Islamic society, for example, concerns the effect of integration of enslaved women on the lives of the free females sharing the same household. How did free wives and concubines get on? It is very difficult to tell. Women’s lives, per se, were likely deemed an inappropriate topic for public discourse. That we read of certain women has largely to do with their success in forging close relations with elite males and participation in male-​dominated public life. As persons of influence, they could not be ignored. The sources, in sum, are skewed. As Denise Spellberg wrote, “[sources] for the study of gender in Islamic society . . . represent a central repository of medieval male definitions about the differences between the sexes, differences that underscore relationships of power in the preservation and creation of a presumably shared Muslim past.”13

4 i

Matthew S. Gordon

The enslaved women, many of whom arrived in these urban households as children and adolescents, derived from at least four sources:  reproduction by slaves in Muslim households; capture in war and other acts of systematic violence, some expressly intended for this purpose; acquisition from busy slave markets operating in all regions bordering medieval Islamic lands; and tribute—​that is, the exchange of persons in the context of political settlements.14 These sources operated in an ebb-​and-​flow fashion relative to time and place. So, for example, acquisition of slaves, including substantial numbers of girls and young women, via conquest and war was likely more prevalent during the early Islamic period. The seventh-​century Arab/​Islamic conquests resulted in the capture and, in cases, physical transfer of subject peoples.15 These same patterns reoccurred at subsequent points in time as well, but the reliance on captive labor, in a variety of areas, likely lessened in extent and impact, the provision of slaves taken up by the other sources. Slave commerce—​that is, the trading of slaves—​which was likely the most important source overall, perhaps reached a peak during the first Abbasid period (c. 750–​950 ce). Southern and eastern Mediterranean urban society flourished at this point with particular vigor, with a concomitant growth in demand for enslaved labor, especially that of girls and women.16 Again, all indications are that the majority landed in domestic positions, making up a labor force serving middle and elite families across the urban landscape. The evidence speaks to specific patterns in the lives and careers of enslaved and freed women. They were licit sexual partners for their male owners; the Quran grants masters access to the bodies of their female slaves, and, throughout centuries of exegesis and debate, the Muslim jurists drew on the Quranic precepts and Hadith (the record of the Prophet Muhammad’s acts and opinions) in formulating normative views regarding concubinage.17 Although there seems little question that the heads of Muslim households acted accordingly, and on a consistent basis across time, the degree to which this occurred at any given point, however, is impossible to quantify. In treating the institution, we must turn to anecdotal references, narrative accounts, and legal texts. It is clear, however, that significant numbers of elite Muslim children were born to concubine mothers and their free male owners. In principle, childbearing transformed the concubine’s legal standing; she was now an umm walad, the “mother of the master’s child.”18 The mothers of nearly all Abbasid caliphs and a number of the Twelver Shi‘i Imams were concubines. It is also impossible to quantify when sexual union was coerced and thus violent, and when it was consensual and thus expressive of emotional ties.19 Given the inequities inherent in such relationships, “consensual” is likely a controversial choice of word. It is the case that medieval Arabic literary sources—​the works of adab—​speak to a spectrum of conduct from sexual assault to close, extended personal relationships.20 And one has also to reference the evidence of a lively sex industry in, for example, Abbasid society, into which, the assumption must be, a good number of enslaved women were forced.21 Sex in medieval Islamic society—​sexual conduct, sexual morality, the weight of Islamic teachings on sexual mores, and so on—​remains poorly studied in modern scholarship.

Introduction

j5

The overlap of feminine slavery and a complex of forms of sexual relations and activity seems nonetheless obvious. Of specific concern to this book is the manner in which concubines gained higher social ranking. They did so over and over again; social mobility on the part of enslaved and freed women constitutes an abiding pattern of Islamic social history. In many cases, it was a matter of linking their fortunes to the careers of male offspring, which is not to ignore the effort with which individual women attended to the construction of networks and the accumulation of personal wealth—​quite the contrary.22 One gathers, from biographies of Ahmad ibn Tulun, an autonomous governor of ninth-​century Abbasid Egypt, that Na`t, the concubine mother of at least one of his sons, administered his unwieldy household.23 Her example was not in the least exceptional; the pivotal work of Nadia El Cheikh, on the tenth-​century Abbasid court, and Leslie Pierce, on the later Ottoman royal household, make all too clear that we are dealing here with long-​term patterns.24 It is in regard to concubinage that religious and legal texts play their part. As made clear by Jonathan Brockopp and Ingrid Mattson, in treating the umm walad in early Islamic legal history, medieval Muslim jurists paid close attention to the institution.25 Brockopp and Mattson point to sustained debate among the jurists over the questions raised by concubinage.26 Such debate notwithstanding, the effort on the part of the jurists was typically to put in place protections for concubine mothers.27 The impulse drew, in no small measure, on Quranic admonitions regarding the treatment of female slaves, as well as more specific statements contained in Hadith.28 But one has to add that leading jurists owned concubines of their own, so to what measure did personal experience and interest drive their legal reasoning? It seems certain, in any case, that the jurists lent their intellectual and cultural weight to the practice. But, as is often pointed out, it remains very difficult to relate evidence contained in Islamic legal works directly to contemporary social practice, including the recourse to concubinage.29 Courtesans are a second category of women about whom the sources provide considerable evidence. They were singers, musicians, and poets employed in royal and princely courts and in other elite households. Modern scholarship has argued that this was particularly the case in early Abbasid Iraq and Umayyad Spain (al-​Andalus). An abiding question is the extent to which employment of highly trained and gifted women was a feature of these particular courts. (The phrase “singing slave girl,” common in modern scholarship, demeans these women given that, in most cases, they were mature professional performers.) It is not yet clear to what extent courtesans graced regional courts and elite households at other points of Islamic history. The question arises when one considers that the Abbasid court, even following the demise of its political supremacy after the mid-​tenth century, remained a model of prestige and organization for local and regional powers. To reiterate a point made earlier: much like the concubines with whom the courtesans must have interacted on a steady basis, these women emerged from a much larger

6 i

Matthew S. Gordon

population of enslaved women. The indications are that many of the Abbasid-​era courtesans derived from the ranks of one category in particular: indigenous slaves. They were born to elite, Arabic-​speaking households and thus were second-​generation, acculturated slaves. Most of the Abbasid singers and poets appear to have originated either from or been educated in such cities as Basra, Ta’if, and Medina, all of which were apparently centers of slave recruitment and production. The singers were consummate performers, versed in a variety of forms of literary arts, music performance, and composition; the prices cited in, for example, al-​Isbahani’s remarkable tenth-​century compendium The Book of Songs (Kitab al-​Aghani) evince the perceived cultural worth of the courtesans, admired on both ends of the Mediterranean.30 Much of their effort went into nurturing relations with their male handlers and counterparts. In sustaining such networks, the women turned not only to repartee and gift sharing, but, needless to say, compelling performance with voice and oud, as well as sexual intimacy. In some cases, they donned the hats of concubine and courtesan alike, although this was not a general pattern.31 The Book of Songs is an archive of fascination. Al-​Isbahani spoke to and for a male audience taken up with, in this case, the figure of the female performer. But we strain to hear the songs and poems themselves; the voices of the women, with important exceptions, went unrecorded. Thus, an abiding task in researching feminine slavery across Islamic history is to take up questions of representation of the enslaved, of a variety of backgrounds and professions, but also at different stages in their careers. It is a wide question of accounting for how enslaved and freed women found a place, not simply in patterns of society, culture, and politics, but also in the Near Eastern and Muslim collective imagination as well. Their guises were of a variety, sectarian in some cases, as in the case of the mothers of the Twelver Shi‘i Imams, and cultural in others, as with the courtesans. But, in each case, it becomes a matter of according the women a place of their own. Notes 1. On the case of Byzantine studies, see Rotman, Byzantine Slavery, 5. 2. See the Bibliography for the recent works by Kecia Ali, Debra Blumenthal, Amy Kallander, William Philips, Leslie Pierce, and Madeline Zilfi. 3. Tucker, “Introduction,” Arab Women, vii. 4. Theliteratureisconsiderable. See the source lists inBabayan, Islamicate Sexualities;Booth, Harem Histories; Kallander, Palace Households; and Marin, Writing the Feminine. See also the new online initiative (http://​rsn.aarweb.org/​articles/​introducingthe-​islam-​gender-​and-​women-​group). 5. Corfield, “History and the Challenge of Gender History,” 241. 6. Ali, Marriage and Slavery, 6–​7. And see her recent essay, “Truth about Islam.” 7. See Hathaway, “Military Household,” on conceiving of the medieval Middle Eastern elite household, and Pierce, Imperial Harem, 132–​136. 8. Brockopp, “Slaves and Slavery.”

Introduction

j7

9. Ragib’s Actes de vente provides 13 bills of slave sales. For an analysis of the documents, see his second volume (Actes de vente, Vol. 2). 10. A fine point of departure are the essays in Booth’s Harem Histories. 11. Richardson, “Singing Slave Girls,” 106. 12. See Bray, “Men, Women and Slaves.” 13. Spellberg, “History Then, History Now,” 4. 14. Lewis, Race and Slavery, 9–​10. 15. Crone, Slaves on Horses, 49–​51, and Hoyland, In God’s Path, 79, 91, 131–​133, 159–​160. 16. On urban commerce in early medieval Islamic history, see Bennison, Great Caliphs, 69–​93, and Berkey, Formation, 119–​123. On the demands of Abbasid-​era urban society for enslaved labor, see McCormick, Origins, 759, 768, 776. 17. See Brockopp, “Concubines.” 18. See Brockopp, Early Maliki Law, 192–​203. 19. For a stark view of sexual exploitation of enslaved women during the Abbasid period, see Ahmed, Women and Gender, 82–​86. 20. On the impact on urban households of the presence of enslaved women, see Bray, “Men, Women and Slaves,” 136–​138. 21. I cite, in my essay, the anecdote of a pimp and his clients in al-​Tanukhi, Nishwar, 2:172–​183. See, also, Caswell, Slave Girls, 26–​32, and Richardson, “Singing Slave Girls,” 111. 22. For a good discussion of the reintegration of women, in this case regarding ancient Greek urban society, see Claire Taylor, “Women’s Social Networks.” 23. Al-​Balawi, Sira, 212, and Ibn Said al-​Andalusi, Mughrib, 93–​94. 24. El Cheikh, “Caliphal Harems,” esp. 91–​93, and Pierce, Imperial Harem, see 28–​50, 57–​112, 267–​285. 25. Brockopp, Early Maliki Law, 192–​193. 26. Mattson, Believing Slave, 126, 179–​182. 27.  Brockopp, Early Maliki Law, 196–​198, 203; Mattson, Believing Slave, 126, 152; and Richardson, “Singing Slave Girls,” 107. 28. On Hadith and slaves, see Juynboll, Canonical Hadith, index, and Wensinck, Handbook, 217–​218. 29. Brockopp, Early Maliki Law, 205, 208–​209; and Mattson, Believing Slave, 180–​181. 30. See Bray, “Men, Women and Slaves,” 136–​137. I include in the bibliography one edition of the Kitab al-​aghani. On reading the Aghani and locating other editions, see Kilpatrick, Making the Great Book of Songs. 31. For one prominent example (the singer Mutayyam), see Caswell, Slave Girls, 248–​254. Also see Richardson, “Singing Slave Girls,” 114–​115.

Bibliography Primary Sources al-​Balawi, Abdallah ibn Muhammad. Sirat Ahmad ibn Tulun. Edited by Muhammad Kurd Ali. Damascus: al-​Maktaba al-​Arabiyya, 1939. Ibn Sa`id al-​Andalusi, Ali ibn Musa. Al-​Mughrib fi hula al-​Maghrib. Edited by Zaki Muhammad Hasan, Sayyida Kashif, and Shawqi Dayf. Cairo: Dar al-​Ma`arif, 1953.

8 i

Matthew S. Gordon

al-​Isbahani, Abu al-​Faraj. Kitab al-​aghani. Edited by Abd al-​Amir Ali Mahanna and Samir Jabir. 27 vols. Beirut: Dar al-​Kutub al-​Ilmiya, 1992 [1412]. al-​Tanukhi, Abu Ali al-​Muhassin ibn Ali. Nishwar al-​Muhadara wa-​Akhbar al-​Mudhakara. Edited by Abbud al-​Shalji, 4 vols. Beirut: Dar Sadir, 1398/​1978.

Secondary Sources Ahmed, Leila. Women and Gender in Islam. New Haven, CT: Yale University Press, 1992. Ali, Kecia. Marriage and Slavery in Early Islam. Cambridge, MA:  Harvard University Press, 2010. —​—​—​. “The Truth about Islam and Sex Slavery History Is More Complicated Than You Think.” 2015. The Huffington Post. http://​www.huffingtonpost.com/​kecia-​ali/​islam-​sex-​slavery_​b_​ 8004824.html (accessed October 1, 2015). Anderson, Glaire. “Concubines, Eunuchs and Patronage in Early Islamic Córdoba.” In Reassessing the Roles of Women as “Makers” of Medieval Art and Architecture, vol. 2, edited by Therese Martin, 633–​670. Leiden: Brill, 2012. Babaie, Sussan, Kathryn Babayan, Ina Baghdiantz McCabe, and Massumeh Farhad. Slaves of the Shah: New Elites of Safavid Iran. London: I.B. Tauris, 2004. Babayan, Kathryn, and Afsaneh Najmabadi, eds. Islamicate Sexualities:  Translations across Temporal Geographies of Desire. Cambridge, MA: Harvard University Press, 2008. Bennison, Amira K. The Great Caliphs:  The Golden Age of the Abbasid Empire. New Haven, CT: Yale University Press, 2009. Berkey, Jonathan P. The Formation of Islam:  Religion and Society in the Near East, 600–​1800. Cambridge: Cambridge University Press, 2003. Blumenthal, Debra. Enemies & Familiars:  Slavery and Mastery in Fifteenth-​Century Valencia. Ithaca, NY: Cornell University Press, 2009. Booth, Marilyn, ed. Harem Histories: Envisioning Places and Living Spaces. Durham, NC: Duke University Press, 2010. Bray, Julia. “Men, Women and Slaves in Abbasid Society.” In Gender in the Early Medieval World, East and West, 300–​900, edited by Leslie Brubaker and Julia M. H. Smith, 121–​146. Cambridge: Cambridge University Press, 2004. Brockopp, Jonathan E. “Concubines.” The Encyclopaedia of the Qur’an. General editor:  Jane Dammen McAuliffe. Leiden: Brill Publishing Online, 2015. —​—​—. Early Maliki Law:  Ibn Abd al-​Hakam and His Major Compendium of Jurisprudence. Leiden: Brill, 2000. —​—​—​. “Slaves and Slavery.” The Encyclopaedia of the Qur’an. General Editor:  Jane Dammen McAuliffe. Leiden: Brill Publishing Online, 2015. Brunschvig, R. “’Abd.” The Encyclopaedia of Islam, 2nd ed. P. Bearma, et  al., eds. Leiden:  Brill Publishing Online, 2015. Caswell, Fuad Matthew. The Slave Girls of Baghdad: The Qiyan in the Early Abbasid Era. London: I.B. Tauris, 2011. Corfield, Penelope J. “History and the Challenge of Gender History.” Rethinking History: The Journal of Theory and Practice 1, no. 3 (1997): 241–​258. Crone, Patricia. Slaves on Horses:  The Evolution of the Islamic Polity. Cambridge:  Cambridge University Press, 1980.

Introduction

j9

El Cheikh, Nadia Maria. “Caliphal Harems, Household Harems: Baghdad in the Fourth Century of the Islamic Era.” In Harem Histories: Envisioning Places and Living Spaces, edited by Marilyn Booth, 87–​103. Durham, NC: Duke University Press, 2010. The Encyclopaedia of Islam, 2nd ed., edited by P. Bearman et  al. Leiden:  Brill Publishing Online, 2015. Encyclopaedia of the Qur’an. General editor: Jane Dammen McAuliffe. Leiden: Brill Publishing Online, 2015. Goitein, S. D. “Slaves and Slavegirls in the Cairo Geniza Records.” Arabica 9, no. 1 (1962): 1–​20. Hathaway, Jane. “The Military Household in Ottoman Egypt.” International Journal of Middle East Studies 27, no. 1 (1995): 39–​52. Hoyland, Robert. In God’s Path:  The Arab Conquests and the Creation of an Arab Empire. Oxford: Oxford University Press, 2014. Juynboll, G. H. A. Encyclopedia of Canonical Hadith. Leiden: Brill, 2007. Kallander, Amy Aisen. Women, Gender, and the Palace Households in Ottoman Tunisia. Austin: University of Texas Press, 2013. Kilpatrick, Hilary. Making the Great Book of Songs: Compilation and the Author’s Craft in Abu al-​Faraj al-​Isbahani’s Kitab al-​aghani. London: Routledge/​Curzon, 2003. Lewis, Bernard. Race and Slavery in the Middle East:  A  Historical Enquiry. Oxford:  Oxford University Press, 1990. Marin, Manuela, and Randi Deguilhem, eds. Writing the Feminine:  Women in Arab Sources. London: I.B. Tauris, 2002. Marmon, Shaun. “Domestic Slavery in the Mamluk Empire: A Preliminary Sketch.” In Slavery in the Islamic Middle East, edited by Shaun E. Marmon, 1–​23. Princeton, NJ: M. Wiener, 1999. —​—​—​. “Slavery, Islamic World.” In Dictionary of the Middle Ages, edited by Joseph R. Strayer, 330–​333. New York: Scribner, 1988. Mattson, Ingrid. “A Believing Slave Is Better Than an Unbeliever:  Status and Community in Early Islamic Society and Law.” PhD diss., University of Chicago, 1999. McCormick, Michael. Origins of the European Economy: Communications and Commerce, A.D. 300–​900. Cambridge: Cambridge University Press, 2001. Müller, Hans. Die Kunst des Sklavenkaufs: nach arabischen, persischen und türkischen Ratgebern vom 10. bis zum 18. Jahrhundert. Freiburg: Schwarz, 1980. —​—​—​. “Sklaven.” In Wirtschaftsgeschichte des vorderen Orients in Islamischer Zeit, edited by Bernard Lewis et al., 53–​83. Leiden: Brill Publishing, 1977. Philips, William D., Jr. Slavery in Medieval and Early Modern Iberia. Philadelphia: University of Pennsylvania Press, 2014. Pierce, Leslie P. The Imperial Harem: Women and Sovereignty in the Ottoman Empire. Oxford: Oxford University Press, 1993. Ragib, Yusuf. Actes de vente d’esclaves et d’animaux d’Égypte medieval. Cairo:  Institut Français d’Archéologie Orientale, 2002 (Vol. 1) and 2006 (Vol. 2). —​—​—​. “Les Marchés aux esclaves en terre d’Islam.” In Mercati e merchant nell’alto medioevo: L’area Euroasiatica e l’area Mediterranea, n.e., 721–​763. Spoleto: Centro italiano di studi sull’alto Medioevo, 1993. Richardson, Kristina. “Singing Slave Girls (Qiyan) of the Abbasid Court in the Ninth and Tenth Centuries.” In Children in Slavery through the Ages, edited by Gwyn Campbell, Suzanne Miers, and Joseph C. Miller, 105–​118. Athens: Ohio University Press, 2009.

10 i

Matthew S. Gordon

Rotman, Youval. Byzantine Slavery and the Mediterranean World. Translated by Jane Marie Todd. Cambridge, MA: Harvard University Press, 2009. Schacht, J. “Umm al-​walad.” The Encyclopaedia of Islam, 2nd ed., edited by P. Bearman et  al. Leiden: Brill Publishing Online, 2015. Schneider, Irene. “Freedom and Slavery in Early Islamic Time (1st/​7th and 2nd/​8th Centuries).” Al-​Qantara 28, no. 2 (2007): 353–​382. —​—​—​. Kinderverkauf und Schuldknechtschaft. Wiesbaden:  Kommissionsverlag Franz Steiner, 1999. Smith, Julia M.  H. “Did Women Have a Transformation of the Roman World?” Gender and History 12, no. 3 (2000): 552–​571. Spellberg, Denise. “History Then, History Now: The Role of Medieval Islamic Religio-​Political Sources in Shaping the Modern Debate on Gender.” In Beyond the Exotic: Women’s Histories in Islamic Societies, edited by Amira al-​Azhary Sonbol, 3–​14. Syracuse, NY: Syracuse University Press, 2005. Taylor, Claire. “Women’s Social Networks and Female Friendship in the Ancient Greek City.” Gender and History 23, no. 3 (2011): 703–​720. Tucker, Judith E., ed. Arab Women:  Old Boundaries, New Frontiers. Bloomington:  Indiana University Press, 1993. Wensinck, A. J. A Handbook of Early Muhammadan Tradition. Leiden: Brill, 1971. Zilfi, Madeline C. Women and Slavery in the Late Ottoman Empire. Cambridge:  Cambridge University Press, 2010.

1 Statistical Approaches to the Rise of Concubinage in Islam Majied Robinson

i  Any discussion of the origins of female slavery in Islam is confounded at the outset by the lack of direct documentary source material.1 Unambiguous references to concubinage do not feature heavily in the Quran,2 and although non-​Muslim sources mention frequently how the Arab conquerors took non-​Arab women as slaves, beyond this little more is ever said.3 As if this were not problematic enough, even when medieval Islamic scholars began to develop a concern for concubinage, they showed little interest in its origins or even its role outside legal discourse.4 This leaves us with virtually no historical narratives of the pre-​Abbasid evolution of concubinage as a social practice. This chapter shows how this lacuna can be filled by supplementing our traditional approaches with a more innovative methodology—​namely, the quantitative analysis of the parental records preserved in a ninth-​century Arab genealogical text. This text—​ the Nasab Quraysh of al-​Zubayri (d.c. 850)—​contains the maternal records of nearly 3,000 Qurayshi tribespeople, the bulk of whom lived from 500 to 750 ce.5 As such, it is potentially a resource of unparalleled quality in terms of preserving the marriage behavior of a Late Antique society. It is a challenging book to read because not only are the marriages and births undated, but also there is the question of whether we can trust any of the information it records. The purpose of this chapter is to show how these issues can be overcome by using basic prosopographical and statistical techniques. In terms of concubinage, extraction and analysis of marriage data reveal patterns that concur clearly with evidence provided in nongenealogical sources, including chronicles. There was a dramatic surge in the number of children born to concubines shortly after the revelation of Islam. This is a strong indicator that our methodology is sound and, as

11

12 i

Majied Robinson

long as they are approached in a prosopographical manner, the data themselves can be considered a documentary record of the unions to which they refer. Our conclusions, again, sit well with the nongenealogical evidence. They conflict, however, with elements of the existing scholarly consensus surrounding the practice of concubinage during the first Islamic period. Contrary to what is often claimed, concubines and their progeny enjoyed a significant presence in elite Arab families well before the Abbasid era, and there is little evidence to suggest that either mothers or their offspring were discriminated against on a systematic basis. Thus, prosopographical methodologies can contribute to current debates when established sources are read in novel ways. Existing Narratives on the Emergence of Concubinage Current scholarship on the origins of concubinage focuses on two broad aspects of the practice. The first type of scholarship concentrates on the social changes, which resulted in the improvement of the status of the umm walad (“the mother of the [master’s] child”) and her children. The most detailed treatment of these changes can be found in Goldziher, Bashear, and Athamina.6 The second strand looks more narrowly at the development of the legal framework that secured these rights. There, the core works are those of Schacht and Brockopp.7 It is the first strand that concerns us here; the second strand, which is rooted more securely in the sources, does not attempt to say a great deal about concubinage before the birth of Islamic legal writing. In terms of its treatment of Islamic concubinage, the first strand can be summarized as follows. Concubinage was uncommon in Arabia before the emergence of Islam.8 Where it did occur, the status of the children of these unions followed that of their mothers from birth, and the children remained slaves for the remainder of their lives unless their fathers recognized them as heirs. The paradigmatic case here is that of Antara.9 With the conquests, however, part of this picture changed: large numbers of slave women became available to the conquerors, and some of these conquerors had children with the slave women. Despite the increasing numbers of children, attitudes toward concubine unions and their progeny remained largely negative, and to marry a non-​Arab woman of slave origin would most likely provoke outrage.10 There was also no question of these children becoming caliphs;11 evidence from this comes from the fact that Maslama ibn Abd al-​Malik was barred from the caliphate thanks to his concubine mother.12 In some places, it is even suggested that the Umayyads believed in a prophecy that linked their dynastic downfall to the arrival of a concubine-​born caliph.13 The argument goes on to indicate that, throughout the course of the Umayyad caliphate, the fundamentally egalitarian nature of Islam led to a change in these attitudes.14 Eventually, increasing numbers of Arab Muslims came to see the existence of a non-​Arab mother as a positive attribute, and by the end of the Umayyad caliphate, we find the first concubine-​born caliphs.15 This process accelerated in the cosmopolitan Abbasid dynasty, where all but three of the caliphs were born of slave women.16 From this point on, slave



Statistical Approaches to the Rise of Concubinage in Islam

j 13

women in the harem became an integral part of both the Muslim and non-​Muslim image of the Islamic court, right up to the twentieth century. Problems with the Existing Narrative There is much that is unsatisfactory about this explanation. First, it looks suspiciously like what many Abbasid-​era Muslim scholars would like us to believe. As Crone has shown, there is a trope in Muslim historical narratives whereby the “true” Islam of Muhammad is corrupted by the self-​interested attitudes of the Umayyads, and only with the arrival of the Abbasids is the “legitimate” faith restored.17 Second, it is uncritical in its use of anecdotal evidence. All the sources used by Goldziher, Bashear, and Athamina are of late provenance (typically mid to late ninth century, and in some cases even later), and although they should not be discarded on this basis alone, the three scholars make little effort to address the issue of reliability. Indeed, there is a worryingly circular element to their argumentation; if a source claiming to be of the Umayyad era is critical of concubines and their children, it is considered to be a genuine tradition because it reflects the attitudes of the time. But, if it claims to be Umayyad while being complimentary to concubines and their children, it is deemed to be a later back-​projection because these enlightened attitudes could only be found after 750 ce.18 In addition, much of the anecdotal evidence is of poor quality, even by its own internal standards. Take, for instance, the supposed refusal of the Arabs to countenance the idea that the otherwise reputable Maslama ibn Abd al-​Malik might ascend to the caliphate or the prophecy that the Umayyad dynasty would be brought down when the son of a concubine took this position.19 There are other reasons to suggest why Maslama ibn Abd al-​Malik did not become caliph:  his responsibility for the disastrous siege of Constantinople (c. 715–​718 ce) and the fact that he died toward the end of the reign of his long-​lived brother Hisham. As for the prophecy, to imply that this existed before the end of the Umayyad dynasty would be to make it one of those rarest of divinations—​one that predicts something successfully. In addition, it should be noted that just a few years after Maslama died, the first concubine-​born caliphs did appear, and this occurrence did not lead directly to a split in the Muslim polity.20 But, there is one significant argument in favor of the traditional narrative: the indisputable fact that the children of concubines do not become caliphs until the very end of the Umayyad period. In a way, this observation is itself a prosopographical study—​one that tracks the category of “mothers of caliphs” over time. Thus, before 744 ce, there were no instances of caliphs born to concubines; between 744 ce and 785 ce, roughly two-​thirds of caliphs were born to slave women (depending on who is counted as a caliph); and, after 785 ce, almost all caliphs were born to concubines. For current scholarship, this is clear evidence that the sons of concubines were discriminated against during the Umayyad period, and is the hook on which the anecdotal reports can be hung.

14 i

Majied Robinson

This chapter contests the existing narrative by demonstrating that the late emergence of the concubine-​born caliph was not a result of discrimination but a consequence of structural changes to the network of relationships that constituted the Muslim polity. We establish this by expanding the prosopography of maternal origins to comprise hundreds rather than dozens of Quraysh. The resulting database is then analyzed statistically. Also shown is that, contrary to our current understanding, concubinage was popular from the very beginning of the Arab military conquests, and there is little evidence that the progeny of these unions were discriminated against on a significant basis. Structuring the Data The first step is to convert the records within the Nasab Quraysh into data suitable for diachronic study. This is a challenge because the work records no dates for marriages or births. To overcome this hurdle, structure is imposed on the data by assigning each relationship a generational number based on the degree to which the fathers of the families are related to a relative.21 The relative in question is Qusayy ibn Kilab,22 who therefore becomes a member of Generation 0.23 For individuals to earn mention in this system, they do not have to be descended from Qusayy, only connected to him genealogically. Table 1.1 demonstrates how this works for Muhammad and one of his contemporaries, both of whom are linked through their common ancestor Murra. This shows that Muhammad and Abu Bakr are not just contemporaries according to the traditional narrative sources but also of the same generation, as suggested in this methodology. This is an important component of establishing the veracity of the genealogical data; a common facet of genealogically minded societies is their propensity to “telescope” genealogies whereby unremarkable ancestors are excised, thus shortening the number of overall links.24 If this were happening on a wide scale in the Qurayshi genealogies, the lineages of these two men would not concur in the fashion shown in Table 1.1. Table 1.1 generational structuring for muhammad and abu bakr Generation Lineage 1 Lineage 2 –​2 Murra Murra –​1 Kilab Taym 0 Qusayy Sad 1 Abd Manaf Kab 2 Hashim Amr 3 Abd al-​Muttalib Amir 4 Abd Allah Abu Quhafa 5 Muhammad al-​Nabi Abu Bakr al-​Saddiq



Statistical Approaches to the Rise of Concubinage in Islam

j 15

Table 1.2 comparison of degrees of separation Name Generation Degree of separation 5 NA Muḥammad ibn Abd Allah Abu Bakr al-​Saddiq Umar ibn al-​Khattab Uthman ibn Affan Ali ibn Abu Talib Talha ibn Ubayd Allah Zubayr ibn Awwam

5 6 6 5 5 5

7 9 5 2 7 5

Estimated years of separation from common ancestor NA 210 270 150 160 210 150

NA, not applicable.

Similar concurrence is found when comparing the generational numbers of other contemporaries of Muhammad, which are detailed in Table 1.2, along with an estimate of the number of years between the births of these men and the births of their common ancestor (these estimates are based on a figure of 30 years for a generation).25 Again, the methodology is producing credible results from the data set. Despite the fact that some of these men are separated from their common ancestors by more than 200 years, they appear in a spread of only two generations.26 The next stage is to extract the relevant marriage data. Because the current study is concerned with the relative popularity of concubinage versus marriage to free women over time, the numbers of children born to both types of women need to be extracted and input into a database. This works as follows. A typical entry in the Nasab Quraysh looks like this: Abd Allah al-​Asghar (the younger) ibn Wahb ibn Zama was born to a concubine (umm walad) . . . his wife was Karima bint al-​Miqdad ibn Amr al-​Bahrani, whose mother was Dubaa bint Zubayr ibn Abd al-​Muttalib. She [Karima] gave birth to: al-​Miqdad ibn Abd Allah who had no descendants and was killed at the battle of Harra; Wahb ibn Abd Allah who had no descendants and was killed at the battle of Harra; and Yaqub, Abu al-​Harith, Yazid, and Zubayr, the sons of Abd Allah al-​Asghar ibn Wahb.27 This extract details the childbearing marriages of two Qurayshi men:  Abd Allah al-​ Asghar (who produced six children by one free woman) and his father, Wahb ibn Zama (who produced Abd Allah through a concubine).28 A simplified database entry for this information is shown in Table 1.3. Data Extraction Replicating this process for the entirety of the Nasab Quraysh yields a database of nearly 3,000 people for whom we know the name of the father and the status of the mother.

16 i

Majied Robinson

Table 1.3 sample database entry Father’s name Generation Wahb ibn Zama Abd Allah ibn Wahb

6 7

Children born to free women 0 6

Children born to concubines 1 0

The decision to count numbers of children rather than women was made because the Nasab Quraysh sometimes records a man as having had a number of named children through “concubines,” without any indication as to how many women to which this might refer.

500 450 Number of children

400 350 300 250 200 150 100 50 0

–6 –4 –2

0

2 4 6 Generation

Total children

8

10

12

Concubine children

Figure 1.1  Children born to concubines (conc.) versus total children. 

This allows us to track the number of children born to free women and concubines over time. The starting generation is that of Fihr, the man whom a number of sources credit as being the person from whom all Quraysh are descended; he was born six generations before Qusayy.29 The endpoint is the final generation that appears in the Nasab Quraysh (Generation 13; the author is Generation 10). The changes in marriage behavior are illustrated in Figure 1.1, which shows two lines: one representing the total number of children born and the other indicating the number of children born to concubines.30 The first observation is that there are no instances of Quraysh children being born to slave women before Generation 3, which is the generation of Muhammad’s grandfather, and it remains uncommon until the generation of Muhammad (Generation 5). This concurs nicely with what we find in the traditional narrative sources and the pre-​Islamic Arab poetry corpus; these sources may be problematic, but in neither type is there any



Statistical Approaches to the Rise of Concubinage in Islam

j 17

500 450 Number of children

400 350 300 250 200 150 100 50 0

3

4

5

6

7 8 9 Generation

Total children

10 11 12 13

Concubine children

Figure 1.2  Focus on children born to Generations 3 through 13.

evidence of significant numbers of slave women producing children for the Arabians of the preconquest Hijaz.31 To understand the evolution of concubinage, it makes sense to focus more narrowly on the generations that bore children through umm walads. This is illustrated in Figure 1.2, which details the maternal origins of children born between Generations 3 and 13. The raw data for these figures can also be presented in tabular form, as shown in Table 1.4. It is clear that, although there are some reports of concubines producing children for the men from before Muhammad’s generation, they are infrequent and found solely in the generations of his father and grandfather. Like the absence of concubinage in earlier generations, this finding concurs with the traditional narrative sources; concubinage was uncommon in pre-​Islamic Arabian society, but this changed for the men of Muhammad’s era as a consequence of military conquests. Although this link may seem obvious, we should recall that it is built on two contentions that are not obvious: (1) that the Nasab Quraysh is filled with accurate maternal data and (2) that the generational tagging suggested here is a viable means of organizing this information. This finding means potentially that, in the instances when statistical analysis diverges from the traditional narrative sources, the conclusions drawn from the former should be considered the more credible. Implications for Scholarship on the Emergence of Concubinage Such divergences appear when we compare the findings as illustrated in Figures 1.1 and 1.2 with elements of the narrative of concubinage as presented by Goldziher and colleagues.

18 i

Majied Robinson

Table 1.4 raw data for figures 1.1 and 1.2 Generation No. of children No. of children Proportion of total children born to concubines born to concubines, % Fihr (–​6) to 2 453 0 0.00 3 250 3 1.20 4 371 7 1.89 5 441 53 12.02 6 439 118 26.88 7 403 143 35.48 8 290 123 42.41 9 223 84 37.67 10 30 6 20.00 11 4 2 50.00 12 0 0 0.00 13 1 1 100.00

The analysis of the data reveals that dramatic shifts in social practice were taking place a long time before the Umayyad dynasty fell, and there is no evidence that the Quraysh as a whole were reticent about taking women as concubines and having children with them after the conquests.32 This appears to have happened as soon as the women were available. There is also no evidence that the Umayyad caliphs were especially averse to concubines either; a subset of the database looking only at the children produced by these caliphs and their sons reveals that these men were actually slightly more likely than average to produce children by slave women.33 In addition, a separate prosopographical investigation reveals that the sons of concubines are recorded as making a number of marriages to Qurayshi women.34 This undermines Athamina’s contention that the “Arab aristocracy was appalled at the idea of their daughters’ marrying princes born to mothers who were umm walad”;35 the prosopography therefore shows that, in marriage terms, the sons of concubines were outdoing non-​Qurayshi Arabs during the postconquest era, and in some cases better than some Qurayshis themselves.36 Figures 1.1 and 1.2 also explain in part the late emergence of the concubine-​born caliph. Even at their most popular, they still accounted for less than half the total children born to Qurayshi men. Their delayed arrival may simply be a matter of demographics, but a better explanation for this change in the maternal origins of the caliphs is found when we extend our prosopography to the marriages made by elite members of the Quraysh to Arab women rather than foreigners. Explaining the Late Appearance of Concubine-​B orn Caliphs Table 1.5 represents the 56 marriages carried out by Umayyad caliphs and their sons to Arab women as recorded in the Nasab Quraysh. The marriages have been categorized



Statistical Approaches to the Rise of Concubinage in Islam

j 19

Table 1.5 marriages of the umayyad caliphs and their sons Generation Non-​Quraysh wives Non-​Umayyad Umayyad wives Quraysh wives 6 3 3 2 7 3 12 4 8 2 1 20 9 0 1 5

Total 8 19 23 6

Table 1.6 maternal status of children of umayyad caliphs and sons by generation Generation Total Total concubine Proportion concubine Qurayshi average by of father children children children, % generation, % 6 21 2 9.52 26.88 7 54 22 40.74 35.48 8 79 36 45.57 42.41 9 27 16 59.26 37.67

as belonging to one of three categories: to non-​Quraysh Arab women, to non-​Umayyad Quraysh women, and to Umayyad women. The chart shows that as the generations of Umayyad caliphs progress through time, their pool of marriage partners narrows. In the first generation (Muawiya and Marwan I), the marriages are split between the three different types of women in roughly equal proportions.37 In the next generation (Abd al-​Malik and Yazid I), the proportion of marriages to Arabs outside the Quraysh tribe falls drastically, and the shortfall is taken up by the other two groups. In the final generations, marriages to non-​Umayyad Quraysh are curtailed significantly—​meaning, that although the caliphs were technically in a position to marry any woman they wanted, they were interested exclusively in other Umayyads. Simultaneously, as Table 1.6 shows, the caliphs were also taking more and more concubines. This seems paradoxical because, in one respect, they are becoming more endogamous through increasing amounts of intra-​Umayyad marriage (in other words, cousin marriage), whereas in a second respect they are taking more and more foreign slave women, who are the most exogamous marriage partners possible. In fact, all three factors—​the late emergence of concubine-​born caliphs, increasing concubinage, and increasing cousin marriage—​may be linked. During the early decades of the caliphate, the structure of the Muslim polity continued to be as tribal as it had been for their ancestors living during the pre-​Islamic period, with loyalty and solidarity

20 i

Majied Robinson

being understood along biological lines. As long as the families were of roughly equal status, marrying Arab women from outside one’s tribe was desirable because the resulting children would be supported by two tribes rather than one, which would have given their lineage a competitive edge. Children of concubines, on the other hand, enjoyed only one set of these relationships, and this handicapped their ascent to the top of the tribal polity. The conquests, however, had consequences that ultimately upset the pre-​Islamic system. As the Umayyad dynasty matured, certain families within the Quraysh became significantly wealthier and more powerful than tribes that had once been equal to them. They now enjoyed the means to secure the latter’s allegiance without the necessity of marriage or blood ties.38 Loyalty could now be earned and maintained through the provision of sinecures, lucrative political appointments, and direct cash payments. In this new environment, the negative sides of exogamous Arab marriage presented themselves—​namely, the cost of dowries, the split loyalties of wives, and the potential for these tribal marriages to destabilize the balance of power among the clans. In this new order, the Muslim elites turned to the cheapest, safest, and most loyal women available to them: cousins and concubines. If the evolution of the pre-​Abbasid social networks can indeed be understood in this way, then they also explain the late emergence of the concubine-​born caliph. The Umayyad son of a slave woman did not need a formal mechanism to bar him from the caliphate throughout much of this period. The fact that he had no maternal relatives in an era when these people were key components of political success was enough to stymie any ambitions he may have had. As the tribal polity gathered the trappings of an empire, these structural barriers broke down, opening the door for the arrival of caliphs born to slave mothers. Marwan II—​the last Umayyad caliph—​epitomizes this change. Born to a slave mother, he proved himself to be an effective governor on the Caucasian frontier. He understood the power of money in securing loyalty, and to this end expanded the number of nontribal military regiments in his forces.39 During the turmoil of the late Umayyad period, he used his experience and assets to seize control of the caliphate and, had he not been swept away by the Abbasid Revolution, he might have eventually consolidated the Umayyad dynasty on much firmer footing.40 Likewise, when we consider his attempt to move the capital eastward, there is a convincing argument that his personal background and political strategies preempted many elements of the Abbasid caliphate. His appearance as the first concubine-​born caliph of note should not, therefore, be seen as the byproduct of the efforts of those who had kept alive the flame of enlightened “true Islam” (as suggested by Goldziher and others), but as a consequence of Marwan II’s ability to make sense of a rapidly changing political environment. None of this should be taken as a denial that, among the early Arab Muslims, there were some who denigrated those with non-​Arab slave mothers. This is certainly true, and the earlier scholarship has done a credible job of marshaling the supporting evidence.



Statistical Approaches to the Rise of Concubinage in Islam

j 21

But, this is not enough to prove these prejudices were ever acted on seriously or held by a critical mass of Muslims. The prosopographical analysis of the Nasab Quraysh provided here indicates that these derogatory remarks may be more akin to a literary exercise in the manner of the shuubiyya than a reflection of a true caste system. Conclusion We have only been able to scratch the surface in terms of what can be said about concubinage and power networks in early Islam. Nonetheless, it is hoped that this chapter demonstrated new methods by which historians can manage the immense corpus of prosopographical data preserved in the Islamic historical sources. Part of this contribution has been methodological; the system of generational organization used here has been shown to be a valid way of studying these data diachronically, and can, potentially, be extended to the study of the non-​Qurayshi tribespeople found in works such as Ibn Sa`d’s Tabaqat or Ibn al-​Kalbi’s Jamharat al-​Nasab. Another contribution of the study has been to add nuance to our understanding of intra-​and extragroup marriage. As has been shown with the marriages of the Umayyad caliphs, when it comes to concubines and cousins, we cannot assume that “exogamy” or “endogamy” are simple antonyms because the underlying social context behind either phenomenon could be the same. Most of all, it is hoped that the investigation presented in this chapter has demonstrated that, by using novel methodologies, we can cast a new light on long-​standing scholarly narratives of social history in the first century of Islam. The figures and tables show the Umayyads and the rest of the Arab elite as being much more dynamic than we may have thought previously and capable of adapting to changing circumstances with remarkable speed. This prosopographical evidence indicates the Arab conquerors took slave women as soon as they could, and they did not discriminate against their progeny. A shift over time in the dynamics of imperial politics saw, too, a shift in marriage patterns, with an increasing turn to concubines and ever-​closer relations. The paucity of information in narrative sources regarding concubines and their children in the first century of Islam led previous scholars to incorporate these meager findings into better attested trends concerning race, politics, and religion of the period (typically the evolving attitudes toward the mawali and the late emergence of concubine-​ born caliphs). But, with the detailed picture of maternal origins provided by the statistical analysis of the Nasab Quraysh,41 we are in a position to examine the concubines of early Islam afresh and to consider the possibility that full Arab Muslims may not have regarded the children of their concubines to be members of a mawali-​like caste of second-​class believers. We would not have learned this had we relied solely on the familiar approach of combing the traditional Islamic historiographies in search of more anecdotes to add to Goldziher’s collection, nor would it have been discovered through the investigation of nonliterary sources or non-​Islamic texts, because these have almost nothing to say on the subject of concubinage. It is only through the use of innovative prosopographical

22 i

Majied Robinson

methodologies that we can critique the existing narratives, and in so doing we can understand more fully the role of the female slave in early Islamic history.

Notes 1. The data referred to in this chapter have been extracted manually and, for the most part, are stored in hard copy. Because of this, although I am confident the general conclusions from this work are secure, it must be emphasized that the figures themselves should be considered preliminary findings for a more detailed study that can use better methods of data extraction and management. 2. Brockopp, “Concubines.” 3. Hoyland, Seeing Islam as Others Saw It, 596, note 9. 4. Brockopp, Early Maliki Law, 204–​207. 5.  Al-​Zubayri, Nasab Quraysh. The introduction to the edition is useful in that it contains extracts from the main biographical notices on al-​Zubayri. For the most important of these, see Ibn al-​Nadim, Fihrist, 160, and al-​Khatib al-​Baghdadi, Tarikh Baghdad, 13:112–​114. For a discussion of al-​Zubayri’s death date, see Levi-​Provencal’s introduction to the Nasab Quraysh (14, note 3) and Pellat, “Muṣʿab.” 6. Goldziher, Muslim Studies, 1:98–​136; Bashear, Arabs and Others; and Athamina, “How Did Islam Contribute.” 7. Schacht, Origins, 264–​266 and “Umm al-​Walad,” and Brockopp, Early Maliki Law, 191–​203. 8. Athamina, “How Did Islam Contribute,” 395. 9. Athamina, “How Did Islam Contribute,” 385–​386. 10. Most scholars writing on concubinage do not differentiate between marriage to a slave woman and having her as a concubine; for our purposes here, this should not be problematic. 11. Athamina, “How Did Islam Contribute,” 395. 12. Rotter, “Maslama b. ʿAbd al-​Malik b. Marwān.” 13. Athamina, “How Did Islam Contribute,” 395 and Lewis, Race and Slavery, 105. 14.  Goldziher, 1:116 (“The old Arabs remained quite untouched by the consequences of Muhammed’s and Islam’s teaching of equality in regard to this question [of Arab men being allowed to marry non-​Arab slaves]); 1:117 (“It cannot be denied, and this has been repeatedly stressed in descriptions of Islam, that the Islamic spirit helped make good treatment of slaves a duty and inner duty and to encourage an attitude that had its roots in the oldest documents of Islam”); Athamina, “How did Islam Contribute,” 395 (where the decline in prejudice against the sons of concubine women is attributed to a number of factors, including “the decline in the basic concepts of the Jāhilī heritage in favor of the basic values of Islam”). A slightly more nuanced view (although still framed in terms of Arabization versus Islamization) can be found in Bashear, 115–​116. 15. Goldziher, Muslim Studies, 1:118, and Athamina, “How Did Islam Contribute,” 395–​396. 16. As counted by Ps.-​Jahiz, Mahasin, 230–​231. 17. Crone, “The First-​Century Concept,” 353. 18.  For example, this is done by both Goldziher, 1:119, and Athamina, “How Did Islam Contribute,” 396–​397, when referring to Zayn al-​Abidin. 19. For other apocalypses that would accompany the arrival of Arabs with foreign mothers, see Bashear, Arabs and Others, 95.



Statistical Approaches to the Rise of Concubinage in Islam

j 23

20.  Doctrinal divisions often preceded political ones; see Crone, Medieval Islamic Political Thought, 21ff. The fact that this did not happen with the ascent of concubine-​born caliphs to the throne is an indicator that few members of Muslim society at the time were convinced of the need for their disenfranchisement. 21.  Generational approaches have, to an extent, been used by other scholars, although not in a quantitative fashion. For examples, see Caskel, Gamharat, in his structuring of family tree diagrams; Bernheimer, “Genealogy,” in her prosopography of the Alid family; and Ahmed, Religious Elite. 22. Qusayy is a figure of some importance in Qurayshi history. He is purported to have not only wrested control of the shrine of the Kaaba from its previous custodians but also established it as a permanent settlement for his tribespeople as well. 23. The Prophet Muhammad would seem a natural choice for Generation 0, but the problem with this is one of clarity; a large number of his relatives would fall into the categories of Generation –​1 and Generation 1 (his uncle Abbas ibn Abd al-​Muttalib alongside his nephews Hasan and Husayn ibn Ali ibn Abi Talib, for example), making for confusing reading. 24. Evans-​Pritchard, “Nuer Time-​Reckoning,” 214. 25. Thirty years to a generation has been adopted as a rule of thumb. This rule is based on a comprehensive recent study of how long a male generation lasts in a society living in largely premodern conditions (see Matsumura and Forster, “Generation Time”). Other scholars have taken different numbers of years for generational length; Ahmed, 8, note 18, uses 20 years, and Bulliet, Conversion, 21, uses 34 years, which is the average number of years between a son’s death and that of his father in his database of Nishapuri scholars. Interestingly, when Ibn Khaldun, Muqaddima, 2:537–​538, was investigating the veracity of genealogical claims, he stated that a credible number of generational links in a lineage was three per century, which falls between the Matsumura and Forster estimate and that of Bulliet. 26. An objection to this conclusion may be that because these are such high-​profile Muslims, generations were added or subtracted to give them a credible number of links to a common ancestor. This is unlikely for two reasons. First, evidence from genealogically organized societies observed by anthropologists shows their members are not aware that mismatched numbers of generations are a cause for concern. Second, if the genealogists were playing around with the generational records for the Quraysh of Muhammad’s era, we would expect to find a substantial amount of conflict within this element of the nasab tradition, which is not the case. See Szombathy, Roots of Arabic Genealogy, 42, 96. 27. Al-​Zubayri, Nasab Quraysh, 228. 28. Miqdad ibn Amr’s marriage is ignored because he was not a member of the Quraysh. 29.  Whether Fihr was widely regarded as the forefather of the Quraysh is not particularly important. More relevant is that fact that, before Fihr, there are very few recorded marriages in the Nasab Quraysh. 30. Occasionally, there are instances when a mother may be a slave but is not listed as being an umm walad. Mariya the Copt is one such example, and there are instances when a woman is described as being a female slave (jariya) but not an umm walad. These instances are rare, however, and the mass-​data approach used in this study means they will not have a significant impact on our overall conclusions. 31. Brockopp, Early Maliki Law, 195, note 151, finds only two unambiguous instances of pre-​Islamic concubinage in the historical sources. Although Athamina, “How Did Islam Contribute,” 385, also

24 i

Majied Robinson

argues for the limited presence of concubines in pre-​Islamic Arabia, at the same time he makes reference to “two lists of Meccan and tribal aristocrats who were born to Christian Greek slave women and to slave women from Ethiopia” (i.e., in the pre-​Islamic period). These occur in Ibn Habib, Kitab al-​ Muhabbar, 305–​309. This glosses over much complication and misunderstanding. In the first place, there are no Rumiyyat to be found in this part of the Muhabbar, as claimed by Athamina, “How Did Islam Contribute,” 385, note 9. Athamina seems to be referring to the list of Nisraniyyat (who are not necessarily Greek, and in the first entry we are told specifically the woman in question is, in fact, Habashiyya). Also, although some of these women are listed as being slaves, we should not assume the rest of them also had this status just because they were Christian or Abyssinian. In addition, many of these men were born during the Islamic period and not the jahiliyya. 32. The Muslims could have had intercourse with slave women without this resulting in a significant number of pregnancies; withdrawal as a contraceptive method can be as effective as condom use in practice. The method was later sanctioned in Islamic law; see Giladi, “Birth Control.” Also see Jones et al., “Better Than Nothing.” 33. Based on average rates across the tribe, we would expect a group with the same number of children (n = 181) across the same spread of fathers by generation to have had 69 of these born to concubines; the Umayyad caliphs and their sons have 76 children born to concubines. Also see Table 1.6. 34. A separate database compiling the marriages of the Generation 7 sons of concubines (who would have been marrying during the late seventh/​early eighth centuries) as recorded in the Nasab Quraysh reveals that they made 15 marriages to Arab women; all but three of these were to Qurayshi women. 35. Athamina, “How Did Islam Contribute,” 394. 36. For instance, Sad ibn Abi Waqqas, who—​although ostensibly a member of the Quraysh tribe, an early convert, a Prophetic Companion, and a cousin of Muhammad’s mother—​is not recorded in the Nasab Quraysh as having married any Qurayshi women. The case is discussed further by Ahmed, The Religious Elite, 46–​47; it is clear from this that Sad’s claim to be a member of the Quraysh tribe was not widely accepted. It should be noted that, although Ahmed does find one Qurayshi wife for Sad in another source, this source cannot name her. 37. This is admittedly a small sample, but if we include the 11 marriages made by the brothers of the two men, we come to similar conclusion; in total, 7 of the 19 marriages of this new sample were to non-​Quraysh Arab women, 6 were to non-​Umayyad Quraysh, and 5 to Umayyads (there is also one woman of unknown origin). 38. See Crone, Slaves on Horses and Banaji, “Late Antique Legacies.” 39. Athamina, “Non-​Arab Regiments,” 368 ff. 40. Hawting, “Marwān II.” 41. The maternal details provided in the Nasab Quraysh are remarkable not just by the standards of Islamic history. No other premodern tribal culture known to this author has ever created anything approaching this density of information. The discovery of a similar data set belonging to a different population would be very welcome, because it would aid comparative studies.

Bibliography Primary Sources Ibn Habib, Abu Jafar Muhammad. Kitab al-​Muhabbar. Edited by Ilse Lichtenstadter. Hayderabad: Matba`at Jam`iyat Da`irat al-​Ma`arif al-​Uthmaniyya, 1942.



Statistical Approaches to the Rise of Concubinage in Islam

j 25

Ibn Khaldun, Abd al-​Rahman ibn Muhammad. Al-​Muqaddima. 3 vols. Edited by Ali Abd al-​ Wahid Wafi. Cairo: Dar Nahdat Misr, 2004. Ibn al-​Nadim, Muhammad ibn Ishaq. Al-​Fihrist. Cairo: Maktabat al-​Tijariyya al-​Kubra, 1929. Ps.-​Jahiz, Abu Uthman. Al-​Mahasin wa’l-​addad. Beirut: Dar Maktabat al-​Irfan, 1955. al-​Khatib al-​Baghdadi, Ahmad ibn Ali. Tarikh Baghdad. Cairo: Maktabat al-​Khanji, 1931. al-​Zubayri, Mus`ab ibn Abd Allah. Kitab Nasab Quraysh. Edited by E. Levi-​Provencal. Cairo: Dar al-​Ma`arif li’l-​Tiba`a wa’l-​Nashr, 1953.

Secondary Sources Ahmed, Asad. The Religious Elite of the Early Islamic Hijaz: Five Prosopographical Case Studies. Oxford: Linacre College Unit for Prosopographical Research, 2010. Athamina, Khalil. “How Did Islam Contribute to Change the Legal Status of Women: The Case of the Jawārī or the Female Slaves.” Al-​Qantara 28, no. 2 (2007): 383–​408. —​—​—​. “Non-​Arab Regiments and Private Militias During the Umayyad Period.” Arabica 45, no. 3 (1998): 347–​378. Banaji, Jairus. “Late Antique Legacies and Muslim Economic Expansion.” In Money, Power and Politics in Early Islamic Syria, edited by John Haldon, 165–​180. Farnham: Ashgate, 2010. Bashear, Suliman. Arabs and Others in Early Islam. Princeton, NJ: Darwin Press, 1998. Bernheimer, Teresa. “Genealogy, Marriage and the Drawing of Boundaries.” In Sayyids and Sharifs in Muslim Societies: The Living Links to the Prophet, edited by Kazuo Morimoto, 75–​91. Abingdon: Routledge, 2012. Brockopp, Jonathan. “Concubines.” In The Encyclopaedia of the Qur’an, Jane Dammen McAuliffe, gen. ed. Leiden: Brill Publishing Online, 2015. —​—​—​. Early Maliki Law:  Ibn ʻAbd al-​Hakam and His Major Compendium of Jurisprudence. Boston: Brill, 2000. Bulliet, Richard. Conversion to Islam in the Medieval Period: An Essay in Quantitative History. Cambridge, MA: Harvard University Press, 1979. Caskel, Werner. Gamharat An-​nasab:  Das Genealogische Werk Des Hisam Ibn Muhammad al-​ Kalbi. Leiden: Brill, 1966. Coquery-​Vidrovitch, Catherine. “Women, Marriage, and Slavery in Sub-​Saharan Africa in the Nineteenth Century.” In Women and Slavery:  Africa, the Indian Ocean World, and the Medieval North Atlantic, edited by Gwyn Campbell, Suzanne Myers, and Joseph Miller, 63–​82. Athens: Ohio University Press, 2007. Crone, Patricia. “The First-​Century Concept of Higra.” Arabica 41, no. 3 (1994): 352–​387. —​—​—​. Medieval Islamic Political Thought. Edinburgh: Edinburgh University Press, 2004. —​—​—​. Slaves on Horses. Cambridge: Cambridge University Press, 1980. Evans-​Pritchard, E. “Nuer Time-​Reckoning.” Africa: Journal of the International African Institute 12, no. 2 (1939): 189–​216. Giladi, Avner. “Birth Control.” In The Encyclopaedia of Islam, 3rd ed., edited by Kate Fleet et al. Leiden: Brill Publishing Online, 2009. Goldziher, Ignac. Muslim Studies. Vol. 1. Translated by C. R. Barber and S. M. Stern. Chicago: Aldine Publishing, 1966. Hawting, G. R. “Marwān II.” In The Encyclopaedia of Islam, 2nd ed., edited by P. Bearman et al. Leiden: Brill Publishing Online, 2015.

26 i

Majied Robinson

Hoyland, Robert. Seeing Islam as Others Saw It. Princeton, NJ: Darwin Press, 1997. Jones, Rachel K., Julie Fennell, Jenny A. Higgins, and Kelly Blanchard. “Better Than Nothing or Savvy Risk-​Reduction Practice? The Importance of Withdrawal.” Contraception 79, no. 6 (2009): 407–​410. Lewis, Bernard. Race and Slavery in the Middle East:  A  Historical Enquiry. Oxford:  Oxford University Press, 1990. Matsumura, Shuichi, and Peter Forster. “Generation Time and Effective Population Size in Polar Eskimos.” Proceedings of the Royal Society B: Biological Sciences 275, no. 1642 (2008): 1501–​1508. Pellat, Charles. “Muṣʿab.” In The Encyclopaedia of Islam, 2nd ed., edited by P. Bearman et  al. Leiden: Brill Publishing Online, 2015. Rotter, Gernot. “Maslama b. ʿAbd al-​Malik b. Marwān.” In The Encyclopaedia of Islam, 2nd ed., edited by P. Bearman et al. Leiden: Brill Publishing Online, 2015. Schacht, Joseph. The Origins of Muhammadan Jurisprudence. Oxford: Clarendon Press, 1950. —​—​—​. “Umm al-​Walad.” In The Encyclopaedia of Islam, 2nd ed., edited by P. Bearman et  al. Leiden: Brill Publishing Online, 2015. Szombathy, Zoltan. The Roots of Arabic Genealogy:  A  Study in Historical Anthropology. Piliscsaba: Avicenna Institute of Middle Eastern Studies, 2003.

2 Abbasid Courtesans and the Question of Social Mobility* Matthew S. Gordon

i  Modern scholarship has long recognized that slavery played a part in the formation of Near Eastern society during the early and middle Islamic periods (c. 650–​1250 ce), but it has yet to take up a number of questions surrounding what was, after all, a significant chapter in the history of the institution across the ancient and medieval world. Thus, much remains to be understood about the extent and nature of early Near Eastern/​ Islamic society’s reliance on slave labor.1 This essay considers one such question: the rise to prominence on the part of enslaved and freed persons resident in the major urban centers of the first Abbasid period (c. 750–​900). It does so using the example of elite female performers at the Abbasid court, and, as evidence, a set of passages concerning three of the women, all of which occur in the tenth-​century Kitab al-​Aghani (Book of Songs) by Abu al-​Faraj al-​Isbahani (d.c. 972).2 The passages voice the same complaint:  that the singer in question was wrongly enslaved. These brief texts speak to the question of upward social mobility. The singers, despite the odds, achieved—​and in certain cases, sustained—​preeminence. The phenomenon is familiar in general terms to historians of the period; a number of early Abbasid-​ era notables were enslaved or the (free) offspring of enslaved parents before achieving elite standing, whether as members of political, commercial, and military circles or at the highest levels of culture and scholarship.3 The evidence indicates that the singers confronted a cluster of obstacles related directly to their standing as enslaved persons. The discussion that follows catches the singers at an early stage in their careers—​that is, at a point when preeminence lay at a discernible reach. *  In memory of Wolfhart Heinrichs, for his generosity and example.

27

28 i

Matthew S. Gordon

The women lived and worked in ninth-​century Baghdad and Samarra, each of which served in turn as the administrative, military, and cultural hub of the early Abbasid Empire. They were, taking contemporary and later Arabic sources as our guide, highly trained and ambitious denizens of Abbasid palaces and fine homes. Their careers entailed the provision to a mostly male cohort of companionship; sexual intimacy, although perhaps as often in the promise than the deed; witty and informed repartee, both verbal and written; a variety of entertainment, including chess and backgammon; and, above all, engaging verse and music performance, instrumental and vocal alike. It is appropriate to refer to the women as courtesans.4 Modern scholarship has shown much interest in these women.5 Too often, however, their experience as enslaved persons and their place in the history of medieval Near Eastern slavery has been considered secondarily.6 This discussion, then, contributes to “writing the feminine,” but has as a principal concern the study of that slave history as it played out in the first Abbasid period.7 The courtesans were a rarefied group, in many ways quite unlike most other female denizens of Samarra and Baghdad, slave or free. There was nonetheless sufficient overlap between their lives and those of the much wider numbers of enslaved women in the period to justify the approach adopted here. This is not to suggest homogeneity in standing of enslaved persons in medieval Near Eastern society (this did not exist), but rather the value of the considerable information on the courtesans to the wider study of Abbasid-​era slavery.8 The three protest passages serve as a small but important example. Each of the passages in the Book of Songs concerns one of the elite singers respectively: Fadl the Poet, Arib al-​Ma’muniya, and Shariya. Al-​Isbahani, in the case of Fadl and Arib, has the courtesans themselves protest their treatment as enslaved persons, and, thus, the manner in which they are handled by slave merchants and owners. The third example involves Shariya. In this case, she does not speak directly, although the text addresses the same underlying question in some detail. One might cite further cases as well, including that of Inan al-​Natifi. The preoccupation with a beating that she endured at the hands of her owner, identified only as al-​Natifi, suggests a view of inappropriate if not illicit conduct toward the enslaved.9 Violence visited by male handlers on enslaved (or freed) women occurs in profiles of others of the courtesans as well. The example concerning Fadl speaks to her sale, either by her father or by a stepbrother. This is al-​Isbahani’s text: Fadl was a slave girl [jariya]10 born in Basra of mixed parentage (muwallada min muwalladat).11 Her mother was a woman of mixed parentage of the Yamama. Born to her mother, Fadl was raised in the house of a man from the Abd al-​ Qays who sold her after seeing to her education and training. She was purchased then gifted to the caliph al-​Mutawakkil. She herself claimed that it was her brother who sold her; her father had slept with her mother who had given birth to Fadl with him, and he had educated and trained her, acknowledging that she was his



Abbasid Courtesans and the Question of Social Mobility

j 29

offspring. But, when his sons, of a woman other than her mother, resolved to sell her, he disavowed her (jahadaha).12 Even after she was manumitted, she was known only as Fadl the Enslaved.13 The passages concerning Arib are more substantial, with the two sections about her in the Book of Songs among the longest and most riveting of the courtesan texts.14 Arib appears as a palpable figure of sharp edges and authority; the pointedness of the references in which she makes an issue of her legal standing suits the overall thrust of the text nicely. In two episodes, perhaps a case of textual doubling, Arib escapes her master for the company of a lover. When the master administers a beating by way of punishment, this in the one episode, she insists he has no right to strike her and adds that, if indeed she is enslaved, then he must sell her. “O, you, would you kill me? I have no reason to put up with this from you, for I am a free woman. But, if enslaved (mamluka), then you must sell me. I shall not endure this indignity!” The second passage has the lover arrested and sent off to be whipped. Arib rides to the prison atop a mule and, her veil pushed aside, proclaims: “I am Arib, and if I am enslaved (mamluka) then have my master sell me but, if I am a free woman, than he has no claim on me.”15 The references to Shariya’s enslavement, our third example, are more involved. They occur early in her profile in the Book of Songs and are nearly as dramatic as those concerning Arib. Al-​Isbahani references a book on Shariya compiled by the late ninth-​century poet and Abbasid family member Ibn al-​Mutazz (d. 908).16 The text sets the question of Shariya’s legal standing against the backdrop of an intricate pas de deux between Shariya’s mother and Ibrahim ibn al-​Mahdi—​Shariya’s owner—​over control of the young singer’s affairs. The key bit is a claim (Arabic qila, “it was said”) that Shariya was born to a slave mother and a Qurayshi father (thus a free male). The implication is that, upon the birth, her mother became an umm walad (“the mother of the master’s child”). Thus, by virtue of her father being free, Shariya herself might also have taken on new legal protections commensurate with her new situation; but the passage states explicitly that her father disavowed her (again, jahadaha) and, thus, like her slave mother (ama), she “entered slavery.” The passage adds (that it was thought) that she was then stolen and sold to a woman of the Hashimite clan.17 Might these singers have asserted themselves in this fashion? Possibly. But we cannot know for certain. Might women of their station have made such claims? The sources would have us believe as much. What matters here is that the three passages speak to issues surrounding Abbasid-​era enslavement and, in particular, the production of trained slave women and a reliance on concubinage—​an institution in wide use across early Islamic/​ Near Eastern society (concubines during this and later periods were nearly always enslaved women).18 They allow one to contribute to a wider discussion on female enslavement by historians of slavery across regions and time periods, particularly in the Mediterranean and Islamic contexts.19 Physical and emotional violence against the enslaved was, it is fair to say, a constant; although, relative to the suffering endured by enslaved persons of, say,

30 i

Matthew S. Gordon

the much later Atlantic slave trade, it appears, in the Near Eastern case, to have been far more moderate in scale.20 The passages serve also as evidence of the particular standing of the courtesans; one understands from the texts that they were situated so as to be able to voice their claims in public and, yet, allowed that standing and privilege by a predominantly male social sphere. Or, to put it differently, to achieve preeminence, it was necessary to seize on every opportunity, with voice and body alike, although such conduct invited all manner of (mostly) male response, from admiration to abuse. Hilary Kilpatrick has translated a strikingly interesting text that speaks to this dynamic. As with the protest passages, it is something attributed to a courtesan. It occurs in another and much shorter of al-​Isbahani’s works, the Ima al-​Shawa’ir (The Female Slave Poets). The courtesan is Murad, described as a jariya belonging to Ali ibn Hisham (d. 832), a ninth-​ century Baghdadi notable and, as the texts would have it, connoisseur of fine singers.21 Murad and Ibn Hisham quarrel, as lovers will, but he feigns disinterest in reconciliation. Reacting to his cruelty, Murad sings, in pained expression of her “own dignity and worth,” now violated: Since I am a slave twice over, to love and to my owner, I must endure, though endurance goes against the grain, close my eyes upon the speck of grit therein, and submit to being humbled and coerced, as befits a slave.22 The Sources The entry of courtesans to high urban society was, in large measure, ensured by the patronage of male owners, slave merchants, and intimate companions. But the sources would have us pay close attention to the women-​as-​actors, in an individual and collective sense. The bulk of evidence, much of it anecdotal, occurs in works of history, poetry, theology, law, and medical science.23 The lion’s share, however, lies in works of Arabic adab or belles-​lettres (a wide rubric, to be sure). Modern scholars most often cite two works: al-​Isbahani’s Book of Songs and the Risalat al-​qiyan (The Epistle on Singing-​Girls) of al-​Jahiz (d. 869).24 There are other excellent sources on hand as well, but these are essential.25 In using them to reconstruct the careers of the courtesans, one has to negotiate a pair of problems. One comes away from the two works with contrasting impressions of the women; the texts work nearly at cross-​purposes.26 Al-​Jahiz was drawn to the courtesans. His essay is one of several writings in which he subjects the women to his sharp gaze, although nearly always in sweeping fashion, with scant reference to individuals. But he was little impressed, viewing the conduct of the women as deleterious to the contemporary social-​ moral order.27 The epistle treats the women at a distance, with al-​Jahiz making the same effort as the writer. He urges his male audience to keep the sirens at bay. The implication is that the women were irredeemable; their male counterparts stand some chance of rescue, although scarred—​morally speaking—​by their dalliances.28 Al-​Isbahani, in contrast,



Abbasid Courtesans and the Question of Social Mobility

j 31

looked on the women with favor, if not abiding affection. I suggest, for example, that he demonstrates his sympathy for the singers in citing the violence visited on them (thus, in this regard, joining al-​Jahiz in condemning the male handlers). The women of the Book of Songs come off as forthright, quick-​witted, and, thus, compelling actors in the myriad stories that make up this most remarkable of books. Al-​Isbahani allows one to hear their voices rise in indignation. The attention devoted to the singers by the two prominent male writers evinces, in its own right, the ascendance of the women to prominence. Al-​Isbahani seems particularly keen to assign the women an active voice—​that is, something close to the “agency” of modern discourse. Arib al-​Ma’muniya, Fadl the Poet, Shariya, Inan, and the other members of their select circle emerge as alert to opportunity, the authors of their own careers, as players with rich repertoires. The women, it is true, owed much to a small industry in the production of fine entertainers.29 One can cite, for example, the activity of Ibrahim (d. 804) and Ishaq (d. 850) al-​Mawsili, brilliant father-​ and-​son musicians of the Abbasid court, but specialists, too, in the trade and training of enslaved female entertainers.30 No less indicative of the production of elite singers are the circles of younger women that accompanied Arib and Shariya. This was a case of freed women, at the peak of their careers, nurturing enslaved apprentices of their own.31 Young slave singers were afforded opportunity to succeed, but this was a fiercely competitive cultural environment.32 To gloss over the choices and determination of the individual courtesans themselves would be to miss an essential feature of the portraits contained in the Book of Songs. Arib, Shariya, and the others drew notice—​from writers and the Baghdadi jet set—​for their skillful play, of instruments and lovers alike, a livelihood to which the women attended closely. The contrasting views of the courtesans bring up a second problem. The sources are replete with literary device; the women are done up in layers of words.33 This is the stuff of the many stories (Ar. khabar, pl. akhbar) that make up al-​Isbahani’s portraits. The highly fashioned character of the evidence obstructs the social historian’s ability to move forward. One option is to consider only how the women are represented: that the courtesans won entry to elite cultural circles seems obvious, but too much language stands in the way of a nuanced description of their achievements.34 But what does occur are patterns in the texts—​that is, a cumulative effect of elements that recur across the stories of the courtesans and that turn, in most cases, on a particular theme, such as that of protest. These patterns suggest something observed (by al-​Isbahani and his sources). They are not, of course, foolproof; rather, they suggest background knowledge of the women and their careers. The patterns cast light on tangible features of the lives of the singers.35 The textual pattern considered here has to do specifically with, on the one hand, the violence and illicit conduct to which traffic in slaves lent itself, and, on the other, the ambiguity of standing insofar as access to, and activity in, the public arena was concerned. Al-​Jahiz, in his sharply worded epistle, decries the courtesans’ many efforts at subterfuge. They would lie to their lovers, for example, claiming they were free persons, this to conceal how exorbitantly expensive they would be to purchase.36 Al-​Jahiz would

32 i

Matthew S. Gordon

have none of it. One wonders if, in this case, he saw the offense to society as a matter that divided owners and owned: the women’s conduct interfered with the proper pursuit of commerce. But it is this same question that arises in al-​Isbahani’s profiles of the courtesans, although, again, in more sympathetic fashion. The two authors signal the appearance in early Abbasid society of a particular source of controversy surrounding the women. It was a controversy of elite society, thought of properly as public scandal. By the same token, the references underscore the rarefied standing achieved by the leading singers and its persistent ambiguity. Evidence provided by two other writers of the period, al-​Yaqubi (d. c.  897)  and al-​ Masudi (d. 956), offers richer context to the protest passages.37 It points to significant facets of the Abbasid-​era slave trade and, specifically, the vigor of that commerce. The information from al-​Yaqubi derives from his Kitab al-​buldan (Book of Lands), an irreplaceable source on ninth-​century Abbasid urban geography. It occurs in the section on Samarra, the second of the great Abbasid cities.38 Contained therein are specific references to slavery, most of which concern the formation of the Abbasid slave military, a new-​style armed force made up of recruits from regions of the Caucasus and Central Asia.39 But, they include as well a reference to women:  al-​Mutasim (r. 833–​842), the Abbasid caliph associated most closely with the slave military’s early history, ordered the distribution of slave women (jawari) to the new recruits. Al-​Yaqubi says little to identify the women, but he adds that the soldiers were to marry their new partners and establish families, and were strictly forbidden from divorcing or abandoning them. It appears as a case of social engineering on an imperial scale.40 Al-​Yaqubi refers elsewhere to a slave market located in the midst of the new capital, with its central location suggestive of a significant part in the commercial life of the city. A passing reference, however, it does not indicate how the site functioned relative to the wider slave trade for which it offers physical representation.41 One question is whether it functioned solely for nonelite commerce in enslaved persons, which is likely—​meaning, it was not a place to acquire the best singers.42 Al-​Masudi’s text concerns Mahbuba, among the best known of the courtesans.43 It works, in its way, as a further protest passage. The single name, bereft of all reference to kin and other social ties, bespeaks her legal standing.44 The text points to the complexity of background of the enslaved women and a shift on Mahbuba’s part from anonymity to prominence. It alludes to gifts sent by notables to the ill-​starred caliph, Jafar al-​Mutawakkil (r. 847–​861), upon his ascent to office. Muhammad ibn Abdallah ibn Tahir (d. 867), governor of Baghdad (and a cousin of Ali ibn Hisham, Murad’s owner and lover), is reported to have sent 200 slaves, male and female. The reference can be joined to that of al-​Yaqubi regarding the caliph’s distribution of enslaved women; both evince the Abbasid court and elite households as engines of a vigorous slave trade. The text indicates that Mahbuba was among the young women offered to al-​Mutawakkil and adds that she had been trained as a singer—​in her case, in Taif, a town in western Arabia. It was one of several such places, including Medina, that served as venues for the production of fine



Abbasid Courtesans and the Question of Social Mobility

j 33

performers.45 Al-​Isbahani, in his profile of Mahbuba, refers to her specifically as a muwallada, the term introduced earlier.46 Yusuf Ragib and Naim Vanthieghem render the term usefully as “born in the home (i.e., of the master).”47 “Of mixed background, free Arab and enslaved Other,” works as well. The text goes on to describe her skill as a singer and wordsmith, relating it directly to her subsequent relations with al-​Mutawakkil: “she held a lofty place in his affections unrivalled by any other person.”48 It then turns abruptly to Mahbuba’s fate following the caliph’s assassination in 861 by officers of the new Central Asian/​Turkish military. Bugha the Elder, a prominent commander, is said to have availed himself of Mahbuba and others of the monarch’s female slaves after the killing. The suggestion is that he plundered the women from the caliphal household.49 In some unspecified social gathering, she is pressed to perform by Wasif, another of the commanders. Inconsolable over the death of her lover (the caliph-​brought-​low), she can sing only of his bloodied remains and her grief.50 Wasif, enraged, orders the singer imprisoned; the text implicates the commanders in the crime of regicide. Mahbuba was never heard from again. An apocryphal text, it is in fact but one version of the story. The version contained in the Nisa al-​Khulafa (The Women of the Caliphs) of Ibn al-​Sa`i, a thirteenth-​century Iraqi writer, has Bugha acquire the singer from Wasif. He manumits her and allows her to reside where she chooses; Mahbuba departs for Baghdad, where she begins a solitary life.51 Both versions sit well with a pattern of references to the highly public face of the courtesans, the fragility of the social ground on which they performed, and a willingness to intervene with one of the few instruments at their disposal: their words/​voices. The texts speak, in sum, to significant features of the Abbasid-​era slave market and the specific experience of enslaved women, especially those groomed for high culture. On the dimensions of the trade, the numbers are not to be read literally, but as markers of scale. Thus, for example, the reference to al-​Mutasim’s distribution of young women: the new army is likely to have numbered in the mid thousands and, even if only some number of the soldiers were assigned women, this adds up to many such individuals.52 Ibn Tahir’s gift is another case in point. But, if the texts indicate large-​scale trading of young women and their consumption by a flourishing imperial/​urban economy, they leave open questions surrounding the nuances of each courtesan’s life, her rise to prominence and, by extension, the culture of Near Eastern/​Islamic enslavement. The Dynamics of Abbasid-​E ra Female Slavery There is, first of all, the question of origins. Michael McCormick has argued for the “insatiable” appetite for enslaved labor on the part of Near Eastern and North African cities.53 Although certainly correct, his stress on Europe as a principal source of such labor risks skewing a larger body of evidence: Arabic texts point to a great variety of regional and, hence, sociocultural background on the part of Abbasid-​era female slaves.54 The Book of Songs, in a valuable reference, has Ibrahim al-​Mawsili (the father)—​slave-​trader and

34 i

Matthew S. Gordon

cherished performer—​as the first to train “beautiful” or “highly valued” women, thus breaking with the practice of training those with darker or tawny skin (Ar. safra, lit. “yellow”).55 At least one modern scholar has understood “beautiful” as referring to “white”—​ thus, presumably, European women.56 Many of the leading courtesans, incidentally, are said to have been safra, which raises a question of the impact of this shift on the careers of the established singers. But this reading—​“European” or “white” women—​needs to be considered. The reference may have more to do with the commodification of women, physically speaking, than with ethnicity.57 The key point, however, is that a seemingly well established trade network imported enslaved women from a variety of regions outside the Arab/​Islamic realm, including Armenia, the Caucasus, Transoxiana, the Indian Ocean, and the East African coast. No less vital a point is that one has to distinguish these women from the muwalladat—​ enslaved individuals who were native-​born. There is good evidence—​such as that of the protest passages—​that controversy surrounded the experience of the women (and men) who were second-​generation enslaved. Their legal standing depended on that of their parents, and it is there that controversy ensued. The point for now is that evidence regarding the courtesans is also evidence that enslaved persons in Near Eastern society reproduced, perhaps in significant numbers.58 Reproduction, for example, was the ostensible aim of al-​Mutasim’s policy of providing women to his Central Asian soldiers; the Central Asian/​Turkish forces did reproduce, although whether as a direct result of this policy is an open question.59 Statistics on the comparative numbers of foreign-​born and domestically produced enslaved persons are beyond our reach. The distinction of categories is nonetheless clear. A key point, one seldom stressed by modern scholarship, is that all of the leading courtesans—​seemingly without exception—​are said to have been born and bred in Near Eastern society. They were all muwalladat. It is safe to presume their mothers were imported from other regions, but the daughters themselves were of Near Eastern origin.60 Their precise legal standing was in question but, in many cases, the daughters were treated as slaves and, thus, two generations of enslaved women lived and worked together in urban Near Eastern households. This can be read as an implication of the protest passages and also points to the relatively humble origins of each singer. It also stands as an example of the textual patterns alluded to earlier. So, for example, if one wishes to dismiss a single case—​ the assertion that Arib was born to Jafar ibn Yahya (d. 803), a member of the renowned Barmakid family, and Fatima, an otherwise obscure concubine in his household—​it is another matter to set aside the fuller body of evidence regarding reliance on concubinage, and its impact on middle and high Abbasid-​era society, to which her example points.61 The distinction of background goes, in turn, to the compelling but sadly elusive subject of cultural competence and engagement on the part of Abbasid-​era slave women. With regard to the foreign-​born women, what linguistic and other cultural memory did they bring to elite Baghdadi and Samarran households? Some numbers of the women surely were employed as nursemaids. In what language did they murmur, hum, and sing to their



Abbasid Courtesans and the Question of Social Mobility

j 35

charges? D. Fairchild Ruggles has raised the question in relation to foreign-​born concubines of Umayyad Spain (al-​Andalus).62 For the enslaved women brought to Abbasid cities, unfortunately, there is almost no such evidence on hand. But, for the native-​born girls, does their skill with the one language, Arabic, stand as evidence of this kind? There are no indications of what languages they might have learned from their mothers apart from Arabic. But, in regard to that language, there is an argument to be made regarding cultural competence on the part of these girls, in both a pedagogical (training) and commercial (marketing) sense. The muwalladat were exposed to Arabic from birth; one presumes they were native Arabic speakers. This must have represented an advantage, one no doubt appreciated by their male handlers and one that worked for the courtesans themselves. A successful career at the highest levels of Abbasid culture turned, in other words, on a display of skills in a variety of forms of that same language—​not only sung and spoken, but also informed by cultural nuance of the sort typically acquired with very early exposure. Consider the frank exchange between Fadl, a leading courtesan associated closely with al-​Mutawakkil, and Abu Dulaf al-​Ijli (d.c. 842), a prominent poet. Testing his younger colleague—​the text implies that Fadl was newly introduced to the imperial court—​he challenges her to match several lines of verse, including “How great the difference between a pierced pearl and the pearl that is not,” to which she replies, without blinking, “And the pearls are useless to their users unless pierced by a borer so they may be properly strung.”63 Many are such passages in which the women join deft handling of language to quick repartee and innuendo, often, as here, sexually charged. One cannot help but suggest, then, that the reference to “properly strung” suggests coercion at the hands of those wielding the bore. The language was courtly Arabic.64 My point is that, by virtue of the fact that they were born to households in which Arabic was a principal tongue, the girls were ready—​setting aside other variables—​for advanced training in poetry, song (writing and performance), and the like. Alongside origin is the matter of selection. The newly enslaved were subject to close physical examination—​a process of culling out, to put it crudely. Al-​Jahiz refers in his epistle to the need for proper selection, and it is worth noting that al-​Isbahani points out routinely the visual impression left by each of the women.65 So, for example, Arib, whom he celebrates for a range of professional accomplishment, was also striking in appearance;66 or Inan, “tawny, with a gorgeous face”;67 or Fadl, “exquisite of face, body and bearing”;68 or Mahbuba, who, despite having an innate gift for (Arabic) poetry that was just surpassed by Fadl, was nonetheless “the more physically striking and chaste” of the two women.69 Close physical examination of enslaved women’s bodies is one dynamic for which there is documentary evidence. Yusuf Ragib has edited and translated a set of Arabic sale receipts from ninth-​and tenth-​century Egypt, several concerning families, but most individual enslaved women. Each comments directly on the condition of the women’s bodies, including their sexual organs (all of which, unsurprisingly, given that the context was one of marketing “goods,” are said to be free of blemish and disease).70

36 i

Matthew S. Gordon

The point is twofold. First, to be judged suitable, the women, including those destined for a life of performance, had to meet a standard of physical allure and, second, undergo intrusive and humiliating inspection. The courtesans-​to-​be shared with their peers in this regard. One wonders, given the particulars of the market in performers, if they were not subject to specific forms of testing, say, of their posture and stance, or powers of voice and mental recall. It needs little reminder, perhaps, that enslaved adolescents (girls and boys) were evaluated as potential sexual partners as well. It is likely that most enslaved persons during the Abbasid period were destined for the Near Eastern domestic market—​that is, households and palaces. The presumption is that as Abbasid urban society flourished, the market for domestic labor did as well.71 Some, perhaps many, young women were destined for concubinage, whether such was the initial intent of their purchasers. Classic Islamic law (and Near Eastern social practice) allowed male owners to sleep with their slaves. In the absence of hard numbers, unfortunately, we have to rely on anecdotal evidence with regard to the prevalence of concubinage. In addition, several key reports (akhbar) in the works of al-​ Tanukhi, a tenth-​century Iraqi writer, suggest a busy Near Eastern sex trade as well.72 The presumption must be that, as a sector of the Iraqi urban economy, the sex trade made considerable demands on slave traders, perhaps to no less a degree than the urban households. Concubinage worked, effectively, as a considerable overlap of the two sectors. Our concern, then, in the case of the singers and poets, is with a particular culture of demand for young women who offered a mix of sexual attraction and ready skills. The requirements of beauty and performance ability issued from the dynastic court and, it seems clear, elite Iraqi households for which the Abbasid majlis (pl. majalis, cultural gathering) offered both a model and a challenge—​that is, a standard to emulate and, when possible, improve.73 This goes to the matter of training—​yet another facet of the luxury slave market. We can turn again to the examples of individual women. Mahbuba was trained by an unnamed individual in Taif, again, one of several Arabian towns known for the production of singers. Arib, also native-​born, is reported to have been trained initially by her first owner, one Abd Allah ibn Ismail al-​Marakibi, an official of Harun al-​Rashid’s administration. The Book of Songs indicates she was schooled in the southern Iraqi city of Basra in poetry, singing, and related arts before her acquisition by al-​Amin (r. 809–​813), an Abbasid caliph (and, like al-​Mutawakkil, the victim of regicide).74 Various references, including one in which she coaches her own enslaved apprentices in the tuning of the oud, suggest that, by late career, she had mastered the instrument and become herself a client of slavers.75 As seen already, Fadl, also a muwallada, is said to have been raised and educated by her father. The same reference, in the Book of Songs, goes on to say that he then sold her and, at some point, she was then gifted to al-​Mutawakkil.76 (The passage follows in which she makes the claim, instead, against her stepbrother who, one supposes, like the father, recognized a source of considerable profit). The many references make clear a web of practice in which the courtesans-​in-​making were produced by local households, prepared for a market in luxury slaves, and sold by



Abbasid Courtesans and the Question of Social Mobility

j 37

“adult familiars”—​a phrase borrowed from Madeline Zilfi.77 Shariya is a case in point, if we can trust a further anecdote. It has the unnamed Hashimite woman take the girl to Baghdad, where she offers her to Ishaq al-​Mawsili. He is sorely tempted, but finds the asking price too high. The woman then offers her to Ibrahim ibn al-​Mahdi, Ishaq’s sworn rival, who pays the sum but then orders the girl to be subjected to intensive training for a year. He offers her then to al-​Mawsili, who eagerly pays a sum 10-​fold higher, at which point Ibn al-​Mahdi reveals her identity, all to al-​Mawsili’s embarrassment.78 This is all contrived, of course, and has mostly to do with the rivalry of the two great male performers. But the text plays with, rather than invents, a pattern of commodification of young, talented women. This same point goes to the part played by the “adult familiars.” The text in which the protest passage is embedded refers to two women who are reported to have been involved closely with the singer-​to-​be in this first part of her life:  the “Hashimite woman” and Shariya’s enslaved mother. The text has, in striking fashion, the mother work, at a later point and behind the scenes, to influence the course of her daughter’s career. She is said to have intervened with influential persons against Ibrahim ibn al-​ Mahdi, Shariya’s owner. She complains he was providing poorly for the girl and demands that she, the daughter, be taken in hand by the caliph (al-​Mutasim).79 It suggests an effort to accrue benefit from the daughter’s skills and appearance beyond the initial sale. The sources offer much other evidence as well. They indicate that instruction of the young women in the language arts, musical performance, and, perhaps, sexual expertise constituted a manufacturing sector for the local economies of Medina, Basra, and other such centers. Why these towns? Part of the answer is that they attracted well-​heeled pilgrims—​that is, a ready market. Ali ibn Hisham, for one, is said to have acquired Murad in Medina while on Hajj.80 Furthermore, the language arts and music are specific disciplines; given the many assertions of the courtesans as gifted artists, a fair assumption is that they received specialized training. It underscores, in its way, the quality and complexity of early Abbasid cultural production.81 It also complicates the question of whether “singing girls,” as a category, should be folded so easily into the larger category of “concubine,” which is to say that the complexity of the Abbasid slave trade ought to be better acknowledged.82 Finally, there is the connection of high culture and the Abbasid slave market. As noted earlier, the al-​Mawsili family produced fine singers and instrumentalists, as did other notable musicians.83 Ibrahim ibn al-​Mahdi (d. 839) trained young slave singers, although in his particular style—​one that stood at odds with that of the al-​ Mawsili family.84 Al-​Jahiz’s epistle contributes here as well; his targets of rebuke are noble sons, the guileless companions of singers, and scions of the best houses. Careers Early On Much awaited the young singers between their rude entry to the urban marketplace and the onset of their adult careers. They were, at this point, a distinct subgroup of an enslaved female population provided to Abbasid society by a complex local and

38 i

Matthew S. Gordon

transregional commerce. They possessed, at least in embryonic form, the appropriate skills and physical presence. They were poised to engage, as well, in the sorts of social networking on which rested most elite Abbasid-​era careers. And the market was ready to receive them. Many references speak of the considerable sums exchanged for the luxury person/​objects but, as well, the sums made available to the singers themselves.85 Ibrahim al-​Mawsili, according to his grandson, profited mightily from the proceeds of his sale of jawari.86 On hand, as well, is the anecdote in which Ibrahim ibn al-​Mahdi bests his rival in shaping Shariya into an expensive prize.87 For his part, and ever dismissive, al-​Jahiz sees the exchange of great sums for the singers as a scandal masked by “mere infatuation.”88 But successful singers stood to profit themselves. The Book of Songs speaks of the “considerable fortune” amassed by Nubayka, a male singer associated with Khumarawayh ibn Ahmad ibn Tulun (d. 896), a governor of ninth-​century Egypt,89 and there are indications that Arib, the great singer and composer, retired to a life of comfort.90 These courtesans-​in-​making were at a point of transition. Success, however, lay on the opposite side of all manner of social, cultural, and legal obstacles. What shape did these take? And how did the young women learn to confront them? It involved, one has to assume, considerable subtlety or, to borrow from Julia Bray, activity by the women that was “relational not absolute . . . they might participate in more than one sort of power structure.”91 They had to make their way, in other words, in and around two different if overlapping social worlds, in a professional capacity and in the more ordinary daily sense. These were, on the one hand, large urban households—​headed variously by caliphs, other Abbasid family members, preeminent merchants, officers, courtiers, and others92—​ and, on the other, elite society at large—​the public and private. The lines, always porous, were particularly so for these women. They were indeed liminal figures.93 The task is to identify the contingent facets of the courtesans’ careers. Some small number of the women managed quite well, but one has to assume a considerable winnowing of their ranks. Abbasid culture, like its politics, could be ruthless. Mahbuba’s disgrace at the hands of the Turkish commanders suggests as much. A great number of women no doubt ended up in less prestigious households or rougher venues—​ brothels, drinking houses, and so on—​or simply plied their trade quietly at court alongside more accomplished peers, among them Arib, Shariya, and others. The Book of Songs names, in passing, a good number of singers about whom nothing is known apart from their participation in the best majalis.94 Surely women must have faced great difficulty in overcoming the many obstacles in question. Therein lies the value of the protest passages. In reading them, a starting point is with patriarchy—​fundamental as it was to the organization of medieval Near Eastern society. The relative and hierarchical subordination of women to male authority was enshrined in Near Eastern social and cultural practice as well as Islamic law and tradition. Reports on the Abbasid majalis bear this out. The courtesans, as women, were a small minority in these cultural circles and were frequently called out on that basis. But, to insist on a patriarchal order is to miss the forest for the trees.95 To borrow again from Zilfi, admittedly



Abbasid Courtesans and the Question of Social Mobility

j 39

on the much later period of Ottoman imperial society, one is better placed in speaking of “multiple patriarchal forms, each supportive of male priority, in norms of social conduct, the distribution of power, and large and small habits of rule.”96 The male dominion remains intact, but proper due is given to a complexity of gender relations—​an intricacy of social dynamics—​that one would expect of the cosmopolitan households of Baghdad and Samarra. It is in negotiating the multiplicity of “patriarchal forms” that the courtesans rose to elite standing. Bray and El-​Cheikh, in their respective works on the writings of al-​ Tanukhi (d. 994), and Delia Cortese and Simonetta Calderini, in a welcome study on women of Fatimid Egypt (c. 909–​1171), offer a nuanced way forward in bringing the singers to center stage. El-​Cheikh, treating the Abbasid period, collects details regarding “the multiplicity of social and moral possibilities available to diverse categories of women.”97 Among these possibilities—​the avenues to social prominence—​were those offered by concubinage. Leila Ahmed has argued for Abbasid society as particularly “androcentric,” with Abbasid-​era males turning early on to concubinage on a vast scale: “The vast majority of the women with whom [elite men] interacted, and in particular those with whom they entered into sexual relationships, were women whom they owned and related to as masters to slaves.”98 A grasp on the dynamics of Abbasid-​era concubinage is vital, but one need not (and ought not) defend concubinage in any sense to raise objections. It is very difficult to establish the scale of reliance on slave labor, concubinage, or, for that matter, exploitation in Abbasid society.99 In addition, critically, the sources say little about the vast majority of Abbasid-​era women. There are, in other words, other reasons why this majority—​wives, daughters, and so on—​never appear in the chronicles and works of adab. They may be silent, but they were certainly not absent. One very likely reads of concubines and courtesans, more than other groupings of medieval Near Eastern women, because of their entry into the male arena. Only then did they become proper subjects. Can we hear their voices? The texts certainly indicate that the future courtesans, as evinced by the protest passages, took up the privilege of speaking out. Setting aside the element of protest for a moment, the three passages are compelling for what they say of the expectation that the young women, having learned to wield language, used it with conviction. It certainly caught the attention of the medieval writers. And it is, after all, what the courtesans were trained to do. The more resourceful they proved to be in this regard, the greater the potential for reward. But at issue here are the constraints that enslaved women faced in pursuance of elite standing. Mahbuba, one can say, chose to step over the line. Rather than perform choice favorites for Wasif al-​Turki, and knowing the cost, she opted to sing her mind, thus not only challenging the expectation that she perform but, more to the point, turning the audience’s attention subtly but unmistakably to the act of regicide and the commanders’ likely guilt. This is to suggest, of course, that we hear from the courtesans directly, but one need not resolve this question. The passage on Mahbuba’s fall from grace can be read alternatively as evidence for the considerable controversy that surrounded the women

40 i

Matthew S. Gordon

performers. This had to do not simply with their public conduct, but also with enslavement and treatment at the hands of male handlers. In this sense, her verse is joined easily to the protest passages, as further indication that something of the lives of the courtesans was awry. It is true that Mahbuba’s verse does not speak to enslavement per se, but the text built around the verse—​and her performance—​do have very much to do with abuse and injustice (visited on caliph and singer alike). Protest voiced by (and/​or attributed to) the courtesans points to enslavement and the violence attendant on that standing as a significant source of constraint on the women’s careers. With regard to the households—​the venue where the singers were first produced—​ many anecdotes relate interaction between the singers and their masters. These offer much evidence regarding Abbasid society’s use of concubinage. But, such interaction must surely have joined myriad other daily interactions with wives, daughters, and other free women.100 One can only suppose that the presence of the highly visible singers and poets gave rise to all manner of social relations, strained and otherwise. Unfortunately, to my knowledge, the sources offer almost no information of this kind.101 Bray and El-​ Cheikh, reading al-​Tanukhi, have done yeoman’s work in gathering references to patterns of domestic life. Many narratives speak to relations of slave women, including singers, with masters and other male counterparts, but, unfortunately, few consider interaction of enslaved and free women.102 There is, as well, the matter of slavery as a source of social stigma. Did enslavement leave a mark? Evidence—​Murad’s response to Ibn Hisham and the protest passages—​suggests as much. The text containing Murad’s verse implies that the poet/​singer directed it at Ibn Hisham at an advanced point in her career. Reference can also be made to the short manual produced by Ibn Butlan (d. 1066), a Baghdadi scholar of Nestorian Christian background and a physician by profession. The treatise addresses the purchase and physical examination of slaves, listing, in particular, the ethnic and regional origins of enslaved women.103 His assessment of the ingrained physical and moral qualities of each amounts to ethnic typecasting—​that is, as stigma that attached to the enslaved. Easy categorization may have served the slaver (nakhkhas) in marketing his product, but did little for the women. But one has also to consider that Abbasid-​era elites were often of enslaved or freed origin; the mothers of nearly every Abbasid caliph were concubines.104 In treating the forms of stigma that attached to enslaved and freed persons in the Abbasid era, it is vital, in other words, to consider both the vitality and variety of the ninth-​and tenth-​ century Near Eastern slave market. In this case, typically, the task is to identify the social and economic standing of the person in question; the powerless presumably deflected such stigma less successfully than their well-​appointed counterparts, enslaved or free.105 It is finally with concubinage that particular constraints come into focus. For a small, select group of Abbasid-​era women, concubinage was a route to respectability and influence. Achievement of this kind turned mostly on the birth, survival, and succession to high office of male offspring. For example, Shaghab, an umm walad of Greek (Byzantine) origin, and known in the sources as Umm al-​Muqtadir—​he, al-​Muqtadir,



Abbasid Courtesans and the Question of Social Mobility

j 41

reigned from 908 to 932—​joined a small cluster of early Abbasid women in gaining entry to high-​level decision-​making—​that is, in her capacity as queen mother.106 This does not appear to have been a route pursued by the courtesans, however, with a few exceptions. Arib, for one, is reported in passing to have had a daughter (illicitly).107 Of real interest is Mutayyam. Her profile in the Aghani reports that she was a muwallada, raised and trained in Basra by the al-​Mawsili father-​and-​son team and Badhl, another of the courtesans, then to have belonged to Lubana, the daughter of Abd Allah ibn Ismail (also Arib’s owner). She was then acquired, still a young girl (juwayriya), by Ali ibn Hisham, with whom she bore a daughter and two sons. The text adds that, on his death, Mutayyam was manumitted, presumably in her capacity as an umm walad.108 She appears to have been exceptional among the singers, in combining full-​time motherhood with a career in song. If lines of verse preserved by Ibn Qutayba (d. 889) hint at anything, it is that concubinage presented the young women with particular problems. The verse speaks approvingly of the widespread use by the singers of contraceptives.109 There is, first of all, the question of their mothers’ standing. It is this question that the protest passages address in particular (although not explicitly). Modern scholarship deals at some length with the umm walad—​the concubine who bears the child of the master. The birth of the child conferred on the mother a new standing. This is clear insofar as classical Islamic law was concerned; the umm walad earned a cluster of protections. She could not be sold, she was manumitted immediately on the master’s death, and her offspring were free automatically at that moment as well. There is much to suggest that the extension of these protections derived from a deeply held impulse to improve the situation of the mother, even at the expense of the father/​master.110 But a reading of the protest passages suggests that reference to the classical law might be misleading, offering, as it does, too simple a picture. Two modern studies of the umm walad indicate that Abbasid-​era jurists and scholars discussed the institution at length.111 Debate of this kind among Abbasid-​era legal scholars turned on at least three topics: acknowledgment of paternity (by the owner/​master), the sale of the umm walad (during the owner’s lifetime), and the position of the children of that union.112 Recognition of paternity was critical, at least in the sense that masters could deny that intercourse took place, which would leave the offspring attached to the enslaved mother’s standing. But, according to Jonathan Brockopp, reviewing the work of a prominent ninth-​century Egyptian jurist, Abd Allah ibn Abd al-​Hakam (d. 829), there was an option whereby the absence of any statement by the master meant that the slave mother still became an umm walad.113 This is but a single early scholar’s opinion, but it does point to ongoing debate on the topic in ninth-​century scholarly circles. Debate surrounded the sale of the umm walad as well. Majority opinion was that the slave mother “enjoyed the expectation of emancipation” and could not be alienated from the master (i.e., through sale); but, clearly, the view that she could be sold was still viable.114 Setting aside the many details of these debates, the following can be said. First, as

42 i

Matthew S. Gordon

suggested earlier, the medieval Muslim jurists appear to have made a concerted effort to provide specific protections to the concubines. This was, for example, the likely intent of the ruling regarding silence on the part of the master regarding paternity. But, equally, a degree of license remained to the master both in relation to paternity and marketing of the concubine mother.115 On neither topic (paternity and sale) does it appear that the jurists reached full accord. The protest passages can be read as an echo of these debates. They signal an apparent pattern of illicit treatment of children whose standing, by virtue of the conduct of their parents, was rendered grimly ambiguous. Brockopp acknowledges that, on the third topic—​the position of children produced by concubinage—​the legal sources offer comparatively little discussion, focusing as they do on the position of the concubine herself.116 What the passages do ask, however, is whether regulations surrounding concubinage in general, and the umm walad in particular, even mattered. Given the evidence that discussion over the standing of slave mothers was ongoing in ninth-​century scholarly circles—​ on the part of jurists, certainly, but more widely as suggested by al-​Jahiz’s comments on the singers—​one can presume that owner-​fathers seeking escape from responsibility could find legal sanction. As likely, however, is that individuals opting simply to disregard the law would find ready protection themselves, if and when it was even needed. The standing of the mothers and, hence, the children, turned not simply on the extension of these protections, but also on the decision to abide by them. And finally, perhaps at the heart of the whole affair, was profit—​the singers promised a handsome return, which was reason enough to skirt the law. Conclusion The protest passages are but a fraction of what the sources tell us about the careers of the courtesans. They suggest wider patterns of Near Eastern society during the early Abbasid period: the dynamics of commerce in enslaved persons and the shaping of an elite, urban cultural scene in which the courtesans, as enslaved women, were closely involved. And the passages are, admittedly, obscure. One can see them as a record of how things were or, more likely, as a token of some manner of debate regarding the women and, perhaps more broadly, practices associated with slavery. They point to an overlap of two elements: the achievement by the courtesans of high social and cultural standing and their ambiguous position as enslaved persons (there is much to suggest, as in Fadl’s case, that the women were manumitted at some stage in their lives). The ambiguity lay in the following. On the one hand, enslavement meant opportunity to gain rigorous training in music and language arts, as well as entry to high society, with its distribution of wealth and ready social contacts. The evidence on the small number of courtesans considered here points to considerable social mobility on their part. On the other hand, enslavement—​the standing of the enslaved person—​raised a set of obstacles around which the women had perforce to negotiate. Controversy surrounded the courtesans, and one need only cite al-​Jahiz’s essay



Abbasid Courtesans and the Question of Social Mobility

j 43

to make the point. It was controversy born of the courtesans’ success and the manner in which they sought to use it. Acknowledgment I thank Renée Baernstein, Julia Bray, Kathryn Hain, and Kristina Richardson for their comments on a previous version of this chapter. Notes 1. On slavery in early Islamic society in general, the sources cited by Brockopp, Early Maliki Law, and Clarence-​Smith, Abolition of Slavery, two very different books, are a good start. On the Abbasid period, on which scholarship on slavery has lagged, see Bray, “Men, Women and Slaves” and Crone, Slaves on Horses, which is largely concerned with military slavery during the period but raises key general questions as well. For two periods/​regions for which scholarship has moved forward much further, see the cited works in Zilfi, Women and Slavery (Ottoman Empire) and Philips, Slavery in Medieval and Early Modern Iberia (Islamic and Christian Spain). 2.  Kilpatrick, Great Book of Songs, is an extended and superb discussion of the Aghani. Kilpatrick points to the difficulty of fixing al-​Isbahani’s death date (p. 20). 3.  See Bray, “Men, Women and Slaves,” 132–​133. Modern scholarship has barely explored the question. This chapter takes up where a previous effort by this author left off; see Gordon, “Preliminary Remarks.” 4. Caswell, Slave Girls, 1, 54–​55, 271–​272, disagrees, citing the enslavement of the Abbasid singers as the feature that distinguishes them from the Japanese geisha or the hetaira of ancient Greece. 5.  A  good starting point is still Pellat, “Kayna.” Also see Athamina, “How Did Islam Contribute”; Bray, “Men, Women and Slaves”; Kilpatrick, “Women as Poets and Chattels”; and Richardson, “Singing Slave Girls.” Caswell, Slave Girls, provides a useful bibliography. His is the most recent discussion of the Abbasid-​era singers. It translates much of the verse attributed to the courtesans but does little to situate the careers of the women against the backdrop of Abbasid social and cultural history. One also has to check his references carefully. 6. Kilpatrick, Great Book of Songs, devotes little attention to the topic. Again, Bray, “Men, Women and Slavery,” and Richardson, “Singing Slave Girls,” offer a way forward. 7. I borrow the phrase from the volume edited by Marín and Deguilhem, Writing the Feminine. 8. Mattson, “A Believing Slave,” 13: “there was never one homogenous class of slaves.” 9. al-​Isbahani, Aghani, 23:94. Myrne and Sharlet comment on Inan in their contributions to this volume (see ­chapters 3 and 14, respectively). 10. One of several terms, including qayna (pl., qiyan), that the Arabic texts use for the courtesans. For a discussion of these terms, see Nielsen’s contribution to this volume (see ­chapter 4). 11. That is, of a free father and an enslaved mother. 12. That is, he withdrew his acknowledgment of paternity. 13. al-​Isbahani, Aghani, 19:314. 14. al-​Isbahani, Aghani, 21:61–​103 and 22:160–​188 (the profile of Arib’s long-​time lover, Ibrahim ibn al-​Mudabbir). I have relied on both texts in writing on Arib. See, for example, Gordon, “`Arib al-​Ma’muniyah (797–​890),”

44 i

Matthew S. Gordon

15. al-​Isbahani, Aghani, 21:72, 76. 16. It is one of at least two books on courtesans that are attributed to him. See al-​Isbahani, Aghani, 16:6 (Shariya), and Kilpatrick, Great Book of Songs, 39 (Arib). 17. al-​Isbahani, Aghani, 16:6–​11. 18. Concubinage is the concern of several other chapters in this volume. On the very rare case of free women as concubines, see Usman Hamid’s chapter in this volume (­chapter 9). 19. On the question of the predominance of women in Mediterranean slave traffic, see Philips, Medieval and Early Modern Iberia, 3–​4. 20. For a particularly compelling account of slavery’s toll on the enslaved of the later Atlantic period, see Brown, The Reaper’s Garden, esp. 13–​59. 21. Arazi, Amour divin, 16–​21; and Kilpatrick, Great Book of Songs, 362, note 14. 22.  Kilpatrick, “Poets and Chattels,” 169. I  have modified Kilpatrick’s translation slightly. Myrne cites the same verse in her contribution to this volume (see ­chapter 3). For the original Arabic, see al-​Isbahani, Ima, 114. 23.  On medieval Arabic medical texts that treat the purchase of slaves, see Ghersetti, “De l’Achat des esclaves.” 24. I cite the English translation by Beeston: al-​Jahiz, Epistle. For discussion of the text, see Cheikh-​Moussa, “La Negation d’Éros” and Gordon, “Yearning and Disquiet.” 25. On the range of sources on the singers and poets, see Cheikh-​Moussa, “Figures de l’esclave chanteuse.” 26. See Cheikh-​Moussa, “Figures de l’esclave,” 31–​32. 27. Cheikh-​Moussa, “La Negation d’Éros,” 118; and Gordon, “Yearning and Disquiet,” 260. 28. On the slave singer as temptress, see Caswell, Slave Girls, 39–​45. 29. See Pellat, “Kayna,” and further references below. 30. Caswell, Slave Girls, 17, cites the Aghani incorrectly. The correct text, 5:170, refers to training (by al-​Mawsili) and three groups of women (designated, it appears, by skin color) but says nothing of Chinese, Indian, or Sindhi women. Also see al-​Isbahani, Aghani, 5:164. 31. al-​Isbahani, Aghani, 14:210 (Arib and Shariya), 21:84, 91–​92 (Arib). 32.  See Gordon, “Place of Competition,” 68–​73; Imhof, “Singing Contest”; and Kilpatrick, “Profiles of Poets,” 112–​113. 33. See Bray, “Men, Women and Slaves,” 136–​138; and Van Gelder, “Slave-​Girl Lost.” 34. See, on ancient Greek sources, Davidson, “Making a Spectacle of Her(self ).’ 35. On “history as it was lived” in the adab literature, see El-​Cheikh, “Women’s History,” 129–​133. 36. al-​Jahiz, Epistle, para. 50. 37. These comments sum up the longer discussion in Gordon, “Preliminary Remarks,” 71–​77. 38. The city functioned as the imperial capital for roughly 60 years (838–​892). See, on the dating, Northedge, Historical Topography, 99, 241. 39.  On the history of the Samarran Turkish military, see Gordon, Thousand Swords, and Kennedy, Age of the Caliphates, 156–​175. 40. al-​Yaqubi, Buldan, 259–​260. Of real interest is the decision by the caliphate to register the young women on the military rolls (diwan). See Ayalon, “Military Reforms,” 29; and Gordon, Thousand Swords, 58, 62, 63. 41.  al-​Yaqubi, Buldan, 260. See Gordon, “Preliminary Remarks,” 73; Rāġib, “Marchés,” 723, note 8; and Savage, “Berbers and Blacks,” 354 (who refers to the market as “vast,” which is not stated explicitly in the original text).



Abbasid Courtesans and the Question of Social Mobility

j 45

42. See Ragib, “Marchés,” 723–​724. On Inan al-​Natifi’s public sale, see ­chapter 3 in this volume. 43. al-​Masudi, Muruj al-​dhahab, 7:281–​286. 44. The naming of slaves in Abbasid society needs further consideration. One question is why slaves and freed persons retained the names even on achieving position of rank. On slave names in the Ottoman period, see Zilfi, Women and Slavery, 156–​158; and, in medieval Iberia, see Philips, Medieval and Early Modern Iberia, 80–​81; and Ruggles, “Mothers,” 72–​73. 45.  See Athamina, “How Did Islam Contribute,” 391, and Cheikh-​Moussa, “Figures de l’esclave,” 33, note 2. 46. al-​Isbahani, Aghani, 22:202. 47. Ragib, Actes de vente, 6–​8, and Vanthieghem, “Quelques contrats.” 48. Gordon, “Preliminary Remarks,” 73. 49. In al-​Isbahani’s version of the story, Aghani, 22:204, it is Wasif who gains possession of the women. 50. For a translation of the verse, see Caswell, Slave Girls, 178. 51. Ibn al-​Sa`i, Nisa al-​khulafa, 96–​98. The passage concludes with pious formulas praising Mahbuba’s constancy of loyalty and affection. I thank Julia Bray for reminding me of this second version of the story. 52. On the size of the Samarran Turkish forces, see Kennedy, Armies of the Caliphs, 126–​128, 205–​208. 53. McCormick, Origins of the European Economy, 759, 768, 776. 54. See Athamina, “How Did Islam Contribute,” 391; and Lewis, Race and Slavery, 9–​12. 55. Athamina, “How Did Islam Contribute,” 391, for the related term sufr, has “fair complexion.” On this same passage, see ­chapter 5 by Dwight Reynolds in this volume. 56. Lewis, Race and Slavery, 56. 57. See the reference in the Aghani, 5:185. The term muthammana, used here in the plural, may refer to “expensive” women rather than those of a particular ethnic origin. 58.  See Athamina, “How Did Islam Contribute,” 391; Lewis, Race and Slavery, 10; and Richardson, “Singing Slave Girls,” 110. The latter two authors downplay this source of Abbasid-​ era slaves. 59. On the second generation, see Gordon, Thousand Swords, 69–​70, 111–​112. 60. See, for a different view, Bray, “Men, Women and Slaves,” 137–​138. Kilpatrick (Great Book of Songs, 49) refers only to the singers’ hometowns. 61. See Gordon, “Third/​Ninth Century Abbasid Courtesan,” 86–​87. 62. Ruggles, “Mothers,” 75–​76. Also see Winer, “Conscripting the Breast.” 63. al-​Isbahani, Aghani, 19:314–​315. 64. A contrasting view on this issue (Toledano, Slavery and Abolition, 36), on the late Ottoman slave trade, indicates that “the Palace” sought out non-​Turkish speakers—​adolescent girls in this case—​to avoid any sort of “corruption” of elite Ottoman education. 65. al-​Jahiz, Epistle, para. 33. Also see Cheikh-​Moussa (“Figures de l’esclave,” 41–​42), who cites the tenth-​century musician and writer Ibn Tahhan al-​Musiqi (d. after 1057). 66. Gordon, “`Arib al-​Ma’muniyah,” 87. 67. al-​Isbahani, Aghani, 23:92. 68. al-​Isbahani, Aghani, 19:314. 69.  al-​Isbahani, Aghani, 22:202. The passage adds that she was gifted to al-​Mutawakkil by Abdallah ibn Tahir (d. 844) and, at that point, was still a virgin (Ar. bikr).

46 i

Matthew S. Gordon

70. Ragib, Actes de vente, 3–​43. On the dynamics of marketing female slaves, including fraud and other patterns of deception used in preparing the women for sale, see Ragib, “Marchés,” 727–​ 743. A fine survey, the article creates nevertheless an impression of uniform practice across the expanse of medieval Islamic society. 71. On urbanism and urban growth during the Abbasid period, see Berkey, Formation of Islam, 119–​123; Bennison, Great Caliphs, 69–​93; and Wheatley, Places Where Men Pray , 87–​102. 72.  See, for example, the detailed account of a pimp and his clients:  al-​Tanukhi, Nishwar, 2:172–​183. See, also, Caswell, Slave Girls, 26–​32; Cheikh-​Moussa, “Figures de l’esclave,” 35–​36; and Richardson, “Singing Slave Girls,” 111. On eroticism and sexuality in Abbasid letters, see Rowson, “Arabic: Middle Ages to Nineteenth Century,” 45–​52. 73. Bray, “Men, Women and Slaves,” 137. For one study of the Abbasid majlis in a broader context, see Ali, Arabic Literary Salons. 74. Gordon, “Arib al-​Ma’muniyah,” 87. 75. al-​Tanukhi, Nishwar, 1:271–​272. 76. al-​Isbahani, Aghani, 19:314. 77. Zilfi, Women and Slavery, 211. 78. al-​Isbahani, Aghani, 16:6–​7 79. al-​Isbahani, Aghani, 16:8–​11. 80. al-​Isbahani, Ima, 113. 81. The many essays in Abbasid Belles Lettres, edited by Julia Ashtiany et al., provide an excellent introduction to Abbasid letters and culture. 82. See El-​Cheikh, “Women’s History,” 134, and Ahmed, Women and Gender, 84. 83. See, for example, the reference, Aghani, 3:248, to the training provided to singing girls by Yazid Hawra. Caswell, Slave Girls, 17, sees it as an indication of a “business partnership,” which is not at all clear. 84. See Bencheikh, “Les Musiciens et la poésie”; Gordon, “Place of Competition,” 61–​62, 72–​ 73; Imhof, “Singing Contest”; and Kilpatrick, “Profiles of Poets,” 122–​124. 85.  On the sums devoted to the purchase of singers, see Cheikh-​Moussa, “Figures de l’esclave,” 45–​54. 86. al-​Isbahani, Aghani, 5:178. 87. al-​Isbahani, Aghani, 16:6–​7. 88. al-​Jahiz, Epistle, para. 38. 89. al-​Isbahani, Aghani, 2:226; and see Kilpatrick, Great Book of Songs, 47. 90. Gordon, “Third/​Ninth Century `Abbasid Courtesan,” 96. 91. Bray, “Men, Women and Slaves,” 146. 92. See Cheikh-​Moussa, “Figures de l’esclave,” 32–​34. 93. See Bray, “Men, Women and Slaves,” 136–​138; El-​Cheikh, “Gender and Politics,” 151; and Richardson, “Singing Slave Girls,” 112–​114. 94.  See the individual singers named in Cheikh-​Moussa (“Figures de l’esclave,” 45–​51) and Pellat, “Kayna,” for the reference to an unpublished thesis (A. Chirane, Sorbonne, 1970) that lists more than 100 names of singers (qiyan). 95. Bray closes her essay (“Men, Women and Slaves,” 146) with this comment on “the images of women that were developed in literature, romantic or historical”: “The imaginative models were one of the means through which free men could reflect on their own moral identity and one what constituted tolerable relations with power.”



Abbasid Courtesans and the Question of Social Mobility

j 47

96. Zilfi, Women and Slavery, 19. 97. El-​Cheikh, “Women’s History,” 136. 98. Ahmed, Women and Gender, 84. 99. It seems a stretch, in other words, to claim “how ordinary [emphasis added] it was to buy and sell women for men’s sexual use” (Ahmed, Women and Gender, 85). But see Ali, Marriage and Slavery, 22–​23, 151. In a subsequent chapter, on the later “medieval” period, Ahmed refers to the high cost of maintaining concubines (Women and Gender, 107–​108). In his contribution to this volume, Majied Robinson (see ­chapter 1) challenges the view that concubinage came into full swing in the Abbasid era. 100. See El-​Cheikh, “Women’s History,” 133–​136. 101. Bray “Men, Women and Slavery,” 136–​142, raises a series of questions in this regard. Also see Richardson, “Singing Slave Girls,” 112–​113. 102. On texts that do so, see ­chapter 3 this volume. 103.  Ibn Butlan, Risala. On the non-​Arab background of well-​known early concubines in Islamic history, see Elizabeth Urban’s work (­chapter 11 this volume). 104. Bennison, Great Caliphs, 114–​115; and Bray, “Men, Women and Slavery,” 134. Also, see ­chapters 1 and 11 this volume. 105. Bray, “Men, Women and Slavery,” 134–​135. 106.  Bennison, Great Caliphs, 112–​117; Bray, “Men, Women, and Slavery,” 134; El-​Cheikh, “Gender and Politics.” On concubines in the Fatimid period, see Cortese and Calderini, Women and the Fatimids, 45–​56, 75–​79. 107. al-​Isbahani, Aghani, 21:78. She is said, at that point, to have been a resident in al-​Mamun’s palace. Following a series of trysts with her lover, a young officer, she bore him a daughter, who goes unnamed. On hearing of the birth, al-​Mamun married Arib to the officer, thus elevating her, presumably, to the standing of umm walad. 108. al-​Isbahani, Aghani, 7:312–​313. 109. Cited by Musallam, Sex and Society, 94. 110. Richardson’s characterization of this facet of concubinage (“Singing Slave Girls,” 106–​107, 114–​115) and, particularly, the new standing that “favored the mother and child over the master” (107). Also see Siddiqui, The Good Muslim, 54–​56. 111. Brockopp, Early Maliki Law; and Mattson, Believing Slave. 112. Brockopp, Early Maliki Law, 192–​203; and Mattson, Believing Slave, esp. 149–​158. 113. Brockopp, Early Maliki Law, 201, 203. 114. Brockopp, Early Maliki Law, 196, and see Schacht, “Umm al-​Walad.” 115. See Mattson, Believing Slave, 180. 116. Brockopp, Early Maliki Law, 196, 200–​201.

Bibliography Primary Sources Ibn Butlan, al-​Mukhtar ibn al-​Hasan. Risala fi shira al-​raqiq wa-​taqlib al-​abid. Edited by Abd al-​Salam Harun in Nawadir al-​makhtutat, no. 4. Cairo: Matba`at Lajnat al-​taʼlif wa’l-​tarjama wa’l-​nashr, 1954 [1373]:  333–​389 (no.  15). Translations:  (German) Hans Müller, Die Kunst des Sklavenkaufs nach arabischen, persischen und türkischen Ratgebern vom 10. bis zum 18.

48 i

Matthew S. Gordon

Jahrhundert. Frieburg: Klaus Schwarz Verlag, 1980, 45–​80; and (Italian) Antonella Ghersetti, Trattato generale sull’acquisto e l’esame degli schiavi. Catanzaro: Abramo, 2001. Partial English translations: Bernard Lewis, Islam from the Prophet Muhammad to the Capture of Constantinople. Oxford:  Oxford University Press, 1987, 243–​251; and Adam Mez, The Renaissance of Islam, translated by S. K. Bukhsh and D. S. Margoliouth. London: Luzac, 1937, 160–​162. Ibn al-​Sa`i, Taj al-​Din Ali ibn Anjab. Nisa al-​Khulafa. Edited by Mustafa Jawad. Cairo:  Dar al-​Ma’arif, 1968. al-​Isbahani, Ali ibn al-​Husayn. Al-​Ima al-​Shawa’ir. Edited by Jalil al-​Atiya. Beirut: Dar al-​Nidal, 1984 [1404]. —​—​—​. Kitab al-​Aghani. Edited by Abd al-​Amir Ali Mahanna and Samir Jabir. 27 vols. Beirut: Dar al-​Kutub al-​Ilmiya, 1992 [1412]. al-​Jahiz, Amr ibn Bahr. The Epistle on Singing-​Girls of Jahiz. Edited and translated by A. F. L. Beeston. Warminster, UK: Aris & Phillips, 1980. Translation: (French) Pellat, Charles. “Les Esclaves-​chanteuses de Gahiz.” Arabica 10, no. 2 (1963): 121–​147. —​—​—​. Rasa’il al-​Jahiz. 4 vols. Edited by Abd al-​Salam Muhammad Harun. Cairo: Maktabat al-​ Khanji, 1965 [1384], 2: 143–​182. al-​Masudi, Ali ibn al-​Husayn ibn Ali. Muruj al-​Dhahab wa’l-​Maadin al-​Jawhar. Edited and translated by C. Barbier de Meynard and Pavet de Courteille as Les prairies d’or. 9 vols. Paris: Imprimerie Impériale and Imprimerie Nationale, 1861–​1877. al-​Nuwayri, Ahmad ibn Abd al-​Wahhab. Nihayat al-​arab fi funun al-​adab. Vol. 5. Cairo:  al-​ Muassasa al-​Misriya al-​Ammah lil-​Talif wa’l-​Tarjamah wa’l-​Tiba`a wa’l-​Nashr, 1964, 38–​116. al-​Tabari, Muhammad ibn Jarir. Tarikh al-​rusul wa’l-​muluk. Edited by M. J.  de Goeje et  al. as Annales. 15 vols. Leiden: Brill, 1879–​1901. al-​Tanukhi, Abu Ali al-​Muhassin ibn Ali. Nishwar al-​muhadara wa-​akhbar al-​mudhakara. Edited by Abbud al-​Shalji. 4 vols. Beirut: Dar Sadir, 1978 [1398]. al-​Yaqubi, Ahmad ibn Abi Yaqub. Kitab al-​buldan. Edited by M. J. de Goeje. Bibliotheca geographorum Arabicorum. Vol. 7. Leiden: Brill, 1892.

Secondary Sources Ali, Kecia. Marriage and Slavery in Early Islam. Cambridge, MA: Harvard University Press, 2010. Ali, Samer M. Arabic Literary Salons in the Islamic Middle Ages: Poetry, Public Performance and the Presentation of the Past. Notre Dame, IN: University of Notre Dame Press, 2010. Arazi, Albert. Amour divin et amour profane dans l’Islam medieval, a travers le Diwan de Khalid al-​Katib. Paris: Maisonneuve & Larose, 1990. Ashtiany, Julia, et al., eds. `Abbasid Belles-​Lettres. The Cambridge History of Arabic Literature. Cambridge: Cambridge University Press, 1990. Athamina, Khalil. “How Did Islam Contribute to Change the Legal Status of Women: The Case of the Jawārī, or the Female Slaves.” Al-​Qantara 28, no. 2 (2007): 383–​408. Ayalon, David. “The Military Reforms of Caliph al-​Mu`tasim.” In Islam and Abode of War, 1–​39. Brookfield, VT: Variorium, 1994. Bencheikh, J. E. “Les Musiciens et la poésie: Les écoles d’Ishaq al-​Mawsili (m. 225 H.) et d’Ibrahim Ibn al-​Mahdi (m. 224 H.).” Arabica 22, no. 2 (1975): 114–​152. Bennison, Amira K. The Great Caliphs:  The Golden Age of the `Abbasid Empire. New Haven, CT: Yale University Press, 2009.



Abbasid Courtesans and the Question of Social Mobility

j 49

Berkey, Jonathan P. The Formation of Islam:  Religion and Society in the Near East, 600–​1800. Cambridge: Cambridge University Press, 2003. Bray, Julia. “Men, Women and Slaves in `Abbasid Society.” In Gender in the Early Medieval World, East and West, 300–​900, edited by Leslie Brubaker and Julia M. H. Smith, 121–​146. Cambridge: Cambridge University Press, 2004. Brockopp, Jonathan E. Early Maliki Law: Ibn `Abd al-​Hakam and His Major Compendium of Jurisprudence. Leiden: Brill, 2000. Brown, Vincent. The Reaper’s Garden:  Death and Power in the World of Atlantic Slavery. Cambridge, MA: Harvard University Press, 2008. Brunschvig, R. “`Abd.” In Encyclopaedia of Islam, 2nd ed., edited by P. Bearman et al. Leiden: Brill, 1954–​2007. Caswell, Fuad Matthew. The Slave Girls of Baghdad:  The Qiyan in the Early Abbasid Era. London: I.B. Tauris, 2011. Cheikh-​Moussa, Abdallah. “Figures de l’esclave chanteuse à l’époque `abbaside.” In Figures de l’esclave au Moyen-​Age et dans le monde moderne, n.e., 31–​76. Paris: L’Harmattan, 1996. —​—​—​. “La Negation d’Éros ou le `isq d’après deux épîtres d’al-​Gahiz.” Studia Islamica 72 (1990): 71–​119. Clarence-​Smith, William Gervase. Islam and the Abolition of Slavery. Oxford: Oxford University Press, 2006. Cortese, Delia, and Simonetta Calderini. Women and the Fatimids in the World of Islam. Edinburgh: Edinburgh University Press, 2006. Crone, Patricia. Slaves on Horses:  The Evolution of the Islamic Polity. Cambridge:  Cambridge University Press, 1980. Davidson, James. “Making a Spectacle of Her(self ):  The Greek Courtesan and the Art of the Present.” In The Courtesan’s Arts, Cross-​Cultural Perspectives, edited by Martha Feldman and Bonnie Gordon, 29–​51. Oxford: Oxford University Press, 2006. El-​Cheikh, Nadia Maria. “Gender and Politics in the Harem of al-​Muqtadir.” In Gender in the Early Medieval World, East and West, 300–​900, edited by Leslie Brubaker and Julia M.  H. Smith, 147–​161. Cambridge: Cambridge University Press, 2004. —​—​—​. “Women’s History: A Study of al-​Tanukhi.” In Writing the Feminine: Women in Arab Sources, edited by Manuela Marín and Randi Deguilhem, 129–​148. London: I.B. Tauris, 2002. Ghersetti, Antonella. “De l’Achat des esclaves:  Entre examen médical et physiognomonie. Le Chapitre 46 du Kitab ad-​Dala’il d’Ibn Bahlul (Xe S.).” In Essays in Honour of Alexander Fodor on His Sixtieth Birthday, edited by Kinga Dévényi and Tamás Iványi, 83–​94. Budapest: Eötvös Loránd University Chair for Arabic Studies, 2001. Gordon, Matthew S. “`Arib al-​Ma’muniya:  A  Third/​Ninth Century `Abbasid Courtesan.” In Views from the Edge:  Essays in Honor of Richard W.  Bulliet, edited by Neguin Yavari et  al., 86–​100. New York: Columbia University Press, 2004. —​—​—. “`Arib al- Ma’muniyah (797–890).” In Dictionary of Literary Biography, Vol. 311: Arabic Literary Culture, 500–925, edited by Michael Cooperson and Shawkat Toorawa, 85–90. Farmington Hills, MI: Thomson Gale, 2005. —​—​—​. The Breaking of a Thousand Swords: A History of the Turkish Military of Samarra (AH 200–​275/​815–​889 CE). Albany: State University of New York Press, 2001. —​ —​ —​ . “The Place of Competition:  The Careers of `Arib al-​ Ma’muniya and `Ulayya bint al-​Mahdi, Sisters in Song.” In `Abbasid Studies:  Occasional Papers of the School of

50 i

Matthew S. Gordon

`Abbasid Studies, 6–​10 July 2002, edited by James E. Montgomery, 61–​81. Leuven: Peeters Publishers, 2004. —​—​—. “Preliminary Remarks on Slaves and Slave Labor in the Third/Ninth Century `Abbasid Empire,” in Slaves and Households in the Near East, edited by Laura Culbertson, 71–84. Chicago: University of Chicago Press (The Oriental Institute), 2011. —​—​—​. “Yearning and Disquiet:  Al-​Jāḥīẓ and the Risālat al-​qiyān.” In Al-​Jāḥiẓ:  A  Muslim Humanist for Our Time, edited by Arnim Heinemann et  al., 253–​268. Würzburg:  Ergon Verlag, 2009. Imhof, Agnes. “Traditio vel Aemulatio? The Singing Contest of Samarra, Expression of a Medieval Culture of Competition.” Der Islam 90, no. 1 (2013): 1–​20. Kennedy, Hugh. The Armies of the Caliphs:  Military and Society in the Early Islamic State. London: Routledge, 2001. —​—​—​. The Prophet and the Age of the Caliphates:  The Islamic Near East from the Sixth to the Eleventh Century. 2nd ed. London: Pearson Longman, 2004. —​—​—​. When Baghdad Ruled the Muslim World: The Rise and Fall of Islam’s Greatest Dynasty. Cambridge, MA: Da Capo Press, 2006. Kilpatrick, Hilary. “Abu l’Farag’s Profiles of Poets: A 4th/​10th Century Essay at the History and Sociology of Arabic Literature.” Arabica 44 (1997): 94–​128. —​—​—​. Making the Great Book of Songs: Compilation and the Author’s Craft in Abû l-​Faraj al-​ Isbahânî’s Kitâb al-​Aghânî. London: RoutledgeCurzon, 2003. —​—​—​. “Women as Poets and Chattels: Abu l-​Farag al-​Isbahani’s ‘al-​Ima’ al-​Shawa’ir.’” Quaderni di Studi Arabi 9 (1991): 161–​176. Lewis, Bernard. Islam from the Prophet Muhammad to the Capture of Constantinople. Vol. II, Religion and Society. Oxford: Oxford University Press, 1987. —​—​—​. Race and Slavery in the Middle East: An Historical Enquiry. Oxford: Oxford University Press, 1990. Marín, Manuela, and Randi Deguilhem, eds. Writing the Feminine:  Women in Arab Sources. London: I.B. Tauris, 2002. Mattson, Ingrid. “A Believing Slave Is Better Than an Unbeliever:  Status and Community in Early Islamic Society and Law.” PhD diss., University of Chicago, 1999. McCormick, Michael. Origins of the European Economy: Communications and Commerce, AD 300–​900. Cambridge: Cambridge University Press, 2001. Mez, Adam. The Renaissance of Islam. Translated by Salahuddin Khuda Bukhsh and D. S. Margoliouth. London: Luzac, 1937. Müller, Hans. “Sklaven.” In Handbuch der Orientalistik, edited by B. Spuler, 53–​83. Div. 1, vol. 6, section 6, part 1. [Wirtschaftsgeschichte des vorderen Orients in Islamischer zeit]. Leiden: Brill, 1977. Musallam, B. F. Sex and Society in Islam. Cambridge: Cambridge University Press, 1983. Northedge, Alastair. The Historical Topography of Samarra. London:  British School of Archaeology in Iraq, 2005. Pellat, Charles. “Kayna.” In Encyclopaedia of Islam, 2nd ed., edited by P. Bearman et  al. Leiden: Brill, 1954–​2007. Philips, William D., Jr. Slavery in Medieval and Early Modern Iberia. Philadelphia: University of Pennsylvania Press, 2014. Ragib, Yusuf. Actes de vente d’esclaves et d’animaux d’Égypte medieval. Cairo: Institut Français d’Archéologie Orientale, 2002.



Abbasid Courtesans and the Question of Social Mobility

j 51

—​—​—​. “Les Marchés aux esclaves en terre d’Islam.” In Mercati e mercanti nell’alto medioevo: L’Area Euroasiatica e l’area Mediterranea, n.e., 721–​763. Spoleto:  Centro Italiano di Studi Sull’alto Medioevo, 1993. Richardson, Kristina. “Singing Slave Girls (Qiyan) of the `Abbasid Court in the Ninth and Tenth Centuries.” In Children in Slavery through the Ages, edited by Gywn Campbell, Suzanne Miers, and Joseph C. Miller, 105–​118. Athens: Ohio University Press, 2009. Rowson, Everett. “Arabic:  Middle Ages to Nineteenth Century.” In Encyclopedia of Erotic Literature, edited by Gaëtan Brulotte and John Philips, 1: 43–​61. New York: Routledge, 2006. Ruggles, D. Fairchild. “Mothers of a Hybrid Dynasty: Race, Genealogy, and Acculturation in al-​ Andalus.” Journal of Medieval and Early Modern Studies 34, no. 1 (2004): 65–​94. Savage, Elizabeth. “Berbers and Blacks:  Ibadi Slave Traffic in Eighth-​Century North Africa.” Journal of African History 33, no. 3 (1992): 351–​368. Schacht, Joseph. “Umm al-​Walad.” In Encyclopaedia of Islam, 2nd ed., edited by P. Bearman et al. Leiden: Brill, 1954–​2007. Siddiqui, Mona. The Good Muslim: Reflections on Classical Islamic Law and Theology. Cambridge: Cambridge University Press, 2012. Toledano, Ehud R. Slavery and Abolition in the Ottoman Middle East. Seattle:  University of Washington Press, 1998. Van Gelder, Geert Jan. “Slave-​Girl Lost and Regained: Transformations of a Story.” Marvels and Tales: Journal of Fairy-​Tale Studies 18, no. 2 (2004): 201–​217. Vanthieghem, Naim. “Quelques contrats de vente d’esclaves de la collection Aziz Atiyya. Journal of Juristic Papyrology 44 (2014): 163–​187. Wheatley, Paul. The Places Where Men Pray Together: Cities in Islamic Lands, Seventh through the Tenth Centuries. Chicago: University of Chicago Press, 2001. Winer, Rebecca Lynn. “Conscripting the Breast: Lactation, Slavery and Salvation in the Realms of Aragon and the Kingdom of Majorca, c.  1250–​1300.” Journal of Medieval History 34 (2008): 164–​184. Zilfi, Madeline C. Women and Slavery in the Late Ottoman Empire:  The Design of Difference. Cambridge: Cambridge University Press, 2010.

3 A  Jariya’s Prospects in Abbasid Baghdad Pernilla Myrne

i  Inan al-​N atifi, a slave woman of early Abbasid Baghdad, flourished in the time of the caliph Harun al-​R ashid (r. 786–​809). She was a jariya1—​a slave courtesan—​and a poet. Her poetry made her famous—​so famous, in fact, that she became one of the most frequently quoted Abbasid women in classical Arabic literature. Free women poets did not rise to fame in the Abbasid Empire, as they had done during the Umayyad era; instead, young, talented slave girls were selected and given an appropriate education to enable them to pursue this art. Thus, they belonged to the same branch of feminine slavery as the slave singers, the qiyan, who were also slave courtesans, and, occasionally, concubines, typically for the elite and growing urban classes. The institution of the slave courtesan opened up possibilities of social mobility for women. If they performed well and their fame spread, they gained access to high society and its wealth. Yet, even if they were much better off than the majority of slaves, they were subject to the same conditions: they could be bought and sold, and were exposed to the whims of their masters. Inan did not belong to the elite courtesans, who lived in the palaces of the imperial family and chief officials, although she no doubt aspired to that standing. Her life story, although fragmented, discontinuous, and episodic, instead conveys a picture of the vicissitudes and possibilities of an ambitious jariya in the flourishing and growing Abbasid imperial center. In reconstructing events in Inan’s life story, I hope to shed light on the general living conditions of jawari in early Abbasid society, and, in particular, their agency, self-​perception, and access to social mobility. In brief, what was it like to be a jariya in early Abbasid Baghdad? What opportunities did a jariya have to improve her living conditions, and how did she perceive her possibilities? 52

A Jariya’s Prospects in Abbasid Baghdad

j 53

Slavery existed long before the Abbasid era, but it was during this period that the jawari rose to fame both as historical actors and fictional characters.2 In contrast, hara’ir (free Arab women) appear to have been less influential than in earlier Islamic societies; in any case, they are seldom mentioned in our sources.3 Modern scholars have inquired after the reasons for this phenomenon and its possible impact on society. One view sees it as having contributed to the development of the more restrictive Islamic regulations for women and gender segregation that emerged during this period. Biases and polemic in the medieval sources, as well as misconceptions in modern scholarship, however, have blurred our understanding of Abbasid slave women such as Inan. Medieval authors and poets often embraced age-​old prejudices against the female slave singers and poets, sometimes as part of a criticism of social ills. Several modern scholars have accepted these prejudices as facts, and even held the institution of slave courtesans and concubines liable for a presumed decline in status for free women during the Abbasid era. Yet, slavery was an age-​old institution, perceived as natural, and freeborn women as well as men profited from it.4 This chapter considers several related problems. It looks at the image of the jawari in modern scholarship, arguing that although medieval Arabic texts convey useful information about slave women, they should be read with caution. It also challenges the view that the trade in women was an entirely male enterprise, by looking at elite women’s relations to slave women in particular. Finally, it takes up the question of how the dichotomy of free/​slave intersected with gender in social hierarchies. Inan al-​N atifi The career of the slave poet Inan al-​Natifi illustrates both the vulnerabilities and possibilities inherent to the lives of jawari. Her main biographer is Abu al-​Faraj al-​Isbahani, who accords her the first and longest biographical entry in al-​Ima al-​Shawair (The Female Slave Poets), and an entry in the Kitab al-​Aghani (The Book of Songs).5 Al-​Isbahani praises her, asserting she was the first famous slave poet in the Abbasid period.6 Other scholars recognize her as a fine poet as well, including Ibn al-​Nadim, who cites her poetic oeuvre in his tenth-​century bibliographical essay, the Fihrist (Index). (She is also one of the few women to whom The Encyclopaedia of Islam devotes an article).7 Most of what al-​Isbahani and other scholars pass on about Inan consists of glimpses from her encounters with male poets who came to visit her in her master’s house and their poetic exchanges. This is typical of the narratives drawn up of women’s lives—​that is, that they are made up of men’s memories of the brief encounters with the women. Everything we know about Inan is related by her male friends and visitors, and we never read of another woman in connection to her. Perhaps she was her owner’s only courtesan, which made the court, with its multitude of cultured and creative women, even more attractive to her. In the elite households, where so many women lived together, there must have been female networks and friendship, but most interactions between women were never recorded

54 i

Pernilla Myrne

and thus remain unknown. Anecdotes concerning the princess Ulayya bint al-​Mahdi, the sister of Harun al-​Rashid, for example, reveal something about women’s oral tradition in the context of palace culture.8 Ulayya was a skilled poet and singer—​her songs sung and transmitted by female slave singers, her own and those of Zubayda, her sister-​in-​law. Some of their jawari taught their songs to male singers, who in turn are quoted in the sources. The male narrators name a few of the female transmitters and teachers, although, for the most part, these women are simply referred to as “jawari,” sometimes in a genitive construction (“someone’s jawari”) or with a possessive pronoun (“our jawari”). These anonymous jawari populate the background of many accounts about Abbasid high society. According to al-​Isbahani, Inan was born a slave in al-​Yamama, located on the Arabian Peninsula east of Najd, where she was raised and educated. She was a muwallada, which indicates she was raised and educated by an Arab family, and thus learned Arabic and other appreciated skills, which raised her price.9 She was purchased in Yamama by a man, one “al-​Natifi,” who brought her to Baghdad in the reign of Harun al-​Rashid. There, she became recognized for her abilities and, over time, in particular, for her fine poetry and quick repartee. She might not have possessed the sophistication of the high-​society jawari who had received the best education and were often raised in elite families; her owner, al-​Natifi, seems to have been a man of modest position. His name denotes “a vendor of natif,” which was a kind of nougat.10 She became friendly with several male poets in Baghdad, who were, unlike her, free men. They used to visit her and sometimes she visited them, encouraged by her owner, who hoped that they would spread the word about her and thus attract wealthy visitors. She is particularly known for her relations with the most famous poet of them all, Abu Nuwas, but other well-​known poets, such as al-​Abbas ibn al-​Ahnaf and Marwan ibn Abi Hafsa, also visited her and admired her ability to improvise poetry. Inan’s owner, al-​Natifi, seems to have abused her physically; there are two occasions when he is said to have beaten her.11 One of these incidents is cited frequently and appears with some variants in numerous sources.12 The earliest source to mention it is al-​Waraqa by Ibn Jarrah (d. 296/​908). His account is to the point:13 Al-​Mubarrad said, “Abu Nuwas came to Inan one day after her master had beaten her and found she was crying.” He also said, “Abu Zayd Umar ibn Shabba commented that Ahmad ibn Muawiya had told him that Marwan ibn Abi Hafsa had said that he entered al-​Natifi’s house after he had beaten Inan, and reported: “Inan wept so that her tears flowed /​as pearls slip off the string”14 She answered, with tears in her voice: “May it wither on the whip /​the right hand of him who struck her unjustly.” Then Marwan said, “By God, she is the most poetic among jinn and humans.” The sources differ as to the name of the male poet who visited her. He is mostly referred to as Marwan ibn Abi Hafsa, as in this version, but sometimes as Abu Nuwas, and a few

A Jariya’s Prospects in Abbasid Baghdad

j 55

times as another poet.15 He finds Inan crying after having been flogged by her owner, which she objects to in her verse. The point of the anecdote is to highlight Inan’s poetic talent. The visitor expresses his admiration for her ability to improvise despite her shock and, indeed, draw on her misfortune in the verse. It is this ability that made her “the most poetic” among jinn (celestial spirits) and humans. Islamic law allowed for flogging, but only in cases in which a crime was committed. As a means of punishment inflicted by the master on a subordinate member of the household, however, the whip might have constituted the dividing line between a free wife and a slave.16 Jurists who were critical of severe beating of wives cited the Prophet’s criticism of men who beat their wives as if they were slaves and then had sexual intercourse with them.17 The implication is that slaves could expect to be beaten. A gifted and celebrated slave courtesan like Inan, however, might have expected to be treated better than a common slave, which possibly is the reason for Inan’s versified claim to have been whipped unjustly. Perhaps the elite slaves were better treated than “common” slaves, but probably they were more exposed to corporal punishment than free women. The Kitab al-​Aghani cites cases of jawari, in their capacity as concubines and courtesans, flogged by their masters. The slave singer Arib, for example, is said to have had 100 lashes inflicted on her after having tried to escape from her master, which is a crime according to Islamic law. Characteristically, she cries while being flogged, “I am a free woman,” claiming that her freeborn status should protect her from her master’s mistreatment.18 The anecdote about Inan raises other questions, especially about the understanding of the incident by medieval authors. Although a modern reader might find it curious that Inan could defy her owner so openly—​in some variants he almost brags about her outspokenness—​al-​Isbahani seeks to understand why al-​Natifi beat Inan. Some 50 years after Ibn Jarrah, he provided a longer version of the episode, asserting that al-​Natifi’s violent treatment of Inan was prompted by her refusal to socialize with a guest to his house. According to this version, Marwan ibn Abi Hafsa ran into al-​Natifi, who invited him to meet Inan. When they arrived at the house, al-​Natifi went in first and told Inan that he had a guest for her. Inan answered that she was sick and did not want to see him. Al-​Natifi responded by striking her with his whip before inviting Marwan to enter. He did so only to find Inan in tears, which prompted him to improvise his verse. Following her quick response, Marwan turned to al-​Natifi and said that he would release his slaves if there were someone more poetic than Inan.19 Al-​Isbahani seems to have had two objectives: first, to prove that Inan was a talented poet and, second, was blameless and thus beaten unjustly. She objected to entertaining her master’s guest only because she was sick. The question remains, however: Why did al-​Natifi not object to Inan’s open defiance? In one variant, he even expresses his admiration.20 Part of the answer is probably that Inan’s quick wit and her biting verse added to her fame (and raised her price), which was worth more to her owner than stating his authority in this situation.

56 i

Pernilla Myrne

Inan uses a well-​known motif in Near Eastern love poetry—​that of the lover who accuses the beloved of unjust treatment. Inan introduces a small twist by complaining not of the beloved’s mistreatment, but of that of her master. The motif of the mistreated lover is also blended with another motif: that of the slave, who has the moral right to demand protection and consideration from her owner precisely because, as a subordinate, she is in his custody. The figure of the slave woman poses as the perfect subordinate. As a woman she submits to her husband, as a slave she submits to her master, and as a subject she submits to the sovereign. This threefold obedience is alluded to in an anecdote about the caliph al-​ Mutawakkil’s favorite concubine, Qabiha.21 She appeared with the caliph’s name written on her cheek, which delighted him. The caliph asked Ali ibn Jahm, who was in attendance, to write a poem about it, but Mahbuba, an accomplished jariya, seized the opportunity in coming up with a beautiful poem before him. The poem praised Qabiha with the words, “Oh what a slave she is, who to the sovereignty of his right hand/​is obedient in that which is hidden and that which is revealed,” which much pleased the caliph.22 The medieval texts feature other jawari who represent themselves as obedient slaves and, yet, mistreated subordinates in poems, improvised verses, or eloquent declarations.23 Abu al-​Faraj has, for example, the poet Nasim accuse her master of mistreating her despite her obedience.24 She declares that she adheres to her obligations as a slave, “If it was not for the submission of slavery, I would not have endured.”25 Another example is Murad, who wrote to her master Ali ibn Hisham: Since I am a slave twice over, to love and to my owner /​I must endure, though endurance goes against the grain /​Close my eyes on the speck of grit in them, submit /​To being humbled and coerced as befits a slave.26 In the anecdotes framing this kind of poem, the woman’s submission to cruel behavior often serves as a moral incentive for the accused to improve his behavior. Hence, in one version of the beating of Inan, al-​Natifi pledges never to beat her again.27 He acknowledges, in other words, Inan’s moral right to demand better treatment from her owner. The Sale and Purchase of a Jariya Inan met poets who were frequent visitors of the court and the elite households, and she must have been attracted by what they described to her. Textual evidence suggests the extent of her ambitions; she wished to be bought by the caliph. Inan’s attempt to be purchased by the caliph indicates she wished to move up from her position in a relatively modest house to the imperial court—​that is, the center of Abbasid culture. There are, all together, four petition poems attributed to her, written to four different men and quoted in different sources. Al-​Isbahani relates that she wrote a letter to Jafar ibn Yahya al-​Barmaki (d. 803), imploring him to ask his father, the vizier Yahya (d. 805),

A Jariya’s Prospects in Abbasid Baghdad

j 57

to put in a good word for her with the caliph and make him buy her.28 To convince Jafar to help her, she sent the letter together with a praise poem. Moreover, she wrote a praise poem to Yahya himself, in which she asked him to do the same favor to her that he did to two other jawari whom Yahya apparently presented to the caliph.29 She also wrote a praise poem to Yahya’s son and Jafar’s brother, al-​Fadl, which may have been part of her efforts to ingratiate herself with the influential Barmakid family to get the caliph’s favor.30 Finally, she wrote a short poem to the succeeding vizier, al-​Fadl ibn al-​Rabi, with an explicit appeal to ask the caliph to buy her.31 There are several versions of the account of Inan’s attempt to be transferred to the court and its outcome. Some authorities confirm that although the caliph, Harun al-​Rashid, wished to buy her, he was dissuaded by Inan’s reputation as being sexually involved with her male visitors, with whom she used to exchange indecent poetry. One source of this reputation, it is told, was a derogatory poem by Abu Nuwas, which he wrote after Inan had fallen out with him, stating: “No one would buy her except for the son of a whore or a pimp.”32 Other authorities report that al-​Natifi demanded too high a price, over which the angry caliph refused to bargain.33 Al-​Isbahani provides an anecdote in which al-​Natifi asks for 100,000 dinars for Inan. The caliph counters with an offer of 700,000 dirhams—​the silver coin of lesser value—​which al-​Natifi rejects.34 It may simply be that he did not wish to sell her. Then, following al-​Natifi’s death, the caliph consulted with jurists, who permitted him to confiscate Inan to cover her owner’s debts. The caliph ordered his eunuch, al-​Masrur, to take her to the slave market by the Karkh Gate in Baghdad and pretend to sell her but put in the highest bid himself. The caliph obviously confiscated Inan with the intention to disgrace her as an act of revenge on her dead owner for not selling her to him for a decent price. At the slave market, Masrur placed the unhappy Inan on a bench, so to display her to potential buyers, and commenced the auction. Inan is said, at this point, to have cursed the caliph, “May God humiliate the one who humiliated me.” The text, unsurprisingly, has Masrur fail to outbid an unnamed rival who purchased her and took her to Khurasan. Al-​Isbahani’s informant relates that she subsequently gave birth to two sons, neither of whom survived infancy, and died in Khurasan sometime after her new owner.35 The anecdote, in assigning Inan a tragic end, is typical; one can cite examples concerning others of the elite singers. But al-​Isbahani, in a variant account, quotes another authority who claims that al-​Natifi freed her and that she moved to Egypt.36 Another of his informants claims she was bought by the general Tahir ibn al-​Husayn, who wanted her to be in charge of his perfumes. Perhaps as an act of repentance for her impious lifestyle, and because she did not want to be exposed to jealousy, she asked to be in charge of his woolen cloths instead.37 According to Ibn Jarrah, al-​Natifi sold her to Abd al-​Malik ibn Salih al-​Hashimi, a cousin of the first and second Abbasid caliphs, al-​Saffah and al-​ Mansur.38 A further informant states she was freed on the death of her owner al-​Natifi, either as part of his will or because she was his umm walad, having given birth to his

58 i

Pernilla Myrne

child.39 And there is yet another version of the auction incident, in which it is al-​Rashid himself who buys her. He fathers her two sons, who, again, die young, and takes her to Khurasan where he himself dies, followed thereafter by Inan.40 Inan was not the only female slave who pleaded with a caliph to buy her. Another example is the accomplished singer and poet Sakan, who wrote to the caliph al-​Mutasim when her owner wanted to sell her.41 Most probably, these ambitious women were attracted to the prosperity and cultural creativity at the court, and, importantly, they believed in their own ability to get there.42 Their slave status notwithstanding, they did not act as mere puppets of their masters. They demonstrated what we might call agency, and did their best to take control over the course of their lives. Inan used the same method that any ambitious poet or writer would, appealing to an influential man to act as mediator to get through to a wealthy patron. Nonetheless, there was a significant difference between her and aspiring male poets: she would be owned by her patron and sexually available to him, in accordance with the Islamic law that allowed and regulated concubinage between men and female slaves. The real destiny of Inan is obscure; she might have been freed, sold at a public auction, or something else. The auction incident, however, involves literary features that make it perhaps less credible. Inan sitting on a bench at the public auction cursing the caliph sets up a dramatic and certainly fictional scene. Nevertheless, regardless of whether the auction really took place, the anecdote demonstrates the conditions—​and possibly also the expectations—​of a jariya in ninth-​century Baghdad. Female elite slaves like Inan, who might have been her master’s concubine, probably expected to be treated almost like free women, and, for example, not displayed in public. When they changed owners, they were usually presented for potential buyers in private or given away.43 Common slaves, in contrast, were exposed for anyone who wanted to look at them at the slave market.44 The setting of the sale was thus an indicator of the slave’s status. What is more, the jawari who were presented to potential buyers in private were encouraged to display their artistic and intellectual abilities, which gave them some agency in influencing the prospective purchase. The presentation of jawari must have been an extremely crucial occasion for them, something for which they had been trained, and their chance to get a better life. In just a few minutes, they had to make an impression that either attracted or put off potential buyers.45 In a highly competitive environment, full of accomplished courtesans, they had to prove their uniqueness. This important moment in the career of a jariya became a literary topic. Classical Arabic belles-​lettres, adab, is rich in anecdotes about witty and eloquent jawari being displayed before a caliph or another man, answering questions not only successfully, but also irresistibly.46 The questions sometimes, but far from always, have sexual connotations. Of course, the purchase might also worsen their conditions, and there are several comical anecdotes about jawari who put off potential buyers instead of attracting them. These anecdotes are told as humorous stories, and the same story often occurs in different versions, which vary with regard to the names of both the women and the buyers. Still, they have some historical value, I suggest, indicating the

A Jariya’s Prospects in Abbasid Baghdad

j 59

efforts jawari made to influence the purchase of them.47 The purchase was a resort for social mobility, and efforts by jawari to influence it demonstrate their limited agency, which was conditioned by their social position. Modern Scholarship and Medieval Texts Nabia Abbott, a pioneer scholar of Arab women’s history, argues in her book on Abbasid imperial women that “the trio of polygamy, concubinage, and seclusion of women” under the early Abbasids, resulted in circumscribed lives for freeborn women and had a degenerative effect on upper class men.48 These social phenomena brought about a decline in social and moral standards that had begun already during the early Umayyad caliphate, when “the presence and attraction of a large number of [enslaved] women” began threatening the standing of free Arab women.49 Abbott insists the Abbasid slave trade was part of an extensive luxury market that catered to “the tastes and vanities of both men and women.”50 Later scholars, however, have often treated the trade in jawari as an exclusively male affair with devastating effects for women (disregarding in the process, however, the investment of women in the trade).51 For Leila Ahmed, the access to slave concubines during the Abbasid era was a contributing factor to gender discrimination in the emerging legal system, which has had a significant impact ever since.52 Men became used to the sorts of relationships with women through which they became, by definition, masters. The result was a blurring of the categories of woman and slave.53 The easy access to slave concubines did not only change men’s perception of women, it had a negative impact on women’s morals as well. The insecure life in a harem led to rivalry between wives and concubines as well as a certain amount of falsehood and manipulation, in contrast to the “earlier Arabian forthrightness.”54 Her examples here are entirely from the caliphal court, especially the account of Zubayda’s reported jealousy of her husband Harun al-​Rashid’s attachment to a concubine. Zubayda was a well-​known and influential queen, remembered in later history writing, whereas the concubine in question is anonymous, and did not seem to have been a real threat to Zubayda’s power. Ahmed has been criticized for disregarding primary sources and drawing a simplistic causality between the mores of elite society and the formation of Islamic law.55 She also ignores upper-​class women’s investment in slavery. Moreover, polygyny was probably equally to blame as concubinage for the insecurity of women’s lives in the harems, in addition to which both institutions had been endorsed long before the Abbasid era. Yet, the fact remains, as Ahmed points out, that the system of concubinage permeated not only elite Abbasid family life, but also that of the middle class as well, with the widespread practice of taking slave women as sexual partners.56 Islamic scholars, who usually did not belong to the elite, owned concubines.57 The system of sexual slavery seems to have been taken for granted on the part of society at large; it is hard to find evidence of any sort of opposition. There is also undeniably a close connection between wife and slave in legal

60 i

Pernilla Myrne

texts from the period, which depicted marriage as a form of ownership of a woman, as shown by Kecia Ali in her meticulous study of marriage and slavery in early Islam.58 It is, on the other hand, very difficult for us to determine the extent to which Abbasid society relied on concubinage. It might have been less widespread than it appears to be. Other scholars have suggested more explicitly, with an ahistorical attitude, that the sexual mores of Abbasid high society came to have an impact on Muslim women’s situation in the modern era. Perhaps the best known of these scholars, at least for a Western audience, is Fatima Mernissi. She provides the following description of the impact of Abbasid slave concubinage on twentieth-​century Muslim societies: The models of hierarchical relationships that the Qur’an imprints in the deepest zones of the Muslim personality would not have retained such influence in the twentieth century had it not been for the expansion of a legendary Muslim empire that allowed sexual inequality to assert itself and to spread through the phenomenon of the jariya.59 Free Arab women in the early Islamic community and the Umayyad caliphate—​so, roughly, the first century of Islamic history—​were, in Mernissi’s view, independent and often exerted political influence. She argues that aristocratic women were not only involved in politics but also spoke out specifically against the veil and polygyny, which they perceived as a threat to their independence.60 Women’s political position, however, changed profoundly during the Abbasid era as a result of the emergence of influential royal concubines:  “From this point on, on the political stage, women were no longer anything but courtesans.”61 Their marital situation changed as well. Elite men not only preferred slave women, but also were “under the spell of jawari.”62 This spell, however, could not have been the result of any real agency on the part of the women themselves, given that their attraction lay in the fact that, in their company, “the man was by definition superior.” The jawari triumphed over free women as “they obeyed more readily than the hurra (free woman).”63 But the jawari in Mernissi’s view were not simply passive and obedient. She refers to a “revolt of the jawari,” in the process of which the royal slave concubines conquered the caliph and his entourage by means of eroticism and sensuality, and, ultimately, even came to share their authority. She labels the moment, “The triumph of the jariya.”64 At this point, Mernissi’s arguments become somewhat murky, painting a picture of (newly enslaved) foreign women doing anything to seduce the men who enslaved them: [Harun al-​Rashid’s] many conquests led to the enslavement of great portions of the conquered peoples, and the palaces swarmed with jawari, who brought with them their culture and exoticism. There were Persians, Kurds, Romans, Armenians, Ethiopians, Sudanese, Hindus, and Berbers  .  .  .  .  The harems became places of greatest luxury where the most beautiful women of the world played their cultural

A Jariya’s Prospects in Abbasid Baghdad

j 61

differences and mastery of diverse skills and knowledge like winning cards for seducing caliphs and viziers.65 Mernissi is not alone in arguing for the jariya as a contributor to the moral degeneration of Abbasid society, in part by having initiated a challenge to free women. Abdelwahab Bouhdiba labels the jawari as “anti-​wives,” exotic and beautiful concubines who brought with them “an exotic perfume of eroticism.”66 They outmaneuvered free wives deliberately by “usurping femininity and taking it over entirely” for themselves. The result was the “revenge” and “victory” of the jawari, which had a devastating impact on the status of the wife in Muslim societies.67 Ahmed also blames non-​Arab influences for the decline in Arab women’s status during the Abbasid era. These influences came partly from enslaved women, as Persian captive women in the households of Arab Muslim men introduced Persian mores into the heart of the Muslim/​Middle Eastern family.68 Mernissi’s claim that the jawari obeyed more readily than free women may at a first glance seem a truism, but the question is hardly that simple. Free wives were also obliged by law to obey their husbands, and their transgressions were punished more severely. Mernissi’s analytic model, in which foreign jawari are opposed to free Arab women, is unfortunate, and although the dichotomy of foreign/​Arab may be unintentional, its premise seems misplaced.69 The assertions presented here make sense mostly from an activist point of view. The search by modern activists for female role models conjoins feminist historians’ search for a “turning point” in history, when civilization chose the path of discrimination, seclusion, and subjugation of women. This is, indeed, also an activist goal; finding the “turning point” would help create a more equal society without having to refute the fundamentals of the Islamic heritage. But, this kind of activist research also leads easily to unreliable historical analysis; it does little to improve our understanding of women’s roles in history. Still more precarious is the tendency to cast blame on the concubines and slave women themselves, which seems to stem from sheer moralism and even ethnic typecasting. Mernissi and Bouhdiba are part of a scholarly tradition that attributes the decline in status of free women to the Abbasid institution of jawari.70 Little more than a century ago, Jurji Zaydan claimed that men’s fondness for foreign concubines, “led to the degradation of the [Arab] woman and the disappearance of her pride and her independence of mind,” and half a century later, Abd al-​Latif Shararah stated, “Arab women lost their struggle with the foreign concubines, and lost their character with time.”71 It is no coincidence that Jurjani and Shararah, who were both Arab nationalists, emphasize that the concubines were “foreign.” They, like Mernissi, do not point out that many of the slave women were descendants of natives of the countries conquered by Arab warriors, who themselves were “foreign” in those countries. That is to say that the dichotomy of foreign/​indigenous is not so simple in this context. There are several claims about jawari that recur in modern studies. First, that the jawari (especially the less cultured of them) had a negative impact on society, spreading

62 i

Pernilla Myrne

moral corruption and decadence.72 These slave women had an inclination to swindle men by means of making them believe they were in love with them.73 The moral corruption caused by the enslaved women had a negative impact on the status of free women, in part because seclusion and other restrictions were imposed on the latter group, and in part because they were obliged to accept a secondary role.74 Moreover, Abbasid men preferred jawari to free women because the former displayed their beauty, which was often exotic, and had been taught how to please men.75 Free women, in turn, were jealous of their husband’s slave concubines.76 The writers referred to here share an uncritical reading of Abbasid sources, especially The Epistle on Singing-​Girls by al-​Jahiz and his Risala fi al-​Ishq wa’l-Nisa (Epistle on Love and Women), as well as Kitab al-​Muwashsha (The Book of Brocade) by al-​Washsha. Their descriptions of female slave singers as dishonest and particularly prone to cheat men out of money, “by training and by innate instinct,” is quoted as fact.77 The deceitfulness of the qayna is, however, an old theme in Arabic poetry, and should not be read as an objective depiction of a real group of women. The censure of singing girls (dhamm al-​qiyan), for example, is a literary motif exploited thoroughly by al-​Washsha.78 Without diminishing the scholarship of al-​Jahiz, relying on him for an objective social history is risky, and we should beware of transmitting medieval prejudices uncritically. Uncritical readings of the sources are particularly problematic with regard to women’s history. The literary sources either exclude women or provide one-​sided accounts, depending on the specific narrative context. The narrators are almost exclusively men. In addition, the preferred narrator in classical Arabic literature is someone who has seen or taken part in the event he narrates. As a consequence, many, perhaps most, anecdotes and historical reports about jawari depict men’s encounters with them at musical and cultural gatherings. Thus, it is easy to get the impression that jawari are women defined wholly by their relationship to men. Considering the apparent numbers of female slaves in Abbasid palaces, however, at least the high-​society jawari must have spent most of their time with other women.79 Furthermore, any information from this time may have been distorted by the nature of transmission and, especially when it comes to this group of women, the prejudices of contemporaneous and later narrators. In recent decades, there has been an increased interest in qiyan and jawari as a historical and literary phenomenon, and scholars, approaching the sources more critically, have examined the female slaves’ impact on musical history as well as their social roles. Two interesting aspects of female slavery have received attention: the impact of domestic slavery on the Abbasid family and the role of the concubine. Julia Bray has published a significant study in which she considers the role of domestic slavery in urban Abbasid society, and Matthew Gordon discusses the background of individual slave singers and soldiers as domestic slaves.80 Medieval sexual slavery and concubinage, however, have been generally understudied, although there is ample material on hand, especially in legal texts, and, thus, Kecia Ali’s Marriage and Slavery in Early Islam is a significant contribution.

A Jariya’s Prospects in Abbasid Baghdad

j 63

Several nuanced portraits of individual influential jawari and qiyan have been composed since Abbott’s pioneering studies of Khayzuran.81 Gordon has portrayed Arib, perhaps the most famous singer of all, and Nadia El-​Cheikh has written about Shaghab, mother of the caliph al-​Muqtadir (295–​320/​908–​932).82 Gordon’s study of the careers of prominent qiyan and their ability to face challenges in the highly competitive milieu of the court is particularly informative about these women’s agency and individual achievements.83 Kristina Richardson, for her part, has examined the agency of the Abbasid qiyan and their approach to gaining personal profit by means of sexual manipulation and by taking advantage of their intimacy with the men in power.84 Lisa Nielson has investigated the role of women slave musicians from a historical perspective, as well as Abbasid authors’ attitudes toward jawari and qiyan.85 Bray, for her part, has discussed the literary prominence of slave women in literature depicting Abbasid culture, as opposed to that of the Umayyad and early Islamic periods, during which most women are likely to have been freeborn. Bray suggests that the coming of Abbasid slave women as prolific literary characters in romantic literature reflects a shift in the dynamics of Abbasid society, with its greater variety of career options for men and greater interest in individuality.86 Romantic love was, additionally, a means for men to develop their individuality and achieve emotional fulfillment. Framed as standing outside the regulated power relationships of the family, it found in “the urban Abbasid slave heroine,” who educates her lovers in culture and fine manners, a means to prepare the young men for a career at court.87 Gender, Slavery, and Social Hierarchy The sources provide some evidence of women’s roles as slave owners in their own right and their relations to the slave women. Elite households kept jawari as ladies-​in-​waiting and entertainers: the ownership of jawari was a mark of prestige and a sign of high standing. In al-​Masudi’s account about Abbada,88 the mother of Jafar Ibn Yahya al-​Barmaki (d. 803 ce), she complains about her loss of status after the destruction of the Barmakids by Harun al-​Rashid. In her own words, her former honorable position is symbolized by the hundreds of jawari who used to accompany her: “There was a time when this same feast saw me escorted by 400 serving-​girls—​and yet I thought my son showed me a lack of respect. Today the feast is here again and I desire no more than two sheep skins, one to serve me as a bed and one to clothe me.”89 Queen Zubayda owned a “large retinue of palace boys (eunuchs) and girls,” and took pride in their accomplishments.90 She is said to have had 100 jawari who were experts in chanting the Quran.91 Free women are also noted for enjoying the company of the most accomplished singers and courtesans. Just like men, they appreciated their artistic and cultural skills. Al-​Azdi relates that Shahik, the grandmother of Ali ibn Hisham, enjoyed listening to his jawari when she came to visit him from Khurasan. She was especially fond of Mutayyam and rewarded her with money.92 There are many other examples of royal women enjoying the

64 i

Pernilla Myrne

company of qiyan in the Kitab al-​Aghani, in which several of the greatest singers are portrayed. The singer Duqaq, for example, was a lady-​in-​waiting for the princess Hamduna, Harun al-​Rashid’s daughter.93 The famous singer Arib was supposedly 14 years old when she became the concubine of the caliph al-​Amin, al-​Rashid’s son.94 When al-​Amin was overthrown by his half-​brother al-​Mamun, Arib was, according to one report, taken back by her former owner, which prompted Zubayda, al-​Amin’s mother, to say, “The worst thing that happened to me after the assassination of my son Muhammad [al-​Amin] was when al-​Marakibi attacked my house and took Arib.”95 Women were not only slave owners but also slave traders, taking part, in particular, in the training of qiyan to sell them.96 A woman of the Quraysh, for example, is reported to have acquired Shariya in Basra, then taught her to sing before selling her in Baghdad.97 The singer Mutayyam, also from Basra, was owned by a woman who sold her as a young child to Ali ibn Hisham.98 The poet and singer Jullanar, from Kufa, was owned by a woman who sold her to her brother for his part of the heritage from their father.99 At least one of the female slave traders is named in the sources, Barbar, who became well known for her slave singers.100 As the preceding examples show, women owned, trained, and traded in jawari. The fact that women invested in slavery, however, does not mean it did not contribute to women’s subordination as a group, inasmuch as concubinage, as well as polygyny and easy divorce, likely weakened the overall position of women. Men did not have access to their wives’ slaves, but a wife might have felt compelled to give an attractive jariya to her husband if he asked her to do so. There are several examples in the sources of wives gifting one or several of their slave girls to their husbands to please them.101 There are also examples of women being jealous of their husband’s potential or real concubines. Zubayda, for example, managed to prevent her husband from purchasing a potential rival. According to a variant of this anecdote, the potential rival was Inan al-​Natifi.102 And Zubayda was, as we have seen, fond of jawari herself. Her own mother was, in fact, a slave concubine. It is likely that she would not have regarded a slave concubine as a more threatening rival than a well-​connected freeborn co-​wife. Zubayda was a very influential woman, however, and the cousin of her husband, the caliph al-​Rashid (whose mother had also been a slave concubine). Thus, the situation of royal women should not be taken as an example for all women. In legal terms, concubines had some advantages over free women. The law stipulated men could divorce their free wives easily—​which meant that free women could be forced to move abruptly from their home and children—​whereas men were not allowed to sell their concubines if they had given birth to their children (and if paternity was acknowledged).103 This, together with the fact that jawari and qiyan were “freer” than freeborn women (using a modern concept of freedom), suggests a more complex reality to the lives and careers of the female slaves. They were usually allowed to move more freely (at least if they were not kept as concubines) and, it appears, they could move up the social ladder more easily than freeborn women. Nonetheless, both slave women and freeborn

A Jariya’s Prospects in Abbasid Baghdad

j 65

women were subordinated, and their legal status was vulnerable. As Bray observes, “The positions of free women and of slaves, male or female, are analogous in that both vary according to circumstance, whereas the legal status of free Muslim male is invariable.”104 All this said, it was still the case that slave women in general were more vulnerable than free women. Most concubines did not have the status of umm walad (mother of her owner’s child), but were kept for sexual enjoyment with the explicit aim of avoiding pregnancy. Thus they could be sold or given away without legal restraint.105 “Luxury” female slaves were often given away in diplomatic exchanges, and there was a “vast movement of [enslaved] women from court to court, from country to country, from continent to continent.”106 Domestic slaves belonging to less influential families likewise seem to have changed hands frequently. Studies of the Geniza records show that in medieval Cairo, female slaves changed owners much more frequently than male slaves.107 Although there are signs of affection between slave-​holding families and their female slaves, the slaves’ position in less wealthy households as being, on the one hand, sexually available to their masters and, on the other hand, providing domestic help to their master’s wives, was much more risky than the position of male slaves, who often were entrusted with important tasks in the family business.108 Furthermore, family connections probably made it difficult for many men to divorce wives, whereas they could sell their concubines easily and legally as long as they did not have a living child together or did not acknowledge paternity. Moreover, although freeborn wives may not have been happy with their husbands’ concubines, the “patriarchal bargain” gave them something in return—​for example, honor and social respect.109 And the immediate effects of slave ownership may not have been perceived as negative for free women. The possession of jawari displayed wealth and gave prestige to the whole family, including free wives, even if their husbands kept some of them as concubines.110 Underlying all systems of slavery are ideas about natural hierarchies of race, ethnicity, and class that position some categories superior to others. To be a freeborn woman in Abbasid society was worthy of social respect in itself, and free wives probably considered themselves superior morally and socially to slave concubines. Piety was an intrinsic quality of free, honorable women, which assured them a reward in the hereafter not attained as easily by slave women. Women’s practice of piety was connected to their willing sacrifice to fulfill their husband’s desires and obey them, represented by the tradition stating that husn al-​taba‘ul (obedience and devotion to the husband) is women’s jihad.111 In light of that, Zubayda’s gift of beautiful slave girls to her husband, Harun al-​ Rashid, as a means of asking him for forgiveness for her jealousy, was an act of piety that would not only bring her husband’s appreciation, but also a reward in the hereafter.112 Concluding Remarks Urban Abbasid society was characterized by a remarkable social mobility that made it possible for the able and diligent individual to change his or her living conditions for the better. Social mobility, however, did not only depend on the ability of the individual

66 i

Pernilla Myrne

but also was subject to a range of social variables, including gender, ethnicity, social class, and degrees of freedom. Free women and slaves were all subordinated, but had different opportunities. Freeborn women had higher status, yet slave women could move up the social ladder more easily and improve their living conditions. Although free women’s family situation could change abruptly if their husbands divorced them, they could at least assume that their paternal family cared for them until they remarried. Female slaves, on the other hand, were more dependent on their ability to adapt to change and to grab the opportunities with which they might occasionally be provided, such as at the time of purchase. The investment that owners made in the education of some slave women provided them with the intellectual and cultural capital that made it easier for them to control their life course to some extent. To succeed, they had to impress by manifesting their skill, wit, and individuality, and thus be noticed by the audience and patrons. Their ability to use poetry and other cultural accomplishments could be decisive for their success. This ability also gave them a chance to exercise some degree of agency and perhaps to influence the course of their lives. They had to exploit opportunities as they arose and demonstrate their skills, which in turn relied on their ability to draw from the situation at hand. The highest form of success was to secure a position in the royal court. Apparently, the feminine milieu of the palaces, with the high number of accomplished and well-​educated women, not only fostered envy and intrigues, as some scholars suggest, but also friendship and a dynamic cultural environment. Perhaps these were a significant part of the court’s attraction to Inan, rather than the prospect of being the caliph’s concubine. Inan’s life story, although discontinuous and episodic, illustrates the career of a slave courtesan and poet in early Abbasid Baghdad. She was educated and accomplished, and appreciated for her poetry. Part of her success was her ability to improvise on her own situation—​a demand of every slave. She attempted to take advantage of her talent and training to change her living conditions. She was ambitious, apparently attracted by the imperial court, and endeavored to be purchased by the caliph. Eventually, however, she depended on her owner’s will. As a slave, her owner could give her away or sell her without consulting her, if he wished. Inan’s life story shows us that, although possibilities of social mobility were within reach for a talented slave courtesan in the imperial center, her opportunities remained subject to her unfree status. Notes 1. Jariya (pl., jawari), “girl,” is the common term for female domestic slaves and courtesans in urban Abbasid society. Although the word jariya often indicates a high-​society slave, whereas a common female slave would be called ama, this differentiation is not absolute. I also use the word qiyan (sing., qayna), a term for skilled slave singers. For a distinction between jawari and qiyan, see Nielson, “Gender and the Politics of Music,” 252–​253. 2. For the literary role of the jariya, see for example Bray, “Men, Women and Slaves,” 136–​139, and “Isnads,” 12–​15; and Van Gelder, “Slave-​Girl Lost and Regained.” For some individual jawari

A Jariya’s Prospects in Abbasid Baghdad

j 67

and their historical roles, see, for example, Abbott, Two Queens; Gordon’s articles on Arib; and El-​Cheikh’s articles on the Abbasid caliphal harem. 3. Cf. Atraqji, Marʾa. 4. Cf. Brockopp, Early Maliki Law, 131, on the concept of slavery as being natural in the Quran. 5. Al-​Isbahani, Ima, 23–​44, and Aghani, 84–​91. For al-​Ima al-​Shawair¸ see Kilpatrick, “Women as Poets and Chattels.” 6. Al-​Isbahani is usually generous with his praise for the female singers and poets that he portrays. 7. Bencheikh, “ʿInan,” 1202; and Ibn al-​Nadim, Fihrist, 361. She is also portrayed by Caswell, Slave Girls, 56–​81, and al-​Heitty, The Role of the Poetess, 113–​124. 8. Cf. al-Suli, Ashar. For another aspect of female palace tradition, see Bray, “Bleeding Poetry.” 9. Muwallada is often translated as “having an Arab father and a slave mother.” The explanation given by Ibn Manzur, Lisan, s.v. w-​l-​d, is more likely, however: “A jariya muwallada is born among Arabs and raised with their children; they [the Arabs] have nourished them with their food and taught them their manners.” 10. On al-​Natifi and his possible low standing, see al-​Heitty, The Role of the Poetess, 113. 11. Ibn Jarrah, Waraqa, 41, and al-​Isbahani, Ima, 27–​28. On the second occasion, Abu Nuwas found Inan crying after al-​Natifi had whipped her. He asked Abu Nuwas to distract her and they engage in a racy poetic duel. 12. The editor of the Ima provides a list of sources. I only refer to some of the more substantial references here. 13. Ibn Jarrah, Waraqa, 41. 14. For this line and Inan’s answer, I rely on Caswell’s translation. He has translated several of her known poems in Slave Girls. 15. Marwan is the visitor in al-​Isbahani, Aghani, 23:85, and Ima, 24–​25; al-​Tawhidi, al-​Basa’ir, 5:14; Ibn al-​Sa`i, Nisa al-​khulafa, 48; and al-​Suyuti, Mustazraf, 39. According to Ibn Abi Rabbih Iqd, 6:61, the male poet was Bakr ibn Hammad al-​Bahili. In Ps.-​Jahiz Mahasin, 193–​194, he is anonymous. 16. Although a few jurists permitted the flogging of disobedient wives as a last resort when they did not obey, most advocated lighter corporal punishment, see Marín, “Disciplining Wives.” 17. Ibn Sad, Tabaqat, 8:165. See Marín, “Disciplining Wives,” 18. 18. Al-Isbahani, Aghani, 21:50. She claimed to be the daughter of Jafar al-​Barmaki; see Gordon, “Arib.” For a less famous, anonymous, jariya being flogged by her owner, see Aghani, 16:283. 19. Al-Isbahani, Ima, 24–​25. This variant is quoted in al-​Isbahani, Aghani, 23:85; al-​Suyuti, al-​Mustazraf, 40; and Ibn al-​Sa`i, Nisa, 49. In Ibn Abd Rabbih (6:61), the narrator, Bakr bin Hammad al-​Bahili, wants to meet Inan after hearing about her skills. Her owner invites him to sit with Inan, but when they entered her room she says she is tired, which makes him strike her with his whip. 20. Ps.-​Jahiz, Mahasin, 194. 21. Al-Isbahani, Aghani, 22:140. 22. The verse is in Caswell’s translation, Slave Girls, 144. 23. Cf. Kilpatrick, “Women as Poets and Chattels,” 167. 24. Al-Isbahani, Ima, 81–​82. 25. The whole poem in Caswell’s translation (Slave Girls, 136) reads, “You lost your temper for no offense of mine /​And it is you who snubs, turns away and shuns /​With the might of your rule

68 i

Pernilla Myrne

you subdued an obedient being /​And but for the submission of slavery I would not bear it /​If you would reflect on what you did apologies are called for /​Or else be unjust and you’ll be forgiven.” 26. Kilpatrick’s translation (169) of Murad’s poem. 27. Ps.-​Jahiz, Mahasin, 193–​194. 28. Al-Isbahani, Ima, 43–​44. 29. Ibn al-​Mutazz, Tabaqat, 421–​422. 30. Ibn Jarrah, Waraqa. 40. 31. Ps.-​Jahiz, Mahasin, 196. 32. Al-Isbahani, Ima, 44. 33. Ps.-​Jahiz, Mahasin, 197; al-​Isbahani, Ima, 41, and Aghani, 23:89. 34.  For this anecdote, see al-​Isbahani, Aghani, 23:89; al-​Suyuti, al-​Mustazraf, 45; al-​Heitty, Role, 118–​119; and Kennedy, Court of the Caliphs, 175–​176. 35. Al-​Isbahani, Aghani, 23:89. The anecdote is ambiguous because of unclear references. Inan’s new owner is only referred to as “he,” which could be Harun al-​Rashid instead of the anonymous bidder; the caliph died in Khurasan in 809. 36. Al-Isbahani, Ima, 42. According to Ibn al-​Sa`i (Nisa’, 53), she died in Egypt in 841. 37. Al-​Isbahani, Ima, 31. 38. Ibn Jarrah, Waraqa, 40. 39. Ibn al-​Sa`i, Nisa’, 48. 40. Al-​Isbahani, Ima, 41–​42; Ibn Abd Rabbih, ʿIqd, 6:60. 41. See al-​Suyuti, Mustazraf, 32–​33; and Ibn al-​Mutazz, Tabaqat, 422. See also Atraqji, Marʾa, 192–​193, on the letters of Inan and Sakan. 42. Cf. Tabbubi, Qiyan, 235–​236. 43. Cf. Ragib, “Marchés,” 724. He concludes that luxury slaves often changed owners in private, without intermediates, but that owners sometimes put them up for public auction only to humiliate them or correct them. Cf. al-​Heitty, “Contrasting Spheres,” 34. 44. However, according to Islamic law, female slaves only had to display their face and hands (Ragib, “Marchés,” 735–​736). 45. Cf. Atraqji, Marʾa, 193; and Kilpatrick, “Women as Poets and Chattels,” 165–​166. 46. Cf. al-​Abi, Nathr al-​Durr, vol. 4, who has gathered several anecdotes of this kind. 47. Al-Isbahani gives some accounts about successful presentations of individual jawari, which may have more historical value (ex. Fadl, Aghani, 19:215; Rayya and Zamya, Ima, 115–​116; Amal, Ima, 125; Nabt, Ima, 130–​131). al-​Mutawakkil seems to be the caliph appearing most frequently in these anecdotes. 48. Abbott, Two Queens, 8. 49.  Abbott, “Women and the State,” 351:  “This situation resulted, in its turn, in a definite class distinction between the free Arab woman of noble race and lineage, haughty but generally virtuous, and the foreign slave woman, singer or concubine, with pride of beauty and talent but easygoing and of comparatively loose morals.” 50. Abbott, Two Queens, 9. 51. Women have certainly always been more affected by slavery—​for example, by sexual abuse. They also appear to have been enslaved in larger numbers; cf. Lerner, The Creation of Patriarchy, Chapter Four (“The Woman Slave”). 52. Ahmed, Women and Gender, 83–​87. 53. Ahmed, Women and Gender, 85, 86.

A Jariya’s Prospects in Abbasid Baghdad

j 69

54. Ahmed, Women and Gender, 84. 55. Meisami, “Writing Medieval Women,” 65; and Ali, Marriage and Slavery, 22. 56. Ahmed, Women and Gender, 85. 57. For al-​Shafiʿi, see Kecia Ali, Imam Shafiʿi, 40–​42. 58. Ali, Marriage and Slavery. 59. Mernissi, Women’s Rebellion, 69. Her view is criticized by Caswell, Slave Girls, 46–​48. 60. Mernissi, Women’s Rebellion, 81–​83. 61. Mernissi, Women’s Rebellion, 84. 62. Mernissi, Women’s Rebellion, 85. 63. Mernissi, Women’s Rebellion, 87. 64. Mernissi, Forgotten Queens, ch. 3, and Women’s Rebellion, 84. She compares them to the zanj, the black labor slaves who rebelled 255/​869: “Unlike the zanj, who tried to seize power from the periphery of the system, the jawari operated within the caliph’s palace itself, in the bed and the heart of the man whom the law set up as absolute master of souls and possessions” (Forgotten Queens, 37–​38). 65.  Mernissi, Forgotten Queens, 58. Nevertheless, the conquests that brought most slaves to the Muslim Empire took place long before Harun al-​Rashid, and many of the slaves in Baghdad were probably born and raised in the Empire. Most of the prominent jawari whose origins are noted by Abu al-​Faraj al-​Isbahani, in Kitab al-​Aghani and al-​Ima al-​Shawa’ir, were muwallidat (i.e., slaves who were born and raised among Arabs). There were also many slaves in Baghdad who were brought from the conquered provinces as a part of the capitulation agreement, and, of course, there was an extensive slave trade from Europe and other places. See Athamina, “How Did Islam Contribute,” 392–​393; and Gordon, “Preliminary Remarks,” 77–​79. In addition, Athamina (“How Did Islam Contribute,” 393) mentions kidnapping as a source of slaves. For the unknown provenance of most domestic slaves and concubines, see Bray, “Family,” 732. 66. Bouhdiba, Sexuality in Islam, 106–​108: “Arab feminism was a victory of these anti-​wives. They alone, indeed, and almost to the exclusion of the legitimate wives, finally got their way over men” (108). 67. Bouhdiba, Sexuality in Islam, 106–​108. 68. Ahmed, Women and Gender, 82. The tendency in modern scholarship to attribute negative influences to the Persians is amply discussed by Massad in Desiring Arabs. 69.  See, for example, Mernissi, Women’s Rebellion, 85–​86, in which she compares the phenomenon with Moroccan men’s “infatuation for foreign women” (i.e., French women, after the Independence, 86–​87); and Forgotten Queens, 44. 70. See Massad, Desiring Arabs, 116–​118. In contrast, Athamina (“How Did Islam Contribute,” 403–​408) seems to claim they had a positive impact. 71. Both quoted in Massad, Desiring Arabs, 117. Zaydan is one of the writers to whom Mernissi refers (see Women’s Rebellion, 91, note 20). 72. See, for example, Amin, Duha, 1:95; Tabbubi, Qiyan, 76; Atraqji, Marʾa, 38; al-​Heitty, “The Contrasting Spheres,” 40; Caswell, Slave Girls, 39. 73. Cf. Atraqji, Marʾa, 39–​44; Tabbubi, Qiyan, 76; Caswell, Slave Girls, 39. 74. Atraqji, Marʾa, 6–​7, 33–​42. She claims that the moral corruption of the foreign jawari resulted in the widespread misogyny of the era and unwillingness to marry among many men. See also al-​Heitty, “The Contrasting Spheres,” 40, and Caswell, Slave Girls, 39. 75. Cf. Amin, Duha, 1:98, Tabbubi, Qiyan, 34–​35, and Caswell, Slave Girls, 45.

70 i

Pernilla Myrne

76. Cf. Atraqji, Marʾa, 126, and Tabbubi, Qiyan, 34. 77. Al-​Jahiz, Singing Girls, 31–​32. 78. See Gordon, “Yearning and Disquiet,” for al-​Jahiz’s view on singing girls. For al-​Jahiz and al-​Washsha on this subject, see Nielson, “Gender and the Politics of Music.” Notably, however, it is the qayna, not the jariya in general, who is censured by these authors. 79.  See Nadia Maria El-​Cheikh’s studies of the harem of Caliph al-​Muqtadir (r. 908–​ 932): “Revisiting the Abbasid Harem”; “Qahramana”; and “Caliphal Harems.” 80. Bray, “Men, Women and Slaves,” 132–​136, and Gordon, “Preliminary Remarks.” See also Bray, “Family,” for relevant research questions and survey. 81. Abbott, Two Queens. 82. Gordon, “ʿArib,” and “Preliminary Remarks” (on Mahbuba); and El-​Cheikh, “Gender and Politics.” For “Arib,” see also Gordon, “Place of Competition,” and “Yearning and Disquiet,” 263–​ 268. For a narrative analysis and translation of Arib’s biography in al-​Isbahani’s Kitab al-​Aghani, see Myrne, Narrative, Gender and Authority. 83. Gordon, “Place of Competition.” 84. Richardson “Singing Slave Girls.” 85. Nielson, “Gender and the Politics of Music.” 86. Nielson, “Gender and the Politics of Music,” 136–​138. 87. Bray, “Men, Women and Slaves,” 138; and also Bray, “Isnads,” 12–​15. 88. A variant spelling of her name is Attaba (see al-​Jahshiyari, Wuzara’, 241). 89. Masudi, Meadows, 126. 90.  Abbott, Two Queens, 160. For Fatimid imperial women and their numerous jawari, see Cortese and Calderini, Women and the Fatimids, 78. 91. Abbott, Two Queens, 160. 92. Al-​Azdi, Bada’i, 335. Ali ibn Hisham was a military commander, patron of singers, and owner of famous qiyan; see al-​Isbahani, Aghani, 17:59–​61. For Mutayyam, see Aghani, 7:222–​233. She was Ali’s favorite concubine, according to al-​Isbahani, and his umm walad (mother of his children). 93. Al-​Azdi, Badaʿiʾ,12:203. 94. Al-​​Azdi, Badaʿiʾ, 21:59. 95. Al-​Azdi, Badaʿiʾ, 21:52. 96. Cf. Bray, “Men, Women and Slaves,” 137, note 44: “for women of good family, training them for resale seems to have been a long-​term home-​based investment.” 97. Al-​Isbahani, Aghani, 16:5. 98. Al-​Isbahani, Aghani, 7:222. 99. Al-​Isbahani, Aghani, 145–​146. 100. Tabbubi, Qiyan, 53. 101. Cf. the example in Abbott, Two Queens, 140. 102. Cf. al-​Isbahani, Ima, 40–​41, and Aghani, 23:88. 103.  The Abbasid family was “an inherently unstable entity” (Bray, “Men, Women and Slaves,” 133). 104. Bray, “Men, Women and Slaves,” 132–​133. 105. Cf. Richardson, 107–​108. Some Sh‘ite hadiths permit men to “lend” their jawari to other men for sexual purposes; see Kulayni, Kafi, 5:468–​470.

A Jariya’s Prospects in Abbasid Baghdad

j 71

106. See Cortese and Calderini (78) on Fatimid slave exchange during the early tenth century, which probably did not differ much from Abbasid luxury slave exchanges. Slave women could be given away as reward or encouragement to civil servants, soldiers, or palace servants. See, for example, Gordon, Breaking, 58, 62, for slave women given to Turkish soldiers as wives. 107. Goitein, “Slaves and Slavegirls,” 3. 108. Goitein, “Slaves and Slavegirls,” 3. 109. For the “patriarchal bargain,” see Kandiyoti, “Bargaining with Patriarchy.” 110. Cf. Bray, “Men, Women and Slaves,” 136. 111. Several hadiths circulating during the ninth century state that women are required to obey their husbands. For the saying that husn al-​taba‘ul is women’s jihad, see al-​Kulayni, Usul, 5:9, 507. 112. Cf. Abbott, Two Queens, 140.

Bibliography Primary Sources al-​Abi, Abu Sa`d Mansur ibn al-​Husayn (d. 421/​1030). Nathr al-​Durr. 7 vols. Edited by Muhammad Ali Qurna, Ali Muhammad al-​Bijawi, Husayn Nassar, et al. Cairo: al-​Hay’a al-​ Misriyya al-​Amma li-​l-​Kitab, 1980–​1991. al-​ Azdi, Ali ibn Zafir. Bada’i al-​bada’ih. Edited by Muhammad Abu al-​ Fadl Ibrahim. Cairo: Maktabat al-​Anglu al-​Misriya, 1970. Ibn Abd Rabbih, Abu Umar Ahmad ibn Muhammad. Al-​Iqd al-​Farid. 6 vols. Edited by Muhammad al-​Tunji. Beirut: Dar Sadir, 2006 [1427]. Ibn Jarrah, Abu Abdallah Muhammad ibn Dawud. Al-​Waraqa. Edited by Abd al-​Wahhab Azzam and Abd al-​Sattar Ahmad Farraj. Cairo: Dar al-​Maarif, 1953. Ibn Manzur, Muhammad ibn Mukarram. Lisan al-​Arab. Beirut: Dar al-​Sadir, 1968. Ibn al-​Mutazz, Abu al-​Abbas Abdallah. Tabaqat al-​Shu`ara. Edited by Abd al-​Sattar Ahmad Faraj. Cairo: Dar al-​Maarif, n.d. Ibn al-​Nadim, Muhammad Ibn Ishaq. Al-​Fihrist. The Fihrist of al-​Nadim. 2 vols. Edited and translated by Bayard Dodge. New York: Columbia University Press, 1970. Ibn al-​Sa`i, Abu Talib Ali ibn Anjab. Nisa al-​Khulafa. Edited by Mustafa Jawad. Cairo:  Dar al-​Maarif, n.d. Ibn al-​Washsha, Abu Tayib ibn Muhammad ibn Ishaq. Le Livre de brocart (The Book of Brocades). Translated by Siham Bouhlal. Paris: Editions Gallimard, 2004. —​—​—​. Al-​Muwashsha, aw al-​zarf wa’l-​zurafa. Beirut: Dar Sadr, 1964. al-​Isbahani, Abu al-​Faraj. Al-​Ima al-​Shawair. Edited by Nuri Hamudi al-​Qaysi and Yunus Ahmad al-​Samarrai. Beirut: Alam al-​Kutub, 1986 [1406]. —​—​—​. Al-​Kitab al-​Aghani. 25 vols. Edited by Ihsan Abbas, Ibrahim al-​Saafin, and Bakr Abbas. Beirut: Dar Sader, 2002 [1423]. Al-​Jahiz, Amr ibn Bahr. The Epistle on Singing-​Girls by Jahiz. Translated and edited by A. F. L. Beeston. Warminster, UK: Aris & Phillips, 1980. —​—​—​. “Risala fi al-​ishq wa’l-​nisa.” In Rasa’il Jahiz, edited by Hasan al-​Sandub, 266–​275. Cairo: Matbaʿa al-​Tijariyya, 1933 [1352]. Ps.-​Jahiz, Abu Uthman. Mahasin wa’l-​addad. Edited by G. Van Vloten. Leiden: Brill, 1898.

72 i

Pernilla Myrne

al-​Jahshiyari, Abu Abdallah Muhammad ibn Abdus. Kitab al-​wuzara wa’l-​kuttab. Edited by M. al-​Saqqa, I. al-​Abyari, and A. al-​H. Shalabi. Cairo: Maktabat Isa al-​Babi al-​Halabi, 1980. al-​Kulayni, Muhammad ibn Ya`qub. Al-​Usul min al-​kafi. 8 vols. Edited by Ali Akbar al-​Ghaffari. Beirut: Dar al-​Adwa, 1985. al-​Masudi, Abu al-​Hasan Ali ibn Husayn. The Meadows of Gold: The Abbasids. Translated and edited by Paul Lunde and Carolie Stone. London: Kegan Paul International, 1989. al-​Suyuti, Jalal al-​Din. Al-​Mustazraf min akhbar al-​jawari. Edited by Salah al-​Din al-​Munajjid. Beirut: Dar al-​Kitab al-​Jadid, 1963. al-​Tawhidi, Abu Hayyan. Al-​Basaʾir wa’l-​dakhaʾir. 10 vols. Edited by Wadad al-​Qadi. Beirut: Dar Sadir, 1988 [1408].

Secondary Sources Abbott, Nabia. Two Queens of Baghdad: Mother and Wife of Harun al-​R ashid. Chicago: University of Chicago Press, 1946. —​—​—​. “Women and the State in Early Islam 2.” Journal of Near Eastern Studies 1, no. 3 (1942): 341–​368. Ahmed, Leila. Women and Gender in Islam. New Haven, CT: Yale University Press, 1992. Ali, Kecia. Imam Shafiʿi: Scholar and Saint. Oxford: Oneworld Publications, 2011. —​—​—​. Marriage and Slavery in Early Islam. Cambridge, MA: Harvard University Press, 2010. Amin, Ahmad. Duha al-​Islam. 3 vols. Cairo: Maktabat al-​nahda, 1977 [1933–​1936]. al-​Asad, Nasr al-​Din. Al-​Qiyan wa-​l-​Ghina fi al-​Asr al-​Jahili. Beirut: Dar al-​Jil, 1960. [Reprinted in 1988.] Athamina, Khalil. “How Did Islam Contribute to Change the Legal Status of Women: The Case or the Jawārī, or the Female Slaves.” Al-​Qantara 28, no. 2 (2007): 383–​408. Atraqji, Wajida. Al-​Mar’a fi Adab al-​ʿAsr al-​ʿAbbasi. Baghdad: Dar al-​Rashid li-​l-​Nashr, 1981. Bencheikh, J. E. “ʿInan.” In Encyclopaedia of Islam, 2nd ed., edited by P. Bearmann et  al. Leiden: Brill Publishing Online, 2015. Bouhdiba, Abdelwahab. Sexuality in Islam. Translated by Alan Sheridan. London: Saqi Books, 1998. Published originally as La sexualité en Islam. Paris: Presses Universitaires de France, 1975. Bray, Julia. “The Family in the Medieval World.” History Compass 9, no. 9 (2011): 731–​742. —​—​—​. “Isnāds and Models of Heroes: Abū Zubayd al-​Ṭā’ī, Tanūkhī’s Sundered Lovers and Abū ‘l-​ʿAnbas al-​Ṣaymarī.” Arabic and Middle Eastern Literatures 1, no. 1 (1998): 7–​30. —​—​—​. “Men, Women and Slaves in Abbasid Society.” In Gender in the Early Medieval World: East and West, 300–​900, edited by Leslie Brubaker and Julia M. H. Smith, 121–​146. Cambridge: Cambridge University Press, 2004. —​—​—​. “Third and Fourth Century Bleeding Poetry.” Arabic and Middle Eastern Literature 2, no. 1 (1999): 75–​92. Brockopp, Jonathan E. Early Mālikī Law: Ibn ʿAbd al-​Ḥakam and His Major Compendium of Jurisprudence. Leiden: Brill, 2000. Caswell, Fuad Matthew. The Slave Girls of Baghdad:  The Qiyan in the Early Abbasid Era. London: I.B. Tauris, 2011. Cortese, Delia, and Simonetta Calderini. Women and the Fatimids in the World of Islam. Edinburgh: Edinburgh University Press, 2006.

A Jariya’s Prospects in Abbasid Baghdad

j 73

El-​Cheikh, Nadia Maria. “Caliphal Harems, Household Harems: Baghdad in the Fourth Century of the Islamic Era.” In Harem Histories: Envisioning Places and Living Spaces, edited by Marilyn Booth, 87–​103. Durham, NC: Duke University Press, 2010. —​—​—​. “Gender and Politics in the Harem of al-​Muqtadir.” In Gender in the Early Medieval World: East and West, 300–​900, edited by Leslie Brubaker and Julia M. H. Smith, 147–​161. Cambridge: Cambridge University Press, 2004. —​—​—​. “The Qahramâna in the Abbasid Court:  Position and Functions.” Studia Islamica 97 (2003): 41–​55. —​—​—​. “Revisiting the Abbasid Harems.” Journal of Middle East Women’s Studies 1, no. 3 (2005): 1–​19. Goitein, D. D. “Slaves and Slavegirls in the Cairo Geniza Records.” Arabica 9 (1962): 1–​20. Gordon, Matthew S. “ʿArib al-​Ma’muniyah (797–​890).” In Arabic Literary Culture, edited by Michael Cooperson and Shawkat Toorawa, 85–​90. Detroit, MI: Thomson Gale, 2005. —​—​—​. The Breaking of a Thousand Swords: A History of the Turkish Military of Samarra (A.H. 200–​275/​815–​889 C.E.). Albany: State University of New York Press, 2001. —​—​—​. “The Place of Competition:  The Careers of ʿArib al-​Maʾmunīya and ʿUlayya bint al-​ Mahdī, Sisters in Song.” In Occasional Papers of the School of `Abbasid Studies, edited by James E. Montgomery, 61–​81. Leuven: Peeters, 2004. —​—​—​. “Preliminary Remarks on Slave Labor in the Third/​Ninth Century ʿAbbāsid Empire.” In Slaves and Households in the Near East, edited by Laura Culbertson, 71–​84. Chicago: University of Chicago Press (The Oriental Institute), 2011. —​—​—​. “Yearning and Disquiet:  Al-​Jahiz and the Risālat al-​Qiyan.” In Al-​Jahiz:  A  Muslim Humanist for Our Time, edited by Arnim Heinemann et al., 253–​268. Beirut: Ergon Verlag, 2009. al-​Heitty, Abdul-​Karem. “The Contrasting Spheres of Free Women and Jawārī in the Literary Life of the Early Abbasid Caliphate.” al-​Masāq 3 (1990): 31–​51. —​—​—​. The Role of the Poetess at the Abbasid Court (132–​347 A.H./​750–​861 A.D. Beirut:  Al Rayan, 2005. Kandiyoti, Deniz. “Bargaining with Patriarchy.” Gender and Society 2, no. 3 (1988): 274–​290. Kennedy, Hugh. The Court of the Caliphs:  The Rise and Fall of Islam’s Greatest Dynasty. London: Weidenfeld & Nicolson, 2004. Kilpatrick, Hilary. “Women as Poets and Chattels: Abu l-​Faraǧ al-​Isbahāni’s “al-​Imā al-​šawā’ir.” Quaderni di Studi Arabi 9 (1991): 161–​176. Lerner, Gerda. The Creation of Patriarchy. New York: Oxford University Press, 1986. Marín, Manuela. “Disciplining Wives: A Historical Reading of Qur’ān 4:34.” Studia Islamica 97 (2003): 5–​40. Massad, Joseph A. Desiring Arabs. Chicago: University of Chicago Press, 2007. Meisami, Julie. “Writing Medieval Women.” In Writing and Representation in Medieval Islam, edited by Julia Bray, 47–​87. London: Routledge, 2006. Mernissi, Fatima. The Forgotten Queens of Islam. Translated by Mary Jo Lakeland. Minneapolis: University of Minnesota Press, 1993. —​—​—​. Women’s Rebellion and Islamic Memory. London: Zed Books, 1996. Myrne, Pernilla. Narrative, Gender and Authority in ‘Abbāsid Literature on Women. Göteborg: Acta Universitatis Gothoburgensis, 2010.

74 i

Pernilla Myrne

Nielson, Lisa. “Gender and the Politics of Music in the Early Islamic Courts.” Early Music History 31 (2012): 235–​261. Ragib, Yusuf. “Les Marchés aux esclaves en terre d’islam.” In Mercati e mercanti nell’alto medioevo: L’area euroasiatica e l’area mediterranea, n.e., 721–​763. Spoleto: Centro Italiano di Studi Sull’alto Medioevo, 1993. Richardson, Kristina. “Singing Slave Girls (Qiyan) of the ʿAbbasid Court in the Ninth and Tenth Centuries.” In Children in Slavery through the Ages, edited by Gwyn Campbell, Suzann Miers, and Joseph C. Miller, 105–​118. Athens: Ohio University Press, 2009. Scott Meisami, Julia. “Writing Medieval Women: Representations and Misrepresentations.” In Writing and Representation in Mediaeval Islam: Muslim Horizons, edited by Julia Bray, 47–​87. Abingdon: Routledge, 2006. Tabbubi, Layla. Al-​Qiyan wa’l-adab fi al-​Asr al-​Abbasi al-​Awwal. Beirut: al-​Intishar al-​Arabi, 2010. Van Gelder, Geert Jan. “Slave-​Girl Lost and Regained: Transformation of a Story.” Marvels and Tales: Journal of Fairy-​Tale Studies 18, no. 2 (2004): 201–​217.

4 Visibility and Performance Courtesans in the Early Islamicate Courts (661–​9 50 ce)

Lisa Nielson

i  The use of slave women as court musicians and their various roles as entertainers, intimate companions, and symbols of status for visiting dignitaries has ancient roots in the Near East and Arabia.1 Because they were familiar symbols of power and ritual, it is not surprising that women musicians were incorporated into the new court culture of early and medieval Islamic states. Referred to in Arabic as “singing girls” (sing., qayna; pl., qiyan), women musicians are referenced in pre-​Islamic poetry and were used in the Umayyad court (661–​750 ce); however, it was during the Abbasid era (750–​1258 ce) that they became essential for court entertainment. Beginning in the eighth century, the demand for singing girls led quickly to the foundation of music centers, specialized trade in musical concubines, and the development of a complex hierarchy among court musicians organized around the intersection of musical prowess, extramusical performance, and gender. Although the textual record reflects their influence on the development of art music, not to mention providing inspiration for poetry and literature, all the sources concerned with singing girls were written by men. Although the sources mention women writing and publishing books of songs and poetry, none of these books are extant. This loss does not mean their voices have been silenced; rather, sources for women’s biographies, poetry collections, and songs attributed to women are filtered through male-​authored literary works. Despite these cautions, representations of singing girls, regardless of presumed accuracy, provide insight into music performance practices, and acceptable and 75

76 i

Lisa Nielson

unacceptable social practices, and they offer a glimpse into the place of singing girls in early and medieval Islamicate court culture.2 One common thread is that there was an elite class of slave courtesans that was trained carefully in music, poetry, and courtly etiquette, and commanded astronomical prices on the market. Because of the exclusivity of their trade, their ownership and patronage was available primarily only to the wealthiest male patrons—​the caliphs and Abbasid family members, urban notables, and wealthy status-​seekers from the merchant classes. By the ninth century, it appears, a courtesan/​patron dynamic begins to emerge. Qiyan came to represent the highest level of courtesanship, with jawari (female slaves) and mughanniyat (female singers) representing a broader category of musical concubines. As such, not only were qiyan able to perform at the most exclusive private events and at court but also their status offered the possibility for social mobility and a degree of agency for talented slave women. Elite women musicians played a key part in early Islamic courts (661–​950 ce). Although their status as slaves and concubines is important in understanding their function in social and cultural terms, it is also vital to locate the singers and performers in the broader historical narrative surrounding the intersection of music, gender, and servitude.3 Because free women of the aristocratic and merchant classes were bound increasingly by rules regulating their visibility as social actors and maintaining propriety, singing girls are represented in the literature transgressing social controls regularly. Singing girls performed in public for mixed company and had relative freedom of movement. Both their visibility and profession meant they lived on the edges of what was deemed proper and improper, and were given considerable social leeway. It is this slippage between legal and social boundaries that enabled women musicians to become active agents of culture, and suggests comparison with women, particularly slave women, who held similar roles in other cultural contexts. References to singing girls can be found in a variety of sources, although relatively few focus specifically on singing girls. The quintessential resource for music and musicians in the medieval Islamic courts is the Kitab al-​Aghani (Great Book of Songs) by the tenth-​ century courtier Abu al-​Faraj al-​Isbahani (c. 897–​969 ce). According to the sources, al-​Isbahani was exceptionally well educated and witty, and amassed a quantity of knowledge throughout his lifetime that impressed even the literary elite.4 He was apparently an eccentric character, given to sloppy dress, questionable hygiene, and poor table manners. His work, however, is far from sloppy. Beginning with the earliest known musicians, the Aghani presents a detailed and often deeply entertaining account of court life, but also provides essential information about music performance, suggested authorship of songs, and debates within the music community regarding compositional style.5 In addition to al-​Isbahani, the Risalat al-​Qiyan (The Epistle of the Singing Girls) by the great satirist al-​Jahiz (776/​7–​868/​9 ce) provides a glimpse at the patronage of singing girls. Although social commentary couched in satire, his vivid descriptions of how singing girls were selected and trained, as well as the depth of their skill, are among the most



Visibility and Performance

j 77

detailed extant. This is not to say we should regard them as accurate or even factual. Al-​ Jahiz was a belle-​lettrist, not a historian or chronicler, and where he gave praise he would often temper it with caution, harsh criticism, philosophy, or clever misdirection.6 The Muwashsha (The Brocaded Book) by Ibn al-​Washsha (d. 936 ce) details the social performance of singing girls in a lengthy chapter dedicated solely to the dangers of patronizing qiyan. He follows this chapter with another on the malady of passion for emphasis.7 Ibn al-​Washsha was a grammarian and scholar who wrote the Muwashsha as an etiquette manual for courtiers and gentlemen to help them acquire necessary skills of refinement. Although the trappings of refinement included proper dress and good manners, Ibn al-​Washsha placed particular emphasis on the ability to write and communicate appropriately. Refinement was also not limited to men. Ibn al-​Washsha includes advice to refined woman regarding dress, ornament, and appropriate modes of expression in mixed company. The Muwashsha, however, is more than a fussy list of manners. Rather, it conveys a philosophy of life based on good behavior, appropriate friendships, and chaste, noble love. Other useful sources for information about qiyan include the Fihrist of Ibn al-​Nadim (935–​990/​1 ce), narratives involving singing girls in the Abbasid histories of al-​Tabari (839–​923 ce) and al-​Masudi (d. 928 ce), anecdotes regarding famous singing girls contained in the works of al-​Tanukhi (940–​994 ce), and legal arguments related to music and listening as reflected in the treatises of the religious scholars Ibn Abi al-​Dunya (823–​ 894 ce) and al-​Adjurri (d. 970 ce). Early Islamic Musical Culture Music was part of Arabian culture before Islam, although it was subordinate to the composition and recitation of poetry.8 During the Umayyad (661–​750 ce) and Abbasid (750–​1258 ce) periods, although poetry continued to represent the highest artistic standards, song soon became a vital aspect of court culture as well. The increased importance of music was a result of several factors, the most fundamental related to trade and the integration of different musical traditions with extant practices. The expansion of the Arab/​Islamic Empire (seventh–​ninth centuries ce) was accomplished through a combination of conquest, annexation, and diplomatic exchange. New urban centers—​notably, the cities of Baghdad and Samarra—​were established, and the caliphate fashioned a lavish court culture to reflect its status as an emerging world power and to provide an appropriate theater for diplomacy. Territorial exchanges and negotiation over trade routes facilitated the relocation of skilled craftspeople and scholars to the growing urban centers of the Islamic territories. Although some were slaves or their services were acquired as part of treaty and trade negotiations, others were invited to the courts. The establishment of an increasingly cosmopolitan court not only brought intellectual and artistic traditions into the urban centers but also created a market for luxury goods, including skilled slaves. Although the Quran emphasized reasonable treatment of

78 i

Lisa Nielson

slaves and manumission, it did not specifically abolish the slave trade. As a result, slavery remained virtually unchanged with the transition to Islam and the expansion of trade networks across the new imperial realm. The growth of the Arab/​Islamic court and expanded merchant wealth increased demand for slave labor (skilled and unskilled) at all levels of society, including the rural economy, urban households, and the military. During the Abbasid era, rulers also began to delegate day-​to-​day running of the Empire to clients (sing., mawla; pl., mawali) and slaves. The use of eunuchs, slaves, and free men was another borrowed practice, and they also became regular functionaries in the palace bureaucracy. Women were acquired solely for domestic services, which encompassed tasks as varying as housekeeping and entertainment. Regardless of their function in the household, all slave women were potentially available sexually. Those women who were also skilled in music or poetry were prized especially as concubines, and demand drove the prices up. By the ninth century ce, highly trained singing girls constituted a highly lucrative slave commodity. As trade and supportive economic policies brought a diverse body of culture into the growing urban centers of the early Islamic states, new instruments, foreign musicians, new styles of singing, and different melodic and rhythmic modes were integrated with existing traditions. Before Islam, there was an extant tradition of accompanied recitation and song that required a class of music professionals, so there was already a pool of skilled performers from which to staff—​and train—​the ranks of court musicians. Although men also worked as musicians, the majority of these professionals were slave women. Women musicians were employed in taverns, brothels, private homes, and in the courts. References to the legendary singing slave girls, the Jaradatan or “Two Grasshoppers,” and the laments of women poets and singers into the last years of the pre-​Islamic era demonstrate that women worked as musicians and professional mourners in Arabia before the advent of Islam.9 In addition to singing girls, pre-​Islamic and early Islamic courts employed free men who were often, but not always, of foreign origin, and cross-​gendered singers called mukhannathun.10 Beginning in the early eighth century ce, there are references to each class of musician performing in the Umayyad court, although singing girls, male musicians, and mukhannathun were also found in pre-​Islamic musical culture.11 By the ninth century ce, the translation of Greek, Indian, Persian, Jewish, and other non-​Arabic texts inspired the formulation of an Arabo-​Islamic music theory comprising new and extant theories of music. The subsequent collaboration and integration of diverse styles, worked out in memory and embodied through performance, led to the development of Islamicate art music.12 Qiyan, Slavery, and Courtesanship When attempting to describe historical slave systems, finding an adequate vocabulary for different gradations of servitude can be complex. Arabic terminology for slave status ranges in meaning from concrete to ambiguous, and, as with other slave cultures,



Visibility and Performance

j 79

embedded in such labels are often implied references to gender, function, ethnicity, and social status. In addition, the meanings of most terms were fluid or highly localized. Such fluidity in degrees of servitude is particularly relevant to the intersection of gender and slavery. Not only is the task of finding an adequate translation for the roles of slave women difficult but also many of the terms originating in the West that are used to designate women’s relationship to men are invested with moral and social implications that can be anachronistic or distort the original meaning. By way of example, the legal designation of “wife” and labels relating to religious orders (nuns, novices, and so forth) are read as morally acceptable, whereas terms that denote nonconjugal relationships, such as mistress, concubine, courtesan, and prostitute, are problematic when applied to non-​Western cultures. Although these roles can be related or are similar to the Western understanding, they are often lumped together and perceived as morally questionable, regardless of whether they were sanctioned by law or religion within their own context. In addition, there are important differences in social class, remuneration, clientele, and recruitment among these categories. Therefore, if a word with too much Western cultural baggage is chosen, the reader (and sometimes the author) will focus on the sexual associations, which may only represent one aspect of the role.13 These issues are all relevant to parsing references to female slaves in sources produced by the cosmopolitan urban culture of the early Abbasid era. They are complicated further by the fact that our current understanding of the Arabic terms may not reflect their meaning. For example, the common Arabic term for a female slave and concubine is jariya (pl., jawari), which carries meanings related to movement or flow. Jariya can denote a slave who is also a musical concubine or entertainer (dancer, poet), or a musical concubine might also be called mughanniya (pl., mughaniyat), or singer. Mughanniyun were not always slaves, or women, so this term could be transitory and descriptive of role only.14 Because all female slaves were potentially available sexually, jariya was likewise not just a referent to concubinage or slavery, but to a slave’s social status and function.15 Qiyan were also slave women, although the label included all skilled women musicians, including free women and the nobility.16 To complicate matters further, the sources often use jawari, qiyan, and mughanniyat interchangeably to indicate “women musicians” or “musical concubines.”17 Jawari and mughanniyat were praised for their skill or noted for their participation in a specific event, but qiyan were apparently a class apart. They were valued beyond their capacity as sexual or musical diversions, and their services could command vast sums of money. Qiyan might be attached to a specific owner or court, but could also live independently.18 Free women who were skilled musicians might also be referred to as qiyan, but were careful to limit their performances to private parties for relatives and to maintain a strictly amateur status to avoid impropriety. One example was the caliph Harun al-​ Rashid’s half-​sister, Ulayyah bint al-​Mahdi (d.c. 825 ce), who was a notable musician and referred to as a qayna. Despite her prowess, she was careful to perform for family only or

80 i

Lisa Nielson

in private, chaperoned functions because of her position as a relative to the caliph and free woman of rank.19 Because many slave—​and free—​women received a certain amount of rudimentary musical training, can a distinction be made among jawari, qiyan, and mughanniyat? Given the usage of these terms and their relationship to descriptions of music performance in the sources, they are not only connected to status and function, but are also indicative of musical skill and accomplishment. Furthermore, they act as a literary reference point for the overall level of virtuosity—​social, musical, or conversational—​within a given situation. In other words, although all three terms were used interchangeably to reference “musical concubine” or woman musician, the key difference may in fact relate to musicianship and, by extension, courtesanship. Although the term “courtesan” has a Western etymology and is not without limitation, its origin as the feminine for “courtier” or “woman of the court” encompasses the function, performance, and physical location of qiyan.20 Establishing a definition for the role and function of courtesans is difficult. Similarities in their training, relationship with the patronate, and representation in the literature can be found cross-​culturally and throughout history. Fundamentally, a courtesan is an unmarried woman, free or unfree, who is viewed primarily as a musician, poet, companion, and entertainer, and is selected and/​ or trained specifically for this purpose. She is visible and expensive, but can choose her clientele. Unlike prostitutes, who have no choice and are paid directly for sex acts, courtesans are compensated indirectly in the form of gifts and favors. Courtesans also have a certain amount of social—​and legal—​freedom, thus choosing to live independently if they have sufficient wealth, or to live in communities/​houses controlled by women. Sexual availability, or the illusion of intimacy, is her secondary function. Courtesans are visible throughout the historical and textual record, often as artistic subjects, but just as often as authors or having works attributed to them. The music and poetry of courtesans often focuses on themes of profane love, frequently co-​opting genres reserved for men.21 In those cultures in which courtesans exist, their artistic presence has had considerable influence on fashion and style, and has contributed to the development of art, music, and literature.22 Because of her dual role of artist and companion, the courtesan treats her body as integral to her performance as much as her instrument or voice. The extrasensory performance of seduction incorporates those signifiers of desire that her culture holds, such as scanty or seductive dress, use of certain fabrics, jewelry, dance, gesture, scents, and trappings of comfort in her home. Seduction also requires a keen ability to read people, so a successful courtesan had to be adept at evaluating and manipulating the emotions and moods of her patrons. The most successful courtesans were not always the most beautiful or the most accomplished; frequently, they were the most intelligent, astute, or ruthless. Therefore, although a courtesan’s performance uses all the standard tropes of a public performance because she often is in the position of actually performing, she must also meet the audience’s expectations of her embodied performance and implied intimacy.



Visibility and Performance

j 81

Because of their complicated social position, courtesans often become symbolic figures in literary arguments on social, religious, and cultural concerns. Not unexpectedly, their supporters laud their skill, beauty, and wit, whereas their detractors point to their use of music, poetry, and seduction to distract men from virtuous living. In many cases, it is not the courtesan at fault, per se, but rather the intimacy of her performance and the subject matter of her music. It is this discomfort that is central to the Risalat al-​ Qiyan and even more evident in Ibn al-​Washsha’s chapters concerned with singing girls in the Muwashsha. Although Ibn al-​Washsha focuses primarily on the social role of singing girls and what he sees as the consequences of their patronage, al-​Jahiz uses the figure of the singing girl on several symbolic levels.23 He represents her as frivolous temptation, skilled musician, a soulless wanton, and, briefly, a pitiable object who, despite her skills at manipulation, is subject to the whims of her patrons. Although his ostensible purpose is a defense of singing girls and their patronage, al-​Jahiz maneuvers skillfully through arguments used by patrons and detractors alike as the essay unfolds. In so doing, he points to fallacies on both sides, ending with criticism of the trade and posturing of status-​seeking patrons. The differences between jawari, mughanniyat, and qiyan become more distinct when understood in the context of references to music performance. Although musicianship was based on aptitude as well as talent, there were additional factors available only to women. These factors might include luck, intelligence, beauty, good patrons, or deft manipulation, but they all related to the woman herself—​or perceptions thereof—​using the system for personal gain. As a result, women musicians at the highest levels of the music hierarchy apparently transgressed boundaries between slave/​slaver, creating a liminal social and legal space between free and unfree. This ability to use their musical role and expectations surrounding that role provided singing slave girls with an unusual degree of agency in the midst of a confluence of patriarchal cultures, transforming a jariya (slave woman/​concubine) into the linguistically slippery qayna. Therefore, the ambiguity of qayna is closest to other terms used to designate courtesanship, rather than concubinage or servitude alone. Like the Chinese ji (singing girl), the Japanese geisha (arts person), or Greek hetaera (comrade), qayna references the education and artistic expression of courtesanship first, juxtaposed with sexual/​emotional intimacy.24 Becoming a Qayna In the Risalat al-​Qiyan, al-​Jahiz makes it clear that not all women were considered capable of becoming qiyan.25 He outlines their training and patronage, suggesting that those at the top of the profession used every imaginable skill to get there. Likewise, stories about singing girls pursuing wealthy patrons, manipulating their own sales to the highest bidders, and using their artistry (or beauty) as a bargaining tool for advancement can be found throughout ninth- and tenth-​century Arabic sources. Such stories, however, are often

82 i

Lisa Nielson

stylized, adhere to specific narrative patterns, and provide the sort of hyperbole considered appropriate to contemporary literary style. Caution is certainly required in reading them as factual depictions. That said, although it is not possible to ascertain the truth behind such representations, the fact there are so many such anecdotes that share themes of singing girls using their musical prowess and beauty for gain is suggestive. What is clear is that, regardless how an individual author might feel about them, singing girls were relentlessly visible as musical, social, and literary figures, and their ability to move through the social hierarchy was understood by author and audience. The primary function of singing girls was to provide musical entertainment, and women trained as singing girls were taught to sing and play musical instruments. Training singing girls was expensive, so there was a considerable amount of upfront speculation based on talent, beauty, and market preferences. Their main instrument was the oud (lute), although they also played zithers and flutes and recited poetry. Memorization of poetic forms and a huge repertoire of extant songs were essential parts of their education, and the most successful were said to have thousands of songs stored in their memory. Singing girls also composed and wrote books of songs, although, again, none of these texts appear to have survived.26 To emphasize their change in status, slaves were given new, Arabic names when captured or acquired—​a common practice in all slave-​owning societies. Names for slave women were often descriptive either of a physical feature or some perceived value by their owner, or acted as simple declarations of beauty. Among such names were Jamila (Beautiful), Dhat al-​Khol (The Girl with the Beauty Mark), Dananir (Shower of Gold), Mahbuba (Beloved), and Bas-​Bas (Caress). Names might also be appended to that of their origin or owner, such as Arib al-​Mamuniyah.27 The use of descriptive or euphemistic names for courtesans can be found in other cultures, such as the Greek Myrtale to reference hetaerae or names related to flowers and willows for Chinese and Japanese courtesans.28 In addition to music, singing girls were trained in Arabic, recitation, and court etiquette. Some of the sources imply that, in addition to music and reading, they were also tutored in what are best termed the seductive arts, including the language of scents, flowers, sexual techniques, and emotional manipulation.29 Their ability to ensnare men is a common thread in stories about singing girls, although opinions are divided with regard to whether such entrapment is good or bad. Al-​Jahiz makes much, in the Risalah, of the intentional manipulation of male patrons by rapacious singing girls. On the one hand, he acknowledges their musical skill, yet faults their education for leaving them bereft of morals. In contrast, Ibn al-​Washsha glosses over their education and dwells primarily on their ability to manipulate their lovers either for financial gain or, more ominously, out of malice. To lend further authority to their claims, both authors include anecdotes and comments by famous singing girls, such as al-​Jahiz’s mention of Sallamah al-​Qass (fl. seventh century ce) and Ibn al-​Washsha’s inclusion of several poems and stories about the famous singing girl Fadl (d. 871 ce).30 Again, one must be cautious with regard to the



Visibility and Performance

j 83

veracity of this material given the perspective of author, audience, and writing conventions of the time. The consistency of representation, however, again serves to underscore that singing girls fulfilled the social role of courtesans. References to the training of singing girls are scattered across diverse biographies of famous women and accounts of notable performances. It seems that those girls and women who had the potential to become singing girls were typically sent to music schools to receive further training. Such schools were not brick-​and-​mortar institutions, but rather constituted the instruction of a specific master and training in his repertoire. According to al-​Isbahani, one of the earliest “schools” was established by the singing girl Jamila, whose students included the mukhannath Tuways (d.c. 721 ce).31 Al-​Isbahani attributes the establishment of the first music school for training singing girls to the famous musician Ibrahim al-​Mawsili (d. 804 ce). Al-​Mawsili apparently wanted to ensure that singing girls received a thorough education as well as to establish standards for music training. He also made an excellent profit by selling them.32 There also seems to have been “in-​house” training provided by the traders themselves. Traders probably had teachers of various disciplines on hand to teach the girls and women what they needed to know. In addition, the likelihood that trainees had some sexual training early on is also suggested in several sources, but no specifics are given.33 Once trained, the most exceptional women were sold to wealthy patrons or the court. If the figures in the Aghani are even remotely accurate, ownership of slave girls was prohibitively expensive for all except the caliph and the very wealthy. Prices are cited as high as 100,000 dirhams, and the initial purchase price was just the beginning of the owner’s expenses. Not unlike geisha, singing girls needed to keep up with their training by learning new songs and continued practice. Along with needing to be given time to practice and study, singing girls took lessons with a music master and required regular supplies of lute strings, plectra, and writing implements.34 In addition, a prized singing girl had the expectation of receiving frequent gifts in the form of expensive clothes, jewelry, favored foods, and scents.35 All additional expenses had to be underwritten by their owner. Some of the cost could be defrayed by loaning her out or, more crudely, pimping her to friends for sexual and/​or musical services. Given that Islamic law prohibits the sale of one’s slaves into prostitution, money did not change hands when a patron offered the services of his singing girl to friends or had her entertain at an event.36 Rather, payment was made in the form of gifts or obligations to the host.37 Although couched in satire, al-​Jahiz confronts this means of getting around the law with evident sarcasm: Furthermore, people send along to her owner’s house presents of all sorts in the way of food and drink, but if they come to visit, they just get a sight [of the girl] and go away frustrated, while her master reaps the fruit of what they have sown, so that he, not they, has the enjoyment of it and is amply provided against the expense of maintaining his [other] slave girls.38

84 i

Lisa Nielson

For those who could not afford to own a singing girl but wanted access to their services, there were public places where singing girls could be heard for a price. Much like the brothels, teahouses, and theaters of the Floating World of Japan and salons of eighteenth-​ and nineteenth-​century Europe and India, a patron visited singing girls to be entertained by music, conversation, and the promise of intimacy.39 It is unclear whether these houses were for the purpose of music training or actual brothels run by flesh merchants, or something of both. Nor were such places owned only by men. Singing girls with enough wealth to buy their freedom sometimes entertained visitors at home, offering their patrons the services of their trainees. Singing girls who spent their lives working in these houses were musical concubines, referred to in the sources as jawari or mughanniyat rather than qiyan. Why they remained rather than being sold or traded is uncertain, although one possibility is that they may have lacked the right mix of talent and beauty (or ruthlessness) to attract a higher bidder.40 Another possibility is that entertaining the men who visited may have been an aspect of their training, and these unknown women did move on to other patrons. Some known singing girls had relatively humble beginnings in schools or entertainment houses.41 The primary patrons of entertainment houses were probably men who could not afford to own singing girls of their own. Even visiting lower level singing girls must have been expensive, however, especially when the cost of food, drink, and gifts was factored into the price of an evening.42 Al-​Tanukhi’s memoirs are filled with anecdotes about men “paying the fees of singing girls” at places of entertainment and individual houses owned by singing girls. His anecdotes depict young trainees being offered to or selected by the patron, and provide detail about additional fees for refreshments and gifts. The compounding of costs could be ruinous. One story tells of a man who decides to buy a singing girl for the price of 1,000 dinars so he no longer needs to pay to visit them, thinking he’ll save money by having his own. In another story, al-​Tanukhi tells about a lengthy party held at a brothel, where the procurer provides an assortment of slave girls, referred to as jawari, for a party of travelers. Several of the women are noted as being singers, and the party apparently carried on for some days.43 Qiyan and Musicianship Although musicianship is defined differently in every cultural context, one common principal is that there is a distinction between technical mastery of an instrument and comprehensive mastery of the broader discipline of music. Technical ability is only one component, because an amateur or nonprofessional might also be an expert instrumentalist or singer. A thorough grounding in music theory, deep understanding of the history of the repertoire, and the ability to interpret familiar and new music flawlessly are essential aspects of musicianship because they enhance technical performance. Therefore, although virtuosity is generally prized, actual performance is only one aspect of mastery, and, in some historical contexts, is considered less important than intellectual mastery.



Visibility and Performance

j 85

Because of these distinctions, the definition of what it meant to be a “musician” and “performer” in the medieval Islamicate courts shifted along with the development of the theoretical structure of, and discourse surrounding, the musical systems that arose from the synthesis of local and new musical systems. Within this discourse, a performing musician, such as a musical concubine or low-​ranking male instrumentalist, was viewed initially as a technician. Not all Arab music theorists held this view, however, and practical musicianship came to be valued as much as theoretical scholarship in some circles.44 By the eighth and into the ninth century, the standards for musicianship at the Abbasid court became more complex, integrating high expectations for technical performance with the need for poetic, theoretical, and social acumen.45 The influx of skilled musicians, new theories of music, and increasing reliance on musical entertainment at court prompted the need for revised standards. At minimum, a court musician was expected to be able to improvise as well as have a vast repertoire of songs committed to memory. All members of the court, including prominent musicians, were also expected to be groomed, elegantly attired, and well versed in courtly behavior.46 Music centers grew up around established artists and traditions, and different schools of musical thought came to be associated with particular musicians. Al-​Isbahani’s Aghani demonstrates amply the tension among different groups, particularly between the adherents to the older or “ancient” musical style and proponents of the new.47 Understanding these different musical styles and making an accurate assessment of the patron’s leanings became an important aspect of musical training. Patrons took sides and encouraged competition among musicians. As a result, singing girls performed and frequently competed for patronage and gifts with other singing girls as well as male and cross-​gendered musicians.48 Despite the high standards for musicianship and the importance of music at court, however, to be a musician remained a dubious profession, and performing musicians continued to be perceived as having lesser social status.49 Singing girls were affected by this (double) standard with men, but had the ability to use their musicianship as a mode of agency and transcend effectively their status as women, slaves, and concubines.50 Although many prominent musicians at the court were men, based on the accounts of their technical and compositional prowess, it is evident that qiyan competed with men to maintain their status within the hierarchy of court musicians. Al-​Isbahani includes many anecdotes about the famous singer Arib al-​ Mamuniyah (d.c. 890 ce) that extol her virtuosity, intelligence, and beauty. In one such passage he comments: Arib was a fine musician and good poet. She had a pleasing hand and literary style. In addition, she was extremely beautiful, accomplished and refined, and endowed with an attractive figure. She was an excellent player of the oud, a very skilled composer, and an expert on tunes and melodic modes. She also had a vast knowledge of poetry and the different branches of culture.51

86 i

Lisa Nielson

By all accounts, Arib’s legacy as a musician was wide reaching, despite her exploits with lovers and keen rivalries with fellow musicians.52 She used her virtuosity as poet and musician, in combination with looks, political acumen, and a degree of ruthlessness, to access the highest levels of patronage and, therefore, as tools of agency. Another singing girl whose musicianship was the source of advancement was Dhat al-​ Khol. A skilled singer, she was owned by a slave trader before catching the eye of Ibrahim al-​Mawsili. He praised her skills to Harun al-​Rashid, who bought her for 70,000 dirhams. Despite his infatuation, when al-​Rashid interrogated her about the nature of her relationship with al-​Mawsili, he became incensed when she admitted it was intimate. Al-​ Rashid then gave her to his slave, Hammawayh, in a fit of pique. The appreciative slave treated her well, but eventually al-​Rashid came to miss her songs and took her back. Dhat al-​Khol asked that Hammawayh be given a high office for his treatment of her, and al-​ Rashid granted her request. Dhat al-​Khol was reinstated at court, and there are several colorful stories of her exploits. One relates how, in a fit of rage over al-​Rashid’s being distracted by another woman while on his way to see her, she cut off (or threatened to) the beauty mark that al-​Rashid found so charming.53 The ability of singing girls to use music as a form of agency can be linked not only to their visibility as performers but also to their performance spaces. Musical events were organized according to the patronage system, with the patron outlining the subject matter, poetic forms, codes of etiquette, and even physical appearance of the performance space. Such gatherings, the majlis (pl., majalis), were held by the court and wealthy private citizens.54 The format of a majlis was highly ritualized and organized, with standing conventions of behavior, appearance, and subject matter. In addition, the visual element was an important aspect of the performance, and a great amount of effort and expense was made to set the proper environment for the musical event.55 Such parties could last for days, and could be either completely spontaneous or require painstaking planning. As during the pre-​Islamic era, beautiful vocal expression was considered the highest, most pure form of music. Musicians generally sang unaccompanied and improvised melodies based on the body of poetry or improvised poetic themes chosen for the majlis. The ability to sing murtajilan, or extemporaneous solo, was the highest standard for a musician.56 Instrumental accompaniment of song was the next level of required skill, with pure instrumental music the third. Depending on the framework, musicians, regardless of gender, had a certain amount of creative control based on their reputation. To step outside the boundaries of the event, however, all aspects of their performance—​music, poetry, and presentation—​needed to be particularly compelling or virtuosic. If a performance met or exceeded the expectations of the patron, the musician could reap considerable monetary rewards and the assurance of continued patronage. If not, they could, at best, be ridiculed or kicked out—​at worst, beaten or sold as slaves. There are many anecdotes about singers having their memory tested, being tricked, or sent away in shame when unable to perform a requested song.57



Visibility and Performance

j 87

Because of the emphasis on improvisation, the majority of musical training was aural; it centered on the internalization of a broad corpus of poetry, melodies, melodic modes (maqam; pl., maqamat), rhythmic modes (iqa), and ornamentation. Having a keen memory was essential to success. To have basic proficiency in improvisation, a musician needed to memorize not only a great number of melodic and rhythmic modes and their possible combinations but also the set repertoire of old and current songs, poems, and the formulaic tools required for composition.58 As poetry and language remained the most esteemed forms of artistic expression, any music accompaniment was in service to the text. Purely instrumental music was used as preludes or postludes to the performance of song cycles or poems. In instrumental performance, the preference was for small groups or a single instrument, such as an oud or other plucked instrument, or a woodwind. A group might contain one or two oud players and possibly a bowed instrument, a woodwind such as a flute, qanun (a large table zither with movable bridges), and percussion. Large ensembles were rare. Although there are accounts of large choruses, they were not common and apparently performed in unison.59 When large groups are mentioned, they are clearly special occasions, as in the visit of a foreign dignitary, celebration, or a particularly lavish party. The Influence of Singing Girls on Early Islamic Musical Culture Although it is impossible to ascertain whether singing girls contributed a specific scale, genre, or style, their ubiquity in a diverse range of sources certainly suggests an overall contribution to the development of Abbasid-​era art music and musical culture. In historical works and early music treatises, women musicians are recognized as the first, most important teachers of traditional songs and new compositions, and many male musicians traced their lineage back to famous singing girls. The institution of specific musical training geared toward the market in singing girls provided advanced training to those women slaves designated to become singing girls. Training in different musical idioms also enabled women musicians to gain stronger facility with improvisation and familiarity with the current repertoire, giving them a wider array of compositional options and making them even more competitive at court. Despite their acknowledged skill as musicians, however, singing girls existed in a morally gray area because of their standing as courtesans. By the time al-​Jahiz wrote the Risalah, patronage of singing girls was at its highest point, and, not unexpectedly, singing girls became targets for social commentary about their morality and the morality of their patrons. Al-​ Jahiz summarizes this concern in his description of their musical and social repertoire: An accomplished singing girl has a repertoire of upwards of four thousand songs, each of them two to four verses long, so that the total amount of poetry contained in it, if one multiplies one figure by the other, comes to ten thousand verses, in

88 i

Lisa Nielson

which there is not one mention of God (except by inadvertence) or of the terrors of future punishment or the attractions of future reward. They are all founded on references to fornication, pimping, passion, yearning, desire and lust. Later on she continues to study her profession assiduously, learning from music teachers whose lessons are all flirting and whose directives are a seduction.60 Although it would appear from reading al-​Isbahani that the most enthusiastic patron of singing girls was the court, notables and merchants no doubt provided patronage to singers as well. From a conservative standpoint, if the caliph indulged in pursuits that were antithetical to religion, it was feared his actions could have apocalyptic consequences for the entire Muslim community. By the tenth century, singing girls and music—​more than wine, gambling, or similar amusements—​were singled out as catalysts for moral bankruptcy. Because music was accompanied frequently by wine, the two were, perhaps not unjustly, featured in many moral tales about the dangers of overindulgence. Concerns about the combination of wine, women, and song were not new, but they became stronger as the courtesan–​patron relationship developed. Although from different perspectives, al-​Jahiz and Ibn al-​Washsha suggest that men were in peril of being entrapped by the wiles of singing girls, and, more ominously, that singing girls exploited men’s weaknesses deliberately to entrap them. Despite growing religious censure of singing girls and music, however, the likely catalyst for such concerns was the ever-​increasing expenditure by the court on frivolous diversions, the majority of which were in opposition to established religious law and prudent social policy. Because of the association of singing girls with music, they quickly became symbols for music and listening in texts concerned with the legality of music in Islam. Legal questions related to music were less concerned with “good” or “bad” music in the artistic sense than the effect music had on the individual. The influence of listening to music was foundational to what is considered the first legal challenge to music by ninth-​century religious scholar Ibn Abi al-​Dunya.61 His influential treatise, the Dhamm al-​Malahi (Censure of Instruments of Diversion) uses qiyan symbolically to reference the illegalities of music listening.62 His followers did the same, and use of singing girls as symbolic referents to music became common in subsequent arguments regarding music.63 The resulting body of literature comprising pro-​and antilistening (sama) arguments grew into a specific genre of adab, commonly referred to as the sama “controversy.”64 Although the anti-​sama faction argued that music was a form of diversion and distracted one from God, the pro-​sama side countered that music not only had the potential for healing and inspiring creativity, but also, in the case of the Sufis, could lead to a closer relationship with God. Sama relates to the act of listening, or hearing, and a distinction was made between hearing music accidentally, which could be forgiven, and seeking out and listening to music actively.65 Listening was the focus as a result of the intimacy of the relationship between audience and performer, and the emotional effect of music. The performer may create the sound, but once created, it could find the ear of the unwary just



Visibility and Performance

j 89

as easily as the active listener. Being unable to control or channel sound was additionally troubling because sound had the ability to penetrate, and therefore influence, both the waking and the unconscious mind.66 The association of music with singing girls meant that music performance was also a gender identifier. Although men could—​and did—​perform as musicians, the traditional affiliation of music with women, and the general taint of music as a profession, could confuse gender boundaries for male performers.67 Just as musicianship could function as a form of agency for singing girls, it could work against men who sought music as a profession.68 Therefore, as the debate over music and listening expanded, it operated within a multivalent landscape of gender, servitude, and power. Conclusion Although the technique and artistry of singing girls were undoubtedly vital to the development of Islamicate art music, their social role was, arguably, even more so. Singing girls not only appear in the literature as musical or social performers, they act as literary signposts through their symbolic appearance in texts. Using references to real—​or imagined—​musicians, authors could not only discuss music and musicianship but also offer social commentary, using a familiar vocabulary of symbolic referents. In so doing, women musicians influenced technical developments as well as music discourse in their capacity as artists and symbols of culture. Throughout the narratives related to singing girls, however, the most fascinating element of their literary performance is reflected in their subversion of social norms. Although singing girls functioned within an apparently controlled and controlling environment, elite courtesans were able to usurp aspects of social power and occasionally shift the submissive role to the patron. In so doing, qiyan challenged regularly the relationship between slaver and slave, despite maintaining a legal and social position of servitude. This ability of singing girls to gain advantage over their patron and assume a dominant role was a source of considerable literary discomfort, and music was considered the focus of their power. Rather than examining gender, music, or slavery as discrete objects, this complication suggests each piece is a component of a broader category, and perhaps necessarily imprecise. Such ambiguity is reflected in the role of courtesans and elite slaves in other cultures, especially those individuals and artists who confused the relationship between owner and owned. To that end, comparisons among qiyan, elite slaves, and courtesans as social and literary figures in other cultures bears further study. Such study would serve several vital purposes. First, it would lend further insight into how slavery was defined, functioned, and justified in different contexts, and how slave artists contributed to cultural development. Second, although the lack of texts authored definitively by women and representing the voice of female slaves is a common lament in slave scholarship, there is still much to be gleaned from the extant texts, particularly the means women had to assert power

90 i

Lisa Nielson

and gain agency. Last, to gain a deeper understanding of the relationship of singing girls to the patronate, more analysis of the rhetoric of discomfort around their social, legal, and gender ambiguity is needed. By focusing on the confluence of their literary and social performance, we may gain additional insight into the intersection of slavery, gender, and the arts not only in the medieval Islamicate world but also in other premodern cultures as well. Notes 1.  The literature on the use of women musicians in the Ancient Near East and pre-​Islamic Arabia is diverse and spread over several disciplines. For an overview of the exchange of musicians in the Ancient Near East, see Franklin, “The Global Economy of Music in the Ancient Near East.” See Hoyland, Arabia and the Arabs, for an overview of cultural practices in pre-​Islamic Arabia, and Farmer, A History of Arabian Music. For a look at the representation of women entertainers in pre-​Islamic Arabia, see MacDonald, “Goddesses, Dancing Girls or Cheerleaders?” The use of women musicians as entertainers, composers of lamentations, and professional mourners is also attested in extant pre-​Islamic Arabic poetry and hadith. For lamentation and elegy, see Stetkevytch, Mute Immortals Speak. 2.  For a detailed analysis of the narratological structure of Abbasid texts concerned with women and gender, see Myrne, Narrative, Gender and Authority in Abbasid Literature on Women. The difficulty of using literary texts to reconstruct the lives of Abbasid-​era courtesans is also noted in Gordon, “Arib al-​Ma’muniya,” 86. 3. I purposely use the term “servitude” because it not only encompasses the different degrees of unfree status in the medieval Islamicate world but also reflects power and relationship dynamics. 4. For a thorough analysis of al-​Isbahani and the Kitab al-​Aghani, see Kilpatrick, Great Book of Songs. 5. Kilpatrick, Great Book of Songs. 6. On al-​Jahiz and the Risalat al-​Qiyan, see Cheikh-​Moussa, “La Negation d’eros ou le ‘ishq d’apres deux epitres d’al-​Gahiz,” 71–​119; and Gordon, “Yearning and Disquiet.” 7. For the chapters on singing girls see Ibn al-​Washsha, Book of Brocades, ­chapters 20, 21. For the Arabic edition, see Ibn al-​Washsha. Muwashsha. There is also an excellent French translation by Bouhlal, Le Livre de brocart. Subsequent references to the Muwashsha include both editions, indicating the French edition as (F) and Arabic as (A). 8. Pre-​Islamic poetry demonstrates the use of song and employment of singing girls. For an overview and discussion of early sources, see Farmer, A History of Arabian Music; and Shiloah, Music in the World of Islam. For a closer look at select, early sources, see Farmer, “Ibn Khurdadhbih on Musical Instruments”; Robson and Farmer, “The Kitab al-​Malahi of Abu Talib Al-​Mufaddal ibn Salama”; and the introduction to Ibn Abi al-​Dunya, Tracts on Listening to Music. For singing girls and their role in the transition from pre-​Islamic musical practices to the courts, see Nielson, “Gender and the Politics of Music.” 9. The Two Grasshoppers, also called the Two Crickets, were the singing girls of Abdullah ibn Ju’dan of the tribe of Ad. They are referred to frequently as the first singing girls. See Ibn al-​Rabbih, Music, 11. Al-​Isbahani mentions them as well: Kitab al-​Aghani, III:327–​331. See also Pellat, “Kayna.”



Visibility and Performance

j 91

10. For the mukhannathun, see Rowson, “The Effeminates of Early Medina.” For male musicians and clientage, see Kilpatrick, “Mawali and Music.” 11. Rowson, “The Effeminates of Early Medina.” See also Farmer, History of Arabian Music, 1–​19; and Shiloah, Music in the World of Islam, 1–​9. 12. The evolution of music styles and changes in certain songs are best seen in the Aghani. Although al-​Isbahani discusses music history and performance in detail, he is less interested in the technical specifics of a particular song than showing its development as it passes through different performative states. See Kilpatrick, Great Book of Songs; George Sawa, Music Performance Practice in the Early Abbasid Era. 13. For a legal perspective on this complication, see Ali, Marriage and Slavery, 23–​25. 14. Mughanni, and the masculine form mughanin, derive from ghina, to sing. 15. There are other Arabic words that designate concubinage or related nonconjugal relationships that derive from root words signifying ownership, secrecy, or desire. These include surriya and hathiya, which are also different from the terms for prostitute, such as baghiy. 16. For the origins of qayna, see Beeston’s Introduction to his translation of al-​Jahiz, Epistle on the Singing Girls, 2.  See also Poche, “Music in Ancient Arabia from Archaeological and Written Sources”; and Caswell, Slave Girls of Baghdad, 2. In the Muwashsha, Ibn al-​Washsha uses taqayyana—​meaning, perfect or skilled—​as a referent to singing girls, although it could also perhaps be a play on the penchant of singing girls to play with their male patrons. For a discussion of Ibn al-​Washsha’s usage, see Pellat, “Kayna,” 4: 820–​824. 17. Al-Isbahani, for example, refers to singing girls interchangeably as qiyan and jawari. Ibn al-​Washsha does the same, although he is careful to use qiyan in the title of his chapter dedicated to the dangers of singing girls. Not surprisingly, qiyan, jawari, and mughanni are also referenced interchangeably in anti-​sama literature, although there is an implied distinction among the three. 18. Independence could be physical, such as living separately from their patron, or social, as in their ability to move between the women’s quarters and men’s spaces with relative freedom. 19. Caswell offers a chapter on free women who were qiyan and dedicates a section to Ulayyah bint al-​Mahdi. See Slave Girls of Baghdad, 191–​209. The section on Ulayyah bint al-​Mahdi begins on p.  197. See also Gordon, “Place of Competition.” For an overview of women musicians in the courts, including noblewomen, see Meyers-​Sawa, “Role of Women in Musical Life,” and “Historical Issues of Gender and Music.” See also al-​Heitty, “Contrasting Spheres of Free Women and Jawari.” 20. Although not unproblematic, “courtesan” is apt when applied to women who use art as well as companionship to secure male patronage outside of marriage. Courtisane, the feminine of courtier, first appeared in 1549 from the French as courtisane and in Italian as cortigiana (prostitute). Literally, “woman of the court,” feminine of cortigiano (one attached to a court), it derives from corte (court) from the Latin, cortem. See “Courtesan,” Dictionary.com, Online Etymology Dictionary, Douglas Harper, Historian, http://​dictionary.reference.com/​browse/​Courtesan (accessed June 28, 2013). 21. As with musical instruments and song genres, poetry forms might also be gendered. For example, sixteenth-​century Venetian courtesans wrote poems in ottava rima, and singing girls might compose epic poetry (pl., qasa’id, sing., qasida ); both forms usually the province of men. 22. For an outline of the characteristics of courtesanship, see Feldman and Gordon, Courtesan’s Arts, 1–​16. Courtesans are ubiquitous worldwide. They are featured in Mughal art, travel books from sixteenth-​century Venice and Rome, Chinese art spanning numerous dynasties, and the

92 i

Lisa Nielson

Ukiyo-​e (or prints from the Floating World) of Japan. In addition to works known to be written by courtesans, such as the poetry of the sixteenth-​century Venetian courtesan Veronica Franco, there are many stories and novels about them. Some of the best known are the stories of the Ming Dynasty writer Feng Meng Long, the novel Umrao Jan Ada by Mirza Ruswa, and Camille by Alexander Dumas (fils). 23. Ibn al-​Washsha begins his commentary on singing girls by stating: “The secret (wicked) intentions of singing girls are demonstrable at a glance. If one sees a young man at a gathering (majlis) with wealth and great fortune, of good cheer and handsome, she would undertake to use her (feminine) charm so that she might draw him to her. In so doing, she breaks him to her will.” See Muwashsha, 136 (F), 134 (A). 24. For studies of the varying roles for court women and concubines cross-​culturally and in history, see Walthall, Servants of the Dynasty. See also the essays in Gavin, Women in the Medieval Islamic World. For courtesanship in India, Japan, China, Korea, and Europe, see the essays in Feldman and Gordon, Courtesan’s Arts. For China, see also Lam, “Reading Music and Eroticism in Late Ming Texts”; and for Japan, Downer, Women of the Pleasure Quarters. Research into courtesanship in North and South India likewise suggests intriguing parallels between qiyan and Indian courtesans. For South India, see Orr, Donors, Devotees and Daughters of God; and Soneji, Unfinished Gestures. For North India, see Qureshi, “How Does Music Mean?” ; and her essay “Female Agency and Patrilineal Constraints: Situating Courtesans in Twentieth-​Century India,” 312–​331. 25. Al-Jahiz is perhaps the most overt in marking the differences in function among women slave entertainers. He references musicians as a group most often as mughanni, but when inserting court entertainers into his satirical work, makes a conscious choice in labels. For example, The Kitab Mufakharat al-jawari wa-’l-ghilman and the Risalah al-​Qiyan both feature musical concubines, but the status for each is quite different. The rough translation of the title for the former is The Book of Debate between the Concubines and Catamites, using jawari for slave women. The focus of this funny and raunchy essay is a debate between girls and boys over their sexual charms. A partial English translation (censored carefully) is available in al-​Jahiz, Nine Essays of al-​Jahiz. In contrast, al-​Jahiz’s Epistle on the Singing Girls is focused specifically on the trade and patronage of top courtesans, and uses qiyan throughout. 26. Al-​Isbahani and al-​Nadim both list books of songs by singing girls. See al-​Nadim, The Fihrist of Ibn al-​Nadim. See also Sawa, Music Performance Practice in the Early Abbasid Era; and Kilpatrick, Great Book of Songs. 27. See Caswell, Slave Girls of Baghdad, 18. See also Gordon, “Arib al-​Ma’muniyah,” 87. 28. Myrtale, or myrtle berry, was also a reference to female genitalia. I am grateful to Dr. John Franklin for pointing out this reference. Tayu, orian, geisha, and ji likewise had evocative stage names, indicating their role as artists and companions. See Zeitlin, “ ‘Notes of Flesh,’ ” 75–​99; and Downer, Women of the Pleasure Quarters. 29. See al-​Jahiz, Risalah on the Singing Girls, 31–​33. 30. Al-Jahiz, Risalah on the Singing Girls, 23; Ibn al-​Washsha, Muwashsha, 138 (F), 139 (A). 31. Al-​Isbahani, Kitab al-​Aghani, vol. III, 27. Farmer summarizes his story in History of Arabian Music, 50–​53. 32. Farmer, History of Arabian Music, 102, 117. 33.  The training of singing girls is similar to the training of Chinese, Japanese, and Indian courtesans. Each system of courtesanship had a process of selection, an apprenticeship stage, and



Visibility and Performance

j 93

formal training. Training likewise included music, dance, recitation, poetics, and communication skills. See Zeitlin, “ ‘Notes of Flesh’ ”; Downer, Women of the Pleasure Quarters; Srinivasan, “Royalty’s Courtesans and God’s Mortal Wives,” 161–​181; and Qureshi, “Female Agency.” For licit sex and slavery as well as the transactional complications of slavery and marriage, see Ali, Marriage and Slavery, 25–​26, 50–​53. 34. Al-Jahiz, Risalah on the Singing Girls, Section 53, 35. 35.  Ibn al-​Washsha provides a list of expected gifts, including lute strings and paper, in the Muwashsha, 136–​137 (F), 135–​136 (A). 36. Awde, Women in Islam, 112. 37. Because the social hierarchy was built on a complex web of favors and obligation, if one person was invited to an event, he or she needed to reciprocate or, at a minimum, provide costly gifts to the host. The cost could be offset if he or she acquired enough gifts and/​or favors, but just as often caused financial ruin. Al-Jahiz, in his essay on the singing girls, 35–37, addresses the system of favors and reciprocity with considerable pungency. 38. Al-Jahiz, Risalah on the Singing Girls, 36. 39. See Downer, Women of the Pleasure Quarters, and Qureshi, “Female Agency,” for Japan and India, respectively. For sixteenth-​and seventeenth-​century Europe, see Feldman and Gordon, Courtesan’s Arts. For eighteenth-​and nineteenth-​century European salon culture, see Hickman, Courtesans. 40. Meyers-​Sawa makes the same suggestion in “Role of Women.” 41. Al-​Jahiz also notes the distinction between houses and individual owners in the Risalah on the Singing Girls. An example of a singing girl who began at a school and later became a favorite of the caliph is the singing girl Shaklah. She was a black slave acquired through conquest who began her studies at a music school in the Hijaz. As a result of her talent and beauty, she came to the notice of caliph al-​Mahdi and is thought to be the mother of prince Ibrahim al-​Mahdi. The story is in al-​Tabari, Early Abbasi Empire, 46. 42.  Fees included the entertainment, but also wine and food. See al-​Tanukhi, Table Talk; Margoliouth, 97–​98, 100, 102. 43. Al-Tanukhi, Table Talk, 100, 101–​105. The story of the brothel party is omitted from the English translation but can be found in the Arabic original of Nishwar al-​Muhadarah, 172–​183. I am grateful to Matthew Gordon for directing me to this amusing anecdote. 44. Al-Farabi was a practical musician and highly respected theorist, advocating practical musicianship in his Great Book of Music. See Sawa, Music Performance Practice in the Early Abbasid Era, 11–​22, and Rhythmic Theories and Practices in Arabic Writings to 339AH/​950ce. See also Shehadi, Philosophies of Music in Medieval Islam, 1–​11, 50–​65; Farmer, History of Arabian Music, 175–​177; Wright, “Musiki,” EI, Vol. 7, and “Music and Verse.” 45. To date, there are still relatively few studies of medieval Islamicate music. For performance, see Sawa, Music Performance Practice in the Early Abbasid Era and Rhythmic Theories and Practices in Arabic Writings to 339AH/​950ce. For connections to modern performance, see Racy, Making Music in the Arab World; Sawa, “Survival of Some Aspects of Medieval Arabic Performance Practice.” See also Farmer, A History of Arabian Music, and “Ghina,” 2:1072–​1075; Shiloah, Music in the World of Islam; Touma, The Music of the Arabs; Neubauer, Musiker Am Hof Der Fruhen ‘Abbasiden; Stigelbauer, Die Sängerinnen am Abbasidenhof um die Zeit des Kalifen Al-Mutawakkil: Nach dem Kitāb al-Agānī des Abu-l-Farağ al Isbahānī u. anderen Quellen dargestellt. For philosophy and theory, see al-​Faruqi, “The Nature of the Musical Art of Islamic Culture”; Shehadi, Philosophies of Music in Medieval Islam.

94 i

Lisa Nielson

46. Courtiers were expected to be able to cook, garden, speak and write well, hunt, dabble in poetry, and understand the language of flowers and scents. These pastimes were associated with culture and refinement, and are the focus of the Muwashsha. For an overview of social life and the court, see Ahsan, Social Life under the Abbasids; Lindsay, Daily Life in the Medieval Islamic World. For cultural transferences between the aristocracy and merchant classes, see Shoshan, “High Culture and Popular Culture in Medieval Islam.” 47.  See Gordon, “Place of Competition”; Sawa, Music Performance Practice in the Early Abbasid Era. 48.  For a detailed analysis of one example of patronage and music, see Gordon, “Place of Competition.” See also Sharlet, Patronage and Poetry in the Islamic World. 49.  See Sawa “Status and Roles of the Secular Musicians in the Kitab al-​Aghani (Book of Songs).” 50. For an analysis of gender and agency with respect to women poets, see Myrne, Narrative, Gender and Authority in Abbasid Literature on Women. 51. Kilpatrick, Great Book of Songs, 53. 52.  See Gordon, “Place of Competition,” “Arib al-​Ma’muniyah,” and “Arib al-​Ma’muniya: A  Third/​Ninth Century Abbasid Courtesan.” For a translation of all the anecdotes from the Aghani related to Arib, see Myrne, Narrative, Gender and Authority, 272–​298. 53. Al-Isbahani, Kitab al-​Aghani, 16:342. The Aghani gives her name as Khunth. 54. From the root, ‫خلس‬. For the use of music in the majlis, see Sawa, Music Performance Practice in the Early Abbasid Era, 111–​144. For a brief article on singing girls at court and in majalis, see al-​Heitty, “The Contrasting Spheres of Free Women and Jawari in the Literary Life of the Early Abbasid Caliphate.” See also Brookshaw, “Palaces, Pavilions and Pleasure Gardens.” 55. Brookshaw, “Palaces, Pavilions and Pleasure Gardens,” and Sawa, Music Performance Practice in the Early Abbasid Era. The Aghani contains the most detail about these gatherings, and Sawa discusses many in Music Performance Practice in the Early Abbasid Era. For another version of the story of the famous “yellow” majlis called by the caliph al-​Mutawakkil, where everything—​from the garden to the musicians—​had to be yellow, see al-​Tanukhi, Table Talk, 160. 56. See Sawa, Music Performance Practice in the Early Abbasid Era, 142–​152. 57. The Aghani is filled with such anecdotes, but for a collection of some of the funnier (and more painful), see Sawa, “Status and Roles of the Secular Musicians” and “Musical Humor in the Kitab al-​Aghani.” See also ­chapters 5, 6, and 7 in Farmer, History of Arabian Music. 58.  There are many references in the Aghani to musicians performing astonishing feats of memory. Such use of memory were also common in the medieval West. See Carruthers, Book of Memory; Busse-​Berger, Medieval Music and the Art of Memory. 59. See Sawa, Music Performance Practice in the Early Abbasid Era, 151–​153. Playing in unison was a necessary aural solution for large, often highly resonant spaces. It also offered a platform for the display of wealth, as in the visual impact of 2,000 singing girls. 60. Al-​Jahiz, Risalah on the Singing Girls, Section 53, beginning of section 54, 35. 61. Farmer and Shiloah both suggest that Ibn Abi al-​Dunya’s treatise was the catalyst for the genre of sama literature, because there was a proliferation of arguments and counterarguments regarding the legality of music that closely followed in response. See Shiloah, “Music and Religion in Islam,” 143–​155, and Music in the World of Islam, 31–​35; Farmer, History of Arabian Music, 146. 62. Ibn Abi al-​Dunya, Dhamm al-​Malahi, 31–​3 2. The Arabic reads ‫( القينات الجواري المغنيات‬alqayyinat al-jawari al-mughaniyyat).



Visibility and Performance

j 95

63. Ibn Abi al-​Dunya and his immediate followers based their arguments against listening on Sura 31:6, defining music as a form of diverting speech. For example, the collection of treatises dedicated to refuting listening held at the National Library of Israel, Jerusalem, has texts from the ninth to the fifteenth century, several of which reflect Ibn Abi al-​Dunya’s arguments and symbolic references to singing girls. See, for example, al-​Ajurri, Al-Jawab ‘an masalat al-sama (Response to Questions Concerning Listening [to Music], NLI, AP. AR., 158/​1 (f. 1a–​10a), copied by Ibn Burayd al Diri, 1451 from a tenth-​century manuscript. 64. Sama is primarily a Sufi genre, as the majority of pro-​sama arguments grew out of Sufi resistance to those legalists who wanted to ban music entirely, such as Ibn Abi’l Dunya. See Gribetz, “The Sama Controversy.” 65. The Arabic root ‫( سمع‬sama), meaning to hear or listen, has additional connotations of paying attention and learning. Form X ‫( استمتع‬istamta) also has the meaning of “to enjoy.” Music and musicians as a literary device were popular in a variety of literary genres during the eighth, ninth and tenth centuries. See Farmer, History of Arabian Music, 127–​128, 152, 168–​169, 172–​175, 214, 221–​222, 224, 226, as well as his extensive bibliography of printed and manuscript sources. 66. Listening to music in medieval Arabia involved more than passive listening. It required the mind, body, and senses. The purpose of a performance was to achieve, at minimum, a heightening of emotion and, at best, ecstasy or tarab (‫)طرب‬. Tarab (‫ )طرب‬is hard to render fully into English. The closest one can get is the acknowledged mind–​body connection that transports one when listening to music. Today, we might refer to such phenomena as getting goose bumps or similar physical reactions to deep, spontaneous emotion. For a detailed analysis of Tarab and saltanah in music, see Racy, Making Music in the Arab World, 96–​100, 133–​144, and passim. 67. The gendering of music was not uncommon in the Ancient Near East. Certain instruments and compositional genres were gendered, and performance using specific instruments conferred or reified a musician’s gender identity. For the medieval Islamic courts, see Nielson, “Gender and the Politics of Music.” 68. See Sawa, “Status and Roles of the Secular Musicians” and “Musical Humor in the Kitab al-​Aghani.”

Bibliography Primary Sources al-​Ajurri, Abu Bakr Muhammad ibn Husayn. Al-​Jawab `an masalat al-​sama (Response to Questions Concerning Listening [to Music]). AP. AR., 158/​1 (f. 1a–​10a). Jerusalem: National Library of Israel. Ibn Abi al-​Dunya, Abu Bakr Abd Allah ibn Muhammad. Dhamm al-​Malahi. Cairo:  Dar al-​Itisam, 1987. —​—​—​. Tracts on Listening to Music: Being the Dhamm al-​Malahi of Ibn Abi Dunya. Translated by James Robson. London: The Royal Asiatic Society, 1937. Ibn al-​Nadim, Abu al-​Faraj Muhammad ibn Abi Yaqub. Al-​Fihrist. The Fihrist of Ibn al-​ Nadim:  A  10th Century Survey of Muslim Culture. 2 vols. Translated and edited by Bayard Dodge. New York: Columbia University Press, 1970. Ibn al-​Washsha, Abu Tayyib ibn Muhammad ibn Ishaq. Le Livre de brocart (The Book of Brocades). Translated by Siham Bouhlal. Paris: Editions Gallimard, 2004. —​—​—​. Al-​Muwashsha aw al-​zarf wa’l-​zurafa. Beirut: Dar Sadir, 1965.

96 i

Lisa Nielson

al-​Isbahani, Abu al-​Faraj. Kitab al-​Aghani. 24 vols. Cairo: Dar al-​Kutub, 1927–​1974. al-​Jahiz, Abu Uthman Amr ibn Bahr. Kitab Mufakharat al-​jawari wa’l-​ghilman. Beirut:  Dar al-​Makshuf, 1957. —​—​—​. Nine Essays of al-​Jahiz. Translated by William M. Hutchins. New  York:  Peter Lang,  1989. —​—​—​. Risalat al-​qiyan. The Epistle on the Singing Girls. Translated by A. F.  L. Beeston. Warminster: Aris & Phillips, 1980. al-​Masudi, Abu al-​Hasan Ali ibn al-​Husayn. The Meadows of Gold: The Abbasids. Translated by Paul Lunde and Caroline Stone. London: Kegan Paul International, 1989. —​—​—​. Muruj al-​Dhahab wa Maadin al-​Jawhar. 4 vols. Beirut: Dar al-​Andalus, 1965–​1966. al-​Tabari, Abu Jafar Muhammad ibn Jarir. The Early Abbasi Empire. Translated by John Alden Williams. Vol. 2, The Sons and Grandsons of al-​Mansur: The Reigns of al-​Mahdi, al-​Hadi and Harun al-​R ashid. Cambridge: Cambridge University Press, 1989. —​—​—​. Tarikh al-​umam wa’l-​muluk. The Early Abbasi Empire. Translated by John Alden Williams. Vol. 1, The Reign of Abu Ja’far al-​Mansur (ad 754–​775). Cambridge:  Cambridge University Press, 1988. al-​Tanukhi, Abu Ali al-​Muhassin ibn Ali. Nishwar al-​Muhadara wa-​Akhbar al-​Mudhakara. Edited by Abbud al-​Shalji, 4 vols. Beirut: Dar Sadir, 1398/​1978. —​—​—​. Table Talk of a Mesopotamian Judge. Translated by D. S. Margoliouth. London: Royal Asiatic Society, 1922.

Secondary Sources Ahsan, Muhammad Manazir. Social Life under the Abbasids. London: Longman Group, 1979. Ali, Kecia. Marriage and Slavery in Early Islam. Cambridge, MA:  Harvard University Press, 2010. Awde, Nicholas, ed. Women in Islam: An Anthology from the Qur’an and Hadiths. New York: St. Martin’s Press, 2000. Brookshaw, Dominic. “Palaces, Pavilions and Pleasure Gardens: The Context and Setting of the Medieval Majlis.” Middle Eastern Literatures, Incorporating Edebiyat 6, no. 2 (2003): 199–​223. Busse-​Berger, Anna Maria. Medieval Music and the Art of Memory. Berkeley:  University of California Press, 2005. Carruthers, Mary. The Book of Memory:  A  Study of Memory in Medieval Culture. Cambridge: Cambridge University Press, 1990. Caswell, Fuad Matthew. The Slave Girls of Baghdad. London: I.B. Tauris, 2011. Cheikh-​Moussa, Abdallah. “La Negation d’eros ou le ‘Ishq d’apres deux epitres d’al Gahiz.” Studia Islamica 72 (1990): 71–​119. Davidson, James. Courtesans and Fishcakes:  The Consuming Passions of Classical Athens. New York: Harper Perennial, 1999. Downer, Leslie. Women of the Pleasure Quarters. New York: Broadway Books, 2001. El-​Cheikh, Nadia Maria. “Describing the Other to Get at the Self: Byzantine Women in Arabic Sources (8th–​11th Centuries).” Journal of the Economic and Social History of the Orient 40, no. 2 (1997): 239–​250. —​—​—​. “Revisiting the Abbasid Harems.” Journal of Middle East Women’s Studies 1, no. 3 (2005): 1–​19.



Visibility and Performance

j 97

Farmer, Henry George. “Ibn Khurdadhbih on Musical Instruments.” Journal of the Royal Asiatic Society of Great Britain and Ireland 3 (1928): 509–​518. —​—​—​. “The Influence of Music: From Arabic Sources.” In Proceedings of the Musical Association 52 (1925): 89–​124. —​—​—​. “Ghina.” In Encyclopedia of Islam. 2nd ed., edited by P. Bearman et  al. Leiden:  Brill Publishing Online, 2012. —​—​—​. “Greek Theorists of Music in Arabic Translation.” Isis 13, no. 2 (1930): 325–​333. —​—​—​. A History of Arabian Music to the 13th Century. London: Luzac, 1929. [Reprinted in 1995.] —​—​—​. Music: The Priceless Jewel. Vol. 1. Frankfurt am Main: Institute for the History of Arabic-​ Islamic Science at the Johann Wolfgang Goethe University, 1942. [Reprinted in 1997.] al-​Faruqi, Lois Ibsen. “The Nature of the Musical Art of Islamic Culture:  A  Theoretical and Empirical Study of Arabian Music.” PhD diss., Syracuse University, 1974. Feldman, Martha, and Bonnie Gordon, eds. The Courtesan’s Arts:  Cross-​Cultural Perspectives. Oxford: Oxford University Press, 2006. Franklin, John. “The Global Economy of Music in the Ancient Near East.” In Sounds of Ancient Music, edited by J. G. Westenholz, 27–​37. Jerusalem: Bible Lands Museum, 2007. Gordon, Matthew. “Arib al-​Ma’muniyah.” In Arabic Literary Culture, 500–​925, edited by Michael Cooperson and Shawkat Toorawa, 85–​90. Farmington Hills, MI: Thomson Gale, 2004. —​—​—​. “Arib al-​Ma’muniya:  A  Third/​Ninth Century Abbasid Courtesan.” In Views from the Edge:  Essays in Honor of Richard W.  Bulliet, edited by Neguin Yavari et  al., 86–​100. New York: Columbia University Press, 2004. —​—​—​. “The Place of Competition:  The Careers of Arib al-​Ma’muniya and Ulayya bint al-​ Mahdi, Sisters in Song.” In Occasional Papers of the School of Abbasid Studies, edited by James Montgomery, 61–​81. Orientalia Lovaniensia Analecta, no. 135. Leuven: Peeters, 2004. —​—​—​. “Yearning and Disquiet: al-​Jahiz and the Risalat al-​Qiyan.” In Al-​Jahiz: A Muslim Humanist for Our Time, edited by Armin Heinemann et al., 253–​268. Wurzburg: Ergon-​Verlag, 2009. Gribetz, Arthur. “The Sama Controversy: Sufi vs Legalist.” Studia Islamica 74 (1991): 43–​62. Hambly, Gavin R.  G., ed. Women in the Medieval Islamic World. New  York:  St. Martin’s Press, 1998. al-​Heitty, Abd al-​Kareem. “The Contrasting Spheres of Free Women and Jawari in the Literary Life of the Early Abbasid Caliphate.” Al-​Masaq 3 (1990): 31–​51. Hickman, Katie. Courtesans:  Money, Sex and Fame in the Nineteenth Century. New  York: HarperCollins Publishers, 2003. Hoyland, Robert G. Arabia and the Arabs:  From the Bronze Age to the Coming of Islam. London: Routledge, 2001. Kilpatrick, Hilary. Making the Great Book of Songs: Compilation and the Author’s Craft in Abu’l Faraj al-​Isbahani’s “Kitab al-​Aghani.” London: RoutledgeCurzon, 2003. —​—​—​. “Mawali and Music.” In Patronate and Patronage in Early and Classical Islam, edited by Monique Bernards and John Nawas, 326–​348. Leiden: Brill, 2005. Lam, Joseph S. C. “Reading Music and Eroticism in Late Ming Texts.” Nan Nu 12 (2010): 215–​254. Librande, Leonard. “Ibn Abi al-​Dunya:  Certainty and Morality.” Studia Islamica 100/​101 (2005): 5–​42. Lindsay, James. Daily Life in the Medieval Islamic World. Westport, CT: Greenwood , 2005. MacDonald, Michael C.  A. “Goddesses, Dancing Girls or Cheerleaders? Perceptions of the Divine and Female Form in the Rock Art of pre-​Islamic North Arabia.” In Dieux et déesses

98 i

Lisa Nielson

d’Arabie:  Images et representations, edited by Isabelle Sachet and Christian Julien Robin, 261–​297. Paris: De Boccard, 2012. Meyers-​ Sawa, Suzanne. “Historical Issues of Gender and Music.” In The Garland Encyclopedia of Music:  The Middle East, edited by Virginia Danielson et  al., 293–​298. New York: Routledge, 2002. —​—​—​. “The Role of Women in Musical Life: The Medieval Arabo-​Islamic Courts.” Canadian Women’s Studies 8, no. 2 (1987): 93–​95. Myrne, Pernilla. Narrative, Gender and Authority in Abbasid Literature on Women. Orientalia et Africana Gothoburgensia, no. 22. Gothenburg, Sweden: Edita Vastra, 2010. Neubauer, Eckhard. Musiker Am Hof Der Fruhen Abbasiden. Frankfurt am Main:  JW Goethe-​Universitat, 1965. Nielson, Lisa. “Gender and the Politics of Music in the Early Islamic Courts.” Early Music History 31 (2012): 233–​259. Orr, Leslie. Donors, Devotees and Daughters of God:  Temple Women in Medieval Tamilnadu. New York: Oxford University Press, 2000. Pellat, Charles. “Al-​Jahiz.” In Abbasid Belles-​Lettres, edited by Julia Ashtiany et  al., 78–​95. Cambridge: Cambridge University Press, 1990. —​—​—​. “Kayna.” In Encyclopedia of Islam, 2nd ed., edited by P. Bearman et  al. Leiden:  Brill Publishing Online, 2012. —​—​—​. The Life and Works of Jahiz. Translated by D. M. Hawke. Berkeley: University of California Press, 1969. —​—​—​. Le Milieu Basrien et la formation de Gahiz (The Environment of Basra and the Influence on Jahiz). Paris: Librairie D’Amerique et D’Orient, 1953. Poche, Christian. “Music in Ancient Arabia from Archaeological and Written Sources.” In The Garland Encyclopedia of World Music: The Middle East, edited by Virginia Danielson et al., 357–​362. New York: Routledge, 2002. Qureshi, Regula. “How Does Music Mean? Embodied Memories and the Politics of Affect in the Indian ‘Sarangi.’” American Ethnologist 27, no. 4 (2000): 805–​838. Racy, A. J. Making Music in the Arab World:  The Culture and History of Tarab. Cambridge: Cambridge University Press, 2004. Robson, James, and Henry George Farmer. “The Kitab al-​Malahi of Abu Talib Al-​Mufaddal ibn Salama.” Journal of the Royal Asiatic Society of Great Britain and Ireland 2 (1938): 231–​249. Rowson, Everett K. “The Categorization of Gender and Sexual Irregularity in Medieval Arabic Vice Lists.” In Body Guards: The Cultural Politics of Gender Ambiguity, edited by Julia Epstein and Kristina Straub, 50–​79. New York: Routledge, 1991. —​—​—​. “The Effeminates of Early Medina.” Journal of the American Oriental Society 111, no. 4 (1991): 671–​693. —​—​—​. “Gender Irregularity as Entertainment:  Institutionalized Transvestism at the Caliphal Court in Medieval Baghdad.” In Gender and Difference in the Middle Ages, edited by Sharon Farmer and Carol Braun Pasternack, 45–​72. Minneapolis: University of Minnesota Press, 2003. Sawa, George. “Musical Humor in the Kitab al-​Aghani.” In Logos Islamikos: Studia Islamica in Honorem Georgii Michaelis Wickens, edited by Rodger Savery and Dionisius Agius, 35–​50. Toronto: Pontifical Institute of Medieval Studies, 1984. —​—​—​. Music Performance Practice in the Early Abbasid Era, 132–​32-​AH/​750–​932AD. Toronto: Pontifical Institute of Medieval Studies, 1989.



Visibility and Performance

j 99

—​—​—. Rhythmic Theories and Practices in Arabic Writings to 339AH/​950CE. Ottawa: Institute of Medieval Music, 2009. —​—​—​. “The Status and Roles of the Secular Musicians in the Kitab al-​Aghani (Book of Songs) of Abu al-​Faraj al-​Isbahani (D.356AH/​967AD).” Asian Music 17, no. 1 (1985): 69–​82. —​—​—​. “The Survival of Some Aspects of Medieval Arabic Performance Practice.” Ethno­ musicology 25, no. 1 (1981): 73–​86. Sharlet, Jocelyn. Patronage and Poetry in the Islamic World:  Social Mobility and Status in the Medieval Middle East and Central Asia. London: I.B. Tauris, 2011. Shehadi, Fadlou. Philosophies of Music in Medieval Islam. New York, NY: Brill, 1995. Shiloah, Amnon. “The Arabic Concept of Mode.” Journal of the American Musicological Society 34, no. 1 (1981): 19–​42. —​—​—​. “Music and Religion in Islam.” Acta Musicologica 69, no. 2 (1997): 143–​155. —​—​—​. Music in the World of Islam: A Socio-​Cultural Study. Detroit, MI: Wayne State University Press, 1995. Shoshan, Boaz. “High Culture and Popular Culture in Medieval Islam.” Studia Islamica 73 (1991): 67–​107. Soneji, Davesh. Unfinished Gestures:  Devadasis, Memory and Modernity in South India. Chicago: University of Chicago Press, 2012. Srinivasan, Doris. “Royalty’s Courtesans and God’s Mortal Wives.” In The Courtesan’s Arts:  Cross-​Cultural Perspectives, edited by Martha Feldman and Bonnie Gordon, 161–​181. New York: Oxford University Press, 2006. Stetkevytch, Suzanne Pinckney. “The Mute Immortals Speak: Pre-​Islamic Poetry and the Poetics of Ritual.” In Myths and Poetics, edited by Gregory Nagy, Ithaca, NY:  Cornell University Press, 1993. Stigelbauer, Michael. Die Sängerinnen am Abbasidenhof um die Zeit des Kalifen Al-Mutawakkil: Nach dem Kitāb al-Agānī des Abu-l-Farağ al Isbahānī u. anderen Quellen dargestellt. PhD diss., Universitat Wien, 1975. Touma, Habib Hassan. The Music of the Arabs. Translated by Laurie Schwartz. Portland, OR: Amadeus Press, 1996. Walthall, Anne, ed. Servants of the Dynasty: Palace Women in World History. Berkeley: University of California Press, 2008. Wright, O. “Music and Verse.” In Arabic Literature to the End of the Umayyad Period, edited by A. F. L. Beeston et al., 433–​450. Cambridge: Cambridge University Press, 1983. —​—​—​. “Musiki.” In Encyclopedia of Islam, 2nd ed., edited by P. Bearman et  al. Leiden:  Brill Publishing Online, 2012. Zeitlin, Judith. “‘Notes of Flesh’: The Courtesan’s Song in Seventeenth-​Century China.” In The Courtesan’s Arts: Cross-​Cultural Perspectives, edited by Martha Feldman and Bonnie Gordon, 75–​99. New York: Oxford University Press, 2006.

5 The Qiyan of al-​Andalus Dwight F. Reynolds

i  Among the many different classes and types of female slaves in the premodern Islamic world, the group that has attracted the most attention from both medieval writers and modern scholars is that of the qiyan (pronounced kee-​yaan), commonly referred to in English as “singing girls” or “singing slave girls,” although neither of these terms is particularly accurate.1 The qiyan were not necessarily young (i.e., “girls”); indeed, some of the most famous among them continued performing into a ripe old age. Nor were they exclusively “singers,” for they were often skilled in a variety of different literary and performing arts, including the composition of both poetry and music, the recitation of akhbar (“accounts” or “anecdotes” of a historical or literary nature), calligraphy, shadow puppetry, and more. They were sexually available to their owners, but as a result of their musical skills and training, they were generally more expensive than female slaves who served only as concubines and were therefore usually accorded a more esteemed social status. As companions and entertainers, they mingled more openly than concubines with male society while performing for their masters’ guests, although this sometimes took place from behind a curtain or screen. The geisha of Japan are perhaps the most comparable form of socially institutionalized female companionship and entertainment for male patrons, although, of course, the differences are also myriad. Although slaves, the qiyan constituted a rather public, even prominent, class of women, especially those qiyan who were associated with the caliphal court. There is, therefore, a comparatively rich body of documentary evidence about their training, their performances, their personalities, and their distinctive characteristics and talents (discussed later). For a small number of individuals, there exist relatively complete biographies, although this information is usually presented as a series of separate anecdotes 100

The Qiyan of al-​Andalus

j 101

rather than as a cohesive narrative. For the majority, however, only brief glimpses of key moments in their lives are preserved. In comparison with other classes of women and other classes of slaves, though, the qiyan offer a unique opportunity for the study of gender, slavery, and social relations during the medieval period. There is no doubt that the practice of owning qiyan was brought to al-​Andalus with the initial Arab/​Islamic conquest, but the social, political, and cultural contexts in medieval Iberia were quite different than those in the East. So the central question posed here is whether the qiyan as an “institution”—​that is, their training, social status, artistic repertory, purchase and sale, and so forth—​developed differently in al-​Andalus than it did in the Mashriq (Eastern Mediterranean), and if so, what were the distinctive characteristics of this institution in the Andalusian context? There is a large body of primary texts and secondary modern scholarship that deals with the qiyan in the East, but not nearly the same amount of resources is available for tracing the history of their presence in al-​Andalus.2 There are no Andalusian texts comparable with al-​Jahiz’s treatise on the qiyan, extended discussions such as those found in Kitab al-​Muwashsha (The Brocaded Book) of al-​Washsha, or the large number of biographies and anecdotes about them that are compiled and preserved in al-​Isbahani’s Kitab al-​Aghani (The Book of Songs), his al-​Ima al-​shawa‘ir (The Slave Poetesses), Ibn al-​Sa‘i’s Nisa al-​khulafa (The Women of the Caliphs), al-​Suyuti’s al-​Mustazraf min akhbar al-​ jawari (Choice Anecdotes from the Accounts of Concubines), and other works. In the case of al-​Andalus, there are, instead, solitary anecdotes scattered through historical accounts such as Ibn Hayyan’s Kitab al-​Muqtabis (lit. “The Quoter”; i.e., a collection of quotes from earlier sources);3 literary biographical compilations such as Ibn Bassam’s al-​Dhakhira fi mahasin ahl al-​jazira (A Treasury of the Virtues of the People of the Iberian Peninsula); biographical dictionaries such as those penned by Ibn Bashkuwal, al-​Humaydi, and others; and the well-​known seventeenth-​century historical description of al-​Andalus, Nafh al-​tib fi ghusn al-​Andalus al-​ratib (The Scented Breeze from the Tender Branch of al-​Andalus), by Ahmad al-​Maqqari. In addition, the two brief chapters on music from Fasl al-​khitab fi madarik al-​hawass al-​khams li-​uli al-​albab (The Final Say on the Perceptions of the Five Senses for Those Possessed of Intelligence) by the thirteenth-​century Tunisian writer Ahmad al-​Tifashi,4 published only during the 1960s, provide some very significant information, and to these known sources may now be added the 18 biographies of Andalusian singers, eight of whom are qiyan, from Ibn Fadl Allah al-​Umari’s (d. 1349) encyclopedia Masalik al-​absar fi mamalik al-​amsar (The Paths of Perception among the Realms of the Great Cities).5 These latter biographies provide not only the earliest accounts of musical life in the Cordoban court but also the first biographical accounts of qiyan in al-​Andalus. Unfortunately, the Kitab fi qiyan al-​Andalus (The Book of the Qiyan of al-​ Andalus), written by female author Umm al-​Fath bint Jafar (fl. eleventh century), in which she is said to have emulated the great Book of Songs of al-​Isbahani, has not survived.6 The portrait of the qiyan in al-​Andalus presented here must therefore be both partial and fragmentary. It consists not of a comprehensive history, but rather of a study

102 i

Dwight F. Reynolds

of several clusters of sources that provide relatively ample documentation for certain historical periods, with little to no information currently available for the time periods in between.7 There is enough evidence, however, to distinguish three broad historical periods: (1) the early period, when all, or nearly all, of the female singers mentioned in the historical documentation were trained in Medina or Baghdad and then imported westward to al-​Andalus and during which there was apparently little difference between the musical repertories in the East and West; (2) a middle period, when qiyan were also (and eventually primarily) trained in the Umayyad capital of Cordoba and the first traces of a distinctive Andalusi musical tradition begin to appear; and (3) a later period, after the sacking of Cordoba in 1013, when singers were trained primarily in Seville, the Andalusian song-​forms of muwashshah and zajal were spreading throughout the Arabic-​speaking Middle East, and qiyan were being exported from Seville throughout al-​Andalus and North Africa. There may also have been a period when qiyan were trained in Nasrid Granada (1212–​1492), but to date no sources that state this specifically have come to light. Mention should also be made of the female “Moorish” singers and dancers from Valencia who performed in the royal court of Aragon and Catalonia during the thirteenth through fifteenth centuries, but these women were not slaves, and also appear to have performed primarily in family groupings along with their husbands, mothers, and fathers, and thus lie beyond the scope of this study.8 Qiyan in both the East and in al-​Andalus were usually women who had become slaves as captives during the ongoing wars of conquest, who had been sold into slavery by their families, or who had been born into slavery in regions already under Islamic control. Although historical sources often mention the price or worth of an individual qayna, few of the logistical details of the market in qiyan are known. In this regard, some of the Andalusian materials examined in this chapter are particularly informative. The slave merchant Ibn al-​Kattani has left us a remarkable, albeit clearly exaggerated, account of the skills he taught female slaves to raise their value, and al-​Tifashi penned an account of how qiyan were examined in the marketplace before being purchased (discussed later). In general, certain female slaves were selected at an early age for training in the musical arts, presumably on the basis of their voice or some other indication of musical talent, along with their looks. A passage in al-​Isbahani’s Kitab al-​Aghani states that Ibrahim al-​Mawsili was the first to train light-​skinned, beautiful female slaves in the art of singing, and that previously only those with darker skin and those who were less good-​looking had become singers: “People did not use to teach beautiful slave-​girls to sing, but instead only taught light brown and black [slave girls to sing]. The first person to teach expensive [= fair-​skinned] slave-​girls to sing was my father. He achieved the highest-​level [of training] of female singers, and thereby raised their value.”9 The implication here appears to be that fair-​skinned female slaves had been sold previously for high prices only for their looks (i.e., as concubines), whereas darker skinned slaves had been trained in music as a means of increasing their worth. Ibrahim al-​Mawsili then began training fair-​skinned young women in music, thereby creating a new class of

The Qiyan of al-​Andalus

j 103

particularly expensive qiyan. The degree to which this claim is historically true is difficult to determine, but the Kitab al-​Aghani also preserves a complaint, in verse, immediately following the previously quoted passage that this practice on al-​Mawsili’s part was inflating the prices of qiyan in the market.10 Before examining the qiyan of al-​Andalus, two aspects of the surviving documentation should be noted. A large percentage of the anecdotes about qiyan in the East offer accounts of situations in which the female performers engage in witty repartee, deliver stinging retorts to male listeners (sometimes even to their owners), and, in general, put male audience members in their place. The most famous qiyan had a reputation for being able to best the men who listened to them not only in the spontaneous composition of poetry and music, but also in banter, wordplay, and witty comebacks. This waspish aspect of the qiyan’s role is almost entirely absent from Andalusian sources, but it is difficult to know if this reflects the actual behavior of qiyan in the elite circles of Andalusian society or the choice of Andalusian authors regarding what type of material to record in their texts. A second theme found in several Eastern sources is condemnation of the entire institution of qiyan. Al-​Jahiz, for example, declares that passion for singing girls is “dangerous,” and speaks of the “addiction” and “infatuation” of the men who frequent them. They are dangerous not only because they lead men to pursue pleasure over all else, but also because the qiyan are quite literally trained to feign love and affection for their devotees: The singing-​girl is hardly ever honest in her passion or sincere in her affection, for she, by training and by disposition, sets traps and snares for her admirers in order that they may plunge into her toils [lit. “noose”].11 He goes on to describe how qiyan exchange provocative glances with a male listener, toy with him in the verses of their songs, correspond with him, and pretend to feel undying love for him above all others. After they have snared a man, they then find fault with him, feign intense jealousy, and seek gifts. Although al-​Jahiz does allow that sometimes this love becomes real, and that a singing girl may occasionally truly fall in love and even renounce her musical craft to lower her price so her lover can afford to buy her, he nevertheless concludes, “Yet for the most part singing-​girls are insincere and given to employing deceit and treachery in squeezing out the property of the deluded victim and then abandoning him.”12 Al-​Jahiz describes a world in which there appear to have been not only higher class qiyan who lived in the households of notables and in the court, entertaining only their master and his guests, but also a somewhat lower class of activity in which owners used their qiyan actively for personal gain by extracting gifts and various forms of payment from admirers. The owner allows the admirer or dupe to visit his home or establishment and to see and hear his singing girls. During their gathering, he turns a blind eye to the winks, smiles, and double entendres she uses to ensnare the admirer, and even pretends to

104 i

Dwight F. Reynolds

doze off so they have a few moments to touch and kiss, until the dupe is hopelessly besotted with her. Having carefully evaluated the financial worth of his “victim,” the owner holds off for a period, during which the admirer sends gifts and continues to visit. Then, when as much as possible has been extracted from him, the owner finally arranges for a tryst, knowing full well that once the admirer has obtained his desire, his passion will abate by nine-​tenths, as well his gifts and other payments to the owner.13 This is a far cry from the elegant world of the court where qiyan are figures of beauty and grace who are only available to their owners! No Andalusian text has yet come to light that connects qiyan with such brothel-​like practices or even that condemns the qiyan as a class or institution at all, although there are some examples of disapprobation of men who devote themselves too much to the life of pleasure. It is clear in these accounts, however, that it is the men who are to blame. The anonymous author of Nubdhat al-​asr fi akhbar muluk Bani Nasr (A Summary of the Age Regarding Reports about the Kings of the Nasrid Dynasty), for example, notes that after the devastating floods in Granada in 1478, Sultan Abu l-​Hasan ‘Ali (r. 1464–​1482, 1483–​1485) began to lead a degenerate life, devoting himself to pleasure, dissoluteness, and frivolity with his “female singers” (al-​nisa al-​mutribat), which the author interprets as a moral indication of the “beginning of the end” for the Nasrids.14 Trained in the East Although there may well have been a handful of qiyan in al-​Andalus during the first decades after the conquest (the period of the Umayyad governors and the reigns of Abd al-​Rahman I and Hisham), no trace of them is found in the historical sources.15 Ahmad al-​Tifashi, writing in the thirteenth century, states that it was only under the reign of al-​ Hakam I (r. 796–​822) that singers from the Mashriq and North Africa began to arrive in al-​Andalus.16 And, indeed, the oldest accounts of qiyan in al-​Andalus are the biographies of six female singers who belonged to the Emir al-​Hakam I, preserved in al-​Umari’s Masalik al-​absar.17 Al-​Hakam is portrayed generally in historical texts as a particularly cold-​blooded and ruthless ruler, so these glimpses of intimate scenes with his concubines and qiyan form an intriguing counterweight to the prevailing image of him. One of the anecdotes cited later includes a possible explanation for the discrepancy between the characterization of al-​Hakam found in other accounts and the one that emerges from these texts. It states that al-​Hakam did not engage in the pleasures of music and wine drinking in public, but only in the strictest privacy (discussed later). There is, however, something peculiar in the organization of these first six biographies in that they are presented in three pairs with identical or nearly identical names: the first two biographies are of qiyan who are both named Aziz, the next two are named Bahja and Muhja, and the last two are Fatin and Fatik.18 It is possible that al-​Hakam enjoyed this playful pairing of names, but another possibility is that al-​Umari was combining material from two separate texts and that there were, in fact, only three singers. Whatever the case

The Qiyan of al-​Andalus

j 105

may be, the information in each of the biographies is distinct and there are no obvious repetitions. Several observations may be made about the unknown author of these texts. First, he was an ardent supporter of the Umayyads, for, at several points, he offers direct praise of the Umayyads and condemnation of the “people of black”;19 second, he was well versed in the history of Arab music, for there are references to and comparisons with several of the famous singers from the East, including Azza al-​Mayla, Badhl, Sallamat al-​Qass, Lubna, and Habbaba, who flourished between the late seventh and early ninth centuries; and, finally, he does not shy from intimate portrayals of sexual union. There is a thematic unity to many of the anecdotes in these six accounts. The primary quality that is extolled in them is the ability of a qayna to sense her master’s emotional state and respond either in poetry or in music to his psychological needs at that moment, for which she is then generously rewarded. The multiple references to the abilities of the qiyan—​their sense of the master’s mood; knowledge of what dwelled in his heart, breast, or soul; and realization of his innermost desires—​may speak to the tastes of the compiler of these biographies, but, more likely, reflects a quality that was appreciated at that time. The qiyan were apparently practitioners of a kind of “musical therapy” for this temperamental ruler who is chiefly known in historical accounts as being vengeful and cruel. The following anecdote, for example, concerns the first singer named Aziz: It is recounted that al-​Hakam loved a certain courtesan whose name was Hayn [lit. “death” or “destruction”]. He went once into the desert when spring had embellished with many colors its flowing garments and adorned the pastures on the edges of the hills. He dismounted when the late afternoon light had grown dim, as if it were complaining about separation from a beloved, and the sun had begun to lower its disc onto the horizon. He had left Hayn behind and therefore lay awake that night—​his eyelids never tasting sleep nor his eyes touching slumber. He sent for his female musicians to entertain him during the night with their singing. Now Aziz was a poetess and singer, intelligent, courteous, clever, and a great teller of anecdotes. She grasped the emotional state of her master and what anguished him due to the separation from Hayn and the anxiety that had taken hold of him from this sudden separation, and composed a tune to poetry from one of [the poets of the tribe of ] Azd—​namely, Abu Adiyy Amir ibn Said, one of the sons of al-​Nimr ibn Uthman. . . . She rehearsed it until she knew it well, and when al-​Hakam called for his female singers, Aziz entered through the door of his pavilion singing this song. Al-​Hakam was so deeply moved that he got off his cushion and said: “By God, O Aziz, how perceptive you are about the places of affliction and how knowledgeable you are about the locations of complaint! Who composed that poem?” She replied that it was by a man from among the sons of al-​Nimr ibn Uthman. He replied, “By God, you are more worthy of it than he is, for you brought it forth as if it were a description of the state we are in!” Then he ordered her to repeat it and he immediately sent someone to bring Hayn. [Once she had joined them,] he stayed on this

106 i

Dwight F. Reynolds

excursion for several days, which [passed] as if they were a dream. He also ordered that Aziz be paid 10,000 [dirhams] and Hayn paid her the same amount. After that she [Aziz] was accorded al-​Hakam’s affection for the rest of his life.20 If the details of this anecdote are to be taken as accurate, al-​Hakam had departed on an excursion into the “desert” (countryside) leaving behind a particular concubine, Hayn, but brought several female musicians/​singers, among whom was Aziz. Her ability to sense and respond to her master’s mood is the focal point of not only this anecdote but also another (see next quote) in which she eventually won a privileged place among al-​ Hakam’s wives and the mothers of his children. In this second anecdote we learn not only of Aziz’s “promotion” to a new social station, but are also told explicitly that she was and remained his sexual companion for many years (a fact that is far more often assumed than actually recorded in qiyan biographies). In this anecdote, it is also clear that al-​Hakam did, in fact, enjoy drinking and listening to music, but was careful to do so only in private: It is also recounted that [al-​Hakam] sent for [Aziz] once at the beginning of the day in which the sun and his cup were raised to greet the morning. Now al-​Ḥakam never partook of pleasures except in secret, nor did he indulge in drink except behind a curtain, nor did he gather drinking-​companions except within his private quarters, nor did he compete in wine-​drinking except within his closest circle, for fear of the scandal of being found out and due to his aversion from having this become known. He spent the day suggesting songs to her and to the rest of his female singers, tossing down [cups of ] wine in private, until night had cracked the glass of day, and satiation appeared in his eyes . . .. Then al-​Hakam said, “Is there anyone among you who can compose a [fitting] poem for this [moment]?” Aziz spontaneously composed [the following verses]: The day has elapsed except for the remains of the rays of the late afternoon. And darkness has come to us from the East—​ Welcome be it, the best of guests! May this [sweet moment] last as long as the hoped for longevity of al-​Hakam, our valiant lord. Her poem pleased him greatly and he bestowed and lavished upon her gifts. Then he ordered her [to set it to music], so she composed a melody for it and sang it to that melody all that night while he [drained] cups one after the other to it and urged her on until the red of the [rising] sun set the coal of the night ablaze. When he began to desire to retire to his quarters in order to recline on his bed, and [in private] allow the matter of his drunkenness to take its course, he ordered that Aziz [be given] 10,000 dirhams and lengths of rich fabric. He increased her salary and transferred her to the elite of his concubines and the mothers of his children, and she remained thus, until she died, his companion in his bed and on his pillow.21

The Qiyan of al-​Andalus

j 107

Although Aziz apparently never gave him a son, at which point she would have become an umm walad (“mother of a son”), a fact that most likely would have been noted, it is clear that she remained one of the Emir’s closest female companions for many years. In an anecdote told about the second singer named Aziz, al-​Hakam had fallen out with one of his concubines after having been quite taken with her. Morosely he asked his female singers to entertain him. This they did until [the second] Aziz stepped forward and sang three verses that caused him to repent and to reconcile himself with his estranged concubine.22 In another account, when one of al-​Hakam’s concubines had asked for and received permission to go on an excursion to a pleasure palace outside the city, he soon began to rue his decision and missed her terribly. The singer, Bahja, “felt what was in his soul” and sang four verses by the poet al-​Buhturi, at the conclusion of which the Emir proclaimed, “It is as if you were the very heart in my breast” (i.e., and knew my sadness firsthand), for which he ordered that she be given 200 dinars and some jewels.23 Not every performance, however, was able to move the Emir so deeply. In several anecdotes, singers attempted but failed to respond in poetry and/​or song to the emotional needs of their master, and only the “heroine” of the anecdote managed to do so. Sometimes a singer was not successful on the first try, but succeeded during her second or third attempt. Al-​ Hakam once suggested to [Muhja] that she set certain verses by Abu Tammam to music:24 I am the son of those among whom Glory sought a wet-​nurse, whose name alone it bears as long as it lives. When [the heroes of the Tayyi tribe] passed on, it seemed that they had made liberality a law, so often did they commend it on their deathbeds. They are as palm and fingers to any glorious act of open-​handedness ( = What glorious acts of open-​handedness have there ever been that did not imitate theirs?) Honorable conduct they entrusted to the wealth of us, [their descendants]; we spent it all, but to our acts of generosity there is no end.25 She composed [a song] in [the rhythm of ] thaqil al-​ramal, but it did not please him. So she said, “I will compose another melody for it.” But he said, “You have ruined it for me. Look instead for something else to set to music.” So she composed [another song] using different verses by Abu Tammam: Race [our tribal hero] Hatim [the Generous] against a [life-​giving] rain-​cloud [= the standard metaphor for generosity], and no-​one would know which was which! This was a man! Some hoard worldly wealth; he gave all his away. Lo! whose is the treasure that has endured?

108 i

Dwight F. Reynolds

Let any who will, boast of acts of generosity; the boast belongs to us alone. In us are conjoined giving and glory, long sundered, as closely as a month conjoins days.26 When she sang this [new composition] al-​Hakam trembled with pleasure so much that he almost fell off his dais and he said to her, “You have done well, by God, and so beautifully! You have surpassed that which my soul desired.” And he ordered that she be given 100 dinars for each verse, which totaled 400 dinars.27 Al-​Hakam was also apparently quite fond of holding contests among his qiyan. In one anecdote he challenges them to create a musical setting for some verses by al-​Farazdaq. Muhja wins the competition and requests an unusual reward: They composed musical settings for these verses, created melodies and set the rhythms, but Muhja struck closest to what was in his soul. He said to her, “Tell me what you would like as your reward.” She replied, “That [the other women] sing only my compositions all day long.” He ordered them to do this and ordered that she teach them until they had memorized her [songs]. Then they sang to him that day [only] her songs and he bestowed on her a generous reward and conferred on [the other women] gifts as well.28 In another competition, al-​Hakam challenged his qiyan to compose a song using a love poem by one of the “Arab lovers of ancient times” that would communicate the poet’s situation to him and bring the poet’s state into his own heart. Bahja eventually won the competition. The Emir was ecstatic and declared, “This is what I have been striving and searching for!” whereupon he “rewarded her for all that she sang and granted her all that she desired.”29 The image that emerges from these six earliest biographies of Andalusian qiyan is of a ruler who, although publicly ruthless, sought companionship and solace with his qiyan. His qiyan are portrayed as having improved their position, financially and/​or socially, by being able to sense and respond to the ruler’s psychological needs. Unlike many later accounts, none of these anecdotes mentions the presence of other men, perhaps because of al-​Hakam’s secrecy regarding his drinking and love of music. Despite his bloody campaigns against all who rebelled against him, he is never portrayed as abusing or beating his female singers, although this is a trope that occurs somewhat frequently in the literature.30 In al-​Umari’s collection of singer biographies, the next three texts concern singers who belonged to the household of al-​Hakam’s son, al-​Mughira (d. 859). Shortly before the death of al-​Hakam, the leaders of al-​Andalus were ordered to swear loyalty to his designated successors. His older son, Abd al-​Rahman, was declared first, and his son al-​ Mughira, second, in line to the throne. Abd al-​Rahman became emir after his father, and al-​Mughira renounced all claims to the throne (although accounts differ with regard

The Qiyan of al-​Andalus

j 109

to whether he did this voluntarily ). In any case, Abd al-​Rahman lived for nearly four decades after his father’s passing and was apparently an avid patron of music. One of these texts is the biography of a male singer, Sulaym, in which a female qayna plays an interesting cameo role, emphasizing the East-​to-​West traffic of qiyan: Sulaym, a client of al-​Mughira, son of al-​Hakam: A man who was fortunate in his patrons and who ascended until he saw the stars draw near; they allowed him to mingle with them and considered him one of their own [lit. they mixed him in their nasab and reckoned him as one of their hasab]. He learned singing from emissaries who came to him [al-​Hakam] from the Christians. He [al-​Hakam] ordered [or: it was ordered] that they be delayed and he [Sulaym] was put in charge of them until their departure. He mastered their art and learned it thoroughly. Then al-​Mughira, son of al-​Hakam, was brought an Iraqi singing-​girl who had been selected for him from the women’s quarters . . .. She taught [Sulaym] [Iraqi] singing until he became proficient and he [then] added the Iraqi singing to what he had gathered [from the Christians]. There occurred sessions between the two of them in the gatherings of al-​Mughira more delicate than dawn breezes and more aromatic than fragrant trees.31 It was customary for the owners of qiyan to send them to more accomplished singers, male or female, for additional training. In an anecdote transmitted by Hammad, from his father, Ishaq, about how his own father, Ibrahim al-​Mawsili (Hammad’s grandfather), ran the finances of his household, Ishaq reports: “At one point we had eighty singing-​girls entrusted into our custody by his friends, and there was not one of them that he did not spend the same amount on in food, clothing, and perfumes, as he did for the most select of his own singing-​girls. And if a singing-​girl was returned to her master, he transported her and clothed her.”32 However, in this Andalusi example we see the reverse: al-​Mughira sent his new qayna to the male singer, Sulaym, so that she would teach him the repertory she had learned in Baghdad, which the author terms “Iraqi singing.” The result is that Sulaym became proficient in several different repertories: the repertory he sang originally at the Cordoban court, the new material he learned from the Christian emissaries, the singing he learned from the unnamed qayna from Baghdad, and possibly also a “hybridized” music in which he brought together one repertory with another, depending on how one interprets the phrase “he [then] added the Iraqi singing to what he had gathered [from the Christians].”33 Abd al-​Rahman II ascended to the throne at the death of al-​Hakam in 822. He is said to have been “[t]‌he first of the Marwanid [Umayyad] caliphs who gave glory to the monarchy in al-​Andalus.”34 In particular, he was known as an appreciator of women: [He] was among the caliphs the one who was most desirous of women and the fondest of sex, acquiring expensive slaves and selecting them after investigating their origins, class, education and conduct. He never took any that was

110 i

Dwight F. Reynolds

not a virgin, even if she surpassed the other women of her era in beauty and excellence.35 Unlike his father, Abd al-​Rahman is portrayed in historical accounts as being publicly known as a lover of music who kept a large number of qiyan, referred to collectively as his sitara (lit. his “curtain”), for they would perform before audiences of male listeners from behind a curtain, and regularly hired the best male musicians both to perform in his court and to instruct his female musicians. Isa ibn Ahmad al-​Razi [d. 989] said: The Emir Abd al-​Rahman ibn al-​Hakam was a great admirer of singing. He was enamored of listening [to it] and placed it above all his other pleasures. He patronized singers who competed in it and had a predilection for the best of them. He sought out the most skillful among them by inquiring after those of the highest rank [in this craft] and he directed his generosity exclusively to [those singers] with liberal gifts, extensive accommodations, and constant support. He turned his attentions to them, away from all that his palace and his private orchestra contained in the way of skillful female singers and excellent slave-​girls. He selected the best among these women [to send] to the male singers he had taken into his service, so that these latter could be their guides in this art, transmitting their artistry, in search of ever greater gratification in listening [to music], always guided by the pursuit of excellence. There are entertaining anecdotes about him in this regard.36 One of his favorite singers was Fadl, who was known as “the Medinese.” This sobriquet was given to qiyan not as an indication of their origin, but rather of where they were trained. Thus, the annals of Andalusian history are filled with many references to “Medinese” singers. Fadl was [e]‌xpert in the “science” of Medina, of superior beauty, skilled in singing and of perfect qualities. It is said that she had belonged to one of the daughters of [the caliph] Harun al-​Rashid, was Baghdadi of origin and education, and from there had gone to Medina where her level in singing improved. There she was bought for Abd al-​Rahman ibn al-​Hakam, together with her companion, Alam the Medinese, and others who were attached to her [her accompanists]. They gave their name to the “House of the Medinese” in the palace, earning great appreciation from the Emir Abd al-​Rahman for the excellence of their singing, their brilliant elegance, refined manner, in all of which Fadl, their leader, was considered superior. This garnered her the favor of the Emir, who, when he went out to some place of pleasure with these Medinese [singers], did not mix with them, out of consideration for his other women. Fadl continued to be his principal favorite and bore him a son, Umar b. Abd al-​Rahman, known as Abu al-​Qasim.37

The Qiyan of al-​Andalus

j 111

The third of these Medinese favorites was known by the name Qalam (“reed pen”): She was of Andalusian origin, a Basque captive, and daughter of one of their leaders. As a young girl she was taken East and ended up in Medina, and there she learned the art of singing until she mastered it, and was then purchased for Abd al-​Rahman, who took her as a concubine and with her completed his set of Medinese favorites. She bore him a son, Aban, known as Abu al-​Walid, who was an author and had no descendants. Muawiya ibn Hisham al-​Shabinisi wrote in his book: Qalam the Medinese, mother of Aban, son of the Emir Abd al-​Rahman was the most skilled of his qiyan in singing, the best in the knowledge of the art of composing melodies, the most versatile in the distinct varieties of song. She had as well a good memory for literature, was a good calligrapher, a reciter/​transmitter of poetry, a memorizer of historical/​literary accounts, and knowledgeable in all genres of literature and etiquette.38 Here, then, is the case of a woman taken captive at a young age who, either through her looks or her voice or perhaps both, was selected for training in Medina as a qayna. She was eventually transported back to al-​Andalus, where she was purchased by the Emir Abd al-​Rahman II and became one of his favorites. Three other qiyan during the reign of Abd al-​Rahman were also known as “the Medinese” and appear in an anecdote in which they encounter the famous singer Ziryab. Soon after they had been purchased for the Emir and arrived in the court, Ziryab asked to hear them sing. They sang nine songs and the Emir was greatly pleased, but Ziryab sat silently, giving no sign of approval. He then picked up the lute and sang the entire sequence of songs with different melodies and techniques. The Emir was so pleased that he was in a state of ecstasy (tarab) and his esteem for Ziryab became even greater than before, eclipsing his satisfaction with his newly acquired qiyan. Ziryab, however, requested that the Emir reward them, which he did.39 In general, music and the Emir’s qiyan took a much more public role in the court of Abd al-​Rahman II than they had in the time of his father, al-​Hakam I. This resulted not only in public performances (although often from behind a curtain), but also in the birth of at least two sons born to the singers Fadl and Qalam; they were, however, but two among his 86 offspring, of which 45 were sons! All these singers appear to have been trained in the East, and nearly all of them had received at least part of their training in the city of Medina. It is noteworthy that none of the anecdotes about these singers portrays them as responding to the emotional needs of their owner. Instead, only their skill and craft as musicians are featured, perhaps as a result of the very large number of singers and concubines in his household. Qiyan Trained in Cordoba In 822, the year of Abd al-​Rahman’s ascension to the throne at the death of his father, the legendary Ziryab is said to have arrived in Cordoba from Baghdad.40 In some sense we

112 i

Dwight F. Reynolds

might date the shift to the period when qiyan began to be trained in Cordoba to the years of Ziryab’s presence in the court. He is said to have passed on both a prodigious repertory as well as his techniques and style to a large number of female singers whom he himself trained. Similar to Ibrahim al-​Mawsili, Ziryab is portrayed as lodging dozens of singing girls in his home, where they were taught music. In one passage Ibn Hayyan lists his qiyan by name, some of whom belonged to him, but many of whom belonged to others and lived in Ziryab’s household while they were being trained: [Ziryab’s singing-​girls were]:  Nur, Itr, Saba, Khajal, Mukhariq, Ruhban, Muallila, Mukhtala, Khilata, Mutayyam, Ghalib, Badhl, Fadl, Sharaf, Ahyaf, Talal, Fawz, Raha, Rayya, Malak, Subh, Sha’n, Ridwan, Humam, Halhal, Amal, Dhayl, Nashr, Bida, Bazi, Aj the singing-​girl of Hashim ibn Abd al-​Aziz, whom Ziryab gave to him and was one of the most prolific transmitters from Ziryab and most skilled in the art [of singing], as well as Tarub, Ward the elder, and al-​Talibiyya, whose name I do not know. Shunayf [lit. “the little ornament or earring”] was a singing-​girl of Ziryab’s who lived for a long time after him such that when [later] singers were confused about something in his repertory—​and they differed in a great deal of it—​they would have recourse to her and seek her knowledge. They listened to her and found her to possess the oldest [i.e., most authentic] style and to be the most reliable transmitter of [his repertory], so they called her “the Imam.” They thronged to learn from her and referred to her regarding anything that troubled them. She had one counterpart [in this], Ward, whose nickname was “Paunch,” the singing-​girl of one of the nobles of the Quraysh. She was skilled in the craft [of singing], and many songs were transmitted by way of her from Ziryab and were corrected by her knowledge, though most of them were of the hazaj genre [a light song genre]. After the children of Ziryab, the three singing-​girls of Ibn Qalqal,41 the distinguished governor, excelled in the styles of Ziryab: Masabih, Ghulam, and Wasif. Many of the best female vocalists learned from them.42 Ibn Hayyan also notes that three additional singers, also known as “the Medinese,” were trained by Ziryab himself: Among those who acquired fame for transmitting [songs] from Ziryab and for their skill in his art were his three singing-​girls: Ghizlan and Hunayda, and they were his two “registers,”43 and their companion, Utba, all of whom sang the Medinese hazaj in his style and were known as “the Medinese.” I do not know if they were [his property] and he singled them out for this style so they became known for it, or if they had been brought to the court from Medina and were known thus for their origin. But all of them were in the palace of Ziryab.44 These anecdotes and additional information found in Ibn Hayyan allow us to ascertain that several generations of qiyan were trained in the repertory and style of Ziryab.

The Qiyan of al-​Andalus

j 113

Al-​Tifashi, a thirteenth-​century scholar, wrote, however, that the style of Ziryab had died out by the end of the eleventh century and had been replaced by a new style created by Ibn Bajja (d. 1139), who “combined the songs of the Christians with those of the East, thereby inventing a style found only in al-​Andalus, toward which the temperament of its people inclined, so that they reject all others.”45 An anecdote preserved in a collection of Andalusi biographies offers a brief glimpse of the training of qiyan on a smaller, more domestic scale. In this brief account, a friend visits the home of the chief judge of Cordoba, Abu Muhammad Ibn Dahhun, and is surprised to hear singing from somewhere nearby. He asked Ibn Dahhun about this, supposing it to be coming from a neighbor’s house. Instead the scholar responded somewhat ruefully: It is [coming from within] my own house, indeed it is my wife, for she is a wealthy woman who purchases slave-​girls, raises them, and teaches them singing. I  am a poor man, my wealth has gone, and I am not able to support myself. So she takes care of me and clothes me, and in this way things are made easier for me.46 What is noteworthy here is not only that singing-​girls are being trained in a private home, but that the wife in question is clearly skilled in singing. Was she, herself, at one point, a qayna as well? Or does this brief account give an indication that even free women of the upper classes were at times musically trained? These questions cannot, in this case, be answered, but the incident gives at least an indication of the buying, selling, and training of qiyan that took place outside the central court. By the eleventh century, references to qiyan trained in the East become scarce. Cordoba was sacked in spring 1013 and the Umayyad caliphate came to an end. Centralized rule from the capital collapsed, and al-​Andalus broke up into a number of smaller kingdoms and principalities ruled by the muluk al-​tawaif (the “petty” or “factional” kings). Similar to the competition among city-​states during the Italian Renaissance, the collapse of central power led to a period of cultural florescence due in part to the multiple sources of patronage now available to scholars, poets, and artisans of all sorts. Several anecdotes preserved in the Dhakhira of Ibn Bassam give indications that this was also a golden age for the training of qiyan and a period during which Andalusian qiyan were sometimes skilled in a remarkably diverse set of fields, including not only the performance arts, but also the physical sciences, Quranic recitation, and even weaponry. Ibn Bassam provides an account of Hudhayl ibn Razin, the founder of a short-​ lived dynasty, the Banu Razin, who ruled a small kingdom from the city of Sahla (known in Spanish as Albarracín, from the dynasty’s name) in Aragon between the sacking of Cordoba and the arrival of the Almoravids. Ibn Bassam quotes his earlier counterpart, Ibn Hayyan, to the effect that Hudhayl was handsome, well mannered, a good companion, manly, and unmatched among the emirs of al-​Andalus in his eloquent speech. He was also, among his peers, the prince who spent the most money on purchasing musical instruments, clothing, and qiyan, and was even criticized by other princes for the high

114 i

Dwight F. Reynolds

prices he was willing to pay, perhaps because by doing so he drove up the market. In one case, he paid 3,000 dinars for an extraordinary qayna, whom he purchased from a certain Ibn al-​Kattani (about whom, see the following quote): She was unique among the qiyan of her age. She had no peer in her class, no one had seen [a woman] more cheerful, more gracious of movement, gentler of gesture, sweeter of voice, better in singing, more excellent than her in writing, more skilled in calligraphy, more refined in manner, and more possessed of all that is good and could be desired. She was free from grammatical error in all that she wrote or sang, with a thorough knowledge of medicine, and her conversation extended even to natural history, the anatomy of the internal organs, and other things about which even many of those who claim to be specialists in those fields do not know well. [Her skills included] the manufacture of weapons [al-​thuqaf—​lit. the art of straightening of weapons such as lances, spears, and javelins], as well as fencing and sparring with swords, spears, and sharpened daggers, and other entertaining pastimes. In this, no one had ever heard of any who was her like, her equal, or her match. To complement her, [Hudhayl] purchased many beautiful female slaves, famous in Quranic recitation, whom he sought from every direction. His sitara [orchestra] was the most glorious of all of the “petty kings” of al-​Andalus. I was told that he had gathered 150 concubines and 60 male castrated Slav servants, the likes of whom had never been assembled by any of his peers.47 Henri Pérès doubted that qiyan were ever trained to recite the Quran, and therefore emended the word tajwid (Quranic recitation) to tajrid, translating this as some sort of armed combat.48 But an anecdote from al-​Humaydi demonstrates that some qiyan were, indeed, trained in Quranic recitation. The qayna described next was apparently trained in Medina and was probably brought to al-​Andalus by her owner, who was originally from the Eastern Mediterranean [the Mashriq], as a member of his household49: A man from the Mashriq known as al-​Shaybani arrived in al-​Andalus and settled in Cordoba on the banks of the river at al-​Uyun. The chief judge, Ibn Salim, when out one day on an errand and a rainstorm overtook him and forced him [to seek shelter], so he led his mount into the entryway of al-​Shaybani and halted him there. Al-​Shaybani welcomed him and asked him to dismount, so he dismounted and al-​Shaybani then ushered him into the house and the two of them spent some time talking. Then al-​Shaybani said to the judge, “My God grant the judge prosperity! I have a Medinese singing girl—​nothing sweeter than her voice has ever been heard, and if you permit it, I will have her recite for you a section from the book of God—​Exalted and Almighty—​and some verses of poetry. So the judge said to him, “Go right ahead.” So he ordered the singing-​girl to do so and she recited [Quran] and then sang [verses of poetry]. The judge was gratified and

The Qiyan of al-​Andalus

j 115

greatly pleased with this. In his sleeve he had some dinars, so he took them out and placed them under the cushion he was sitting on, though al-​Shaybani did not notice this. When the rain let up and the judge mounted and said farewell to al-​ Shaybani, the judge called out to him and the singing-​girl saying, “I left something there, and it is for the singing-​girl, to help her with whatever she needs.” Al-​ Shaybani exclaimed, “God be praised! [meaning, “You didn’t need to do this”].” But the judge replied, “You have no choice but to accept it. I have sworn that you will do so.” When al-​Shayban went back into the house and took the purse, he found in it twenty dinars! If Hudhayl ibn Razin purchased many additional qiyan to complement his prized female singer, it certainly seems more likely that he acquired beautiful female slaves skilled in Quranic recitation than in the martial arts. The previous description of the extraordinary qayna whom Hudhayl ibn Razin purchased from Ibn al-​Kattani is presented as credible. But in another passage, Ibn Bassam offers more information about this same slave merchant, noting that he was a bit of a scoundrel and a charlatan: Muhammad ibn al-​Kattani al-​Mutatabbib50 was unique in his time, a “sly dog” of his era, a seller of qiyan in the market, who would teach them Quran and classical Arabic and other things from the literary/​courtly arts. He used artful means and was given to frequent use of all manners of skillful lies and falsehood. He would on occasion compose treatises and attribute them to qiyan, whom he then sold at a higher price. In the entry on Ibn Razin we [= Ibn Bassam] mentioned that he sold him a qayna for three thousand dinars, according to what Abu Marwan [Ibn Hayyan] recounted. There is a passage on a scrap of paper by Ibn al-​Kattani in which he describes how he taught his qiyan: I am capable of calling forth [intelligence] from stones, to say nothing of what I  can do with dullards and ignoramuses! Consider that I  have in my possession four Christian women who were but yesterday ignorant, and now are learned, wise, schooled in logic, philosophy, geometry, music, skilled in the use of the astrolabe, in astronomy,51 astrology, grammar, prosody, literature, and calligraphy. The proof of this for those who do not know them are the large collections which have appeared in their script about the meanings of the Quran, its unique vocabulary, and other topics in the field of Quranic studies, as well as the sciences of the Arabs in rising and setting of the stars,52 prosody, grammar, and books of logic and geometry, and all the remaining types of philosophy. They mark with grammatical case-​endings everything that they copy, and they correct [the text] according to its meaning, due to their careful revision of it. This is the greatest testimony that I am indeed unparalleled in my era, one of a kind; I have exhausted this epoch of experience and this era of enlightenment. So recognize—​may God bless you!—​my rank, and give me my due, and do not hope of finding an expert like me, or one endowed with my

116 i

Dwight F. Reynolds

qualities, even if you should travel the horizons, inquire of all your friends, walking all the way to Iraq, [searching] alleyway by alleyway.53 Obviously, we have here a case of hyperbole, but the interest of the passage lies in that fact that what he is boasting about is the scholarly merits of these qiyan—​the breadth of their knowledge and education—​rather than their looks, their voices, or their musical repertory. He is even accused of having falsified some of these skills by, for example, composing treatises himself and passing them off as creations of his qiyan, so whether his slaves ever reached the levels of education he claims for them, the account only makes sense if there was indeed a market for qiyan who had such skills and that, by falsifying such achievements, one could in fact earn more money. Rich men clearly purchased such female slaves, even if they sometimes later learned they had been duped and the slaves did not possess all the qualities advertised. The remarkable element of these anecdotes, and others from this same period, is the focus on the intellectual achievements of qiyan. It appears that for several centuries in al-​Andalus, the value of these women could be increased greatly by serious study and the mastery of a number of fields that were normally restricted only to men. Although it may not have been the norm for an average singing girl also to be educated in the sciences and act as a copyist and corrector of texts—​let alone be skilled in the crafts of weaponry—​ the fact that this existed at all demonstrates that some of these female slaves had reason and opportunity to develop their minds and scholarly skills, and that their owners had motives (financial, among others) to allow them do so. Qiyan Trained in Seville Only one of the Andalusian singer biographies found al-​Umari’s Masalik is devoted to a qayna, Suada, from the court of al-​Mutamid of Seville (r. 1069–​1091), and unfortunately this text is rather brief and uninformative. After the usual praise of her beauty there is an anecdote in which al-​Mutamid decided to move all of his concubines from one town to another. They traveled at night and he sent Suada two verses of poetry about this night journey, asking her to set it to music, which she did. A second anecdote begins with the text of one of her famous songs and then recounts the story behind it, in which al-​Mutamid went to visit a favorite concubine, but found her ill and feverish.54 From this period, however, comes a fascinating account of the buying and selling of qiyan found in a surviving fragment from a much larger, no longer extant, work by the thirteenth-​century writer al-​Tifashi. After giving a long list of the most famous songs of his era, he notes the following: Among the cities of al-​Andalus, this music is primarily located in Seville where older professional women singers teach singing to female slaves whom they own,

The Qiyan of al-​Andalus

j 117

as well as to hired, mixed-​race [muwalladat] female servants. These girls are sold from Seville to the various rulers in the Maghrib and Ifriqiya. A singing-​g irl is sold for a thousand Maghribi dinars, sometimes more, sometimes less, according to her singing, not her face, and she must be sold with a document that lists what she has memorized, and most of these [songs] are among the poems that we have mentioned above. Some of these are “light” songs that are good for the beginning of a performance, and others are “heavy” songs that a skilled performer only sings at the end [of a performance], such as “al-​Kumayt Complained”55 and “The Palm Tree and the Palace,”56 for these songs and others like them are only sung by experts. For that reason they are considered obligatory for the sale, and the lack of such songs would necessarily lower the price. Among the Andalusians a singing slave girl must also be skilled in calligraphy. She must submit the document listing what she has memorized to someone who makes sure that it is all in good Arabic. Her buyer reads what is in her document, indicates what he wishes to hear, and she sings it for him on the instrument specified in her bill of sale. She might also be skilled on all of the different instruments, as well as all forms of dance and shadow-​puppetry,57 and possess her own instrument and her own servant-​g irls who accompany her with percussion and wind instruments, in which case she is known as “complete” and is sold for several thousand Maghribi dinars.58 Ahmad al-​Tifashi was originally from Tunisia, but traveled to the East and eventually settled in Cairo. Like the authors of so many of the other surviving works about al-​Andalus (al-​Maqqari, al-​Humaydi, Ibn Dihya, and others) he was an Andalusian who was writing for an Eastern readership. It is interesting, therefore, that he specifically says that, “among the Andalusians a qayna must also be skilled in calligraphy,” which may imply that this was a skill less commonly found among qiyan in the Mashriq. He goes on to note that she might also be skilled on several instruments, as well as in dance and shadow puppetry. Yet, he also mentions that she might have to submit her “register” to someone who could correct her Arabic, if this were faulty, implying that even during the thirteenth century, qiyan in al-​Andalus were still being acquired from non-​Arabic-​speaking communities. Equally worthy of note is the description of a “complete” qayna, who is purchased with her own entourage of musicians who accompany her. In other words, in passages where a text says that a nobleman or an emir purchased a qayna, this may in fact mean that he purchased a small ensemble in which the qayna was the lead singer and instrumentalist. Conclusion The documentation on qiyan in al-​Andalus is fragmentary, and we are safer when characterizing individual sources rather than making broad historical generalizations. The

118 i

Dwight F. Reynolds

earliest biographies of qiyan in al-​Andalus, preserved in al-​Umari’s fourteenth-​century Masalik, place an emphasis on the ability of a qayna to sense her master’s emotional state and respond to his needs in poetry and song. Indeed, a number of the anecdotes from this text demonstrate that a qayna’s skill at doing so was key to both monetary reward and upward social mobility, in the sense of being moved into the ranks of her master’s wives and the mothers of his sons. This focus may represent the taste of the compiler of the biographies, the psychological needs of this particular ruler, al-​Hakam I, who enjoyed music and wine in private, or it may represent a quality that was appreciated more broadly. The many anecdotes about qiyan owned by al-​Hakam’s two sons, al-​Mughira and Abd al-​R ahman II, offer few examples of this type of emotional intimacy and instead focus on their musical skills. During the ninth century, the famous musician Ziryab trained a large number of qiyan, who then went on to train succeeding generations of singers, establishing a particular style that lasted for two centuries or so, according to Ahmad al-​Tifashi. In addition, there is evidence that young girls were being bought, trained as qiyan, and later sold for substantial profit even in smaller households. A rather unexpected development is found during the tenth to twelfth centuries—​ namely, that at least some Andalusian qiyan appear to have possessed remarkable intellectual and scholarly training. Although it is difficult to determine how many such qiyan there were, some had skills far beyond those needed to fulfill the social function of entertainment (music, song, the recitation and composition of poetry, and charming conversation) and were educated in scholarly fields otherwise deemed to be exclusively the domain of men. Although there is probably no direct link, this is also the period in time when the two strophic song forms that subsequently become the best-​known cultural products of al-​Andalus, the muwashshah and the zajal, emerge. Perhaps it is because of these new forms that we, for the first time, hear of the exportation of qiyan from al-​ Andalus to the nobles and rulers of North Africa. By the thirteenth century, mentions of qiyan in both Eastern and Andalusi sources become very rare, to the point that the institution as a whole seems nearly to have disappeared by the fourteenth century. There were almost certainly still female singers, but the international trade and traffic in female slaves who had received specialized training in singing, poetry, the playing of musical instruments, calligraphy, the memorization of literary and historical anecdotes, and other topics that would make them scintillating conversational partners seems to have come to an end. Acknowledgment I thank Julia Bray, Eckhard Neubauer, and Everett Rowson for their extremely helpful comments and suggestions on early drafts of this chapter.

The Qiyan of al-​Andalus

j 119

Notes 1. The terms qiyan (less frequently, qaynat; sing., qayna) and jawari (sing., jariya) are often used interchangeably in medieval texts, with the distinction, however, that the term qayna refers unambiguously to a female slave who performed musically, whereas the term jariya could refer to other types of female slaves, concubines, and servants who were not necessarily musicians. For simplicity’s sake, only the term qayna (pl., qiyan) is used in this chapter, except in citations of texts that use the terms jariya or jawari. 2.  For an overview, see Pellat, “Kayna.” 3.  The work as a whole has not survived, but various fragments have been published in facsimiles, Arabic editions, and in translation. The section referred to in this chapter is Ibn Hayyan, al-​Sifr al-​thani min kitab al-​muqtabis. 4.  See al-​Tanji, “Tara’iq.” 5.  The 10th volume of Masalik al-​absar is available in three editions:  Ed. Abd al-​Allah ibn Yahya al-​Sarihi (UAE:  al-​Majma al-​Thaqafi, 2003); Ed. Ghattas Abd al-​Malik Khashaba (Cairo:  Dar al-​Kutub wa’l-​Wathaiq al-​Qawmiyya, 2005); and Ed. Kamil Sulayman al-​Juburi (Beirut: Dar al-​Kutub al-​Ilmiyya, 2010). 6. See Delgado and Vilchez, 1:342–​343. 7. Perhaps the best coverage of this topic up to this point is Peres, Poesie, 383–​393. 8. See Muntane, La musica. 9. Al-Isbahani, Aghani, 5:170. 10. Al-Isbahani, Aghani, 5:170. 11.  Al-Jahiz, Epistle, 31–​32. 12. Al-Jahiz, Epistle, 34. 13.   Al-​Jahiz, Epistle, 36–​37. 14. Nubdhat, Arabic text, 5; Spanish text, 7. Note that the Spanish passage “cantoras y danzaderas” has added a reference to dancers; the Arabic text makes no mention of dancers. 15. There are a handful of references to concubines and slave girls in texts that deal with these early years. For example, in History of Ibn al-​Qutiyya, the daughter of Yusuf al-​Fihri, the last governor of al-​Andalus, gives the newly arrived Umayyad emir, Abd al-​Rahman I, a slave girl named Hilal, who later became the mother of his first-​born son, Hisham. None of these women, however, are said to have been singers. See James, Early Islamic Spain, 71. 16. Al-​Tanji, “Al-​Tara’iq wa-​l-​alhan al-​musiqiyya,”114. 17. For a discussion of al-​Umari’s sources, see Reynolds, “Lost Virgins.” 18.  The Aya Sofya and Topkapı manuscripts list the first two singers as Aziz; in the later Bibliotheque Nationale manuscript, the second name has been changed to Gharir (a change of the placement of dots in the Arabic script). 19. Black was the color of the Abbasid regime that overthrew the Umayyads of Damascus and later moved the capital of the Islamic Empire to Baghdad. 20. Al-Umari, Masalik, ed. Khashaba, 466; Masalik, ed. al-​Juburi, 414–​415; Masalik, ed. al-​Sarihi, 576–​577. The three editions differ on many points, but I have, in general, followed Khashaba, who gives the soundest readings. 21. Al-Umari, Masalik, ed. Khashaba, 467; Masalik, al-​Juburi, 415–​416; Masalik, al-​Sarihi, 577–​578.

120 i

Dwight F. Reynolds

22. Al-​Umari, Masalik, ed. Khashaba, 468–​469; Masalik, al-​Juburi, 416–​417; Masalik, al-​ Sarihi, 579–​580. 23. Al-Umari, Masalik, ed. Khashaba, 471–​472; Masalik, al-​Juburi, 419–​420; Masalik, al-​ Sarihi, 580–​581. 24. My thanks to both Julia Bray and Everett Rowson for their assistance in understanding these two poems. In the first, al-​Hakam has selected four nonconsecutive verses from a much longer poem, in which the poet Abu Tammam, Diwan, lines 21 (584), 25, 26 (586), and 27 (587), is boasting about the generosity and good deeds of his tribe, the Tayyi. 25. Translation by Julia Bray (pers. comm., August 27, 2014). 26. Translation by Julia Bray (pers. comm., August 27, 2014). 27. Al-Umari, Masalik, ed. Khashaba, 476; Masalik, al-​Juburi, 422–​423; Masalik, al-​Sarihi, 588. 28. Al-Umari, Masalik, ed. Khashaba, 476–​477; Masalik, al-​Juburi, 423; Masalik, al-​Sarihi, 588–​589. 29. Al-​Umari, Masalik, ed. Khashaba, 470–​471; Masalik, al-​Juburi, 418–​419; Masalik, al-​ Sarihi, 588–​582. 30. The final pages of Ibn al-​Qutiyya’s History, for example, recount the sorry fate of Bazia, a qayna so famed for her singing that she was known as the “Imam.” She was a slave of the Abd al-​Rahman III (r. 912–​961). When the latter’s brother, Ibrahim, came to visit him, he called on Bazia, who was behind a curtain, asking her to sing for his brother. She sang the verses, “It delights my heart to see your visitors /​Having him who loves you near, augments my pleasure.” After the brother and his retinue exited, Abd al-​Rahman had her whipped for having chosen those verses, declaring that they showed that she was secretly in love with his brother. She narrowly escaped being beaten a second time when Ibrahim next visited the court (see James, Early Islamic Spain, 141). Another example is the singing girl, a gift from Yusuf ibn Tashfin, who was thrown into a river and drowned by al-​Mutamid when he took offense at some verses that she sang for him (see Peres, Poesie, 13, and al-​Maqqari, Analectes II, 620–​621). 31. Al-Umari, Masalik, ed. Khashaba, 486–​488; Masalik, al-​Juburi, 431–​433; Masalik, al-​Sarihi, 601–​604. The remaining portion of the biography contains the lyrics to two of Sulaym’s compositions—​the first by the poet al-​Ahwas ibn Muhammad al-​Ansari, the musical setting of which was in the rhythm of thani al-​ramal, and the second is labeled merely madih (i.e., praise to the Prophet Muhammad)—​followed by an anecdote in which al-​Hakam asks Sulaym to compose a song to verses by al-​Musayyib b. Alas ibn Malik, which he does and thereupon receives as a reward a shawl (mitraf) in violet silk lined with marten fur, and 200 dinars. 32. Al-​Isbahani, Aghani, 5:164. 33. Thus, in the Aya Sophia and Bibliotheque Nationale manuscripts as well as Sarihi, and al-​ Juburi; the word ma‘a is missing in Khashaba, perhaps because it appears in garbled form in the Topkapı Ms: wa-​jama‘a min al-​ghina’ al-​‘iraqi mahma‘an [sic] jama‘a. For a more detailed analysis of this passage, see Reynolds, “Music, Poetry, and Lingua Franca”. 34. Ibn Hayyan, al-​Muqtabis, 280. 35. Ibn Hayyan, al-​Muqtabis, 303. 36. Ibn Hayyan, al-​Muqtabis, 307. 37. Ibn Hayyan, al-​Muqtabis, 306. 38. Ibn Hayyan, al-​Muqtabis, 306–​307. 39. Ibn Hayyan, al-​Muqtabis, 331. 40. The figure of Ziryab that is most widely known, which springs from al-​Maqqari’s Nafh al-​ tib, is more legend than fact. The earliest accounts portray him as an excellent musician, but not

The Qiyan of al-​Andalus

j 121

much more; these sources contain a number of disdainful, even insulting, anecdotes about him. Over the course of several centuries, his biography was added to by later writers and, after the fall of the Umayyads, he emerged as a symbol of their glory. See Reynolds, “Al-​Maqqari’s Ziryab,” and Davila, “Fixing a Misbegotten Biography.” 41. Possibly Qulqul; al-​Maqqari, Nafh, 3:131, cites him as Abu Hafs Umar ibn Qalhil. 42. Ibn Hayyan, al-​Muqtabis, 329. These are “stage names” chosen for their sound and image, which are themselves of some interest. The list begins with names such as Light, Perfume, Sheba, Modesty, and so on. The final singer is referred to only as al-​Talibiyya, a nisba probably derived from her owner’s name (i.e., “the one owned by al-​Talibi”) and the author then adds that he cannot remember her stage name. 43. They were nicknamed the two “registers” [diwanan] because they would memorize the new songs that Ziryab composed in the middle of the night and he would consult them in the morning because he himself had forgotten the compositions. These nighttime compositions were said to have been inspired by the jinn. 44. Ibn Hayyan, al-​Muqtabis, 328–​329. 45. Liu and Monroe, Ten Hispano-​Arabic Strophic Songs, 42, and al-​Tanji, “Al-​Tara’iq,” 114–​115. 46. Abu Talib al-​Marwani, Kitab uyun al-​imama, 81–​82. My thanks to Maribel Fierro for calling my attention to this text. 47. Ibn Bassam, al-​Dhakhira 3:70–​71, and Peres, Poesie, 385. 48. Peres, Poesie, 385. 49. al-​Humaydi, Jadhwat al-​muqtabis, 43–​44, in the biography of Judge Muhammad ibn Ishaq ibn Salim Abu Bakr of Cordoba. 50. The term al-​mutatabbib in this context clearly means “the would-​be physician” or one who “passed himself off as a physician.” although it is not always used derogatorily. In some printings of Ihsan ‘Abbas’s edition of al-​Dhakhira there appears a footnote regarding the identity of Ibn al-​Kattani: “The biographical notice of Muhammad ibn al-​Hasan al-​Mudhhaji [?]‌al-​Kattani al-​Tabib [the physician] appears in Tabaqat Said, Ibn Abi Usaybi`a, al-​Safadi, and Jadhwat al-​ muqtabis under the name Muhammad ibn al-​Hasan and Muhammad ibn al-​Hasayn—​see the Introduction to Kitab al-​Tashbihat. But I  doubt that he is the same figure as the merchant of qiyan whom Ibn Bassam describes here as ‘given to frequent use of all manners of skillful lies and falsehood.’ ” (My translation, Ibn Bassam, 319.) 51. Lit., skilled in applying equations to the mean positions of the sun and planets to derive their true positions (see “Ta‘dil,” Encyclopaedia of Islam). 52. The ancient Bedouin system of calendrical units (approximately 28) marked by the setting of different stars (see “Anwa‘,” Encyclopaedia of Islam). 53. Ibn Bassam, al-​Dhakhira, 3:202, and Peres, Poesie, 384. The entire passage is written in rhymed prose (Ar. saj); the rhymes of the final phrases have been included to give some sense of this style. 54. Al-Umari, Masalik, ed. Khashaba, 499–​500; Masalik, al-​Juburi, 443–​445; Masalik, al-​ Sarihi, 616–​618. 55. Lyrics by Ibn Abi al-​Rabi (d. 940), perhaps the best-​known poet of the Umayyad period. 56. Lyrics by Abu Qatifa Amr (d. before 73/​693), the son of al-​Walid ibn Uqba (d. 61/​680), and melody by Ma`bad (d.c. 125/​743). 57. Shadow puppetry is an art form still practiced in the Middle East and as far east as Indonesia. The audience sits in front of a cloth screen and the flat puppets are manipulated by the puppeteer with backlighting so that the shadow of the puppets is projected onto the screen.

122 i

Dwight F. Reynolds

58. al-​Tanji, “Al-​Tara’iq,” 103, and Liu and Monroe, Ten Hispano-​Arabic Strophic Songs, 37–​38. The final phrase in brackets has been crossed out in the original manuscript.

Bibliography Primary Sources Abu Tammam, Habib ibn Aws. Diwan Abi Tammam bi-​sharh al-​Khatib al-​Tibrizi. Vol. 4. Edited by Muhammad Abduh Azzam. Cairo: Dar al-​Ma‘arif, 1965. Anonymous. Nubdhat al-​asr fi akhbar muluk Bani Nasr. Edited by Alfredo Bustani, translated by D. Carlos Quiros. Larache, Morocco: Artes Graficas Bosca, 1940. al-​Humaydi, Muhammad. Jadhwat al-​muqtabis fi dhikr wulat al-​Andalus. Cairo:al-​Dar al-​ Misriyya li’l-​Taʼlif wa’l-​Tarjama, 1966. Ibn Bashkuwal, Abu al-​Qasim. Kitab al-​Sila. Cairo: Dar al-​Misriyya li’l-​Talif wa’l-​Tarjama, 1966. Ibn Bassam, Abu al-​Hasan Ali. Al-​Dhakhira fi mahasin ahl al-​jazira. Edited by Ihsan Abbas. Beirut: Dar al-​Kutub al-​Ilmiyya, 1998. Ibn Hayyan, Mahmud Ali Makki, and Federico Corriente. Cronica de los emires Al-​Hakam I y Abdarrahman II entre los anos 796 y 847 [Almuqtabis II-​1]. Translated by Mahmud Ali Makki and Federico Corriente. Zaragoza: Instituto de Estudios Islamicos y del Oriente Próximo, 2001. —​—​—​. Al-​Sifr al-​thani min kitab al-​muqtabis. Edited by Maḥmud Ali Makki. Riyadh: Markaz al-​Malik Faysal lil-​Buhuth wa’l-​Dirasat al-​Islamiya, 2003. Ibn Sa`i, Abu Talib Ali ibn Anjab. Nisa al-​khulafa. Cairo: Dar al-​Ma`arif, 1960. al-​Isbahani, Abu l-​Faraj. Al-​Ima al-​shawa’ir. Beirut:  Alam al-​ Kutub:  Maktabat al-​ Nahda al-​Arabiyya, 1984. —​—​—​. Kitab al-​aghani. 23 vols. in 15. Cairo:  al-​Muassasa al-​Misriyya al-​Amma li’l-​Talif wa’l-​ Tarjama wa’l-​Tibaa wa’l-​Nashr, 1963–​1964. al-​Jahiz, Amr ibn Bahr. Risalat al-​qiyan. The Epistle on Singing-​Girls of Jāhiz. Edited and translated by A. F. L. Beeston. Warminister, UK: Aris & Philips, 1980. al-​Maqqari, Ahmad. Analectes sur l’histoire et la littérature des Arabes d’Espagne [Nafh al-​tob min ghusn al-​Andalus al-​ratib]. Edited by Reinhart Dozy. Amsterdam: Oriental Press, 1967. —​—​—​. Nafh al-​tib min ghusn al-​Andalus al-​ratib. Edited by Ihsan Abbas. Beirut: Dar Sadir, 1968. al-​Marwani, Abu Talib. Qit‘a min kitab uyun al-​imama wa-​nawazir al-​siyasa. Edited by Bassar Awwad Maruf and Salah Muhammad Yarrar. Tunis: Dar al-​Gharb al-​Islami, 2010. al-​Suyuti, Jalal al-​Din. Al-​Mustazraf min akhbar al-​jawari. Beirut: Dar al-​Kitab al-​Jadid, 1963. al-​Tanji, Muhammad ibn Tawit. “Al-​Tara’iq wa’l-​alhan al-​musiqiyya fi Ifriqiya wa’l-​Andalus.” Al-​ Abhath 21 (1968): 93–​116. al-​Umari, Ibn Fadl Allah. Masalik al-​absar fi mamalik al-​amsar. Edited by Abd al-​Allah ibn Yahya al-​Sarihi. UAE: al-​Majma al-​Thaqafi, 2003. —​—​—. Masalik al-​absar fi mamalik al-​amsar. Edited by Ghattas Abd al-​Malik Khashaba. Cairo: Dar al-​Kutub wa-​l-​Wathaiq al-​Qawmiyya, 2005. —​—​—​. Masalik al-​absar fī mamalik al-​amsar. Edited by Kamil Sulayman al-​Juburi, Beirut: Dar al-​Kutub al-​Ilmiyya, 2010. al-​Washsha, Muhammad ibn Ishaq. Das Buch des buntbestickten Kleides. Translated by Dieter Bellman. Leipzig: G. Kiepen­heuer, 1984. —​—​—​. Kitab al-​Muwashsha. Edited by Rudolf-​Ernst Brünnow. Leiden: Brill, 1886.

The Qiyan of al-​Andalus

j 123

—​—​—​. El Libro del Brocado. Translated by Teresa Garulo. Madrid: Alfaguara, 1990. —​—​—​. Le livre de brocart, ou, La societe raffine. Translated by Siham Bouhlal. Paris: Gallimard, 2004. —​—​—​. Al-​Muwashsha aw al-​zarf wa’l-​zurafa. Beirut: Dar Sadir, 1998.

Secondary Sources Caswell, Fuad Matthew. The Slave Girls of Baghdad. London: I.B. Tauris, 2011. Davila, Carl. The Andalusian Music of Morocco. Al-​ Ala:  History, Society and Text. Wiesbaden: Reichert Verlag, 2013. —​—​—​. “Fixing a Misbegotten Biography: Ziryab in the Mediterranean World.” Al-​Masaq 21, no. 2 (2009): pp. 121–​136. Delgado, Jorge Lirado, and Jose Miguel Puerta Vilchez, eds. Biblioteca de al-​Andalus. 7 vols. Almería: Fundacion Ibn Tufayl de Estudios Arabes, 2012–​2013. James, David. Early Islamic Spain: The History of Ibn al-​Qutiyya. New York: Routledge, 2009. Liu, Benjamin, and James Monroe. Ten Hispano-​Arabic Strophic Songs in the Modern Oral Tradition. Berkeley: University of California Press, 1989. Muntane, Maria del Carmen Gomez. La musica en la Casa Real catalano-​aragonesa durante los anos 1336–​1432. Barcelona: A. Bosch, 1979. Pellat, Charles. “Kayna.” In Encyclopaedia of Islam. 2nd ed., edited by P. Bearman et al. Leiden: Brill Publishing Online, 2012. Peres, Henri. La Poesie andalouse en arabe classique au XIe siecle. Paris: Adrien-​Maisonneuve, 1937. Reynolds, Dwight F. “Lost Virgins Found:  The Arabic Songbook Genre and an Early North African Exemplar.” Quaderni di Studi Arabi n.s. 7 (2012): 69–​105. —​—​—​. “Al-​Maqqari’s Ziryab:  The Making of a Myth.” Middle Eastern Literatures 11, no. 2 (2008): 155–​168. —​—​—​. “Music, Poetry, and Lingua Franca in Medieval Iberia.” Philological Encounters 2, nos. 1–​2 (2017): 76–​94.

6 The Ethnic Origins of Female Slaves in al-​Andalus Cristina de la Puente

i  Slavery, race, and gender are invariably joined in the Western imagination. Two broad events gave rise not only to a full historiography but also to a colorful and sensually rich iconography as well, one that did much to further a collective historical memory: the Mediterranean trade in captives during the Ottoman–​Spanish wars of the sixteenth and seventeenth centuries, and the Atlantic slave trade—​the historical, political, and social consequences of which are still very much in evidence today. It is thus widely held, and naively so, that slaves had to possess a skin color or be of a geographic origin different from that of their owners. It is perhaps too easily forgotten, however, that slavery, throughout the course of its long history, has not always been rooted in ethnic difference, or that it has sometimes had that connection, but only partially or occasionally. The risk is also that one ignores the fact that membership in a specific ethnic grouping, with its particular racial, linguistic, and cultural characteristics, did not necessarily have to be a reason for slavery or its opposite. Slavery, although a constant in human society to the present day, has adopted highly diverse forms, and often the differences in systems of slavery coexisted during the same historical period or, conversely, systems have experienced diachronic evolution that alters them substantially, even within a short time frame.1 A key question concerns how specific cultures describe and refer to slavery in written texts. Sources do not always reflect social reality, but rather, in many cases, an idealized image of reality. Arabic sources of the Islamic West often do not distinguish between male and female slaves, but refer to slaves (abid) generically. So, too, they often do not distinguish between domestic slaves and concubines, using, for example, the term jariya for both categories of women and, occasionally, also free young women. On the one 124



The Ethnic Origins of Female Slaves in al-Andalus

j 125

hand, concubinage in texts from al-​Andalus (Islamic Spain) seems usually interesting to their authors from one of two perspectives: First, from a legal point of view, in which lawyers try to resolve conflict situations related to coexistence within the home, often in relation to the legitimate wives of the master, or in cases of sales, where jurists need to describe the slave sold. On the other hand, historical chronicles mention concubines who have given birth to elite sons. In this case, as seen in this chapter, ethnicity is one of the few elements that allow the identification of the mother. The sources usually specify whether the caliph’s mother was free or slave, and mention her name and geographic origin. In the first case, lawyers deal with concubinage in a general way and typically do not give examples of real concubines, whereas in the second case we are dealing with individual women, about whom, however, the texts provide little personal or biographical information.2 There is a need to catalogue what is known of the ethnic origins of women slaves in al-​ Andalus (eighth–​fourteenth centuries ce)—​that is, the territories of the Iberian Peninsula under Muslim domination. This requires grappling with various methodological problems, because the topic is one for which Andalusian Arabic sources offer comparatively little information.3 The extant written evidence does not allow for a recreation of the overall history of ethnicity in al-​Andalus, nor can firm conclusions be reached regarding the ethnic origin of female or male slaves for any period or place in al-​Andalus. Nonetheless, a wide range of genres of Andalusian Arabic sources, which are plentiful and immensely rich where other types of information are concerned, do yield valuable data. The details referring to the ethnic origin of slaves appear mostly in isolated form and are few in number, but are not insignificant. The fact that they are omitted frequently is also significant, and the point deserves consideration as well. Texts are important because they provide information about slavery, but also because they omit some relevant details, sometimes on purpose. For various reasons, this project might seem overly ambitious, because of the chronological length of the period chosen, the long history of al-​Andalus, and the geographic complexity of the terrain. As is well known, al-​Andalus endured for nearly eight centuries, from the start of the eighth to the close of the fifteenth century (711–​1492 ce),4 and, indeed, for even a century more, if we take into account the communities of Muslims—​ the so-​called Moriscos—​who continued to live in Spain after that date until their definitive expulsion in 1609 ce. 5 The history of al-​Andalus was also neither politically nor administratively homogeneous. Andalusian territory changed hands any number of times as a result of both internal convulsions and external invasions. The territorial expanse of al-​Andalus also shifted during these 800 years, ostensibly because of the southward expansion of the Christian kingdoms during the so-​called Reconquest (Sp. reconquista). The size of al-​Andalus was reduced gradually and significantly, and thus, over long stretches of time, even during periods of relative stability, the border fluctuated in one direction or another.6 Last, Andalusian territory did not always remain cohesive, and the power of different polities was quite uneven. During the eleventh century, even, there were various Andalusian

126 i

Cristina de la Puente

kingdoms at one and the same time, governed by figures of different ethnic and political origin, some among them of slave origin, as in the case of certain of the Taifa kingdoms.7 The history of al-​Andalus, like that of all other territories of the Islamic world, must be studied, therefore, in terms of shifting geographic and political realities. Moreover, it is necessary to bear in mind its particular standing as a borderland; many Arabic sources describe al-​Andalus as the “land of jihad.” Various hadiths (Prophetic teachings) that are cited in later works refer to al-​Andalus as “the gate of Paradise” or “the house of jihad and the home of the ribat (frontier fortress).” Although these hadiths are also related to other places in the Islamic world—​North Africa for example—​it must not be forgotten that the westernmost extreme of the vast Islamic realm would live through incessant secular conflict with the Christians of the peninsular North, a changing and difficult relationship with huge cultural and social repercussions.8 All of these circumstances are to be taken into account when undertaking a study of the ethnic origin of women slaves in al-​Andalus. War against the infidel and the capture of slaves were closely interconnected activities. To approach the subject, we must certainly take into account the ethnic origin of the rulers and the characteristics of the administrative systems they imposed on their respective territories;9 and, second, the extent of their political, and thus military, power. Their capacity to conquer determined how and where captivity occurred during each historical period. These factors will be decisive in contextualizing the references to slaves in Andalusian Arabic sources. A related issue concerns the manner in which Andalusians conceived of different ethnic groups. Descriptions of peninsular Christians, for instance, are vague and have more to do with perceptions of dress and (a lack of ) personal hygiene than with skin color. Knowledge of the Other seems to have been slight and, grosso modo, Arabs divided the peninsular lands that did not belong to al-​Andalus into three large territories with blurred borders. They call the inhabitants of the northwest, “Galicians” ’ those of the central north, “Basques” (Vascones); and those in the northeast, “Franks.”10 The origin of individuals, both free and enslaved, becomes even more difficult to investigate as one tracks the Arabization and Islamization of the Iberian Peninsula. For one thing, at particular moments, a great many individuals of Hispano-​Roman and Berber origins claimed false Arab genealogies to gain access to prestigious social and economic rank. These many issues inform the following discussion. The evidence offered by the different genres of Arabic/​Islamic historiography is considered, drawing particular attention to the methodological problems that stem from their use to evince conclusions regarding the complex subject of the interconnection of race, gender, and slavery in al-​Andalus. Slavery in the Arabic Sources of al-​A ndalus Works either written by Andalusian authors or that mention al-​Andalus are numerous and belong to many diverse genres. During the past century or so, many of these works



The Ethnic Origins of Female Slaves in al-Andalus

j 127

have been edited carefully, often translated, and, of course, studied closely. The often impressive results provide us with a broad view of many aspects of Andalusian political, cultural, and social history. Nevertheless, and despite this abundance of material, the sources have limitations, which are, in most cases, common to Arabic texts from the premodern period from other regions of the Islamic world. Their authors were typically urban dwellers, attached to their specific elite, intellectual circles, composing works for their peers. This circumstance means there is scarcely any information about the Andalusian rural world or smaller population centers, about which often only the name is known. Furthermore, the authors tend to leave out information known to their readers, taking for granted facts and circumstances that would be a source of enrichment for us today. In relation to the ethnic origins of female slaves, further stumbling blocks arise. In some genres, for instance, as seen later in this chapter, it is clear that Andalusian authors often relied heavily on Eastern (e.g., Iraqi and Egyptian) sources. Thus, at times, it is difficult to work out whether the content consists of original narratives and descriptions from the Andalusian world or if anecdotes and texts from the Arab/​Islamic Near East were simply copied in an exercise of literary transmission well known in the Islamic field and well studied in the case of Andalusian historiography.11 The practice of slavery is reflected in all the genres, although the results are highly diverse depending on the type. Andalusian historical chronicles typically deal with elite politics; they provide important details about fortress life and governing elites, but say comparatively little about the rest of Andalusian society.12 The slaves described, therefore, are those of the court: concubines, servants, eunuchs, and others. This information, however, is never systematic; these individuals are scarcely ever the protagonists in such accounts, appearing usually only in passing. Interest in the origin of given slaves is shown on rare occasions, although enormous confusion still exists regarding the terminology used by the chroniclers. Debate surrounds, for example, the term saqlabi (pl., saqaliba), literally “Slav”; the word alludes, in other words, both to the geographic and ethnic origin of these slaves.13 Historians have long been inclined to think that the term designated palace eunuchs, and this must have been the case initially.14 But, systematic study of these individuals shows that, over time, the term saqlabi was used to identify “white” slaves, who performed important roles, regardless of whether they were eunuchs. In al-​Andalus, they were possibly captive Christians from the peninsular North. One of the saqaliba, Mujahid, after the fall of the caliphate, became king of the small eleventh-​century Taifa kingdom of Denia and had a son who succeeded him in office. There are almost no textual references that show that slaves were imported from such far-​off lands. Perhaps the only such reference comes from Eastern geographer Ibn Hawqal (d. after 973 ce), whose information on al-​ Andalus proves at times to be highly unreliable:15 One very well-​known export article consists of slave boys and girls who have been taken from France and Galicia, as well as slave eunuchs. All the eunuch slaves to

128 i

Cristina de la Puente

be found on the surface of the earth come from al-​Andalus. They are made to undergo castration near this country; the operation is carried out by Jewish traders. The slaves are descended from Japheth: their home country, which is vast, extends over a great length. The warriors of Khurasan come into contact with them in the region of the Bulgars. They are taken as prisoners to this province; their virility is left intact, and their bodily integrity is preserved. The slave territory is immense. The text is widely cited in modern scholarship.16 Unfortunately, however, there is little further evidence, whether in Arabic or Christian sources from the Iberian Peninsula, to support the claims contained in the passage. The term abid also raises questions surrounding the ethnic origins of slaves. The term appears in accounts of the Umayyad caliphate of Cordova to designate units of military slaves. Strangely enough, Western historians have typically translated abid in this context as “black slaves,” despite the fact that the color of their skin is not mentioned in any chronicle. In fact, careful reading of the historical sources shows that when Arab authors wished to indicate that a particular slave was black they typically added the adjective aswad.17 It may simply be that modern scholars have projected backward the modern sense of the term—​that is, in identifying all abid as black. When the chronicles mention female slaves, they typically do not specify origin, with the exception of the mothers of lords (amirs) and caliphs.18 In Andalusian courts, female slaves were often Berbers imported from North Africa and were also highly valued in the Middle East because they were stereotyped as being ideal sexual partners, whereas their Christian counterparts were particularly prized for their capacity for domestic work.19 Of the first Umayyad ruler, Abd al-​Rahman I (r. 756–​788 ce), for example, it was said that “his mother was an umm walad (mother of the master’s child) named Rah, a captive of the Zanata tribe”—​that is, a Berber of the Nafza tribe, taken to Syria.20 Likewise, the mother of Abd al-​Rahman II (r. 821–​852 ce) was “an umm walad named Halawa, a Berber muwallada” (an individual of mixed slave and free origin).21 Although there were free Berber women in al-​ Andalus, it is also known that slaves of this origin were traded at every point in its history.22 An especially striking anecdote is told regarding another Berber slave mother, Athl, whose son was the ruler al-​Mundhir (r. 886–​888 ce): His mother was a Berber named Athl. Her intention had always been to become the mother of a caliph, and so she despised and disdained members of her tribe. One day her maternal uncle seized her and took her to Cordova, where he sold her. Her buyer was Sakan, the mother of Hashim ibn Abd al-​Aziz, the vizier. Sakan sent Athl to her son Hashim, a gracious and handsome man, who found himself before a haughty and honourable slave who aspired to the highest rank. He felt affection for her and tried to approach her, but the slave turned away. He then attempted to lie with her, but she rejected him in disgust; he desired her, but she refused and told him: “I neither desire nor need men; I am not glad to be your slave or that of anyone



The Ethnic Origins of Female Slaves in al-Andalus

j 129

like you; I desire a caliph, for this belly must be made pregnant by a caliph and you are not one, or even the descendant of one.” This made him furious; he turned from her then called her name, but she did not reply. When he touched her, she moved away, so he struck her, making her bleed. Seeing the blood, she broke down in tears and said: “I do not believe that you will escape the hand of one whose mother was brought to such a state,” which made him laugh, and he left her. She gathered her clothes and headed for the house of Ibn al-​Salim, to whom she recounted what had occurred, and he informed her that she was free. The imam Muhammad23 heard about this and decided to marry her. Eventually he made her pregnant with al-​ Mundhir, who was to reign after him. That is how God fulfilled the desires of this woman. Al-​Mundhir was born seven months after his conception and he was the one who would eventually kill Hashim ibn Abd al-​Aziz.24 As in other cases, the anecdote about the maternity of Athl serves simply as a vehicle for the chronicler to present a premonitory tale in which the real protagonists are the prince (al-​Mundhir) and the favorite vizier of his father, Muhammad (r. 852–​886 ce)—​that is, Hashim ibn Abd al-​Aziz. On the one hand, the arrest and death of the vizier are justified by the ill treatment he meted out to the mother of the Umayyad ruler;25 and, on the other, the destiny of sovereigns is shown, as on so many occasions, to have been written in advance. For present concerns, it is interesting that the woman is mentioned to have turned her back on her tribe; her Berber origin is shown not to have stood in the way of her achieving high rank at court or giving birth to a prince. Other mothers of princes were Christians. Differing accounts exist concerning the mother of al-​Hakam I  (r. 798–​821 ce). The Dhikr bilad al-​Andalus26 states that “his mother was an umm walad named Zukhruf, who had been given as a present to his father by Charlemagne, the son of Pipino, when he signed a treaty with Abd al-​Rahman al-​Dakhil.”27 The work’s editor, Luis Molina, asserts, however, that this version of the story appears only in the one source and is contradicted by others. Indeed, elsewhere, we are told that Charlemagne had proposed a marriage covenant to her but it had been rejected.28 Molina considers that this must have been true, because the sources would not have omitted the news of a Frankish princess in the harem of Hisham I (r. 788–​798 ce). The mother of the first Umayyad caliph of al-​Andalus, Abd al-​Rahman III (r. 912–​961 ce), was also a Christian (rumiyya), named Muzayna, Muzna, or Hazm.29 In other instances, the name of the slave is mentioned, as in the case of Tahr, the slave mother of Emir Muhammad,30 who died in childbirth, but no reference is made to her origins31—​Ashshar, the mother of Abd Allah (r. 888–​912 ce),32 and Murjan, mother of al-​Hakam II (r. 961–​976 ce).33 In the case of the Almoravid rulers of the Maghrib, chronicles refer to various black concubine mothers who were the mothers of the rulers themselves or the concubines of their sons.34 References to female slaves outside of elite political circles are rare; rarer still are references to their origins. In a few cases, allusion is made to the possession of a particularly

130 i

Cristina de la Puente

valuable slave, this in anecdotes in which the aim is to demonstrate the power and wealth of the family to which the slave belonged. The best-​known example concerns Ibrahim ibn Hajjaj (d. 900 ce), a notable of Seville and Carmona; he has a female slave singer and poet from Baghdad named Qamar brought to him. The text quotes examples of her verse, and describes her as having been always sumptuously dressed.35 Matthew Caswell has studied the text in his work on the palace slave singers of Baghdad.36 The sources also indicate that, after the ninth century, the court of Abd al-​Rahman II (r. 822–​852 ce) acquired slave singers from the Islamic East. Al-​Maqqari states that the Umayyad ruler attached to his palace a wing called Dar al-​madaniyyat (“the residence of the women from Medina”).37 Part of the royal household, it housed slave girls imported from the city in question, among them Fadl and Alam, both of whom were of Baghdadi origin, sold to the Umayyad house in Medina.38 It was also during this period that Andalusian slaves were sent East for training—​a case in point, Qalam al-​Bashkunsiyya.39 In these instances, the references are to geographic, not ethnic, origin—​the latter being of no interest and scarcely ever receiving mention. These slaves were highly coveted because of their exquisite training as singers; it also appears that the model of the Abbasid court was of particular interest, with the more modest Andalusian court seeking to emulate the behavior of its counterpart in the Iraqi imperial capital. Information on slavery in biographical dictionaries, a genre that was widely disseminated in al-​Andalus and of which lengthy works of great value have been preserved, is unfortunately thin and purely anecdotal. The dictionaries are devoted mainly to charting the transmission of the Islamic religious and legal tradition through biographies of ulama (religious scholars, experts in the religious sciences). The authors tend to be interested mainly in the intellectual biographies of their subjects, leaving little space for vital personal questions. Few allusions to slaves occur apart from the occasional remark that a scholar possessed a male or female slave. Typically, these references concern the service provided by the slaves, with little further description of them as individuals. When authors highlight such information, it is because they deem it to be genuinely exceptional, although they might not express it directly. There is mention, for instance, in the biography of Abu Hafs al-​Qu`ayni, a man of letters, that he fell in love with a black slave.40 The anecdotal references to male and female slaves in biographical dictionaries, although scarce in number and scattered across works of many different periods, seem to furnish real information regarding Andalusian figures. Works of other genres, such as Andalusian literary sources, do not appear to deal in concrete social realities, reliant as they are on all manner of well-​worn literary convention. Literary sources, both in prose and verse, sometimes represent the odd slave, but often do so in a hackneyed way. In addition, these texts show a clear Near Eastern influence, so that it is hard to distinguish what is properly Andalusian from a literal copy of texts produced a great distance away. We may add as a generalization that they do not pay attention to the ethnic origins of the women slaves described, not even in the infrequent listings of their physical features.41



The Ethnic Origins of Female Slaves in al-Andalus

j 131

Arabic legal sources provide the greatest amount of information on slavery. These are texts produced by scholars of the Maliki tradition (“school”) of jurisprudence, the predominant Islamic legal tradition of al-​Andalus and North Africa. These works alone make it possible not only to classify categories of slaves but also to consider the obligations and rights that shaped their role in society. As largely normative texts, however, the legal works cannot be counted on to shed light on social patterns rooted in specific periods and locales.42 The very partial evidence contained in these works goes to such issues as the ethnic identity of slaves, for example, and the problematic issue of the forms of labor demanded of slaves.43 The question of ethnic origin is particularly difficult. Ethnicity, whether of slaves or free persons, has no relevance as a category in Muslim legal doctrine. Although other factors, such as gender or religious affiliation, were fundamental in determining the rights and obligations of the distinct categories of slaves and, consequently, for understanding the social roles they performed, ethnicity/​race lacked legal meaning because theoretically “all men are equal in the eyes of God.” These are issues, of course, that relate to the theoretical and ethical conceptions of slavery. The evidence in question typically occurs in passing details. In some instances, for example, the Maliki texts mention the color of slaves or their geographic origin as an example of what they wish to spell out in theoretical or doctrinal terms. One of the richest sources in this sense are the collections of notary forms—​that is, documents intended as procedural models (watha’iq). Details contained therein inform us about local customs and peculiarities.44 The jurist Ibn al-​Attar (d. 1008 ce), in a document dealing with the transfer of ownership of a slave (and thus his labor), includes the following passages: Salam sale contract model:45 Salam sale of some slaves for others So-​and-​so, son of So-​and-​So, has paid in advance (sallafa) So-​in-​so, son of So-​and-​so, his Galician (jilliqi) slave of the following name, [by profession] carpenter or bricklayer, in exchange for foreigners (ajami), or Berber or black slaves, whose descriptions are as follows. The debtor (musallaf) So-​in-​so has received the slave whose personal details were given in the moment of contract [owing in exchange] the two slaves mentioned supra whom he will have to deliver to [creditor] So-​and-​So on the first of x month in x year, in accordance with the traditional use of Muslims in relation with salam contracts concerning slaves. Creditor (musallif) [So-​and-​So] will have to be believed with regard to what he demands from [Such-​and-​Such], without need to swear an oath, in the event of legal proceedings being instigated. [The witnesses] testify to this. Then you [the notary] must complete the agreement. Before the date, the following must be said: “[This act takes place] in the presence of the slave delivered in salam, whose personal details have been given, and once he has acknowledged his condition as a slave vis-​à-​vis the creditor (musallim) [So-​and-​So] until the time of delivery (taslim) mentioned [above]. [This deed] is [drawn up] on x date. Legal context ( fiqh): The advance (taslif) of [a slave] for [others] is only lawful when [the latter pursue] different trades. For example, giving an artisan in exchange

132 i

Cristina de la Puente

for two slaves who are not, or two non-​artisan slaves in exchange for an artisan, under the obligation of describing both the slave and his trade. Should neither of them have a trade, and there only be differences in [the degree of ] their beauty, the advance payment of one for the other would not be lawful. [Accordingly], delivery will be made of the secretary, the accountant and the grammarian in exchange for slaves who are not grammarians, secretaries or accountants. [Likewise, there will be] advance payment of one female slave for another: [delivery will be made of ] the cook and the baker in exchange for two female slaves who are not qualified as cooks, or bakers. There [cannot be agreed on] an exchange of female slaves when [there only exist] differences in their beauty and their skin color (tabayun).46 The two related passages contain significant information. Although the notary documents are theoretical, as models to be used in relevant cases, the notary often makes evident the actual, practical circumstances to which the legal act applied. So, for example, in referring to the slave who is to be the object of exchange, the latter’s ethnic background is to be specified. Ibn al-​Attar, in drawing up the model, suggests that the most frequent origins among them during the tenth century were “Galician,” Berber, and black. The importance attached in these texts to the origins and physical features of slaves arises in the model documents used for other types of contracts as well. So, for instance, Ibn al-​Attar provides a notary model for the case of transferring the services of a slave, a practice known as ikhdam: So-​and-​so, son of So-​and-​so, grants to Such-​and-​such, son of Such-​and-​such, the services of his Galician slave [name provided], whose details are such-​and-​such (or his female Galician or Frankish slave, of this name and whose details are as such-​ and-​such). [This being] a valid transfer of the usufruct of services (ikhdam), without right of retraction or option.47 Race does not appear as a factor influencing the jurisprudence of the document, nor does it involve a legal principle. What the texts imply clearly, however, is the following. First, the beauty of the slave and his or her physical qualities influenced the way in which he or she was assessed, although beauty per se was not related directly to a particular ethnicity. Second, the slave’s ethnic origin served as a descriptor and, therefore, was used to classify the individual slave and, thus, distinguish him or her from other slaves. Typically, legal contracts require detailed descriptions to prevent fraud or a procedural error from taking place. In a fifteenth-​century claim document from Granada for the sale of a country house (masriyya),48 one reads the following description of individuals: In the name of God, the Merciful, the Compassionate. God bless our lord Muhammad and his family. The black man, Mubarak, a freedman of Muhammad al-​Shamhani, on behalf of himself and Faraj and Mas`ud, his two youngest sons,



The Ethnic Origins of Female Slaves in al-Andalus

j 133

who find themselves in his custody, and the school teacher, Ali ibn al-​Ahsan al-​ Husayni, on behalf of the black woman, Zumurrud (Emerald), a freewoman of the warden Muhammad al-​Balansi.49 The text has historiographical value because it is a real document and not a standard form or model to be filled in, like those collected in the books of watha´iq mentioned earlier. Extant documents of this kind come uniquely from fifteenth-​century Granada; these survive precisely because of the Christian conquest of the territory.50 This fragment, as in the previous cases, references a slave’s color as a distinctive physical feature. In some documents, however, the notaries indicate, by implication rather than design, that color was more than simply a way to describe an individual. These references suggest that social prejudices were at play that fell outside the concerns of theoretical jurisprudence but could have practical consequences when different legal acts were enforced. In such cases, ethnicity acted, if not as a determining factor, then certainly as a feature that was taken into account when determining the legality or illegality of the contract or transaction. Thus, in a form referring to a possible fraudulent sale of a slave by reason of belonging to a different ethnic grouping, Ibn al-​Attar proposes the possibility of someone holding a particular origin in abhorrence and recording this in a document or through an oath: If [the slave] was not of the type (jins) that it was held up to be [i.e., that it belonged to a “superior” type or group], there are no grounds for action [on the part of the purchaser], unless it was previously known that he detested this type, through an oath or something similar. If such should be the case, [the slave] was to be returned, even if its jins were better than the kind that motivated its purchase.51 Skin color was unquestionably a distinctive feature deemed relevant in a marriage. Notary Ibn Mughith of Toledo (d. 1067), for instance, records a legal discussion regarding whether a black woman can be repudiated after marriage when the husband did not previously know she was of that color.52 This is a sample case of hypothetical fraud; the bride was not described correctly before the marriage. Although the legal response is that such repudiation is unlawful because, according to Maliki doctrine, the only cause for inequality between spouses was to be religion, it can be seen in this case that jurisprudence acted as a brake on situations that must have occurred with relative frequency both where free women, as in possibly the case of this spouse, and slave women were concerned. Documentation from the kingdoms of Aragon and Castile during the medieval period is comparatively rich. Notary documents contained therein, for instance, provide very thorough descriptions of slaves and do so using a genuine classification of these individuals according to skin color: black, almost mixed race, mixed race, light mixed race, white, and so on. The issue arises whether, if real documents and not only models had been

134 i

Cristina de la Puente

preserved in al-​Andalus, there would not be similar documentation that provided precise descriptions of this kind, and not just of a general nature.53 Maliki legal texts included market treatises (hisbat al-​suq) as well,54 which contain significant information on the sale of slaves. A particularly interesting example was written in al-​Andalus by al-​Saqati, an early thirteenth-​century market inspector from Malaga. Pilar Coello has studied the text closely.55 It contains the usual section on the rules to be followed to prevent fraud in slave trading, but includes, as well, a separate, lengthy section that provides detailed information about different categories of slaves, their ethnic origin, and the usefulness they had in connection with their perceived origin. Coello shows, however, that the author of the text drew almost verbatim from the earlier treatise of Ibn Butlan al-​Baghdadi (d. 1066), the Risala fi shira’ al-​raqiq wa-​taqlib al-​abid, in which specific guidelines are provided for the purchase of slaves. The latter text is devoted in large part to the discovery of physical defects in slaves through physical examination.56 The following is a fragment of al-​Saqati’s text, and it must be treated with great caution where the history of al-​Andalus is concerned: Among the types of fraud and deception [by these merchants] is the selling of slaves of one type (i.e. origin) for another, those of one race for another. Much has been said about slaves (mamalik), their categories, appearance and nature, and what is useful about each kind, with all sorts of discussion on the subject. They say that the female Berber servant [is suitable for] physical pleasure; the Christian (rumiyya) for looking after money and accounts; the Turk for producing sons; the Ethiopian (zanjiyya) for breastfeeding; the Meccan for singing; the Medinan for physical appearance; and the Iraqi for gaiety and flirtation.57 The text shows the significance attached to ethnicity and race in determining the price of a slave and how deception in this area constituted fraud (thus rendering a sale null and void). In another passage in the same work, al-​Saqati states that it was a frequent occurrence for a slave to be sold as rumiyya—​in other words, as a Christian, when she was not of that religion.58 Al-​Saqati’s text also suggests clearly that slaves, and particularly in regard to their ability to perform different trades, were typecast (a practice one finds still today in countries with large migrant worker populations). The fragment, however, like references elsewhere to Hindu and Yemeni slaves, raises a serious question: Were Ethiopian, Meccan, and Turkish slaves traded in al-​Andalus? Not only does no other source provide evidence of this kind, but one has also to question whether “exotic” slaves would have been imported to the territory when Christian, black, and Berber captives were widely available; it is they who are referred to in the different categories of texts cited earlier. Not even in the Alcazar, apart from the saqaliba and singing slaves (qiyan) mentioned earlier, do references occur to slaves from such faraway places and, as has been shown, the word saqlabi soon ceased to designate Slavs as such, as occurred in Western languages. Also, in



The Ethnic Origins of Female Slaves in al-Andalus

j 135

regard to singers, it took little time before it was decided to send peninsular slaves to the East for their training. Slaves in al-​Andalus were typically the product of wars, either through direct capture or indirectly in the sense of being the descendants of captives; the majority of these slaves came from regions close at hand. This does not contradict the fact that the color of their skin or the stereotypes that were in circulation regarding their (possible) origins had a significant impact on the prices that were fixed for their sale, as is shown in short passages in different Andalusian sources. So, for example, the odd illustrative text from different historical moments demonstrates that the prices of Christian slaves sometimes plummeted, so plentiful were they. This occurred, for instance, under the government of al-​Mansur (d. 1002),59 chamberlain to the caliph Hisham II, who, at the end of the tenth century, waged numerous campaigns in the North, bringing to al-​Andalus great numbers of captives that were, in turn, sold in local markets. The devaluation of their price was so sharp that the chronicles mention it as a significant circumstance.60 Furthermore, in an extract from a legal agreement, Ibn Sahl (d. 1093), an eleventh-​ century jurist, discusses the case of a black woman who was returned to her seller because, 70 days after the sale—​that is, when the legal waiting period for having sexual relations with the new owner (istibra) had transpired—​she had not menstruated. The point is that Ibn Sahl specifies that she was a black slave whose selling price had been extremely high.61 The fact that her membership in a particular ethnic grouping might vary the price is confirmed by evidence contained in other types of sources. An anecdote is told, for example, of a Jewish trader who upped the price of the female slaves he was bringing from the Christian North when it became clear to him that the governor of the city of Merida, who would later become the Emir Muhammad,62 had taken a fancy to them.63 And, as Marín has pointed out, a treatise on medicine, Book of the Pillow by Ibn Wafid of Toledo (d. 1067), refers to a medicine designed to remove the paleness of the complexion of slaves64 (although, in this case, the reason might not be to lend them a more beautiful color, but rather a more healthy appearance). Conclusion Information on male and female slaves in Andalusian Arabic texts is scarce and fragmentary. It stands to reason, of course, that slaves would be acquired from regions closest at hand. Thus, it was that, until the mid eleventh century, the majority of slaves appear to have come from the peninsular North or were Berbers captured in North Africa.65 Later, the Christians became the victors and, consequently, captors of captives (from the twelfth century onward, references to “Saracens” abound in Christian sources). The Andalusians of the twelfth and thirteenth centuries, governed by North African dynasties, very likely supplied themselves in North African markets, with, presumably, a corresponding increase in the number of sub-​Saharans among the slaves.

136 i

Cristina de la Puente

This is information regarding captives. Unfortunately, we know little about individuals born into slavery. Nor do the sources refer to the color of freed men and women; in many cases, it seems, manumitted slaves did not stand out physically from the rest of the populace. In the case of female slaves, it is known that their integration into host families was greater through concubinage and acquisition of the legal status of umm walad, which, of course, required acknowledgment of paternity by the owner. Christian/​Latin sources, as noted earlier, reveal that female slaves ended up integrating to such a degree that the matter of their origins became blurred—​that is, as they merged with host families through marriage or manumission. Other kinds of analysis could be tested, although there is reason to be skeptical about the possible results. An onomastic study, for instance, of Andalusian women slaves ought to provide interesting information about their origins, yet the sources very rarely mention their names. When they do, especially in the case of slaves working in the court, everything seems to indicate that they were given Arabic names, either because they were born as slaves and their owner had named them or because their original, given names had been changed. Typically, as in the case of slavery in other societies, it was traders themselves who changed the names. Women slaves are frequently given a poetic name—​Qamar, Ajab, Subh, and so on—​which tends to indicate their slave condition, but only on rare occasions do the sources provide further information about the individual.66 In the case of Subh, a favorite of al-​Hakam II and the mother of Hisham II, the sources do add the appellation al-​bashkunsiyya—​that is to say, “Basque”—​which would suggest she was born in that region, where she would have been captured, or she was already born into slavery in the Fortress of Cordova and was the descendant of Basque slaves.67 Significantly, her brother, who also resided at the Cordovan court, is referred to as a fata (“page”), with no further information on his origins. This can be a clear example of how sometimes authors consider ethnicity interesting in relation to female slaves but not to male slaves. It is highly likely that Subh’s race was considered significant because she was a concubine and the caliph’s mother. Race probably had sexual connotations in Andalusian society, related directly to beauty and color, as in other Muslim societies. The critical question is why Andalusian sources tend to mention only in passing the ethnic origin of female slaves, and do not explain how this ethnic origin was perceived or appreciated socially. One reason could be that authors considered it unnecessary because prejudices about race, slavery, and gender were widely known. Nevertheless, the most likely reason has to do with the Andalusian authors themselves. We know most of them received a thorough education in Islamic sciences—​Hadith, law, Quranic studies, and so on—​and they were deeply influenced by Muslim religion and by religious social ideals, where ethnicity does not make juridical sense. Maliki jurists insist repeatedly that physical characteristics do not have legal consequences and it can be a reason for omitting descriptions or making comments against the ideals of religious jurisprudence.



The Ethnic Origins of Female Slaves in al-Andalus

j 137

Notes 1. On race and slavery in the Islamicate world, see Lewis, Race and Color, and Braude, “Cham et Noé.” 2.  On concubines and slave mothers, and their integration into the Muslim–​Andalusian family, see de la Puente, “Los límites legales del concubinato,” “Slaves in al-​Andalus,” and “Free Fathers, Slave Mothers and Their Children.” 3. A first result of this work was presented at a congress organized by Shaun Marmon at Princeton in March 2012. I take the opportunity here to express my thanks for the invitation to take part in the meeting, where I had the chance to discuss its contents with various specialists in the field, whose remarks and suggestions I am also sincerely grateful for. 4. Kennedy, Muslim Spain and Portugal; Constable, Medieval Iberia; Jayyusi, The Legacy of Muslim Spain. 5. Garrido García, “La esclavitud morisca en el Reino de Granada”; Benítez Sánchez-​Blanco, “El cautiverio de los moriscos.” 6. Manzano, “Creation of a Medieval Frontier.” 7. Wasserstein, Rise and Fall of the Party Kings. 8. On captives in al-​Andalus, see, for example, Vidal Castro, “El cautivo en el mundo islámico”; and de la Puente, “Mujeres cautivas en ‘la tierra del islam.’ ” 9. Fierro, “Genealogies of Power in al-​Andalus,” 29–​56. 10. Molina, “Las campañas de Almanzor a la luz de un nuevo texto,” 261, note 109; Carballeira Debasa, Galicia y los gallegos en las fuentes árabes medievales. 11. Makki, Las aportaciones orientales. 12. Linehan, History and Historians. 13. Meouak, Saqaliba, eunuques et esclaves à la conquête du pouvoir; Meouak, “Esclaves noirs et esclaves blancs en al-​Andalus umayyade et en Ifriqiya fatimide,” 47–​48. 14. Kentaro, “Slave Elites and the Saqaliba in al-​Andalus in the Umayyad Period.” 15. Ibn Hawqal, Surat al-​ard, 62. 16. See Trabelsi, “Commerce et esclavage dans le Maghreb oriental (VIIe–​Xe siècles),” 19. 17. Felipe Maíllo observes similar confusion on the translation of the word wasif. See his translation into Spanish of Ibn Idhari, Bayan al-​mugrib, 53 (Arabic text); de la Puente, “Sin linaje, sin alcurnia, sin hogar,” 181–​182. Meouak draws this distinction between black slaves (abid) and white slaves, but he does not give any example from Andalusi sources where the term abd was used for a black slave, see Meouak, “Esclaves noirs et esclaves blancs,” 25–​26. 18. Concerning the families of Andalusi Umayyads, see Vallvé, “Sobre demografía y sociedad en al-​Andalus.” 19. Marín, Mujeres en al-​Andalus. 20. See Dhikr bilad al-​Andalus, 90, 116. Luis Molina points out in a note that some authors made her a native of Nafza (Ibn al-​Abbar, Kitab al-​hulla al-​siyara´, 1:35). In the Bayan, 47, 49, 73, it says that she was called Rah or Ridah and that she came from North Africa. Also see Marín, Mujeres en al-​Andalus, 123. 21.  Dhikr bilad al-​Andalus, 112, 145. About the legal status of slave mothers, see Blanc and Lourde, “Les conditions juridiques de l´accès au statut de concubine-​mère en droit musulman malékite.” 22. Marín, Mujeres en al-​Andalus, 123.

138 i

Cristina de la Puente

23. The Umayyad Emir Muhammad of al-​Andalus (r. 852–​886 ce), son of Abd al-​Rahman II and father of the amir al-​Mundhir. 24. Dhikr bilad al-​Andalus, 126–​127, 161–​162; and 124–​158, where the author only mentions her name. 25. See the account of the imprisoning of the vizier and his death in Ibn `Idhari, Bayan al-​ mugrib, 2:115–​116, 118–​119, 190–​191. It is of interest to note that the vizier wrote poems from prison to his slave `Aj, some verses of which are cited. 26. See Dhikr bilad al-​Andalus, 90, 116. 27. Dhikr bilad al-​Andalus, 104, 133. 28. al-​Maqqari (d. 1041/​1632), Nafh al-​tib, 1:330. 29. Dhikr bilad al-​Andalus, 131, 169. Molina says that all the authors call her Muzna, except Ibn Hazm (d. 1064), who calls her Hazm. 30. See Dhikr bilad al-​Andalus, 126–​127, 161–​162; and 124–​158. 31. Dhikr bilad al-​Andalus, 120, 155. 32. Dhikr bilad al-​Andalus, 127, 163. 33. Dhikr bilad al-​Andalus, 141, 178; and Ibn Idhari, Bayan, 2:233. 34.  See the description of the women of this North African court in Marín (Mujeres en al-​Andalus, 127), who provides examples. 35. Ibn Idhari, Bayan al-​mugrib, 2:128–​129. 36. Caswell, Slave Girls, 137, where Ibrahim ibn Hajjaj is called an amir. 37. Al-​Maqqari, Nafh, III:140. 38. Caswell, Slave Girls, 265. See Ibn Hayyan, Muqtabis. 39. Caswell, Slave Girls, 265. 40. Ibn Bassam, Dhakhira. Even more extraordinary is the case of Abida, a black slave given as a present to Dahhun (d. 815 ce), who became his concubine and a wise transmitter of oriental prophetic traditions. See Ávila, “Las mujeres sabias.” 41. On women in the literary sources, see Del Moral, “Contribución a la historia de la mujer a partir de las fuentes literarias andalusíes.” 42. Cahen, “Considérations sur l’utilisation des ouvrages de droit musulman para l’historien.” 43. De la Puente, “Mano de obra esclava en al-​Andalus.” 44. Concerning Andalusian kutub al-​watha´iq, see Ortiz, “Formularios notariales de la España musulmana.” See also the description of these texts provided by Tyan, Le Notariat et le régime de la preuve par écrit dans la pratique du droit musulman; and Zomeño, “Abandoned Wives and Their Possibilities for Divorce in al-​Andalus,” 111–​113. 45. Salam is a type of contract according to which the price of a clearly defined item is paid in advance on the understanding that the contract will be executed and the sold item received at a later date. 46. Ibn al-​Attar, al-​Watha´iq wa-​l-​sijillat, 55–​57 (Arabic text). 47. Ibn al-​Attar, al-​Watha´iq, 115–​116. 48. Al-​Masriyya means a country house or a storeroom built on the upper floor of a house. 49. Seco de Lucena, Documentos arábigo-​granadinos, note 36; see also Zomeño, “Del escritorio al tribunal: Estudio de los documentos notariales en la Granada nazarí.” 50. Shatzmiller has studied these documents in Her Day in Court. 51. Ibn al-Attar, Watha´iq, 35. 52. Ibn Mughith, al-​Muqni` fi `ilm al-​shurut, 28–​29, 64. When this is a case of marriage, a free black woman would most usually be the subject. Regarding this assumption, see also Lewis, Race and Color, 92; and Marín, Mujeres en al-​Andalus, 128.



The Ethnic Origins of Female Slaves in al-Andalus

j 139

53. Ferrer i Mallol and Mutgé, De l’esclavitud a la llibertat. 54. On market treatises in al-​Andalus, see Chalmeta, El señor del zoco en España. 55.  Coello, “Las actividades de las esclavas según Ibn Butlan”; see al-​Saqati, Kitab fi adab al-​hisba. 56. Ibn Butlan, “Risala fi shira´ al-​raqiq wa-​taqlib al-​`abid.” 57. Al-Saqati, Kitab adab al-​hisba, 49. 58. Al-​Saqati, Kitab adab al-​hisba, 47–​53; and see Marín, Mujeres en al-​Andalus, 383. 59. Bariani, Almanzor. 60. De la Puente, “La caracterización de Almanzor: Entre la epopeya y la historia,” 388. 61. Ibn Sahl (d. 486/​1093), “Ahkam,” 559, in Marín, Mujeres en al-​Andalus. 62. See Dhikr bilad al-​Andalus, 126–​127, 161–​162, and 124–​158. 63. Al-​Nubahi, Marqaba al-​ulya, 56. 64. Marín, Mujeres en al-​Andalus, 384. 65. At an even later period, captures of Christians occasionally took place with economic repercussions that were of relevance to the chroniclers, such as the Almohad expedition to Evora in 577/​1181–​ 1182, in which 400 women were seized and then put up for sale in Seville. Ibn `Idhari, Bayan al-​mugrib. 66. Meouak, “L’onomastique des personnages d’origine ‘slave’; Marín, Mujeres en al-​Andalus, 41–​44. 67. Bariani, “De las relaciones entre Subh y Muḥ̣ammad”; Marín, “Una vida de mujer: Subh.”

Bibliography Primary Sources Dhikr bilad al-​Andalus li-​mu’allif majhul (Una descripción anónima de al-​Andalus). 2 vols. Edited, studied, and translated by Luis Molina. Madrid: CSIC, 1983. Ibn al-​Abbar. Kitab al-​hulla al-​siyara´. 2 vols. Edited by H. Mu´nis. Cairo:  Al-​ Shirkat al-​`arabiyya, 1963. Ibn al-​Attar. Al-​Watha´iq wa’l-​sijillat. Edited by Pedro Chalmeta and Federico Corriente. Madrid: IHAC, 1983. Ibn Bassam. Al-​Dhakhira fi mahasin al-​Jazira. Edited by I. `Abbas. Beirut: Dar al-​thaqafa, 1979 [1399]. Ibn Butlan. al-​Mukhtar ibn al-​Hasan. Risala fi shira al-​raqiq wa-​taqlib al-​abid. Edited by Abd al-​Salam Harun. Nawadir al-​makhtutat, no.  4.  Cairo:  Matba`at Lajnat al-​taʼlif wa’l-​tarjama wa’l-​nashr, 1954 [1373]: 333–​389 (no. 15). Ibn Hawqal. Surat al-​ard. Edited by J. H. Kramers. Leiden: Brill, 1967. [Translated into Spanish in Ibn Hawqal, Configuración del mundo ( fragmentos alusivos al Magreb y España), traducción e índices por María José Romaní Suay. Valencia: Universidad de Valencia, Textos Medievales, 1971]. Ibn Hayyan. Al-​Sifr al-​thani min kitab al-​muqtabis. Edited by Maḥmūd Alī Makkī. Riyadh: Markaz al-​Malik Faysal lil-​Buhuth wa’l-​Dirasat al-​Islamiya, 2003. Ibn Idhari. Bayan al-​mughrib fi akhbar al-​Andalus wa’l-​Maghrib. 2 vols. Edited by G. Colin and E. Lévi-​Provençal. Leiden: Brill Publishing, 1948–​1951. —​—​—​. Bayan al-​mughrib fi akhbar al-​Andalus wa’l-​Maghrib: Qism al-​muwahhidin. Vol. 5. Edited by M. I. Al-​Kattani, M. Ibn Tawit, M. Znaybar, and `A. Q. Zamama. Beirut: Dar al-​Thaqafa, 1985. —​—​—​. Bayan al-​mughrib fi akhbar muluk al-​Andalus wa-​l-​Maghrib: Al-​Juz´ al-​thalith. Edited by E. Lévi-​Provençal. Paris: Geuthner, 1930.

140 i

Cristina de la Puente

Ibn Mughith. Al-​Muqni` fi ilm al-​shurut. Edited by F. J.  Aguirre Sádaba. Colección Fuentes Arábico-​Hispanas 5. Madrid: CSIC, 1994. al-​Maqqari. Nafh al-​tib. Edited by Ihsan `Abbas. Beirut: Dar Sadir, 1968. al-​Nubahi, Ibn al-​Hasan. Kitab al-​marqaba al-​ulya fi-​man yastahiqqu al-​qada (Histoire des juges d’Andalousie). Edited by E. Lévi-​Provençal. Cairo: Dar al-​Katib al-​Misri, 1369/​1948. Saqati. Kitab fi adab al-​hisba. Edited by G. S. Colin & E. Lévi-​Provençal. Paris:  Leroux, 1931. [Translated into Spanish by Pedro Chalmeta. “El Kitab fi adab al-​hisba de al-​Saqati = Libro del buen gobierno del zoco.” Al-​Andalus 33 (1968): 125–​434].

Secondary Sources Arnaldez, Roger. “Statut juridique de sociologique de la femme en Islam.” Cahiers de Civilisation Médiévale 20 (1977): 131–​143. Ávila, María Luisa. “Las mujeres ‘sabias’ en al-​Andalus.” In La mujer en al-​Andalus:  Reflejos históricos de su actividad y categorías sociales, edited by María Jesús Viguera, 139–​184. Madrid: Ediciones de la Universidad Autónoma de Madrid, 1989. Bariani, Laura. “De las relaciones entre Subh y Muhammad b. Abi `Amir al-​Mansur con especial referencia a su ruptura (wahsha) en 386/​388/​996–​998.” Qurtuba 1 (1996): 39–​57. Benítez Sánchez-​Blanco, Rafael. “El cautiverio de los moriscos.” Manuscrits 28 (2010): 19–​43. Blanc, François-​Paul, and Albert Lourde. “Les Conditions juridiques de l’accès au statut de concubine-​mère en droit musulman malékite.” Revue de l´Occident musulman et de la Méditerranée 36 (1983): 163–​175. Botte, Roger, and Stella Alessandro, eds. Couleurs de l’esclavage sur les deux rives de la Méditerranée (Moyen Âge–​XXe siècle). Paris: Éditions Karthala, 2012. Braude, Benjamin. “Cham et Noé: Race, esclavage et exégèse entre islam, judaïsme et christianisme.” Annales HSS 1 (2002): 93–​125. Brunschvig, Robert. “De la Filiation maternelle en droit musulman.” Studia Islamica 9 (1958): 49–​59. Cahen, Claude. “Considérations sur l’utilisation des ouvrages de droit musulman para l’historien.” In Atti . . . Congr. Studi Arabi Islamici, 239–​246. Ravello: Istituto Universitario Orientale, 1966. Carballeira Debasa, Ana María. Galicia y los gallegos en las fuentes árabes medievales. Cuadernos de Estudios Gallegos Anejo XXVI. Madrid: CSIC, Xunta de Galicia, 2007. Caswell, Fuad Matthew. The Slave Girls of Baghdad:  The Qiyan in the Early Abbasid Era. London: I.B. Tauris, 2011. Chalmeta, P. El señor del zoco en España. Madrid: IHAC, 1973. Coello, Pilar. “Las actividades de las esclavas según Ibn Butlan (s. XI) y al-​Saqati de Málaga (ss. XII–​XIII).” In La mujer en al-​Andalus: Reflejos históricos de su actividad y categorías sociales, edited by María Jesús Viguera, 201–​210. Madrid: Ediciones de la Universidad Autónoma de Madrid, 1989. Constable, Olivia Remie, ed. Medieval Iberia:  Readings from Christian, Muslim and Jewish Sources. Philadelphia: University of Pennsylvania Press, 1997. de la Puente, Cristina. “La caracterización de Almanzor:  Entre la epopeya y la historia.” In Biografías y género biográfico, edited by M. L. Ávila and Manuela Marín, 367–​402. , Estudios Onomástico-​Biográficos de al-​Andalus, VIII. Madrid: CSIC, 1997.



The Ethnic Origins of Female Slaves in al-Andalus

j 141

—​—​—​. “Free Fathers, Slave Mothers and Their Children: A Contribution to the Study of Family Structures in al-​Andalus.” Imago Temporis Aevum 7 (2013): 27–​44. —​—​—​. “Judicial Sources for the Study of Women:  Limitations of the Female Capacity to Act According to Maliki Law.” In Writing the Feminine, edited by Manuela Marín and R. Deguilhem, 95–​110. London: I.B. Tauris, 2002. —​—​—​. “Los límites legales del concubinato: Normas y tabúes de la esclavitud sexual.” Al-​Qantara 28 (2007): 409–​433. —​—​—​. “Mano de obra esclava en al-​Andalus.” Espacio, Tiempo y Forma 23 (2010): 135–​147. —​—​—​. “Mujeres cautivas en ‘la tierra del islam.’” Al-​Andalus-​Magreb 14 (2007): 19–​37. —​—​—​. “Sin linaje, sin alcurnia, sin hogar: Eunucos en al-​Andalus en época omeya.” In Identidades marginales, edited by Cristina de la Puente, 147–​193. Estudios Onomástico-​Biográficos de al-​ Andalus, XIII. Madrid: CSIC, 2003. —​—​—​. “Slaves in al-​Andalus through Maliki watha’iq Works (4th–​6th Centuries H./​10th–​12th Centuries CE):  Marriage and Slavery as Factors of Social Categorisation.” Annales Islam­ ologiques 42 (2008): 187–​212. Del Moral, Celia. “Contribución a la historia de la mujer a partir de las fuentes literarias andalusíes.” In La sociedad medieval a través de la literatura hispanojudía, edited by Ricardo Izquierdo Benito and Ángel Sáenz-​Badillos, 101–​121. Cuenca: Ediciones de la Universidad de Castilla-​La Mancha, 1998. Ferrer i Mallol, Maria Teresa, and Josefina Mutgé, eds. De l’esclavitud a la llibertat: Esclaus i lliberts a la Edad Mitjana. Barcelona, Institución Milà i Fontanals, 2000. Fierro, Maribel. “Genealogies of Power in Al-​Andalus: Politics, Religion and Ethnicity during the Second/​Eighth–​Fifth/​Eleventh Centuries.” Annales Islamogiques 42 (2008): 29–​56. Garrido García, Carlos Javier. “La esclavitud morisca en el Reino de Granada: El caso de la Villa de Fiñana.” Miscelánea de Estudios Árabes y Hebreos (sección árabe) 50 (2001):107–​131. Ghersetti, Antonella. “De l’Achat des esclaves:  Entre examen médical et physiognomonie. Le chapitre 46 du Kitab al-​dala´il d’Ibn Bahlul (Xe S.).” The Arabist: Budapest Studies in Arabic 23 (2001): 83–​94. Jayyusi, Salma Khadra, ed. The Legacy of Muslim Spain. Leiden: Brill, 2000. Johansen, Baber. “The Valorization of the Human Body in Muslim Sunni Law.” In Law and Society in Islam, edited by Devin J. Stewart, Baber Johansen, and Amy Singer, 71–​112. Princeton, NJ: Markus Wiener, 1996. Kennedy, Hugh. Muslim Spain and Portugal. London: Longman, 1996. Kentaro, Sato. “Slave Elites and the saqaliba in al-​Andalus in the Umayyad Period.” In Slave Elites in the Middle East and Africa: A Comparative Study, edited by Miura Toru and J. E. Philips, 25–​40. London: Kegan Paul International, 2000. Lewis, Bernard. Race and Color in Islam. New York: Harper & Row, 1971. Linehan, Peter. History and Historians of Medieval Spain. Oxford: Clarendon, 1993. López Ortiz, José. “Formularios notariales de la España musulmana.” La Ciudad de Dios 145 (1926): 260–​273. Makki, Mahmud Ali. Las aportaciones orientales y su contribución a la formación de la cultura hispanomusulmana hasta finales del siglo x. Madrid: Instituto Egipcio de Cultura Islámica, 1967. Manzano, Eduardo. “The Creation of a Medieval Frontier: Islam and Christianity in the Iberian Peninsula, Eighth to Twelfth Centuries.” In Frontiers in Question:  Eurasian Borderlands, 700–​1700, edited by Daniel Power and Naomi Standen, 32–​52. Hampshire: Macmillan Press, 1999.

142 i

Cristina de la Puente

Marín, Manuela. Mujeres en al-​Andalus. Estudios Onomástico-​Biográficos de al-​Andalus, XI. Madrid: CSIC, 2000. —​—​—​. “Una vida de mujer: Subh.” In Biografías y género biográfico, edited by M. L. Ávila and Manuela Marín, 425–​445. Estudios Onomástico-​Biográficos de al-​Andalus, VIII. Madrid: CSIC, 1997. Meouak, Mohamed. “Esclaves noirs et esclaves blancs en al-​Andalus umayyade et en Ifriqiya fatimide.” In Couleurs de l’esclavage sur les deux rive de la Méditerranée (Moyen Âge–​XXe siècle), edited by Roger Botte and Stella Alessandro, 25–​53. Paris: Karthala, 2012. —​—​—​. “L’onomastique des personnages d’origine ‘slave’ et ‘affranchie’ en al-​Andalus à l’époque califale (IVe/​ Xe siècle):  Premières approximations documentaries.” Onoma 31, no. 1–​ 3 (1992–​1993): 17–​28. —​—​—​. Saqaliba, eunuques et esclaves à la conquête du pouvoir: Géographie et histoire des élites politiques “marginales” dans l’Espagne umayyade. Helsinki: Academia Scientiarum Fennica, 2004. Molina, Luis. “Las campañas de Almanzor a la luz de un nuevo texto.” Al-​Qantara 3 (1982): 467–​472. Ragib, Yusuf. Actes de vente d’esclaves et d’animaux d’Égypte médiévale, 2 vols. Cahiers des Annales Islamologique. Cairo: Institut français d’archéologie orientale, 2006. Safran, Janina M. “Identity and Differentiation in Ninth-​Century al-​Andalus.” Speculum 76 (2001): 573–​598. Seco de Lucena, Luis, ed. Documentos arábigo-​granadinos. Madrid:  Instituto de Estudios Islámicos, 1961. Shatzmiller, Maya. Her Day in Court:  Women’s Property Rights in Fifteenth-​Century Granada. Cambridge, MA: Harvard University Press, 2007. Trabelsi, Salah. “Commerce et esclavage dans le Maghreb oriental (VIIe–​Xe siècles).” In Couleurs de l’esclavage sur les deux rive de la Méditerranée (Moyen Âge–​XXe siècle), edited by Roger Botte and Stella Alessandro, 9–​23. Paris: Karthala, 2012. Tyan, Émile. Le Notariat et le régime de la preuve par écrit dans la pratique du droit musulman. Beirut: Imprimerie Saint-​Paul, 1959. Vallvé, Joaquín. “Sobre demografía y sociedad en al-​Andalus.” Al-​Andalus 42 (1977): 323–​340. Vidal Castro, Francisco. “El cautivo en el mundo islámico: Visión y vivencia desde el otro lado de la frontera andalusí.” In II Estudios de frontera, actividad y vida de frontera (Congreso celebrado en Alcalá la Real, del 19 al 22 de noviembre de 1997), edited by Francisco Toro Ceballos, 771–​823. Jaén: Diputación, 1998. Wasserstein, David. The Rise and Fall of the Party Kings:  Politics and Society in Islamic Spain (1002–​1086). Princeton, NJ: Princeton University Press, 1985. Zomeño, Amalia. “Abandoned Wives and Their Possibilities for Divorce in al-​Andalus:  The Evidence of the Watha´iq Works.” In Writing the Feminine: Women in Arab Sources, edited by Manuela Marín and Randi Deguilhem, 111–​126. London: I.B. Tauris, 2002. —​—​—​. “Del escritorio al tribunal:  Estudio de los documentos notariales en la Granada nazarí.” In Grapheîon:  Códices, manuscritos e imagines:  Estudios filológicos e históricos, edited by J. P. Monferrer Sala and M. Marcos Aldón, 75–​98. Col. Studia Semitica 2. Cordova: Servicio de Publicaciones de la Universidad de Córdoba, 2003.

7 The Mothers of the Caliph’s Sons Women as Spoils of War during the Early Almohad Period

Heather J. Empey

i  In his monograph on Celia, a nineteenth-​century Missouri slave woman who killed her abusive master, the American historian Melton A. McLaurin makes the insightful observation that the lives of lesser figures, men and women who lived and died in virtual anonymity, often better illustrate certain aspects of the major issues of a particular period than do the lives of those who, through significant achievement, the appeal of the orator, or the skill of the polemicist, achieve national prominence.1 This chapter is a study of just such lesser figures:2 women who were taken as spoils of war (ghanima) and then distributed as concubines or sold into slavery, during the rise to power of the Almohads, the dynasty that ruled the Islamic West (i.e., al-​Maghrib)3 from 1147 to 1269. The Almohads were the most powerful dynastic entity to rule over the Maghrib during the medieval period.4 They were al-​Muwahhidun, those who “proclaimed God’s Unity” in accordance with the teachings of their spiritual father, Muhammad ibn Tumart (d. 1130).5 Ibn Tumart was an ascetic from the Souss region of the Farthest Maghrib who believed in the Quranic mandate to command good and forbid wrong (amr bi’l-​maruf wa’l-​nahy an al-​munkar). He railed against the corruption of the rulers of his day: the Zirids (r.c. 972–​11590/​1160) of Tunis, the Hammadids (r.c. 1007–​1153) of Bejaia, and, 143

144 i

Heather J. Empey

especially, the Almoravids (r.c. 1070–​1147) of Marrakesh.6 Ibn Tumart claimed he was an infallible messianic redeemer (i.e., Mahdi), but the rebellion he instigated against the Almoravids met with little success. After his death, leadership of the Almohad movement passed to the man known to history as his most beloved disciple: Abd al-​Mu’min ibn Ali al-​Kumi (d. 1163). Abd al-​Mu’min, the first Almohad caliph, led a veritable revolution across North Africa in the name of Almohadism. One by one, he overthrew the Almoravids, the Hammadids, and the Zirids.7 Throughout the course of his wars, he took several concubines from among the Berber tribes he conquered.8 Indeed, the Almohads frequently took captive the womenfolk of their opponents, selling them off into slavery if their people refused to embrace the religious doctrines of Ibn Tumart. These women bore Abd al-​Mu’min the many sons, who helped him govern his empire and marginalize his political rivals.9 The subject of female spoils of war during Abd al-​Mu’min’s reign has yet to be explored by scholars, despite the fact that we find numerous references to female prisoners and slaves in the chronicles. The goal of the current work, therefore, is to scour the historical record in search of traces of the lives and impact of these women. As we shall see, their stories have something to tell us not only about female slavery under the Almohads in particular but also about the very nature of Almohadism as well as about the large historical importance of seemingly small historical actors. Indeed, like McLaurin’s Celia, these women open a unique window onto the wider political and ideological revolution in which they were key pawns, and Almohad concubines and female slaves offer us an intimate glance at the conduct of Almohad warfare that we do not find elsewhere in the primary sources. Despite the fact that their stories come down to us through the texts of their Almohad conquerors and, more often than not, we do not hear their voices directly or learn their names, this chapter seeks to demonstrate that the study of the lives of women, even lacking in agency and relegated to the margins of history, can make unique and surprising contributions to our understanding of the past. Sources and Voices Although the historical record dealing with the rise of the Almohads can be sparse and contradictory,10 this study draws on two important chronicles written close to the time period in question:  the Mann bi-​l’Imama by the Andalusian bureaucrat Ibn Sahib al-​ Sala (d.c. 1204)11 and the Kamil fi’l-​Tarikh by the famed Iraqi historian Ibn al-​Athir (d. 1233).12 Even more extensive use, however, is made of the Memoirs of al-​Baydhaq, a candid, firsthand account of early Almohad history.13 The Memoirs provide us with key insights about Almohad-​era women that we do not find in any other medieval source. Abu Bakr al-​Sanhaji (d. after 1163), nicknamed al-​Baydhaq, or “the Pawn,” was an early and devoted companion of Ibn Tumart and went on to accompany Abd al-​Mu’min on his campaigns. And, although the Memoirs can be considered a collection of credible



The Mothers of the Caliph’s Sons

j 145

eyewitness reports, they are also highly apologetic.14 In the first half of his work, al-​ Baydhaq draws parallels between the life of Ibn Tumart and that of Prophet Muhammad to bolster Ibn Tumart’s claim to being the infallible spiritual guide of his age and awaited Mahdi. This, in turn, foreshadows the central piece of apologia in the Memoirs—​namely, that the purpose of Ibn Tumart’s messianic preaching was to pave the way for the arrival of his successor, Abd al-​Mu’min.15 Al-​Baydhaq reiterates throughout that Abd al-​Mu’min’s decisions and actions were modeled on those of Ibn Tumart in an attempt to strengthen his position as leader of the Almohads.16 A comparison with sources from outside of the Almohad Empire, such as Ibn al-​Athir’s Kamil fi’l-​Tarikh, suggests that this apologia was designed to squelch rumors that Abd al-​Mu’min was something of a usurper and not the inevitable successor to the movement founded by Ibn Tumart. A clear example of said apologetic can be found in al-​Baydhaq’s narration of Abd al-​ Mu’min’s offensive against Almoravid targets in the Souss and High Atlas at the beginning of the Almohad Revolution. Abd al-​Mu’min’s chief foe at the time was the Reverter, a Christian mercenary in the service of the Almoravid sultan, Ali ibn Yusuf ibn Tashfin (r. 1106–​1140). In 1140/​1141, Almohad forces raided several large towns, returning to their home base at Tinmal17 with the booty, including many women.18 The aftermath is worth examining in detail because it introduces us to Tamagunt, who cleverly frees herself and her fellow prisoners in the only “speaking part” given to a woman in the whole of the Memoirs: The Amir [i.e., Abd al-​Mu’min] designated a dwelling for the women. [Meanwhile], the Reverter attacked the Gaiga and carried away their women.19 Among the captives was the wife of Yaʿazza ibn Makhluf.20 Then [back at Tinmal], Tamagunt spoke. She said, “Oh people! Is the Commander of the Faithful (amir al-​mu’minin) here?” “Yes,” they told her. She said: “Oh, Commander of the Faithful! My father Yintan ibn Umar interceded in favor of the Mahdi!” He replied: “You are right. You are free!” Then she said, “Is it right to free me alone amongst 400 others?” He replied, “You are right!” He ordered that the women be set free and treated with respect until they arrived in Marrakesh. When they arrived, Ali ibn Yusuf took the wife of Yaʿazza ibn Makhluf and the other women of Gaiga and sent them back such that their security and dignity were safeguarded. When they arrived, the Caliph said, “our actions were only repaid in kind. We are a people who do not violate women’s honor (la naʿmalu ala hatk al-​ird)”.21 Several years before, in one of the seminal moments of early Almohad history, Ibn Tumart had barged into Ali ibn Yusuf ’s Friday Mosque in Marrakesh to debate with the sultan’s men of religion. One of Ali ibn Yusuf ’s courtiers advised the sultan to have Ibn Tumart killed. In contrast, Yintan ibn Umar, another courtier, pleaded for the Mahdi’s life and then advised him to leave Marrakesh before the sultan could change his mind.22

146 i

Heather J. Empey

Abd al-​Mu’min, as Ibn Tumart’s loyal successor, thus makes a show of repaying the former kindness of Yintan by freeing his daughter, Tamagunt. Almohad apologetic is also reflected in Abd al-Mu’min’s key sentence:  “We are a people who do not violate women’s honor.” The Memoirs’ insistence on the fact that the captive Almoravid women were treated with respect can also be interpreted as a defense against accusations to the contrary. Contemporaries of the Almohads would have been well aware that many women were taken captive and sold into slavery during the course of the Almohad Revolution, arousing suspicion that they were not as mindful of women’s honor as they claimed. Tamagunt’s story, in fact, seems to have had a special propagandistic value for the Almohads, because it is repeated again toward the end of the Memoirs.23 The Mothers of the Caliph’s Sons Although the episode concerning Tamagunt provides us with an interesting glimpse into Abd al-​Mu’min’s treatment of enemy women, the question of the composition of the caliph’s own female household is even more relevant to his rise to power. Once again, only al-​Baydhaq’s Memoirs shed significant light on this issue. His brief digressions about the concubines seized by the caliph throughout the course of his campaigns can be used to piece together some important aspects of Abd al-​Mu’min’s reign. These are not made explicit elsewhere in Almohad historiography. Thus, shortly after freeing Tamagunt and early on during his wars against the Almoravids, al-​Baydhaq tells us that Abd al-​Mu’min took a great deal of booty from the fortress of Dai. The Sanhaja occupants of the fortress came to tell him they had become Almohads. As soon as they converted to Almohadism, the caliph sent back all of the Sanhaja captives, except for the future mother of his son, Abu Said Uthman.24 Although the text does not call this woman a concubine explicitly, this is the most plausible status for her, given that she was originally part of the Dai booty.25 As the mother of a male child, she would have earned the status of umm walad (i.e., mother of a child),26 mitigating her status as a concubine over time. It is important to note that the dominant interpretation of Almohad history—​ popularized by famed Maghribi historian Ibn Khaldun (d. 1406)—​holds that the Almohads’ rise to power was the result of solidarity among the Masmuda of the High Atlas and their ultimate triumph over their rivals, the Sanhaja of the plains.27 Ibn Tumart is often depicted as an implacable foe of the Sanhaja as a whole, not just the Almoravids. Unlike Ibn Tumart, Abd al-​Mu’min was not Masmuda, but rather a Zanata Berber from the Eastern Maghrib and “stranger” to the High Atlas.28 As we shall see, Abd al-​Mu’min’s struggle to solidify his position as absolute ruler of the Almohads eventually involved the liquidation of many of the Masmuda tribes that had supported him during the initial stages of the Almohad Revolution. In turn, the caliph’s personal life mirrored his



The Mothers of the Caliph’s Sons

j 147

attitude toward tribal politics. Although he initially took a Masmuda wife, he subsequently added women from rival groups to his household, such as the mother of Abu Said Uthman, eventually using their diverse ethnic and regional affiliations to strategic advantage. By the time of Abd al-​Mu’min’s aforementioned raid on Dai, at least two other women were (or had been) part of his household. Al-​Baydhaq refers to one of them as a woman “from Izurba in the Sous.”29 She bore the caliph the initial heir to the throne, Abu Abdallah Muhammad, as well as another son, Musa. Abd al-​Mu’min had also married Safiyya bint Abi Imran.30 She was the daughter of Abu Imran Musa ibn Sulayman al-​Kafif, an early companion of Ibn Tumart’s from Tinmal. Abu Imran Musa later became one of Abd al-​Mu’min’s chief judges.31 The caliph’s marriage to Safiyya likely took place in the shadowy period after Ibn Tumart’s death, when he desperately needed the support of the people of Tinmal to ensure his succession—​a goal he could achieve more easily by forming alliances with Ibn Tumart’s earliest Masmuda companions. Ibn Khaldun speculated that Abd al-​Mu’min hid Ibn Tumart’s death deliberately from the Masmuda because he knew that, as an outsider, he did not have their support to become caliph. Ibn Khaldun also claims that Abd al-​Mu’min then married the daughter of Abu Hafs Umar al-​Hintati, Abd al-​Mu’min’s second-​in-​command. Although the famed historian errs with regard to the identity of Abd al-​Mu’min’s father-​in-​law, his interpretation of events is probably not far from the truth of why Abd al-​Mu’min chose to marry into the family of one of Ibn Tumart’s early Masmuda followers and, thus, into a Masmuda clan.32 Safiyya bint Abi Imran bore Abd al-​Mu’min at least two sons: the future caliph, Abu Yaqub Yusuf, and Abu Hafs Umar, who acted as vizier to both his father and brother.33 None of this, of course, would have prevented Abd al-​Mu’min from marrying the Dai mother of Abu Said Uthman. Yet, given that the caliph acquired several more women in quick succession, there would have presumably been too many living women in his household for all of them to be his wives. The possibility exists, of course, that all of Abd al-​Mu’min’s women were wives and that he simply had more than the legally allotted four, or that he had a series of marriages and divorces, but the Almohad sources are not forthcoming on these points. An important piece of evidence to consider is that only the daughter of Abi Imran is singled out as a free woman (imra’a hurra),34 and her father is referred to as Abd al-​Mu’min’s kinsman by marriage.35 Moreover, the sources do not mention any other marriages or fathers-​in-​law for the caliph. The most plausible solution, then, is that, like so many sultans and caliphs before him, Abd al-​Mu’min contracted few marriages and kept a household of mostly concubines.36 In fact, shortly after the fall of Dai, Abd al-​Mu’min made his way north and conquered the town of Azru, where he added yet another woman to his household. Al-​Baydhaq relates that, from the conquered people, the caliph seized (akhadha) the mother of his son Abu Muhammad Abdallah.37 Once again, circumstantial evidence points to the fact that this woman was a concubine and, eventually, an umm walad.

148 i

Heather J. Empey

Al-​Baydhaq gives a more detailed glimpse at the distribution of female captives when the Almohads conquer several cities in the north, including Melilla: The Caliph distributed the booty. We had taken one hundred virgins (bikr) who were safe with us. The Caliph distributed them amongst the Almohads, who married them (tazawwajuhunna). Only Fatima, daughter of Yusuf, the Zanata woman, and the daughter of Maksan ibn al-​Mu’izz, lord of Melilla were left. The Caliph cast lots for Fatima with Abu Ibrahim38 and Abu Ibrahim won her. The Caliph took (akhadha) the daughter of Maksan ibn al-​Mu’izz, the mother of the princes Ibrahim and Ismail.39 This passage raises several interesting questions. First, the fact that the women in question were virgins, or girls who had not been married before, might have influenced the Almohads’ willingness to marry them. Indeed, as in the story of Tamagunt, al-​Baydhaq takes care to emphasize the fact that virtuous women were safe at the hands of the Almohads. Al-​Baydhaq also states that Abd al-​Mu’min took (akhadha) Bint Maksan ibn al-​Mu’izz, when he could have stated that he married her, indicating, perhaps, that no marriage was contracted. It is also noteworthy that Abd al-​Mu’min cast lots for the Zanata woman he wanted, which was a common method of distributing the booty during the early Islamic conquests.40 One might otherwise presume that, as caliph, Abd al-​Mu’min had his pick of the booty. The incident seems to demonstrate that it was important for him to show he was on equal footing with his men. Finally, it is somewhat puzzling that he preferred Fatima, the Zanata woman, to the daughter of Maksan ibn al-​Mu’izz, who likely had the higher social standing. The fact that Fatima was Zanata like Abd al-​Mu’min and that Bint Maksan ibn al-​Mu’izz was the daughter of an Almoravid official might have influenced his choice. Indeed, the sons of Bint Maksan ibn al-​Mu’izz were not given governorships during Abd al-​Mu’min’s lifetime.41 Once again, circumstantial evidence seems to indicate she became only another of his concubines. The casting of lots for Bint Maksan ibn al-​Mu’izz is the last time that al-​Baydhaq relates a story about the caliph adding a woman to his household. Nevertheless, a few more details about the mothers of Abd al-​Mu’min’s children emerge when the Memoirs discuss how the caliph distributed governorships to his sons in 1156.42 Abu Hafs Umar Inti (d. 1175/​1176), an early Masmuda adherent to Almohadism and Abd al-​Mu’min’s second-​in-​command, was originally meant to succeed him as caliph. But, as Ibn al-​Athir and others suggest, Abd al-​Mu’min had secretly been grooming his sons for the succession. In the end, Abu Hafs Umar and other Almohads felt pressured to allow him to name his eldest son, Abu Abdallah Muhammad, as heir.43 Shortly thereafter, the caliph also appointed Abu Yaqub Yusuf to the governorship of Seville, capital of al-​ Andalus. His brother, Abu Hafs Umar, became governor of Tlemcen, Abd al-​Mu’min’s native region.44 Abu Said Uthman became governor of Granada and its region,45 which, during the eleventh century, had been ruled by a branch of the Zirids, Sanhaja Berbers like



The Mothers of the Caliph’s Sons

j 149

his mother from Dai.46 The son of the woman from Azru, Abu Muhammad Abdallah, participated in the overthrow of the Hammadids and was appointed governor of their capital, Bejaia.47 The caliph’s son, Abu al-​Hasan Ali, was appointed governor of Fez. Al-​Baydhaq notes that his mother was a woman named Fatima from Fez and she is the only woman besides Safiyya bint Abi Imran whose name is provided by al-​Baydhaq.48 There is no way of knowing when the caliph met her or whether she was a concubine or wife. Al-​Baydhaq also mentions “a woman of the Lamta,”49 who was the mother of Abu Zayd Abd al-​Rahman. This son was still too young to occupy his post when his father appointed him governor of Souss, his mother’s native region.50 As with the Sanhaja mother of Abu Said Uthman, Abd al-​Mu’min seems to have had no qualms about taking women from the main rivals of the Masmuda into his household. Finally, we come to Abu Rabi Sulayman, who was appointed governor of the region of Tadla.51 He is the only son of Abd al-​Mu’min’s whose mother is not mentioned by al-​ Baydhaq. Notice of his appointment follows that of Abu al-​Hasan Ali, governor of Fez, possibly indicating the two were full brothers. Ibn Sahib al-​Sala, for his part, describes Abu Rabi Sulayman as the full brother (shaqiq) of the caliph’s sons Abu Ali al-​Hasan and Abu Ali al-​Husayn.52 The Spanish scholar Huici Miranda was not able to confirm the existence of any Abu Ali al-​Hasan, but notes the frequent confusion in the chronicles between Abu al-​Hasan Ali (son of Fatima of Fez) and Abu Ali al-​Husayn.53 If we admit the nonexistence of a son named Abu Ali al-​Hasan, it could be that Abu Rabi Sulayman, Abu al-​Hasan Ali, and Abu Ali al-​Husayn were all full brothers and, thus, all sons of Fatima of Fez. To add to the confusion, however, Ibn al-​Qattan makes Abu Rabi Sulayman the full brother of Abu Said Uthman and the nonexistent Abu Ali al-​Hasan.54 This would then make Abu Rabi Sulayman, Abu Said Uthman, and possibly even Abu Ali al-​Husayn all full brothers and sons of the Sanhaja woman of Dai. Dai, in fact, was located in the Oum Rbia River Valley, in the Tadla region, the seat of the governorship of Abu Rabi.55 Regardless of the potential justifications for each appointment, we can see that Abd al-​Mu’min was using his sons to his strategic advantage. The mere fact that several sons were sent to regions to which they were connected through their mothers indicates the caliph was hoping these tribal ties would make it easier for them to consolidate Almohad power over the diverse peoples and provinces of the Maghrib.56 That Abd al-​Mu’min had been grooming his sons all along for the succession seems to have been no secret, as the well-​thought-​out plan he had for the upbringing of his children illustrates. The Almohad-​era historian Ibn al-​Qattan (d. after 1266) reports that the caliph fed his sons excellent food, disciplined them harshly, and sometimes made them march rather than ride horseback. He made sure they knew the Quran, said their canonical prayers, and learned the intricacies of the Almohad creed. He also trained them thoroughly in the arts of war.57 One could even speculate that Abd al-​Mu’min deliberately set out to have as many sons as possible to help expand his dynasty. Indeed, in their turn,

150 i

Heather J. Empey

his son Abu Yusuf Yaqub (r. 1163–​1184) had about 18 sons, and his grandson Yaqub al-​ Mansur (r. 1184–​1199) had 14 sons, ensuring the stability of the Almohad dynasty for the duration of their reigns.58 Muslim Women as Almohad Spoils of War Beyond Abd al-​Mu’min’s household, the Almohad Revolution had a profound effect on women across the empire. Both Almohad and non-​Almohad historians corroborate the unfortunate fate of these female spoils of war. These brief reports also show that Abd al-​Mu’min manipulated Masmuda nationalism along with Almohad ideology to his own advantage and they underscore the fact that conversion to Almohadism was a means of avoiding despoilment and slavery. The main cities of the Farthest Maghrib fell to the Almohads in rapid succession after the death of the Almoravid sultan, Tashfin ibn Ali ibn Yusuf, in 1145. Ibn al-​Athir reports that, when the city of Tlemcen was conquered, most of its inhabitants were put to the sword. He adds, “The children and womenfolk were taken as captives. . . . Those who were not slain were sold at minimal prices.”59 The people of Tlemcen were being punished for resisting the Almohad siege and protecting part of the Almoravid army within their walls. In 1147, Marrakesh fell after resisting the Almohads for about a year, and most of the Almoravid leadership was put to death. As al-​Baydhaq reports, “Everything that was in the city was taken to the treasury. The women were sold and everything [all the booty from the conquest] went back to the treasury.”60 Ibn Sahib al-​Sala provides us with a similar report, but adds an important caveat: [Abd al-​Mu’min] distributed the houses [of the Marrakshis] to [the Almohads]. The families of Marrakesh were sold and their children were sold into slavery, except Zaynab the daughter of Ali ibn Yusuf, who was not allowed to be sold because of the position of her husband, the Amir Yahya ibn Ishaq al-​Masufi known as Wanzamar,61 because he had left his tribe and accepted the call of Abd al-​Mu’min. So, his house was exempted from the booty.62 Thus, as in the example of the fortress of Dai, those who converted to Almohadism by choice were treated leniently, even though they were affiliated to the deposed and presumably despised Almoravids. A slightly different version of this report, also on the authority of Ibn Sahib al-​Sala, can be found in the Maghribi chronicle Al-​Bayan al-​Mughrib. This version explains that Abd al-​Mu’min forbad the sale of Zaynab and of the members of Yahya ibn Ishaq al-​Masufi’s household, possibly including all of the “daughters” of Ali ibn Yusuf. Despite this prohibition, the brothers of Ibn Tumart dared to take two of them (i.e., the “daughters”).63 This important turn of events foreshadows the conflict that took place subsequently between Abd al-​Mu’min and the family of Ibn Tumart over control of the



The Mothers of the Caliph’s Sons

j 151

Almohad movement. Indeed, several years later, Ibn Tumart’s brothers plotted, unsuccessfully, to kill Abd al-​Mu’min just as he was planning to hand over governorship of the empire to his sons.64 Women remained important spoils of Almohad warfare as Abd al-​Mu’min struggled to keep the dynasty together. In 1149/​1150, the caliph put down an uprising from the tribes of the Dukkala region. The Dukkala had submitted to the Almohads before the fall of Marrakesh, but later apostatized and became allies of the rebellious Almoravid warlord al-​Sahrawi.65 Al-​Baydhaq mentions that, upon defeating the Dukkala, the caliph took a great deal of booty and sold their women into slavery.66 Ibn al-​Athir adds there were so many Dukkala women for sale that “handsome girls were sold for a few dirhems.”67 And the anonymous Andalusian chronicle, al-​Hulal al-​Mawshiyya, confirms that, “a woman was sold for a dirhem and a boy slave (ghulam) for half a dirhem.”68 Rebellion was also fermenting among the founding tribes of the Almohad movement, and Abd al-​Mu’min sent several of his most loyal men to carry out purges of dissident Almohads in an event known as “the recognition.” By this time, nonloyalty to Abd al-​ Mu’min seems to have been equated with apostasy from Almohadism. Some 32,730 men were killed69 and their women risked being sold into slavery. After the defeat of a group of rebellious Almohads in the region of Tadla, their women were brought to Abd al-​ Mu’min, who had the intention of selling them, until Abu Bakr ibn Jabr, a prominent Almohad, interceded on their behalf.70 Although the women were spared, it is significant that Abd al-​Mu’min was willing to use enslavement as a punishment for disobedience, even against former Almohads. This purge of the Almohads was followed by one of Abd al-​Mu’min’s greatest military triumphs: the overthrow of the Hammadid dynasty in the Eastern Maghrib in 1152/​1153. The caliph duly enslaved the women and children of any Hammadid loyalists who came out to fight him, but did not harm or take booty from the people of the Hammadid capital Bejaia, because they surrendered to him willingly.71 This conquest made Abd al-​ Mu’min the master of a vast Maghribi empire with the added advantage of fortifying his army with Arab tribesmen, such that he became even more independent from the Almohad/​Masmuda tribes of the High Atlas. This was no small triumph, given that, during the preceding century, the nomadic Arab tribes of the Banu Hilal had settled in the Eastern Maghrib, disturbing the economic, social, and political stability of the Hammadids and Zirids. Although the Bedouin Arabs soon became prized as mercenaries, they were not ones to submit completely to outside authority.72 Yet, Abd al-​Mu’min subdued these indomitable fighters and did not hesitate to take their women prisoner when he needed. To wit, Al-​Baydhaq relates that, after defeating the Hammadids and Arabs, “The Caliph and the Almohads left for Marrakesh, with the booty, the female Arab captives and the camels, full of joy and happiness.”73 The Arab women were not sold into slavery, but rather were used as pawns to lure their men over to the Almohad side. The Arab leaders were obliged to come to Marrakesh, where Abd al-​Mu’min gave them gifts and

152 i

Heather J. Empey

told them to go back to the Eastern Maghrib with their families and wait for him to call on them. Although al-​Baydhaq’s account is fairly cryptic, Ibn al-​Athir provides us with a more politically savvy interpretation of events. According to Ibn al-​Athir, after defeating the Arabs at the Battle of Setif, [Abd al-​Mu’min] left the women and children under guard and put them in the charge of eunuch servants who served them and looked after their needs, ordering them to be well protected. When they arrived with him at Marrakesh, he lodged them in spacious houses and supplied them with generous funds. He ordered his son Muhammad to write to the Arab emirs to tell them that their women and children were under guard and protected and to order them to come so that his father could hand everything back. He added that they were offered guarantees of security and honorable treatment. When Muhammad’s letter came to the Arabs, they hastened to set out for Marrakesh. Upon their arrival, Abd al-​Mu’min gave them their women and children, treated them well and gave them large sums of money. Thereby, he captivated their hearts, they stayed with him and were welcomed by him. He sought their help to establish his son Muhammad as his heir, as we shall relate under the year 551 [1156/​1157].74 Ibn al-​Athir’s analysis of Abd al-​Mu’min’s motives for making allies of the Arabs can be corroborated by other sources as well. Letters from the Almohad chancery attest to the fact that the tribes of the Banu Hilal were among those who urged the caliph to ignore the wishes of the Almohad shaykhs and appoint Abu Abdallah Muhammad, his son by the woman from Izurba, as heir.75 Once incorporated into the ranks of the Almohad army, the Arab tribes also greatly lessened Abd al-​Mu’min’s dependence on the Masmuda for the rest of his reign. The Arabs eventually settled in the Farthest Maghrib and continued to wield considerable military and political influence throughout Almohad history.76 It is interesting to note that, as Abd al-​Mu’min’s reign came to a close, there are notably fewer reports about women being taken as spoils of war in the chronicles. For one thing, al-​Baydhaq’s narrative trails off, depriving us of his unique insights into Almohad affairs. After Abd al-​Mu’min’s death in 1163, the caliph’s sons concentrated their military efforts on al-​Andalus, and so there are more reports about their acquisition of female Christian captives,77 rather than non-​Almohad Muslim ones. We can also presume that, by this time, the Muslims of the Maghrib all presumably identified themselves as “Almohad” to avoid problems and persecution. That being said, Ibn al-​Athir provides us with yet one more incident of female enslavement, dating from the reign of the caliph’s grandson Yaqub al-​Mansur. According to the Kamil fi’l-​Ta’rikh, in 1199, the Almohad governor of the city of Mahdiyya defeated a Bedouin Arab tribe known as the Banu Awf, taking much booty, including their families. The Arabs then went to the governor of Tunis to embrace Almohadism, in the hopes of regaining their families and possessions. The governor of Tunis agreed to return the



The Mothers of the Caliph’s Sons

j 153

booty immediately, but the governor of Mahdiyya was loath to relinquish the women and booty. It was only when Yaqub al-​Mansur sent a large army against him that he switched his loyalty back to the Almohads and returned the women to the Banu Awf.78 Although this is only a minor incident in later Almohad history, it nevertheless demonstrates that Abd al-​Mu’min’s policies had not disappeared entirely and that conversion to Almohadism still carried weight with Almohad rulers, even after the heady days of the Almohad Revolution were over. Almohad Ideology and the Conquered People Viewed in historical perspective, the continuing enslavement of non-​Almohad Muslim women during the Almohad period is particularly significant because it is a rare example of Muslim women being enslaved by other Muslims—​a practice forbidden by Islamic law.79 The enslavement of women broadly speaking was not unusual in the history of the Islamic West, particularly in wars between Muslims and non-​Muslims. Both Muslim and non-​Muslim women were commonly taken as slaves during the early Arab conquest of North Africa as well as throughout the history of al-​Andalus.80 Classical Islamic law permits only the enslavement of non-​Muslim women and forbids the enslavement of freeborn Muslims.81 Yet, as the chronicles have shown, the Almohads often sold Muslim women into slavery if their tribe refused to submit to Almohadism, and freeborn Muslim women were taken as concubines by Abd al-​Mu’min. In essence, the Almohads dealt with the Muslim populace of the Maghrib as though they were non-​Muslim enemy combatants or apostates.82 This violation of Islamic legal norms and custom is particularly difficult to justify, given the Almohads’ claim to being more law-​abiding than their rival Maghribi rulers. At the very start of the Almohad movement, Ibn Tumart had tried to justify his armed rebellion against the Almoravids by arguing that the latter were guilty of anthropomorphism and attributed human characteristics to God, thus violating his essential Unity.83 The Almoravids’ anthropomorphism consequently made them guilty of unbelief (kufr) and meant they were unfit to rule a Muslim land (dar al-​Islam). Although there is little evidence to substantiate Ibn Tumart’s accusations, this was the only way he could argue that his rebellion against the Almoravids was valid legally.84 A letter attributed to him claims that jihad against the Almoravids is a legal obligation for all Almohads and that fighting them is even more important than jihad against the Christians.85 The accusation of unbelief (takfir) against the Almoravids was then extended to all non-​Almohad Muslims under Abd al-​Mu’min. For example, in 1152, six years after the fall of Marrakesh, the caliph wrote to the people of Constantine in the Eastern Maghrib, exhorting them to submit willingly to the Almohads to protect their property and lives.86 There is no evidence, however, that the Muslims of Constantine were guilty of the kinds of severe religious transgressions that would have condemned them as apostates according to Islamic law.87 In contrast, it was the Almohads’ de facto jihad against their fellow

154 i

Heather J. Empey

Muslims which transgressed the law and became one of the most notorious features of their religious ideology. Indeed, at least one scholar has proposed that the Almohad belief system was influenced by the doctrines of the radical Kharijites, who persecuted non-​Muslims and considered most other Muslims to be sinners and unbelievers.88 Few radical Kharijite groups populated North Africa during the Almohad period, making Kharijite influence difficult to prove, leaving the Almohads’ harsh treatment of their fellow Muslims still in need of some kind of explanation.89 One potential answer is that this is simply a gray area of medieval Muslim history. As Wael B. Hallaq notes, classical Islamic legal discourse only dealt with wars between Muslims and non-​Muslims and did not address wars between Muslims themselves. This might help to explain why the Almohads felt their treatment of their fellow Maghribi Muslims was justifiable. As Hallaq explains: Nowhere in the entirety of juristic discourse do the legists deal with wars and conflict between and among Muslim principalities themselves, a feature of considerable importance given the fact that, historically, wars between these principalities were as frequent, if not more so, as wars between Muslims and non-​Muslims.90 But the Almohads’ enslavement of Muslim women cannot be attributed solely to a lack of legal knowledge or to their ability to manipulate juristic discourse to their own advantage. As Hallaq further observes with regard to Islamic juristic discourse on jihad: The theory of jihad is just that—​theory  .  .  .. [I]‌t is one thing to speak of legal doctrine and quite another to speak of how this law interacts with the mundane context. A consideration of particular importance here is the proximity of jihad’s normative law to the interests of the rulers . . . the larger the investment of the ruling elite in a particular sphere of law, the more likely the application of this sphere will be tinged with non-​Shariʿa, government-​dictated elements.91 It is precisely such “non-​Shariʿa, government-​dictated elements” that have emerged throughout the course of this study of early Almohad female spoils of war. Although Ibn Tumart’s polemic against the Almoravids may have justified Abd al-​Mu’min’s initial policy toward the conquered people, his own dynastic ambitions soon became the driving force behind his treatment of his enemies. As the sources have shown, he used the womenfolk of his opponents to coerce, to manipulate, to fill his household, to form alliances, and even to add revenues to his treasury. Final Thoughts This chapter began by arguing that female spoils of war during the Almohad period would tell us more about the political and ideological context of their times.



The Mothers of the Caliph’s Sons

j 155

Certainly, the sources do not reveal much about the public and private lives of women taken as war spoils, or how they felt about their captivity or about the frustrations of the menfolk, who were unable to protect them. Even Tamagunt’s brave resistance to her captivity can be interpreted as serving a broader propagandistic purpose for the Almohads. Instead, and not insignificantly, the historical record about slave and concubine women during the Almohad period sheds light on the personal life and ambitions of the powerful figure at the center of the Almohad movement: Abd al-​Mu’min. This is no small feat, given that almost nothing is known about Abd al-​Mu’min the man. Almohad historians, such as al-​Baydhaq and Ibn Sahib al-​Sala, seem to have concealed deliberately any mention of Abd al-​Mu’min’s family background or personal motivations that might conflict with his role as successor to the Mahdi Ibn Tumart. Through his concubines and sons, however, we begin to pierce through the veneer of Almohad legend. Perhaps most important, reports about slave women in the chronicles shed considerable light on Almohad jihad and Almohadism’s complex relationship to normative medieval Islam. In fact, the Almohads’ heterodox treatment of their fellow Muslims, including Muslim slave women, can be linked to the equally controversial question of their persecution of Jewish and Christian minorities. Many medieval non-​Almohad historians assert that Abd al-​Mu’min abolished the special protected status (dhimma) of the Jews and Christians—​whereby they could reside in lands under Muslim rule without forfeiting their religion, customs, or property—​and forced them to convert to Almohadism. The Almohads’ prejudicial treatment of non-​Almohad Muslim women can be considered another piece of evidence that Almohad persecution of religious “Others” was not simply a byproduct of their wars of conquest, but part of a far-​reaching and consciously implemented policy.92 Indeed, it is only through a study of marginal figures such as female concubines and slaves that we are able to catch a glimpse of this attitude that is otherwise suppressed or forgotten in our sources. Likewise, several intriguing and as yet untapped subjects have been brought to light by studying Almohad female slaves and concubines. For example, many sources have indicated that children were taken as spoils along with women, and it would be fruitful to analyze how they, too, became pawns of the Almohad Revolution. Indeed, the last Almoravid sultan, Ishaq ibn Tashfin, was still a boy when the Almohads conquered Marrakesh, and his execution seems to have weighed heavily on Abd al-​Mu’min’s conscience.93 Both Almohad and non-​Almohad historians have also made mention of boy slaves and eunuchs. And, relatively little is known about male slavery in Almohad times, although al-​Baydhaq and others make mention of the Abid al-​Makzan, the “state slaves,” who had an established role in the Almohad hierarchy.94 The subject of both female and male slavery during the Almohad period could also be examined from an economic perspective by asking how the Almohads continued to fill their treasury after their wars of conquest were over and they had less booty to fund the needs of the state.95 Hence, female spoils of war may still help us write more chapters of Almohad history.

156 i

Heather J. Empey

Finally, however, there can be no question that female slavery under the Almohads was very different from female slavery under other Muslim dynasties during the same time period. As we have seen, the Almohads’ lack of hesitation to take female Muslim slaves and concubines is a proof of their deviance from normative medieval Islamic norms. One might expect to see similar examples of deviance among newly Islamicized peoples along the medieval Muslim world’s eastern and southern frontiers, but less so in the Maghrib, where Islamic laws and norms began to make inroads as early as the seventh century ce. The revolution in who could be enslaved and under what pretense testifies to the profound revolutionary nature of Almohadism itself and should not be seen as an emerging trend among medieval Muslims. At the same time, Almohad elites’ treatment of female Muslim slaves and concubines cannot be taken as representative of all female slavery in the Maghrib during the same time period. Further research is needed to evaluate whether Almohad dynastic practice had an impact on how female slavery was understood by jurists and the common people, and whether their precedent had an influence on later centuries of Maghribi history or on the broader history of female slavery in the Muslim world. Notes 1. McLaurin, Celia, A Slave, ix. 2. On women in the history of the Muslim West, see Viguera Molíns, “Borrowed Space.” 3. Islamic West or Maghrib refers to Islamic Spain (al-​Andalus) and North Africa. By “Farthest Maghrib,” I  mean the region that corresponds roughly to modern-​day Morocco. “Eastern Maghrib” refers to the western half of Algeria; “Ifriqiyya,” to eastern Algeria and Tunisia. 4. The best scholarly overview of the Almohad period remains: Miranda, Historia política del Imperio Almohade. 5. On Ibn Tumart as a religious reformer, see García-​Arenal, Messianism and Puritanical Reform. 6.  On the Zirids, Hammadids, and Almoravids, see Fierro, The New Cambridge History of Islam. 7. For Abd al-​Mu’min’s campaigns, see Merad, “ʿAbd al-​Mu’min à la conquête de l’Afrique du Nord (1130–​1163). 8. For a history of the Berbers, see Brett and Fentress, The Berbers. 9. The current work owes a great deal to Évariste Lévi-​Provençal’s and Huici Miranda’s earlier studies of Abd al-​Mu’min’s 15 or more sons: Lévi-​Provençal, “Appendice I,” 225; and Miranda, Historia política del Imperio Almohade, 2:613–​630. 10. On Almohad-​era historiography, see Viguera Molíns, “Historiografía.” 11. See Ibn Sahib al-​Sala, Ta’rikh al-​Mann bi-​l-​Imama. I also use citations from missing portions of the Mann bi-​l-​Imama reproduced in the Bayan al-​Mughrib of Ibn ‘Idhari (d.c. 1295) and in the anonymous Hulal al-​Mawshiyya, composed c. 1381–​1382: Ibn ‘Idhari, al-​Bayan al-​Mughrib fi; Kitab al-​Hulal. 12. Ibn al-​Athir, al-​Kamil fi’l-​Ta’rikh. 13. Al-​Baydhaq, Ta’rikh al-​Muwahhidin, 50–​133 (Arabic), 75–​224 (French). 14. Viguera Molíns, “Historiografía,” 9. 15. Al-​Baydhaq, Ta’rikh al-​Muwahhidin, 57 (Arabic), 86–​87 (French).



The Mothers of the Caliph’s Sons

j 157

16. For Ibn Tumart’s own military campaigns, see al-​Baydhaq, Ta’rikh al-​Muwahhidin, 74–​79 (Arabic), 119–​127 (French). For Ibn Tumart’s views toward Almoravid women, see Marín, “On Women and Camels,” and Mujeres en al-​Ándalus. 17.  Tinmal (or Tinmallal) was the Almohads’ fortified base in the High Atlas:  Miranda, Historia política del Imperio Almohade, 1:71–​78. 18. Al-​Baydhaq, Ta’rikh al-​Muwahhidin, 88 (Arabic), 142 (French). 19.  The Gaiga were Almohad allies:  Ibn Sahib al-​Sala, al-​Kitab al-​Ansab, 41 (Arabic), 2 (French). 20. Yaʿazza ibn Makhluf was an early member of the Almohad movement: al-​Baydhaq, Ta’rikh al-​Muwahhidin, 129 (Arabic), 220 (French). 21. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 88 (Arabic), 142–​143 (French). Here and elsewhere, I use Lévi-​Provençal’s French translation of al-​Baydhaq’s text as a guide for my English translations, making adaptations where necessary. 22. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 68–​69 (Arabic), 108–​112 (French). Cf. Ibn ‘Idhari, al-​Bayan al-​Mughrib, 24–​25, 28. 23. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 129–​130 (Arabic), 219–​220 (French). 24. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 89 (Arabic), 144 (French). 25. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 89 (Arabic), 3 (French). 26. Schact, “Umm al-​Walad,” and Puente, “Mujeres,” 25–​26. 27. On Ibn Khaldun’s view of the Berbers, see Brett and Fentress, The Berbers, 116–​119 and 131–​ 142; Hamès, “De la Chefferie,” 101–​137; and Hamès, “Le Pouvoir dynastique almohade.” 28. For Abd al-​Mu’min’s origins, see Hamès, “De la Chefferie tribale, 101–​137; and Fierro, “Las genealogías de Abd al-​Mu’min.” 29. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 117 (Arabic), 193 (French). 30. Al-​Baydhaq, Ta’rikh al-​Muwahhidin, 192 (Arabic), 116 (French), who does not call Safiyya bint Abi ‘Imran a wife, however. 31. On Abu ‘Imran Musa, see Ibn Sahib al-​Sala, al-​Kitab al-​Ansab, 34 (Arabic), 51 (French); and Ibn Sahib al-​Sala, al-​Mann bi-​l-​Imama, 157. 32. Ibn Khaldun, Histoire des Berbères. 33. On the careers of Abu Hafs ‘Umar and Abu Yaʿqub Yusuf, see Miranda, Historia política del Imperio Almohade, 2:613–​618, 624–​627. 34. Cf. her designation as a free woman in Abd al-​Wahid al-​Marrakushi, al-​Muʿjib fi Talkhis Akhbar al-​Maghrib, 345–​346; and Ibn Abi Zarʿ, al-​Anis al-​Mutrib bi-​R awd, 269. 35. Ibn Sahib al-​Sala, al-​Mann bi-​l-​Imama, 157. 36.  On the small number of marriages contracted by the Abbasid caliphs, for example, see Kennedy, The Court of the Caliphs, 165–​168. 37. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 89 (Arabic), 145 (French). 38. He was one of the highest-​ranking members of the Almohad hierarchy: al-​Baydhaq, Ta’rikh al-​Muwahhidin, 89 (Arabic), 114, note 2 (French). 39. Al-​Baydhaq, Ta’rikh al-​Muwahhidin, 94 (Arabic), 152 (French). 40. For the practice of casting lots during the early Islamic period, see Crone and Silverstein, “The Ancient Near East and Islam.” 41. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 117 (Arabic), 193 (French); and Miranda, Historia política del Imperio Almohade, 2:620–​621. 42. Lévi-​Provençal, Un Recueil de lettres officielles almohades, 36.

158 i

Heather J. Empey

43. Lévi-​Provençal, Un Recueil de lettres officielles almohades, 36. Cf. Ibn al-​Athir, al-​Kamil fi’l-​ Ta’rikh, 9:407–​408 (sub anno 551 AH). 44. Al-​Baydhaq, Ta’rikh al-​Muwahhidin, 116 (Arabic), 192 (French). 45. Al-​Baydhaq, Ta’rikh al-​Muwahhidin, 116 (Arabic), 192 (French). 46. Cf. Miranda, Historia política del Imperio Almohade, 2:618. 47. Al-​Baydhaq, Ta’rikh al-​Muwahhidin, in Lévi-​Provençal, Documents inédits d’histoire almohade, 116 (Arabic), 192 (French). 48. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 116–​117 (Arabic), 193 (French). 49. The Lamta were an Almoravid tribe: see Chaker, “Lemtouna, Lamtûna, Lemta, Lamta/​ Ilemteyen.” 50. Al-​Baydhaq, Ta’rikh al-​Muwahhidin, 116–​117 (Arabic), 193 (French). 51. Al-​Baydhaq, Ta’rikh al-​Muwahhidin, 116 (Arabic), 193 (French). 52. Ibn Sahib al-​Sala, Ta’rikh al-​Mann bi-​l-​Imama, 155–​156. 53. Miranda, Historia política del imperio almohade, 2: 619–​620. 54. Ibn al-​Qattan, al-​Nazm al-​Juman, 206–​207. Cf. Ibn al-​Sahib al-​Sala, al-​Mann bi-​l-​Imama, 155–​156. 55.  Gauthier, “Medinat-​ou-​Dai.” Cf. Lévi-​Provençal, Documents inédits d’histoire almohade, 144, note 2 (French). 56. For other sons of Abd al-​Mu’min not mentioned here, see Miranda, Historia política del imperio almohade, 2:613–​623. For the Caliph’s daughters, ‘A’isha and Safiyya, see Ibn Sahib al-​Sala, Ta’rikh al-​Mann bi-​l-​Imama, 156. 57. Ibn al-​Qattan, al-​Nazm al-​Juman, 172. 58. Miranda, Historia política del imperio almohade, 2:624–​630. 59. Ibn al-​Athir, The Chronicle of Ibn al-​Athir, 224. Cf. Ibn al-​Athir, al-​Kamil fi’l-​Ta’rikh, 9:203 (sub anno 514 AH). I have reproduced Dr. Richards’s English translation for direct citations from Ibn al-​Athir, except for the addition of Arabic terms. 60. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 106 (Arabic), 174 (French). 61. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 97–​98 (Arabic), 158–​159 (French); and Ibn Idhari, al-​Bayan al-​Mughrib, 24–​25. 62. Kitab al-​Hulal al-​Mawshiyya, 143–​144. 63.  This passage is confusing because there are lacunae in the original manuscript of the text: Ibn ‘Idhari, al-​Bayan al-​Mughrib, 30. Cf. Ibn Idhari, al-​Bayan al-​Mugrib, 287–​288. 64. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 118–​120 (Arabic), 195–​199 (French). See also Lévi-​ Provençal, Un Recueil de lettres officielles almohades, 32–​33; and Le Tourneau, “Du Mouvement almohade à la dynastie mu’minide,” 2:111–​116. 65. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 102 (Arabic), 168 (French); 109 (Arabic), 179–​180 (French); 105–​109, 120 (Arabic), 176–​180, 200 (French). 66. Al-​Baydhaq, Ta’rikh al-​Muwahhidin, 109 (Arabic), 180 (French). 67. Ibn al-​Athir, Chronicle of Ibn al-​Athir, 226. Cf. Ibn al-​Athir, al-​Kamil fi’l-​Ta’rikh, 9:206 (sub anno 514 AH). 68. Kitab al-​Hulal al-​Mawshiyya, 147. 69.  The total number is probably exaggerated:  al-​Baydhaq, Ta’rikh al-​Muwahhidin, 112 (Arabic), 185 (French). 70. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 110–​111 (Arabic), 182 (French). 71. Ibn al-​Athir, al-​Kamil fi’l-​Ta’rikh, 9:373 (sub anno 547 AH).



The Mothers of the Caliph’s Sons

j 159

72. On the Banu Hilal, see Brett and Fentress, The Berbers, 131–​139. 73. al-​Baydhaq, Ta’rikh al-​Muwahhidin, 115 (Arabic), 190 (French). 74.  Ibn al-​Athir, The Chronicle of Ibn al-​Athir, 63. Cf. Ibn al-​Athir, al-​Kamil fi’l-​Ta’rikh, 9:390–​391. 75. Cf. Lévi-​Provençal, Un Recueil de lettres officielles almohades, 30–​35. 76. On the Arabs’ importance as Almohad allies, see Aguilar Sebastián, “Politica de Abd al-​ Mu’min con los Arabes de Ifriqiya,”, 17–​30; and Buresi, “D’une Péninsule à l’autre,” 17–​23. 77. Cf. Ibn al-​Athir, al-​Kamil fi’l Ta’rikh, 10:238 (sub anno 591 AH). 78. Ibn al-​Athir, al-​Kamil fi’l Ta’rikh, 10:259–​260 (sub anno 595 AH). 79. See Brunschvig, “ʿAbd.” A notable exception was the continued enslavement of “black” Muslims (e.g., Ethiopians and Sub-​Saharan Africans), even if freeborn:  Williams, Slaves and Slavery. 80. Brett and Fentress, The Berbers, 82–​92; Marín, Mujeres en al-​Ándalus, 121–​140 and de la Puente, “Mujeres cautivas,” 19–​37. 81. Brunschvig, “ʿAbd.”; Khadduri, War and Peace in the Laws of Islam, and Hallaq, Shariʿa, 324–​341. 82.  Compare my line of argumentation here with Bennison, “Almohad Tawhid and Its Implications for Religious Difference,” esp. 211–​213. See also Urvoy, “La Pensée d’Ibn Tumart,” esp. 32. 83.  See “Rasā’il lil-​Imām al-​Mahdī wa-​Abd al-​Mu’min,” 1–​17 (Arabic), 1–​24 (French), for numerous repetitions of this accusation. Cf. Lagardère, “Le ğihad almohade,” 2:617–​631; and Serrano, “¿Por qué llamaron los almohades antropomorfistas a los almorávides?” 2:815–​852. 84. On rebellion in Islamic law, see Abou El Fadl, Rebellion and Violence in Islamic Law. 85. “Rasā’il lil-​Imām al-​Mahdī wa-​Abd al-​Mu’min,” 9 (Arabic), 15 (French). Cf. Lagardère, “Le ğihad almohade,” 2: 618. 86. Lévi-​Provençal, Un Recueil de lettres officielles almohades, 27–​28 ; and Hallaq, Shariʿa, 328. 87. Al-​Baydhaq, Ta’rikh al-​Muwahhidin, 51–​52 (Arabic), 77–​78 (French). 88. Urvoy, “La Pensée d’Ibn Tumart,” 1974 ; and Della Vida, “ʿKharidjites.” 89. On religion under the Almohads, see Fierro, “La Religión,” 437–​546. 90. Hallaq, Shariʿa, 333. 91. Hallaq, Shariʿa, 334. 92.  For a recent collection of studies on this issue, see Bennison and Angelles Gallego, Religious Minorities under the Almohads, 143–​289. See also Empey, “al-​Dhahabi on the Almohad Persecution of the Jews and Christians.” 93. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 102–​103 (Arabic), 171–​172 (French). 94. Al-Baydhaq, Ta’rikh al-​Muwahhidin, 77 (Arabic), 125 (French); and Kitab al-​Ansab, 46 (Arabic), 70 (French). 95. See Løkkegaard, “Fay’.”

Bibliography Primary Sources Abd al-​Wahid al-​Marrakushi. Al-​Muʿjib fi Talkhis Akhbar al-​Maghrib. Edited by Muhammad Said al-​Iryan and Muhammad al-​Arabi al-​Ilmi. Beirut: Dar al-​Kitab, 1978.

160 i

Heather J. Empey

al-​Baydhaq, Abu Bakr ibn Ali al-​Sanhaji. Tarikh al-​Muwahhidin li-​Abi Bakr ibn Ali al-​Sanhaji al-​Mukannan al-​Baydhaq. In Documents inédits d’histoire almohade, edited by Évariste Lévi-​ Provençal, 50–​133 (Arabic), 75–​224 (French). Paris: Librairie Orientaliste Paul Geuthner, 1928. Ibn Abi Zar`, Ali ibn Abdallah. al-​Anis al-​Mutrib bi-​R awd al-​Qirtas fi Akhbar Muluk al-​ Maghrib wa-​Ta’rikh Madinat Fas. Edited by Abd al-​Wahhab Bin Mansur. Rabat: al-​Matbaʿa al-​Malakiyya, 1999. Ibn al-​Athir, Izz al-​Din Abu al-​Hasan. The Chronicle of Ibn al-​Athir for the Crusading Period from al-​Kamil fī’l-​Ta’rikh. Part 1: The Years 491–​541/​1097–​1146: The Coming of the Franks and the Muslim Response. Translated by D. S. Richards. Surrey, UK: Ashgate Publishing, 2005. —​—​—​. The Chronicle of Ibn al-​Athir for the Crusading Period from al-​Kamil fī’l-​Ta’rikh. Part 2: The Years 541–​589/​1146–​1193: The Age of Nur al-​Din and Saladin. Translated by D. S. Richards. Surrey, UK: Ashgate Publishing, 2007. —​—​—​. The Chronicle of Ibn al-​Athir for the Crusading Period from al-​Kamil fi’l-​Ta’rikh. Part 3:  The Years 589–​629/​1193–​1231: The Ayyubids after Saladin and the Mongol Menace. Translated by D. S. Richards. Surrey, UK: Ashgate Publishing, 2008. —​—​—​. Al-​Kamil fi’l-​T’rikh. Edited by Muhammad Yusuf al-​Daqaqa. 10 vols. Beirut:  Dar al-​ Kutub al-​Ilmiyya, 1987. Ibn Idhari, Muhammad. Al-​Bayan al-​maghrib fi akhbar al-​Andalus wa’l-​Maghrib:  Qism al-​ Muwahhidin. Edited by Muhammad Ibrahim Kattani, Muhammad ibn Tawit, Muhammad Znaybar, and Abd al-​Qadir Zamama. Beirut: Dar al-​Gharb al-​Islami, 1985. —​—​—​. Al-​Bayan al-​Mugrib: Nuevos fragmentos almorávides y almohades. Translated by Ambrosio Huici Miranda. Valencia: [Gráficas Bautista], 1963. Ibn Khaldun, Abd al-​Rahman ibn Muhammad. Histoire des Berberes et des dynasties musulmanes de l’Afrique Septrionale. 2 vols. Translated by Le Baron de Slane. Edited by Paul Cassonova. Paris: Librairie Orientaliste Paul Geuthner, 1925. Ibn al-​Qattan, Abu Muhammad Hasan ibn Ali. Al-​Nuzum al-​juman li-​tartib ma salafa min akhbar al-​zaman. Beirut: Dar al-​Gharb al-​Islami, 1990. Ibn Sahib al-​Sala, Abd al-​Malik ibn Muhammad. “Al-​Kitab al-​Ansab.” In Documents inédits d’histoire almohade, edited by Lévi-​Provençal, 18–​49 (Arabic), 25–​74 (French). Paris: Librairie Orientaliste Paul Geuthner, 1928. —​—​—​. Ta’rikh al-​Mann bi-​l-​Imama ʻala al-​mustadaʻin bi-​anna jaʻalahum Allah aʼimmah wa-​ jaʻalahu al-​warithin. Edited by Abd al-​Hadi al-​Tazi. Baghdad: Wizarat al-​Thaqafa wa’l-​Funun, 1979. Kitab al-​Hulal al-​Mawshiyya fi Dhikr al-​Akhbar al-​Marrakushiyya. Edited by Suhayl Zakkar and Abd al-​Qadir Zamama. Casablanca: Dar al-​Rashad al-​Haditha, 1979. Lévi-​ Provençal, Évariste, ed. and trans. Documents inédits d’histoire almohade:  Fragments manuscrits du “Legajo” 1919 du fonds arabe de l’Escurial. Paris:  Librairie Orientaliste Paul Geuthner, 1928. —​—​—​. Un Recueil de lettres officielles almohades: Étude diplomatique, analyse et commentaire historique. Paris: Librairie Larose, 1942. al-​Nuwayri, Ahmad ibn Abd al-​Wahhab. Nihayat al-​Arab fi Funun al-​Adab. 33 vols. Edited by Mufid Qumayha et al. Beirut: Dar al-​Kutub al-​‘Ilmiyya, 2004. “Rasa’il li’l-​Imam al-​Mahdi wa-​Abd al-​Mu’min.” In Documents inédits d’histoire almohade, edited by Évariste Lévi-​Provençal, 1–​17 (Arabic), 1–​24 (French). Paris:  Librairie Orientaliste Paul Geuthner, 1928.



The Mothers of the Caliph’s Sons

j 161

Secondary Sources Abou El Fadl, Khaled. Rebellion and Violence in Islamic Law. Cambridge: Cambridge University Press, 2001. Aguilar Sebastián, Victoria. “Politica de Abd al-​Mu’min con los Arabes de Ifriqiya.” In Actas del II Coloquio Hispano-​ Marroquí de Ciencias Históricas:  “Historia, ciencia y sociedad.” Madrid: Agencia Española de Cooperación Internacional, 1989. Bennison, Amira K. “Almohad Tawḥīd and Its Implications for Religious Difference.” Journal of Medieval Iberian Studies 2, no. 2 (2010): 195–​216. Bennison, Amira K., and Maria Angelles Gallego, eds. Religious Minorities under the Almohads. A Special Issue of the Journal of Medieval Iberian Studies 2, no. 2 (2010): 143–​289. Brett, Michael, and Elizabeth Fentress. The Berbers. Oxford: Blackwell Publishing, 1996. Brunschvig, Robert. “`Abd.” In Encyclopaedia of Islam, 2nd ed., edited by P. Bearman et  al. Leiden: Brill Publishing Online, 2012. Buresi, Pascal. “D’une Péninsule à l’autre: Cordoue, Utman (644–​656) et les Arabes à l’époque almohade (XIIe–​XIIIe siècles).” Al-​Qantara 31 (2010): 17–​23. Chaker, S. “Lemtouna, Lamtûna, Lemta, Lamta/​Ilemteyen.” In Encyclopédie berbère, 29 vols., edited by Gabriel Camps et al. Aix-​en-​Provence: Édisud, 1984–​2008. Cressier, Patrice, Maribel Fierro, and Luis Molina, eds. Los Almohades: Problemas y perspectivas. Madrid: Consejo Superior de Investigaciones Científicas, 2005. Crone, Patricia, and Adam Silverstein. “The Ancient Near East and Islam:  The Case of Lot-​ Casting.” Journal of Semitic Studies 55, no. 2 (2010): 423–​450. de la Puente, Cristina. “Mujeres cautivas en ‘la tierra del Islam.’” Al-​Andalus Magreb 14 (2007): 19–​37. Encyclopaedia of Islam, 2nd ed., edited by P. Bearman et al. Leiden: Brill Publishing Online, 2012. Fierro, Maribel. “La Religión.” In El retroceso territorial de al-​Andalus, edited by Maria Jesus Viguera Molíns, 437–​546. Madrid: Espasa Calpe, 1997. —​ —​ —​ . “Las genealogías de Abd al-​ Mu’min, primer califa almohade.” Al-​Qantara 24 (2003): 77–​108. —​—​—​, ed. The New Cambridge History of Islam. Vol. 2, The Western Islamic World: Eleventh to Eighteenth Centuries. Cambridge: Cambridge University Press, 2010. García-​Arenal, Mercedes. Messianism and Puritanical Reform:  Mahdis of the Muslim West. Leiden: Brill, 2006. Gauthier, E. F. “Medinat-​ou-​Dai.” Hespéris 6 (1926): 5–​25. Hallaq, Wael B. Shariʿa:  Theory, Practice, Transformations. Cambridge:  Cambridge University Press, 2009. Hamès, Constant. “De la Chefferie tribale à la dynastie étatique: Généalogie et pouvoir à l’époque almohado-​hafside (XIIe–​XIVe siècles).” In al-​Ansâb. La quête des origines: Anthropologie historique de la société tribale arabe, edited by Pierre Bonté et al., 101–​137. Paris: Éditions de la Maison des Sciences de l’Homme, 1991. —​—​—​. “Le Pouvoir dynastique almohade entre parenté berbère, arabe et islamique.” In Los Almohades:  Problemas y perspectivas, edited by Patrice Cressier, Maribel Fierro, and Luis Molina, 425–​450. Madrid: Consejo Superior de Investigaciones Científicas, 2005. Huici Miranda, Ambrosio. Historia política del Imperio Almohade. Facsimile ed. 2 vols. Granada: Universidad de Granada, 2000.

162 i

Heather J. Empey

Kennedy, Hugh. The Court of the Caliphs:  The Rise and Fall of Islam’s Greatest Dynasty. London: Weidenfeld and Nicolson, 2004. Khadduri, Majid. War and Peace in the Laws of Islam. Baltimore, MD: Johns Hopkins University Press, 1955. Lagardère, Vincent. “Le ğihad almohade: Théorie et pratique.” In Los Almohades: Problemas y perspectivas, edited by Patrice Cressier, Maribel Fierro, and Luis Molina, 617–​631. Madrid: Consejo Superior de Investigaciones Científicas, 2005. Le Tourneau, Roger. “Du Mouvement almohade à la dynastie mu’minide: La Révolte des frères d’Ibn Tumart de 1153–​1156.” In Mélanges d’histoire et d’archéologie de l’Occident musulman, 2:111–​116. Algiers: Imprimerie Officielle, 1957. Levi Della Vida, Giorgio. “Kharidjites.” In Encyclopaedia of Islam, 2nd ed., P. Bearman, et al., eds. Leiden: Brill Publishing Online 2012. Lévi-​Provençal, Évariste. “Appendice I:  Les Fils du Calife Abd al-​Mu’min.” In Évariste Lévi-​ Provençal, Documents inédits d’histoire almohade, 225. Paris: P. Geuthner, 1928. Løkkegaard, Frede. “Fay`.” In Encyclopaedia of Islam, edited by P. Bearman et al. Leiden: Brill Publishing Online 2012. Marín, Manuela. Mujeres en al-​ Ándalus. Madrid:  Consejo Superior de Investigaciones Científicas, 2000. —​—​—​. “On Women and Camels: Some Comments on a Hadith.” In O Ye Gentlemen: Arabic Studies on Science and Literary Culture, edited by Arnoud Vrolijk and Jan P. Hogedijk, 485–​ 493. Leiden: Brill, 2007. McLaurin, Melton A. Celia, A Slave. Athens: University of Georgia Press, 1991. Merad, Ali. “ʿAbd al-​Mu’min à la conquête de l’Afrique du Nord (1130–​1163).” Annales de l’Institut d’Études Orientales 15 (1957): 109–​160. Najjar, Abd al-​Majid. Al-​Mahdi Ibn Tumart, Abu ʻAbd Allah Muhammad Ibn ʻAbd Allah al-​ Maghribi al-​Susi al-​Mutawaffi Sanat 524/​1129:  Hayatuhu wa-​Araʼuhu wa-​Thawratuhu al-​ Fikriyah wa-​l-​Ijtimaʻiyah wa-​Atharuhu bi-​l-​Maghrib. Beirut: Dar al-​Gharb al-​Islami, 1983. Schacht, Joseph. “Umm al-​Walad.” In Encyclopaedia of Islam, 2nd ed., edited by P. Bearman et al. Leiden: Brill Publishing Online, 2012. Serrano, Delphina. “¿Por qué llamaron los almohades antropomorfistas a los almoravides?” In Los Almohades: Problemas y perspectivas, edited by Patrice Cressier, Maribel Fierro, and Luis Molina, 815–​852. Madrid: Consejo Superior de Investigaciones Científicas, 2005. Urvoy, Dominique. “La Pensée d’Ibn Tumart.” Bulletin d’Études Orientales 27 (1974): 19–​44. Viguera Molíns, María Jesús. “A Borrowed Space: Andalusi and Magribi Women in Chronicles.” In Writing the Feminine:  Women in Arab Sources, edited by Manuela Marín and Randi Deguilhem, 165–​180. London: I.B. Tauris, 2002. —​—​—​. “Historiografía.” In El retroceso territorial de al-​Andalus: Almorávides y almohades, siglos XI a XIII, edited by María Jesús Viguera Molíns, 3–​37. Madrid: Espasa Calpe, 1997. —​—​—​, ed. El retroceso territorial de al-​Andalus: Almorávides y almohades, siglos XI a XIII. Madrid: Espasa Calpe, 1997. Williams, John Ralph, ed. Slaves and Slavery in Muslim Africa. 2 vols. London: Frank Cass, 1985.

8 Concubines on the Road: Ibn Battuta’s Slave Women Marina A. Tolmacheva

i  Female slavery received close attention from medieval Muslim jurists and scholars because of the women’s sexual and childbearing capacity. Inevitably, legal treatises regulating slave ownership focus on sexual relations involving slaves and their progeny. In general terms, Islamic principles prohibit enslavement of Muslims and dhimmis, or those members of monotheistic faiths living under Muslim government. Thus, production of slaves during the medieval period was limited legally to only two major means: warfare (captives, even those held for ransom, had the status of slaves) and birth of children to slave parents or slave mothers (except in cases of acknowledged parentage by the mother’s owner).1 Travel for new slaves in the Islamicate world usually began with captivity on the frontiers of Islam, followed by transportation to the slave market or markets. The travels of female slaves did not stop upon crossing the border of the Dar al-​Islam, or the Islamic realm, where most of them found their new, although not always permanent, home. The following is a case study of the slave women belonging to the wandering household of a Muslim slave owner who garnered fame as the most prominent world traveler of the medieval period. Abu Abdallah Muhammad ibn Battuta (1304–​1368 ce) was a serial slave owner, master of numerous concubines, and author of the Book of Travels (Rihla) completed in 1357.2 This Moroccan globetrotter lived during the era of the later Bahri Mamluk dynasty in Egypt and Syria (1250–​1382), the Mongol Golden Horde in the southern Russian steppe (c. 1250–​1480), and the early Delhi Sultanate in northern India (1206–​1526). Travel by premodern Muslims has become a subgenre of travel studies, and much has been written about Arab and Muslim travel literature. Ibn Battuta’s highly autobiographical book is a rare work of the rihla (travel) genre that tells readers more about the people 163

164 i

Marina A. Tolmacheva

the author met than the places he visited. The Tuhfat al-​nuzzar fi ghara’ib al-​amsar wa aja’ib al-​asfar (Gift to Those Eager to Observe the Wonders of Cities and Marvels of Journeys) provides multiple examples of slave women encountered, owned, and transported across the vast spaces of Africa and Eurasia within one person’s lifetime.3 Ibn Battuta departed his home in Tangier in 1325 to perform the pilgrimage to Mecca (hajj). He possessed, at that point, background in at least the fundamentals of the law (fiqh) of the Maliki “school,” the practice of which would later inform his career on the road. On his arrival in Egypt, he experienced firsthand the culture, wealth, and power of the Mamluk state, the major political and military force in the Near East of the period after the Mongol conquest of Baghdad in 1258 and before the Ottoman conquest in 1516 to 1517. It is a peculiar historical coincidence that not only was Ibn Battuta a foreigner wherever he went on his extended journeys, but so, too, the rulers of the three dynasties that dominated the regions in which Ibn Battuta spent these many years. Moreover, the Mamluk elite were manumitted slaves themselves—​a tiny minority in the Arab countries who maintained their authority by application of military might supplied by soldier slaves purchased continuously from southern Eurasia. The Mamluk era produced outstanding records of history, biographies, and legal cases pertaining to elite women.4 Beginning as early as the eighth century ce, famous concubines became mothers of Muslim rulers; by the fourteenth century, many state leaders in central Islamic lands were sons of slave mothers, and historians and chroniclers included in their books accounts of royal concubines who played important roles in court politics or were known for their piety, charitable gifts, and patronage. Once acquired by the palace, or gifted directly to the ruler, these women were unlikely to leave their privileged cage, unless they proved barren.5 Female slaves were expected to serve the owner’s sexual needs, even if the woman had not been originally acquired for purposes of concubinage. If she produced a child, the concubine acquired certain privileges granted by the law and custom alike. The concubines of rich men could have comforts unavailable to wives of poor men, but in legal terms, as Paul E. Lovejoy wrote, “There is no doubt that being a wife was preferable to being a concubine.”6 What we do know about the prominent slave women is usually predicated on their stable attachment to an important individual or their being part of an institution and remaining located in a certain household or social environment for a considerable length of time. In contrast, very little is known about individual concubines of the less prominent masters and households. Although women are relatively well represented in Mamluk autobiographies, the authors tend to talk more about their mothers and daughters than about their wives and concubines, much less about their perspectives on their own relationships with them.7 Many writers of the Mamluk period mention the presence of slave girls and manumitted slaves in the streets and public buildings of Egypt and Syria,8 but few speak of such figures in their own homes. At the time of Ibn Battuta’s travels, the Islamic realm was undergoing massive expansion through conquest and conversion; consequently, and important for his experience



Concubines on the Road

j 165

of concubinage, the first half of the fourteenth century marked a peak in the supply of slave girls to the Muslim markets.9 The Rihla details his encounters with queens and princesses who provided occasional patronage and protection. But Ibn Battuta’s memoir is unusual in adding to the chatty narrative of encountered celebrities his reminiscences about his relations with certain individuals and even about his feelings for his wives, concubines, and children. He married and divorced the daughters of his learned friends and colleagues, or female relatives of patrons in the service. Such women often declined to accompany him on his journeys, and he resorted to the company of slave girls, purchased or presented to him. Unlike the women of ruling elites, the women of his household are described rarely and briefly. We rarely learn even their names.10 A sweeping calculation by Li Guo estimates that Ibn Battuta “married and divorced over a period of thirty years of globetrotting more than twenty women and fathered, and eventually abandoned some seventy children. But, there is little if any information in Ibn Battuta’s account about these women and children, most of whom remain nameless.”11 Roxanne Euben is more cautious: As his stature grew over the course of his journeys, he acquired numerous male and female slaves who often traveled with him, as well as several wives (ten are mentioned in the Rihla, but there may well have been more), and fathered an unknown number of children (five are mentioned, but there may have been more).12 Suffice it to point out here that Ibn Battuta not only sought female companionship constantly, but acknowledged his habit openly to travel with his womenfolk.13 He contracted two marriages on his first journey before reaching Egypt, and thereafter acquired slave girls and wives at various stops along his itineraries. We will follow him along the complex and extended route, exploring various aspects of his slave women’s spatial mobility through the prism of travel, migration, and shifting domicile. Women in the Shadow of the “Traveler of the Arabs” It has been estimated that Ibn Battuta traveled three times the distance of Marco Polo’s journeys and visited the equivalent of 44 contemporary countries.14 From Cairo he headed north, planning to join the pilgrim caravan from Damascus. Ibn Battuta explored Palestine and Syria, made several detours, and finally completed his first hajj a year and a half after leaving home. The chronology of his travels is somewhat uncertain. He returned to Mecca in 1328 for an extended stay after exploring Iraq and western Iran. In 1330 or earlier, he traveled to Yemen, and sailed from Aden to East Africa before returning north and visiting Mecca once more. Ahead lay journeys to Anatolia, the Black Sea, and the Crimea, from where he entered the domain of Ozbek, the Mongol Khan of the Golden Horde. After a side trip to Constantinople in the train of one of Ozbek’s wives, a Byzantine princess, Ibn Battuta

166 i

Marina A. Tolmacheva

traveled southeast through Central Asia and Afghanistan, arriving in India in 1333 or 1335.15 After almost eight years in Delhi and about two years in Malabar, he visited Ceylon (Sri Lanka), Bengal, the Maldives, and Sumatra (1344–​1345). Failing in an embassy mission to China, he went there on his own (1347) and finally returned to Morocco via Arabia in 1349 in the wake of the Black Death. His devotion to protecting the Islamic realm made him join a group of jihad volunteers going to defend Gibraltar from the army of Alfonso X of Castile, but the threat of attack receded, so Ibn Battuta continued to Malaga and Valencia, reaching Granada before returning to Morocco in late 1350. After a few months there, he decided to cross the Sahara desert to the West African kingdom of Mali (1351–​1354),16 the description of which made Ibn Battuta’s book a unique source of eyewitness information on the history, culture, and society of Muslim black Africa at the time of its greatness. Travels did not make Ibn Battuta a learned geographer, and he made no claims to academic learning beyond religion and law. He was, however, a keen observer of people, places, local cultures, and elites, and his narrative provides rich, vivid, and usually reliable information about the places visited, individuals encountered, and stories heard. Among other things, he was interested in marriage practices, local women, gendered behavior, customs, and sexual mores. The literature exploring the issues of seclusion, confinement, and spatial mobility for women in Islamic societies has focused especially on regulations and customs pertaining to veiling and the royal harem of the past centuries. It has been noted that women are far more visible in Mamluk–​period sources than in those from earlier periods.17 Mamluk studies have shown a considerable degree of authority and independence exercised by women of the Mamluk elite, despite pressure for stricter regulations.18 Expectations for females of slave status were somewhat less strict and depended both on the status of her owner and her main function in the household (such as concubine vs. cook or personal attendant). If captivity was a result of warfare, so was the subsequent dislocation and relocation of captives and prisoners of war. Women and children were favored as booty in raids, whereas men were more often captured in warfare. Slave status changed the equation both for men and women. “It is a curious irony,” write Hunwick and Powell, “that the best path to an easier life for male slaves was to be deprived of their sexuality,19 whereas for females it was to have their sexuality exploited.”20 Involuntary geographic mobility was thus a constant aspect of slave status, reinforced by ethnic and linguistic alienation, despite frequent (although not required) conversion to Islam. Ibn Battuta’s case as an individual slave owner is exceptional in its geographic scope; the case of his slaves shows the extent of enforced spatial mobility and illustrates the impact of specific circumstances on the outcome for each woman connected to this itinerant master. Warfare and captivity remained parts of the background, whether in central Islamic lands or on the frontiers of Islam. As Ibn Battuta’s accounts show, there was considerable difference in the level of spatial mobility between free and slave women, but neither group likely followed a



Concubines on the Road

j 167

predictable path. In general terms, Arab culture equates lack of free spatial movement with the feminine, whereas lack of ownership capacity is associated with an animal-​ like inability to be fully human in one’s own right, socially and legally.21 Referring to the Hanafi legist al-​Sarakhsi (d. 1096), for example, Agostino Cilardo states:  “It is precisely because the manumitted slave is able to possess, Sarakhsi says, that he is brought back to life, revivified.”22 In light of this logic, women who traveled might be seen more fully in possession of their faculties and even rights, although not fully so if they were concubines and therefore slaves. If they traveled with their master, it was he who made the decisions and exercised his legal rights—​including property rights over his slaves. If a slave woman traveled on her own, it was with her master’s permission; his very possession of her was a precondition and guarantee of her, sometimes privileged, mobility.23 Ibn Battuta—​master/​owner—​was received abroad as a fellow-​Muslim, although foreign, wherever he went. In contrast, his slaves were strangers not only to him personally, but also alienated from their new environments by their race, color, language, or religion. The ties of ownership legitimized their presence wherever their master’s domicile was, whereas their isolation from local society likely ensured tighter control over their disposition and livelihood. According to Rapoport, “fifteenth-​century literary sources indicate that men of modest background kept a concubine as a substitute for a wife.”24 Ibn Battuta left for his journey from the Near East to Turkestan and India without a wife and before he had made his fortune. It is apparent from the frequency with which slave women and his own concubines are mentioned in the book, both before his arrival in India and after his departure from the country, that he found concubines a suitable alternative, and sometimes a replacement, for his wealthier and well-​connected wives.25 Ibn Battuta’s three-​year sojourn in Mecca (c. 1328–​1331) acquainted him with numerous religious scholars in residence there and their marital history. Coming from all directions of the expanding Muslim world to the center of pilgrimage and learning, some of them neglected their womenfolk, went through divorce, remarried, purchased slave girls or took foreign wives, and shared with Ibn Battuta stories of seduction, infidelity, and runaway slaves. Ibn Battuta discovered quickly that families, tribes, and social customs often prevented women from leaving their homeland, and he learned to inquire about local marital practice soon upon arrival at a new destination. Islamic law typically did not compel wives to travel against their will, but such freedom of choice did not apply to slaves, including concubines.26 Ibn Battuta mentions pointedly societies in which women “never leave their country”:  he found them in Yemen (in Zabid), in the Maldives, in China, and in Africa.27 “Traveling Man” is an easy label to affix to a roaming figure like Ibn Battuta, as James Rumford has done,28 but dislocation, geographic and social, was not welcome to most women.29 Yet, whether in Syria, India, or sub-​Saharan Africa, Ibn Battuta was often, if not constantly, in the company of women.30 At various times his traveling party included some male friends, associates, and companions, as well as concubines, female and male

168 i

Marina A. Tolmacheva

attendants, and eunuchs. Although Mecca was a major slave market, Ibn Battuta first reports purchasing slave girls in Anatolia,31 and it appears that, although he lost his fortune and possessions more than once, he never traveled without a concubine if he could help it. As he prepared for the major venture of his life, the journey to India, Ibn Battuta’s modest household was already a cultural mosaic. From North Africa he was accompanied by a male Tunisian friend, al-​Tuzari;32 from Iran, Amir-​amiran al-​Kirmani, who later died in India;33 and other male companions who remain unnamed. Of the two Greek slave girls that Ibn Battuta purchased in Anatolia, he tells us the name of one, acquired at Bali Kasri:  “I bought in this city a Greek slave-​girl named Margalita.”34 A  few days earlier, at Ephesus (Aya Suluq), he had bought another Greek slave girl, a virgin, for 40 gold dinars.35 Ibn Battuta does not describe their looks, but we may assume he was attracted by the girls’ fair complexion.36 Among the reasons for special attraction may have been the young girl’s virginity. Female virginity appears as an important motif in recorded male dreams of the Mamluk period. As Huda Lutfi has shown, the female body is described as beautiful, fertile, corpulent, unveiled, and most importantly as virginal. Male dreams depicted in the bird narrative seem to confirm that virginity is a highly prized trait in the female. Imagery of possessing, hunting, or slaughtering the virginal slave or woman abounds in dreams of domestic birds. And as we shall see later, this is also a common feature in the dream chapters dealing with man, woman, slaves, and sexuality.37 Ibn Battuta describes some of his dreams, although not about women in his case. It appears from his lasting fondness of his Greek concubine that the young woman made him happy. It is difficult to say whether Ibn Battuta thought the price of the Greek virgin too high or a great bargain. At the time of the purchase, as during much of the first half of the fourteenth century, the supply of Eurasian slave girls reached a possible high point.38 The Mamluk sultan al-​Malik al-​Nasir (r. 1299–​1340), whose pilgrimage of 1331 coincided with one of Ibn Battuta’s own, increased purchases of male and female slaves from Rum (Byzantine Anatolia) and Central Asia.39 Rapoport, referring to an important archive in Jerusalem, indicates that during the Mamluk period, the prices of slave-​girls appear to have been fairly stable, except for a possible rise at the end of the fifteenth century. The Haram documents show that at the end of the fourteenth century one could still buy an Ethiopian female slave for a mere 300 dirhams (about 12 dinars), while the highest price mentioned is 550 dirhams (about 22 dinars). These prices are slightly lower than those quoted for the first half of the century.40



Concubines on the Road

j 169

We do not know if Ibn Battuta purchased the girl in public from a slave market, although it is more likely he used an intermediary, an agent, because slave dealers were officially considered disreputable.41 Because Muslims could not be enslaved for debts or crimes, and because Arabs could no longer be enslaved legally by Muslims,42 importation of slaves from the non-​Islamic world became an ever-​expanding business through the nineteenth century. As a result, any slave possessed by a Muslim was of a different ethnicity from its owner, and one possessed by a Muslim Arab was definitely a foreigner.43 Ibn Battuta more often refers to the ethnicity of his slaves than their names, and he never discusses the religious background of his slaves, so conversion to Islam may be presumed, however recent. The Greek girls are likely to have been Orthodox Christians.44 They were obliged to learn their master’s language, although, in fact, it is not clear if any of them spoke or understood Arabic at this point. Complicating matters is that Ibn Battuta’s party now found themselves in Turkish-​speaking territory deprived of communication with the locals after the hired translator had left them. Ibn Battuta complained bitterly about this problem to a young member of the Akhi religious order at Yanija.45 His slaves were now in the land of their captors, they had a new master who was foreign to them, himself a stranger away from his homeland and preparing to enter foreign territory as unfamiliar to him as to them. Frontiers of the Islamic realm were a busy zone in the production of slaves and served as both the venue for captivity and a transit area.46 The area through which Ibn Battuta traveled after leaving Arabia and heading toward Asia Minor and the Black Sea was in the northern Middle East. For decades since the Mongol conquest of Baghdad in 1258, this region had been a theater of war between the Mamluk dynasty (1260–​1517), then in control of Egypt and Syria, and the Mongol rulers of Persia (1256–​1357), the Ilkhans. They finally negotiated a peace during the early 1320s. Princessess on both sides celebrated the event by participating in the 1326 hajj to Mecca, as witnessed in person by Ibn Battuta.47 When he saw it, Anatolia was still at war, with Turks threatening Byzantium and fighting among themselves. If no longer a military border area (thaghr),48 it was definitely still a frontier zone engaged in raid warfare and holy war against unbelievers—​in this case, primarily against Byzantium. Ibn Battuta’s interest in the borderlands may have been related to participation in jihad on the part of religious scholars (an effort attempted eventually by Ibn Battuta in Spain).49 Naomi Standen has noted that “frontiers and the mechanisms by which they are created and maintained, are wholly and absolutely constructions made by human beings.”50 Speaking of the standard perceptions of the region, destined to become an Ottoman frontier later during the fourteenth century, Colin Heywood finds that it “fit the stereotypes of other frontier societies: there exist in it (or appear to exist, following the sources) few women . . . and no children. Obviously this was not the case in reality.”51 Indeed, as Ibn Battuta saw, Anatolia and the Black Sea coasts were not originally denuded of women and children. Rather, young captives were being drained from the frontier zones to the slave markets and major urban centers south and east. Greeks, such as the young virgin

170 i

Marina A. Tolmacheva

he bought, could be captured either by Turks or the Genoese and then sold to the other side or put up for international purchasers at established centers of the Mediterranean slave trade. Caffa (present-​day Feodosia), where Ibn Battuta visited in 1332, had been a Genoese colony for half a century by that time and exported about 1,500 slaves annually.52 Once Ibn Battuta’s Greek slaves had been torn from their Byzantine roots and sold, they started their involuntary travels in Ibn Battuta’s train. The party included Ibn Battuta and his male companions as well as a number of servant women and young boys. Ibn Battuta describes his route (although not always in sequence), so we know the slaves he purchased were transported over the winter steppe and difficult mountain passes. After his acquisition of the Greek women came the journey to Constantinople,53 still the heart of the much-​reduced Byzantine Empire, and to Saray, the capital of the Golden Horde on the Volga.54 From there, braving snow and winter cold in 1332, the party progressed to Central Asia. The most privileged passenger was Ibn Battuta’s favorite concubine; she rode in a heavy wagon and camel litter with her master. We do not know how she kept warm “in the depth of winter” when Ibn Battuta himself had to wear “three fur coats and two pairs of trousers, one of them quilted.”55 If the drops of hot water for his ablutions froze “instantly,” so also, presumably, froze the water for the women’s use. At this time, Ibn Battuta’s favorite (whose name we never learn) was already heavily pregnant. She was attended by other slave women traveling with their owner when the approaching labor made them camp near Bukhara. Ibn Battuta had intended to travel as far as Samarqand and wait there for her delivery, but labor pains necessitated an earlier stop. Thus, after leaving Bukhara the traveler was forced to seek patronage of the local Mongol ruler who, however, was away on a hunt. I met his deputy, the amir Taqbugha, who assigned me a camping ground close to his mosque, and gave me a kharqa—​this is a kind of tent, a description of which we have given previously. I put the slave-​girl into this kharqa, and she gave birth to a child that same night. They told me that it was a male child, although it was not so, but after the [ceremony of the] aqiqa one of my companions informed me that the child was a girl. So I summoned the slave-​girls and questioned them, and they confirmed the statement. This girl was born under a lucky star, and I experienced everything to give me joy and satisfaction from the time of her birth.56 The child’s mother appears to be one of the Greek slave girls; the description clearly indicates that other slave women traveled in Ibn Battuta’s train, in simpler accommodations. We are not told if Ibn Battuta purchased any additional slave girls in Central Asia, where both Samarkand and Bukhara boasted significant slave markets.57 Despite the hardships of travel, the mother’s health allowed her to carry the pregnancy to term, and a baby was born to everyone’s satisfaction. A ceremony of naming and haircutting followed, traditionally held on the seventh day since birth. Many years later, after his return from travels



Concubines on the Road

j 171

and long after the child’s death, Ibn Battuta recalled his daughter with warm fondness; but, when she was first born, he had not approached the child in the first days of her life, much less bonded with her. It fell to another male in the party, a free man closer in status and education to Ibn Battuta, to advise him that the child was a girl. The female slaves concealed the truth from the master, afraid of his disappointment, or even wrath, at not having a son. We are not told in the Rihla what the mother’s sentiments were, but we may believe with confidence in her happily greeting the child, not only as a mother embracing her baby, but also as someone who had thereby acquired the legally unassailable status of umm walad, “mother of [the master’s] child.” Giving birth to her master’s offspring, male or female, protected her from ever being sold or gifted away—​that is, when the father acknowledged his paternity. Being Greek and, likely, a recent captive, the young mother may or may not have known about her legal rights, but she probably had gathered from the other slave women’s advice and attitude that her position as a master’s favorite was now enhanced. Her child was born free; unfortunately, this daughter died a month and a half after Ibn Battuta’s arrival in India, when she was not yet one year old.58 Fortunately for her mother, the loss of her master’s child did not reverse the mother’s status as umm walad, and the woman most likely would have remained in Ibn Battuta’s household until her death even if she were manumitted. This, however, was not to be her story. From Danger to Disaster Travel everywhere was fraught with danger, and Ibn Battuta and his party suffered the hardships of inclement weather, discomfort on the road and in camp, shipwreck, piracy, and robbery. The birth of a child evidently did not slow down Ibn Battuta’s mobile household. The women who had already experienced travails in the steppes of the Black Sea region and the deserts of Central Asia still had to cross mountains and more desert before reaching their master’s goal:  India. There Ibn Battuta served for six years as a Maliki judge, followed the sultan of Delhi Muhammad ibn Tughluq (r. 1325–​1351) on military campaigns, and retreated periodically from political intrigue to hermit caves. Although Ibn Battuta had contracted a number of marriages at the Delhi court, we know from his remarks interspersed through later episodes that in addition to wives he regularly owned a number of concubines, often received as gifts (in turn, he also gave some slave women away to various high-​ranking officials). When Delhi politics felt threatening, Ibn Battuta engineered a dignified departure as a leader of an embassy to China. He divorced his wives so they could stay at home, safe and with a chance to remarry, but he took with him two of his slave women. Intending to sail for China from Calicut (Qaliqut) he stated plainly to the captain: “I want a cabin to myself because of the slave-​girls, for it is my habit never to travel without them.”59 Ibn Battuta discusses in detail the ships’ size, interior structure, and conveniences, as well as issues of privacy onboard.

172 i

Marina A. Tolmacheva

Safety was an ever-​present concern. On this occasion, his party was fortunate to survive a shipwreck on the Coromandel coast without loss of life. Ibn Battuta provides a dramatic account of rescue efforts in which his selfish priorities are challenged, and the slave master yields to the needs of his slaves and fellow travelers: We cut down the mast and threw it overboard, and the sailors made a wooden raft. We were then about six miles from the shore. I set about climbing down to the raft, when my companions (for I had two slave girls and two of my companions with me) said to me, “Are you going to go on the raft and leave us?” So I put their safety before my own and said, “You two go and take with you the girl that I like.” The other girl said, “I am a good swimmer and I shall hold on to one of the raft ropes and swim with them.” So both of my companions and the one girl went on the raft, the other girl swimming. The sailors tied ropes to the raft and swam with their aid. I sent along with them all the things that I valued and the jewels and ambergris, and they reached the shore in safety because the wind was in their favor. I myself stayed on the ship. The captain made his way ashore on the rudder. The sailors set to work to make four rafts, but night fell before they were completed, and the ship filled with water. I climbed on the poop and stayed there until morning, when a party of infidels came out to us in a boat and we went ashore with them in the land of Ma`bar.60 In this scene, Ibn Battuta is generous and selfish in turn. Chastised, he puts the safety of his male companions ahead of his own (“you two go”), but seeing that the raft can only take three persons (and his valuables), he chooses to send his favorite slave woman to safety. The lives of all are at risk, and Ibn Battuta is willing to take a chance and stay on board, but the principle of “women and children first” is unknown to him. He was apparently prepared to leave his party behind when he saw the raft as a means to save himself; he obviously does not expect the sailors building the raft to use it for themselves. As the passenger paying for the passage of his whole party, he probably felt entitled—​at first to flee to safety ahead of the others and, when thwarted, to choose whom he prefers to be saved. We see him prepared to sacrifice the second woman to the safety of both his male companions and his preferred concubine. Sending the latter to safety, he openly displays his preference and, possibly under the stress of the shipwreck, calls her “the girl that I love.”61 It was fortunate for the other woman that she could swim and did not hesitate to jump into the water. We do not know if the woman for whom Ibn Battuta foregoes his privilege and risks his life by staying onboard the wrecked ship was the Greek umm walad, but all ends well in this episode. The next time things were not so fortunate. Feeling obligated to persist on his diplomatic mission, Ibn Battuta traveled on. This time, luck turned against him and he became separated from his effects, companions, state officials, and male and female slaves, including his favorite concubine. For reasons of



Concubines on the Road

j 173

privacy, Ibn Battuta had his party moved from the Chinese junk to a smaller boat called a kakam. In the meantime, he stayed ashore to attend mosque on Friday. A storm destroyed the junk as it lay anchored in the harbor. Ibn Battuta records that a sailor rescued his concubine; he refers to her as his favorite and the property of a local merchant.62 If danger awaited women who had to follow wealthy merchants or important dignitaries like Ibn Battuta traveling with Chinese ambassadors, circumstances were more precarious for less-​prominent travelers. Indian Ocean shipping carried a mix of passengers in vessels large and small, including, on occasion, single females who had to be able to fend off unwanted suitors.63 This particular sea disaster had multiple negative consequences for Ibn Battuta himself and his entire party: When those on the kakam saw what had happened to the junk they spread their sails and went off, with all my goods and slave-​boys and slave-​girls on board, leaving me alone on the beach with but one slave whom I had enfranchised. When he saw what had befallen me he deserted me.64 We sympathize with Ibn Battuta, who had lost his friends and all his valuables including his slaves—​among them his favorite concubine—​but we know that he survived the misfortune to return home and tell us his story. The kakam ship reached Sumatra (which Ibn Battuta calls Jawa), and further permutations ensued. His friends went on their way (he met one of them again on his journey to China), and in a few months he was reunited with two of his male slaves. The outcome was much more distressing for his female companions. When several months later Ibn Battuta found himself at Calicut, he was told that the slave-​girl who had been pregnant, and on whose account I was much upset, had died, and that the ruler of Jawa had taken the rest of my other slave-​girls, that my goods had been seized by various hands, and that my companions were scattered to China, Sumatra and Bengal.65 It is to Ibn Battuta’s credit that he had been “much upset” over parting from his favorite concubine, and possibly also over losing his unborn child. The tragic outcome for the woman is beyond regret. She was on board because her master’s habit was never to travel without concubines. His pleasure came before her safety, whether we speak of wanderlust or sexual relations. His other slave girls were now the property of the ruler of Sumatra, who could place them in his own harem, gift them to his courtiers, or turn them into serving women for his other concubines. They were disposable valuables, and could be appropriated as abandoned goods from the ship. In this situation, illegal in some other parts of the Muslim world, gender discrimination is evident in the fact that two male slaves belonging to Ibn Battuta were let go and managed to return to Calicut, where they reported the events to their master. His “other slave-​girls” at least

174 i

Marina A. Tolmacheva

escaped with their lives, but they, too, paid a terrible price of never knowing what life held for them in the future and having no power over directing its course even if they did. Pleasing the “Traveling Man” The death of the pregnant concubine after repeated misfortunes at sea and while pregnant was Ibn Battuta’s most grievous loss. Never again does he speak with such tender regret of the women with whom he cohabited, free or slave. As Bouhdiba indicates, “it is of utmost importance to realize that the relations between masters and female slaves were also carnal relations.”66 Ibn Battuta’s fondness for this particular “slave-​girl” shows that, although the purpose of concubinage was to cater to the male master’s sexual needs, personal relations could develop some depth of attachment. Nevertheless, he did have other slave girls besides his favorite, not only for travel (when he might not be accompanied by a wife), but also concurrently with wives or between marriages. A later adventure in the Maldives reveals another preference for a new favorite who was even preferred to the high-​born wives whom Ibn Battuta married locally. After staying in the islands for about a year and a half and contracting four marriages, local politics and rivalries among the viziers combined to undermine Ibn Battuta’s interest in high-​level marital connections. From this point on we hear about fewer and possibly less prestigious marriages, as Ibn Battuta seemed to prefer the company of concubines.67 He first divorced two of his wives, then the remaining two. He explains the transition to his temporarily single (but not celibate) state in this way: I now set sail and reached the island of Wazir `Ali. Here my wife was attacked by severe pains and wished to go back, so I divorced her and left her there, sending word to that effect to the Wazir, because she was the mother of his son’s wife. I divorced also the wife for whom I had set a fixed term, and sent for a slave-​girl of whom I was very fond [emphasis added], and we continued to travel through the island from one district to other.68 Ibn Battuta’s narrative is frank in sexual matters; satisfying his male sexual appetite is, for him, a major bonus of having female companionship, to be enjoyed lawfully.69 In concubinage, legal agency was reserved for the male, but sexual and emotional manipulation allowed the woman, whether wife or concubine, a degree of control. Khalil Athamina speculates: [F]‌emale slaves apparently filled a void in the sexual life of the Muslim man, since the Arab woman, or the Oriental one, bound as she was by conservative traditions and a complicated system of values of dignity and rules of modesty, could not give her husband what the female slave could.70



Concubines on the Road

j 175

It remains unsaid that, under the law, a slave woman is not under the same obligations for modesty and ritual purity as the free woman.71 Athamina points out that slave concubines were less particular about modesty and dignity during sexual relations than the Arab woman, who was bound by “a heavy burden of conservative tradition.”72 It was understood implicitly that concubines were expected to provide the male a greater degree of sexual pleasure than a wife (or wives). South Asian concubines pleased our traveling man most of all: “I never saw [women] better than these in companionship in the [whole] world.”73 In the Maldives, he was offered a choice of girls for his harem by a local wazir: Then he sent me a slave-​girl. His servant told me: The Wazir says to you, if you like this one, she is yours, but if no, I’ll send you a Marathi girl.” I really liked the Marathi girls, so I said to him, “Indeed I would like a Marathi.” So he sent her to me. Her name was Gul-​i-​stan, meaning Garden Flower, and she had learned the Persian language. I liked her. People in those islands have their own language, but I did not know it. Then he sent me the next day after that a Coromandel slave-​girl called “Scent of Ambergris.”74 “Garden Flower” and “Scent of Ambergris” were typical names for slave girls and, in addition to encountering a sexually pleasing Persian speaker, the wordplay in the latter case clearly appealed to Ibn Battuta. The Arabic word for Coromandel is Ma`bar and for amber, ambar (spelled anbar). So, the Tamil girl’s complete identification was the alliterated and rhymed ambariyya ma`bariyya—​that is “the amber-​scented Malabar girl.” The names of slave girls had caught his fancy in Yemen as well, where he found that inhabitants of Zafari of all men most closely resemble the people of the Maghrib in their ways. I lodged in the house of the preacher (khatib) in its principal mosque, who was `Isa b. `Ali, a man of great distinction and a generous soul. He had a number of slave-​girls, who were called by the same names as the female slaves (khuddam) in the Maghrib; one was named Bukhait [“good luck”] and another Zad al-​Mal [“may wealth increase”], and I have never heard these names in any other country.75 The generous host’s name in Yemen is remembered, as is the wazir’s in the Maldives, but the women’s “slave” names are merely a linguistic curiosity—​their original names never revealed or even inquired about (of his estimated 10 wives, Ibn Battuta names only one76). Everywhere Ibn Battuta lived or traveled, he inquired about women and sexual mores, and sought advice and remedies to enrich and improve his sexual experience. In India, he discovered the aphrodisiac effect in betel leaves and the coconut. In his

176 i

Marina A. Tolmacheva

narrative, the betel leaves’ medicinal properties seem overshadowed by their effect on sexual potency: Their specific property is that they [betel leaves] sweeten the breath, remove foul odors of the mouth, aid digestion of food, and stop the injurious effect of drinking water on an empty stomach; the eating of them gives a sense of exhilaration and promotes cohabitation, and a man will put some by his head at nights so that, if he wakes up or is awakened by his wife or slave-​girl, he make take a few of them and they will remove any foul odor there may be in his mouth. I have been told, indeed, that the slave-​girls of the sultan and of the amirs in India eat nothing else.77 Clearly, Ibn Battuta did not feed his own slave girls nothing but betel leaves, unlike the diet of the concubines of the sultan and amirs, but in his retelling, the (literally) intimate nature of detail suggests that he probably himself had experienced “a sense of exhilaration” and enhanced intercourse after chewing some leaves. When it comes to coconuts, he was informed their benefits were multiple—​as food, drink, and aphrodisiac—​and affirmed his positive experience consuming the nuts over a period of some duration: Among the properties of this nut are that it strengthens the body, fattens quickly, and adds to the redness of the face. As for its aphrodisiac quality, its action in this respect is wonderful. One of the marvelous things about it is that at the beginning of its growth it is green, and if one cuts out a piece of its rind with a knife and makes a hole in the head of the nut, he drinks out of it a liquid of extreme sweetness and coolness, but whose temperament is hot and aphrodisiac. After drinking this liquid, he takes a piece of the rind and fashions it like a spoon, and with this he scoops out the pulp, which is inside the nut. The taste of this is like that of an egg, which has been broiled but not fully cooked, and one uses it for food. This was what I lived on during my stay in the islands of Dhibat al-​Mahal for a period of a year and a half. These useful discoveries and ready availability of betel and coconut in South Asia did not prevent Ibn Battuta from seeking further knowledge of aphrodisiacs. Vegetable resources aside, the useful discovery in the Maldives was a kind of fish called qulb al-​mas. “This fish which provides their nourishment has an extraordinary and unequalled effect on their sexual prowess,” Ibn Battuta recalled, “and the islanders perform incredible feats because of it. I myself when I lived there had four wives, and concubines as well. I did the rounds of all of them every day, and then spent the night with the wife whose turn it was.” The story is both a boast and a plaint about the lost access to a resource no longer available to Ibn Battuta after his return to Morocco. The women of his Maldives harem knew to keep themselves available and ready for his nightly “rounds” that, according to Ibn Battuta, would have included all the concubines each evening, which means he did not linger on his visits to slave women. A slave does not have hill (freedom of action in



Concubines on the Road

j 177

sexual matters). In contrast, the four wives may have had a break of three days between spousal overnight stays in their beds. Giving and Buying Some of Ibn Battuta’s slave girls may have been among the gifts received by him from his local patrons, as in the story from the Maldives related earlier. Gift exchange has always been an important part of diplomacy. We have extensive records of gifts sent and received by Mamluk and Ottoman rulers and members of their entourage. Slaves were often an exotic means of impressing foreign rulers with gifts, and female slaves were valued both by givers and receivers. A subject might offer a ruler the gift of a concubine as “an optimal expression of gratitude and a very friendly gesture to gain the good will [sic] of the ruler.”78 Ibn Battuta habitually took care to provide accommodations and solicit hospitality for his traveling companions, and often shared the gifts he had received, handing down a horse here, a slave there.79 From a superior personage, one expected to receive more generous presents in return. Competition among the givers could be expected (and fostered). In the Maldives, the amir Dawlasa came to Ibn Battuta’s dwelling “bringing two slave-​girls and two men servants, and said to me ‘The sultan says to you that this present is in proportion to his means, not those of Sultan Muhammad [of India].’ ”80 Ibn Battuta remembered for years the gifts he received and gave in return. To the governor of Multan, Qutb al-​ Mulk, he presented “a white slave, a horse, and some raisins and almonds.”81 To the “king, Mughith al-​Din Muḥammad, son of the ‘king of kings” `Imad al-​Din al-​Simnani” he sent five horses, two slave girls, and two mamluks.82 He lost the ship with diplomatic gifts sent with his embassy from India to China, but by comparison with the official exchange of gifts, the gifting of slaves brings us closer to the concept of property as relationship.83 The transfer of ownership creates a bond between the giver and the recipient; the slave who is being gifted has no choice in the matter. Whether received as a gift or purchased, the slave woman is compelled to transfer her most intimate services to the new owner, having no agency in this transaction. One of Ibn Battuta’s lasting ambitions, as a legal scholar and self-​important opportunity seeker, was to own an educated concubine. Athamina points out that, although the education of jawari “was intended primarily as a means of increasing the slave traders’ profits,” it contributed to raising the level of education among women.84 Ibn Battuta never indicates if he had ever contributed to the education of his womenfolk.85 He was finally able to achieve this goal, if only temporarily, in Takedda (Tagadda) en route to Morocco from Mali. The episode, as described, touches on points of Islamic law with regard to purchase contracts and the importance of possession for the owner, both as buyer and seller: We now entered the territory of the Bardama, who are a tribe of Berbers. No caravan can travel [through their county] without a guarantee of their protection, and for this purpose a woman’s guarantee is of more value than a man’s. Their

178 i

Marina A. Tolmacheva

women are most perfect in beauty and the most shapely in figure of all women, of a pure white colour and very stout; nowhere in the world have I seen any who equal them in stoutness. The inhabitants of Tagadda have no occupation except trade . . .. They live in luxury and ease, and vie with one another in regard to the number of their slaves and serving-​women. The people of Malli and Iwalatan do the same. They never sell the educated female slaves or but rarely and at a high price. When I arrived at Tagadda I wished to buy an educated female slave, but could not find one. After a while the qadi sent me one who belonged to a friend of his, and I bought her for twenty-​five mithqals. Later on her master repented [of having sold her] and wished to have the sale rescinded, so I said to him, “If you can show me where to find another, I shall cancel it for you.” He suggested a servant belonging to `Ali Aghyul, who was that very Maghrabin from Tadala who had refused to carry any of my effects when my camel broke down, and to give my boy water when he was thirsty. So I bought her from him (she was better than the former one) and cancelled the sale with the first man. Afterwards this Maghrabin too repented of having sold the servant and wished to have the sale cancelled. He was very insistent about it but I refused, simply to pay him back for his vile conduct. He was like to go mad or die of grief, but afterwards I cancelled his bargain for him.86 This is one of the few times when Ibn Battuta refers to the bargaining involved in the acquisition of slaves, and to contractual obligations and guarantees. In much of Muslim Africa, Maliki law prevailed, so Ibn Battuta as purchaser as well as the scholars selling him their slave women are well informed about and comfortable with existing regulations. Purchase contracts under Maliki law (Ibn Battuta’s own madhhab) allowed three days for examination of slave goods for quality and defects (the same as for cattle buys).87 It was the seller who was responsible for the maintenance of the slave during the waiting period.88 In addition to the “three-​day guarantee,” there were additional rules for the sale of women and girls, who were valued for their sexual and reproductive capacity. Purchase could be canceled during the waiting period for men or women, but after that became legal; the new owner then could enter into full possession, including sexual rights. As property, any female slave could be used as a sexual partner by her master, not only those purchased for the purpose of cohabitation. Clearly, Ibn Battuta enjoyed possession of both the women at least for a while, because he says the newly acquired slave woman was “better than the former one.” Both male sellers in Takedda soon came to regret parting with their slave women, and Ibn Battuta canceled the sale, first of one woman then the other (selling them back to former owners would have involved yet another waiting period). The owners’ decision-​ making, negotiation, and their means determined the outcome for the women. However fond of them the owners might be, greed moved them easily to go through with the deal when Ibn Battuta offered them a chance to cash in on the slave’s value.



Concubines on the Road

j 179

For his part, Ibn Battuta, who had been driven by vanity and curiosity to seek ownership of an educated female, parted with his female companions, however pleasing, without regret (and received back the purchase price). The circular transaction in this case left all three male acquaintances satisfied. The matter was almost routine in legal terms, although we get a glimpse of the male owners’ psychology. However, we can only speculate about the thoughts and feelings of the women. The two women had been yanked first from one owner to another and, within days—​the days that already had involved sexual intimacy with the new master—​were taken back to the former household. They may thus have been reassured of their Takedda masters’ fondness, but, in fact, been discarded by the new owner. Moreover, their return may not have been possible had Ibn Battuta acquired them just before he departed from Takedda. They would have been bound to follow him and his caravan, away to unknown country and uncertain future, dependent on his interest in them and compassion regarding their fate, ability to support them, or his willingness to swap them with his friends and mere acquaintances. West African Muslim slaves sometimes reproached their owners for not providing Islamic instruction, and for not living up to the legal and ethical guidelines scholars provided for the treatment of slaves.89 Education of slaves, including concubines, in matters of religion took place in large cultural centers such as Cairo and Damascus.90 If awarded certificates, such slaves could then participate further in the transmission of religious knowledge alongside free women and children,91 despite resistance by conservative religious authorities.92 By comparison with Cairo, Damascus played the role of a “poor relative,”93 with fewer prominent figures resident and active there.94 Yehoshua Frenkel points out “most slaves newly arrived to Syria were from foreign origin and first-​generation Muslims. In other words, they had converted to Islam only after settling in the Mamluk sultanate.”95 Education improved chances for social mobility and was sometimes available for slaves of the elite, such as those at the court of Delhi or Ibn Battuta’s slave women who ended up at the court of the ruler of Sumatra. But it may have been of questionable value depending on whether the slaves were incorporated into the royal household and treated well or sold by the treasury to the higher bidder or a slave dealer. The outlying margins of the world of Islam were precarious legal territory. Thus, for example, it was among the Beja on the Red Sea that Ibn Battuta encountered an Arab boy in captivity96 and in Mali where he met an Arab slave girl from Damascus owned by the local savant Farba Sulayman, who, as an Islamic scholar, presumably should have known better, given that enslavement of Arabs by Muslims was illegal.97 But what the Takedda episode illustrates quite vividly is that even educated concubines, a rarity on the slave market, were interchangeable for the owners, although they were held out as a proud badge of ownership. Possession was, in such cases, more important than sexual pleasure; Ibn Battuta finds one bedmate more satisfying than the preceding one, but lets her revert to the previous owner after having taught him a

180 i

Marina A. Tolmacheva

lesson. The slave woman at Takedda may have been educated, at his own expense, by her previous master or the slave dealer. Of course, when Ibn Battuta says that the newly acquired slave woman was “better than the former one,” he may be referring not to her better education or intelligence, but to his experiencing better sexual pleasure in her company.98 In shaping Islamic law, the Muslim jurists adhered to the Quranic principle that limits sexual relations between a man and a woman to marriage and concubinage. In real terms, this meant that, in addition to the convenience of not having interfering kin of the wives and having no claim to inheritance, slave women—​including concubines—​were disposable. Ibn Battuta is very clear in distinguishing between free and slave women. The Arabic word he uses for free women is always nisa’, whereas slave women are jawari. The sexual role of the jawari was central to their social function and subject to legal regulation. Wives’ sexual needs and limitations (such as shyness and the need to appear modest) had to be satisfied by the husband,99 whereas the concubine was expected to serve her master’s pleasure.100 Some legal authorities ruled coitus interruptus “permissible with the bondmaid but not with the free woman,” although al-​Ghazali, the great eleventh-​century Muslim scholar, allowed it with a woman who might be afraid of pregnancy or childbirth, distraction from pious observances, or the effects on her beauty (which enhances man’s enjoyment of her company). Among men’s concerns, the legally acceptable ones were having to provide for too many children, the previously openly undesirable position of having to marry off his daughters, and having the pregnancy of a concubine detract from her value as sexual property, especially with a view toward possible manumission if she bore a child,101 for it would have meant loss of property.102 Ibn Battuta was a careful owner where living female property was concerned. If he gave away a given slave woman gifted to him, it was because he was conscious of the ritual value of prestige gift-​giving and probably did not wish to carry the expense of maintenance for the slave women he did not fancy as bedmates. He also returned the purchased Takedda woman to avoid a tense situation with an educated colleague. But although Ibn Battuta once mentions a male slave whom he had set free, we never read of him either selling or manumitting a slave woman.103 Conclusion Tracking Ibn Battuta helps place the womenfolk of this individual Moroccan Muslim slave owner within the whole Dar al-​Islam:  Africa, the western Mediterranean, the Indian Ocean basin, Iraq, Iran, and the Turco-​Mongol world, including Central Asia and northern India.104 None of the principalities and empires where Ibn Battuta visited or settled was walled off from the outside world—​East, West, North or South. The relatively narrow focus on slavery or gender issues chosen here connects the Dar al-​Islam to the surrounding world.



Concubines on the Road

j 181

Exploring the subject of “concubines on the road” through the story of one, admittedly major, traveler allows us to uncover a complex layering of legal considerations, social and cultural criteria, and personal and interpersonal expectations. It becomes clear that concubines must be considered a separate category, or subclass, not only of women or slaves, but also of travelers. Their case is different from that of wives or other female slaves, as well as from free and enslaved males, whether their owners or unrelated men. In addition to their own status, duties, responsibilities, and opportunities, they emerge as not only a complementary presence in the household or traveling party, but also a competitive, if unintentional, entity that compensates for the lack or absence of spousal companionship. Bouhdiba cites a curious prayer, which expresses the male frustration with some real-​ life females’ reluctance to venture abroad: “My God, procure for me a djinnia who will keep me company wherever I go.”105 Moreover, the vast picture gleaned from the actions and records of the male masters of concubines reveals a surprising dynamic. By taking female slaves on the road in their sexual capacity, Muslim men made their spouses’ travel not only less necessary, but also less appealing socially and culturally to the free women. Road trips could be relegated to the lower class of female, unless undertaken on a grand, or at least comfortable, scale (as in the train of a Mamluk princess). Dislocation, geographic and social, was not welcome to most women, and the hardships of travel in the premodern world made journeys difficult and hazardous for all. A wife, as a free woman, could avoid the danger and discomfort by declining to accompany her husband or asking for a divorce. The concubine’s slave status trumped all other aspects of her identity. She lacked the legal agency reserved for the male owner, although she could improve her lot by sexual and emotional manipulation and subterfuge, or by producing a child for her master. If the owner was on the move, she might have to give birth to her child in a traveling camp or onboard a ship. And when all was said and done, the master, even one who loved her and cherished her as a person, also remained ever mindful of her as a valued possession. Notes 1. See Brunschvig, “`Abd.” 2. See “Ibn Battuta” in the Bibliography for editions used in this study. 3.  For evidence of women’s travel in the Rihla, see Tolmacheva, “Ibn Battuta on Women’s Travel.” 4. See, for example, Rapoport, “Women and Gender in Mamluk Society”; Lutfi, “Manners and Customs of Fourteenth-​Century Cairene Women,” and “Al-​Sakhāwī’s Kitāb al-​Nisā’ ”; Frenkel, “Women in Late Mamluk Damascus”; Marmon, “Domestic Slavery in the Mamluk Empire.” See also Roded, Women in Islamic Biographical Collections. 5. Bon, Sultan’s Seraglio, 120; Burckhardt, Travels in Arabia, 187. 6.  Lovejoy, “Concubinage and the Status of Women Slaves in Early Colonial Northern Nigeria.”

182 i

Marina A. Tolmacheva

7. Rapoport. Marriage, Money and Divorce in Medieval Islamic Society, 13–​17; Guo, “Tales of a Medieval Cairene Harem,” 120. 8. Frenkel, “Women in Late Mamluk Damascus,” 419. 9. Rapoport, “Women and Gender in Mamluk Society,” 46. 10. Remke Kruk makes a point of discussing some of these names. Kruk, “Ibn Battuta,” 372. 11. Guo, “Tales of a Medieval Cairene Harem,” 101–​102 (with multiple references to Dunn, Adventures of Ibn Battuta). 12. Euben, Journeys to the Other Shore, 64. Remke Kruk also gives the count of 10 wives and 5 children (Kruk, “Ibn Battuta,” 381). 13. Seeking passage to China suitable for women’s privacy, Ibn Battuta tells a junk factor: “I want a cabin to myself because of the slave-​girls, for it is my habit never to travel without them” (Ibn Battuta, Selections, 236). 14. Dunn, Adventures of Ibn Battuta, Introduction. 15. Dunn, Adventures of Ibn Battuta, 183. 16. Tolmacheva, “Ibn Battūtah,” 129–​131. 17. Humphreys, “Politics of the Mamluk Sultanate,” 230. 18. Cilardo, “Transmission,” 31–​52; El-​Azhari, “Role of Salĝūqid Women in Medieval Syria,” 111–​126; Guo, “Tales of a Medieval Cairene Harem,” 101–​121; Lutfi, “Manners and Customs of Fourteenth-​Century Cairene Women”; and Toru, “Urban Society in Damascus.” 19. In becoming eunuchs. 20. Hunwick and Powell, African Diaspora, 99. 21. Rosen, Varieties of Muslim Experience, 154. 22. Cilardo, Transmission, 32; cited in Rosen, Varieties of Muslim Experience, 51. 23. The Rihla narrative of his later years indicates Ibn Battuta’s possession of concubines while married. Examples of substitution are common at the times of Ibn Battuta’s departure from established households for travel to further destinations (see about his departure from the Maldives in Ibn Battuta, Travels, 4:845). 24. Rapoport, “Women and Gender in Mamluk Society,” 15. 25. The Rihla narrative of his later years indicates Ibn Battuta’s possession of concubines while married. Examples of substitution are common at the times of Ibn Battuta’s departure from established households for travel to further destinations (see about his departure from the Maldives in Ibn Battuta, Travels, 4:845). 26. Tolmacheva, “Ibn Battuta on Women’s Travel,” 132; see also 125–​126, “Wives Who Would Not Travel.” 27. Ibn Battuta, Riḥla, 578. 28. Rumford, Traveling Man. 29. Tolmacheva, “Medieval Muslim Women’s Travel.” 30. Miloslavskii, Ibn Battuta, 49, calls the Rihla a “virtual encyclopedia on women of the East”; Tazi, “Al-​Rahâla al-​arab wa’l-​muslimun, iktishaf al-​akhar,” 56–​57, says, “aucun des voyageurs arabes et musulmans n’a parlé de la femme, à ma connaissance autant que le voyageur marocain” [no Arab or Muslim traveler, to my knowledge, spoke about women as much as the Moroccan traveler]. 31. Ibn Battuta, Travels, 2:445, 449. 32. Ibn Battuta, Travels, 2:473. 33. Ibn Battuta, Travels, 3:723.



Concubines on the Road

j 183

34. Ibn Battuta, Travels, 2:449, and Rihla, 307. 35. Ibn Battuta, Travels, 2:445. 36. In early Islam, marrying Greek or Persian women (jawari) was designed to lighten their descendants’ swarthy skin. Athamina, “How Did Islam Contribute to Change,” 397. 37. Lutfi, “Construction of Gender Symbolism,” 144. 38. Rapoport, “Women and Gender in Mamluk Society,” 46. 39. Tsugitara, “Slave Traders and Karimi Merchants during the Mamluk Period,”144. 40. Rapoport, “Women and Gender in Mamluk Society,” 14. 41. Lapidus, Muslim Cities in the Later Middle Ages, 82, 122–​123; Tsugitara, “Slave Traders and Karimi Merchants,” 141. 42. Brunschvig, “`Abd,” I:26. 43. Non-​Muslims in the Dar al-​Islam were not allowed to own Muslim slaves (Brunschvig, “`Abd,” I:27); also observed by Ottavio Bon in Constantinople in the seventeenth century (120). 44. At this point in his travels, Ibn Battuta also acquired (by gift) two male Greek slaves, with Christian names Mikha’il and Nicholas (Ibn Battuta, Travels, 2:444, 446). 45. Ibn Battuta, Travels, 2:455. The younger brother, Ibn Battuta concluded, only knew the word na`am (yes) in Arabic. 46. Heywood, “Frontier in Ottoman History,” 244. 47.  Ibn Battuta (Travels, 1:245) reports:  “The date of my first Standing [at `Arafa] was a Thursday, in the year [7]‌26 [November 6, 1326], the commander of the Egyptian caravan at that time being Arghun the dawadar, the lieutenant of al-​Malik al-​Naṣir. In the same year, the daughter of al-​Malik al-​Naṣir, wife of this Arghun’s son Abu Bakr, came on pilgrimage, and likewise the wife of al-​Malik al-​Naṣir, called al-​Khunda, she being the daughter of the exalted Sultan Muhammad Uzbak, king of al-​Sara and Khawarazm.” Uzbak (r. 1312–​1341) was the Khan, often called sultan, of the Golden Horde. See also Behrens-​Abouseif, “The Mahmal Tradition and the Pilgrimage of the Ladies of the Mamluk Court.” See also Johnson, “Mamluk Accounts of the Pilgrimage to Mecca.” 48. According to Lambton, “the thaghr is what divides the dar al-​Islam from the dar al-​harb. The Arab authors are very clear in this respect.” Lambton, State and Government, 13, 18–​19. By the early fourteenth century, the Byzantine thaghr had moved to the Aegean. 49. About the early concept of jihad conducted by the ulema, see, for example, Touati, Islam and Travel in the Middle Ages. 50. Standen, “Introduction,” 17. 51. Colin, “Frontier in Ottoman History,” 228–​250. 52. Epstein, Purity Lost, 59. Cuman was the language of the Kipchak Turks. Usually written in Uighur script, in this dictionary these words appear in Latin letters. So important was the Eurasian traffic through the Crimea that, during the early fourteenth century, a trilingual dictionary (Latin, Persian, and Cuman Turkish) was compiled by a Genoese merchant or missionary, and later developed into what became known as the Codex Cumanicus. See Kuum, Codex Cumanicus. 53. Tolmacheva, 1993, especially 128–​131. 54. Ibn Battuta, Travels, 2:515–​517. 55. Ibn Battuta, Travels, 2:514. 56. Ibn Battuta, Travels, 3:555–​556, and Rihla, 369. This is the child whom Kruk identifies mistakenly as “The girl born in Samarkand” (Kruk, “Ibn Battuta,” 376). 57. Athamina, “How Did Islam Contribute to Change,” 393. 58. Ibn Battuta, Travels, 3:738–​739.

184 i

Marina A. Tolmacheva

59. Ibn Battuta, Selections, 236. 60. Ibn Battuta, Travels, 4:857. 61. Ibn Battuta, Travels, 4:815. 62. Ibn Battuta explains that the slave woman’s master, the merchant, promised 10 dinars to whoever saves her, but the sailor from Hormuz who had extricated the woman declined the reward (Ibn Battuta, Travels, 4:815). 63. Stories told in the Book of the Marvels of India (953 ce) illustrate what could happen to those women who did not have a wealthy protector or a strong guard. See item LXXXIII in [Marvels of India]. Ibn Shahriyar, Book of the Marvels. 64. Ibn Battuta, Selections, 236–​238; Ibn Battuta, Travels, 4:816. 65. Ibn Battuta, Selections, 240; Ibn Battuta, Travels, 4:821. 66. Bouhdiba, Sexuality in Islam, 105. 67. While spending 10 weeks at Muluk island shortly afterward, Ibn Battuta reports two new marriages. He finally left Muluk on August 22, 1344 (Ibn Battuta, Travels, 4:846). 68. Ibn Battuta, Travels, 4:845. 69. Sexuality is regarded in Islam as a natural aspect of human relations that must be regulated. The Quran limits sex relations between a man and a woman to marriage and concubinage. Any other framework for sexual relations leads to fornication (zina) and is explicitly forbidden (IV:15, XXIV:2). The Prophet Muhammad had a regular sex life with his wives and concubines, and his superior sexual prowess is regarded as one of the signs of his excellence among men (Roded, Women in Islamic Biographical Collections, 159). 70. Athamina, “How Did Islam Contribute to Change ,” 394. 71. Brunschvig, “`Abd.” I:27. 72. Athamina, “How Did Islam Contribute to Change,” 397. 73. Ibn Battuta, Selections, 244, and Riḥla, 578. 74. Ibn Battuta, Riḥla, 584. Ibn Battuta (Selections, 227) admired Marathi women’s “special beauty, particularly in their noses and eyebrows.” 75. Ibn Battuta, Travels, 2:385, and Riḥla, 261. Gibb (Ibn Battuta, Travels) elaborates in note 79: “Bukhait is the diminutive of bakht, ‘good luck.’ Zad al-​Mal appears to be a phrase, ‘May wealth increase.’ ” 76. She is the pious and devout Ḥurnasab, sister of the Sharif Ibrahim, an official at the Delhi court. Ibn Battuta, Travels, 3:719, and Riḥla, 488. 77. Ibn Battuta, Travels, 2:387–​388, and Riḥla, 263. 78. Athamina, “How Did Islam Contribute to Change,” 390. 79. Ibn Battuta, Travels, 3:723. The gift of a war horse received from the sultan proved fatal to Ibn Battuta’s friend Kirmani when the latter could not control the animal and was thrown. 80. Ibn Battuta, Selections, 275. 81. Ibn Battuta, Travels, 3:605. 82. Ibn Battuta, Travels, 3:765. 83. Rosen, Varieties of Muslim Experience, 154. 84. Athamina, “How Did Islam Contribute to Change,” 403. Citing Qalqashandi (1355–​1418), he says: “Toward the beginning of the fourth/​tenth century, there were women who demanded the right to occupy positions which had heretofore been reserved for men.” 85. Ibn Battuta points only to one of his wives who was somewhat educated. She is the same Hurnasab who was also the only wife to be named. “She was a pious woman who used to spend



Concubines on the Road

j 185

part of the night in prayers and to compose wirds for the recollection of God (High and mighty is He) and she bore a daughter by me. I do not know what God has done with them both. She used to read but not to write” (Ibn Battuta, Travels, 3:719). 86. Ibn Battuta, Selections, 335–​336, and Riḥla, 697. 87. Hunwick and Powell, African Diaspora, 112; Lydon, “Slavery, Exchange and Islamic Law,” 125 The right of ownership applied to runaway slaves, but Ibn Battuta (Selections, 237)  only reports two male slave defections from his household (and one by a freed man). 88. Lydon, “Slavery, Exchange and Islamic Law,” 145. 89. Ware, “Slavery in Islamic Africa, 1400–​1800,” 55. 90. See, for example, Berkey, Transmission of Knowledge. 91. Frenkel, “Women in Late Mamluk Damascus,” 410, 416. 92. About al-​Ghazali’s position, see, for example, Farah, Marriage and Sexuality in Islam: If the husband is diligent in teaching her, she will not have to go out and ask the ulema about it . . .. Otherwise, she has the right to go out and ask; in fact, she is obligated to do so, and the husband would be in defiance should he prevent her. No matter how much she learns about her obligations, she should not go out to attend a dhikr nor to receive instruction in superfluous knowledge without the consent of her husband . . .. If he has several wives, then he should deal equitably with them and not favor one over the other; should he go on a journey and desire to have one [of his wives] accompany him, he should cast lots (aqra`) among them. For such was the practice of the Messenger. (103) Also in Frenkel, “Women in Late Mamluk Damascus,” passim. 93. Martel-​Thoumian, “Review of Carl F. Petry,” 213. 94. Frenkel, “Women in Late Mamluk Damascus,” 417–​418: “Some scholars even project a male world-​vision, giving the impression that writers felt threatened by women. Their writings attest to the conspicuous presence of women in public and can be interpreted as a clear sign of women’s active participation in the urban scene.” For pursuit of education among the Mamluks see Mauder, Gelehrte Krieger. See also Chamberlain, Knowledge and Social Practice in Medieval Damascus. 95. Frenkel, “Women in Late Mamluk Damascus,” 417. 96. Ibn Battuta, Travels, 2:414. 97. Ibn Battuta, Selections, 334, and Rihla, 695. 98. African slaves were common and sometimes highly prized in the Near East and South Asia, but it appears that light-​skinned, or white, concubines were always more desirable and more expensive because they were more rare. However, some dark women were thought to possess special sexual attraction. 99. Ghazali in Roded, Women in Islamic Biographical Collections, 162, 166. 100. Athamina, “How Did Islam Contribute to Change,” 394. 101. Roded, Women in Islamic Biographical Collections, 165–​166. 102. The master’s child by a slave mother would be entitled to inherit from his father, but the freed concubine was not entitled to inherit from her former master. 103. Ibn Battuta, Selections, 237. 104. Humphreys, “Politics of the Mamluk Sultanate,” 229. 105. al-​Kafi, al-​Masā’il al-​Kafiyya, 8. Cited in Bouhdiba, Sexuality in Islam, 69.

186 i

Marina A. Tolmacheva

Bibliography Primary Sources Ibn Battuta. al-​Riḥla. Edited by Hasan Abd al-​Sami Muhsin. Cairo:  Majm` al-​lugha al-​ arabiyya, 1965. —​—​—​. Travels in Asia and Africa, 1325–​1354. Translated and selected with an introduction and notes by H. A. R. Gibb. London: Routledge & Kegan Paul, 1929. [Reprinted in 1983]. —​—​—​. The Travels of Ibn Battuta: A.D. 1325–​1354. 5 vols. Cambridge: Haklyut Society, 1958–​2001. [Vols. I, II, and III, edited by C. Defrémery and B. R. Sanguinetti, and translated with revisions and notes from the Arabic text by H.  A. R.  Gibb. Vol. IV, translated by C.  F. Beckingham. Vol. V: index, compiled by A. D. H. Bivar. Haklyut Society, second series, vols. 110, 113, 143, 178, 190.] Ibn Shahriyar, Buzurg. The Book of the Marvels of India. From the Arabic by L. Marcel Devic. Translated into English by Peter Quennel. London: G. Routledge & Sons, 1928. al-​Kafi, Muhammad ibn Yusuf. Al-​Masa’il al-​Kafiyya. Cairo: Matba`at Hijazi, 1934.

Secondary Sources Abd ar-​Raziq, Ahmad. La Femme au temps des Mamlouks en Égypte. Le Caire: Institut Français d’Archéologie orientale du Caire, 1973. Amitai-​Preiss, Reuven. “Northern Syria between the Mongols and Mamluks: Political Boundary, Military Frontier, and Ethnic Affinities.” In Frontiers in Question: Eurasian Borderlands, 700–​ 1700, edited by Daniel Power and Naomi Standen, 128–​152. London: Macmillan, 1999. Athamina, Khalil. “How Did Islam Contribute to Change in the Legal Status of Women: The Case of the Jawārī or the Female Slaves.” Al-​Qantara 27, no. 1 (2007): 383–​408. Behrens-​Abouseif, Doris. “The Mahmal Tradition and the Pilgrimage of the Ladies of the Mamluk Court. Mamluk Studies Review 1 (1997): 87–​96. Belarbi, Mokhtar. “Rhétorique de la difference dans Présent à ceux qui aiment réfléchir sur les curiosities des villes et les merveilles des voyages d’Ibn Battûta.” In Tropes du voyage: Les Rencontres, edited by Aboubakr Chraïbi, 107–​122. Paris: L’Harmattan, 2010. Berkey, Jonathan. The Transmission of Knowledge in Medieval Cairo. Princeton, NJ:  Princeton University Press, 1992. Bon, Ottaviano. The Sultan’s Seraglio: An Intimate Portrait of Life at the Ottoman Court ( from the Seventeenth-​Century Edition of John Withers). Introduced and annotated by Godfrey Goodwin. London: Saqi Books, 1996. Bouhdiba, Abdelwahab. Sexuality in Islam. Translated from the French by Alan Sheridan. London: Routledge & Kegan Paul, 1985. Brack, Yoni. “A Mongol Princess Making hajj:  The Biography of El Qutlugh Daughter of Abagha Ilkhan (r. 1265–​82).” Journal of the Royal Asiatic Society (Third Series), 21, no. 3 (2011): 331–​359. Brockopp, Jonathan E. Early Maliki Law: Ibn `Abd al-​Hakam and His Major Compendium of Jurisprudence. Leiden: Brill, 2000. Brunschvig, Robert. “`Abd.” In The Encyclopaedia of Islam, 2nd ed., edited by P. Bearman et al. Leiden: Brill Publishing Online, 2012.



Concubines on the Road

j 187

Burckhardt, John Lewis. Travels in Arabia:  Comprehending an Account of Those Territories in Hedjaz Which the Mohammedans Regard as Sacred. London, 1829. [Reprinted in 1968, London: Frank Cass.] Chamberlain, Michael. Knowledge and Social Practice in Medieval Damascus, 1190–​ 1350. Cambridge: Cambridge University Press, 1994. Chraïbi, Aboubakr, ed. Tropes du voyage: Les Rencontres. Paris: L’Harmattan, 2010. Cilardo, Agostino. “The Transmission of the Patronate in Islamic Law.” In Miscellanea Arabica et Islamica, edited by F. DeLong, 31–​52. Leuven: Uitgeverij Peeters en Department Orientalistiek, 1993. Dunn, Ross E. The Adventures of Ibn Battuta, a Muslim Traveler of the 14th Century. 2nd ed. Berkeley: University of California Press, 2012. El-​Azhari, Taef K. “The Role of Salĝūqid Women in Medieval Syria.” In Egypt and Syria in the Fatimiod, Ayyubid and Mamluk Eras, IV:  Proceedings of the 9th and 10th International Colloquium Organized at the Katholieke Universiteit Leuven in May 2000 and May 2001, edited by U. Vermeulen and J. Van Steenbergen, 111–​126. Leuven–​Dudley, MA:  Uitgeverij Peeters, 2005. Encyclopaedia of Islam. 2nd ed., edited by P. Bearman et al. Leiden: Brill Publishing Online, 2012. Epstein, Stephen A. Purity Lost: Transgressing Boundaries in the Eastern Mediterranean, 1000–​ 1400. Baltimore, MD: Johns Hopkins University Press, 2006. Euben, Roxanne L. Journeys to the Other Shore:  Muslims and Western Travelers in Search of Knowledge. Princeton, NJ: Princeton University Press, 2006. Farah, Madelain. Marriage and Sexuality in Islam:  A  Translation of al-​Ghazali’s Book on the Etiquette of Marriage from the Ihya’. Salt Lake City: University of Utah Press, 1984. Ferrand, Gabriel. “Le Tuḥfat al-​albāb de Abu Hamid al-​Andalusi al-​Ğarnaṭi édité d’après les MSS. 2167, 2168, 2170 de la Bibliothèque Nationale et le MS. d’Alger.” Journal Asiatique July–​ September (1925): 1–​148; October–​December (1925): 193–​204. Frenkel, Yehoshua. “Women in Late Mamluk Damascus in the Light of Audience Certificates (samaa`aat).” In Egypt and Syria in the Fatimiod, Ayyubid and Mamluk Eras, IV: Proceedings of the 9th and 10th International Colloquium organized at the Katholieke Universiteit Leuven in May 2000 and May 2001, edited by U. Vermeulen and J. Van Steenbergen, 409–​424. Leuven–​ Dudley, MA: Uitgeverij Peeters, 2005. Ghazoul, Ferial J. “Medieval Anthropology:  Arab Travelers Venturing beyond the Abode of Islam.” In Tropes du voyage:  Les Rencontres, edited by Aboubakr Chraïbi, 89–​ 105. Paris: L’Harmattan, 2010. Griffel, Frank. “Review of Euben, Roxanne L.  Journeys to the Other Shore:  Muslims and Western Travelers in Search of Knowledge. Princeton, NJ: Princeton University Press, 2006.” International Journal of Middle Eastern Studies 44, no. 1 (2012): 172–​174. Guo, Li. “Tales of a Medieval Cairene Harem:  Domestic Life in al-​Biqa`i’s Autobiographical Chronicle.” Mamluk Studies Review 9, no. 1 (2005): 101–​121. al-​Harithy, Howayda. “The Patronage of al-​Nasir Muhammad ibn Qalawun, 1310–​1341. Mamluk Studies Review 4 (2000): 220–​224. Heywood, Colin. “The Frontier in Ottoman History: Old Ideas and New Myths.” In Frontiers in Question:  Eurasian Borderlands, 700–​1700, edited by Daniel Power and Naomi Standen, 228–​250. London: Macmillan, 1999.

188 i

Marina A. Tolmacheva

Hunwick, John, and Eve Troutt Powell. The African Diaspora in the Mediterranean Lands of Islam. Princeton, NJ: Markus Wiener, 2002. Johnson, Kathryn. “Mamluk Accounts of the Pilgrimage to Mecca of the Khawand al-​Kubraa (Senior Wife of the Sultan).” Studia Islamica 91 (2000): 107–​131. Keddie, Nikki R. “Introduction.” In Women in Middle Eastern History: Shifting Boundaries in Sex and Gender, edited by Nikki R. Keddie and Beth Baron, 3–​22. New Haven, CT:  Yale University Press, 1991. Keddie, Nikki R., and Beth Baron, eds. Women in Middle Eastern History: Shifting Boundaries in Sex and Gender. New Haven, CT: Yale University Press, 1991. Kruk, Remke. “Ibn Battuta: Travel, Family Life, and Chronology: How Seriously Do We Take a Father?” Al-​Qantara 16, no. 2 (1995): 369–​384. Lambton, A. K.  S. State and Government in Medieval Islam:  An Introduction to the Study of Islamic Political Theory: The Jurists. Oxford: Oxford University Press, 1981. Lapidus, Ira M. Muslim Cities in the Later Middle Ages. Cambridge, MA: Harvard University Press, 1967. Lewis, Bernard. Race and Slavery in the Middle East:  An Historical Enquiry. Oxford:  Oxford University Press, 1990. Lovejoy, Paul E. “Concubinage and the Status of Women Slaves in Early Colonial Northern Nigeria.” Journal of African History 29, no. 2 (1988): 245–​266. Lutfi, Huda. “The Construction of Gender Symbolism in Ibn Sīrīn’s and Ibn Shāhīn’s Medieval Arabic Dream Texts.” Mamluk Studies Review 9, no. 1 (2005): 123–​161. —​—​—. “Manners and Customs of Fourteenth-​Century Cairene Women:  Female Anarchy Versus Male Shar`i Order in Muslim Prescriptive Treatises.” In Women in Middle Eastern History: Shifting Boundaries in Sex and Gender, edited by Nikki R. Keddie and Beth Baron, 99–​121. New Haven, CT: Yale University Press, 1991. —​—​—​. “Al-​Sakhāwī’s Kitāb al-​Nisā’ as a Source for the Social and Economic History of Muslim Women during the Fifteenth Century A.D.” Muslim World 71, no. 2 (1981): 104–​124. Lydon, Ghislaine. “Slavery, Exchange and Islamic Law: A Glimpse from the Archives of Mali and Mauritania.” African Economic History 33 (2005): 117–​148. Mackintosh-​Smith, Tim. Landfalls:  On the Edge of Islam with Ibn Battutah. London:  John Murray, 2010. Marmon, Shaun. “Domestic Slavery in the Mamluk Empire: A Preliminary Sketch.” In Slavery in the Islamic Middle East, edited by Shaun Marmon, 1–​24. Princeton, NJ: Markus Wiener, 1999. —​—​—​, ed. Slavery in the Islamic Middle East. Princeton, NJ: Markus Wiener, 1999. Martel-​Thoumian, Bernadette. “Review of Carl F. Petry: The Criminal Underworld in a Medieval Islamic Society: Narratives from Cairo and Damascus under the Mamluks.” Chicago: Middle East Documentation Center, 2012. Bulletin of the School of Oriental and African Studies 77, no. 1 (2014): 212–​214. Mauder, Christian. Gelehrte Krieger: Die Mamluken als Träger arabischsprachiger Bildung nach al-​Afadi, al-​Maqrizi und weiteren Quellen. Arabistische Texte und Studies. Hildesheim, Zurich: Georg Olms Verlag, 2012. Metcalf, Barbara D. “Ibn Battuta Meets Shah Jalal al-​Din Tabrizi in Bengal.” In Islam in South Asia in Practice, edited by Barbara D. Metcalf, 139–​143. Princeton, NJ: Princeton University Press, 2009.



Concubines on the Road

j 189

Mez, Adam. The Renaissance of Islam. London: Luzac, 1937. [Reprinted in 1975, New York: AMS Press.] Miloslavskii, G. V. Ibn Battuta. Moscow: Mysl’, 1974. Power, Daniel, and Naomi Standen, eds. Frontiers in Question: Eurasian Borderlands, 700—​1700. London: Macmillan, 1999. Power, Daniel. “Introduction.” In Frontiers in Question: Eurasian Borderlands, 700–​1700, edited by Daniel Power and Naomi Standen, 10–​31. London: Macmillan, 1999. Rapoport, Yossef. Marriage, Money and Divorce in Medieval Islamic Society. Cambridge: Cambridge University Press, 2005. —​—​—​. “Women and Gender in Mamluk Society: An Overview.” Mamluk Studies Review 11, no. 2 (2007): 1–​47. Reese, Scott S. The Transmission of Learning in Islamic Africa. Leiden: Brill, 2004. Roded, Ruth. Women in Islam and the Middle East: A Reader. London: I.B. Tauris, 1999. —​—​—​. Women in Islamic Biographical Collections:  From Ibn Sa`d to Who’s Who. Boulder, CO: Lynne Rienner, 1994. Rosen, Lawrence. Varieties of Muslim Experience: Encounters with Arab Political and Cultural Life. Chicago: University of Chicago Press, 2008. Tazi, Abdelhadi. “Al-​Rahâla al-​arab wa’l-​muslimun, iktishaf al-​akhar.” In Actes du colloque organisé par le Ministère de la Culture et le Projet géographique arabe les 14, 15,16 et 17 novembre 2003. Casablanca: Imprimerie al-​Najah al-​Jadida, 2005. Tolmacheva, Marina. “Ibn Battuta on Women’s Travel in the Dar al-​Islam.” In Women and the Journey:  The Female Travel Experience, edited by Bonnie Frederick and Susan H. McLeod, 119–​140. Pullman: Washington State University Press, 1993. —​—​—​. “Ibn Battūtah (Abū `Abd Allāh Muhammad Ibn Battūtah).” In Essays in Arabic Literary Biography, edited by Joseph Lowry and Devin Stewart, 127–​137. Wiesbaden: Harrassowitz, 2008. —​—​—​. “Medieval Muslim Women’s Travel:  Defying Distance and Danger.” World History Connected 10, no. 2 (2013). http://​worldhistoryconnected.press.illinois.edu/​10.2/​forum_​tolmacheva.html. Touati, Houari. Islam and Travel in the Middle Ages. Translated by Lydia G. Cochrane. Chicago: University of Chicago Press, 2010. Toru, Miura. “Urban Society in Damascus as the Mamluk Era Was Ending.” Mamluk Studies Review 10, no. 1 (2006): 157–​193. Tsugitara, Sato. “Slave Traders and Karimi Merchants during the Mamluk Period: A Comparative Study.” Mamluk Studies Review 10, no. 1 (2006): 141–​156. Vermeulen, U., and Van Steenbergen, J., eds. Egypt and Syria in the Fatimid, Ayyubid and Mamluk Eras, IV: Proceedings of the 9th and 10th International Colloquium organized at the Katholieke Universiteit Leuven in May 2000 and May 2001. Leuven–​Dudley, MA: Uitgeverij Peeters, 2005. Ware, Rudolph T, III. “Slavery in Islamic Africa, 1400–​1820.” In The Cambridge History of Slavery, Vol. 3:  AD 1420–​AD 1804, edited by David Eltis & Stanley Engerman, 81–​110. New York: Cambridge University Press, 2011. Whalen, Brett Edward, ed. Pilgrimage in the Middle Ages:  A  Reader. Toronto:  University of Toronto Press, 2011.

9 Slaves in Name Only Free Women as Royal Concubines in Late Timurid Iran and Central Asia

Usman Hamid

i  The possibilities of what it meant to be a concubine in a late medieval setting were several. Conventional wisdom holds that concubinage was a gendered function of slavery—​that is, a legal arrangement in which slave owners enjoyed licit sexual relations with their female property. Although this was certainly the view of Muslim jurists and a historical reality for many enslaved women, this understanding of concubinage fails to take into account the different ways in which it was practiced in history. One notable example is found in the households of Timurid princes, who, as descendants of the famous conqueror Temur, or Tamerlane (1336–​1405 ce), ruled collectively over the regions of Khurasan in eastern Iran and Transoxiana in Central Asia during the fifteenth century. Timurid sources from the later period list numerous women as royal concubines who were not slaves. Instead, they belonged to notable Muslim families whose pedigrees contemporaries deemed prestigious enough to record for posterity. This unusual feature of Timurid concubinage raises a series of questions. First, what motivated the later Timurids to take freeborn Muslim women as concubines? The highly competitive political environment of late fifteenth-​century Iran and Central Asia required would-​be potentates seek out a wide network of supporters, which was achieved in part through securing marital alliances with important amirs and local notables. Given the Islamic proscription of taking more than four wives, Timurid princes felt more at ease in engaging in a legal fiction that permitted them to take freeborn Muslim women as concubines rather than exceeding the number of legal marriages allowed to them by the law. 190



Slaves in Name Only

j 191

Second, if concubinage was a means of securing power and authority, was it possible for concubines themselves to participate in the broader political economy of Timurid Iran and Central Asia? If so, how did contemporaries look on their involvement? Here, the career of Khadija Begi Agha, a concubine who eventually became the legal wife of a Timurid prince, serves as an instructive example of a politically active consort who cultivated a diverse network of allies and supporters that she mobilized in times of dynastic conflict. The strong reaction she engendered from her contemporaries suggests that, although the Timurids are well known for their inclusion of women as active participants in dynastic politics, women who traced their origins as concubines in royal households nevertheless faced censure for their involvement. Freeborn Muslim Women as Concubines Genealogical records for Timurid princes list a number of freeborn Muslim women as concubines. For example, in the anonymous genealogy Mu‘izz al-​Ansab fi Shajarat al-​ Ansab (The Glorifier of Genealogies Concerning the Ancestral Tree), the compiler includes the parentage of a number of concubines for Sultan-​Husayn Bayqara (r. 1469–​1506), the penultimate ruler of Timurid Herat.1 Among these were daughters of important military commanders such as Latifa-​Sultan Aghacha, the daughter of Amir Sultan-​Husayn of Chaharshamba, and Zubayda-​Sultan, the daughter of Amir Khvaja Hasan-​Shaykh Temur Jalayir. Others belonged to notable families that had close ties to the royal household. For instance, Papa Aghacha, the daughter of one Khvaja Ahmad Ataka, was the foster sister of Sultan-​Husayn’s wife Apaq Begim, herself the daughter of the important military commander Amir Taj al-​Din Hasan ibn Amir Nizam al-​Din Charkas. Sultan-​Husayn was not alone in taking freeborn Muslim women as his concubines. His predecessor, Sultan-​Abu Saʿid Mirza, who had ruled both Samarqand and Herat until his death in 1469, is known to have had at least 35 wives and concubines, many of whom belonged to notable Muslim households, based on their fathers’ names and titles.2 Before interpreting this information, it is fair to ask whether there is a possibility that these lineages could have been fabricated. After all, embellishing a royal consort’s pedigree was not entirely unheard of during the late Timurid period. Take, for example, the case of Qara Koz Begim, the wife of Umar-​Shaykh Mirza, the ruler of Timurid Fergana. In his memoirs, Umar-​Shaykh’s son, Zahir al-​Din Muhammad Babur (1483–​1530), the founder of the Mughal dynasty in India, wrote that on account of the love shown by his father to Qara Koz Begim, those who wished to please him attributed to her a royal Timurid lineage.3 However, it is unlikely that the information in Mu‘izz al-​Ansab was completely fictitious. Babur, who was Sultan-​Husayn’s third cousin once removed, was a scrupulous recorder of family lineages,4 and his memoirs, together with the work of the late Timurid historian Ghiyas al-​Din Khvandamir (d.c. 1535), verify the information provided in Mu‘izz al-​Ansab.5

192 i

Usman Hamid

It is worth pointing out that the inclusion of freeborn Muslim women did not preclude the presence of female slaves who served as concubines in royal households. In fact, a number of sources mention concubines who had been gifted as slaves. For example, at the time of Sultan-​Husayn’s marriage to Shahr-​Banu Begim, the daughter of Sultan-​Abu Said, she presented him with one of her slaves (mamluka), Mengli Bey Aghacha, who later became his concubine and mother to five of his children.6 Similarly, while campaigning in India, Babur received a gift of two Circassian slave girls (qiz) from the Safavid ruler of Iran, Shah Tahmasp I (r. 1524–​1576), during an exchange of embassies.7 More often than not, however, barring certain exceptions like Mengli Bey Aghacha, concubines who were slaves simply escaped the notice of Timurid authors because of their low status. Indeed, when recording the genealogies of various Timurid rulers, Babur mentioned frequently that he only recorded the names of important concubines and not the many others who were of no consequence.8 The inclusion of free women among the ranks of royal concubines, despite the ready availability of female slaves, suggests that Timurid practices of concubinage served a purpose other than simply ensuring dynastic continuity. The Timurid household was, at its core, a political institution. Household relations, whether based on kinship or service, served a clear function to legitimate and bolster a prince’s claim to power in a highly competitive environment. This was particularly true of Timurid marriages. With the adherence to the principle of corporate sovereignty, which permitted all of Tamerlane’s male heirs to lay a claim to his patrimony, princes sought to forge advantageous alliances with the families of important amirs by contracting marriages with women of Timurid and amirid birth.9 Because Islamic law permitted a maximum of four legal marriages at any given time, it appears that Timurid princes used concubinage to secure alliances and built a network of loyal supporters in lieu of marriage. Instead of flouting the Islamic proscription of taking more than four wives, the Timurids appear to have felt more at ease with perpetuating a legal fiction that permitted freeborn Muslim women to be taken as concubines.10 Although Timurid sources generally omit the circumstances under which they were acquired, a clear example of a political alliance being forged through the taking of a concubine is found in events of 1519, when Babur took Afghani Aghacha, the daughter of the Yusufzai chief Malik Shah Mansur, to secure her tribe’s support.11 Late fifteenth-​century authors appreciated the importance of well-​placed marital alliances for princes and extolled the virtues of marriage to women of high birth as opposed to union with female slaves. Perhaps the clearest exposition of this opinion is to be found in a work of political and ethical advice written by the leading theologian and jurist Jalal al-​Din Davani (d. 1502). Although the work was not written for a Timurid audience, but instead for the Aq Qoyunlu ruler, Uzun Hasan (d. 1478), and his son Sultan-​Khalil, it is nevertheless instructive because the political structure of the two neighboring polities were remarkably similar.12



Slaves in Name Only

j 193

Davani begins his discussion on the management of the household with a description of the ideal wife. He sets forth a laundry list of qualities she should possess: “the best women are those who are adorned with intelligence, piety, chastity, cleverness, modesty, kind-​heartedness, manners, an inclination to please their husband, and a dignified bearing.” Then, after emphasizing the importance of her ability to bear children, Davani argues that, “a freewoman is better than a slave for included [in the marriage] is the acquisition of followers and support from her close relations, conciliation with one’s enemies, mutual assistance in matters of livelihood, and guards against mean pedigree.”13 Here, Davani clearly lays out the benefits of a well-​placed marriage, particularly for rulers. Through marriage, he argues, one may gain supporters, broker peace with one’s opponents, forge mutually beneficial alliances with notable families, and ensure that one’s heirs bear a lineage that is beyond reproach. The disregard for Muslim legal codes regulating marriage and concubinage did not go uncommented on by contemporaries. In his memoirs, Babur disapproved of the practice of taking free Muslim women as concubines, deeming the relationships to be unlawful. Reflecting on Sultan-​Husayn’s legacy, Babur mused that it was strange that such a great ruler should have only three legitimate sons with the remaining 11 being the result of unlawful sexual relations (valad al-​zina).14 In Babur’s estimation, only Sultan-​Husayn’s children born to legally wed wives were legitimate. The three were Badi al-​Zaman Mirza, the son of Sultan-​Husayn’s first wife, a Timurid princess named Bika-​Sultan Begim (d. 1488);15 Haydar-​ Muhammad Mirza, the son of Payanda-​Sultan Begim, who was the daughter of Sultan-​Abu Said;16 and Muzaffar-​Husayn Mirza, Sultan-​Husayn’s second son with Khadija Begi. Meanwhile, the sons born to his freeborn Muslim concubines—​that is, Latifa-​Sultan Aghacha and Papa Aghacha—​were deemed illegitimate.17 Babur’s judgment is not surprising given that sexual relations between unmarried free men and women were considered unlawful in the eyes of Muslim jurists. Also excluded, however, were Sultan-​Husayn’s sons by the Uzbek slave Mengli Bey Aghacha. This omission is curious given that Islamic law deemed the children born to concubines as legitimate. Babur’s memoirs perhaps suggest that Timurid use of free Muslim women as concubines had cast the entire institution in disrepute in the eyes of some contemporaries. Given that Timurid concubines included women of different social status, with daughters of important amirs on the one hand and female slaves on the other, their standing in royal households must have reflected this difference to some degree. After all, members of Timurid court society were keen observers of social protocol and hierarchy.18 However, the task of reconstructing the hierarchies of Timurid concubinage remains a difficult one because contemporary authors rarely included information about concubines and low-​born women, even when they were mothers of prominent princes.19 One measure of social rank is the order in which concubines appear in Timurid genealogies such as the Mu‘izz al-​Ansab. Although a genealogy cannot convey the dynamic social hierarchies of the royal household, it can nevertheless serve as a barometer of status, painting in broad strokes the obscured complexities of Timurid concubinage.

194 i

Usman Hamid

The entry for Sultan-​Husayn is particularly instructive because it divides clearly his wives and concubines into various strata.20 Listed under the heading “wives and concubines” (khavatin va qumayan) are the names of his six wives, two of whom he divorced, as well as of two concubines. The two concubines were both daughters of important amirs: Zaynab-​Sultan Aghacha, the daughter of Amir Taj al-​Din Hasan b. Amir Nizam al-​Din Charkas; and Latifa-​Sultan Aghacha, the daughter of Amir Sultan-​Husayn of Chaharshamba. Although the Mu‘izz al-​Ansab clearly follows two patterns in listing the eight women—​that is, it places wives before concubines and orders them according to their natal family’s political rank—​the genealogy does not differentiate explicitly between wives and concubines. If a historian were to rely on the Mu‘izz al-​Ansab alone, it would be impossible to state with absolute certainty which of the eight women were among Sultan-​Husayn’s wives (khavatin) and which were among his concubines (qumayan). Thus, although acknowledging the subtle difference in status among the eight women, the text appears to suggest that concubines who were daughters of important amirs were comparable in rank with legal wives. Unlike the daughters of important amirs, women belonging to notable Muslim families and female slaves were listed separately under the subheading sarariy, the Arabic term for concubines. These included the Uzbek slave Mengli Bey Agacha; Papa Aghacha, the daughter of Khwaja Muhammad Ataka; and Biki Sultan, whose father’s name is not provided. One notes that the Mu‘izz al-​Ansab uses the Mongol term for concubines, qumayan (sing., quma), for daughters of amirs, whereas women of lower rank were listed under the Arabic sarariy (sing., surriya). This use of different languages in categorizing concubines was not accidental. After all, the Timurids often used Turko-​Mongolian titles to denote high-​ranking positions within their royal households.21 This polyglotism was a way for the Mu‘izz al-​Ansab’s compilers to indicate a difference in status. Although the concubines of amirid extraction were distinguished from those having a less prestigious background, no pattern can be discerned for how the text orders the sarariy. The genealogy lists the Uzbek slave Mengli Bey Agacha before Papa Aghacha, who was certainly a woman of significantly more substantial means. After all, her family was integrated closely with the royal household; she was a member of Sultan-​Husayn’s wife Apaq Begim’s foster family. Nor does it order things according to the number of children borne by each woman; Papa Aghacha mothered more sons and more daughters than Mengli Bey Agacha.22 It is possible the Mu‘izz al-​Ansab lists the names of the women based on the order in which they became concubines, but this hypothesis cannot be confirmed because the sources do not provide the relevant dates. The clearest indication that the Mu‘izz al-​Ansab preserved the different statuses of these women is seen in the case of Zubayda-​Sultan. The lone name mentioned under the subheading dukhtaran-​i khana—​Persian for slave girl or, alternatively, house-​born female slave; the book’s placement of Zubayda-​Sultan seems to be a mistake. As the daughter



Slaves in Name Only

j 195

of a high-​ranking amir, she should have been included in the list of qumayan along with other freeborn Muslim women of amirid background. However, the case of Zubayda-​ Sultan reflects the uncertainties of concubinage for even high-​born women and serves as a cautionary tale of what could happen when allegiances and alliances shifted in a fluid political environment. It is unclear when exactly Sultan-​Husayn took Zubayda-​Sultan as a concubine, but it must have been in the days when the Timurid ruler wished to cement his relationship with her father, Khvaja Hasan-​Shaykh Temur Jalayir. The amir was Sultan-​Husayn’s pick for the governor of Astarabad after the conquest of Herat in 1469. However, within the span of three years, Sultan-​Husayn charged Hasan-​Shaykh Temur with treason and had him executed in 1472. According to the historian Khvandamir, although Hasan-​Shaykh Temur was executed for his attempts to gather men and arms to instigate strife, he had, on numerous previous occasions, displayed disloyalty and ingratitude, not the least of which was when he threw in his lot with Sultan-​Husayn’s rival Mirza Yadgar-​Muhammad (d. 1470).23 Hasan-​Shaykh Temur, who had been a paramount amir, was also no longer a politically expedient ally for Sultan-​Husayn. Even if his loyalty was assured, Hasan-​Shaykh Temur had taken a leading role in opposing the Timurid ruler’s newly fiscal reform program intended to replenish his depleted coffers.24 With her father executed, Zubayda-​ Sultan no longer served any political purpose and found herself disgraced, judging from her relegation to the very end of the Mu‘izz al-​Ansab’s list of Sultan-​Husayn’s consorts. Becoming a concubine was thus a high-​stakes gamble. But, not all shared Zubayda-​ Sultan’s fate or that of the many others of whom little is written. Some engaged actively in dynastic politics, and none more so than Khadija Begi Agha. The rest of this chapter is devoted to her career. Royal Concubines in Timurid Politics and Their Critics:  The Case of Khadija Begi Agha No royal concubine had as successful a career as Khadija Begi Agha. The daughter of a Muslim nobleman, Khadija Begi Agha was the concubine of two different Timurid rulers before becoming Sultan-​Husayn’s wife and a powerful figure in the political landscape of late fifteenth-​century Iran. As his consort, Khadija Begi Agha cultivated a wide network of political allies who supported her in her bid to ensure that her son, Muzaffar-​Husayn, would succeed Sultan-​Husayn. A formidable individual, she engendered a variety of reactions from the members of the royal household and the Timurid literary elite. An examination of her life demonstrates the degree to which social mobility was possible in the royal household for a concubine, and the means through which a woman could shape dynastic politics. According to Babur, Khadija Begi Agha was initially a concubine of Sultan-​Husayn’s predecessor, Sultan-​Abu Said, with whom she had a daughter named Aq Begim.25 After Sultan-​Abu Said’s death in 1469, Khadija Begi Agha arrived at court, where Sultan-​Husayn

196 i

Usman Hamid

is said to have fallen in love with her and taken her as his concubine.26 Timurid sources make clear that although Khadija Begi Agha would later become Sultan-​Husayn’s legal wife, she was initially his concubine. The compiler of the Mu‘izz al-​Ansab was sure to include a note under her name declaring that Khadija Begi Agha “was at first a concubine (quma) and later entered into a legal marriage (nikah).”27 Similarly, Babur notes that she advanced from the rank of a concubine to that of begim—​that is, wife.28 There is little evidence to suggest that Khadija Begi Agha’s success was on account of her family or that she was taken as a concubine to forge an important alliance. At best, she came from a middling amirid family.29 The Mu‘izz al-​Ansab identifies her as the daughter of one Amir Muhammad Sariq, who was the son of Amir Muhammad Khvaja—​both of whom have yet to be identified clearly.30 Although the Mu‘izz al-​Ansab notes the use of the high-​ ranking title amir for her father and grandfather, the fact that no source makes note of them suggests they were not particularly important figures at court. Instead, they were probably local notables of some sort. In fact, even Babur, who recorded assiduously the genealogies of others, makes no mention of her family in his notice on Khadija Begi Agha.31 It is therefore unlikely that Sultan-​Husayn’s initial impetus to take Khadija Begi Agha as a concubine was driven by a desire to inherit an alliance forged between his predecessor and her family. After all, the Mu‘izz al-​Ansab omits her name from among Sultan-​Abu Said’s wives and concubines. Although a woman by the name of Khadija Begi is listed, her father’s name is given as Mawlana Nasr al-​Din.32 Thus, her absence from the section on Sultan-​Abu Said in the Timurid genealogy suggests her family was not important enough to merit mentioning the relationship. Such omissions were not unusual. Babur, for example, does not include the names of the concubines who mothered his cousins, Rajab-​Sultan and Muhibb-​Sultan, both of whom were Timurid princesses.33 Babur was not the only author to provide a partial list of a Timurid prince’s concubines. The Mu‘izz al-​Ansab does not list any names for the wives or concubines of Sultan-​Abu Said’s sons—​Sultan-​Ahmad (d. 1494), Sultan-​Mahmud (d. 1495), and Umar-​Shaykh (d. 1494)—​despite providing the names of their amirs and administrators.34 And although it is true that the Mu‘izz al-​Ansab does not list anyone by the name of Aq Begim among the daughters of Sultan-​Abu Said,35 it is unlikely that Babur’s information about Khadija Begi Agha and her daughter would have been incorrect. After all, Sultan-​Abu Said was Babur’s paternal grandfather and Aq Begim was therefore his aunt. Furthermore, Babur’s daughter Gulbadan Begim attests numerous times to Aq Begim’s parentage in her own memoirs.36 Khadjia Begi Agha came to prominence only after her marriage to Sultan-​Husayn. Although contemporary sources do not mention the date of their wedding, it likely took place sometime after the birth of their first son, Shah-​Gharib Mirza, in 1471, and before the birth of their second, Muzaffar-​Husayn, a few years later.37 That it was skipped over by contemporaries suggests that they viewed it as hardly worth notice. Khadija Begi Agha’s elevation to the status of a legally wed wife was neither because she gave birth to a son nor simply because of Sultan-​Husayn’s affections. Other concubines, who belonged to noble



Slaves in Name Only

j 197

families and had sons by Sultan-​Husayn or were dearly beloved by him, never became wives.38 Whatever his reasons for marrying Khadija Begi Agha, they were unique to her. After her marriage, Khadjia Begi Agha focused her energies on ensuring the succession of Muzaffar-​Husayn, her only son to survive Sultan-​Husayn.39 Although their eldest son, Shah-​Gharib (d. 1497), was given the governorship of Herat during the reign of Sultan-​Husayn, there is no evidence on hand to suggest Khadija Begi Agha had a role in his appointment.40 Despite the prestigious position, Shah-​Gharib was never a viable contender for his father’s throne. No well-​placed marriage was contracted for him and, although it is true that Babur and Khvandamir remembered him fondly, thus describing him as good-​natured and a patron of men of learning, he was noted for having a physical disability and being generally frail.41 So Khadija Begi Agha focused her attention on Muzaffar-​Husayn’s political future, but in doing so she set in motion events that precipitated a major political crisis in late-​fifteenth-​century Iran.42 In 1497, not long after Shah-​Gharib’s death, Khadija Begi Agha secured successfully the governorship of Astarabad for her son Muzaffar-​Husayn.43 In doing so, she not only obtained for him an important concession, but also undermined his main rival, Sultan-​ Husayn’s eldest son, Badi al-​Zaman, because the governorship had been promised previously to his son Muhammad-​Muʾmin Mirza at his circumcision feast.44 Feeling threatened, Badi al-​Zaman went into active rebellion against his father,45 which only served to bolster Muzaffar-​Husayn’s position further, the latter already Sultan-​Husayn’s favorite.46 In attempt to put an end to the growing rift, Sultan-​Husayn’s close adviser Ali Shir Navai (1441–​1501) offered to mediate between father and son. He secured from Sultan-​ Husayn the promise to offer clemency to Badi al-​Zaman if he were to cease his rebellion. At first the negotiations appeared successful, but they fell apart when Badi al-​Zaman intercepted a secret letter that incriminated Sultan-​Husayn in a plot against the prince. According to Khvandamir, Sultan-​Husayn’s duplicity was on account of the advice given to him by members of his inner circle, particularly his vizier, Khvaja Qivam al-​ Din Nizam al-​Mulk Khvafi.47 No explanation is given for why Khvaja Nizam al-​Mulk, a longtime political ally and personal friend of Ali Shir Navai,48 sabotaged the negotiations. However, it seems likely that he acted on behalf of Khadija Begi Agha, with whom he developed a close relationship, as evidenced by what transpired next. In the absence of a peaceful resolution, two sets of fathers and sons faced off on the battlefield. Sultan-​Husayn marched against Badi al-​Zaman and Muzaffar-​Husayn set off to fight Muhammad-​Mumin at Astarabad. On May 3, 1497, Muzaffar-​Husayn defeated Muhammad-​Mumin and sent him as a captive to Herat in silver chains.49 Astonishingly, the inhabitants of the city offered him a hero’s welcome. He appears to have been so beloved that when he was imprisoned in the Ikhtiyar al-​Din Fort, a mob formed quickly. Fearful that if Muhammad-​Mumin lived on, it might jeopardize her son’s life, Khadjia Begi Agha arranged his elimination. She turned to Khvaja Nizam al-​Mulk and his sons to support her cause, which they did.50 That evening, heeding the advice of Khvaja Nizam al-​Mulk, an intoxicated Sultan-​Husayn issued an order to Bibi Muhibb-​Jangi, the warden

198 i

Usman Hamid

of the Ikhtiyar al-​Din Fort,51 to carry out the execution. News of the prince’s death caused widespread mourning among the inhabitants of Herat and panic among some of Sultan-​ Husayn Bayqara’s other sons who, out of fear, rose up in rebellion.52 The relative ease with which Khadija Begi Agha dealt with Muhammad-​Mumin demonstrates the efficiency with which she could mobilize members of her political network. Although she was not held responsible for the death of Muhammad-​Mumin, Ali Shir Navai saw to the punishment of Khvaja Nizam al-​Mulk and his sons, and steps against Khadija Begi Agha’s network. Less than a year after Muhammad-​Mumin’s death, Ali Shir Navai orchestrated the downfall and execution of Khvaja Nizam al-​Mulk and his family in June 1498.53 Although there was a host of reasons behind Khvaja Nizam al-​Mulk’s execution, the inhabitants of Herat believed it to be connected to Muhammad-​Mumin’s murder.54 The deaths of Khvaja Nizam al-​Mulk and his family did not compromise entirely Khadija Begi Agha’s ability to shape the course of Timurid politics. After all, she had cultivated relations with various factions at court that included bureaucrats and military men. The extent of her network became clear at the time of Sultan-​Husayn’s death and the succession of Badi al-​Zaman and Muzaffar-​Husayn. During the final years of Sultan-​Husayn’s life, Khadija Begi Agha became particularly vigilant of the resources available to her son in his bid for the throne. When Sultan-​Husayn fell ill and Badi al-​Zaman called off his rebellion to offer allegiance to his sick father, Khadija Begi Agha entreated her husband to postpone the reunion until he was better. She convinced him that the absence of Muzaffar-​Husayn and the ruler’s weakened condition might lead his army to shift their loyalties from him to his eldest son. Sultan-​Husayn heeded his wife’s counsel and delayed meeting Badi al-​Zaman until he was in better health and Muzaffar-​Husayn was able to join them.55 At some point it became clear to Khadija Begi Agha that Badi al-​Zaman was unlikely to be passed over and so, in 1504, before Sultan-​ Husayn was to reunite with his son, she and Muzaffar-​Husayn met with him publicly outside Herat and “washed away with the water of amity and cordiality the dust of rancor that had clouded their relations.”56 This reconciliation did not mean Khadjia Begi Agha gave up on her ambitions for her son. Instead, in what was an unusual turn of events, Muzaffar-​ Husayn and Badi al-​Zaman would be declared co-​regents after their father’s death. Khvandamir, the historian, records the intense debate that followed the death of Sultan-​Husayn over his succession. Some argued that Badi al-​Zaman should rule independently, whereas others put forth that Muzaffar-​Husayn should be made co-​regent and have his name included in coinage and in the sermons for Friday prayers. Here, Khadija Begi Agha’s network of allies played a pivotal role. In addition to having great authority and sway, she was supported by Sultan-​Husayn’s chief amir Shuja al-​Din Muhammad Burunduq Barlas and his sons, particularly Mirza Ali Beg. It was on account of the fact that most of the army followed her and the Barlas amirs that the partisans of Badi al-​ Zaman capitulated and Muzaffar-​Husayn was declared co-​regent.57 In ensuring the accession of her son, Khadjia Begi Agha earned the condemnation of other royals, including Babur and his relative Muhammad Hayder Dughlat. Babur



Slaves in Name Only

j 199

described the co-​regency as “a strange arrangement; never has a joint kingship been heard of.” It was against the received wisdom encapsulated in Shaykh Sadi’s words in the Gulistan: “Ten poor men can sleep under one blanket, but two kings cannot fit into one clime.”58 Although royal etiquette meant they continued to show due deference to her in public and they enjoyed her hospitality in Herat, they castigated her privately in their writings. Muhammad Hayder Dughlat was unsparing in his criticism, describing Khadjia Begi Agha as “the instigator of all mischief.”59 Her actions engendered intense reactions. When Babur’s brother Jahangir Mirza fell sick after drinking too much in Herat, Muhammad Hayder Dughlat was quick to point the finger at her in his memoirs, claiming, “it was commonly rumored that Khadija Begi Agha was up to her old tricks and had poisoned his wine.”60 Not surprisingly, as royals who had their own political aspirations, men like Babur and Muhammad Hayder Dughlat found no fault with the ambitions of her son Muzaffar-​Husayn. They had, of course, a personal stake in the political maneuverings undertaken by royal concubines, which no doubt colored their perspective. Nowhere was this more apparent than in Babur’s harsh criticism of Zuhra Begi Agha, who attempted to ensure her son remained the ruler of Samarqand on the eve of the Uzbek takeover of Central Asia.61 Following the death of Sultan-​Mahmud Mirza in 1495, his son Sultan-​Ali Mirza managed to a gain control of Samarqand, the old Timurid capital, but his rule was precarious, facing as he did serious challenges from two sides. The first was his cousin Babur who, although previously cooperative, now coveted Samarqand for himself and was supported by local notables such as Khvaja Qutb al-​Din Yahya. The second was the formidable founder of the Uzbek Khanate, Muhammad Shibani Khan (d. 1510). In previous years, he had taken on a prominent role in Timurid politics. Believing Babur’s ambitions to be the greater threat, and perhaps seeing the Uzbek conquest of the city as inevitable, his mother and Sultan-​Mahmud Mirza’s favorite concubine, Zuhra Begi Agha, decided it was more expedient to forge an alliance with Muhammad Shibani Khan. Both Timurid and Uzbek sources claim that Zuhra Begi Agha offered herself in marriage to Muhammad Shibani Khan with the hope that her son would join the Uzbek Khan’s household and maintain control of Samarqand. The plan succeeded only briefly. Within a few days after meeting with the Uzbek conqueror, the Timurid prince was executed. Zuhra Begi Agha herself is said to have been pushed aside quickly.62 Babur, whose own ambitions were thwarted by her actions, disparaged his cousin, saying that, “by listening to the words of women, he removed himself from the circle of those of good repute.”63 Conclusion The study of royal concubinage in Timurid Iran and Central Asia demonstrates that institutions and practices regulating relations between the sexes manifested in different ways across history. These variations reflected the particular circumstances in which they operated. Thus, consideration of royal concubinage sheds further light not only on the

200 i

Usman Hamid

role of women in ruling households, but also on the broader political culture of which the practice was an integral part. For the Timurids, royal concubinage was a solution to the problem of uncertain loyalties and intense political competition. Thus, it was possible to find women of amirid birth, daughters of notable Muslim families, and female slaves all listed together as concubines in Timurid genealogies. The practice was not approved of universally, but, as is often the case with politics, pragmatism trumped legalism. Enmeshed in politics, concubinage allowed for rare moments of mobility and exercise of power, but was also a fraught venture, where women could be cast aside when they, or their relations, were no longer of use. Timurid concubinage also invites renewed reflection on the relationship between marriage and slavery. In her provocative study of Islamic legal thought, Kecia Ali has argued that slave ownership and its attendant discussions on property were central to how Muslim jurists understood and conceptualized the institution of marriage. According to this argument, “both were forms of control or domination exercised by one person over another.” This similarity is said to have often led to frequent comparisons between the two in Muslim legal thought.64 For the Timurids, the opposite dynamic appears to have been the case; the political expectations of marriage and its benefits came to define the terms by which royal concubinage was practiced. Although concubinage was never disentangled from slavery entirely, being a slave was no longer the defining feature of what it meant to be a concubine. Acknowledgments I extend my deepest thanks to Maria Eva Subtelny for providing generous feedback to an earlier draft of this chapter, as well as sharing her insights on late Timurid Iran and Central Asia. For his editorial suggestions, I am indebted to Matthew Gordon. Notes 1. First commissioned by the Timurid ruler Shahrukh in 1426 to record Tamerlane’s genealogy, the Muʿizz al-​Ansab fi Shajarat al-​Ansab (Glorifier of Genealogies Concerning the Ancestral Tree) was updated to include his descendants, with the final entry being Sultan-​Husayn’s son Badiʿ al-​ Zaman (d.c. 1514). A facsimile of one of the manuscripts for the Muʿizz al-​Ansab (MS, Bibliothèque nationale de France, Ancien fonds persan 67) has been published alongside a Russian translation, for which see Vokhidov, Istoriia Kazakhstana v persidskikh istochnikakh. All references to the text follow the folio numbers indicated by the editor. On the Muʿizz al-​ansab, see Woods, Timurid Dynasty, 1–​15. For Sultan-​Husayn’s concubines see Muʿizz al-​Ansab, fols. 159b–​160a. 2. Muʿizz al-​Ansab, fols. 154b–​155a; and Woods, Timurid Dynasty, 35–​36. 3. Babur, Baburnama, 1:24–​25 (=fol. 12b). 4. For a modern scholar’s estimation of Babur as a source for Timurid genealogy and the nature of his relationship with Sultan-​Husayn see, respectively, Subtelny, Timurids in Transition, 44, 46, and “Babur’s Rival Relations,” 109.



Slaves in Name Only

j 201

5.  For instance, Latifa-​Sultan Aghacha’s amirid background is mentioned in Khvandamir, Habib al-​Siyar, 4:320. For the English translation see Khvandamir, Habibu’s-​siyar, 2:511. That Papa Aghacha was Apaq Begim’s foster sister is mentioned by Babur, for which, see Baburnama, 2:352–​353 (=fol. 169b). Sultan-​Husayn would also marry Apaq Begim’s sister Zaynab-​Sultan. For Apaq Begim’s lineage and the circumstances surrounding her marriage see Subtelny, Timurids in Transition, 54, 174. For Hasan-​Shaykh Temur’s appointment see Khvandamir, Habib al-​Siyar, 4:142, and Habibu’s-​siyar, 2:424. Also see Subtelny, Timurids in Transition, 64. 6. Khvandamir, Habib al-​Siyar, 4:321, and Habibu’s-​siyar, 2:511. 7.  Babur, Baburnama, 3:654–​655 (=fol. 305a). Annette Beveridge has suggested tentatively that their names may have been Gulnar Aghacha and Nazgul Aghacha. See Begim, The History of Humayun, 122, note 3. 8. Babur, Baburnama, 1:24–​25 (=fol. 12b), 1:56–​57 (fols. 28a–​28b), 2:352–​353 (=fol. 169b). 9. For a discussion on the role of corporate sovereignty in Timurid politics, see Subtelny, “Babur’s Rival Relations,” 102–​118, and Timurids in Transition, 36–​38. For the significance of amirid wives, see Manz, “Women in Timurid Dynastic Politics,” 125, and Power, Politics and Religion in Timurid Iran, 174. One of the most influential and powerful female figures in fifteenth-​century Timurid politics was Gawharshad (d. 1457), the daughter of the important amir Ghiyas al-​Din Tarkhan and wife to Tamerlane’s eventual successor and son Shahrukh (r. 1409–​1447). See Manz, “Gowhar-​Šād Aga.” 10. It was not unusual for Muslim jurists made accommodations for Turko-​Mongolian customs that ran counter to Islamic legal precedent—​see Subtelny, Timurids in Transition, 27. 11.  Babur, Baburnama, 2:468–​469 (= fols. 220a–​220b); and Gulbadan Begim, History of Humayun, 10 (Persian), 91 (English). Although both Thackston and Beveridge translate Babur’s acquisition of Afghani Begim as a “marriage,” it is clear from Babur and Gulbadan Begim’s wording that she became his concubine and not his wife. Also see Parodi, “Of Shaykhs, Bibis and Begims,” 121–​138, 135. 12. Woods, Aqquyunlu, 11–​23. 13. Davani, Akhlaq-​i Jalali, 188. 14. Babur, Baburnama, 2:352–​353 (=fol. 169b). 15. Bika-​Sultan Begim was the daughter of the Timurid prince Sanjar Mirza (d. 1459). Sultan-​ Husayn’s marriage to Bika-​Sultan Begim was an unhappy one. According to both Babur and Khvandamir, she was an ill-​tempered woman, prone to fits of anger and jealousy, eventually resulting in their divorce. See Khvandamir, Habib al-​Siyar, 4:115–​116, and Habibu’s-​siyar, 2:412–​ 413; Babur, Baburnama, 2:350–​351 (=fol. 168b); and Subtelny, Timurids in Transition, 52–​53. 16.  Sultan-​Husayn was married previously to Payanda-​Sultan Begim’s sister Shahr-​Banu Begim. The wedding celebration, which took place not long after his first conquest of Herat in 1469, must have been a great public spectacle, for Khvandamir records a large celebration with the city’s notables in attendance and a generous distribution of gifts. See Khvandamir, Habib al-​Siyar, 4:136, and Habibu’s-​siyar, 2:421. Sultan-​Husayn’s decision to marry the daughter of his predecessor, Sultan-​Abu Saʿid, was consistent with the established pattern of Timurid princes securing marital connections with the family of the rulers they displaced. See Manz, “Women in Timurid Dynastic Politics,” 123–​124. In this case, the wedding appears to have been a piece of political theater meant to emphasize the continuity of rule between the well-​respected Sultan-​Abu Saʿid and newly enthroned Sultan-​Husayn. See Subtelny, Timurids in Transition, 61. The political significance of the marriage would explain why he was so quick to marry her sister after he divorced

202 i

Usman Hamid

Shahr-​Banu Begim for her support of Sultan-​Husayn’s opponent, her brother Sultan-​Mahmud (d. 1495), at the Battle of Chakman in 1471. See Babur, Baburnama, 2:350–​351 (=fol. 169a). Although this edition is missing Thackston’s translation of the section on Shahr-​Banu Begim, it is included in Babur, The Babur-​nama, 211. 17. Khvandamir, Habib al-​Siyar, 4:320–​321, and Habibu’s-​siyar, 2:511. 18. For example, Gulbadan Begim noted assiduously the order in which female guests were seated at banquets, including the wedding feast of Babur’s son Hindal. See Begim, History of Humayun, 31–​33 (Persian), 118–​123 (English). Also, see Soucek, “Timurid Women,” 200. 19. Manz, “Women in Timurid Dynastic Politics,” 131. 20. Muʿizz al-​Ansab, fols. 159b–​160a. 21. Subtelny, Timurids in Transition, 34. 22. Khvandamir, Habib al-​Siyar, 4:320–​21, and Habibu’s-​siyar, 2:511. 23. Khvandamir, Habib al-​Siyar, 4:160–​161, and Habibu’s-​siyar, 2:433. 24. Subtelny, “Centralizing Reform and Its Opponents in the Late Timurid Period,” 134. 25. Babur, Baburnama, 2:350–​351 (= fols. 169a–​169b). 26. Babur, Baburnama, 2:350–​351 (=fol. 169a). Although Thackston translates aldi as married, the text simply implies he “took” her. This reading is confirmed by the late sixteenth-​century translation of the text into Persian by Abd al-​Rahman Khan-​i Khanan, included by Thackston in his edition, where it is rendered as girifta (taken). See Babur, Baburnama, 2:350. For a similar translation, see Babur, The Babur-​nama: Memoirs of Babur, trans. Annette Susannah Beveridge (London: Luzac, 1912), 1:268. Sultan-​Husayn would eventually marry Khadija Begi Agha after the birth of their first son, Shah-​Gharib. 27. Muʿizz al-​Ansab, fol. 159b. 28. Babur, Baburnama, 2:350–​351 (=fol. 169a). 29. She was not, as has been suggested elsewhere, a slave. See Balabanlilar, Imperial Identity in the Mughal Empire, 24. 30. It is unlikely that she was the daughter of Muhammad Sariq, the agent of the Timurid prince Iskandar b. ʿUmar Shaykh (1384–​1415), given the fact that he was killed in 1414 and Khadija Begi Agha was still of childbearing age at least as late as the early 1470s. On Muhammad Sariq, see Manz, “Local Histories of Southern Iran,” 277–​278. 31. Babur, Baburnama, 2:350–​351 (=fols. 169a–​169b). 32. Muʿizz al-​Ansab, fol. 155a. 33. Babur, Baburnama, 1:55–​57 (=fols. 27b–​28b). 34. Muʿizz al-​Ansab, fols. 157a–​158a. 35. The list of Sultan-​Abu Saʿid’s daughters is given in Muʿizz al-​Ansab, fols. 156a–​156b. 36. Begim, History of Humayun, 14, 23, 31 (Persian), 97, 107, 118 (English translation). 37. Khvandamir does not mention Shah-​Gharib’s date of birth, but it coincided with Sultan-​ Husayn’s arrival at Akhta Akhur, which took place sometime between March 23, 1471 (his first Īd al-​Fiṭr celebrations after reclaiming Herat), and his return to the city for the winter later that year. His comment that Shah-​Gharib was born of “that lofty cradle, Khadija Begi Agha, who, in accordance with the glorious Shariʿa had attained felicity by way of marriage (nikah) to the Khaqan,” is somewhat misleading given Babur’s previously discussed comment that only three of Sultan-​ Husayn’s sons were legitimate. See Khvandamir, Habib al-​Siyar, 4:154, 158–​159, and Habibu’s-​ siyar, 2:430, 432. That Khadija Begi Agha was elevated after Shah-​Gharib’s birth is also Annette Beveridge’s understanding, for which see The Babur-​nama, 1:268, note 4. Muzaffar-​Husayn’s date



Slaves in Name Only

j 203

of birth can only be approximated. According to Khvandamir, when Muzaffar-​Husayn reached the age of maturity, he was married to his cousin Khanzada Khanim, the daughter of Sultan-​Husayn’s sister Badiʿ al-​Jamal Begim. Shortly before the wedding, which took place at the beginning of 1487, Muzaffar-​Husayn had been the recipient of a lavish two-​month celebration, with considerable feasting, drinking, and partying to mark his circumcision. See Khvandamir, Habib al-​Siyar, 4:178–​ 179, and Habibu’s-​siyar, 2:442–​443. It is unclear how old Muzaffar-​Husayn may have been when he got married or reached the age of maturity, but it is possible to speculate based on the little that is known about princely circumcisions in the late Timurid milieu. One of the best-​documented ceremonies to mark a royal circumcision was the celebration held by Sultan-​Abu Saʿid for his sons’ rite of passage. Held over a period of two months at the Bagh-​i Zaghan (from May 17–​July 19, 1466), it is described by Khvandamir as having been exceptionally spectacular and extravagant. See Khvandamir, Habib al-​Siyar, 4:83–​84, and Habibu’s-​siyar, 2:396–​397. The ages of Sultan-​Abu Saʿid’s sons ranged between 10 years and 15 years at the time. For their years of birth, see Babur, Baburnama, 35 (=fol. 18a), 1:50–​51 (=fol. 25b), and 1:12–​13 (=fol. 6b). This age range seems consistent with what has been approximated by Beatrice Forbes Manz for the age at which the Timurid prince ʿAbd al-​Latif (d. 1450), the patricidal son of Ulugh Beg (d. 1449), had his circumcision feast—​that is, between 8 years and 12 years. See Manz, Power, Politics, and Religion, 248, note 16. It is likely that Muzaffar-​Husayn’s circumcision and wedding took place while he was between the ages of 10 years and 15 years, thereby making his date of birth sometime between 1472 at the earliest (a year after his elder brother Shah-​Gharib was born) and 1477. It can be reasonably deduced, then, that Khadija Begi Agha was married legally to Sultan-​Husayn sometime during the mid 1470s. 38.  For instance, Latifa-​Sultan Aghacha and Papa Aghacha. See, for the latter, Babur, Baburnama, 2:352–​353 (=fol. 169b). 39. According to the Muʿizz al-​Ansab, she had a total of five children by Sultan-​Husayn, three of whom died in infancy. These included two sons, Sultan-​Jahangir and Jahangir-​Husayn, and a daughter whose name is simply given as Khanim. See Muʿizz al-​Ansab, fol. 163b. 40. Babur, Baburnama, 2:343 (=fol. 166a). 41.  Babur, Baburnama, 2:343 (=fol. 166a); Khvandamir, Habib al-​Siyar, 4:207, and Habibu’s-​siyar, 2:456. 42.  Khadija Begi Agha’s role in the political crisis that unfolded during the final decade of Sultan-​Husayn’s rule was first pointed out byBarthold, Four Studies on the History of Central Asia, 65. 43. Vasifi, Badayiʿ al-​Vaqayiʿ, 2:333. 44. Babur, Baburnama, 1:33 (=fol. 41a). 45. Khvandamir, Habib al-​Siyar, 4:207, and Habibu’s-​siyar, 2:456–​457. 46. Babur, Baburnama, 2: 344–​345 (=fol. 166a). 47. For the entire episode, see Khvandamir, Habib al-​Siyar, 4:208, and Habibu’s-​siyar, 2:457. 48. To appreciate fully the changing dynamics of ʿAli Shir Navaʾi and Khvaja Nizam al-​Mulk’s relationship see Subtelny, Timurids in Transition, 82–​101, but especially 93, 96, 100–​101. 49. Khvandamir, Habib al-​Siyar, 4:213–​214, and Habibu’s-​siyar, 2:459–​460. 50. Khvandamir, Habib al-​Siyar, 4:214, and Habibu’s-​siyar, 2:460. 51.  Although no other source mentions her, Khvandamir records a certain Bibi Muhibb as a patron of public buildings in Herat during the reign of Sultan-​Husayn. However, there is no evidence to suggest that these two were the same. See Subtelny, Timurids in Transition, 157; and Khvandamir, Maʾasir al-​Muluk, 178.

204 i

Usman Hamid

52. Vasifi, Badayiʿ al-​Vaqayiʿ, 2:334. Those who rebeled included Latifa-​Sultan Aghacha’s sons, Abu al-​Muhsin Mirza, then governor of Marv, and Muhammad-​Muhsin Mirza. See Khvandamir, Habib al-​Siyar, 4:239, and Habibu’s-​siyar, 2:472. Others, however, remained loyal, such as Payanda-​Sultan Begim’s son, Hayder-​Muhammad Mirza, and Papa Aghacha’s son, Ibn-​i Husayn Mirza. See Khvandamir, Habib al-​Siyar, 4:245, and Habibu’s-​siyar, 2:474. 53. Subtelny, Timurids in Transition, 100–​101. 54. Vasifi, Badayiʿ al-​Vaqayiʿ, 2:337. It is true that Vasifi’s memoirs, which were completed at the Uzbek court at Tashkent in 1538 or 1539, need to be read judiciously for their account of Timurid intrigues, for although he purports to have had intimate knowledge of high politics, there remain doubts about his veracity. See Subtelny, “Scenes from the Literary Life of Timurid Herat,” 140; and Pistoso, “Taste for Ambiguity,” 167. Nevertheless, as Subtelny alludes to in her article, although the historicity of his reports may be questioned, they reflect more accurately the perceptions of a contemporary litterateur. 55. Khvandamir, Habib al-​Siyar, 4:301, and Habibu’s-​siyar, 2:502. 56. Khvandamir, Habib al-​Siyar, 4:309, and Habibu’s-​siyar, 2:506. 57. Khvandamir, Habib al-​Siyar, 4:364, Habibu’s-​siyar, 2:532; and Babur, Baburnama, 2:381–​ 383 (=fols. 182b–​183a). 58. Babur, Baburnama, 2:382 (=fol. 183a). 59. Dughlat, Tarikh-​i-​R ashidi, 1:168–​169, 2:133. 60. Dughlat, Tarikh-​i-​R ashidi, 1:171, 2:135. 61. Dale, Garden of Eight Paradises, 63–​64. 62. Babur, Baburnama, 1:161–​163 (=fols. 79b–​80b); Khvandamir, Habib al-​Siyar, 4:277–​281, and Habibu’s-​siyar, 491–​492; and Salih, Shibani-​nama, 56–​57. I am grateful to Eliza Tasbihi for her reading of Muhammad Salih’s Chaghatay Turkish. However, Mulla Binaʾi makes no mention of Zuhra Begi Agha’s role in his section on the conquest of Samarqand. See ʿAli Binaʾi, “Shibani-​nama,” 66–​77. 63. Babur, Baburnama, 1:163 (=fol. 80b). 64. Ali, Marriage and Slavery in Early Islam, 8.

Bibliography Primary Sources Babur, Zahir al-​Din Muhammad. Baburnama. 3 parts. Turkish transcription, Persian edition, and English translation by W. M. Thackston, Jr. Sources of Oriental Languages and Literatures 18. Cambridge, MA: Department of Near Eastern Languages and Civilizations, Harvard University, 1993. —​—​—​. The Babur-​nama: Memoirs of Babur. 2 vols. Translated by Annette Susannah Beveridge. London: Luzac, 1912. Begim, Gulbadan. The History of Humayun:  Humayun-​ nama. Edited and translated by Annette S. Beveridge. Oriental Translation Fund, n.s., 1. London: Royal Asiatic Society, 1902. Binaʾi, Kamal al-​Din Ali. “Shibani-​nama.” In A Synthetical Study on Central Asian Culture in the Turco-​Islamic Period, edited by Kazuyuki Kubo, 1–​93 (Persian), 61–​67 ( Japanese and English). Research report, Ministry of Education, Science, Sports, and Culture, Japan; project no. 6301043. Kyoto: Kyoto University Press, 1997.



Slaves in Name Only

j 205

Davani, Jalal al-​Din. Akhlaq-​i Jalali. Edited by Abd Allah Masudi Arani. Tehran: Instisharat-​i Ittilaat, 2012 [1391]. Haydar Dughlat, Muhammad. Tarikh-​i-​R ashidi: A History of the Khans of Moghulistan. 2 vols. Edited and translated by W. M. Thackston. Sources of Oriental Languages and Literatures 37–​38. Cambridge, MA:  Harvard University, Department of Near Eastern Languages and Civilizations, 1996. Khvandamir, Ghiyas al-​Din ibn. Humam al-​Din al-​Husayni. Habibu’s-​siyar: Tome Three. 2 parts. Translated by W. M. Thackston. Sources of Oriental Languages and Literatures 24. Cambridge, MA: Department of Near Eastern Languages and Civilizations, Harvard University, 1994. —​—​—​. Tarikh-​i Habib al-​Siyar fi Akhbar-​i Afrad-​i Bashar. 4 vols. Edited by Jalal al-​Din Humaʾi. Tehran: Kitabkhana-​i Khayyam, 1954 [1333]. —​—​—​. Maasir al-​Muluk bi-​zamima-​i Khatima-​i Khulasat al-​Akhbar va Qanun-​i Humayuni. Edited by Mir Hashim Muhaddis. Tehran: Rasa, 1994 [1372]. Muʿizz al-​Ansab fi Shajarat al-​Ansab. MS, Bibliothèque nationale de France, Ancien fonds persan 67. Facsimile of the manuscript published in Sh. Kh. Vokhidov, edited and transcribed. Istoriia Kazakhstana v persidskikh istochnikakh. T. 3, Muʿizz al-​ansab (Proslavliaiushchee genealogii). Almaty: Daik-​Press, 2006. Salih, Muhammad. Shibani-​nama: Die Scheïbaniade; ein özbegisches Heldengedicht in 76 Gesängen. Edited and translated by Hermann Vambéry. Vienna: K.K. Hof-​und Staatsdrukerei, 1885. Vasifi, Zayn al-​ Din Mahmud. Badayiʿ al-​Vaqayiʿ. 2 vols. Edited by Aleksandr Boldyrev. Tehran: Bunyad-​i Farhang-​i Iran, 1970–​1972 [1349–​1350].

Secondary Sources Ali, Kecia. Marriage and Slavery in Early Islam. Cambridge, MA: Harvard University Press, 2010. Balabanlilar, Lisa. Imperial Identity in the Mughal Empire:  Memory and Dynastic Politics in Early Modern South and Central Asia. Vol. 1, Library of South Asian History and Culture. London: I.B. Tauris, 2012. Barthold, V. V. Four Studies on the History of Central Asia. Vol. 3, Mir ʿAli-​Shir and A History of the Turkman People. Translated by V. Minorsky and T. Minorsky. Leiden: Brill, 1962. Dale, Stephen F. The Garden of Eight Paradises: Babur and the Culture of Empire in Central Asia, Afghanistan and India (1843–​1530). Brill’s Inner Asian Library, vol. 10. Leiden: Brill, 2004. Manz, Beatrice Forbes. “Gowhar-​Šād Āga.” Encyclopædia Iranica, XI/​2, 180–​181. —​—​—​. “Local Histories of Southern Iran.” In History and Historiography of Post-​Mongol Central Asia and the Middle East: Studies in Honor of John E. Woods, edited by Judith Pfeiffer and Sholeh A. Quinn, in collaboration with Ernest Tucker, 267–​281. Wiesbaden:  Harrassowitz Verlag, 2006. —​—​—​. Power, Politics and Religion in Timurid Iran. Cambridge Studies in Islamic Civilization. Cambridge: Cambridge University Press, 2007. —​—​—​.“ Women in Timurid Dynastic Politics.” In Women in Iran from the Rise of Islam to 1800, edited by Guity Nashat and Lois Beck, 121–​139. Urbana: University of Illinois Press, 2003. Parodi, Laura E. “Of Shaykhs, Bibis and Begims: Sources on Early Mughal Marriage Connections and the Patronage of Babur’s Tomb.” In Medieval and Modern Iranian Studies: Proceedings of the 6th European Conference of Iranian Studies, Held in Vienna on 18–​22 September 2007 by

206 i

Usman Hamid

the Societas Iranologica Europaea, edited by Maria Szuppe, Anna Krasnowolska, and Claus V. Pedersen, 121–​138. Cahiers de Studia Iranica 45. Paris: Association pour l’Avancement des Études Iraniennes, 2011. Pistoso, Maurizio. “A Taste for Ambiguity. Reconsidering Mahmud Vasefi’s Memoirs.” In La Civiltà Timuride Come Fenomeno Internazionale, Volume 1 (Storia—​I Timuridi e. l’Occidente), edited by Michele Bernardini. In Oriente Moderno n.s., 15, no. 2 (1996): 165–​172. Soucek, Priscilla P. “Timurid Women: A Cultural Perspective.” In Women in the Medieval Islamic World: Power, Patronage, and Piety, edited by Gavin R. G. Hambly, 199–​226. New York: St. Martin’s, 1999. Subtelny, Maria Eva. “Babur’s Rival Relations:  A  Story of Kinship and Conflict in 15th–​16th Century Central Asia.” Der Islam 66, no. 1 (1989): 102–​118. —​—​—​. “Centralizing Reform and Its Opponents in the Late Timurid Period.” Iranian Studies 21, nos. 1–​2 (1988): 123–​151. —​—​—​. “Scenes from the Literary Life of Timurid Herat.” In Logos Islamikos: Studia Islamica in Honorem Georgii Michaelis Wickens, edited by Roger M. Savory and Dionisius A. Agius, 137–​ 155. Toronto: Pontifical Institute of Medieval Studies, 1984. —​—​—​. Timurids in Transition: Turko-​Persian Politics and Acculturation in Medieval Iran. Brill’s Inner Asian Library, vol. 19. Leiden: Brill, 2007. Woods, John E. The Aqquyunlu: Clan, Confederation, Empire. Rev. ed. Salt Lake City: University of Utah Press, 1999. —​—​—​. The Timurid Dynasty. Papers on Central Asia, no. 14. Bloomington: Research Institute for Inner Asian Studies, Indiana University, 1990. —​—​—​. “Timur’s Genealogy.” In Intellectual Studies on Islam: Essays Written in Honor of Martin B.  Dickson, edited by Michel M. Mazzaoui and Vera B. Moreen, 85–​125. Salt Lake City: University of Utah Press, 1990.

10 A Queen Mother and the Ottoman Imperial Harem Rabia Gülnuş Emetullah Valide Sultan (1640–​1 715)

Betül İpşirli Argıt

i  On his deathbed, the renowned Ottoman statesman, the Grand Vizier Köprülü Mehmed Pasha (d. 1661) gave the following advice to the reigning sultan, Mehmed IV (r. 1648–​1687): “Never give ear to the counsel and advice of women, always keep a full treasury, and always keep the troops occupied.”1 These words reflected the mindset of a statesman who changed the balance of power in Ottoman politics. As vizier, Köprülü Mehmed Pasha assumed much of the power wielded previously by Ottoman sultans, ruling the empire without interference until his death. Ironically, although Köprülü Mehmed Pasha criticized women’s interference in the political sphere, his assumption of power in 1656 was made possible by the “counsel and advice” of a queen mother, Hatice Turhan Sultan (d. 1683), a notorious figure whose efforts against enemies in the harem (the royal household) allowed her to retain her influence and, thus, keep her nine-​year-​ old son safely on the throne. Ottomanist scholarship has generally held that Köprülü Mehmed Pasha’s ascendance put an end to the influence of the powerful valide sultans (queen mothers of the reigning sultans) and thus terminated an era of Ottoman history known as “the sultanate of women.”2 The period, one during which valide sultans wielded exceptional authority in social and political life, lasted roughly from the mid-​sixteenth to the mid-​seventeenth centuries.3 A closer look at late seventeenth-​and early eighteenth-​century sources, however, suggests this perception of lessening feminine power does not reflect the historical reality. This chapter challenges the perception of a waning of women’s influence during this period by examining the life and career of Rabia Gülnuş Emetullah, Valide Sultan (d. 1715). 207

208 i

Betül İpşirli Argıt

Gülnuş Sultan began her career as the slave concubine of Mehmed IV before rising to political power as the mother of two sultans, Mustafa II (r. 1695–​1703) and Ahmed III (r. 1703–​1715).4 Her example deserves careful consideration. Although clearly a figure of sweeping influence, she has been overshadowed largely by the preceding valide sultans, and her considerable political influence thus mostly ignored in modern scholarship. A close study of her career, drawing on contemporary historical sources, chiefly European accounts alongside Ottoman chronicles and archival documents, allows us to situate her properly in the history of the Ottoman court. More specifically, it shows how the former slave girl, making use of the opportunity to rise in the social and political hierarchy, wielded power even after the so-​called “the sultanate of women” had ended. Thus, as the career of Gülnuş Sultan makes clear, the influence of the queen mother continued after the period of the Köprülü family. In no sense were such women passive politically. The following discussion first considers Gülnuş Sultan’s origins and career before becoming valide sultan, before turning to her political career as the queen mother during the reign of Mustafa II and her continued influence in the court of Ahmed III. It turns, finally, to an assessment of her part in the major political developments of the day, and especially her activity in the so-​called Great Northern War (1700–​1721) between Russia and the Swedish Empire, a conflict into which the Ottoman Empire was quickly drawn. Köprülü Mehmed Pasha’s statement regarding the “interference” of women in politics should come as no surprise, given that—​from the end of the sixteenth century onward—​ the role of women in state affairs was criticized routinely and roundly by Islamic religious scholars (ulema) and statesmen.5 Such critics regarded women in positions of political influence as representing an unnatural state of affairs; thus, they marshaled a variety of sources to indicate the impropriety of a ruler deferring to the will of women.6 Despite the fact that the influence of royal women appears to have receded by degree after the period of “the sultanate of women,” their position not only made intervention in politics inevitable, but, in an important sense as well, they often brought a certain balance to Ottoman political life. The contribution of the valide sultans to the well-​being of the dynasty and that of the empire itself lasted for more than 600 years. In this sense, then, Köprülü Mehmed Pasha’s advice rang hollow. Origins and Early Career Gülnuş was born during the 1640s in Rethymno, Crete, the island, at that point under Venetian rule.7 Of Greek origin, her given name was Eugénie.8 The Ottoman army had seized the island during the Crete War (1645–​1669); the girl was enslaved and sent to the palace either as a gift or plunder after the conquest of the island in 1646 by Deli Hüseyin Pasha.9 No information is available concerning when she became the haseki (favorite concubine) of Mehmed IV. It is clear, however, that Gülnuş Sultan served for a comparatively long and privileged period as his consort, bearing him two sons in the process: Shahzade Mustafa and his brother Ahmed. She no doubt learned to negotiate



A Queen Mother and the Ottoman Imperial Harem

j 209

harem politics during this period, observing how best to influence the greater political scene beyond the imperial household that, at this point, was subject to the “reign” of the valide sultan, the Slavic Ruthenian slave Hatice Turhan Sultan. The latter is remembered, in part, for having participated in the murder of her mother-​in-​law, Kösem Valide Sultan; she is said to have joined the conspiracy to protect her sons. As the sultan’s favorite concubine, and with the survival of her own sons as a paramount concern, Gülnuş developed the skills she would later need as valide sultan in her own right. Thus, her political career commenced during the haseki period. The lessons she derived during these early years ensured that, in the long term, she would have a better command of court life. So, for example, her participation in the imperial campaign procession (sefer alayi), an event of great pomp and circumstance in which all members of the bureaucracy took part, offered her the opportunity to witness royal political culture up close—​especially, in this case, royal ceremonials and protocols. She accompanied Mehmed IV, Hatice Turhan Sultan, Shahzade Mustafa, and the sisters of Mehmed IV, together with a large entourage, in processions marking the Polish War in 1672 and 1673.10 She later joined a similarly large entourage in a procession marking the siege of Vienna in 1683.11 Gülnuş Sultan, during this first period, became involved in politics through a variety of means. She devoted much effort to protecting the position of Mehmed IV, and her sons, Mustafa and Ahmed. The presence of two sons of Sultan İbrahim (r. 1640–​1648) by other concubines, Süleyman and Ahmed, made it a difficult task. During the reign of Mehmed IV, members of the military and ulema threw their support to Süleyman, Mehmed IV’s half-​brother, making several attempts to bring him to power.12 It was in this context that Gülnuş Sultan took up the challenge of countering all threats to her consort, Sultan Mehmed IV. At one point, she resorted to giving money from her own fortune to calm troops that had risen up against the monarch.13 Efforts on her part to secure the succession to the throne of her sons brought Gülnuş Sultan into conflict with Hatice Turhan Sultan; as suggested earlier, the latter sought to protect Süleyman and Ahmed from Gülnuş Sultan’s considerable ambitions regarding her own sons.14 Gülnuş Sultan also established networks of support within the imperial court—​one of several routes to influence available to women of the royal household. Given the importance of the chief eunuch in the seraglio, ensuring a close relationship with this official was essential.15 The chief harem eunuch at the time, Yusuf Agha, was one of the most significant people to whom Gülnuş Sultan was allied within the imperial harem.16 Their relationship appears to have dated to a relatively early point. Yusuf ’s long-​standing importance in the royal household was evident already by 1675, when Shahzade Ahmed was circumcised; the boy is reported as having sat in his lap during the ceremony.17 Yusuf Agha, in addition, was the administrator of the waqf (pious foundation) that Gülnuş Sultan founded in 1680 and that provided income for a hospital and public kitchens in Mecca.18 The responsibility of representing the queen mother in public was one of several factors that contributed to their relationship. A further sign of Gülnuş Sultan’s close

210 i

Betül İpşirli Argıt

ties to Yusuf Agha is that her chamberlain, Mehter Osman Agha, was his apprentice and protégé.19 Gülnuş also enjoyed close relations with Feyzullah Efendi, who served as tutor to her son Mustafa. From 1669 onward, with the help of Vani Efendi, at that point the sultan’s personal religious advisor, the career of Feyzullah Efendi advanced rapidly. Even after the disastrous collapse of the siege on Vienna in 1683, as a result of which his influence at court fell sharply, his close relations with Gülnuş Sultan remained. Thus, in one incident dating to 1686, when Feyzullah Efendi let his horse graze in the Royal Garden (Has Bahçe), a royal property, it was decided he had to be punished. Although his record was erased from the ulema register, Gülnuş Sultan intervened to save him and he was subsequently assigned a new post.20 Gülnuş Sultan and Feyzullah Efendi would encounter one another at a later point as well, together with Amcazade Hüseyin, Köprülü Mehmed Pasha’s nephew, with whom she had a close acquaintance too. Having met with Mehmed IV and Gülnuş—​still, at this point, in her role as haseki—​on the way to the Polish War in 1672, Hüseyin would later join her household, serving therein for an extended period and finally becoming her konakçı (chief billeting officer) in 1682.21 Gülnuş Sultan also played a role in determining the careers of various statesmen, including the unfortunate Grand Vizier Kara Mustafa Pasha (1676–​1683). Charged with the failure of the Vienna campaign in 1683, he was first brought before an unhappy Mehmed IV; then, after a considerable lobbying effort on the part of Gülnuş Sultan and the court eunuchs, he was stripped of his office and executed.22 She continued to play an active part in the rise and downfall of several viziers thereafter. Von Hammer-​Purgstal sees Gülnuş Sultan’s influence as having lessened during the vizierate of Fazil Ahmed Pasha (1661–​ 1676), then having risen again after Kara Mustafa Pasha’s fall from grace—​that is, during the tenures of Kara İbrahim Pasha (1683–​1685) and Süleyman Pasha (1685–​1687).23 The defeat at Vienna had fueled unrest that culminated in the deposition of Mehmed IV in November 1687. An attempt to crown Mustafa, Gülnuş Sultan’s eldest son, failed with the accession to power of Mehmed IV’s half-​brother, Shahzade Süleyman (r. 1687–​ 1691).24 Süleyman proved incompetent, with the affairs of state falling largely to those who had overseen the dethronement of Mehmed IV. The same influential bloc then arranged the succession of Shahzade Ahmed, Süleyman’s brother, in 1691, this despite renewed efforts on the part of Mustafa’s supporters. It was only with the collapse of Ahmed II’s reign (r. 1691–​1695) in February 1695, that Mustafa finally rose to office, thus making Gülnuş Sultan the new valide sultan.25 Gülnuş remained, during this eight-​year period (1687–​1695), in the Old Palace—​the traditional residence of women of nonreigning members of the royal house.26 During her haseki period, she had cut her teeth on court politics, gaining particular experience after the death of Turhan Sultan in 1683. Von Hammer-​Purgstal writes that, from that point onward, Gülnuş Sultan ruled the harem with absolute authority, influencing the fate of the empire.27 Now, as the sultan’s mother, she would exercise only greater influence still.



A Queen Mother and the Ottoman Imperial Harem

j 211

Queen Mother: The Reign of Mustafa II Gülnuş Sultan’s political career would last another two decades until her death during the reign of her second son, Ahmed III. The fact that Mustafa II was 31 years old when he ascended the throne did not prevent his mother from exercising influence over him. She took on a decisive role, for example, in the selection of her son’s staff. Members of the dynasty whom she had protected while still a haseki, alongside several officials with whom she had long been in contact, were among the new appointees. Considering the close relationship of previous valide sultans with the chief harem eunuchs, the selection of the chief eunuch was particularly sensitive. The new valide sultan saw to the transfer of Ali Agha (or, as he was known, Yapraksiz) from Egypt and his appointment as the chief eunuch of the Imperial Harem.28 Feyzullah Efendi had served as şeyhülislȃm during the era of Süleyman II, but had been dismissed for his ties to Mehmed IV and his family. He was now reappointed to his former office.29 The valide sultan developed close ties as well to Amcazade Hüseyin, yet another significant figure in Mustafa II’s reign. On Feyzullah Efendi’s suggestion, Gülnuş worked to advance his career, seeing to his appointment as grand vizier in 1697. When accusations arose of his involvement in plots against the sultan, she effectively saved his life by making entreaties on his behalf in 1702.30 Correspondence—​still extant in the Ottoman archives—​between the valide sultan and her steward (kethüda), Mehmed Efendi, probably the most influential member of her household, reveals that Gülnuş obtained considerable information from him regarding the performance of a variety of officials.31 She also followed military developments closely, about which her steward kept her up-​to-​date as well.32 Thus, news of the campaigns in Hungary in 1696 was forwarded to her.33 Rycaut wrote that, in 1697, her son, the sultan, sent dispatches to Gülnuş about developments on the battlefield and a promise to send her gifts of young slave girls captured in Transylvania.34 At the end of the war, the sultan then dispatched a black eunuch to inform the valide sultan of the unhappy outcome of the fighting and dispel rumors concerning his death.35 She also corresponded with statesmen during this period. In an encouraging letter to the Crimean Khan (prince), she spoke of the conflict as a time of strife and sacrifice, asking that they work together.36 The sense of crisis and loss of territory that resulted from the War of Zenta deeply saddened the valide sultan. Zinkeisen writes that she sought to pressure Mustafa II to continue the war, seeing this as the most reliable means to protect her own influence.37 After the losses at Zenta in 1697, the internal balance of power in the palace shifted, with a particular rise in authority of the şeyhülislȃm, Feyzullah Efendi. Thus, an evaluation of the valide sultan’s place in the palace during the second half of Mustafa II’s reign is in order. Sir Robert Sutton, the British council from 1701 to 1716, speaks of Gülnuş’s increasing access to Mustafa II. Noting that Feyzullah Efendi also enjoyed a significant influence on the sultan, Sutton comments that Feyzullah Efendi found the secret to managing the valide sultan herself; the two joined forces and she now only took heed of Feyzullah

212 i

Betül İpşirli Argıt

Efendi’s interests.38 Rycaut, for his part, states that the valide sultan was a “woman of intrigues” and that she exerted great power over her son; Mustafa II fell completely under her control, surrendering himself entirely to her recommendations.39 In a similar vein, the Italian envoy Carlo Ruzzini, who represented Venice during the Karlowitz Treaty in 1699, wrote that action was taken according to the advice received from the sultan’s mother and Feyzullah Efendi.40 Their close relations were not to last. In 1703, rioters in Istanbul demanded that the şeyhülislȃm be dismissed.41 On realizing that the people of Istanbul did not intend to back down and were indeed preparing to march on Edirne, the site of the imperial court, Gülnuş intervened to persuade the sultan to deliver Feyzullah Efendi to the rioters. When he rejected her request, the valide sultan said, “Should you create problems among the Muslims for just one person, then neither you nor I will remain.” She continued to insist that Feyzullah Efendi be removed, by which point, on receiving an edict to this end, she saw to his dismissal. 42 It appears, then, that the valide sultan’s handling of the crisis, and her influence on Mustafa II’s decision-​making, was a determining factor in shaping the course of events. Despite her intervention, however, the rebellion continued; the rioters are reported to have called for the dismissal of not only the şeyhülislȃm, but of the sultan and valide sultan as well.43 It was widely held that the valide sultan was in league with Feyzullah Efendi, and, through her efforts, Gülnuş was working to keep the dethroned sultan in Edirne.44 Elias Habesci, the secretary of the grand vizier during the reign of Mustafa III, wrote that if the sultan were to return to Istanbul, his presence would have stopped the progress of the rebellion. The efforts of the valide sultan and şeyhülislȃm, whose influence depended on keeping the sultan secluded in the harem, prevented this from taking place.45 After the deposition of Mustafa II, efforts by Gülnuş in determining his successor proved critical. Rioters refused her second son, Shahzade Ahmed, on the grounds that he was Mustafa’s brother and son of the same mother. They instead demanded a cousin with equal rights to the throne:  Ahmed II’s son, İbrahim.46 Gülnuş came out against Shahzade İbrahim and in favor of her son Ahmed, whose future had been much on her mind throughout the reign of his brother, Mustafa II.47 If Dimitrie Cantemir has it right, the valide sultan had long been aware of the possibility that her son, Ahmed, could become sultan, and had worked through Mustafa’s reign to prepare him for this eventuality.48 In the end, the ulema, including Mehmed Efendi, who had been declared the new şeyhülislȃm by the rebels, decided that Ahmed was more suitable than the 11-​year-​old İbrahim. After the revolt, and in keeping with long-​held practice, she was informed by Mehmed Efendi that the rioters had selected Shahzade Ahmed to succeed to the throne and, in a formal letter, asked for her approval.49 She confirmed the enthronement of her younger son following her older son’s abdication.50 Problems remained, however, even after Ahmed III’s accession. Pointing to her continued influence, the rioters insisted on Gülnuş’s removal; she was to “go straight to the Old Palace and not be permitted into the New Palace (the residence of the reigning sultan).”51



A Queen Mother and the Ottoman Imperial Harem

j 213

Aubry de la Motraye reports that after Mustafa II was stripped of office, the valide sultan was removed to the Old Palace until things calmed down—​a development that seems to have surprised him, given that she was mother to the new sultan. He explains that she had supported the former şeyhülislȃm’s decision to keep the dethroned sultan in Edirne.52 In the end, Gülnuş, under the protection of the second vizier, Morali Hasan Pasha, was accompanied to the New Palace with a royal procession.53 The Court of Ahmed III (r. 1703–​1736) Gülnuş’s intervention during the reign of Mustafa II reveals the extent to which authority remained with the imperial harem. She took up her duties immediately as mother of the new sultan, paying, for example, an enthronement gratuity from her own funds—​a tradition used to retain the loyalty of the military and palace staff.54 Foreign observers comment on her continued power during this period (in which there was no effective figure like Feyzullah Efendi). So, for example, in 1706, the Italian envoy Carlo Ruzzini wrote that, in keeping with the wishes of the valide sultan’s son, she intervened in administrative matters less than she had in the reign of her other son, although without loss to her vaunted position at court.55 Voltaire notes that Gülnuş, once ill used by her son, had gained new influence in the imperial court.56 Baron Fabricius, of the Holstein Palace and envoy in charge of relations with the King of Sweden, Charles XII, wrote that the valide sultan gradually gained greater authority.57 Fabricius adds, more generally, that Ahmed III was strongly influenced by women and statesmen close to him—​all individuals recommended to him by women.58 Indeed, Ahmed III’s court included a number of individuals with whom the valide sultan had been in contact for years and whose careers had benefited from her support. Such was the case of Kalaylikoz Ahmed Pasha, a close associate from Gülnuş’s haseki period. In hiding for three years because of an incident in which he was involved in 1699, he was forgiven by the valide sultan.59 At a later point, intervention on her behalf would lead to his appointment as governor of Crete.60 Finally, in 1704, he became the grand vizier, which lasted three months. Ahmed Pasha, in effect, owed Gülnuş Sultan his career. Her difficult relationship with Grand Vizier, Çorlulu Ali Pasha (1706–​1710), deserves special mention. The valide sultan, working alongside Silahdar Ali Agha (discussed later) and the chief harem eunuch, Uzun Süleyman Agha, prepared the way for Grand Vizier Çorlulu Ali Pasha to be dismissed in 1710. She turned against him on learning that he sought to bring down her son, Ahmed III, and replace him with Shahzade İbrahim.61 Her influence in the dismissal of Çorlulu Ali Pasha is evidence of her continuing engagement with the political factions at court. Gülnuş was also influential in the career of Baltacı Mehmed Pasha, who twice held the office of grand vizier (1704–​1706, 1710–​1711). Baltacı Mehmet Pasha’s wife was a former slave of the valide sultan.62 Ahmed III gave the Circassian girl, of whom he was very fond,

214 i

Betül İpşirli Argıt

to Baltacı Mehmed when he became sultan.63 Members of the dynastic family, including valide sultans, arranged marriages for their female slaves to members of the imperial court at various levels. These marriages had political implications in terms of establishing networks and alliances within and outside the court, and securing the loyalty of the dynasty’s servants. This was another form of securing “power from the harem”; Gülnuş Sultan expanded her political influence through the marriages of household members. In this case, the effect was to tie Baltacı Mehmed Pasha more firmly to the dynasty and, more important, to reinforce his relations with Gülnuş Sultan herself. Baltacı Mehmed Pasha’s lengthy letter to Gülnuş, dated 1711, contains important evidence regarding his relations with the valide sultan as well as her political position in the imperial court. 64 He refers to Gülnuş as his patron and benefactor and, in detailing his career, provides examples of his loyalty to the family of Mehmed IV. The letter, in which Baltacı Pasha appeals for help from the valide sultan, becomes clear in the political context of the period. After the war over Russia (discussed later), the Swedish king and his chief Crimean ally criticized Baltacı Mehmed Pasha for having failed to take proper advantage of the victory. As a result, Ahmed III decided to dismiss him. Mehmed Pasha turned immediately to the valide sultan in an effort to save his position, apparently believing that Gülnuş was aware of political currents at court and could influence the sultan’s decision. It is not known, however, whether the valide sultan tried to intervene on his behalf after receiving the letter. Despite his best efforts, Baltacı Mehmed Pasha was removed from his post November 20, 1711. Gülnuş also did much to advance the career of Silahdar Ali Agha, who had remained in office after the enthronement of Ahmed III. Mustafa II, the preceding sultan, had used the valide sultan to protect Silahdar Ali. The effect of her mediation was the latter’s appointment as grand vizier in 1713.65 These various examples make clear that Gülnuş played a decisive role in the court of Ahmed III. The “Great Northern War” In addition to using her influence on the careers of her male counterparts, it is also clear that Gülnuş followed international political developments closely. The case of Baltacı Mehmed Pasha leads us to examine her part in the Russo-​Swedish war. The conflict saw Karl XII of Sweden (d. 1718) fall to Peter I of Russia (d. 1725) in the Poltava clash of 1709, then seek refuge in Ottoman territory. Russian officials, regarding this as a violation of their treaty with the Ottoman state, demanded the king be returned. The Ottomans opted instead for confrontation, declaring war on Russia in 1711. After the defeat of the Russian army, the conflict ended with the signing of the Prut Treaty on July 21, 1711. The Swedish king remained in Ottoman territory for five years, between 1708 and 1714; the controversy thus remained high on the Ottoman agenda for years, with the Swedish monarch in close contact with the Ottoman court.66



A Queen Mother and the Ottoman Imperial Harem

j 215

Gülnuş Sultan played a role throughout, providing funding, establishing contact with a network of people in and outside the court, and engaging in frequent correspondence. Cantemir writes that, during the Moscow expedition, she recruited troops using personal funds.67 Developments surrounding the Prut campaign were reported in detail to the valide sultan by her steward (kethüda).68 She was also in contact with a number of Ottoman statesmen as events progressed. In a detailed letter dating to 1711, Baltacı Mehmet Pasha, then grand vizier, informed the valide sultan on the state of the war and its happy conclusion.69 The Pasha’s letter offers an extremely detailed report of the progress of the war, the negotiations that occurred between the various sides, and the actual fighting. All this speaks to how well the valide sultan herself followed events. Gülnuş, as Voltaire notes, was openly in favor of the war against the Russian czar, a position she shared with the chief harem eunuch and the agha of the Janissaries.70 She was involved in diplomatic traffic between Ottoman and Swedish officials, and corresponded with foreign political leaders and statesmen. In addition, the Crimean Khan wrote to inform her of the innermost workings of the diplomatic talks that took place between himself and the Swedish king (Karl XII) during spring 1710.71 Gülnuş’s letters to the King of Sweden underscore not only the extent of her political involvement but also the intriguing nature of her role during the war. She wrote two secret letters to the Swedish monarch; the letters are warm and sincere, and, at one point, she addresses him as “my strong and lofty son, my lion.”72 Throughout the conflict, she protected the interests of the Swedish king, often mediating on his behalf with the Ottoman court. She thus provided a vital, if informal, connection between the two monarchs and their courts. Her letters gain greater clarity against the backdrop of Ottoman–​Swedish relations after the Prut expedition. The Treaty of Prut (1711) did not resolve fully the tensions dividing the Ottoman–​Swedish and Russian sides. In early 1713, the Ottoman court put pressure on the Swedish king to return home, and even issued orders that he be returned by force if necessary. Tatar forces plundered the Swedish camp, surrounding the king in the process, and killing some of his retinue. The attacking troops also captured a number of Swedish soldiers, compelling some of them to convert to Islam, and, when Janissaries opened fire on the king’s residence, the monarch was injured on the ear.73 Given the damage to the reputation of the Ottoman court, it became important to mend relations with the Swedes. Letters written by Gülnuş make clear that she intervened at this point with the Ottoman court on Karl XII’s behalf. Again, Gülnuş used relations with a network of individuals to retain her influence. Valide sultans had used female palace slaves routinely in this manner, and Gülnuş was no exception. An unnamed female courtier is reported to have been deeply effective in the Ottoman–​Swedish talks; she and her husband are said to have enabled communication between the imperial court and the valide sultan. References in the first of the two secret letters indicate she spoke for Gülnuş and, thus, thanks to her, in large part, the valide sultan became aware of the king’s problems. In his letter to the French king, Louis XIV, Das

216 i

Betül İpşirli Argıt

Alleurs gives detailed information regarding the unnamed woman.74 He states she was someone close to the valide sultan in the palace, that the sultan had met her before coming to the throne, and that, at one point, she had been one of Ahmed III’s favorites and, as a result, despite having married someone else, she retained easy access to palace circles. This is probably the same individual to whom Cantemir refers as having once been in love with Shahzade Ahmed but who later married Baltacı Mehmed Pasha. The second of the secret letters reveals another person with whom Gülnuş dealt at this point. In his letter to the French king, Das Alleurs notes that she was on good terms with a certain General Poniatowski; the latter had been sent to Istanbul by the Swedish court to represent the king’s interests, although in an unofficial capacity.75 Poniatowski had an obvious hand in the king’s relations with the Ottoman court, activity in which his correspondence with the valide sultan played a key role.76 At one point, for example, Poniatoswki, who himself was helped by an anonymous woman at court, wrote a letter to Gülnuş in which he explained the terrible situation in which Karl XII found himself upon arriving in Edirne; he states this situation was not in keeping with the glory of the sultan.77 A third of her contacts, a Portuguese doctor named Fonseca, is reported to have assisted Poniatowski with his plans. Well placed in political circles, Fonseca was associated with the palace and got on well with the viziers. According to what Poniatowski told Voltaire, Fonseca was active in transmitting letters to the valide sultan. Voltaire also mentions a Jewish woman, perhaps a merchant, bringing merchandise for customers in the seraglio. Voltaire notes the valide sultan established relations with Poniatowski in part through this same, unnamed woman. She is reported, specifically, to have influenced Gülnuş by her constant mention of the qualities of the Swedish king.78 At the end of the crisis, as reported by Gülnuş herself, Karl XII returned safely to Sweden. Evidence regarding the entire episode makes clear the extent of Gülnuş Sultan’s role in foreign relations. The Swedish problem finally resolved, the valide sultan continued her activity in the political arena. Zinkeisen writes that, on the matter of the Morean campaign (1714), for example, she took part in efforts—​alongside the şeyhülislȃm—​to resolve the conflict with Venice peacefully.79 Works of Charity Beginning during her haseki period, Gülnuş funded a number of charitable projects, which enhanced the public-​oriented efforts of the imperial state. Much in the way of previous valide sultans, she supported a range of institutions, including mosques, fountains, libraries, soup kitchens, primary schools, and services for the twin holy cities of Mecca and Medina, and thus gained wide popular appreciation. Among her projects was a complex of buildings (külliye)—​mosque, soup kitchen, school, fountain, and tomb—​built in Üsküdar.80 Construction of the complex tied Gülnuş to a tradition established by previous female members of the dynasty who had



A Queen Mother and the Ottoman Imperial Harem

j 217

ordered the construction of charitable complexes in various parts of Istanbul. In addition, Gülnuş sponsored the transformation of a church in Galata into a mosque, and the building of five fountains with which clean water was finally made available to the area.81 She pursued similar efforts in regions outside Istanbul, including Edirne, Chios, Mecca, Medina, Kastamonu, and Menemen. After the reconquest of the island of Chios in 1695, a church was converted to a mosque in her name.82 Her undertaking resembles the example of Hatice Turhan Sultan; during Gülnuş’s haseki period, after the conquest of Candia, a church had been converted into a mosque in the name of Turhan Sultan.83 Gülnuş Sultan also constructed a fountain next to the mosque in Chios, responding to local needs, as stated in the endowment deed. 84 Thus, her charity works contributed to the architectural, social, urban, and religious development of the later Ottoman Empire. Conclusion A reconstruction of Gülnuş Sultan’s career, set against the backdrop of the political developments of the era, suggests that limiting the influence of the valide sultans on the workings of the imperial court and Ottoman internal and international politics to the era of “the sultanate of women” does injustice to historical reality. Shared patterns tied Gülnuş’s career to a tradition of queen mothers that had existed since antiquity, thus well before the Ottoman period itself. The women sought to promote their own interests and those of their sons, but also devoted themselves to the maintenance of order and the well-​ being of dynasty and empire. In the case of Gülnuş Sultan, much of her activity served the best interests of the Ottoman sultan and the state as a whole. Her advisory, mediatory, and negotiating activity at the imperial court, and her investments, both in the army and charitable works, contributed to the prestige of the dynasty as well as the public good. Her comment during the 1703 revolt—​“if fighting breaks out within the ummah of Muhammad, neither I  nor you will remain”—​reflects a recognition on her part of the relationship of public attitudes, the fortunes of the dynasty, and her own standing at court. The experience of queen mothers may have varied across time and space, but, as a close scrutiny of the participation of these women in political life suggests, compelling continuities across late-​medieval and premodern Turco-​Mongol and Islamic history persisted as well. The position of the queen mother, in other words, drew on deeply rooted institutionalized structures and well-​established social patterns. Thus, although Gülnuş Sultan’s career was dictated by social, cultural, and political circumstances unique to seventeenth-​and eighteenth-​century Ottoman life, it relied on much longer term practice as well.85Although the degree of influence of Ottoman women varied from period to period, by necessity and tradition, they played an active part in imperial governance. When evaluating the part played by women in Ottoman political life, it is crucial to pay as much attention as the evidence allows to individual personalities.86 Kösem Sultan

218 i

Betül İpşirli Argıt

ruled the empire as official regent during the reigns of her sons, Murad IV (r. 1623–​1640) and İbrahim I (r. 1640–​1648). She was contested by her grandson’s concubine mother, Hatice Turhan Valide Sultan. Kösem Sultan was attempting to replace the younger woman, an obvious rival, when Turhan Sultan arranged her assassination. Her example takes on greater significance when the stakes and outcome of the rivalry are taken into account. Intent on preserving her power, Kösem Sultan chose Dilaşub Sultan, the mother of İbrahim’s son Shahzade Süleyman, over Hatice Sultan. She expressed her opinion that the new mother would be more pliable: “Dilaşub is a credulous woman who cannot be passionate about filling the position of valide sultan properly.”87 Thus, it can be said that Gülnuş Sultan was not as ambitious a character as Kösem Sultan, yet she had the perspicacity to follow developments closely and intervene when necessary. She had an impact on the careers of statesmen, used her personal wealth for the benefit of the state, stepped in at critical moments, corresponded with foreign statesmen, and thus played a notable role in shaping imperial foreign policy. The power of women clearly did not end with the period of “the sultanate of women”; Gülnuş Sultan, as haseki and valide sultan, carried on a long tradition that provided a critical balance of power within the imperial court. The essential role played by women in the Ottoman dynastic household was also reflected in the organization and operation of households of lesser status. Foreign authors of the period write that Köprülü Mehmed Pasha’s wife ensured that her son, Ahmed Pasha, became grand vizier.88 Interestingly, although Köprülü Mehmed Pasha advised Mehmed IV not to listen to women, his own wife intervened in political affairs when it suited her, in an effort to use her influence for the benefit of her family. Acknowledgments This chapter is a revised version of a paper delivered at the Middle East Studies Association (MESA) meeting, Denver, Colorado, November 23, 2012. It has been much improved by the critical and constructive comments of Kathryn Hain, Matthew Gordon, Ali Akyildiz, Nukhet Varlik, and Nalan Turna. This chapter is part of a larger research project supported by TUBITAK. Notes 1. Rycaut, History of the Turkish Empire, 113. 2. The source of this assumption is mainly a result of the work of Refik, Kadınlar Saltanatı. 3.  This period is subject to many studies, but the most distinguished study on this issue is Peirce, Imperial Harem. 4. During the fourteenth and fifteenth centuries, Ottoman rulers and their sons contracted legal marriages. By the mid-​fifteenth century, legal marriages had lapsed and the preference for slave concubinage emerged. From then on, consorts—​and, therefore, all queen mothers—​were originally slaves. Peirce, Imperial Harem, 28–​56.



A Queen Mother and the Ottoman Imperial Harem

j 219

5.  İpşirli, “Hasan Kafi el-​Akhisarî,” 250. For Piri Pasha’s advice to Yavuz Sultan Selim, see Yücel, 30. For criticism of Sheikh-​ul-​Islam Sunullah Efendi brought in 1599, see Selaniki, Tarih-​i Selaniki, II:826. 6. Pierce’s work (Imperial Harem) contributes a great deal to refuting the arguments about imperial women’s interference in politics. 7. According to Sieur de la Croix, she was born in 1642: de la Croix, Le Sérail des empereurs turks, 165. Sutton notes that she was in her 90s in 1712. Sutton, Despatches of Sir Robert Sutton, 148. But, the information given by Silahdar seems to be more reliable because of his affinity to the family of Mehmed IV. He states that she was 75 when she died in 1715. Silahdar, “Nusretname,” 837. 8. Chassepol, History of the Grand Visiers, 168–​169; von Hammer-​Purgstal, Histoire, XI:165. Rycaut notes that her father was bishop of the region. Rycaut, The History of Turks Beginning with the Year 1679, 522. According to Sieur de la Croix, she was from the Verzizi family, Le Sérail, 165. 9. Von Hammer-​Purgstal, Histoire, XI:164–​165; de Saint-​Georges, Account of a Late Voyage to Athens, 233; de la Croix, Le Sérail, 165. 10. Türek and Derin, “Feyzullah Efendi’nin Kendi Kaleminden Hal Tercümesi,” 69. 11. Von Hammer-​Purgstal, Histoire, XII: 80; Ağa, Silahdar Tarihi, II:1–​5. 12. Chassepol, History, 234. de Saint-​Georges, Account of a Late Voyage to Athens, 348–​349; von Hammer-​Purgstal, Histoire, XII:240–​241. 13. Le Noble notes that the Haseki Gulnus Sultan gave 2,000 purses of 500 crowns each. Le Noble, Abra-​Mulè, 119. 14. De la Croix, Le Sérail, 166–​167, and Memoires Sieur de la Croix, 392–​393; de Saint-​Georges, Account of a Late Voyage to Athens, 348–​349; von Hammer-​Purgstal, Histoire, XI:413; Jorga, Geschichte des Osmanischen reiches, IV:225. 15. Eunuchs were employed at royal courts and in great households, where they entered services as slaves. They functioned as intermediaries between the imperial women and the members of the imperial court, persons outside the imperial court. 16. In a letter to Yusuf Agha, Le Noble writes that he knows the fact that Yusuf Agha was supported by Gülnuş Sultan. Le Noble, Abra-​Mulè, 30. 17. For 1675 festivals, see Nutku, IV: Mehmed’in Edirne Şenliği (1675); Coke, True Narrative. 18. Topkapı Palace Museum Archive (TSMA), D 1416. 19. Bakkalzade Defterdar, 251. 20. Silahdar, Silahdar Tarihi, II: 242. 21. Silahdar, Silahdar Tarihi, I:574, II:10; Uzunçarşılı, Osmanlı Tarihi, 444;Köprülü, “Hüseyin Paşa Amcazade,” 646; Aktepe, “Amcazade Hüseyin Paşa,” III:8–​9. 22. Jorga, Geschichte, IV:197, 200. These eunuchs might be the chief harem eunuch Yusuf Agha and Bosnak Sari Sileyman Agha, both of whom were enemies of Kara Mustafa Pasha. See Özcan, “Merzifonlu Kara Mustafa Paşa,” XXIX:246–​248. 23. Von Hammer-​Purgstal, Histoire, XII: 242–​243. 24. Jorga, Geschichte, IV:226. 25. Silahdar, “Nusretname,” 21; Cantemir, History of the Growth and the Decay, 397. 26. Efendi, Uşşâkîzâde Tarihi, I:179; Özcan, Anonim Osmanlı, 108. 27. Von Hammer-​Purgstal, Histoire, XII:242–​243. 28. Baltacı Mehmet Pasha gives this information in his letter to valide sultan (TSMA, E 2989). For transliteration see Kurat, Prut seferi ve Barışı, II:815–​823. 29. Türek and Derin, “Feyzullah Efendi’nin,” 79.

220 i

Betül İpşirli Argıt

30. Silahdar, “Nusretname,” 546. 31. TSMA, E 79/​33–​34. 32. In an undated document, Kethüda gave her news of the conquest of the castles. TSMA, E 145/​30. 33. Silahdar, “Nusretname,” 77, 202. 34. Rycaut, History of the Turks, 550–​551. At this point, it should be noted that in the Ottoman hierarchy, many of the elite were slaves but owned their own slaves. Gülnuş Sultan owned slaves purchased or given to her while she was still a slave during her haseki period. The elite would often train these slaves and then marry them to elite men to form ties of patronage. 35. Rycaut, The History of the Turks, 553. 36. TSMA, E 79/​15. 37. Zinkeisen, Geschichte des Osmanischen Reiches in Europa, V:202. 38. Quoted in Setton, Venice, Austria, and the Turks, 413–​414. 39. Rycaut, History of the Turks, 522, 527. 40. Pedani-​Fabris, Relazioni di ambasciatori veneti, 760. 41. The reason for this event was the influence of the political, economic, and social depression into which the Ottomans fell after the Vienna defeat. The sheikh-​ul-​Islam, Seyyid Feyzullah Efendi, who had great influence on Mustafa II, interfered in state affairs and placed his own men in high ranks throughout the imperial court, and hence increased his influence over the sultan. For details of the incident see Özcan, Anonim Osmanlı, 223–​245; Naima, Tarih-​i Naima, IV:1857–​1892; Andreasyan, “Balatlı Georg’a göre Edirne Vakası,” 47–​64; Meservey, “Feyzullah Efendi”; and Rifa’at Ali Abou-​El-​Haj, The 1703 Rebellion. 42. Özcan, Anonim Osmanlı, 232. 43. Andreasyan, “Balatlı Georg’a göre Edirne Vakası,” 50. 44. de la Motraye, Voyages du Sir A. de la Motraye, I:334. 45. Habesci, Present State of the Ottoman Empire, 77–​78. 46.  Defterdar, Zübde-​i Vekaiyat, 808–​809; Silahdar, “Nusretname,” 619; Özcan, Anonim Osmanlı, 238–​239. 47. Von Hammer-​Purgstal, Histoire, XIII:129. 48. Cantemir, History, 442–​444. 49.  TSMA, E 79/​18. A  document signed by Sheikh-​ul-​Islam and six kazaskers states that Shahzade Ahmed was selected to the throne. This document was sent to the valide sultan (TSMA, E 3065). Uzunçarşılı, “Osmanlı Tarihinde Gizli Kalmış Vesikalar.” 50. TSMA, E 79/​20. 51. Özcan, Anonim Osmanlı, 253. 52. La Motraye, Voyages, I:334. 53. Defterdar, Zübde-​i Vekayiat, 821; Silahdar, “Nusretname,” 639. 54. Özcan, Anonim Osmanlı, 250. 55. Pedani-​Fabris, Relazioni di ambasciatori veneti, 767. 56. Voltaire, The History of Charles XII, trans. W. Todhunter, 189. 57. Fabricius, Genuine Letters, xiv–​xv. 58. Fabricius, Genuine Letters, xiv–​xv. 59. Silahdar, “Nusretname,” 418. 60. Von Hammer-​Purgstal, Histoire, XIII:160. 61. Voltaire, The History of Charles XII, trans. W. Todhunter, 196–​198.



A Queen Mother and the Ottoman Imperial Harem

j 221

62. Members of the imperial dynasty including valide sultans possessed slaves and trained them and married them well. This established system of patronage could be used by the members of the dynasty to increase their influence. 63. Voltaire, The History of Charles XII, trans. W. Todhunter, 211. 64. TSMA, E 2989. 65. Özcan, Anonim Osmanlı, 270. 66. Kurat, İsveç Kralı XII. Karl’ın Türkiye’de Kalışı. 67. Cantemir, History, 297. 68. TSMA, E 833. 69. The Prime Ministry Ottoman Archive (or, BOA), C. SM 102/​5143. 70. Voltaire, The History of Charles XII, trans. W. Todhunter, 211. 71. Kurat, İsveç Kralı XII. Karl’ın Türkiye’de Kalışı, 218. 72.  For the first letter, see Paris, Archives des Affaires Étrangères (AE) Correspondance Politique (cor.pol.) vol. 52: 96–​97. In Kurat, Metinler, 196–​197. For the second letter, dated April 2, 1713, see Paris, AE cor. pol. Turquie, 52:115. In Kurat, Metinler, 197–​198. 73. Mahmud, Mecmua, 12–​13. 74.  Paris, Archives des Affaires Étrangères (AE) Correspondance Politique (cor.pol.) vol. 52:  96–​97. In Kurat, Metinler, 197. 75.  Paris, Archives des Affaires Étrangères (AE) Correspondance Politique (cor.pol.) vol. 52:  96–​97. In Kurat, Metinler, 198. 76. The valide sultan’s correspondence with the king was a disputed subject even at that time. Aubry de la Motraye criticizes Voltaire because he claimed that the king and the valide sultan were in contact. See de la Motraye, Remarques historiques, 1732. But, Baron Fabricius’s letters verify the secret correspondence between two (Fabricius, Genuine Letters, 238). For disproof of Motraye’s claims see François-​Marie Arouet de Voltaire, The History of Charles XII, trans. Antonia White, 252. 77. Kurat, İsveç Kralı XII. Karl’ın Türkiye’de Kalışı, 646. 78. Voltaire, The History of Charles XII, trans. W. Todhunter, 189; von Hammer-​Purgstal also notes that Poniatowski caught the valide sultan’s attention with the affair of the king via the Portuguese doctor Fuseca and a Jewish woman. von Hammer-​Purgstal, Histoire, XIII:212. 79. Zinkeisen, Geschichte, V:463. 80. Silahdar, “Nusretname,” 708; Efendi, Uşşâkîzâde Tarihi, 953; Ayvansarâyî, Hadîkatü’l-​Cevâmi’, 594. 81. Pedani-​Fabris, Relazioni di ambasciatori veneti, 800–​801; Raşid, Tarih-​i Raşid, II:391, 398; Silahdar, “Nusretname,” 143; Defterdar, Zübde-​i Vekâyiat, 607–​608. 82. Silahdar, “Nusretname,” 17. 83. Silahdar, Silahdar Tarihi, I:525–​526, 544–​546. 84. TSMA, D 10090. 85. Based on the fact that the institution of the valide sultanate was not unchanging and timeless, examining the valide sultans who lived after Gülnuş Valide Sultan would be useful. Among them were Saliha Sultan (mother of Mahmud I), who became valide sultan in the aftermath of the 1730 revolt and lived with the influential chief harem eunuch Beşir Agha, and Mihrişah Valide Sultan (mother of Selim III), who witnessed the period of reforms. 86. For the importance of personal experience apart from gender, see Downs, “What Future for Gender History,” 69–​94. 87. Naima, Tarih-​i Naima, III:1326. 88. de Saint-​Georges, Account of a Late Voyage to Athens, 359.

222 i

Betül İpşirli Argıt

Bibliography Archival Sources Başbakanlık Osmanlı Arşivi (BOA) (The Ottoman Archives of the Prime Minister’s Office), C. SM (Cevdet Saray) 102/​ 5143. Topkapı Sarayı Müzesi Arşivi (TSMA) Topkapı Palace Museum Archive, E 79/​15, E 79/​18, E 79/​20, E 79/​33–​34, E 833, E 3065, E 2989; D 1416.

Primary Sources Ayvansarâyî, Hüseyin Efendi. Hadîkatü’l-​Cevâmi‘ (İstanbul Câmileri ve Diğer Dînî-​Sivil Mimârî Yapılar). Edited by A. N. Galitekin. Istanbul: İşaret Yayınları, 2001. Defterdar Sarı Mehmed Paşa. Zübde-​i Vekayiat:  Tahlil ve Metin:  1066–​1116/​1656–​1704. Ankara: Türk Tarih Kurumu, 1995. Cantemir, Dimitrie. The History of the Growth and the Decay of the Ottoman Empire. Translated by N. Tindal. London: John James, Paul Knapton, 1734. Chassepol, Francois. The History of the Grand Visiers, Mahomet and Achmet Coprogli, of the Three Last Grand Signiors, Their Sultana’s and Chief Favourites, with the Most Secret Intrigues of the Seraglio besides Several Other Particulars of the Wars of Dalmatia. London: H. Brome, 1677. Coke, Thomas. A True Narrative of the Great Solemnity of the Circumcision of Mustapha, Prince of Turkie, . . . Together with an Account of the Marriage of His Daughter to His Great Favourite, Mussaip at Adrianople, as It Was Sent in a Letter to Person of Honour, by Mr. Coke. London:  W. Crook, 1676. de la Croix, Sieur. Memoires Sieur de la Croix. Paris: Cellier, 1684. —​—​—​. Le Sérail des empereurs turks, Relation manuscrite du sieur de la Croix à la fin du règne du sultan Mehmed IV. Edited by Corienne Théoaut-​Cabasset. Paris: CTHS, 2007. de la Motraye, Aubry. Remarques historiques et critiques sur l’Histoire de Charles XII de M.  de Voltaire. London: Chez Pierre Dunoyer, 1732. —​—​—. Voyages du Sir A. de la Motraye en Europe, Asie and Afrique. Edited by J. van Duren and T. Johnson. Lahaye: T. Johnson and J. Van Duren, 1727. de Saint-​Georges, Georges Guillet. An Account of a Late Voyage to Athens Containing the Estate Both Ancient and Modern Famous City, and of the Present Empire of the Turks, the Life of the Now Sultan Mahomet with the Ministry of the Grand Vizier Coprogli Achmet Pacha:  Also the Most Remarkable Passages in the Turkish Camp at the Siege of Candia and Divers Other Particularities of the Affairs of the Port. London: John Macock, 1676. Evliya Çelebi. Evliya Çelebi Seyahatnamesi, Topkapı Sarayı Bağdat 307 Yazmasının Transkripsiyonu-​ Dizini. Edited by Y. Dağlı, S. A. Kahraman, and İ. Sezgin. Istanbul: Yapı Kredi, 2001. Fabricius, Baron. The Genuine Letters of Baron Fabricius, Envoy from His Serene Highness the Duke Administrator of Holstein to Charles XII of Sweden. London: T. Becket, 1761. Habesci, Elias. The Present State of the Ottoman Empire Containing a More Accurate and Interesting Account of the Turks. London: R. Baldwin, 1784. Jorga, Nicolae. Geschichte des Osmanischen Reiches. Gotha: Academie Imperiale des Science, 1908. Le Noble, M. Abra-​Mulè: Or, A True History of the Dethronment of Mahomet IV. Translated by J. P. London: R. Clavel, 1696.



A Queen Mother and the Ottoman Imperial Harem

j 223

Naima. Tarih-​i Naima. Edited by M. İpşirli. Ankara: Türk Tarih Kurumu, 2007. Özcan, Abdülkadir, ed. Anonim Osmanlı Tarihi (1099–​1116/​1688–​1704).Ankara:  Türk Tarih Kurumu, 2000. Pedani-​Fabris, Maria Pia. Relazioni di Ambasciatori veneti al senato. Vol. XIV: Constantinopoli Relazioni İnedite (1512–​1789). Padova: Bottega D’Erasmo, 1996. Raşid. Tarih-​i Raşid. Istanbul: Matbaa-​i Âmire, 1282. Rycaut, Paul. The History of the Turkish Empire from the Year 1623 to the Year 1677: Containing the Reigns of the Three Last Emperours, viz. Sultan Morat or Amurat IV, Sultan Ibrahim, and Sultan Mahomet IV His Son, the XIII: Emperour Now Reigning. London: R. Clavell, 1687. —​—​—​. The History of Turks Beginning with the Year 1679. London: R. Clavell, 1700. Selaniki. Tarih-​i Selaniki. Edited by Mehmet İpşirli. Ankara: Türk Tarih Kurumu, 1999. Silahdar, Fındıklılı Mehmed Ağa. “Nusretname: Tahlil ve Metin (1106–​1123/​1695-​1721).” Edited by M. Topal. Unpublished PhD diss., Marmara University, 2001. —​—​—​. Silahdar Tarihi. İstanbul: Devlet Matbaası, 1928. Sutton, Sir Robert. The Despatches of Sir Robert Sutton Ambassador in Constantinople: 1710–​1714. Edited by Akdes Nimet Kurat. London: Royal Historical Society, 1953. Uşşâkîzâde İbrahim Efendi. Uşşâkîzâde Tarihi. Edited by Raşit Gündoğdu. Istanbul:  Çamlıca Basım Yayın, 2005. Voltaire. The History of Charles XII King of Sweden. Translated by W. Todhunter. London:  J. M. Dent, 1762. von Hammer-​Purgstal, Joseph Freiherr. Histoire de l’Empire Ottoman. Translated by J. J. Hellert. Paris: Bellizard, 1838. Yücel, Yaşar, ed. Osmanlı Devlet Teşkilatına Dair Kaynaklar, Kitab-​ı Müstetab-​Kitabu Mesalihi’l Müslimîn ve Menâf ’i’il Müminin-​Hırzu’l Mülûk. Ankara: Türk Tarih Kurumu Yayınları, 1988. Zinkeisen, Johann Wilhelm. Geschichte des Osmanischen Reiches in Europa. Gotha:  Friedrich Andreas Perthes, 1857.

Secondary Sources Abou-​El-​Haj, Rifa’at Ali. The 1703 Rebellion and the Structure of Ottoman Politics. Leiden: Nederlands Historisch-​Archaelo, 1984. Aktepe, Münir. “Amcazade Hüseyin Paşa.” In İslâm Ansiklopedisi. 44 vols. , Vol. 3:  8–​9. Istanbul: Türkiye Diyanet Vakfı. İslâm Ansiklopedisi Genel Müdürlüğü, 1988–​2013. Andreasyan, Hrand. “Balatlı Georg’a göre Edirne Vakası.” İstanbul Üniversitesi Edebiyat Fakültesi Tarih Dergisi 15 (1960): 47–​64. Harris, Ruth, and Laura Lee Downs. “What Future for Gender History: Is Gender Dead?” In Writing Contemporary History, edited by Robert Gildea and Anne Simonin. 69–​94. London: Bloomsbury, 2008. İpşirli, Mehmet. “Hasan Kafi el-​Akhisarî ve Devlet Düzenine ait Eseri Usûlü’l-​hikem fî Nizâmi’l-​ Alem.” Tarih Enstitüsü Dergisi 10–​11 (1979–​1980): 239–​278. İslâm Ansiklopedisi (İA). 44 vols. Istanbul:  Türkiye Diyanet Vakfı. İslâm Ansiklopedisi Genel Müdürlüğü, 1988–​2013. Köprülü, Fuat. “Hüseyin Paşa Amcazade.” In İslâm Ansiklopedisi. 44 vols. Vol. 5:  646-​650. Istanbul: Türkiye Diyanet Vakfı. İslâm Ansiklopedisi Genel Müdürlüğü, 1988–​2013.

224 i

Betül İpşirli Argıt

Kurat, Akdes Nimet. İsveç Kralı XII. Karl’ın Türkiye’de Kaldığı Zamana ait Metinler ve Vesikalar. Ankara: Ankara Üniversitesi Dil Tarih Coğrafya Fakültesi, 1943. —​—​—​. İsveç Kralı XII. Karl’ın Türkiye’de Kalışı ve bu sıralarda Osmanlı İmparatorluğu. Ankara: Ankara Üniversitesi Dil ve Tarih Coğrafya Fakültesi, 1943. —​—​—​. Prut seferi ve Barışı:  1123 (1711). Ankara:  Ankara Üniversitesi Dil ve Tarih Coğrafya Fakültesi Tarih Enstitüsü, 1951. Meservey, Sabra. “Feyzullah Efendi:  An Ottoman Şeyhülislam.” Unpublished PhD diss., Princeton University, 1965. Nutku, Özdemir. IV: Mehmed’in Edirne Şenliği (1675). Ankara: Türk Tarih Kurumu, 1972. Özcan, Abdülkadir. “Merzifonlu Kara Mustafa Paşa.” In İslâm Ansiklopedisi. 44 vols. Vol. 29: 246–​ 248. Istanbul: Türkiye Diyanet Vakfı. İslâm Ansiklopedisi Genel Müdürlüğü, 1988–​2013. Peirce, Leslie. The Imperial Harem:  Women and Sovereignty in the Ottoman Empire. New York: Oxford University Press, 1993. Refik, Ahmed. Kadınlar Saltanatı. 4 vols. Istanbul: Kitabhane-​i Hilmi, 1923 [1332]. Setton, Kenneth. Venice, Austria, and the Turks in the Seventeenth Century. Philadelphia: American Philosophical Society, 1991. Türek, Ahmed, and F. Çetin Derin. “Feyzullah Efendi’nin Kendi Kaleminden Hal Tercümesi.” İstanbul Üniversitesi Edebiyat Fakültesi Tarih Dergisi 23 (1969): 69–​92. Uzunçarşılı, İ. H. Osmanlı Tarihi, III/​II. Ankara: Türk Tarih Kurumu, 1988. —​—​—​. “Osmanlı Tarihinde Gizli Kalmış Vesikalar.” Belleten 163 (1977): 507–​554. Voltaire, François-​Marie Arouet de. The History of Charles XII King of Sweden. Translated by Antonia White. London: The Folio Society, 1976.

11 Hagar and Mariya Early Islamic Models of Slave Motherhood

Elizabeth Urban

i  According to early Arabic historical sources, the Shi‘i rebel Zayd ibn Ali rose up against the Umayyad state in 740 ce and claimed the caliphate for himself. The Umayyad caliph Hisham (r. 724–​743) learned of Zayd’s actions and scoffed, “You claim the imamate [i.e., the caliphate] when your mother is a slave? The imamate is not suitable for the children of slave women.”1 Zayd responded by invoking the figure of Hagar (Arabic, Hajar), concubine of Abraham (Arabic, Ibrahim) and mother of the Prophet Ishmael (Arabic, Ismail). If God deemed Ismail worthy of prophethood while his mother was a slave, why should Zayd not be worthy of the caliphate? Zayd’s revolt proved abortive, but the question of whether a slave-​born man could legitimately hold the caliphate remained open. After the Abbasid revolution brought an end to the crumbling Umayyad state in 132/​749, the second Abbasid caliph, Abu Jafar al-​Mansur, also defended his right to rule as the son of a slave woman. In this case, the Shi‘i rebel Muhammad al-​Nafs al-​ Zakiyya (“The Pure Soul”) renounced the fledgling Abbasid state in 145/​762–​763 and asserted his own authority as the noblest Arab, “the purest of both mother and father.” Al-​Mansur responded by calling on the figure of Mariya the Copt, concubine of the prophet Muhammad and mother of his son Ibrahim, who died in infancy. If the Prophet himself had deemed a slave woman worthy of bearing his son—​who has the best lineage of all—​why should al-​Mansur not be worthy of the caliphate? Thus, two prominent men living in the mid-​eighth century invoked two paradigmatic slave mothers from the Islamic religious tradition to justify their own political pretensions. These invocations capture an important moment in Islamic history, when slave mothers became intimately entwined in rhetoric about legitimacy and identity. This 225

226 i

Elizabeth Urban

article investigates these invocations of Hajar and Mariya to show just how important slave women were, not only in shaping early Islamic social and political institutions but also in shaping the very ideologies that underlay those institutions as well. There are two possible angles for analyzing this topic, and this chapter combines them in fruitful ways. One angle is male-​centric, for the protagonists here are powerful men who use female figures as pawns in their political games. This angle helps us discern the ways men construct idealized female models that do not take the women’s actual experiences into account, but rather serve their own purposes. (This is also the easiest angle to find in the primary source material.) However, this angle strips the women of their lived reality and agency. I thus complement this male-​centric angle with a more female-​centric angle, because these invocations reveal the pivotal role that slave women played in early Islamic history. This gendered analysis uncovers aspects of early Islamic history that might otherwise go unnoticed. Slave women had an impact not just on family structures and notions of marriage and sexuality that people often associate with “women’s history” but also on official ideologies often seen as part of the masculine sphere. Slave women may seem like the lowliest element in society, but here they inform the loftiest notions about who deserves to rule and who counts as a noble Arab. Ultimately, one cannot understand the changing notions of political legitimacy or Arab identity in the eighth century without understanding the deep influence of slave women on Islamic society. Hajar and Mariya The hagiographies of Hajar and Mariya are elaborated in several genres of Arabic-​Islamic writing, including biographical dictionaries, Quranic exegeses, and Tales of the Prophets literature. Scholars have long scrutinized these sources,2 and this chapter does not aim to assess the historical accuracy of these women’s biographies. However, learning how Hajar and Mariya are represented in the Islamic tradition will help us understand more fully the purpose they serve—​and do not serve—​in the historical accounts under consideration here. Particularly, it will show us that Zayd and al-​Mansur only called on these women because of two aspects of their biographies: (1) they were slaves and (2) they bore children to prophets. The men expunged all other aspects of these women’s memories, which apparently had no utility for their political agendas. If we read only these historical narratives, we would miss out on myriad other ways Hajar and Mariya were important in the Islamic tradition. Hajar was an enslaved Egyptian handmaiden given to Sara, wife of Ibrahim, by the Pharaoh. Because Sara was still barren at that time, she suggested that Ibrahim sleep with Hajar so that she might provide him with a child. Hajar then bore Ibrahim his son Ismail, with whom he would later rebuild the Kaaba and thus establish the roots of Abrahamic monotheism in Arabia. Hajar began to act haughtily toward Sara, who responded by cutting Hajar’s flesh—​piercing her ears and circumcising her. Despite these reprisals, the



Hagar and Mariya

j 227

presence of Hajar and Ismail continued to provoke Sara’s jealousy, so she eventually asked Ibrahim to remove the pair from the household. God directed Ibrahim to lead Hajar and the baby Ismail to the Arabian desert, where he prayed for their safety: O Lord, I have made some of my offspring settle in an unfruitful valley by your Sacred House, O Lord, so that they might establish the prayer. Therefore make the hearts of some people incline toward them and provide them with fruits, so that they may give thanks. (Q 14:37) The mother and son wandered in the wilderness until Ismail became thirsty, at which point Hajar began to search frantically for water. Her memory is enshrined in the rituals of the Hajj, as Muslims run between the hills of Safa and Marwa to commemorate Hajar’s search for water for Ismail. Gabriel miraculously opened up the well of Zamzam (the spring that waters Mecca) for them; not only did Zamzam save their lives, but this act also attracted a caravan from the local Bedouin tribe of Jurhum. Hajar helped Ismail find a wife among the Jurhum, and he assimilated into his new tribe. In this way, Ismail became the ancestor of the “Arabized” Arabs, which includes the prophet Muhammad’s tribe of Quraysh. Hajar is consequently known as the “mother of the People of the Water of Heaven.” As a pious believer who suffered for her faith and was ultimately vindicated, she can be considered “one of the pillars of Islamic consciousness” and “a symbol of Islamic identity.”3 As for Mariya, her story begins in the year 628 ce, when Muhammad reportedly sent letters from his polity in Medina to all the rulers of his day, calling them to Islam. One of the letters went to al-​Muqawqis, the Melkite patriarch who governed Alexandria for the Byzantine Empire. In response to this letter, the patriarch sent gifts to Muhammad, including a donkey and a mule, some fine robes and gold pieces, a eunuch, and two Coptic slave sisters named Mariya and Sirin. Mariya was reportedly very beautiful, with curly hair and fair skin, and she also converted to Islam on her way to Medina. For these reasons, Muhammad favored her and took her as his own concubine. In the year 629, Mariya bore Muhammad a son, Ibrahim. Muhammad declared that her son had made her free—​that is, Mariya was manumitted on the birth of Ibrahim. Aside from Ibrahim, all of Muhammad’s children had been born of Khadija (d. 619), his first wife, long-​time companion, and first believer in his message. The birth of Mariya’s child, coupled with her beauty and the fact that Muhammad clearly enjoyed spending time with her, provoked jealousy among his wives. He thus relocated her to a different part of Medina, farther away from his wives, which thereafter became known as the Portico of the Mother of Ibrahim. The domestic disturbances surrounding Mariya also resulted in the revelation of the first verses of Sura 66, which reassured Muhammad that his relations with her were acceptable. Sadly, Ibrahim died when he was about 16 months old. Muhammad died shortly thereafter, in 632, and the Muslim community helped provide for Mariya until her own death five years later.

228 i

Elizabeth Urban

There are some clear parallels between the biographies of these two women. They were both Egyptian slaves. They bore children to prophets, who also had children through their free wives. Because of their motherhood, they provoked jealousy in their prophets’ respective households. They were pious believers in their respective prophet’s monotheistic message, and they both occasioned the revelation of particular Quranic verses (14:37, Hajar; and 66:1–​4, Mariya). They are associated with certain geographic places (the Safa–​Marwa passage in Mecca for Hajar, and the Portico of the Mother of Ibrahim in Medina for Mariya). Finally, they are also both associated with a prophetic saying that encourages good Muslim relations with the Coptic population in Egypt.4 These parallels have led some scholars to suggest that Mariya never existed, or that her biography has been manipulated beyond recognition to match that of Hajar. Such scholars have argued that early Islamic authors constructed Mariya’s biography to bolster Muhammad’s connection to Abraham’s legacy and/​or to ancient Persian models of kingship.5 It is indeed likely that Mariya’s story was modeled after Hajar’s to some extent, and both women’s biographies contain certain religious themes that should not be considered strictly historical. But, I wish to highlight here what is historical about their invocations in the eighth century. Zayd and al-​Mansur do not engage in hagiographic elaboration, and they do not invoke these women to bolster Islam’s credentials regarding other religions or to explain verses of the Quran. Rather, these invocations serve as part of a political debate within the Islamic community about whether the children of slave women deserved to hold the caliphate. The debate, and the rhetoric surrounding Hajar and Mariya, emerged in a new context that developed approximately 120 years after the advent of Islam. For the first time in Islamic history, children born of slave women and free Arab men could hope to attain the caliphate. To understand why Hajar and Mariya suddenly became such salient symbols, we must place them against the background of this political ascent of the children of slave mothers. The Political Ascent of the Children of Slave Mothers For the first century of Islamic history, it seems to have been taken for granted that the caliph would be born of a free Arabian father and mother. Not only were all of the Rightly Guided and early Umayyad caliphs born of free mothers, but it has also long been asserted that the talented Umayyad prince Maslama ibn Abd al-​Malik (d. 738) was barred from the caliphate because his mother was a slave.6 However, by the end of the Umayyad period, starting in 740, when Zayd ibn Ali made his unsuccessful bid for the caliphate, the political tide was turning in favor of the children of slave mothers. By 744, the first Umayyad caliph born of a slave mother, Yazid III (r. 744), had taken control of the caliphate. Thereafter, the final three Umayyad caliphs and most of the Abbasid caliphs were born of slave women.7



Hagar and Mariya

j 229

The reasons for the political ascendancy of the children of slave mothers are still under investigation today.8 The new situation is likely attributable to a widespread political change that was taking place throughout the Umayyad period. Beginning with the first Umayyad caliph, Muawiya (r. 661–​680), the Umayyad state relied on networks of local Arabian and Syrian tribal elites to keep the government running smoothly. A marriage alliance between two tribes (such as Quraysh and Kalb, an alliance much favored by Muawiya) would produce a child who enjoyed widespread support among the tribal elites. By the end of the Umayyad period, however, the Islamic conquests had so stretched these tribal networks—​geographically across a vastly expanding empire, and demographically through the incorporation of outsiders into Arabian tribes—​that they ceased to function as efficient political tools. Instead of local tribal networks, there now emerged two great military factions that dominated the eastern regions of the empire. Members of these military factions competed fiercely for government positions and ultimately plunged the Umayyad state into civil war. The factions were based on an ideology of shared descent, through the paternal line, from one of two distant ancestors: Qahtan for the southern faction and Adnan for the northern faction.9 In such a climate, where the most important political identity was one’s place within one of the two major factions, one’s maternal kinship networks came to matter less than they had previously. In fact, if a man’s mother was from one faction and his father was from the other, his political support might be hopelessly divided. It was better to have a mother with no social ties—​in other words, a slave mother—​than to have a mother from the “wrong” faction.10 This political change probably resulted ultimately from the increasing numbers of slave women entering Islamic society through conquest, and bearing children within that society. The Quran speaks about the acceptability of taking slave women as sexual partners; male Muslims are considered chaste if they limit their sexual activities to their legal wives and to “those whom their right hands possess,” widely understood as a euphemism for slaves.11 Beyond the Quranic text, early Islamic law developed the category of the umm walad (literally “mother of a child”), a female slave who bears a child for her master and thereby gains certain legal rights. Although there are many subtle differences among the Islamic law schools concerning the rights and responsibilities of the umm walad,12 the general rules are as follows. First, when a woman becomes an umm walad by bearing a child for her master, she cannot be sold away. Second, she is freed automatically on the death of her master, unless he chooses to manumit her during his lifetime—​that is, motherhood is her path to freedom. And third, her child is considered freeborn and legitimate. The child is a full member of the father’s family, heir to the father’s estate, and equal legally to any of the father’s freeborn children. Built on this bedrock of religious and legal sanction for using slave women to bear children, it seems the prevalence of umm walads and their children increased steadily through the early Islamic period.

230 i

Elizabeth Urban

Whatever the reason for the improving political situation of the children of slave mothers, it remains that their ascendancy was complete by the end of the Umayyad period. This ascendancy was accompanied by rhetoric about their mothers, used either to justify or to lament their rise to the caliphate. One rhetorical strategy was to exalt the lineage of the slave mother by claiming that she was a foreign princess descended from great kings. For instance, rather than apologizing for his slave mother, the first slave-​born Umayyad caliph, Yazid III, hailed her as a Persian princess from the royal Sasanid line who had been captured during the Islamic conquests.13 He bragged of his doubly noble lineage: “I am the son of Khusraw; my father is Marwan. One grandfather is a Caesar, the other a Khaqan.”14 Another rhetorical strategy, in this case serving the opposite purpose of delegitimizing the children of slave mothers, casts doubt on the paternity of slave-​born children and suggests further that using slave women to bear children is a recipe for complete sociopolitical chaos. For instance, the enemies of the final Umayyad caliph, Marwan II ibn Muhammad (r. 744–​750), suggested he was not actually the son of the Umayyad prince Muhammad ibn Marwan, but that his slave mother was already pregnant when she was taken from the hostile camp of either Ibn al-​Ashtar or Musab ibn al-​Zubayr.15 This uncertainty about the paternity of slave-​born children fostered a measure of apocalyptic hysteria over the end of the Umayyad caliphate, as in the (historically dubious) account in which the caliph Uthman warns:  “This community will fall into innovation when three things happen: complete material prosperity, the attainment of adulthood by the children of concubines, and the recitation of the Quran by both Arabs and non-​Arabs.”16 The richest rhetorical strategy that emerged at this time is the explicitly “Islamic” model of slave motherhood created by the invocation of Hajar and Mariya. This model justifies the political aspirations of the children of slave mothers by citing prophetic precedent—​it frames political legitimacy in religious language. Indeed, this model does more than merely give religious sanction to slave-​born caliphs. It also creates a new notion of how lineage works and who counts as an “Arab.” That is, if lineage works in such a way that the child of a prophet deserves to be a prophet regardless of his mother’s identity, or the child of an imam deserves to be an imam regardless of his mother’s identity, then the child of an Arab father deserves to count as a full Arab regardless of his mother’s identity. These ramifications can be extracted from the several versions of Zayd’s invocation of Hajar and the single version of al-​Mansur’s invocation of Mariya. The Accounts Zayd’s invocation of Hajar represents the earliest moment in the development of a new religious ideology of motherhood. Zayd’s invocation comes in six versions, which I  present here in full before turning to their analysis. In one anomalous account, Zayd’s interlocutor is his cousin Abdallah ibn Hasan ibn Hasan. Abdallah challenges



Hagar and Mariya

j 231

Zayd: “Do you want to obtain this while you are the son of an Indian slave?” Zayd responds: 1. Ismail was the son of a slave woman, and he obtained more than that.17 The other five accounts are more elaborate, and they are presented as a conversation between Zayd and the Umayyad caliph Hisham. After Hisham scoffs at Zayd’s aspirations for the caliphate, Zayd says: 2. O Commander of the Faithful, this Ismail ibn Ibrahim—​upon them both be peace—​was the son of a slave woman, and prophethood was right for him, and he was true to his promise, and he was pleasing to his Lord. And prophethood is greater than the imamate! Hisham said to him: O Zayd, God will never combine prophethood and kingship in one person. Zayd said: What is this [nonsense] O Commander of the faithful? For God Almighty said: “Or do they envy the people to whom God has granted His bounty? For We have given the family of Ibrahim the Book and the wisdom, and we have given them a great kingdom.” (Quran 4:54)18 3. If there were some deficiency in umm walads, then God would not have sent Ismail as a prophet while his mother was Hajar. And is the caliphate greater or prophethood?19 4. Nobody is closer to God or higher in rank with Him than a prophet whom He has sent. Ismail was one of the best prophets, and he produced the very best of them, Muhammad. Ismail was the son of a slave woman, while his brother [Ishaq] was born of a free woman, just like you. But God chose Ismail in preference to [Ishaq] and made the best of mankind to come forth from him. No one has any [standing] over one whose ancestor was the Prophet of God, whoever his mother was.20 5. Woe to you! The status of my mother puts me in [a good] position. By God, Ishaq was the son of a free woman while Ismail was the son of a slave woman, and God Almighty singled out the descendants of Ismail and made the Arabs from them, and that continued until the Messenger of God was from them. Fear God, O Hisham!21 6. Mothers serve no purpose for men other than to reach the goal [of bearing children]. For the mother of Ismail was a slave, and that did not prevent God from sending him as a prophet, or from making him the father of the Arabs, or from sending out Muhammad from his loins. So do you say this to me, when I am a son of Fatima and my grandfather is Ali?22 The first account is, in some sense, the most straightforward and, in some sense, the most veiled. Zayd’s response is stripped down radically; it mentions nothing other than the fact

232 i

Elizabeth Urban

that Ismail was born of a slave. Despite this simplicity, it is difficult to say exactly what is at stake in the account, for the topics at hand are only alluded to obliquely as “this” and “that.” The debate between Abdallah and Zayd seems to be about the leadership of the Alid family. The Alids considered themselves the “Family of the Prophet” who deserved to rule because they were the closest heirs of Muhammad. According to Alid doctrine, Muhammad had designated Ali ibn Abi Talib—​his cousin, son-​in-​law, and early believer in his message—​to succeed him as leader of the Muslim community. Religious-​political authority continued to pass down Ali’s bloodline, particularly through al-​Hasan and al-​Husayn, his sons from his marriage to the Prophet Muhammad’s daughter Fatima.23 Thus, Zayd here invokes Ismail/​Hajar to justify his own attempt to become the guardian of the Alid estate, to be recognized as the Shii Imam of the day, and perhaps to seize the caliphate. The remaining accounts present Zayd’s invocation not as part of a debate within the Alid household, but as a debate between Zayd and the Umayyads. Almost half of the second account is taken up by a discussion about the relationship between secular power and religious authority. Hisham seems to be advocating here for a separation of powers. The Umayyads are secular kings but not heirs of the prophets, whereas the heirs of the prophets (presumably, the Alids) have no claim to political power. Zayd counters by invoking the family of Ibrahim, to which God granted both prophethood and kingship. By analogy, Zayd is arguing for an office of the imamate/​caliphate that carries both religious and political authority. Zayd locates himself within the tradition of the prophet-​kings through his own slave mother and advocates for his own position as imam-​king. This account is historically suspect because it exhibits a clear anti-​Umayyad bias. That Hisham would imply that he and the Umayyads are simply “kings” with no religious legitimacy seems unlikely; Crone and Hinds have shown that the Umayyads saw themselves as agents of God.24 Especially given that all the other accounts are directed more against Hisham’s perceived Arab chauvinism than his secular conception of rule, historians can dismiss this second account as anti-​Umayyad polemic. The third version of Zayd’s speech also highlights the connection between prophethood and the imamate, but its upshot has broader implications:  “If there were some deficiency in umm walads, then God would not have sent Ismail as a prophet while his mother was Hajar.” The discussion has shifted from the nature of the caliphate and the relationship between religious and secular rule to a justification of using slave women to bear children. The conclusion drawn from this account potentially concerns not just the political elite, but anyone whose mother was a slave concubine or who wished to use a slave concubine for procreation. As mentioned earlier, slave women had long been used for sexual purposes, and their children were considered legitimate legally. However, it seems that, during the earliest years of Islamic history, slave women were not always considered desirable procreative partners, and their children seem to have experienced some degree of cultural (if not legal) bias. This cultural bias against the children of slave



Hagar and Mariya

j 233

mothers had been waning throughout the Umayyad period, resulting in part from the appearance of great slave-​born generals such as Maslama ibn Abd al-​Malik and scholars such as Ali ibn al-​Husayn (Zayn al-​Abidin). Here, the invocation of Hajar deals this bias a religious deathblow by creating an explicitly Islamic identity for the children of slave mothers. The fourth account brings in two new elements. The first is the notion that the Prophet Muhammad is an heir of Ismail, which is implied but not spelled out in the previous accounts. By naming Muhammad directly, Zayd makes it clear that Umayyad obsession with “purity” not only disregards the ancient practice of Ibrahim, but also it insults the Prophet Muhammad as a descendant of Hajar. Moreover, the mention of Muhammad reveals that Zayd is advocating for the Alids as the closest descendants of Muhammad, against the Umayyads who have no such claim. Zayd insists that his slave mother does not invalidate his position as the best descendant of Muhammad, just as Ismail’s slave mother does not invalidate his position as the best descendant of Ibrahim. The doctrine of descent from Muhammad also lies at the heart of al-​Mansur’s invocation of Mariya described later, as we shall see. The more controversial idea brought up in this fourth account is that the slave-​born Ismail was to be preferred over the freeborn Ishaq (Isaac). Other scholars have investigated the stories of Ismail and Ishaq in Islamic religious literature, particularly the debates over which son was almost sacrificed by Ibrahim.25 This religious literature serves largely to express Islam’s relationship to the other Abrahamic religions; Ismail is seen as the ancestor of the Arabs/​Muslims and Ishaq is seen as the ancestor of the Jews. Moreover, in the Shu`ubiyya literary controversy of the Abbasid era, Shu`ubi partisans apparently boasted that the Persians were descended from the freeborn Ishaq and denigrated the Arabs as descendants of “the uncircumcised woman,” meaning Hajar.26 Here, however, we find a different context for debating the relative merits of Ismail and Ishaq, one not concerned with interfaith relations or with cultural superiority, but with an intra-​Islamic debate about the nature of politics. Ismail asserts that the children of slave mothers can actually be more suitable for the prophethood (and, by implication, the caliphate) than freeborn children. The final two accounts draw an even broader generalization about the descendants of Ismail. It is not only Muhammad as an individual, but the Arabs more generally, who are descended from Ismail. “God Almighty singled out the descendants of Ismail and made the Arabs from them,” and “the mother of Ismail was a slave, and that did not prevent God . . . from making him the father of the Arabs.” These accounts assert that having a slave ancestress does not dilute one’s Arab lineage, and they imply further that Arab lineage is passed down through the male line alone. According to this doctrine of male descent, Zayd is just as much of an “Arab” as Hisham; they were both (theoretically) the patrilineal descendants of Adnan, the ancestor of the northern Arabs, who was himself supposedly a distant descendant of Ismail. Thus, we find this invocation of Hajar is not merely about religious ideology; it also advocates subtly a new definition of “Arabism.” In

234 i

Elizabeth Urban

al-​Mansur’s invocation of Mariya described later, this new patrilineal concept of Arabism is highlighted even more deliberately. The sixth and final version is quite explicit that this invocation of Hajar serves to devalue the contribution of mothers to their children’s genealogies. Here, mothers appear as nothing more than a means to an end—​namely, the passing along of the paternal lineage. Hajar was simply an instrument that God gave Ibrahim to provide him with a son. Her own identity does not affect Ismail’s prophetic lineage. By implication, mothers do not affect caliphal lineage, Arab lineage, or any kind of lineage whatsoever. However, this invocation of Hajar stands in an uneasy tension with another mother whom Zayd mentions at the end of the account: Fatima, the daughter of the Prophet Muhammad, wife of Ali ibn Abi Talib, and mother of the Prophet’s two beloved grandsons al-​Hasan and al-​Husayn. By saying that the Prophet Muhammad was his “ancestor,” Zayd appears to advocate for a doctrine of descent from the Prophet through Fatima. According to this doctrine, Fatima is unique—​the one single instance in which lineage was passed down through the maternal line (for Muhammad had no sons who survived into adulthood). Because Fatima is a singular exception to the rule of male descent, Hajar’s presence need not challenge Fatima’s status directly. Yet, by invoking Hajar to demonstrate that women are simply vessels of the male seed, and that lineage is passed down through the male line only, Zayd advances implicitly a model of motherhood that stands as a counterexample to Fatima. Ultimately, the awkward mention of Fatima at the end of this account, and her absence from the other accounts, suggests that her presence here is a later interpolation. In contrast to the multiple variations of Zayd’s invocation, I have only been able to locate one version of al-​Mansur’s invocation of Mariya the Copt, transmitted by both al-​ Tabari and al-​Baladhuri. The stable text of this account represents a sophisticated piece of Abbasid propaganda. In this account, the rebel al-​Nafs al-​Zakiyya, who was descended from the Prophet’s grandson al-​Hasan, had sent caliph al-​Mansur a letter in which he claimed to be “the center of the Banu Hashim in terms of genealogy.” He bragged: “I am the purest of both mother and father—​the non-​Arabs have put down no roots in me.”27 To this, al-​Mansur, whose own mother was a slave named Sallama, responded: You are proud that the non-​Arabs never bore you, meaning that you have no blood from umm walads. But you have boasted over someone who is better than you in genealogy from beginning to end: Ibrahim, the son of the Messenger of God (peace be upon him), whose mother was Mariya the Copt. Moreover, there has not been born among you [Alids] anyone better than Ali ibn al-​Husayn [Zayn al-​Abidin, the fourth Shii Imam], who was born to an umm walad. He was better than your grandfather Hasan ibn Hasan. And after [Ali ibn al-​Husayn], there was no one better among you than his son Muhammad ibn Ali ibn al-​Husayn [al-​Baqir, the fifth Imam], whose grandmother was an umm walad. And as for your claim that you are the sons of the Prophet Muhammad, God most high said: “Muhammad



Hagar and Mariya

j 235

is not the father of any of your men, but rather the messenger of God and the seal of the Prophets” (Q 33:40). You are the sons of the daughter of the Messenger of God—​may God have mercy on her—​but she does not preserve your inheritance or bequeath the trusteeship [of the Prophet’s family].28 This heated exchange between al-​Mansur and al-​Nafs al-​Zakiyya was part of an ideological disagreement between the Abbasids and the Alids over what it meant to belong to the “Family of the Prophet,” which both groups used as their basis for claiming political authority. As we have seen, Alids supported the descendants of the Prophet’s cousin and son-​in-​law Ali, particularly through the two sons he had by Fatima. Al-​Mansur and the Abbasids argued that, on the contrary, they were the closest descendants of the Prophet Muhammad because of their descent from the Prophet’s paternal uncle, al-​Abbas.29 To refute the Alids’ claims of descent from the Prophet through Fatima, al-​Mansur cites Mariya the Copt to prove that the Prophet Muhammad produced no male heirs who reached adulthood. In such cases in which immediate male kinsmen are not present, the next closest heir is the paternal uncle, not the daughter. The presence of Mariya, along with verses of the Quran and legal statutes, serves to demonstrate that the Abbasids, not the Alids, are the true descendants of the Prophet Muhammad. As we saw before in one of Zayd’s accounts, Fatima stands in uneasy tension with Hajar; here, al-​Mansur makes the ideological importance of Mariya as a counterpoint to Fatima perfectly clear. In addition to this ideological debate about the nature of inheritance from the Prophet Muhammad, al-​Mansur’s invocation of Mariya serves to justify his noble lineage. We have already seen that Zayd’s invocation of Hajar justified his identity not only as an imam but also as an Arab descended from Ismail. Al-​Mansur uses Mariya even more strongly to condemn al-​Nafs al-​Zakiyya for his concern about “non-​Arab blood” and his pride in his “pure” Arabic lineage on his mother and father’s side. By linking the children of slave women to the “non-​Arabs,” al-​Nafs al-​Zakiyya was trying to exclude them from a restricted definition of Arab that took both the mother and the father’s identity into account. Al-​Mansur’s invocation of Mariya, on the other hand, serves to expand the notion of Arabism to include anyone who was descended through the male line from an Arab ancestor. Somewhat paradoxically, this rhetorical move both highlights and erases Mariya; al-​Mansur calls on her precisely so that he can discount her genealogical contribution. In addition, the presence of Mariya gives this expanded notion of Arab identity an Islamic justification, based on the example of Ibrahim ibn Muhammad. Al-​Mansur seems to be saying: I am not advocating Arab chauvinism as practiced by the Umayyads,30 I am advocating an Islamically sanctioned notion of genealogy and noble lineage based on the model of the Prophet’s own family. In fact, these new religious models of slave motherhood invoked during late Umayyad and early Abbasid history were a necessary step toward the classical definition of “Arab” that surfaces during the Abbasid period, expressed particularly in the emergent genre of genealogical literature.31 Historical anthropologists, including Zoltan Szombathy,

236 i

Elizabeth Urban

Edouard Conte, and Daniel Martin Varisco, have studied this genealogical literature, which is almost always organized purely according to male descent.32 Although wives and mothers might be mentioned in such works, matrilineage does not matter for their organization or systematization.33 That is, there are no categories that might depend on the mother’s identity—​such as the old tribal concepts of “pure” or “half-​blood”—​but there is only “Arab,” determined solely by patrilineal descent. Moreover, like the political rhetoric about Mariya and Hajar under consideration here, these genealogies are political tools. They are used to bolster the political claims of those claiming descent from the Prophet (most notably the Abbasids),34 or to trumpet the political ascendancy of the Arabs over the non-​Arabs.35 Thus, I suggest that the accounts about Hajar and Mariya that reduce the role of the mother to mere “vessel” to justify the increasing political ambitions of the children of umm walads is a key step toward this simplified, expanded, and politicized view of who counts as an Arab in the genealogical genre.36 This analysis indicates that umm walads and their children are crucial for understanding the types of Arab identity expressed during the Abbasid era, and it also calls into question the idea that a coherent political “Arabism” existed before the end of the Umayyad period. A Word about Authenticity of the Accounts When researching topics in early Islamic history, one must use the available sources carefully. The narrative histories—​such as those of al-​Tabari (d. 923), al-​Yaqubi (d.c. 905), and Khalifa ibn Khayyat (d. 854); and biographical dictionaries such as those of Ibn Sad (d. 845), al-​Baladhuri (d. 892), and Ibn Asakir (d. 1176)—​were all written at least a century after Zayd and al-​Mansur’s reported invocations. Moreover, most of these authors held generally negative views of the Umayyads and often maligned them in subtle and not-​so-​ subtle ways. These authors were also steeped in the ideologies and concepts of their own days and were apt to back-​project anachronistically their worldviews onto earlier periods in history. Thus, one must ask: Are these invocations of Hajar and Mariya authentic to the eighth century? Or, do they reflect the interpolations of the later authors living in the ninth and tenth centuries? Although it is difficult to determine the historical veracity of these accounts with full certainty, they reflect a glimmer of historical fact that Zayd and Mansur did, indeed, invoke Hajar and Mariya in this way. The first feature that speaks for their authenticity is that these are the earliest invocations I  could find of Hajar and Mariya in the annals of Islamic history. I  expected to find references to Hajar and Mariya in earlier years, for slave women had been bearing children for Arabian tribesmen for generations. Hajar and Mariya seemed like “obvious” or “natural” figures to invoke against perceived Umayyad racial discrimination, or in the service of slave-​born identity narratives in early Islamic society. However, such invocations do not seem to exist before the mid-​eighth century, whether in the historical annals or in biographies of earlier slave-​born men, such as Muhammad ibn al-​Hanafiyya



Hagar and Mariya

j 237

(d. 700–​701), Ali Zayn al-​Abidin (d.c. 713), Muhammad ibn Marwan (d. 719–​720), and Maslama ibn Abd al-​Malik (d. 738). If these invocations of Hajar and Mariya had been fabricated to reflect the beliefs and symbols of later authors, one might have expected to find references to these women inserted earlier in the narratives of Islamic history. The second feature that indicates these accounts capture an actual historical moment is that they undergo an ideological development, from messy origins to crystallized doctrine. As we have seen, Zayd’s invocation of Hajar comes in six different versions that vary fairly significantly. All of the accounts emphasize that Ismail was a prophet although he was born of a slave woman; but beyond that, they diverge. One account contains an extended discussion about the nature of the caliphate, some accounts state that Ismail was preferred over Ishaq, some of them mention Muhammad and/​or “the Arabs” specifically as descendants of Ismail, one advocates bluntly for the practice of using slave women to bear children, and yet another mentions Fatima, who seems to somewhat undercut the original message. I interpret this variability to mean there was some vague historical memory that Zayd invoked Hajar, but different transmitters and authors had different ideas about what he said, to whom he said it, and what he meant by it. By the time of al-​Mansur’s invocation of Mariya a mere 23 years later, the language has become more stable—​an official Abbasid ideology wielded against the Alid Shia. Thus, I suggest that these accounts represent a generally sound documentation of the birth and maturation of a new rhetoric of motherhood that emerged for political reasons during the mid-​eighth century. Who Speaks for Slave Women? These invocations of Hajar and Mariya create a complex identity for the slave mother herself. On the one hand, by mentioning Hajar and Mariya as paradigmatic slave mothers, Zayd and al-​Mansur provide an “Islamic” identity for the umm walads of their own day. Hajar is remembered in the Islamic tradition as a pious woman, beloved by the prophet Ibrahim, and devoted to her son Ismail. Likewise, Mariya was chosen as the consort of the prophet Muhammad because of her beauty and her early conversion to Islam. Thus, invoking Hajar and Mariya can be seen as a way of exalting female slaves and providing them with positive religious role models, which may also have been an effective tool in encouraging umm walads to convert to Islam. However, the deeper effect of these invocations is to reduce mothers to the role of mere vessels of the male lineage. Although this focus on patrilineal descent served to expand the notion of “Arab” to include the children of slave women, this expansion seems to have come at the cost of women’s genealogical contributions to the identity of their children.37 In addition, it is worth repeating that Hajar and Mariya are considered appropriate motherhood models in these accounts only because they were slaves. People today are sometimes uncomfortable with the fact that Hajar and Mariya were slaves—​preferring to

238 i

Elizabeth Urban

think of them as free wives, perhaps even members of the royal family of Egypt. Scholars such as Aisha Hidayatullah and Kecia Ali use Mariya as an example of how Muslims and non-​Muslims alike continue to struggle with the legacy of slavery in Muslim sexual ethics. However, in these eighth-​century invocations, their slave status is central. These women are not invoked to debate the legality or ethicality of slave concubinage,38 but only to buttress the position of rising political leaders. The men who called on Hajar and Mariya had a vested interest in showing that the children of umm walads could have noble genealogies and legitimate political authority. The slave status of Hajar and Mariya was thus useful for the political aims of powerful men. Instead of blaming Zayd and al-​Mansur for their utilitarian invocation of these women, we should learn from their example. Even today, when people speak of Hajar, Mariya, Aisha, Fatima, or other women from the Islamic tradition, they are often less concerned about these women’s actual lived experiences and more interested in using these women as a barometer for the health of society or as a locus for their own interests. When anyone posits some “truth” about who Mariya and Hajar “really” were, they are closing their eyes to the ever-​unfolding, contextual nature of history. Different aspects of historical actors’ lives become salient in different contexts, including the early Islamic political context, and including our own diverse contemporary contexts. Noticing how Zayd and al-​Mansur invoke these women to fit their own concerns helps us better to appreciate the ways historical memory is constructed, and to acknowledge our own continued participation in that construction. These accounts also help us discern the difference between a specific historical memory invoked in a particular moment, and the broader historical importance of a person or group. This chapter has indicated there is a discrepancy between the limited way Hajar and Mariya are invoked in the narrative histories under consideration here, and their more capacious influence on the Islamic tradition. As Hajar and Mariya stand as symbolic representatives of the innumerable slave mothers of early Islamic history, we can apply this sense of discrepancy more broadly. The historical narratives mention the slave mothers of early Islamic history in a fairly limited capacity, and almost always in relation to their famous masters and sons. But, that should not trick us into thinking that the only thing about these women that matters is their status as mothers to powerful politicians and scholars. The memory of Hajar and Mariya stretches into the realm of theology, social customs, and interfaith relations; by implication, we can appreciate that early Islamic slave mothers likewise had a far-​ranging impact on Islamic history, even if the historical narrative sources only give them scant treatment. Conclusion According to the early Arabic historical sources, Hajar and Mariya emerged as paradigms of slave motherhood only after the children of slave mothers had attained a



Hagar and Mariya

j 239

certain level of power in the Islamic polity. This late emergence does not mean the stories of Hajar and Mariya were nonexistent or unknown before the mid-​eighth century; for the first time in Islamic history, their images became politically expedient. Invoking these women highlighted the political legitimacy of slave-​born caliphs, and it also created a new “Islamic” model of Arabism and male descent. The rhetoric about model slave mothers from the prophetic tradition potentially exalts the umm walad as a pious believer and integral member of Islamic society, but it also reduces her contribution to her children’s identity, making her into a mere vessel of the male line of descent. Ultimately, the salience of the models of Hajar and Mariya indicates that slave women had an enormous, if indirect, impact on the course of early Islamic history. These two women, as representatives of countless other slave mothers in early Islamic society, underscore how crucial it is to consider the role of women when investigating changing notions of belonging, identity, and authority in the first centuries of Islamic history. Notes 1. Ibn Asakir, Tarikh Madinat Dimashq, 19:368. “Imamate” refers to the office combining religious and political authority over the Muslim community; in this context, it is synonymous with “caliphate.” 2.  Hajar’s story has been elaborated particularly in the Tales of the Prophets literature, see Firestone, Journeys in Holy Lands, 37–​43, 63–​70, 74–​75. See also Lindsay, “Sarah and Hagar in Ibn Asakir’s History of Damascus”; and Stowasser, Women in the Quran, 43–​49. Mariya’s story is absent from the Sira of Ibn Ishaq, but it is elaborated in the Tabaqat of Ibn Sa’d and in exegesis of Q 66:1–​4. See particularly Aisha Hidayatullah, “Mariyya the Copt.” 3. Stowasser, Women in the Qur’an, 46, 49. 4. For instances of this saying that refer to Hajar, see al-​Tabari, Tarikh, I:270, 2585–​2586; Ibn Abd al-​Hakam, Futuh Misr, 2–​3; and Muslim, Sahih Muslim, 4:2–​3. For instances of this saying that refer to Mariya (or Mariya and Hajar), see al-​Baladhuri, Futuh al-​Buldan, 219; and Ibn Sad, Tabaqat, 8:154. 5.  See particularly Cannuyer, “Mariya, la concubine copte de Muhammad”; and Öhrnberg, “Mariya al-​Qibtiyya unveiled.” See also David Powers (Muhammad Is Not the Father of Any of Your Men) for the argument that Muhammad did not really have a son with Mariya who died in infancy, but this narrative was invented so Muhammad could fulfill his role as “seal of the Prophets.” 6. On Maslama, see Lammens, “Maslama”; and Borrut, Entre mémoire et pouvoir. Likewise, Maslama’s half-​brother, the caliph Sulayman, nominated a cousin as his heir to the caliphate, rather than nominate one of his own sons by a slave woman. 7. Some sources do not consider Yazid III and Ibrahim ibn al-​Walid to be proper caliphs (Ibn Asakir, for instance, leaves them out of his Tarikh Madinat Dimashq). In any case, the universally recognized final Umayyad caliph, Marwan II ibn Muhammad, was the son of an umm walad. 8. See c­ hapter 1 in this volume. For a different interpretation, see also Athamina, “How Did Islam Contribute to Change,” esp. 394–​398.

240 i

Elizabeth Urban

9.  For a thorough analysis of the advent of the military factions in Umayyad politics, see Crone, Slaves on Horses. 10. Such seems to have been the case with the Umayyad prince al-​Hakam ibn al-​Walid, whose political hopes were dashed by the fact that his mother was from the southern tribe of Kalb whereas his father’s supporters were from the northern tribes of Qays and Mudar. al-​Tabari, Tarikh, II:1892. 11. See Q 4:3, 23:6, and 33:50 for the permissibility of sexual intercourse with “those whom your right hands possess.” 12.  For some of the complicated details of the status of the umm walad in Islamic law, see Schacht, “Umm al-​Walad”; Brockopp, Early Maliki Law; and Marmon, “Concubinage, Islamic.” 13. Her exact identity in the sources is unclear; she is described alternatively as an unnamed daughter of the family of al-​Mudkhaj ibn Yazdagird (al-​Baladhuri, Ansab al-​Ashraf, 7:540); the daughter of Yazdagird ibn Kisra (Usfuri, Tarikh Khalifa ibn Khayyat, 386–​387); or Shahi Afrid, daughter of Firuz ibn Yazdagird (ibn Shahriyar) ibn Kisra (al-​Tabari, Tarikh, II:1874; al-​Yaqubi, Tarikh, 2:401). A similar rhetorical technique was also applied to the fourth Shii imam Ali ibn al-​Husayn Zayn al-​Abidin. See Amir-​Moezzi, “Šahrbanu.” In fact, because of the unknown provenance of most slave women, many concubines in history have been seen as “secret princesses.” A famous example of this phenomenon is Nakhshidil, mother of the Ottoman Sultan Mahmud II. According to legend, Nakshidil was actually Aimée du Buc, a French heiress who had been captured as a young girl by Barbary pirates. See Isom-​Verhaaren, “Royal French Women in the Ottoman Sultans’ Harem.” 14. al-​Tabari, Tarikh, II:1874. According to one account, she was so noble that she turned the notion of “half-​blood” on its head. When the man who bought her asked if any child of hers would be half-​blood, the people answered: “Yes, he will be half-​blood through his father.” That is, her blood was so royal that any man with whom she procreated would dilute her lineage. al-​ Tabari, Tarikh, II:1247. 15.  al-​Baladhuri, Ansab al-​Ashraf, 7:561. See also Ibn Asakir, Tarikh Madinat Dimashq, 57:319–​323. 16. al-​Tabari, Tarikh, I:2803–​2804. 17. al-​Tabari, Tarikh, II:1672. 18. Ibn Asakir, Tarikh Madinat Dimashq, 19:368. 19. Ibn Asakir, Tarikh Madinat Dimashq, 19:471. 20. al-​Tabari, Tarikh, II:1676. 21. al-​Yaqubi, Tarikh, 2:390. 22. al-​Baladhuri, Ansab al-​Ashraf, 7:363–​364. 23.  During the first centuries of Islamic history, some Alids supported descendants of Ali alone, without regard to Fatima. See Amir-​Moezzi, Divine Guide in Early Shiʻism; and Dakake, Charismatic Community. However, here we have a debate between two descendants of both Ali and Fatima: one Husaynid (Zayd) and one Hasanid (Abdallah). 24. See Crone and Hinds, God’s Caliph. 25. Particularly in Firestone, Journeys in Holy Lands, 135–​151. 26. The Shu`ubiya was a literary controversy of the ninth and tenth centuries ce, in which Muslims of Persian descent took pride in the great cultural legacy of Persia and made fun of the “uncivilized” history of the Arabs. Ibn Qutayba mentions this Shu`ubi view in his Fadl al-​Arab, 48. He refutes the Shu`ubi claim by defending Hagar, by pointing out that the Jews (not the Persians)



Hagar and Mariya

j 241

are descended from Ishaq, and by providing examples of noble, prominent slave-​born men from early Islamic history. See also Savant, New Muslims, 102; and Savran, “Eloquent Tribesmen,” 90. 27. See al-​Baladhuri, Ansab al-​Ashraf, 2:419–​421; al-​Tabari, Tarikh, III: 211–​213. 28. al-​Baladhuri, Ansab al-​Ashraf, 2:422; al-​Tabari, Tarikh, III:212–​213. 29. Or at least, this was their eventual stance. The Abbasids began as a proto-​Shi‘i movement themselves, claiming descent from Ali’s son Muhammad ibn al-​Hanafiyya, whose mother was a slave. On the origins of Abbasid doctrine, see particularly Sharon, Black Banners from the East. 30. The Umayyads are generally remembered for valuing Arabs over non-​Arabs, forcing non-​ Arab Muslims to pay unjust taxes, discriminating against the children of slave women, and exhibiting other forms of un-​Islamic ethnic chauvinism. However, the sources that present this ugly picture are usually hostile to the Umayyads, fail to represent accurately the Umayyad point of view, and lack the perspective that the Umayyads were operating before the elaboration of many “classical” Islamic norms and are thus being judged in the unforgiving light of hindsight. 31. Works of genealogical literature, known in Arabic as nasab literature, are essentially giant family trees written in narrative form. 32. A notable exception being Ibn Habib’s Ummahat al-​Nabi. 33. Conte particularly emphasizes that we should not be misled by the patrilineal organization of genealogical works into thinking that maternal kinship ties do not matter for day-​to-​day social life, in both historical and contemporary settings. Conte, “Agnatic Illusions,” 18. 34. For instance, Varisco has noted the discrepancy between the elaborate genealogies of the Prophet Muhammad and the social reality that Muhammad faced as an orphan who fell primarily through the cracks of the tribal support system. The genealogy of Muhammad was used to justify the political claims of his relatives—​the Abbasids and Alids—​rather than to reflect the reality of the kinship ties that operated for him. Varisco, “Metaphors and Sacred History.” 35. Szombathy, “Genealogy in Medieval Muslim Societies,” 10–​12. Indeed, the old tribal registries do seem to have recorded the clients of a tribe, unlike the classical Arabic genealogical works. See, for instance, al-​Baladhuri, Ansab al-​Ashraf, 3:105: We were in Kufa with Muhammad ibn Sulayman, and he asked about Ibrahim al-​Nakhai (an Umayyad-​era scholar, d. 96/​717), whether he was an Arab or a client. They disagreed about the answer, so they sent for the genealogists of al-​Nakha, and they brought him their registry, and he found him in the registry as a client. 36. I derive my idea about the simplification of Arabism from Szombathy, who says that the science of genealogy represents “a highly simplified pattern with which extremely variegated social facts and relationships can be explained in a manner understandable to all.” Szombathy, “Genealogy in Medieval Muslim Societies,” 11. 37. I wish to clarify that the devaluation of motherhood I am talking about here is political—​in other words, that the political ascendancy of the children of slave mothers was justified by devaluing the genetic contribution of mothers. I do not mean to imply that mothers had no actual influence on their children’s identities. Surely, mothers always transmitted some of their customs and values—​not to mention their DNA—​to their children. See for instance Ruggles, “Mothers of a Hybrid Dynasty.” 38.  As Joseph Schacht has previously noted (“Umm al-​Walad”), Mariya the Copt does not appear as a “proof text” in the legal debates about umm walads; nor, it seems, does Hajar.

242 i

Elizabeth Urban

Bibliography Primary Sources al-​Baladhuri, Ahmad ibn Yahya. Ansab al-​Ashraf. Edited by Mahmud al-​Fardaws al-​Azm. Damascus: Dar al-​Yaqazah al-​Arabiyah, 1997–​. —​—​—​. Liber expugnationis regionum (= Futuh al-​Buldan). Edited by M. J. de Goeje. Leiden: Brill, 1968. Ibn Abd al-​Hakam, Abu al-​Qasim Abd al-​Rahman. Futuh Misr wa-​Akhbaruha. Baghdad: Maktabat al-​Muthanna, 1968. Ibn Asakir, Ali ibn al-​Hasan. Tarikh Madinat Dimashq. 80 vols. Edited by Umar ibn Gharamah al-​Amrawi. Beirut: Dar al-​Fikr, 1995–​2000. Ibn Sa`d, Muhammad. Biographien Muhammeds, seiner Gefährten und der späteren Träger des Islams, bis zum Jahre 230 der Flucht (= al-​Tabaqat al-​Kubra). 9 vols. Edited by Eduard Sachau. Leiden: Brill, 1904–​1940. Muslim ibn al-​Hajjaj al-​Qushayri. “Sahih Muslim.” In Jam` Jawami al-​Ahadith wa’l—​Asanid, vol. 4. Vaduz: Jamiyyat al-​Maknaz al-​Islami, 2000–​2001. al-​Tabari, Muhammad ibn Jarir. Annales quos scripsit Abu Djafar Mohammed ibn Djarir al-​Tabari (= Tarikh al-​Rusul wa-​al-​Muluk). Edited by M. J. De Goeje. Leiden: Brill, 1964–​1965. al-​Usfuri, Khalifa ibn Khayyat. Tarikh Khalifa ibn Khayyat. 2nd ed. Edited by Akram Diya al-​ Umari. Damascus: Dar al-​Qalam, 1977. al-​Yaqubi, Ahmad ibn Abi Yaqub. Tarikh Ahmad ibn Abi Yaqub ibn Jafar ibn Wahab ibn Wadih al-​Maruf bi-​al-​Yaqubi. 2 vols. Edited by M. T. Houtsma. Leiden: Brill, 1969.

Secondary Sources Amir-​Moezzi, Mohammed Ali. The Divine Guide in Early Shiʻism: The Sources of Esotericism in Islam. Translated by David Streight. Albany: State University of New York Press, 1994. —​—​—​. “Šahrbanu.” In Encyclopaedia Iranica, Ihsan Yarsheter, gen. ed. New York: Encyclopaedia Iranica Foundation, 1982–​(or online at iranicaonline.org). Athamina, Khalil. “How Did Islam Contribute to Change the Legal Status of Women: The Case of the Jawari or Female Slaves.” Al-​Qantara 28, no. 2 (2007): 383–​408. Borrut, Antoine. Entre mémoire et pouvoir: L’Espace syrien sous les derniers Omeyyades et les premiers Abbassides (v. 72–​193/​692–​809). Leiden: Brill, 2011. Brockopp, Jonathan. Early Maliki Law:  Ibn ‘Abd al-​Hakam and His Major Compendium of Jurisprudence. Leiden: Brill, 2000. Cannuyer, Christian. “Mariya, la concubine copte de Muhammad:  Réalité ou mythe?” Acta Orientalia Belgica 21 (2008): 251–​264. Conte, Édouard. “Agnatic Illusions:  The Element of Choice in Arab Kinship.” In Tribes and Power: Nationalism and Ethnicity in the Middle East, edited by Faleh Abdul-​Jabbar and Hosham Dawod. London: Saqi, 2002: 15–​49. Crone, Patricia. Slaves on Horses:  The Evolution of the Islamic Polity. Cambridge:  Cambridge University Press, 1980. Crone, Patricia, and Martin Hinds. God’s Caliph:  Religious Authority in the First Centuries of Islam. Cambridge: Cambridge University Press, 1986.



Hagar and Mariya

j 243

Dakake, Maria Massi. The Charismatic Community: Shi‘ite Identity in Early Islam. Albany: State University of New York Press, 2007. Encyclopaedia Iranica. Ihsan Yarsheter, gen. ed. New  York:  Encyclopaedia Iranica Foundation, 1982 (or online at iranicaonline.org). Encyclopaedia of Islam. 1st ed. Edited by M. T. Houtsma et al. Leiden: Brill Publishing Online, 2012. Encyclopaedia of Islam. 2nd ed. Edited by P. Bearman et al. Leiden: Brill Publishing Online, 2012. Firestone, Reuven. Journeys in Holy Lands:  The Evolution of the Abraham–​Ishmael Legends in Islamic Exegesis. Albany: State University of New York Press, 1990. Hidayatullah, Aysha. “Mariyya the Copt: Gender, Sex and Heritage in the Legacy of Muhammad’s umm walad.” Islam and Christian–​Muslim Relations 21, no. 3 (2010): 221–​243. Isom-​Verhaaren, Christine. “Royal French Women in the Ottoman Sultans’ Harem: The Political Uses of Fabricated Accounts from the Sixteenth to the Twenty-​First Century.” Journal of World History 17, no. 2 ( June 2006): 159–​196. Lammens, Henri. “Maslama.” In Encyclopaedia of Islam, 1st ed., edited by M. T. Houtsma et al. Brill Publishing Online, 2012. Lindsay, James E. “Sarah and Hagar in Ibn ‘Asakir’s History of Damascus.” Medieval Encounters 14, no. 1 (2008): 1–​14. Marmon, Shaun. “Concubinage, Islamic.” In Dictionary of the Middle Ages, Joseph R. Strayer, gen. ed. New York: Charles Scribner and Sons, 1982–​1987. Öhrnberg, K. “Mariya al-​Qibtiyya Unveiled.” Studia Orientalia 55 (1983–​1984): 295–​303. Powers, David S. Muhammad Is Not the Father of Any of Your Men:  The Making of the Last Prophet. Philadelphia: University of Pennsylvania Press, 2009. Ruggles, D. Fairchild. “Mothers of a Hybrid Dynasty: Race, Genealogy, and Acculturation in al-​ Andalus.” Journal of Medieval and Early Modern Studies 34, no. 1 (2004): 65–​94. Savant, Sarah Bowen. The New Muslims of Post-​Conquest Iran: Tradition, Memory, and Conversion. Cambridge: Cambridge University Press, 2013. Savran, Scott. “Eloquent Tribesmen, Dignified Sheikhs, and Pompous Kings: Conceptualizing Early Islamic Historical Accounts of Arab–​Sasanian Encounters in the Context of the ‘Abbasid High Culture.” Unpublished PhD diss., University of Wisconsin-​Madison, 2011. Schacht, Joseph. “Umm al-​Walad.” In Encyclopaedia of Islam, 2nd ed., edited by P. Bearman et al. Leiden: Brill Publishing Online, 2012. Sharon, Moshe. Black Banners from the East: The Establishment of the ‘Abbasid State. Jerusalem: Magnes Press, Hebrew University, 1983. Stowasser, Barbara F. Women in the Quran, Traditions, and Interpretation. New  York:  Oxford University Press, 1994. Szombathy, Zoltán. “Genealogy in Medieval Muslim Societies.” Studia Islamica 95 (2002): 5–​35. Varisco, Daniel Martin. “Metaphors and Sacred History: The Genealogy of Muhammad and the Arab ‘Tribe.’” Anthropological Quarterly 68, no. 3 (1995): 139–​156.

12 Between History and Hagiography The Mothers of the Imams in Imami Historical Memory

Michael Dann

i  When the 11th of the 12 Imams recognized by Imami Shi‘is Hasan al-​Askari, died in 260/​874 without any publicly known offspring, the Imami community was faced with a crisis of unprecedented magnitude. The aftermath of the crisis precipitated a profound transition in the history of Imami Shi‘ism and was thus preserved in the community’s historical memory. Imami authors of the fourth/​tenth and fifth/​eleventh centuries recorded a large number of reports regarding Hasan al-​Askari’s death and the ensuing controversies over his estate, his succession, and especially the possibility of his having a son. Hasan’s mother, an umm walad1 named Hudayth, and one of his concubines, the purported mother of his son, played a central part in these developments. This first part of this chapter sifts through a variety of polemical accounts regarding these events to uncover the exceptional historical role played by these two women. The second part analyzes the construction of an extensive hagiographical portrait that would come to define popular memory of Hasan’s concubine in particular. I conclude that the phenomenon of dynastic succession that characterized Imami Shi‘ism and Abbasid imperial practice alike was the central determining factor both in the emergence of the two women onto the stage of history and the construction of the elaborate hagiographical tradition that came to surround them. The Historiographical Context Scholarship on the mothers of the Imams has been focused primarily on either the historical or literary roles they have played in the development of Imami Shi‘ism. Studies 244



Between History and Hagiography

j 245

of the aftermath of Hasan al-​Askari’s death have, necessarily, given ample attention to the role of his mother and his concubine, carefully comparing conflicting accounts to establish a probable outline of events.2 With respect to the mothers of previous Imams, it is generally difficult to place them in a historical event other than the birth of their sons with any confidence. Studies of these women, and especially of the figure of Shahrbanu, the concubine of Husayn (d. 61/​680) and mother of Ali Zayn al-​Abidin (d. 95/​713), have accordingly treated them primarily as literary figures, tracing the emergence and development of different themes in the hagiographic portrayals of them produced by a diverse and evolving Imami community.3 Although these women certainly cannot be understood outside the context of Imami Shi‘ism, there is much to recommend expanding this context by viewing their lives against the wider backdrop of the practice of female slavery in the early Arabo-​Islamic Empire. The influx of slaves of foreign origin was one of the most fundamental dynamics that shaped the sociopolitical evolution of this polity, but historical investigations of the slaves’ variegated roles pale in comparison with their significance. This is especially the case with female slaves, whose avenues to social distinction were greatly restricted in comparison with those available to their male counterparts. In particular, the male counterparts of the figure of the umm walad, corresponding roughly to those non-​Arab freedmen who became clients of their former masters (mawali), availed themselves of the limited social latitude that they enjoyed during the late Umayyad period and emerged as significant individual actors in at least two respects. They distinguished themselves as transmitters and interpreters of religious knowledge, becoming major authorities in newly emergent traditionalist milieus, and as a political force to be reckoned with, undertaking military exploits in the service of both revolutionary movements and the Umayyad dynasty itself.4 Female slaves, on the other hand, are known to us in the literary traditions concerned with the Umayyad period almost exclusively as undifferentiated stock characters.5 The fate of concubines could be deployed as a barometer of justice or inequity in political polemics, and the collective status of their upstart children was a source of great consternation to those who wished to preserve unadulterated Arab purity and supremacy.6 With the transition from the Umayyads to the Abbasids, the upward swell of subaltern demographics thrust individual concubines unambiguously into the realm of elite politics. Whereas only the last three Umayyad caliphs were born to concubines, the great majority of the early Abbasid caliphs were sons of this heretofore nameless class of women.7 Political prominence came hand in hand with historiographical attention, and the chronicles of the period admit of full monographs dedicated to the concubines of early Abbasid caliphs.8 Another class of female slaves, trained as poetesses and musicians, was also beloved to the Abbasid literary tradition, which preserves reams of material on the qiyan. Even in the case of the qiyan, however, marriage to a caliph, or at least a close connection to the Abbasid court, was the surest means of their memory being preserved for posterity. Away from the imperial center, female slaves of the early Abbasid era still received scant attention.9

246 i

Michael Dann

A partial exception to this rule is found in the concubine mothers of the last six Imams recognized by Imami Shi‘is. The historiographical attention received by these women is unique in that it is found almost exclusively in Imami literature, as opposed to Abbasid chronicles or belles-​lettres. It is also unique among the Alids in general. Although genealogical literature suggests that a large portion of Alids during this period were born to concubines, these concubines are rarely individuated, let alone celebrated, in contrast to the mothers of the Imams.10 Yet while the question of provenance sets the mothers of the Imams apart from the mothers of caliphs, and the question of literary prominence sets them apart from the mothers of other Alids, the political landscape of the era suggests that, in all of these cases, we are dealing with different gradations of a single elite. Throughout the first century of Abbasid rule, the dynasty was haunted by the prospect of an Alid counter-​caliphate. The details of the Abbasids’ fraught relationship with Shi‘ism and the Imams cannot be pursued here, but the fact of the relationship itself is central to our discussion. In schematic terms, the Imams constituted a sort of quietist shadow dynasty, held close by the Abbasids in times of friendship and closer in times of suspicion and enmity. The fate of those concubines who became embroiled in a crisis of succession can only be understood properly in light of this dynamic. This chapter begins by analyzing four separate reports on the roles of Hasan al-​ Askari’s mother and concubine in the events following his death, with the intent of identifying the “historical kernel” on which they agree and the polemical concerns that gave rise to differences between them. I also examine what the reports reveal regarding the Imami community’s assumptions concerning the role of slaves in Hasan al-​Askari’s household. The latter portion of the chapter places the aforementioned hagiographical account regarding the mother of the 12th Imam in its intertextual context of hagiography of the mothers of previous Imams. I show that the account in question draws heavily on topoi contained in these earlier hagiographical reports to construct a biography worthy of the mother of the messianic figure of the Mahdi.11 Although these reports can hardly be relied on for historical data about these women, they are instructive in that they reveal popular assumptions about various aspects of the slave trade during the early Abbasid era. The Crisis of Succession and the Birth of the 12th Imam Although establishing a precise chronology for the emergence and consolidation of the doctrines of Imami Shi‘ism remains a subject of considerable debate among scholars, it is certain that well before the middle of the third/​ninth century, Imamis believed that legitimate leadership of the Muslim community after Muhammad was passed from one Imam to another by explicit designation (nass or wasiyya) in a line of patrilineal succession extending back to Husayn ibn Ali.12 Both Hasan ibn Ali (d. 50/​670) and his younger brother Husayn had been Imams, but the Imamate was restricted to patrilineal



Between History and Hagiography

j 247

succession in the line of Husayn after his death. Despite the theoretically indisputable mechanism of explicit designation, succession to the Imamate did not proceed without considerable controversy. Unexpected deaths of appointees, their occasional youth at the time of their succession, competition among potential successors, differing expectations regarding the proper role and qualifications of the Imam, and beliefs in reincarnation and occultation that circulated in Shi‘i milieus from the first/​seventh century onward all posed challenges to a smooth transition from one Imam to another.13 The succession of Hasan al-​Askari was particularly controversial for a number of reasons. In addition to doubts regarding his qualification for the office raised by his conduct and apparently limited mastery of religious knowledge, his succession was troubled by his brother Jafar’s competing claim to the Imamate, which was supported by influential agents of their father, Ali al-​Hadi (d. 254/​868), who had fallen out with al-​Hadi during his own lifetime.14 The full implications of the two brothers’ competing claims would only take shape in the wake of Hasan’s death without any publicly known offspring—​an event that would pose a greater challenge to the integrity of the Imami community than any before. Given that the majority of the community had rejected Jafar’s claim to the Imamate for that of Hasan, the latter’s death without any known offspring threatened to leave the community without an Imam for the first time. Imami doctrine held the validity of the individual Muslim’s faith and the religious guidance of the community to be entirely dependent on the existence and recognition of the Imam. So essential was he to the fabric of human life that, if only two men remained on the earth, one of them would have to be the Imam.15 The theological centrality of the office of the Imam was by no means divorced from substantial worldly interests, both actual and potential. The Imams presided over a vast financial network through which the charitable contributions of pious Shi‘is flowed to them. Although the network began informally during the Imamate of Jafar al-​Sadiq (d. 148/​765), its administration became increasingly sophisticated and precise over the course of the following century. By the middle of the third/​ninth century, the network was managed by numerous local agents who collected an increasing variety of charitable donations and taxes on behalf of the Imams, who accrued and disbursed considerable sums.16 In addition to such concrete interests, the Imami community was also animated by the typical, simmering Shi‘is expectations of a millenarian triumph over the usurpers of power that belonged rightfully to the Alids. However, unlike many of their Alid relatives, the Imams largely preached quietism and never leveraged their extensive following to launch a revolutionary effort. This quietism no doubt placated the Abbasids, but did not dispel the need for prudence altogether; Hasan acquired his nickname “al-​Askari” (“of the military camp”) as a result of spending his life in a state of semicaptivity in the Abbasid military capital of Samarra after his father was summoned there by al-​Mutawakkil in 233/​848.17 In light of both the worldly and otherworldly interests attached to the person of the Imam, it

248 i

Michael Dann

comes as no surprise that Hasan’s death without an heir plunged the community into a crisis of epic proportions. According to a late third/​ninth century Imami heresiographer, the community split into 14 sects after Hasan’s death, two of which believed that Hasan would come back to life or had not died at all, and four of which believed in Jafar’s Imamate.18 The various responses to the problem of succession that arose in the wake of Hasan’s death left an indelible mark on the widely differing and often contradictory reports regarding the events that followed. These reports are preserved in Imami sources from the fourth/​tenth and fifth/eleventh centuries, but their origins likely stem from the turbulent 70-​year period between Hasan’s death and the community’s adoption of the doctrine of the “greater occultation” of the 12th Imam.19 Despite the difficulties associated with these reports, they concur in their affirmation of a conflict between Hasan’s mother, Hudayth, and his brother Jafar after Hasan’s death. They are also unanimous in their recognition that one of Hasan’s concubines was involved in the conflict due to the possibility of her having been pregnant, or of having given birth to a son before his death. The greatest amount of detail regarding these events is contained in four separate reports. Report 1 Both al-​Kulayni (d. 329/​939) and Ibn Babawayh (d. 381/​991) quote an almost identical report attributed by separate authorities to Ahmad ibn Ubayd Allah ibn Yahya ibn Khaqan, an Abbasid tax official in Qumm, whose father had been a vizier under both al-​ Mutawakkil (r. 232–​247/​847–​861) and al-​Mutamid (r. 256–​279/​870–​892).20 According to the report, al-​Mutamid sent Ubayd Allah with a retinue of slaves,21 medical specialists and certified witnesses to keep watch over Hasan. Upon his death, the governor of Samarra initiated a relentless search for any children Hasan might have had, and sent midwives to inspect his female slaves for pregnancy. One of them believed a certain slave to be pregnant and she was put under the charge of one of al-​Mutamid’s chief slaves. The report describes the prayer recited over Hasan’s bier as a momentous event in Samarra. It was performed by an Abbasid prince, Abu Isa ibn al-​Mutawakkil, who uncovered Hasan’s face, showed him to “the Banu Hashim from both the Alids and the Abbasids, the commanders, secretaries, judges, jurists and court officials and said, ‘This is Hasan ibn Ali ibn Muhammad ibn al-​Rida. He died of natural causes on his bed.’ ” After describing the funeral prayer, the report goes on to describe the controversy that followed Hasan’s death: When he was buried and the people had dispersed, the governor and his men busied themselves with searching for his child. There was much searching in homes. They suspended the distribution of his inheritance and those who were put in charge of watching over the slave-​girl whom they had imagined to be pregnant remained



Between History and Hagiography

j 249

with her for two years and more, until it became clear to them that she was not pregnant. So his inheritance was divided between his mother and his brother Jafar. His mother claimed to have received a bequest from him and the judge confirmed this, but the governor continued to search for his child.22

Report 2 Ibn Babawayh provides a second report in which Hasan al-​Askari prophesied his death to one of his messengers and informed him of the signs by which he might recognize the next Imam. He would be the individual leading the funeral prayer and would demonstrate knowledge of the unseen by knowing the contents of a certain purse. The messenger returned to Hasan’s house after a journey to find that he had died, and saw the Shi‘is offering their condolences to Jafar, his brother and rival. When Hasan’s slave, Aqid, invited Jafar to perform the funeral prayer, a young boy emerged from the house, ordering Jafar to step aside on grounds that he had a greater right to perform the funeral prayer. Upon the completion of the last rites, a delegation arriving from Qumm, a city in which Imami Shi‘ism predominated from an early date, asked Jafar to inform them of the contents of a purse they were carrying, upon which Jafar balked at the notion that he should know the unseen. The boy, however, knew the exact contents of the purse, leading the messenger to recognize him as the next Imam. The messenger proceeded to describe Jafar’s recourse to the Abbasids and the ordeal that followed: Jafar entered upon al-​Mutamid and informed him of [the existence of a son], so al-​Mutamid sent his slaves and they detained the slave girl Saqil and demanded the boy from her. She denied the boy and claimed that she was pregnant in order to conceal his state. She was turned over to the judge Ibn Abi al-​Shawarib.23 The sudden death of Ubayd Allah ibn Yahya ibn Khaqan24 and the revolt of the leader of the Zanj in Basra25 surprised them such that they were distracted from the slave girl and so she escaped from them.26 Report 3 Ibn Babawayh includes a third report that opens with a description of the services performed by Aqid, Hasan’s slave, and Saqil, his concubine, at the point of his death. The relevant portion of this account is as follows: Abu Muhammad’s [Hasan’s] mother, whose name is Hudayth, came from Madina to Samarra when the news [of his death] reached her. Stories that would take a long time to explain transpired between her and his brother Jafar—​he demanded

250 i

Michael Dann

his inheritance from her, brought her before the governor and revealed that which God ordered to be concealed [the existence of a son]. At that time, Saqil claimed that she was pregnant. She was brought to the palace of al-​Mutamid, and the latter’s women and slaves, as well as those of al-​Muwaffaq and the women of the judge Ibn Abi al-​Shawarib, began to watch over her constantly until the affair of al-​ Saffar,27 the sudden death of Ubayd Allah ibn Yahya ibn Khaqan, their move from Samarra,28 the affair of the leader of the Zanj in Basra, and other matters distracted them from her.29 Report 4 Husayn ibn Hamdan al-​Khasibi (d. 346/​958 or 358/​969) records a report in which a group of Shi‘is visited Hasan’s father, Ali al-​Hadi, after differing among themselves over whether Jafar or Hasan would be his successor. In this rather odd report, both al-​Hadi and an apple that he causes to speak in a thaumaturgical feat inform the confused group that Hasan is the true Imam. Of particular concern to us is the fact that al-​Hadi prophesies that “Jafar will report [Hasan’s] slave-​g irls and turn them over to the tyrant.”30 These four reports become more intelligible if they are read in light of the conflicts that occurred within the Imami community after Hasan’s death. The first report recorded by al-​Kulayni and Ibn Babawayh clearly seems to have been designed to emphasize the indubitable nature of Hasan al-​Askari’s death to those who rejected it. The presence at his deathbed of witnesses, medical experts, and government officials; the drama surrounding the funeral procession; and Abu Isa’s flat pronouncement of his death before an impressive assembly of the elite of the Abbasid court all seem designed to make denial of Hasan’s death impossible.31 Conflicting interests, however, produced conflicting reports, as can be seen clearly in the second report recorded by Ibn Babawayh, which states that Hasan’s young son and successor performed the funeral prayer in his own home without any fanfare. This report served the purpose not of convincing the reader of Hasan’s death, but of the actual existence of Hasan’s son and of Jafar’s unsuitability for the Imamate. The same polemical interests are apparent but less overt in the third report, also recorded by Ibn Babawayh, which bears a strong resemblance to the second report aside from its omission of any discussion of last rites. Although the final report, recorded by al-​Khasibi, about the confused Shi‘is who visited Ali al-​Hadi makes no mention of a son, it was also transparently intended to support Hasan’s claim to the Imamate against that of Jafar. The Women and Slaves of Hasan’s Household Whether their aim was to confirm the death of Hasan al-​Askari, support the existence of a son, or challenge the legitimacy of Jafar’s claim to the Imamate, the transmitters



Between History and Hagiography

j 251

of the reports listed above all acknowledged the dispute between Jafar and Hudayth over Hasan’s inheritance, as well as the detention of one of his concubines after it was claimed that she was pregnant.32 Beyond the kernel shared by the accounts, the chief difference among them consists in which party actually claimed that one of Hasan’s concubines was pregnant. Given the clear polemical interests and dubious historicity of the account in which the governor brings women to examine Hasan’s concubines, it seems unlikely that these women were responsible for the claim. It is also difficult to imagine how the concubine’s own claim of pregnancy could have effectively concealed the state of her already-​born son, as suggested by the account ascribed to one of Hasan’s messengers. Modarressi offers the most plausible interpretation: For Jafar to make such a claim, as al-​Khasibi’s account suggests,33 would have in fact been inimical or irrelevant to his own interests. According to Sunni law, Jafar would only have had a right to a share of Hasan’s inheritance if Hasan did not have any children, whereas according to Shi‘i law, Jafar had no right to a share in Hasan’s inheritance whether Hasan had a son or not. Thus, the most plausible candidate for advancing the claim of pregnancy was, in fact, Hasan’s mother Hudayth, who in doing so may have intended to prevent Jafar from receiving any share of Hasan’s inheritance. Ironically, this version of events is recorded only in an anonymous manuscript dating from the fifth/​eleventh century.34 It is supported, however, by the reports indicating that Hasan’s inheritance was divided between Hudayth and Jafar only after it became certain that the slave girl was not pregnant.35 Regardless of whether it was, in fact, Hudayth who first claimed that one of Hasan’s concubines was pregnant, her dispute with Jafar demonstrates that she, an umm walad whose master was dead, enjoyed a degree of prominence and influence. According to the third report listed above, she traveled from Madina to Samarra upon hearing the news of Hasan’s death—​a journey that would have likely required a retinue and considerable expenditures. Her ability to travel long distances and her presence in the Hijaz at the time of Hasan’s death is also supported by a report in which Hasan orders her to bring his son along with her for the pilgrimage the year before his death.36 In a report recorded by al-​Khasibi, she provides generously for Jafar’s large family after he has descended into poverty.37 Although the report is likely a piece of anti-​Jafar polemic, it nonetheless may preserve a genuine memory of Hudayth’s wealth. In addition, although the Abbasids’ direct involvement in Hudayth’s dispute with Jafar almost certainly owed more to political than legal concerns, the fact that she was the principal representative of one side in a dispute that garnered the personal attention of several important officials is indicative of the prominence that accrued to her as the mother of an Imam. Hasan’s concubine Saqil, on the other hand, does not appear to have exercised any significant agency in the process of her detention—​until perhaps when she escaped—​and the contrast between her abject dependency on the will of other parties and the agency of Hudayth in the same episode could hardly be sharper. This is in conformity with the

252 i

Michael Dann

general patterns of social mobility for female slaves in early Abbasid society. A woman of slave origins could gain considerable power and influence in elite circles, but generally only after progressing from slavery to freedom (whether through the birth of her owner’s child or through manumission) and living long enough to accrue the necessary social and material capital.38 Khaizuran (d. 173/​789) and Shaghab (d. 321/​933), both concubine mothers of Abbasid caliphs, eventually exercised considerable authority at court and intervened in factional politics on behalf of their sons.39 Arib al-​Mamuniyya (d. 277/​ 890), perhaps the most famous qayna of the Abbasid era, is portrayed as triumphantly displaying her own cohort of qiyan at court competitions in her golden years.40 Hudayth may even have acted as a sort of spiritual regent, alongside the more official figure of Uthman b.  Said al-​Amri, after Hasan’s death. According to a report recorded by Ibn Babawayh and al-​Khasibi, when a confused follower approached Hasan’s aunt Hakima two years after Hasan’s death, she named Hudayth as the principal recourse for the Shi‘is in Hasan’s absence.41 Hakima herself seems to have occupied an important position within the community after Hasan’s death, as demonstrated by the numerous reports attributed to her regarding the birth of Muhammad al-​Mahdi, the 12th Imam. Ibn Babawayh, al-​Khasibi, and al-​Masudi all record lengthy reports from Hakima regarding the birth of Hasan’s son. In all of the reports, Hasan displays a miraculous foreknowledge that the Mahdi will be born to a certain slave girl. According to several reports, this slave girl in fact belonged to Hakima, who offered her to Hasan after she noticed him looking at her. Hasan accepted the offer after explaining that he was staring at her because of his knowledge that God had chosen her to be the mother of the Mahdi. According to other reports, the slave girl, in fact, originally belonged to Hasan. Three reports give the slave girl’s name as Narjis, while the other two give it as Saqil (the name of the concubine detained in previous accounts). The reports recount the child’s birth and the initial interactions between father and child with a heavy emphasis on the miraculous qualities of the Imams. They conclude by offering varying explanations of the child’s occultation.42 Notably, in one of the accounts recorded by Ibn Babawayh, Hakima states that although the Shi‘is had become divided, she was still in communication with the Imam, indicating her stature within the community.43 Hakima was not the only member of Hasan’s household to be accorded the privilege of seeing the Mahdi as an infant. Accounts of events in his infancy are also attributed to a number of Hasan’s slaves. At the end of one of the reports attributed to Hakima, the narrator states that he verified her story with a slave named Uqba.44 Although there is no way to ascertain the number of Hasan’s slaves or their names, it seems likely that “Uqba” is to be read as “Aqid,” who is mentioned in a number of other accounts.45 In one such account, Aqid reports the date of the Mahdi’s birth and gives his mother’s name as Saqil.46 Other reports in which slaves of the household claim to have seen the child are attributed to two slave girls named Mariya and Nasim.47 Yet another report is attributed to an unnamed slave girl who was gifted to Hasan and fled from his house to her former owner when



Between History and Hagiography

j 253

Jafar instigated an Abbasid intervention. She married her former owner and informed him of the birth of Hasan’s son to a slave girl named Saqil. In this account, however, Saqil died before Hasan.48 Although none of these reports is reliable at the individual level, taken together they are informative. The relatively large number of reports attributed to Aqid, the important roles attributed to him in the events surrounding Hasan’s death, and his ability to corroborate others’ stories all suggest that he was the most prominent slave in the household and that he was held in some regard by the Imami community after Hasan’s death.49 Beyond Aqid, nothing can be said regarding the role or identity of other slaves to whom reports regarding the birth of the 12th Imam are attributed, but the fact that slaves constitute the majority of those credited with having seen the young Mahdi is significant in itself. Whether or not they actually narrated anything on the subject, the Imami community likely viewed them as individuals who would have had privileged access to events in Hasan’s household and were thus important sources of information regarding family affairs. The Hagiography of an Umm Walad Based on the accounts examined thus far, the mother of the 12th Imam is still surrounded by ambiguity. Neither her name nor her place of origin is clear.50 Nor is it clear whether she was the same slave girl who was implicated in the conflict between Hudayth and Jafar, or even whether she survived Hasan al-​Askari himself. One account in particular, however, surpassed all others in shaping the image of the Mahdi’s mother that would become enshrined in popular memory. This romantic account, first recorded by Ibn Babawayh, is attributed to a certain Bishr ibn Sulayman (likely a fictional character),51 who is identified as a slave dealer and a descendant of the Companion Abu Ayyub al-​Ansari (d. 52/​672).52 Bishr states that Ali al-​Hadi, who had instructed him regarding commerce in slaves, sent for him one day with detailed instructions on purchasing a slave girl on his behalf. Al-​Hadi provided him with a letter in a “Rumi” language (presumably Greek) and 220 dinars, telling him to go to a location in Baghdad where agents of the Abbasid officers and small groups of young men awaited the unloading of prisoners from boats. He tells Bishr that when he sees a slave trader named Umar ibn Yazid, he should observe from afar until a slave girl of a certain description emerges to be shown to potential buyers. Al-​Hadi informs him that “she will be wearing two thick pieces of silk and will refuse to unveil her face or submit to anyone who tries to touch her or busy his gaze with admiring that which is exposed of her behind the thin veil. The slave dealer will strike her and she will respond, shouting in Rumi. Know that she is saying, ‘Oh, my honor is violated!’ One of the buyers will say, [Sell her] to me for 300 dinars, for her chastity has increased my desire for her.”53 The girl rebukes the man, insulting him in Arabic. The frustrated slave dealer asks what he is to do, since he must sell her, to which she replies that she must be confident in the religious commitment and trustworthiness of her buyer.

254 i

Michael Dann

At this, Bishr is instructed to present her with the letter. Upon reading the letter, the girl cries bitterly and threatens suicide if the slave dealer will not sell her to its author. After bargaining, he agrees to sell her to Bishr for 220 dinars. On the way to Bishr’s residence in Baghdad, the slave girl kisses the letter, places it on her eyes and wipes it on her body. Bishr, aghast at these antics, expresses his disapproval, for which Narjis in turn rebukes him, informing him that she is the granddaughter of the Byzantine emperor and the daughter of a woman descended from Peter, “the delegate (wasi) of the Messiah.” She tells him that she was betrothed to her grandfather’s nephew, and that the elite of the Byzantine Empire were assembled for her marriage ceremony. As the ceremony was about to begin, crosses crashed to the ground, pillars crumbled, and the prince careened to the floor from his seat atop a raised throne. The Byzantine priests recognized these portents as signaling the destruction of the Christian religion and begged the emperor to desist from the marriage, but he insisted stubbornly, only to have the same events repeat themselves. That night, the young girl had a dream in which Muhammad approached Jesus to ask for her hand on behalf of his descendant, Hasan al-​Askari. After this dream, she was so overtaken by the love of Hasan that she could no longer eat or drink and became extremely ill. When her grandfather asked what might aid in her cure, she instructed him to release all the Muslim prisoners. When he granted the request, she “tried hard” to show that she was well, although she was still lovesick. After four nights passed, Narjis had another dream, this time of Fatima, in which she complained that Hasan al-​Askari would not visit her. Fatima informed her that he would not visit her as long as she remained a Christian and instructed her to say the Muslim testimony of faith (shahada), after which she would send him to visit. Nightly dream visits began in due course, and in one of them Hasan instructed her to disguise herself as a servant and follow the slave girls of her grandfather’s army. She did so only to be captured by Muslim troops, which was how Bishr found her. When Bishr finally brings her to Ali al-​Hadi, al-​Hadi asks her how God showed her the might of Islam and the ignominy of Christianity, to which she replies, “How should I explain to you, O son of the Messenger of God, that which you already know more about than me?” After informing her that she is to marry Hasan, al-​Hadi instructs his slave Kafur to summon none other than his sister Hakima, who is to take Narjis to her house and teach her the basic rituals of worship.54 Of all of the accounts examined thus far, the latter is perhaps the least useful in determining anything concrete about Hasan al-​Askari’s concubine—​that is, the woman to whom a child was attributed. Nonetheless, it provides a fascinating example of how established topoi in Shi‘i hagiographical literature were employed in constructing the biography of the Mahdi’s mother. Royal lineage had already been ascribed to the mother of the fourth Imam, Ali Zayn al-​Abidin, whose mother was eventually claimed to be the daughter of the Sasanid emperor Yazdagird III.55 A report recorded by al-​Saffar (d. 290/​902) and al-​Kulayni regarding her origins displays a



Between History and Hagiography

j 255

sufficient number of similarities to Ibn Babawayh’s report on Narjis’s origins to suggest a partial dependence of the latter on the former. Like Narjis, she is reported to have covered her face when the Caliph Umar looked at her and to have insulted him in Persian. She also ultimately chose among her potential buyers at the suggestion of Ali ibn Abi Talib, selecting his son Husayn by putting her hand on his head.56 The mothers of both Musa al-​Kazim (d. 183/​799) and Ali al-​Rida (d. 203/​817) were also said to have been from “the nobles of the non-​Arabs,”57 and the Nubian mother of the ninth Imam, Muhammad al-​Jawad (d. 220/​835), was said to have been related to Mariya the Copt, one of the Prophet’s concubines.58 The nearly ubiquitous desire to lay claim to matrilineal nobility for the Imams seems to have been relatively anomalous and, in all likelihood, stemmed from the superhuman conception of the Imams that became dominant over time. Although descent from non-​Arab slave girls was no longer an impediment to social standing by the advent of the third/​ninth century, and isolated examples of claiming noble origins for concubines exist in later periods, I am unaware of any parallel example in which noble origins are ascribed to concubine mothers with such frequency.59 Even in the case of Hasan al-​Askari, who is not reported to have married at all, some remembered the mother of his son as the sister of the Zaydi activist, Muhammad ibn Zayd, who ruled in Tabaristan in the latter half of the third/​ninth century.60 Another hagiographical theme of Ibn Babawayh’s account of Narjis’s origins concerns the Imams’ miraculous foreknowledge of their concubines, or those of their sons, and the order of their purchase. This theme also finds its first expression in a report recorded by al-​Kulayni, in which Muhammad al-​Baqir (d. 114/​732) informed a visitor that a Berber slave dealer would come to Madina and that he would buy from him a slave girl for his son, Jafar al-​Sadiq. When the slave dealer arrived, al-​Baqir gave the man a sack of money and told him to purchase the slave girl, which he proceeded to do, and found that the sack contained precisely the slave dealer’s asking price. The young woman was then brought to al-​Baqir, and when he learned from her that she was a virgin, he expressed his astonishment because “nothing falls into the hands of the slave dealers except that they spoil it.” She explained that whenever her captor approached her, a man with white hair and a white beard would beat him until he retreated.61 Al-​Kulayni also records that Musa al-​Kazim ordered someone to buy a slave girl on his behalf. When the man did so reluctantly, the slave dealer informed him that when he bought this girl in the Maghreb, a woman from the People of the Book told him that she should only belong to “the best of the people of the earth” and that she would be the mother of a child to whom the east and west would submit.62 The ninth Imam, Muhammad al-​Jawad, also informed someone that a caravan containing slave girls would be approaching and ordered him to purchase one of a certain description on his behalf.63 The final topos of Ibn Babawayh’s account of Narjis’s origins is that of the divine benediction conferred on events by dreams. Musa al-​Kazim was reported to have

256 i

Michael Dann

had a dream of his concubine Takattum before purchasing her, similar to the dream the Prophet Muhammad reportedly had of his wife Aisha before their betrothal.64 Alternatively, Musa’s mother, Humayda, was also reported to have had a dream in which the Prophet ordered her to give Takattum to Musa as a gift.65 Shahrbanu, the mother of Ali Zayn al-​Abidin, was said to have received a marriage proposal from the Prophet on behalf of his grandson Husayn in a dream and to have converted at the hand of Fatima the following night, exactly as in Ibn Babawayh’s report about Narjis. Fatima informed her that the Muslims would prevail in battle but that she would not be harmed.66 Although these hagiographic accounts are certainly dominated by a desire to confer the best maternal pedigree possible on the Imams, they are also inflected with certain mundane social realities of slavery, some of them quite grim, inasmuch as these were assumed in the cultural production of inhabitants of the early Abbasid Empire. For example, in their descriptions of the relationship between the Imams and their concubines, the foregoing accounts occasionally replace the terminology of concubinage with that of marriage. Shahrbanu is freed at the suggestion of Ali ibn Abi Talib and given the opportunity to choose among her “suitors.” Although Narjis is clearly “purchased,” she is purchased for a “husband” rather than a master and is “proposed to” in her dreams. This shift derives undoubtedly from the same impulse that caused royal origins to be ascribed to these women in the first place. If the mothers of the Imams are to be romanticized, so, too, should the relationships that produced them. The conceptual slippage involved may also have been facilitated by a certain level of ambiguity and overlap in the legal and social statuses of concubines—​umm walads in particular—​and freeborn wives. Although the legal differences entailed by free and servile statuses were undoubtedly of great consequence, there was also clearly a measure of terminological and conceptual overlap in medieval jurists’ descriptions of marriage and slavery.67 At the social level, concubines were occasionally known to outstrip freeborn wives in winning their masters’ affections, to be manumitted and married by their erstwhile masters, and even to occupy an uncertain space between slavery and freedom.68 While the extent of the legal and social parallels between wives and concubines should not be overstated, similarities did exist, and the possibility that they enabled the sort of shifts examined here cannot be easily dismissed. The sorts of transactions depicted in these hagiographical accounts also bore a strong resemblance to social practice. With the exception of the miraculous elements, patterns in the accounts that depict relatives purchasing a concubine for an Imam or gifting one to him are widely attested in other sources.69 The prices of the mothers of the Imams, ranging from 70 to 300 dinars, were well above the average price of 20 dinars, but not so high as to be unheard of, implying that these were not ordinary women.70 The accounts also include portrayals of some of the more brutal aspects of the slave trade. The miraculous deliverance of Narjis and the concubine of Jafar al-​Sadiq from harassment and rape at the



Between History and Hagiography

j 257

hands of potential buyers and slave dealers reflects an assumption that such occurrences were part of the passage from captive to slave as much as it enhances the dramatic effect of their stories. Likewise, the poorly understood and frantic exchange that took place between Narjis and her seller, in both Greek and Arabic, was, in all likelihood, reflective of the sort of stifled resistance that many newly enslaved women engaged in as they were displayed to potential buyers.71 Conclusion The lives of the mothers of the Imams as we know them today are composed of interwoven elements: memories of real events, hagiography, and the inevitable transformations that occur as a community’s recollections of its past are refracted through the challenges of the present. The lives of Hudayth and Saqil/​Narjis can be perceived only through the lens of accounts shaped irrevocably by the Imami community’s struggle to come to terms with the question of succession to the Imamate. Nonetheless, they constitute an invaluable resource for historians of female slavery in the early Abbasid era. The mothers of the Imams are among the few female slaves outside the Abbasid court who appear as individuated historical actors in the sources—​a fact best explained by reference to the corresponding centrality of dynastic succession in the Imami community. It is clear that the prominence of these women derived primarily from their sons and from their owners only to a lesser extent. As much is confirmed by the trajectory of their lives. The freedom and influence of the mother of an adult Imam, as represented by Hudayth, contrasts sharply with the abject dependence on other parties of the mother of an Imam-​to-​be, as represented by Saqil. This largely paralleled the dynamics of the Abbasid court, where the most frequent means of female slaves attaining prominence, barring the exceptional charm and talent of the elite qiyan, was to have a son attain the throne. Ironically, it was the failure of the principle of dynastic succession that ultimately resulted in a historically obscure concubine being forgotten as such and remembered as a saintly woman of both royal and apostolic lineage. The Narjis of later Imami tradition is the subject of hagiographic legend, but to fulfill its purpose, her legend had to reflect a world familiar enough to its audience to be believed and appreciated. In doing so, it not only made for an enthralling story, but portrayed, “behind a thin veil,” as it were, realities of the slave trade and slave life, in all of their harshness and complexity. Acknowledgments I am grateful to Michael Cook and Jelena Radovanonic for their comments on a draft of this chapter, and especially to Shaun Marmon, Matthew Gordon, and Kathryn Hain for reading and commenting on successive drafts.

THE TWELVE IMAMS AND THEIR CONCUBINE MOTHERS Ali b. Abi Talib (d. 40/661)

Hasan b. Ali (d. 50/670)

Husayn b. Ali (d. 61/680)--Shahrbanu/Sulafa/Ghazala

Ali Zayn al-Abidin (d. 95/713)

Muhammad al-Baqir (d. 114/732)

Jafar al-Sadiq (d. 148/765)--------------Hamida

Musa al-Kazim (d. 183/779)---------------Takattum

Ali al-Rida (d. 203/817)-------------Hadith/Susan

Muhammad al-Jawad (d. 220/835)------Samana/Susan

Ali al-Hadi (d. 254/868)-------------------Hudayth

Jafar ibn Ali

Hasan al-Askari (d. 260/874)-----Saqil/Narjis

Muhammad al-Mahdi

Figure 12.1  The 12 Imams and those of their concubines who gave birth to the next Imam.



Between History and Hagiography

j 259

Notes 1. On the umm walad, see Brockopp, Early Maliki Law, 192–​203. On the differences in Imami law see Schacht, Origins, 264–​265. 2. For studies in this vein see Modarressi, Crisis and Consolidation, 75–​82; Hussain, Occultation, 67–​78. See also the excellent study of Edmund Hayes, which offers a penetrating analysis but unfortunately appeared too late to be incorporated effectively into this chapter: Hayes, “Envoys of the Hidden Imam,” 202–​248. 3.  For Shahrbanu see Amir-​Moezzi, “Shahrbanu, Lady of the Land,” 45–​100, esp.  45–​83; Pinault, “Zaynab Bint Ali,” 80–​82, 95–​96; Savant, The New Muslims, 102–​108. For the mothers of other Imams, see now Pierce, Twelve Infallible Men, 134–​141. Pierce’s study also appeared too late to be fully incorporated into this chapter. 4. On the prominent role of the mawali in the transmission of religious knowledge see for example Juynboll, “Role of Non-​Arabs,” 355–​386. On the social status of the mawali during the late Umayyad period in general, with special reference to their role in military conflicts, see Urban, “Early Islamic Mawali,” 86–​137. 5. The general anonymity of concubines throughout the Umayyad period is displayed clearly in a recent prosopographical study of elite families of the early Islamic Hijaz, in which concubines are ubiquitous but rarely identified by name. See Ahmed, The Religious Elite. 6. Elizabeth Urban demonstrates that despite the anonymity of concubines, their social position can be traced through collective representations of them and their children in the literature. See Urban, “Early Islamic Mawali,” 141–​158. Given the long-​standing focus on the chauvinism displayed toward the sons of concubines, Urban places her emphasis elsewhere, but the point still comes through in many of the examples she cites. 7. For a list of mothers of caliphs during this period see Ibn Hazm, Ummahat al-​khulafa, 17–​24. Ibn Hazm lists three Umayyad caliphs born to concubines, but as noted by Urban, only one of them was universally acknowledged. Cf. Urban, “Early Islamic Mawali,” 148; Hitti, History of the Arabs, 332–​333. 8. See, for example, Abbott, Two Queens. 9. Exceptions include Husn, the concubine of Ahmad ibn Hanbal, and possibly Rabia al-​ Adawiyya, who in some accounts was said to be of slave origin. Regarding the former, a handful of reports on her purchase, her children, her piety, and her domestic chores are recorded (Ibn al-​Jawzi, Virtues of the Imam Ahmad ibn Hanbal, 2:59–​62). Rabia is a semilegendary character regarding whom there is an ample body of literature but it is of limited use for the purpose of social history (Smith, “Rabia al-​Adawiyya al-​Kaysiyya.”) 10.  On the large portion of concubines among the mothers of Alids see Scarcia Amoretti, “Historical Atlas on the Alids,” 100. 11. On the figure of the Mahdi, see, for example, Sachedina, Islamic Messianism. Imami Shi‘is eventually came to identify the 12th Imam as the Mahdi. 12.  On the difficulty of characterizing the beliefs of the community during its formative period, see the comments of Moomen, Introduction, 61–​62, 73–​75; Cooperson, Classical Arabic Biography, 75–​76; Haider, “Prayer, Mosque and Pilgrimage,” 153–​154, note 11. On the concepts of nass and wasiyya see Moomen, Introduction, 153–​155. 13. For a detailed exposition of the early history of the Imami community with emphasis on the difficulties associated with the transition from one Imam to another, see Modarressi, Crisis and Consolidation.

260 i

Michael Dann

14. Modarressi, Crisis and Consolidation, 70–​77. 15. See the reports in al-​Kulayni, Usul al-​Kafi, 1:198–​208; Moomen, Introduction, 147–​159. 16. On the financial administration of the Imamate see Modarressi, Crisis and Consolidation, 12–​15. 17. Modarressi, Crisis and Consolidation, 15; Moomen, Introduction, 44. 18. Al-Nawbakhti, Firaq al-​Shia, 96–​100, 106–​108, 112. Notwithstanding the tendency of heresiographical literature to impose strict categories on fluid and ambiguous identities, such a claim is indicative of the deep confusion that ensued in the wake of Hasan’s death. 19.  On the transition from the doctrine of “lesser occultation” to that of “greater occultation,” see Kohlberg, “From Imamiyya to Ithna-​Ashariyya,” 521–​534; Modarressi, Crisis and Consolidation, 79–​105. During the period of lesser occultation, the 12th Imam was in communication with his community through a succession of “emissaries,” who occupied a quasi-​official office in their spiritual and administrative mediation between the Imam and his community. When the last of these emissaries died, the period of “greater occultation” commenced, in which the Imam is held to be alive but inaccessible to the community until his awaited triumphant return. 20.  On Ahmad and his father Ubayd Allah, see Gordon, “Khaqanid Families,” 244–​247. Sourdel (cited by Gordon) considers Ahmad to have been a Shi‘i author on the basis of al-​Tusi’s Fihrist kutub al-​Shia. See Sourdel, Le Vizirat Abbaside, 1:283, note 5. He is, indeed, identified by al-​Tusi and al-​Najashi as an author or transmitter of a laudatory description of Hasan al-​Askari (see al-​Khui, Mujam rijal al-​hadith, 2:158–​159, and the sources cited there). However, the reports at hand describe him as fierce in his animosity toward the family of the Prophet, which would be in agreement with the pro-​Muawiya views of his brother cited by Sourdel. Cf. Ibn Babawayh, Kamal, 40; al-​Kulayni, Usul al-​Kafi, 1:574. 21. The word translated here as slaves is “khuddam” (pl., khadim), which could refer to a slave, a free servant, or, frequently, a eunuch—​the latter technically being termed khasi. On the debate over the extent to which the terms khadim and khasi should or should not be considered identical, see Daniel’s review of Eunuchs, Caliphs and Sultans: A Study in Power Relationships by David Ayalon, 345–​348. Given the ambiguity and lack of explicit evidence, I have chosen to translate khadim throughout this chapter simply as “slave,” although the fact that a young woman was placed under the guard of these particular khuddam would suggest that they may have been eunuchs. In this regard, see Marmon, Eunuchs and Sacred Boundaries, esp. 3–​23. 22. Ibn Babawayh, Kamal, 42–​43; al-​Kulayni, Usul al-​Kafi, 1:576–​577. My translation is based on the version quoted by Ibn Babawayh. 23. Al-​Hasan ibn Abi al-​Shawarib (d. 261/​875) served as a judge in Samarra under al-​Mutamid beginning in 240/​854-​5 until he died. Assuming that the chronologies we have are correct, Ibn Abi al-​Shawarib would have died before the concubine’s escape. See al-​Dhahabi, Siyar, 12:518; Waki, Akhbar al-​qudat, 3:303. 24.  According to al-​Tabari, he died in 262/​875–​6, which is in rough agreement with most authors’ account of the concubine’s confinement lasting two years. See Gordon, “The Khaqanid Families,” 246. 25. This is a reference to the revolt of African slaves in southern Iraq from 255/​869 to 270/​883, on which see Popovic, Revolt. 26. Ibn Babawayh, Kamal, 475–​476.



Between History and Hagiography

j 261

27. Misspelled in Ibn Babawayh (Kamal, 474) as “al-​sighar,” this is a reference to Yaqub b. al-​ Layth al-​Saffar, who had overthrown Abbasid vassals in Persia and briefly invaded Iraq during the period in question. See Bosworth, “Saffarids.” 28. The removal of the Abbasid army from Samarra. 29. Ibn Babawayh, Kamal, 473–​474. 30. Al-Khasibi, Al-​Hidaya, 237. Al-​Khasibi was a central figure in the development of the Nusayri sect, but the work in question preserves many Imami narrations and was dedicated to the Buyid ruler of Aleppo, Sayf al-​Dawla. See Friedman, “al-​Husayn ibn Hamdan al-​Khasibi,” 94, 97–​109. 31. A similar report is also found in al-​Khasibi, al-​Hidaya, 290, although it does not mention the conflict between Jafar and Hudayth. 32. Al-​Khasibi’s report does not refer explicitly to all the events mentioned in the other reports, but can be read as an allusion to them. The broad outlines of these events are also confirmed by later Sunni and Imami authors. See Ibn Hazm, al-​Fisal, 4:158; al-​Mufid, al-​Fusul al-​ashara, 348, 353, 357. 33. In addition to the account quoted earlier, al-​Khasibi states explicitly elsewhere that it was Jafar who made this claim (al-​Khasibi, al-​Hidaya, 182). 34. This paragraph summarizes the version of events argued for by Modarressi in Crisis and Consolidation, 78–​79. In contrast to Hudayth’s claim of a pregnant concubine, which was likely directed externally toward the Abbasids, Uthman ibn Said al-​Amri, the chief financial agent of both Hasan and his father, assured the Imami community internally of the existence of a son born already during Hasan’s lifetime. On his role see Modarressi, Crisis and Consolidation, 77–​81. 35. Ibn Babawayh, Kamal, 42–​43; al-​Kulayni, Usul al-​Kafi, 1:576–​577. 36. See al-​Masudi, Ithbat al-​wasiyya, 247–​248. There is some question regarding the attribution of this work to al-​Masudi. See Pellat, “Masudi et l’Imamisme,” 77–​90, in which Pellat concludes tentatively that the attribution is correct. 37. Al-Khasibi, al-​Hidaya, 288–​289. 38. See the comments of Gordon, “Place of Competition,” 74–​75 and esp. 76. 39. On Khaizuran, an umm walad of al-​Mahdi and mother of Harun al-​Rashid, see Abbott, Two Queens, 54–​134. On Shaghab, an umm walad of al-​Mutadid and the mother of al-​Muqtadir, see El-​Cheikh, “Caliphal Harems, Household Harems,” 92–​93. For further examples see Athamina, “How Did Islam Contribute,” 404–​408. The most vivid picture of the authority exercised by concubines in dynastic politics in an Islamicate empire comes from the Ottoman context. See Peirce, Imperial Harem, esp. 17–​18, 109–​112, 284–​285. 40. On Arib, see Gordon, “Place of Competition,” 61–​62 and esp. 65–​66. 41.  Ibn Babawayh, Kamal, 501, 507; al-​Khasibi, al-​Hidaya, 275–​276; Modarressi, Crisis and Consolidation, 78, note 126. 42.  Ibn Babawayh, Kamal, 424–​430; al-​Khasibi, al-​Hidaya, 264–​267; al-​Masudi, Ithbat al-​wasiyya, 248–​251. 43. Ibn Babawayh, Kamal, 429. 44. Ibn Babawayh, Kamal, 424. 45. See three separate reports in Ibn Babawayh, Kamal, 474–​475. 46. Ibn Babawayh, Kamal, 474–​475. 47.  Ibn Babawayh, Kamal, 430, 441; al-​Khasibi, al-​Hidaya, 267–​268. Nasim is identified explicitly as a slave girl by Ibn Babawayh, but the male form of the verbs qala and haddathana is always used by al-​Khasibi.

262 i

Michael Dann

48. Ibn Babawayh, Kamal, 431. 49. See reports 2 and 3, in which he is described as orchestrating events at Hasan’s house after his death and as having been present with Saqil at Hasan’s last moments. Aqid fits closely the profile of a eunuch (see the sources cited earlier in note 21, esp. Marmon), but in the absence of explicit evidence, it is impossible to state conclusively that he was one. 50. In addition to the names of Narjis and Saqil already encountered, Rayhana and Sawsan are also mentioned (Ibn Babawayh, Kamal, 432). None of the previous accounts mention a place of origin, although in one of them Hakima states that Narjis “spoke to her in Sindiyya,” which would imply she was from the region of Sind in the Indian subcontinent. See Khasibi, al-​Hidaya, 265. 51. Bishr b. Sulayman appears in Shi‘i sources only as the narrator of this account. The name also appears in Sunni hadith works, but chronology precludes the identification of this Bishr with that of the Sunni reports. 52. This identification bears some significance in the report. Aside from the few well-​known companions accepted by Imamis on account of their support for Ali, Abu Ayyub al-​Ansari was among those who received relatively favorable reviews in some sources. See al-​Kashshi, Ikhtiyar, 169–​190. 53. Ibn Babawayh, Kamal, 419. 54. Ibn Babawayh, Kamal, 417–​423. 55. See Amir-​Moezzi, “Shahrbanu, Lady of the Land,” 45–​51. 56. Al-Saffar, Basair al-​darajat, 355; al-​Kulayni, Usul al-​Kafi, 1:530–​531. I originally found many of the references to reports on the mothers of other Imams in al-​Shirazi, Ummahat, for which I am indebted to Nebil Husayn. I cite only the original sources wherever I have been able to trace them. 57. Al-​Shirazi, Ummahat, 246, 257. 58. Al-Kulayni, Usul al-​Kafi, 1:561. There was apparently some confusion with regard to whether Muhammad al-​Jawad was actually the son of Ali al-​Rida, because of his dark skin color. On this, see Modarressi, Crisis and Consolidation, 63. A number of reports state the mother of the Mahdi/​qaim is a black slave girl (al-​Kulayni, Usul al-​Kafi, 1:360–​361; Ibn Babawayh, Kamal, 329), which would seem to reflect an expectation that al-​Jawad should have been the Mahdi. 59. A similar claim of royal Sasanian lineage was made by the last Umayyad caliph, Yazid III, on behalf of his concubine mother. See Urban, “Early Islamic Mawali,” 158–​159. For examples of assertions of noble origins for concubines during the Mamluk and Ottoman periods, see Guo, “Tales of a Medieval Cairene Harem,” 109–​110; Pierce, Imperial Harem, 31–​32. 60. Al-Khasibi, al-​Hidaya, 244. 61. Al-​Kulayni, Usul al-​Kafi, 1:542–​543. 62. Al-Kulayni, Usul al-​Kafi, 1:554–​555. 63. Al-Tabari al-​Saghir, Dalail al-​imama, 410. 64. Al-Tabari al-​Saghir, Dalail al-​imama, 348–​349. For the dream of Aisha see Muslim b. al-​ Hajjaj, Sahih Muslim, 2: 1041. 65. Al-Tabrisi, Ilam al-​wara, 2:41. 66. Al-​Rawandi, al-​Kharaij wa al-​jaraih, 2:751. Because this account of Shahrbanu’s conversion is first reported in a sixth-​century source, it is quite possible it is dependent on Ibn Babawayh’s story of Narjis and not the other way around. In any case, the intertextual linkage of the biographies of the mothers of the Imams is reinforced. On the theme of dreams and conversion to Islam, see Schimmel, Die Träume des Kalifen, 119–​123. 67. This point is demonstrated amply in Kecia Ali’s recent work, Marriage and Slavery in Early Islam. See also Johansen, “Valorization of the Human Body,” esp. 78–​79, 93–​94.



Between History and Hagiography

j 263

68. See Abbott, Two Queens, 25, 38, 67; El-​Cheikh, “Caliphal Harems, Household Harems,” 92. The interstices of the legal consequences of slavery for marriage were the subject of considerable disagreement among jurists. See Ali, Marriage and Slavery, 166–​169; Johansen, “Valorization of the Human Body,” 90, 93–​94; de la Puente, “Esclavitud y Matrimonio,” 324. For the ambiguous status of Arib, see Gordon, “Place of Competition,” 75; Nicola Lauré al-​Samarai, Die Macht Der Darstellung, 66, 68. See also the comments of Gordon in “Yearning and Disquiet,” 267–​268. 69. Gordon, “Place of Competition,” 67, 74; Abbott, Two Queens, passim. 70. On the range of prices of slave girls in a medieval market see Ragib, “Les Marchés,” 758–​759; Goitein, “Slaves and Slavegirls,” 8–​11; Lewis, Race and Slavery, 13. Ragib states that the price of first-​ rate slaves began at 60 dinars. On the mothers of al-​Sadiq and al-​Hadi, see al-​Kulayni, Usul al-​Kafi, 1:542–​543; al-​Tabari al-​Saghir, Dalail, 410. 71. The humiliating public displays and invasive physical examinations to which slaves could be subjected are described by Ragib, “Marchés,” 734–​743.

Bibliography Primary Sources al-​Dhahabi, Muhammad b. Ahmad. Siyar alam al-​nubala. Beirut: Muassast al-​risala, 1993. Ibn Babawayh, Abu Jafar Muhammad b. Ali. Kamal al-​din wa tamam al-​nima. Tehran: Maktaba al-​Saduq, 1970. Ibn Hazm, Abu Muhammad Ali b.  Ahmad. Al-​Fisal fi al-​milal wa al-​ahwa wa al-​nihal. Jedda: Maktabat Ukaz, 1982. —​—​—​. Ummahat al-​khulafa Beirut: Dar al-​Kitab al-​Jadid, 1980. Ibn al-​Jawzi. Virtues of the Imam Ahmad ibn Hanbal. Vol. 2. Translated by Michael Cooperson. New York: New York University Press, 2015. al-​Kashshi, Muhammad b. Umar. Ikhtiyar marifat al-​rijal. Qumm: Muassasat Al al-​Bayt li-​Ihya al-​Turath, 1404. al-​Khasibi, Husayn b. Hamdan. Al-​Hidaya al-​kubra. Lebanon: Diyar Aql, 2007. al-​Khui, Abu al-​Qasim. Mujam rijal al-​hadith. Beirut: Dar al-​Zahra, 1983. al-​Kulayni, Muhammad b. Yaqub. Usul al-​Kafi. Tehran: Dar al-​Uswa, 1418. al-​Masudi, Ali b. al-​Husayn. Ithbat al-​wasiyya. Qumm: Maktabat Basirati, n.d. Mufid, Muhammad b. Muhammad al-​Ukbari. Al-​Fusul al-​ashara fi al-​ghayba. In Iddat rasail li’l-​ Shaykh al-​Mufid, 345–​388. Qumm: Maktabat al-​Mufid, n.d. Muslim b. al-​Hajjaj. Sahih Muslim. Stuttgart: Thesaurus Islamicus Foundation, 2000. al-​ Nawbakhti, Abu Muhammad al-​ Hasan b.  Musa. Firaq al-​Shia. Najaf:  Al-​Matba’a al-​Haydariyya, 1936. al-​Rawandi, Qutb al-​Din. Al-​Kharaij wa al-​jaraih. Qumm: Muassasat al-​Imam al-​Mahdi, 1409. al-​Saffar, Muhammad b. al-​Hasan. Basair al-​darajat. Tehran: Muassasat al-​Alami, 1404. al-​Shirazi, Muhammad al-​Husayni. Ummahat al-​masumin. Beirut: Markaz al-​Jawad, 2003. al-​Tabari al-​Saghir, Muhammad b. Jarir b. Rustum. Dalail al-​imama. Qumm: Muassasat al-​Bitha, 1412. al-​Tabrisi, Amin al-​Islam. Ilam al-​wara bi-​alam al-​huda. Qumm:  Muassasat Al al-​bayt li-​ihya al-​turath, 1417. Waki, Muhammad b. Khalaf. Akhbar al-​qudat. Cairo: Matbaat al-​istiqama, 1947.

264 i

Michael Dann

Secondary Sources Abbott, Nabia. Two Queens of Baghdad. Chicago: University of Chicago Press, 1946. Ali, Kecia. Marriage and Slavery in Early Islam. Cambridge, MA: Harvard University Press, 2010. Amir-​Moezzi, Mohammed Ali. “Šahrbanu.” In Encyclopaedia Iranica, edited by Ehsan Yarshater et al. New York: Brill, 1996–​. —​—​—​. “Shahrbanu, Lady of the Land of Iran and Mother of the Imams: Between Pre-​Islamic Iran and Imami Shi‘ism.” In The Spirituality of Shi‘i Islam, 45–​100. New York: I.B. Tauris, 2011. Asad, Ahmed. The Religious Elite of the Early Islamic Ḥijāz. Oxford: Unit for Prosopographical Research, 2011. Athamina, Khalil. “How Did Islam Contribute to Change the Legal Status of Women: The Case of the Jawari or the Female Slaves.” Al-​Qantara 28, no. 2 (2007): 383–​408. Bosworth, C. E. “Saffarids.” In Enyclopaedia of Islam, 2nd ed., edited by C. E. Bosworth et al. Leiden: Brill, 1960–​2007. Brockopp, Jonathan E. Early Maliki Law:  Ibn Abd al-​Hakam and His Major Compendium of Jurisprudence. Boston: Brill, 2000. Cooperson, Michael. Classical Arabic Biography: The Heirs of the Prophets in the Age of al-​Mamun. New York: Cambridge University Press, 2000. Daniel, Elton. Review of Eunuchs, Caliphs and Sultans: A Study in Power Relationships by David Ayalon. Journal of Islamic Studies 12 (2001): 345–​348. de la Puente, Cristina. “Esclavitud y Matrimonio en al-​Mudawwana al-​Kubrá de Sahnun.” Al-​ Qantara 16, no. 2 (1995): 309–​333. El-​Cheikh, Nadia Maria. “Caliphal Harems, Household Harems: Baghdad in the Fourth Century of the Islamic Era.” In Harem Histories, edited by Marilyn Booth, 87–​103. Durham, NC: Duke University Press, 2010. Friedman, Yaron. “al-​Husayn ibn Hamdan al-​Khasibi: A Historical Biography of the Founder of the Nusayri–​Alawite Sect.” Studia Islamica 93 (2001): 91–​112. Goitein, S. D. “Slaves and Slavegirls in the Cairo Geniza Records.” Arabica 9 (1962): 1–​20. Gordon, Matthew. “The Khaqanid Families of the Early Abbasid Period.” Journal of the American Oriental Society 121 (2001): 236–​255. —​—​—​. “The Place of Competition:  The Careers of Arib al-​Mamuniya and Ulayya Bint al-​ Mahdi, Sisters in Song.” In Abbasid Studies: Occasional Papers of the School of Abbasid Studies, Cambridge, 6–​10 July 2002, edited by James Montgomery, 61–​82. Dudley, MA:  Peeters Publishers & Department of Oriental Studies, 2004. —​—​—​. “Yearning and Disquiet:  Al-​Jahiz and the Risalat al-​Qiyan.” In Al-​Jahiz:  A  Muslim Humanist for Our Time, edited by Arnim Heinemann et al., 253–​268. Würzburg: Ergon, 2009. Guo, Li. “Tales of a Medieval Cairene Harem:  Domestic Life in al-​Biqai’s Autobiographical Chronicle.” Mamluk Studies Review 9, no. 1 (2005): 101–​121. Haider, Najam. “Prayer, Mosque and Pilgrimage: Mapping Shi‘i Sectarian Identity in 2nd/​8th Century Kufa.” Islamic Law and Society 16 (2009): 151–​174. Hayes, Edmund. “The Envoys of the Hidden Imam: Religious Institutions and the Politics of the Twelver Occultation Doctrine.” PhD diss., University of Chicago, 2015. Hitti, Philip. History of the Arabs. London: St. Martin’s Press, 1970. Hussain, Jassim M. The Occultation of the Twelfth Imam:  An Historical Background. London: Muhammadi Trust, 1982.



Between History and Hagiography

j 265

Johansen, Baber. “The Valorization of the Human Body in Muslim Sunni Law.” In Law and Society in Islam, edited by Devin Stewart, Baber Johansen, and Amy Singer, 71–​112. Princeton, NJ: Markus Weiner, 1996. Juynboll, G. H. A. “The Role of Non-​Arabs, the Mawali, in the Early Development of Muslim Hadith.” Le Muséon: Revue d’Études Orientales 118 (2005): 355–​386. Kohlberg, Etan. “From Imamiyya to Ithna-​Ashariyya.” Bulletin of the School of Oriental and African Studies 39, no. 3 (1976): 521–​534. Lewis, Bernard. Race and Slavery in the Middle East: An Historical Enquiry. New York: Oxford University Press, 1990. Marmon, Shaun. Eunuchs and Sacred Boundaries in Islamic Societies. New  York:  Oxford University Press, 1995. Modarressi, Hossein. Crisis and Consolidation in the Formative Period of Shi‘i Islam. Princeton, NJ: Darwin Press, 1993. Moomen, Mojan. An Introduction to Shi‘i Islam. New Haven, CT: Yale University Press, 1985. Peirce, Leslie. The Imperial Harem:  Women and Sovereignty in the Ottoman Empire. New York: Oxford University Press, 1993. Pellat, Charles. “Masudi et l’Imamisme.” In Le Shi‘isme Imamite, edited by Toufic Fahd, 69–​90. Paris: Presses Universitaires de France, 1970. Pierce, Matthew. Twelve Infallible Men:  The Imams and the Making of Shi‘ism. Cambridge, MA: Harvard University Press, 2016. Pinault, David. “Zaynab Bint Ali and the Place of the Women of the Households of the First Imams in Shi‘i Devotional Literature.” In Women in the Medieval Islamic World:  Power, Patronage and Piety, edited by Gavin R. G. Hambly, 69–​98. New York: St. Martin’s Press, 1998. Popovic, Alexandre. The Revolt of African Slaves in Iraq in the 3rd/​9th Century. Princeton, NJ: Markus Weiner, 1999. Ragib, Yusuf. “Les Marchés aux exclaves en terre d’Islam.” In Mercati e Mercanti nell’alto medioevo: L’area Euroasiatica e l’area Mediterranea, 721–​765. Spoleto: Presso la sede del Centro Italiano di Studi Sull’alto Medioevo, 1993. Sachedina, Abdulaziz. Islamic Messianism: The Idea of the Mahdi in Twelver Shi‘ism. New York: State University of New York Press, 1981. al-​Samarai, Nicola Lauré. Die Macht Der Darstellung: Gender, sozialer Status, historiographische Re-​Präsentation: zwei Frauenbiographien aus der frühen Abbasidenzeit. Wiesbaden: Reichert, 2001. Savant, Sarah Bowen. The New Muslims of Post-​Conquest Iran: Tradition, Memory, and Conversion. New York: Cambridge University Press, 2013. Scarcia Amoretti, Biancamaria. “A Historical Atlas on the Alids: A Proposal and a Few Samples.” In Sayyids and Sharifs in Muslim Societies: The Living Links to the Prophet, edited by Kazuo Morimoto, 92–​122. New York: Routledge, 2012. Schacht, Joseph. The Origins of Muhammadan Jurisprudence. Oxford: Clarendon Press, 1975. Schimmel, Annemarie. Die Träume des Kalifen. Munich: C.H. Beck, 1998. Smith, Margaret. “Rabia al-​Adawiyya al-​Kaysiyya.” In Enyclopaedia of Islam, 2nd ed., edited by C. E. Bosworth et al. Leiden: Brill, 1960–​2007. Sourdel, Dominique. Le Vizirat Abbaside. Damascus: Insitiut Français, 1959. Urban, Elizabeth. “The Early Islamic Mawali: A Window onto Processes of Identity Construction and Social Change.” PhD diss., University of Chicago, 2012.

13 Are Houris Heavenly Concubines? Nerina Rustomji

i  Ideas about marriage, in Islamic theological texts, have structured expectations of how Muslim men and women ought to interact. By the Abbasid era, marriage not only had an Islamic ethic,1 but its variations and domestic configurations were considered, discussed, and regulated. These discussions created a system in which gender was the most defining aspect of “legal personality.”2 As the idealized form of social interaction between adult males and females, marriage presupposed and projected a Muslim society built on a gendered code. The development of these gendered ethics are found in Quranic verses and hadiths (traditions of the prophet Muhammad). The contractual nature of marriage brought personal relationships within the legal framework. As a result, theological and legal texts formed a discourse about marriage that presented marriage as an Islamic institution and a cornerstone of Islamic society. Marriage was both a project and a social practice. In the legal framework, marriage was a contract in which a man agreed to provide for and maintain a woman; in exchange, she promised to obey her husband and to provide sexual access.3 Yet, texts about marriage also articulated the expectations and models of behavior for free Muslim men and Muslim women. The historical practices concerning marriage, however, did not always reinforce the ethics found within theological texts about marriage and divorce. As a result, the relationship between men and women within the contract of marriage continues to be a rich area for historical study.4 The intense focus on gender relations through marriage, however, detracts from the fact that Muslim men did not only experience feminine companionship through marriage. The possession of slave girls (jawari; sing., jariya) and singing girls (qiyan; sing., qayna) was another kind of gendered relationship not reflected in theological literature, 266



Are Houris Heavenly Concubines?

j 267

but was displayed and celebrated frequently in courtly literature (adab). To be certain, not every man owned slaves, and even fewer maintained singing girls, but Muslims lived in a society in which the ownership of people was common practice. Companionship with slaves, then, constituted one form of social relations that was reflected in the Quran and regulated by law. However, given the asymmetric nature of the relationship between master and servant,5 and the fact that slaves were not meant to be Muslim, texts did not typically consider an ethics of slave relationships. Marriage may have been the foundation of societal relations between free men and women, but Muslim men of means were able to experience a different kind of female companionship with their slaves. Understanding gender relations in medieval Islamic society, then, requires consideration of legal station and social placement. Whether a female companion was free or enslaved affected expectations of feminine behavior, obligations to her husband or owner, and position in the household. The distinction between wife and slave, however, was not fixed. A man could marry his slave and shift her station legally and socially. Furthermore, different categories of slaves gave rise to differing social expectations. Slave girls constituted a part of the household, but were also available for sexual access. Singing slave girls were a subset of slaves recognized for their diversions, and may or may not have been available for sexual access. They constituted a category of companionship that may have involved sexuality, but was defined by the capacity for musical performance. Muslim males, especially if they were wealthy, could experience different forms of female companionship. They could contract marriage with a woman and thus ensure their lineage. They could own female slaves, treat them as concubines, and thus experience sexual relationships. They could also own singing slave girls to provide artful entertainment and diversions to guests and households. Within this matrix of female companionship stands the ideal of the hur’in, the houris of Islamic paradise or the Garden (al-​janna). The Quran mentions the houris as beings in Islamic paradise. Although the houris are described in terms of purity and untouched manner (Q 55:56), the nature of their reward was not fully understood. The Quran never answers what exactly a houri is. Instead, it lists some attributes, and houris are mentioned sometimes in apposition with slave boys who serve drink. As a result, there is a tradition of interpreting the houris as sexually pure virgins of paradise who constitute one of the rewards of the otherworldly realm. In this interpretation, the reward of the houri is based on her sexual purity. This chapter reassesses the houri by focusing on her vocal capacity, and it suggests her reward be considered in light of her ability to speak and sing like a qayna. The following discussion treats four facets of the relationship between the houris and qiyan. The section titled “The Idealized Houri” considers the houri as a female who is introduced in the Quran and is developed further in the hadith collections of al-​Bukhari (d. 870 ce) and Muslim (d. 875 ce), and eschatological passages in al-​Muhasibi’s (d. 875 ce) Kitab al-​Tawahhum, Ibn Habib’s (d. 852 or 853 ce) Wasf al-​Firdaws, and al-​Qurtubi’s (d. 1273 ce) Jami‘li-​ahkam al-​Qur’an. In the sections, “Moral Matters” and “Aesthetic

268 i

Nerina Rustomji

Matters,” I compare descriptions of the houri in hadiths and eschatological manuals with al-​Jahiz’s (d. 869 ce) Risalat al-​Qiyan. The last section, “The Vocalized Houri,” turns to the hadith collection of al-​Tirmidhi (d. 892 ce), al-​Ghazali’s (d. 1111 ce) Ihya ulum al-​din, and Muhasibi’s Kitab al-​Tawahhum to highlight that the houri provides the sole musical expression found in Islamic paradise, and the attribute of melodic voices aligns the houri in heaven with the qayna on Earth. As a result, if we focus solely on the houri’s attribute of sexual purity, then she may resemble a heavenly concubine, but if we take note of her musical production, then the houri is not just a heavenly concubine who provides sexual favors; instead, she is a heavenly singer whose art transcends the limitations of the earthly world. The reward of the houri, then, is a not just a sexual reward. It is an aesthetic one. The possibility that the qiyan affected the eschatological promise of the heavenly houri is striking, and this type of dynamic of secular pleasures on religious texts has been demonstrated in the example of the saqi or cup bearer.6 Aside from the connection between adab and theology or kalam is the significance for women in the Abbasid period and gender relations in Arab texts. Looking at the linkages between otherworldly promise and earthly pleasures offers an opportunity to investigate Islamic gendered relationships outside the institution of marriage, and to study the ways that female companionship was configured in Islamic society. The exploration of feminine companionship outside marriage, then, sheds direct light on the nature of gendered relationships between free Muslim men and their feminine possessions. If houris are heavenly qiyan, then Islamic models of feminine companionship also should be understood outside the framework of marriage. The Idealized Houri Houris represent an untouched purity in the Quran. They are “hur’in, like pearls well-​ guarded” (Q 56:22–​23) or restrained (maqsurat) (Q 55:72). Two Quranic verses refer to “companions with big beautiful eyes” (zawwajnahum bi-​hur’in) (Q 44:54, 52:20). Other verses convey purity and modesty, but do not link them explicitly with houris: “restraining their glances” (Q 37:48, 38:52, 55:56), “like gems and small pearls whom no man or jinn before them has touched” (Q 55:56), “companions of equal age” (Q 78:33), “virginal” (Q 56:35–​377), and “pure companions” (Q 2:25, 3:15, 4:57). In the Quran, the houris and their fellow slave boys (wildan and ghilman) do not stand out as the most explicit markers of an opulent otherworldly realm; but, as beings and not mere objects, they occupy an intriguing station within the sumptuous paradise. In some respects, they are labor and their presence acknowledges a rare glimpse into the logistics of paradise. Slave boys serve drink; females provide companionship. Yet, the houris’ inclusion is not for practical reasons. After all, there are many practical matters that are not discussed in Islamic paradise. Given the descriptions about purity



Are Houris Heavenly Concubines?

j 269

and gems, the slave boys and houris function in another way. In the Quran and hadiths, both are exquisite to look at. In the later eschatological literature, however, slave boys develop a more economic role whereas houris emerge as the prime female companions within paradise.7 If houris are companions of paradise, what is their constitution and what kind of companionship do they provide? Theologians have grappled with the mystery of the houri in Quranic commentaries and eschatological narratives. The houris are never mentioned explicitly as slaves, but they are also not identified as earthly wives. Some traditions puzzle through these earthly categories and ask, “Which is better—​houris or wives?,” thus reinforcing that God’s heavenly and earthly creations are not made of the same substance and do not signal the same kinds of piety.8 As beings in paradise, houris were not bound by social or cosmological expectations. They were not married to men contractually and they did not elect contractually to submit to the will of God. Classical Islamic texts do not consider the relationship between houris and female slaves. Theological texts are more concerned with understanding what the houri promised and how the houri compared with earthly women. Furthermore, there is no agreed-​ on definition of concubines or courtesans in Islamic texts. In this sense, the question posed in the title of this chapter is more invested in a cultural history that seeks to investigate whether eschatological promises are based on earthly realities. The question also inherits a long-​standing orientalist fascination with how Islamic texts depict females whose beauty depends on a sensual or sexual quality. It is a tricky question faced by methodological challenges. It intersects with discourses that focus on gender in ways that the sources do not, and critiques and thus engages European textual and visual registers in which all women in Islamic realms are types of odalisque. Yet, the question has value because it allows us to focus on how gender and class created distinctions in earthly societies that became projected on heavenly imaginations. To understand the houri is challenging; one can see Quranic commentators grappling with the celestial being’s meaning and composition. But, beyond the relationship among houris, earthly wives, concubines, and qiyan is the larger investigation of feminine possession and reward. To understand the nature of the reward, we have to reflect on two questions: What does it mean for a woman to be a reward? What does it mean for a perfected feminine being to be a cosmological reward? In Islamic discourses, a woman is a reward when she is beautiful inside or outside. A houri is a reward because she is beautiful and pure cosmologically. Yet, although the canonical sources insist on the purity, there is a popular theological tradition that understands that purity within sexual terminology; the houris are, after all, abkaran or virginal, “untouched by men or jinn” (56:35–​37). Here is where the houri diverges from the singing slave girl. A heavenly reward could have a literal shape, but what makes it a reward is its transmuted material. In Islamic paradise, material luxuries are evidence of rightly configured purity. Pleasure in the next world is pleasurable, in other words, but it is so much

270 i

Nerina Rustomji

more. Pleasure is a spiritual experience. A houri, then, is both a female who is awarded to male believers and an embodiment of cosmological beauty. Qiyan and houris are similar in many respects; yet, their differences are even more striking. Both operate within the realm of pleasure by serving males, providing entertainment, and offering visual spectacles of feminine beauty. Although their functions were similar (albeit in different cosmological realms), their moral standing varied. Qiyan were cast as morally suspect, whereas the houris were upheld as the pinnacle of feminine beauty. This chapter argues that the houris, like earthly singing girls, were celebrated for their physical beauty and melodic voices; however, the pleasure that houris provided ultimately transcended the figure of the qiyan because houris were imbued with spiritual purity that was denuded of earthly appetitive taint. Moral Matters To compare houris and qiyan presupposes that the two kinds of females may be in apposition, if not in opposition. Certainly, there is a difference in placement. Houris are ethereal and cosmic; qiyan are worldly possessions. Houris are seen as purified of earthly matters; qiyan represent earthly appetites and pleasures. Ostensibly, the two females occupy different meanings for feminine models. Yet, the differences are a result of how the ideals of companionship are configured for different realms. When looking at the figures as possessions, both the houris and the qiyan offer a moral ambiguity. Qiyan have limited social recourse. They are possessed, exchanged through commerce, and their social mobility can proceed only through the quality and station of their owners. As a result, qiyan acquire a reputation for flirtation, scheming, and insincerity, but these outward charges obfuscate their rather limited position and their need to develop a social network.9 Nonetheless, the qiyan are diversions as opposed to women whose legitimacy is conveyed through piety, lineage, or marriage. Qiyan do not occupy social stations as wives. Their need to sustain networks of patronage makes them vulnerable, and it also makes them illicit. Houris, on the other hand, embody a kind of piety that is so exquisite that it is within their substance. As a result, it does not have to be exhibited or performed. The houri is designated for the believer. As abkaran, or virginal, she is untouched by men and jinn (Q 55:56). Although she is also a possession (albeit a cosmic one), she is never exchanged. As a result, she is accorded purity in terms of sexuality and commerce. Both females, then, provide companionship and are meant to delight the senses. Yet, although the qiyan can be owned and exchanged, the houri is the cosmic reward designated for the male believer. The taint of the qayna is not just her reputation, but the transaction that is required to acquire her. Yet, the moral valence of the singer and houri is not as straightforward as it appears. A more complex way of understanding female companionship is elaborated by al-​Jahiz in his Risalat al-​Qiyan. Al-​Jahiz’s complex, comedic, and rather biting essay focuses



Are Houris Heavenly Concubines?

j 271

on the dangers of male passion for the slave girl. In the opening, al-​Jahiz indicates the natural relationship between men and women. “Women are a ‘tillage ground’ for men, just as herbage is a provision for those animals for which it has been made a provision.”10 Drawing on the Quran, al-​Jahiz reasserts the argument about the natural relations between men and women where men hold dominion, especially over the sexuality of women. Yet, al-​Jahiz also offers a societal perspective. It is the distinctions between illicit and licit that pave way for consideration of legal matters such as inheritance and parentage. Without these injunctions, “no man would be more entitled than any other to any one woman:  just as no beast is better entitled than another to graze the pastures watered by the rains”11 In introducing the question of licit and illicit, al-​Jahiz focuses on the religious and legal codes that shape social interactions between the sexes. The interactions are shaped by the possibility of limited visual access to women through glances. In contrast, the direct visibility of the qayna is her hallmark. She has a more public presence at court. More important, slaves, when purchased, are typically examined and valued.12 However, when the good cannot be quantified easily, it is “judged by beauty or ugliness.”13 The central problem for al-​Jahiz, then, is a lack of judgment. A  good judge understands that beauty is about attaining completeness and moderation. Beauty is the result of a moderate proportion that establishes a kind of equilibrium. Al-​Jahiz discusses the cases of mismatched proportions—​a tiny-​eyed man with a big nose or a large face with a skinny body—​and he extends the question of proportional moderation and balance to “buildings, rugs, embroidery, clothes or canals where water flows.”14 Although women can be jealous and frail, the central problem for al-​Jahiz is men’s passions. Just as al-​Jahiz makes a strong case for the importance of completeness and moderation with objects, he also notes the importance of moderation in passion. The passion for love, focused on the singing girl, constitutes the heart of al-​Jahiz’s critique. For al-​Jahiz, love-​sickness is a malady that is tricky to cure because it compounds so many other problems: “Passion is compounded of love and infatuation and natural affinity and habitude of association. It begins with a growing intensity, reaches a climax, and then falls off by natural progression to the stage of complete dissolution and the point of positive revulsion.”15 Yet, what makes the passion for the singing girl particularly dangerous is that the singing girl invokes pleasure through many senses. The eye appreciates the beautiful girl. The ear appreciates her voice, and touching her leads to desire for sex. The qayna offers such a spectacle that the senses compete in sending “its message about her to the heart before the other.”16 The problem with this sensual overload is that it is not balanced and complete even though the soul finds satisfaction in these pleasures.17 The singing girl takes advantage of the male imbalance. Within the context of men’s passion, al-​Jahiz enumerates the singing girl’s moral orientation. She is not sincere, she ensnares men, she plays games, and she employs “deceit and treachery in squeezing out

272 i

Nerina Rustomji

the property of the deluded victim and then abandoning him.”18 The reputation of the singing girl is that of seducer and temptress; a moral taint surrounds her. If the singing girl is a possession driven by imbalanced pleasure and not bound by marriage (although an owner can choose to marry and manumit his slave girl), why is the houri who is also a possession and not bound by marriage considered to be the epitome of purity? The houri is a reward promised to men in paradise, and she invokes all the same senses as the singing girls. Her voice is so melodious that no one has heard anything like it on Earth. The sight of her is marvelous, and she smells of luxurious scents such as musk and amber. In the traditions, however, there are no developed scenes of touching houris. If companionship with the singing girl is imbalanced and perhaps immoral, does the companionship with the houri reinforce or invalidate the sense of balance and morality? To a certain extent, the promise of the houri is balanced pleasure. The houri is enjoyed not on Earth, but in a different cosmological realm. That realm dictates the context of her being and meaning. The houri is not just pure, but as a being of paradise, she is beyond the judgment of morality. Yet, the houri herself could hardly be called moral. She is not an agent who exercises will and makes choices. She is not even a Muslim who submits to the will of Allah. She is merely a creation unique to the cosmological realm. It is this sense of extramoral or amoral capacity that allows the houri to invoke the same senses as the singing girl, but exist within a different register of acceptable behavior. The same senses that appreciate the qayna are activated with the houri, but the dictates of the different realm allow her to be enjoyed without the possibility of immoderation and incompleteness. The passion for the houri, then, is not a complicated earthly passion. Instead, it is something all together more elevated. It is the elevation of the earthly appetites that defines the Garden as the realm of true spiritual pleasure. Aesthetic Matters Although the moral valance of qiyan and houris differs, in terms of aesthetics, they are both considered to be beautiful. Houris and qiyan are defined by similar aesthetic standards. They are not defined by their personality, temperament, or upbringing. Instead, they must be pleasing to the senses. One caveat is that the singing girl is intelligent, but that intelligence, particularly in verse, is not necessarily celebrated. The houri, in contrast, does not exhibit any singular thoughts. Furthermore, what is understood as beautiful is slightly different in the two cases. Although both females are appreciated visually, the effect of the appreciation leads to different moral standings. The qayna entices and potentially seduces a man away from his social placement. The houri elevates pleasure into a cosmological reward. Al-​Jahiz’s different components of beauty reveal a philosophy of aesthetics. Beauty, as al-​Jahiz counseled, has an ethical framework to it. It is a system of sorts; one cannot be out of proportion on the inside or outside. How is the qayna perceived as beautiful? Because the singer invokes the senses in a variety of ways, her owner can experience her



Are Houris Heavenly Concubines?

j 273

through a variety of senses. The eye sees the beautiful girl, and the viewer does not have to contend with her intelligence (“since cleverness and beauty are hardly ever simultaneously possessed by a single object of enjoyment and delight”19). The ear hears her delightful voice, and touching her leads to longing and desire. What is striking about the qayna is how these attributes are combined for al-​Jahiz: When the girl raises her voice in song, the gaze is riveted on her, the hearing is directed attentively to her, and the heart surrenders itself to her sovereignty. Hearing and sight race each other to see which of the two can transmit its message about her to the heart before the other, and they arrive simultaneously at the heart’s core and pour out what they have observed.20 The singer invokes the different senses and offers an engaged flirtation that captures the man’s attention and focuses his efforts on the need to shower the qayna with gifts and attention. The houri, in contrast, requires no work. She is, in fact, a reward for the work already conducted. There is no insincerity with the houri because she is true to her nature to serve and provide companionship. Unlike the qayna, the houri does not have to exhibit intelligence or accomplishments because she is perfect ontologically. The form of companionship affirms the spiritual beauty of the (male) believer and reinforces that affirmation with a transformed aesthetic register. Both the houris and qiyan wear brocade, enjoy silks, and smell of musk and amber. Although the opulent attire of the singer allows for a sensual event, the houri’s similar opulent attire demonstrates that the believer has earned a place in the Garden through good works. For that reason, even the plants bow when the houri passes.21 She is such an exquisite production that even the landscape of the Garden recognizes her presence. Beyond the landscape of the Garden, the wonders of the houris can be judged in terms of their earthly affects. For example, if she happened to spit in the world, the entire world would have been filled with musk.22 These forms of earthly amplification are ways to articulate the aesthetic nature of the houri, who becomes a literal embodiment of the Garden. Although the qiyan invoke the senses, houris invoke the transformed senses of each believer. Those senses take in the very stuff of the Garden: saffron, musk, amber, and camphor, and raw silk. From knee to toe, they are made of saffron; from knee to breast, musk; breast to neck, amber; and neck to head, camphor.23 From one glimpse, one scent, one touch of the houri, believers will experience all the wonders of the Garden: gold, silver, light, saffron, musk, silk. What emanates from the touch is not sexual longing, but the affirmation that the believer has been awarded the supreme reward and can appreciate the rewards through sublime experience. This is not to say that believers forego sex. The purified amplification extends to the otherworld, where male believers are able sustain the sexual power of 100 men, but they do not have to contend with the impurities of bodily fluid such as semen.24

274 i

Nerina Rustomji

Although the beauty of houris is described in terms of cosmic pureness and sumptuary materials, the qiyan are understood through attributes and affects. Yet, the houris have one constitutive attribute. They are fair and they have big eyes. Is the houri based on any particular model of beauty or a kind of female slave? Ibn Butlan, in his Risala fi shira’ al-​raqiq wa-​taqlib al-​abid, offers a rubric for female slaves; however, the descriptions of houris do not align with other descriptions of slaves. If there is one group of women who have similar characteristics, it is the women of Mecca who are “languorous, feminine, with supple wrists and of a white color tinged with brown. Their figures are beautiful, their bodies lissome, their mouths clean and cool, their hair curly, their eyes sickly and languid.”25 However, to try to find the earthly model of the houri is to miss that her aesthetic nature is not only physical. It also signals a spiritual register. Her fairness and the ability to see through her bones does not make her like a Meccan. Instead, these features suggest that the world has literally not left its mark. In both cases, aesthetics is what defines the singer and the houri. What links the two cases is not just the figure of the females, but also the sumptuous objects that are used. For the houri, the stuff of the Garden—​brocade, silk, musk, and amber—​singles her unique composition. In fact, the use of adornment and amorous messages may have been the same in descriptions of the houri and qayna. Ibn Habib reveals that written on the houris’ bodies is the phrase, “You are my love, I am your love, my eyes are only for you and my soul leads to you.”26 The message for the believer is embodied literally in the houri. Interestingly, slave girls may have adopted bodily ornamentation and inscribed poetry on their hands, feet, and forehead with oils, henna, and scents.27 Ibn Washsha (d. 936 ce), in Kitab al-​Muwashsha, even describes how qiyan adorned their bodies with poetry written in musk and amber.28 These bodily testaments of love reinforce and fuse the ideals of love and possessions. It is no wonder, then, that the qiyan were once referred to as the “qiyan of the houris.”29 The Vocalized Houri The houri embodies beauty in Islamic paradise. Composed figuratively of the materials of the exquisite realm, she illustrates the potential of the sensual, spiritual cosmic realm. Yet, what sets the houri apart from the sensual landscape is her voice. Within the eschatological manuals, the houri becomes an active part of paradise and beckons to the believer or speaks directly to him.30 Although paradise is mostly a visual spectacle, the houris are the sole actors who offer a voice in paradise. What do they say? The message of the houri reaffirms the love she has for the believer: “We live forever and never pass away, we are affluent and never austere, we are content and never discontent. Blessed are those who belong to us and to whom we belong.”31 In this sense, the houri is the sole vocalized agent of the Garden. Like her other attributes, her voice is one that embodies beauty. It is soft and melodic, and when the houri’s voice is raised, it is unlike the creation of the world.32 In fact, the vocalization of the houris is so singular that Mundhiri (d. 1258 ce), in al-​Targhib wa’l-​tarhib min al-​hadith



Are Houris Heavenly Concubines?

j 275

al-​sharif, follows his section on the “Description of the Women of al-​Janna,” with “songs of the hur’in whose melodic voices are heard by humanity and jinn.” Their messages for believers are conveyed through the verb “to sing.”33 The correspondence of the houri with song, then, is the site where we see an intersection between feminine beauty on heaven and Earth. The houris act as the qiyan of paradise. Their beautiful voices fill paradise, and their messages to the believers are conveyed through melody and song that is so marvelous that is exceeds the beauty of the created world. Male believers receive this reward without commerce and enjoy the reward without passions that signal imbalance in the soul. The particular attribute of melodic voices points to the qiyan as the model of feminine beauty. Yet, the houris are not limited by the practice and taint of earthly qiyan. They are spiritual qiyan who are promised to men and whose potential stands in apposition, opposition, or support (depending on your interpretation of the houri) of women. In this respect, they are not just heavenly concubines. They are feminine works of art. Yet, the reward of the houri cannot be understood fully because the cosmological world operates by different rules and can be understood only through earthly realities, experiences, and institutions. The houris are not possessions to be purchased; they are possessions to be earned. They are not intelligent women cultivated for the arts to please their masters; they are embodiments of the wonders of paradise that invoke the senses. Quranic commentators long grappled with the meaning of the houri, the nature of her attributes, and how to understand her in relation to earthly wives. Although the discussions were never conclusive, the confusion over the nature of the cosmological houri suggests that commentators saw houris as separate from earthly wives who were transformed cosmologically (although this interpretation has been contested by twenty-first-​century commentators). The earlier Quranic commentators intimated there is an opposition between houri and wife. Placing the texts about the houri alongside the literature about the qayna suggests another possibility. The houris could not be wives because they were beings of paradise, as opposed to transformed Muslims; however, the form of the companionship they provided is reminiscent of the pleasures offered by the qiyan. Houris were a type of heavenly possession and their descriptions draw on a social earthly world in which women were bought and exchanged for men’s diversions. The houri, however, is not tainted by the immoderation of earthly life or earthly commerce. Instead, her promise was a kind of pure companionship that transcended both contractual relationships such as marriage and possessive relationships such as slaveholding. Studying the houri, then, allows us to see the different forms of companionship experienced during the Abbasid era. These alternate relationships may not have been discussed at length in theological texts. Yet, they were not merely frivolous courtly entertainments. Instead, the qiyan, even if slaves, left an imprint on the eschatological promises of the Islamic afterlife. In doing so, they helped develop the understanding of a Quranic eschatological promise, and they influenced the way that Muslim men would understand the possibilities for female companionship on heaven and Earth.

276 i

Nerina Rustomji Notes

1. Al-Ghazali, Marriage and Sexuality in Islam. 2. Ali, Marriage and Slavery in Early Islam, 47. 3. Ali, Marriage and Slavery in Early Islam, 96. 4. Rapoport, Marriage, Money, and Divorce in Medieval Islamic Society, 1–​11. 5. Richardson, “Singing Slave Girls,”, 105–​106. 6. Stetkevych, “Intoxication and Immortality,” 210–​232. 7. Rustomji, Garden and the Fire, 111–​115. 8. Al-Qurtubi, Jami‘ li-​ahkam al-​Qur’an, 44–​54. 9. Gordon, “Yearning and Disquiet,” 268. 10. Al-Jahiz, Risalat al-​qiyan, 15. 11. Al-Jahiz, Risalat al-​qiyan, 15. 12. Al-Jahiz, Risalat al-​qiyan, 24. 13. Al-Jahiz, Risalat al-​qiyan, 24. 14. Al-Jahiz, Risalat al-​qiyan, 26. 15. Al-Jahiz, Risalat al-​qiyan, 28. 16. Al-Jahiz, Risalat al-​qiyan, 31. 17. Al-Jahiz, Risalat al-​qiyan, 30. 18. Al-​Jahiz, Risalat al-​qiyan, 34. 19. Al-Jahiz, Risalat al-​qiyan, 31. 20. Al-Jahiz, Risalat al-​qiyan, 31. 21. Al-​Muhasbi, Kitab al-​Tawahhum, 149. 22. Al-Ghazali, Remembrance of Death and Afterlife, 130. 23. Al-Ghazali, Remembrance of Death and Afterlife, 130. 24. Al-Tirmidhi, al-​Jami‘al-​sahih, 4:2536. 25. Lewis. Islam, From the Prophet Muhammad to the Capture of Constantinople, 249. 26. Ibn Habib, Wasf al-​Firdaws, 223. 27. I am grateful to Lisa Nielson for bringing this passage to my attention. Nielson, “Diversions of Pleasure,” 161–​162. 28. Nielson, “Diversions of Pleasure,” xx. 29. I am grateful to Matthew Gordon for bringing this passage to my attention. Caswell, The Slave Girls of Baghdad, 27. 30. Al-​Muhasbi, Kitab al-​Tawahhum, 160–​163. 31. Al-Tirmidhi, al-​Jami‘ al-​sahih, 4:2564. 32. Al-Tirmidhi, al-​Jami‘ al-​sahih, 4:2564. 33. Mundhiri, al-​Targhib wa-​al-​tarhib, 130.

Bibliography Primary Sources al-​Ghazali, Abu Hamid Muhammad. Marriage and Sexuality in Islam: A Translation of al-​ Ghazali’s Book on the Etiquette of Marriage from the Ihya ulum al-din. Translated by Madelaine Farah. Salt Lake City: University of Utah Press, 1984.



Are Houris Heavenly Concubines?

j 277

—​—​—​. The Remembrance of Death and Afterlife: Selections from Ihya ulum al-din. Translated by T. J. Winter. Cambridge: Islamic Texts Society, 1989. Ibn Habib, Abd al-​Malik. Wasf al-​Firdaws. Beirut: Dar al-​Kutub al-​Ilmiyya, 1987. al-​Jahiz, Abu Uthman Amr ibn Bahr. Risalat al-​qiyan: The Epistle on Singing-​girls of Jahiz. Edited by A. F. L. Beeston. Warminster, UK: Aris & Phillips, 1980. al-​Muhasibi, Abu Abdallah Harith ibn Asad. Kitab al-​Tawahhum: Une Vision humaine des fins dernières: Le Kitab al-​tawahhum d’al-​Muhasibi. Translated by Andre Roman. Paris: Librairie Klincksieck, 1978. al-​Mundhiri, Abd al-​Azim. Al-​Targhib wa’l-​tarhib min al-​hadith al-​sharif. Beirut: Dar al-​Kutub al-​Ilmiyah, 1986. al-​Qurtubi, Abu Abdallah Muhammad. Jami‘li-​ahkam al-​Qur’an. Amman: Royal Al al-​Bayt Institute for Islamic Thought (Digital Islamic Library Project), 2000. al-​Tirmidhi, Abu Isa Muhammad ibn Isa. Al-Jami‘al-​sahih. 4 vols. Edited by Ibrahim Atwa ‘Awd, I. Cairo: Dar Ihya al-​Turath al-​Arabi, n.d.

Secondary Sources Ali, Kecia. Marriage and Slavery in Early Islam. Cambridge, MA: Harvard University Press, 2010. Caswell, Fuad Matthew. The Slave Girls of Baghdad:  The Qiyan in the Early Abbasid Era. London: I.B. Tauris, 2011. Gordon, Matthew. “Yearning and Disquiet: al-​Jahiz and the Risalat al-​qiyan.” In al-​Jahiz: A Muslim Humanist for our Time, edited by Arnim Heinemann et  al., 253–​268. Beirut:  Ergon Verlag Würzburg, 2009. Lewis, Bernard. Islam, From the Prophet Muhammad to the Capture of Constantinople: Religion and Society. New York: Harper & Row, 1974. Nielson, Lisa. “Diversions of Pleasure:  Singing Slave Girls and the Politics of Music in Early Islamic Courts (661–​1000 ce): Their Influence, History and Cultural Roots as Seen through the Kitab al-​Muwashsha (Book of Brocade) of Ibn al-​Washsha, the Risala al-​Qiyan (Epistle of the Singing Girls) of al-​Jahiz, and the Dhamm al-​Mahahi (Censure of Instruments of Diversion) of Ibn’l Dunya.” PhD diss., University of Maine, 2010. Rapoport, Yossef. Marriage, Money, and Divorce in Medieval Islamic Society. Cambridge: Cambridge University Press, 2005. Richardson, Kristina. “Singing Slave Girls (qiyan) of the Abbasid Court.” In Children in Slavery through the Ages, edited by Joseph Miller et al., 105–​118. Athens: Ohio University Press, 2009. Rustomji, Nerina. The Garden and the Fire:  Heaven and Hell in Islamic Culture. New  York: Columbia University Press, 2009. Stetkevych, Suzanne Pinckney. “Intoxication and Immortality: Wine and Associated Imagery in al-​Ma’arri’s Garden.” In Homoeroticism in Classical Arabic Literature, edited by J. W. Wright and Everett Rowson, 210–​232. New York: Columbia University Press.

14 Educated Slave Women and Gift Exchange in Abbasid Culture Jocelyn Sharlet

i  Although educated slave women play a significant role in Abbasid literature, their portrayal has received much less attention than that of their free male counterparts. The Abbasid-​era stories of gift exchange that feature the slave women Utba and Inan demonstrate how these slave women participated in the negotiation of their evolving status in the context of patriarchy in general, and educated female slavery in particular. Two stories considered here treat the participation of such women in episodes of gift exchange. The stories illustrate the dynamics of the educated slave woman’s subjective agency and objectification in the competition between different sectors of elites, including rulers and subordinate notables. As a theme of Abbasid literature, the exchange of material gifts contributes to a reconstruction of elite networks and hierarchies. The slave woman may be objectified as a gift, but she may also display subjective agency by interfering with her exchange or by giving a gift herself. Educated female slavery fits into a broader pattern of change and continuity in slavery. Slavery in Islamic law demonstrates continuity as well as significant contrasts with slavery in the legal systems that prevailed in the regions that later became the medieval Islamic world, including Sasanian, Jewish, Egyptian and Greco-​Egyptian, and Christian and Roman provincial law.1 Islamic law regulated slavery in new ways. Debt bondage (including self-​dedition) as well as the selling and pawning of one’s children were often legal in these other legal systems, although the selling of one’s wife was often illegal. Likewise, foundlings could be enslaved legally. During the earliest stage of Islamic law (seventh and eighth centuries), debt bondage and the enslavement of foundlings was sometimes permitted, but the sale of free persons was illegal.2 Subsequently, during the period when many of the foundational texts were written (ninth and tenth centuries), debt bondage 278



Educated Slave Women and Gift Exchange in Abbasid Culture

j 279

had become illegal and the sale of family members was forbidden, although foundlings were presumed to be free and were raised by Muslims, Christians, or Jews, depending on the religion of the neighborhood in which they were found.3 As in other systems, the enslavement of people captured in military conquests of new territory was deemed legal in Islamic law. The acquisition of slaves as part of ongoing capitulation agreements was also legal.4 Similarly, people who were captured outside the domain of Islamic law and brought to slave markets in Muslim-​ruled cities could often be sold legally.5 We can conclude that Islamic law maintained the practice of slavery while also circumscribing it, for example, with new restrictions on the sale of free persons into slavery. In addition to the rejection of debt bondage and the sale of family members into slavery, Islamic law imposed regulations on the status of a female slave who worked as the concubine of her master. Jurists used overlapping concepts in their discussions of wives and concubines.6 However, a wife—​in contrast with a concubine or other slave—​could not be sold to someone else. Like marriage, concubinage entailed one man’s exclusive dominion.7 The concubine who bore her owner’s child could not be sold. Thus, Islamic law defined a special status for female slaves whose work generated legally recognized kinship relations. The restrictions on the sale of free persons into slavery and the special legal value placed on female slaves with legally recognized kinship relations to their masters parallel the special cultural and social value assigned to educated female slaves in Muslim-​ruled societies. In all three cases, slavery becomes a context in which new dimensions of the slave’s humanity—​kinship to one’s own family, kinship with a master, or education—​are recognized legally or culturally alongside the slave’s commodification. In the case of concubines and educated female slaves, masters could enjoy privileges with slave women they could not enjoy with free women, so that exploitation of slave labor went hand in hand with opportunities for slave women. Educated female slaves who were used for entertainment played an increasingly prominent role in society in the context of the increasing number of slaves resulting from Islamic conquests. They were viewed as a sign of status for their owners and were also exchanged as gifts to maintain alliances between rulers and among different kinds of elites.8 Although such slaves during the first centuries of Islam were educated in Medina and other centers, Basra appears to have become the center of their education during the Abbasid period. Their owners hosted cultural salons that often included different kinds of elites so that the owner and the educated female slave each had the opportunity to build networks. Thus, it is not surprising that educated female slaves sometimes rose to positions of power. At this time, education involved, in large part, the Arabic language, Quran, poetry, and music. Their role in entertainment contributed to the expansion of women’s education in general.9 Regulations for slavery in general and for concubines in particular, as well as social practices that valorized women’s education in what we know as the humanities, led to developments in women’s subjectivity and human rights within the exploitative context of slave labor.

280 i

Jocelyn Sharlet

In the stories about Utba and Inan, the slave woman’s role in the competition between different sectors of elite society not only sheds light on the authority of the ruler and other high-​ranked persons, but says much about the manner in which the activities of subordinates undercut that authority. These individuals worked in and around the court, and included elite slave women, poets, medical specialists, merchants, and high officials. In this context, each woman was objectified, yet emerged as a subjective actor who influenced the outcome of elite competition.10 These women intervened in the terms of their ownership and trade; gave gifts great and small; composed poetry and sang, in response to their often violent treatment; and pursued extraslavery affairs. In doing so, the women publicly performed their roles as slaves who, on occasion, challenged the coercion inherent in slavery. In the gift exchange stories of the two women, who were in some sense commodities themselves, we encounter an indirect approach to competition between elites. The stories also take on social tensions and political crises, and provide an important perspective on the direct and public impact of elite women on Abbasid social life. Viewed from this perspective, Utba, Inan, and women like them become figures of Abbasid history. The stories of the two elite slave women Utba and Inan fit into a broader literary pattern of historical characters as a source for Abbasid history. Abbasid writers transmitted, wrote, and compiled anecdotes in which they used a combination of fact and fiction, as well as juxtaposition of different versions of stories and rhetoric, to make arguments about culture, society, and elite politics.11 Many of these stories concern historical figures linked to the political sphere; they and others like them were probably among the audience for the stories. Meanwhile, religious scholars circulated sayings of the Prophet Muhammad (hadith) and biographies of hadith transmitters, exegesis of the Quran (tafsir), and law (fiqh). Religious and court writing flourished in urban, cosmopolitan networks through social exchange and the dissemination of written texts, and each group took part in defining social values.12 The stories circulated by court writers offered an opportunity to contemplate, comment on, and criticize aspects of social life, including the careers of historical figures, events, or practices.13 As a result, these stories can be a rich source for social history. Among the preoccupations of these writers were patriarchy, the caliphate, and the caliph’s reliance on different sectors of elite male society—​from merchants and doctors to soldiers and bureaucrats. But, elite slave women—​and, to a lesser extent, other categories of women—​also play a significant though limited role in this literature. The portrayal of elite slave women is related to earlier representations of free women in early Arabic love stories. Most scholars agree that Abbasid-​era stories, composed during the eighth and ninth centuries, about Umayyad-​era poets and lovers of the preceding century and a half, convey conflicting reactions to the social dislocations produced by Islamic conquests, and articulate anxieties about the role of the individual in social life.14 Further social dislocations that followed on the development of urban, cosmopolitan society—​including the expanding role of elite slave women in Abbasid literature, new



Educated Slave Women and Gift Exchange in Abbasid Culture

j 281

roles for male elites, and increasingly elaborate ideas of refinement—​produced new ways of thinking about identity and individuality for elite slave women and their male handlers.15 Julia Bray, commenting on this period, observed: The diversification of male roles, and of the self, in response to the diversification of career opportunities, involved a sliding scale of authority and autonomy in the different areas of a male individual’s life, and is one of the most significant developments of metropolitan Abbasid society compared to earlier Muslim societies. Abbasid analysts saw it as the last phase in the digestion of the conquests and of a process of social transformation which had begun some two centuries before.16 The Abbasid nexus of themes of identity, individuality, love, female enslavement, and the refined use of language and gift exchange thus developed as a long-​term outcome of the Islamic conquests and cosmopolitan urbanization. Two stories, among many others, concern Utba and Inan. These accounts of gift exchange and confiscation, featuring slave women and their male counterparts, suggest that transactions between them played a crucial role not only in the fashioning of the elite male self but also in crafting new forms of authority and autonomy for the elite female slave. Utba, the Gift Not Given The competition over Utba between elite male actors, and in accounts of which she plays an active role, features her master, the caliph al-​Mahdi (r. 775–​785 ce), and her suitor, the poet Abu al-​Atahiya (d. 826), as well as al-​Mahdi’s son, the future caliph, Harun al-​Rashid (r. 786–​809). The episode of gift exchange in which Utba plays a part, and that contributes to the articulation of her subjectivity, involves al-​Mahdi’s promise to give her to Abu al-​Atahiya, his decision to renege on the promise, and Abu al-​ Atahiya’s attempt to secure the gift (Utba) by giving al-​Mahdi modest material gifts on which he has written praise poetry. The prospect of al-​Mahdi giving her to the poet entails her objectification as a slave, as does the poet’s repeated demand for al-​Mahdi to give Utba to him. Utba’s subjective agency, on the other hand, appears in her public rejection of the poet, her authoritative use of an elite code of refined love to reject him, and her influence on al-​Mahdi’s decision not to give her away. Moreover, the revelation in the longest version of the story of a secret extraslavery affair between Utba and Abu al-​Atahiya contributes to Utba’s display of subjective agency and resistance to her master, whereas her refusal to accept Abu al-​Atahiya after he reveals their secret love demonstrates her defiance of her lover. The activities of Utba and Abu al-​Atahiya speak to the prestige of the Abbasids—​al-​Mahdi and his son Harun al-​Rashid—​but also the limitations on their power, as Utba and Abu al-​Atahiya challenge their authority and the two Abbasids undermine one another.

282 i

Jocelyn Sharlet

Abu al-​Atahiya’s well-​known love poetry on Utba draws on conventions that typically occur in praise poetry for rulers during this period. It features a fictional/​legendary lost or departing woman of pre-​Islamic Arabia who serves as the figure of the poet’s desire for the patron’s support. However, his unsolicited love poetry, which is understood to be about Utba, is considered scandalous. Utba, or a matriarch who supervises her, pressures the caliph not to give her away to the poet so as not to ruin her reputation further and undermine the caliph’s dignity. Abu al-​Atahiya gets in trouble for presenting love poetry publicly about Utba, and when the caliph forbids him from composing love poetry, he turns to ascetic poetry, a genre of Arabic literature for which he is best known. This only lands him in further trouble, this time with the caliph’s son Harun al-​R ashid. The genre of ascetic poetry was perceived to have links to heretical beliefs; that is, it was seen as a context in which poets explored a range of often controversial ideas that were circulating in Islamic and other religious and philosophical circles at the time. The circulation of poetry for and about rulers, and the use of love and ascetic poetry in court, elite gatherings (majalis), and written work—​especially in citations in prose works—​was a form of public performance in the marketplace of ideas. If, in some versions of the story, Utba resists being given to the poet, in most versions she complains about him for failing to live up to the elite code of refined love. It was a code that emphasized secrecy, rejection of material needs, monogamy, and chastity. So, for example, in one version of the story, she elopes secretly with the poet then denies it in accordance with the code of refined love. What follows is a discussion of three principal versions of the story. It begins with the version recorded in the two-​volume Kitab al-​Zahra (The Book of the Flower) by Ibn Dawud al-​Isbahani (d. 909), a Baghdadi jurist and writer of literary works.17 The book is a compilation of poetry and prose on chaste love and friendship; the story of Utba, al-​Mahdi, and Abu al-​Atahiya opens the chapter on gift exchange. It continues with the version recorded in the multivolume Kitab al-​Aghani (Book of Songs) by Abu al-​Faraj al-​ Isbahani (d. 967), a renowned figure in Arabic letters of Iranian background also based in Baghdad.18 The story appears in a short chapter about the male singer who sang the poet’s verses about Utba for her master. The discussion concludes with the longer version of the story that occurs in the Zahr al-​Adab (The Blossoms of Refined Literature) by al-​Husri, a North African scholar (d. 1022).19 This work is a two-​volume compilation of prose and poetry that, like many works written in North Africa and Andalusia, draws heavily on the Arabic heritage of the eastern Arab world. In Ibn Dawud al-​Isbahani’s version of Utba’s story, the caliph al-​Mahdi summons Abu al-​Atahiya and requests poetry about Utba. He then promises to marry her to the poet. But, no sooner does the poet depart, than al-​Mahdi is told, “The women said that he composed love poetry about her and made her famous (in a scandalous way)—​if you marry her to him, you will show that the rumors are true.”20 The caliph immediately reneges on his promise.



Educated Slave Women and Gift Exchange in Abbasid Culture

j 283

“Poetry about Utba” might be understood as love poetry or it could take the form of a love introduction to a praise poem about the caliph, and it is central to the poet’s request for Utba from the caliph. As her master, the caliph controls her, but the story underscores the complexity of their relationship. Publicity about Utba in poetry, especially if it were followed by praise of the caliph, might have worked to enhance the caliph’s standing by demonstrating his refined approach to women. Accidental reference to a woman of the palace by name, however, posed a problem; publicity about Utba in Abu al-​Atahiya’s love poetry—​that is, before al-​Mahdi requested it—​and the women’s awareness of it, lowers the caliph’s standing by undermining his authority over the women of his household. Abu al-​Atahiya sought Utba again by delivering three fans on which he wrote lines of verse. He gave them to the caliph’s servant Masrur, explaining they were gifts for the caliph. The caliph looked at each in turn and liked the examples of poetry; each comprised a miniature praise qasida (a metered, monorhyme ode): “upon one of them was written: I perceived the fragrance of the success of my request, for its breeze wafts from your palms And when he saw it he said, ‘He did well.’ Then he read the second one upon which was written: In my hope for you, I am committed to a journey on a mount that hastens in carrying me to you Then he read the third one upon which was written: And if I sometimes despair, I then say to myself, “No, the one who has guaranteed success is noble and generous. And then he said ‘He did well’.”21 Al-​Mahdi replied there was no way he would give him Utba, but ordered 50,000 of an unspecified unit of money as recompense. In the poetry borne on fans, the poet implies that the gift echoes his request for Utba. The verses connect to the motif of the fragrant breeze of success from al-​Mahdi’s palms: the motion of a waved fan creates a breeze, but only an artificial one. In response, al-​Mahdi reasserts possession of Utba and offers the poet a large monetary gift to make him go away. The verse concludes that he is generous, but the monetary gift is a poor substitute for fulfilling the promise. The artificial breeze in lieu of real generosity echoes the monetary gift in lieu of Utba, the poet’s real object of desire. Abu al-​Atahiya’s second attempt to obtain Utba from al-​Mahdi involved a gift of a perfumed garment contained in a jar on which he wrote: “My soul is attached to something worldly that God and al-​Mahdi, who executes the orders of God, can supply for me. /​ I despair of attaining it but then your contempt for this world and everything in it renews my ambition for it.”22 At this point, al-​Mahdi considers exchanging Utba for the poetry

284 i

Jocelyn Sharlet

(about her), which had been his original intention, but realizes that a scandal was brewing. Utba objects, however, saying “Even with all of my dignity and service, you would send me off to an ugly jar-​seller who profits on love?” Al-​Mahdi responds by ordering the jar to be filled with money, again in a gesture meaning “Please go away.” Abu al-​Atahiya replies, “He didn’t order anything but [gold] dinars,” to which the secretaries say, “Maybe you are right, but if you want we can fill it with [silver] dirhams until we get the invoice for dinars.” As they argue, Utba reiterates her complaint about the poet as unrefined, saying, “If he were a lover as he says, he would not worry about the difference between dirhams and dinars.”23 Thus, Abu al-​Atahiya fails initially to obtain Utba from al-​Mahdi using the ode, a genre crucial to gift exchange in patronage—​in this case, written on the gift of fans. Praise in an ode puts pressure on the patron to be generous in a worldly sense, but al-​Mahdi does not give Utba away, offering money instead. The poet raises the stakes when he offers two verses that promote al-​Mahdi’s role as “God’s caliph” who holds worldly wealth in contempt. The motif refers to a contentious issue about who holds authority over the Muslim community—​the imperial court or the religious scholars.24 Abu al-​Atahiya’s praise evokes the ideological argument for divine rule by the caliph, but he uses it for his individual and worldly goal—​to convince al-​Mahdi to give him Utba as a gift. The poet puts the caliph in a corner by calling him God’s caliph. If he fails to provide the poet with what he seeks, he undercuts the public statement of his righteous standing. No wonder al-​Mahdi is ready to give Utba away. But al-​Mahdi’s decision, to submit to Utba’s objection to being married to the poet, implies he is prepared to sacrifice his claim to legitimacy to satisfy her wishes. She backs up her objection with the reference to refined love by which she seeks to assert her elite standing. The poet is but a jar seller and, more important, he argues about the money offered to him, so is undeserving of the respect accorded to refined lovers, which might entail al-​Mahdi giving her away to him. Thus, at a crucial juncture, Utba herself intervenes to prevent al-​Mahdi from giving her away, emphasizing her subjective status. The second version of Utba’s story occurs in Abu al-​Faraj al-​Isbahani’s Book of Songs. It comes in the entry on the musician Yazid al-​Hawra, from whom Abu al-​Atahiya seeks intercession with al-​Mahdi in the effort to win Utba. The friend sings the three-​verse mini ode that appears in the version from the Kitab al-​Zahra discussed earlier, but adds an additional verse of praise that suggests that rain will follow (from “the fragrant breeze”). The caliph again replies, “As for Utba, no way—​her [female] master (mawlatuha) has forbidden it. But take 50,000 dirhams and buy one better than her.”25 In this version of the story, the poet’s demand for fulfillment of the promise is amplified by the verses that he wrote on fans (as noted earlier), the intervention of the poet’s friend, the additional verse on the falling of rain, and the friend’s musical setting of the verses. In this second version of the story, it is the matriarch, not Utba herself, who intervenes to prevent al-​Mahdi from giving the young woman to Abu al-​Atahiya. Here, the monetary gift is definitely dirhams, and is offered more literally as compensation for the woman-​not-​ given, with instructions to purchase another slave woman. The idea that the money—​the



Educated Slave Women and Gift Exchange in Abbasid Culture

j 285

compensation for Utba—​was to be used for another woman, as well as the decision made between the caliph and the matriarch without input from Utba, emphasizes the latter’s objectified status. In the third and final version of Utba’s story, preserved in al-​Husri’s Zahr al-​Adab, Abu al-​Atahiya asks his friend Yazid to talk to al-​Mahdi about Utba. The friend replies, “I cannot talk to him, but compose some poetry for me to sing.”26 The poet composes the two verses about al-​Mahdi as the caliph of God, cited in the first version discussed earlier. Al-​Mahdi hears the verses and wants to know who wrote them. The poet’s friend tells him, and al-​Mahdi says he will look into it. Yazid relays this to Abu al-​Atahiya, who waits a few months before returning to Yazid to ask for news. He presses Yazid to sing yet more verses and include another appeal to the caliph’s generosity. At this point, al-​Mahdi is ready to give Utba away with a gift for the bride and groom, but Utba says, “My [male] master (mawla) knows his obligation to my [female] master and I want to mention it to her.”27 Al-​Mahdi instructs her to do so. Yazid reports this to Abu al-​Atahiya. After a few days, the poet follows up, and, on Yazid’s urging, he composes the four verses from the second version of Utba’s story as described earlier. Yazid sings them and al-​Mahdi summons Utba, who says that her (female) master does not like it but the caliph can do as he pleases. Al-​Mahdi says, “I won’t do anything she does not like.”28 Yazid notifies Abu al-​Atahiya, who composes three new verses about giving up and being cheated. These last verses document the poet’s failure to win his beloved, despite his success in winning the monetary reward. In the first version, from the Kitab al-​Zahra, the caliph reacts to Utba’s protest by refusing to give her away. In the second version, which occurs in the Kitab al-​Aghani, the matriarch sides with the caliph in opposing the poet’s request. And, finally, in the version in the Zahr al-​Adab, first Utba, then Utba and the matriarch, and finally the pliant caliph, refuse to give Utba away. The caliph, despite his being head of the household and Utba’s master, will not oppose the women, just as he would not oppose Utba in the version in the Kitab al-​Zahra. Each of the versions of the story rewrites the negative reaction to the match in a way that, on the one hand, leaves the poet pining and, on the other, suggests the constraints on the caliph’s authority. The agents of this limitation on al-​Mahdi’s authority are the women who gossip, and either Utba alone or Utba and the matriarch. Thus, the story is about the subjective and even authoritative interventions of the slave woman in her enslavement as well as the broader issue of the limits placed on the caliph’s control as a patriarch and slave owner. As al-​Husri’s version continues, the marriage is off and the story of Abu al-​Atahiya’s encounter with the caliph turns ugly.29 Al-​Mahdi orders the poet to be lashed and exiled for reciting a line in which a gazelle belonging to the caliph pursues the poet. This was presumably the verse that the gossiping women deemed to be (unsolicited) love poetry for Utba. The poet Abu Duhman comments, “If not for the fact that the caliph beats lovers /​I would reveal the name of the one that I love, but I am a man who cowers in fear.”30 The view of this second poet lends broader significance to the story of Utba in reminding us we are dealing with a study of the ruler’s coercive conduct and attempts by his subjects

286 i

Jocelyn Sharlet

to limit his decision-​making powers. But, the poet’s comment refers as well to the refined code of love in which secrecy takes priority. The comment offers an interpretive frame for the women’s rumors and Utba’s affair with Abu al-​Atahiya. Al-​Husri’s version goes on to explain that al-​Mahdi’s son, Harun al-​Rashid, facilitated an affair between Utba and the poet following the latter’s return from exile, thus reviving the love about which the women had gossiped. Utba’s refusal of Abu al-​Atahiya thus fits into the refined custom of keeping love secret even when it required separation and, ironically, through her rejection of the poet on grounds that he was an unsuitable lover.31 In al-​Husri’s version, Utba displays subjectivity, first, in refusing the match and, second, when it is revealed she had conducted an affair with the poet and kept it secret. The limits of the caliph’s power over her are laid bare, and he appears as a cuckold. Scholars have identified, in Abbasid literature on Umayyad-​era lovers, an ideal form of chaste love as a pastoral affair that begins in childhood between a boy and a girl. Their families or tribes prevent their union, leading to obsessive but innocent love, sometimes with a caliph or other official hovering in the background.32 Again, as noted earlier, stories of this sort seem rooted in anxieties over new ways of thinking about individuality in social life.33 Likewise, the stories of slave women like Utba build on this representation of the social dislocations of Islamic conquest to explore related social dislocations of urbanization and new gender roles for elite slave women and the men who own them or seek them out. Al-​Husri’s version has the secret affair between Utba and Abu al-​Atahiya continue until al-​Mahdi learns of it. He confronts Abu al-​Atahiya about his indirect reference to al-​Mahdi’s mother, which, for the caliph, is clearly offensive. The poet tries to make up by offering an ode that begins with the well-​known line, “The caliphate came to him, trailing her robes.” Once again, the poet gets into an ambiguous situation with al-​Mahdi by way of poetry about “his” women—​an indirect reference to his mother, and then a direct reference to the caliphate as bride. Al-​Mahdi has the poet first beaten then rewarded for the words of praise or, alternatively, rewarded only for the praise; violence and reward were considered analogous forms of coercion. Abu al-​Atahiya escapes the beating but retreats from composing poetry on themes of praise, invective, and love, and shifts to ascetic poetry. The juxtaposition of the parts of al-​Husri’s version of the story enhances Utba’s status by comparing her both with the caliph’s mother and the caliphate, as if to suggest she is the source of his power. It is important to keep in mind that status is a matter of appearances and, in this case, it is in the play of appearances. First, the caliph derives status as a refined person in preparing to give her away to the poet in exchange for (solicited) love poetry about her. Second, he derives his status as a representative of patriarchy from his refusal to give her away to the poet after he learns the women had spread rumors. Third, he gets more credit as a refined person for being ready to give her away at the cost of his (divine) legitimacy.



Educated Slave Women and Gift Exchange in Abbasid Culture

j 287

Utba, for her part, displays her subjective resistance to the caliph in three ways. She consults her female master about the match when the caliph is ready to go through with it; participates in a secret affair with the poet; and, finally, acts to keep it secret. He may be God’s caliph, but he is not omniscient, and her secrecy trumps his divine power. Al-​ Mahdi was ready to give her up, just as the poet was ready to give up the secret by reciting the requested love poetry in exchange for Utba. But, Utba takes the lead in reasserting the code of love. In this way, and only in this way, could she be a star in a way that transcends her enslaved status. The elaboration of a code of love in the urban court occurred in conjunction with a growing role for elite slave women in Abbasid court writing. Elite slave women such as Utba, separated from their male kin and from tribal or communal ties, were on their own in their limited urban and cosmopolitan networks when it came to negotiating for power in relationships with masters and lovers alike—​hence the crucial role for the code of love in a spectacular yet private performance of individual desire. Following Utba’s lead, al-​Mahdi maintains the code of love by submitting to her objection at the cost of his legitimacy as a caliph of God. The problem with the poet is not that he is a jar seller and not even that he is haggling over money; he fails to get Utba because of his willingness to give up the secret. Al-​Rashid not only facilitates the “secret” love affair, he also imprisons Abu al-​Atahiya for refusing to compose love poetry, the poet having now turned solely to ascetic verse. Al-​Rashid says, “Al-​Mahdi forbade you from composing love poetry but you insisted, and now I order you to compose it and you refuse.”34 The two Abbasids effectively undermine one another. Abu al-​Atahiya points out that he was young then but has grown older, and sends poetry on this theme from prison until al-​Rashid releases him. What sounds like a crazy contradiction is actually an articulation of the anxiety about individualism expressed in the discourse both on love (the code of love) and ascetic practice, which came to use the language of love. Both seem to have been perceived as challenges to political authority if for different reasons. Al-​Rashid’s facilitation of the secret affair, now revealed, complements his reputation for a luxurious lifestyle and support for poets. He is also infamous for his violence toward the Barmakid family of ministers, and Jafar ibn Yahya al-​Barmaki in particular, despite the fact that Jafar’s mother nursed both men, which was understood as a form of foster brotherhood.35 This dark side to al-​Rashid provides a context for his decision to punish the poet for choosing ascetic over love poetry. In addition to her role in articulating the perception of male political elites, Utba is crucial to the perception of Abu al-​Atahiya, a major and—​in many ways—​transgressive court poet.36 His love poetry for Utba and his ascetic poetry demonstrate resistance to political authority, yet the poetry appears in the context of Utba’s secret love affair with the poet and her rejection of him. The poet’s rejection of love poetry in Utba’s story conceals yet another form of resistance to authority. This is not a focus of the story itself, but appears in the poet’s biographies: the accusation of holding heterodox beliefs expressed

288 i

Jocelyn Sharlet

in themes of his ascetic poetry and his conduct. Utba’s cooperation and resistance to the Abbasids intersects with the poet’s cooperation and resistance to them. Utba’s objectified status as a commodity—​unexchanged—​merges with her subjective status when she rejects the suitor. Thus, Utba is both a subjective authority on refined love and a negotiator of the conditions of her relationships to master and poet alike, and an objectified body the poet seeks in exchange for poetry and a symbolic material object, and for which he receives a gift of money as compensation. Her story with the poet, the caliph, and the latter’s son may be intended as a lighter, shorter version of contentious developments in poetry as a marketplace of ideas. More important, in the long run, Utba’s story works through ideas about the subjectivity, objectification, and public role of elite slave women. Inan and the Gift Taken Back The competition between male elites over Inan (d. 841), an episode in which she plays an active role, features her master, the merchant al-​Natifi, alongside the poet Abu Nuwas (d. 814), and Harun al-​R ashid, as caliph. As in the story of Utba, the account revolves around an elite slave woman, her master, and the transgressive poet whom she loves and who seems to love her back, although he falls short of her expectations. The gifts in this story are rings exchanged by Inan and her lover Abu Nuwas. The exchange of gifts merges with competition between elite actors. Inan demonstrates her active role by using poetry to criticize her master, articulate a professional and personal relationship with Abu Nuwas, and perform a working relationship with other poets. As a slave and victim of violence in the households of the merchant and later the caliph, Inan is objectified; her attempt to initiate relationships with Abu Nuwas and Harun al-​R ashid display her subjectivity. As in the case of Utba and Abu al-​Atahiya, the illicit activities of Inan and Abu Nuwas, on the one hand, serve as reminder of the patriarchal power of merchant and caliph alike, and, on the other, the limits of that authority, while the merchant’s resistance to the caliph’s quest for Inan undermines the caliph’s authority further. The main biographies of Inan appear in al-​Isbahani’s Book of Songs and at the beginning of his shorter work, al-​Ima al-​shawaʿir (Slave Women Poets). The texts make clear her successful career as a poet. The two biographies depict her experience as an objectified slave as well as her subjective interventions, often in the form of poetry, against the authority of her male handlers and their competition for her. As in the case of Utba and Abu al-​Atahiya, gift exchange provides the setting for the display of these themes. Described as being of mixed non-​Arab and Arab descent, Inan was educated in Arabia before her purchase by al-​Natifi. The latter subjects her to beatings; the portrait of her in Slave Women Poets also refers to threats by the merchant master against any who sought to visit her and the despair it caused her. Her talent and his marketing of it soon attracted the attention of the libertine wine poet Abu Nuwas, who exchanged poetry with Inan as



Educated Slave Women and Gift Exchange in Abbasid Culture

j 289

well as affection and playful gifts. She pursued the caliph Harun al-​Rashid successfully as well, although he was only able to purchase her after the death of her master.37 Al-​ Isbahani asserts in the shorter of these two works that Inan was the first person to gain preeminence for poetry during the Abbasid period.38 His biography of her in the much longer Book of Songs describes her as a successful competitor with the best of poets.39 It continues with obscene poetry that Abu Nuwas composed for her and to which she responded, although they also exchanged love poetry. The collection of anecdotes about Abu Nuwas by Abu Hiffan (d.c. 870) mentions her in passing, but there she appears as only one of his many female interlocutors. The most substantial anecdote concerns a playful exchange of obscene poetry in which Abu Nuwas conceals his identity from her.40 The suggestion is that Inan was not very important to Abu Nuwas, and that their relationship turned mostly on the exchange of ribald verse. The reputation of most women poets was based not on extensive compositions, but on the ability to improvise with skill. The sources depict their poetry in the context of interaction with men, and this is true of Inan as well.41 Inan, although a major figure of early Abbasid culture, was also objectified as a carefully guarded commodity in her master’s house. In her transgressive exchange of obscene poetry with Abu Nuwas, we encounter her simultaneously as both subjective and objectified. Abu Nuwas heard of the threat by her master, al-​Natifi, that he would attack anyone who visited her without his approval, and responded with verses at once wistful and defiant. Meanwhile, al-​Natifi invited the more conventional poet, Marwan ibn Abi Hafsa, announcing that he was the best of poets. Thus, the problem for the master is not his slave woman’s career as a poet, but his need to control it. Marwan, finding Inan ill, composes get-​well verses to which she responds with verses of her own decrying her master’s violent treatment. Marwan observes: “Inan cried, her tears flowing like pearls falling from a string.” She responds: “Let the right hand of the one who oppressively beats her dry upon his whip.”42 Likewise, when Abu Nuwas was finally allowed to see her, he finds her crying, locked to an iron rail. When the master orders Abu Nuwas to compose something, the lovers exchange a few wistful verses referring to religion in which they imagine how an observer would react to her unfortunate situation. Inan’s embodied and objectified status intersects with her articulate and subjective status, in Abu Nuwas’s obscene poetry and her response, as well as the exchange with Marwan about her master’s violence. It seems that the master not only wants to publicize her talent but also wants to make a spectacle of his own violence against her. His conduct objectifies her and undermines her subjectivity, but also provides the context for her articulate resistance. These parallel encounters with poets demonstrate that, although Inan’s master resented the playful and often obscene poet Abu Nuwas, whom he viewed as a rival, he was also eager to publicize his refinement as Inan’s master. He does so by arranging for an exchange of poetry with the panegyric poet Marwan ibn Abi Hafsa and the love poet Abbas ibn al-​Ahnaf. The range of poets with whom she exchanged verse speaks to her ability to maneuver in different genres of Arabic poetry. Although Inan, like most women poets,

290 i

Jocelyn Sharlet

interacts mainly with men in our sources, she also has verses in which she is said to have borrowed motifs from the Abbasid princess and poet Ulayya bint al-​Mahdi, who was known in part for her relations with a male slave (Bunan).43 Inan’s subjective experience appears in her poetry, even in the exchanges arranged by her master and including the obscene verse sent to Abu Nuwas. Inan’s use of verse in resisting patriarchy complements her subjective agency in the give and take of material gifts. Abu Nuwas uses verse to resist Inan’s master, but nothing changes except that his intervention highlights the limits of the master’s power. Inan, for her part, takes matters into her own hands by giving Abu Nuwas a ring with a red stone as a token of her transgressive love for the (transgressive) poet. The exchange of rings, however, takes on a somewhat ridiculous cast. Abu Nuwas sends the same ring to a male friend, Ahmad ibn Khalid Haylawah, Inan asks to have it back, and he responds by giving her a ring with a green stone. She thus displays the limits of her master’s power and that of the lover as well. When Abu Nuwas sends a long poem to his friend explaining his predicament, the friend returns the ring with the red stone along with two thousand dirhams. The ring is finally returned to Inan as requested, but that is the end of her love for him. Inan again exerts her own influence, but this time in a more serious vein. Al-​Isbahani records a letter in verse in which Inan seeks intercession to persuade Harun al-​Rashid to buy her.44 She is successful and he grows interested. Her merchant master, however, demands an exorbitant price. Although she suggests to al-​Rashid that he just pay the price, he says he could never satisfy the merchant. Meanwhile, al-​Rashid cites an obscene verse, most likely by Abu Nuwas, about whoever buys her and asserts that he cannot meet the price. The thought of submitting to the demands of a merchant was perhaps too much for the caliph to contemplate. It may also be that Inan’s reputation for exchanging obscene verse with Abu Nuwas undermined her chances with him. Although it was her idea to appeal to the caliph, the poetic power of the transgressive poet and the mercantile power of the merchant display the limits on the caliph’s political power and on Inan’s ability to determine her own fate. In the other “gift exchange” in this story, the merchant gives pious gifts in gratitude for keeping Inan from the caliph, a further act of circumscribing the caliph’s power. Thus, both the ring exchange with the poet and the merchant’s pious gift display Inan’s efforts to use new relationships to resist her master—​and the failure of these efforts. Al-​Rashid is only able to purchase her from the merchant’s estate. The anecdotes featuring Inan are of a piece with the extensive discourse about al-​ Rashid as a figure of luxury, pleasure seeking, and violence.45 The caliph is subject to the rule of law, in which the slave trade is legal, so he cannot just confiscate her, although he tries. Masrur, the caliph’s servant, beats Inan, then secures her purchase by the caliph from her master’s estate. She is effectively back where she started. Al-​Rashid’s Abbasid wife, Zubayda, a free woman, tries to distract al-​Rashid from Inan with poetry that she pays for to “cure” him of his love. One wonders if this helps to explain the caliph’s inability to come up with the money to pay for her.



Educated Slave Women and Gift Exchange in Abbasid Culture

j 291

In addition to her role in displaying the competition of mercantile and political elites, Inan contributes to the depiction of Abu Nuwas and his work in love and obscene poetry, although the relationship was not significant for him.46 The transgressive nature of her work as a female poet complements the better-​known work of Abu Nuwas in poetry about love, sex, and wine. The comic “exchange” of rings between them echoes the homoerotic and heteroerotic poems by and stories about Abu Nuwas. Inan’s story positions the playful poet as a contender of sorts with the mercantile and political elites of the time, and shows how he moved along the margins of elite social life even as he defined central features of that society in his poetry. In the case of Inan, the eloquent poet, one is reminded of the limits of discourse. Despite her eloquence, she remains with the violent master, fails to secure the devotion that she feels for her lover, fails as well to convince the caliph when she appeals to him, and finally moves to a new abusive environment in the palace. Oppressed or rejected by her master, her poet lover, the caliph’s servant, and his jealous wife, Inan becomes a pawn at the center of conflicts between the different elite actors. But she manages, nonetheless, to achieve fame as a poet and pursues the people that she desires. She remains defiant despite her misery. Conclusion Gift exchange in the stories of Utba and Inan is a material expression of the value of each slave woman. In addition, their role at the moment of gift exchange augments each woman’s subjectivity. In the context of these stories, gifts cannot be read at face value; the story infuses the gift with the subjectivity of the people who take part in the act of exchange. The merchant’s pious gift in gratitude for keeping Inan despite her appeal to the caliph is ironic, given his cruel treatment of her. The stories also help to define the transformation of genres in the work of Abu al-​Atahiya and Abu Nuwas, and the significance of these genres in social life. Some of their more transgressive poetry presented an alternative to the literary market for political patronage, although their participation in literary patronage made these alternatives possible. The level of violence in these stories of elite slave women and their male handlers can be shocking, but it, too, should not be read at face value. The physical abuse of individual people offers a critical perspective on social hierarchy and the abuse of power in general. Abu al-​Atahiya fell victim himself to violent beatings and incarceration by al-​Mahdi and al-​Rashid. Inan was not only beaten by her master, but by al-​Rashid’s servant as well when he finally purchased her. The slave women, however, manage to assert their subjectivity despite the various forms of coercion. Utba conducted an affair with Abu al-​Atahiya and expressed authoritative views on the code of refined love in her rejection of the poet. Inan, for her part, composed poetry about the beating, exchanged obscene poetry as well as rings with Abu Nuwas, and it was she who sought out al-​Rashid. Although these stories are, in their way, light interludes that feature slave women and transgressive poets,

292 i

Jocelyn Sharlet

the scenes of violence suggest a dark side to the celebration of refinement in Abbasid culture.47 The development of women’s subjectivity within the exploitative context of slave labor in these stories of educated slave women and gift exchange help shed light on the relationship between the concepts of slavery and freedom. In Islamic law, freedom was understood as the opposite of slave status beginning during the eighth century.48 Jurists understood freedom as a default setting for human beings, but they also believed that free status could be lost. In this way, they fashioned a legal system in which fundamental freedom and slavery coexisted. The stories of educated slave women and gift exchange discussed here display the role of education in women’s subjectivity and resistance to the obligations of slavery, which can be understood as a dimension of the fundamental freedom of human beings. Thus, the education that makes these women highly valued slaves is also the key to understanding their capacity to redefine themselves. Roy Mottahedeh has articulated the crucial role of patronage in hierarchy and social mobility in premodern Islamic societies, especially for highly skilled men.49 Other scholars have investigated the role of client relationships in hierarchy and social mobility for non-​Arabs in early Islamic society.50 In contrast, Mohammed Ennaji focuses on the importance of slavery for understanding authority in premodern Islamic societies.51 Protégés, highly skilled slaves, and clients often exchanged labor for financial support and status in a bond between individuals. In all three cases, the sponsor, patron, or master depended on this skilled labor for status and social networking among elite peers. In narratives of patronage, clientage, and slavery—​and in the language used to depict each type of relation—​the three systems sometimes seem to overlap, with the important distinction that a protégé or client could seek out or terminate a relationship whereas a slave did not enjoy this privilege. All three systems are ways to integrate new talent into elite society. Existing elites maintain control of newcomers, but they also must negotiate their own status and that of their subordinates. The importance of educated slave women, so much more extensive than the depiction of educated free women in our sources, suggests that slavery is crucial for understanding not only authority, but also social mobility circumscribed by slavery in premodern Islamic societies. The stories about Utba and Inan help to frame the tensions, conflicts, and crises among male elites, along with the public perception of these men. In their participation in gift exchange and love affairs, and their use of poetry and rhetoric, the women carve out a subjective role despite their objectification in and by elite male networks. They delineate the limits of power in patriarchy, and thus demonstrate that they, too, are agents of history. Notes 1. Schneider, “Freedom and Slavery,” 353–​382, 361. 2. Schneider, “Freedom and Slavery,” 377. 3. Schneider, “Freedom and Slavery,” 373, 376. 4. Athamina, “How Did Islam Contribute,” 383–​408, 391.



Educated Slave Women and Gift Exchange in Abbasid Culture

j 293

5. Athamina, “How Did Islam Contribute,” 392. 6. Ali, Marriage and Slavery, 50. 7. Ali, Marriage and Slavery, 52. However, the owner of a concubine was allowed to practice withdrawal for birth control without the concubine’s permission, although it was illegal to do so with a wife (Katz, “Concubinage, in Islamic Law”). 8. Athamina, “How Did Islam Contribute,” 389–​390. 9. Athamina, “How Did Islam Contribute,” 400–​404. 10. Compare the slave women’s participation in gift exchange with the widespread practice of men giving slave women as gifts. See Gordon, “Arib al-​Ma’muniya,” 86–​100, esp. 88. 11.  On the relationship between fact and fiction in such stories, see the essays in Section B (“Story-​Telling in Adab Literature and Theory”) in Leder, Story-​Telling, 61–​263. 12.  The use of writing was more extensive in the wake of the acquisition of Chinese paper technology in Central Asia during the late eighth century. El-​Hibri, Reinterpreting Islamic Historiography, 7. On the interconnectedness of the oral and the written, see Schoeler, Oral and the Written in Early Islam. 13.  On the use of narratives about historical characters to articulate arguments about culture and society and the development of adab, see Khalidi, Arabic Historical, 83–​110; and Bray, “ʿAbbasid Myth and the Human Act.” 14. Jayyusi, “Umayyad Poetry,” 387–​426; Bauer, Liebe un Liebesdichtung, 45–​47. 15.  On the representation of women in medieval Arabic sources, see Meisami, “Writing Medieval Women,” 47–​88. On slave women and their men, see Bray, “Men, Women and Slaves,” 121–​146. 16. Bray, “Men, Women and Slaves,” 141. 17. Al-​Isbahani, al-​Zahra, 2:744–​753. 18. Al-​Isfahani, Kitab al-​Aghani, 4:3, 4:5. 19. Al-Qayrawani al-​Husri, Zahr al-​Adab, 1:326–​330. 20. Al-​Isbahani, Kitab al-​Zahra, 2:744. 21. Al-​Isbahani, Kitab al-​Zahra, 2:744–​745. 22. Al-Isbahani, Kitab al-​Zahra, 2:745. 23. Al-Isbahani, Kitab al-​Zahra, 2:745. 24. Crone and Hinds argue that “Deputy of God” was, in fact, the official title of the caliphs from the start, and it went along with the religious significance of the names they took, although the religious scholars argued that the title meant “Deputy [of the Prophet] of God” so as to assign the role of “deputy of God” to the function of religious scholars—​in effect, an argument for rule of law rather than divinely ordained rule by one man. Crone and Hinds, God’s Caliph, 4–​23. 80–​96. 25. Al-​Isfahani, Kitab al-​Aghani, 3:249. 26. Al-Husri, Zahr al-​Adab, 1:326. 27. Al-​Husri, Zahr al-​Adab, 1:326. 28. Al-Husri, Zahr al-​Adab, 1:327. 29. Al-Husri, Zahr al-​Adab, 1:326–​330. 30. Al-​Husri, Zahr al-​Adab, 1:327. 31. Ruqayya Khan has argued that the secret as a theme in a range of genres of literary and religious texts is a crucial way of articulating a sense of the self. Khan, Self and Secrecy. Also see Vadet, L’Esprit courtois. 32. Dayf, al-​Hubb al-​ʿUdhri ‘inda al-Arab.

294 i

Jocelyn Sharlet

33. Jayyusi, “Umayyad Poetry”; Bauer, Liebe und Liebesdichtung, 45–​47. 34. Al-​Husri, Zahr al-​Adab, 1:329. 35. See El-​Hibri, Reinterpreting Islamic Historiography. 36. On the development of ascetic poetry in the work of Abu al-​Atahiya, see Hamori, “Ascetic Poetry (zuhdiyyat),” 265–​274; and Rowson, “Abu al-​Atahiyah,” 12–​20. 37. Al-Isfahani, Kitab al-​Aghani, 23:92–​101. Also, see Ibn al-​Saʿi, Nisaʾ al-​Khulafaʾ, 47–​53; and al-​Suyuti, al-​Mustazraf min Akhbar al-​Jawari, 38–​46. Also see Caswell, Slave Girls of Baghdad, 56–​80. 38. Al-​Isbahani, al-​Imaʾ al-​Shawa’ir, 23–​44, 23. 39. Al-Isfahani, Kitab al-​Aghani, 23:92. 40. Abu Hiffan, Akhbar Abi Nuwas, 110–​111. 41. Bencheikh, “Les musiciens et la poesie,” 114–​152, 142; Kilpatrick, “Women as Poets and Chattels,” 161–​176, 163. 42. Al-​Isfahani, Katib al-​Aghani, 23:94. 43. Al-Isbahani, al-​Ima’, 40. 44. Al-​Isbahani, al-​Ima’, 43. 45. For the dynamics of these aspects of al-​Rashid’s identity, see El-​Hibri, Reinterpreting Islamic Historiography, 17–​58. 46. On the role of Abu Nuwas in love and obscene poetry, see Kennedy, “Abu Nuwas,” 121–​132 or the monograph by the same author, Abu Nuwas: A Genius of Poetry. 47. For the dynamics of the relationships of Rashid and Yahya and Jafar, see El-​Hibri, Reinterpreting Islamic Historiography, 17–​58. 48. Schneider, “Freedom and Slavery,” 356. 49. Mottahedeh, Loyalty and Leadership. 50. Bernards, “Contribution of Mawali and the Arabic Linguistic Tradition,” 426–​453. 51. Ennaji, Slavery.

Bibliography Primary Sources Abu Hiffan, Abdullah ibn Ahmad ibn Harb al-​Mihzami. Akhbar Abi Nuwas. Edited by Abd al-​ Sattar Ahmad Farraj. Cairo: Maktabat Misr, n.d. al-​Ghuzuli, Ala al-​Din Ali ibn Abd Allah al-​Bahaʾi. Matali` al-​Budur fi Manazil al-​Surur. n.e. Cairo: Maktabat al-​Thaqafa al-​Diniyya, 2000 [1419]. al-​Husri, Abu Ishaq Ibrahim ibn Ali al-​Qayrawani. Zahr al-​Adab wa-​Thamar al-​Albab. 2 vols. Edited by Ali Muhammad al-​Bajawi. Cairo:  Dar Ihya al-​Kutub al-​Arabiyya, Isa al-​Babi al-​ Halabi wa-​Shurakaʾuhu, 1953. Ibn Abi Usaybiʿa, Muwaffaq al-​Din Abu al-​Abbas Ahmad ibn al-​Qasim al-​Saʿdi al-​Khazraji. Uyun al-​Anba fi Tabaqat al-​Atibba. Edited by Muhammad Basil Uyun al-​Sud. Beirut: Dar al-​ Kutub al-​Ilmiyya, 1998 [1419]. Ibn al-​Saʿi, Taj al-​Din Abu Talib Ali ibn Anjab. Nisa al-​Khulafa al-​Musamma Jihat al-​Aʾimma al-​ Khulafa min al-​Harair wa’l-​Ima. Edited by Mustafa Jawad. Cairo: Dar al-​Maarif, 1968. Ibn Zubayr, Ahmad ibn al-​Rashid. Kitab al-​Dhakhaʾir wa’l-​Tuhaf. Edited by Muhammad Hamid Allah and Salah al-​Din al-​Munajjid. Kuwait: Dairat al-​Matbuat wa’l-​Nashr, 1959.



Educated Slave Women and Gift Exchange in Abbasid Culture

j 295

al-​Isbahani, Abu Bakr Muhammad ibn Dawud. Kitab al-​Zahra. 2 vols. Edited by Ibrahim al-​ Samarraʾi. Al-​Zarqa, Jordan: Maktabat al-​Manar, 1985 [1306]. al-​Isbahani, Abu al-​Faraj. Al-​Ima al-​Shawa’ir. Edited by Nuri Muhammadi al-​Qaysi and Yusuf Ahmad al-​Samarraʾi. Beirut: Alam al-​Kutub, 1984 [1404]. al-​ Isbahani, Abu al-​ Qasim Husayn ibn Muhammad al-​ Raghib. Muhadarat al-​Udaba wa-​Muhawarat al-​Shuara wa’l-Bulagha. 2 vols., n.e. Beirut: Manshurat Dar Maktabat al-​Hayat, n.d. al-​Isfahani, Abu al-​Faraj. Kitab al-​Aghani. 27 vols. Edited by Abd Ali Muhanna and Samir Jabir. Beirut: Dar al-​Kutub al-​Ilmiyya, 2002 [1422]. al-​Jawziyya, Ibn Qayyim. Akhbar al-​Nisa. Beirut: Dar al-​Marifa, 1997 [1417]. al-​Masudi, Ali ibn al-​Husayn ibn Ali. Muruj al-​Dhahab wa-​Ma`adin al-​Jawhar. 4 vols. Edited by Mufid Muhammad Qumayha. Beirut: Dar al-​Kutub al-​Ilmiyya, n.d. al-​Suyuti, Jalal al-​Din. Al-​Mustazraf min Akhbar al-​Jawari. Edited by Ahmad Abd al-​Fattah Tammam. Cairo: Maktabat al-​Turath al-​Islami, 1989. al-​Tanukhi, Abu Ali al-​Muhassin ibn Ali. Nishwar al-​Muhadara wa-​Akhbar al-​Mudhakara. 8 vols. Edited by Abbud al-​Shalji. Beirut: Dar Sadir, 1971 [1391–​].

Secondary Sources Ali, Kecia. Marriage and Slavery in Islam. Cambridge, MA: Harvard University Press, 2010. Athamina, Khalil. “How Did Islam Contribute to Change the Legal Status of Women: The Case of the Jawari, or the Female Slaves.” Al-​Qantara 28, no. 2 (2007): 383–​408. Bauer, Thomas. Liebe un Liebesdichtung in der arabischen Welt des 9.  und 10. Jahrhunderts. Wiesbaden: Harrassowitz Verlag, 1998. Bencheikh, J. E. “Les Musiciens et la poésie: Les Écoles d’Ishaq al-​Mawsili (m. 225 H.) et d’Ibrahim Ibn al-​Mahdi (m. 224 H.).” Arabica 22, no. 2 (1975): 114–​152. Bernards, Monique. “The Contribution of Mawali and the Arabic Linguistic Tradition.” In Patronate and Patronage in Early and Classical Islam, edited by Monique Bernards and John Nawas, 426–​453. Leiden: Brill, 2005. Bray, Julia. “ʿAbbasid Myth and the Human Act: Ibn ʿAbd Rabbih and Others.” In On Fiction and Adab in Medieval Arabic Literature, edited by Philip Kennedy, 1–​50. Wiesbaden: Harrassowitz, 2005. —​—​—​. “Men, Women and Slaves in Abbasid Society.” In Gender in the Early Medieval World East and West, 300–​900, edited by Leslie Brubaker and Julia M. H. Smith, 121–​146. Cambridge: Cambridge University Press, 2004. Caswell, Fuad Matthew. The Slave Girls of Baghdad: The Qiyan in the Early Abbasid Era. London: I.B. Tauris, 2011. Crone, Patricia, and Martin Hinds. God’s Caliph:  Religious Authority in the First Centuries of Islam. Cambridge: Cambridge University Press, 1986. Dayf, Shawqi. al-​Hubb al-​ʿUdhri ʿinda l-​ʿArab. Cairo: al-​Dar al-​Misriyya al-​Lubnaniyya, 1999. El-​Hibri, Tayeb. Reinterpreting Islamic Historiography: Harun al-​R ashid and the Narrative of the ʿAbbasid Caliphate. Cambridge: Cambridge University Press, 1999. Ennaji, Mohammed. Slavery, the State and Islam. Cambridge:  Cambridge University Press,  2013.

296 i

Jocelyn Sharlet

Gordon, Matthew S. “Arib al-​Ma’muniya: A Third/​Ninth Century Abbasid Courtesan.” In Views from the Edge: Essays in Honor of Richard W. Bulliet, edited by Neguin Yavari, Lawrence G. Potter, and Jean-​Marc Ran Oppenheim, 86–​100. New York: Columbia University Press, 2004. —​—​—​. The Breaking of a Thousand Swords: A History of the Turkish Military of Samarra (A.H. 200–​275/​815–​889 C.E). Albany: State University of New York Press, 2001. Hammond, Marle. Beyond Elegy: Classical Arabic Women’s Poetry in Context. Oxford: Oxford University Press, 2010. al-​Jawari, Ahmad ʿAbd al-​Sattar. al-​Ḥubb al-​‘ Udhrī:  Nash’atuhu wa-​Taṭawwaruhu. Beirut:  al-​ Mu’assasa al-​Arabiyya li’l-​Dirāsa wa’l-​Nashr, 2006. Jayyusi, Salma K. “Umayyad Poetry.” In Arabic Literature to the End of the Umayyad Period, edited by A. F.  L. Beeston, T. M. Johnstone, R. B. Serjeant, and G. R. Smith, 387–​426. Cambridge: Cambridge University Press, 1983. Joseph, Suad. “Connectivity and Patriarchy among Urban Working-​Class Arab Families in Lebanon.” Ethos 21, no. 4 (1993): 452–​484. Katz, Marion. “Concubinage, in Islamic Law.” In Encyclopedia of Islam, 3rd ed., edited by Kate Fleet et al. Leiden: Brill Publishing Online, 2015. Khalidi, Tarif. Arabic Historical Thought in the Classical Period. Cambridge: Cambridge University Press, 1994. Khan, Ruqayya Yasmine. Self and Secrecy in Early Islam. Columbia: University of South Carolina Press, 2008. Kilpatrick, Hilary. “Women as Poets and Chattels: Abu al-​Farag al-​Isbahani’s al-​Imaʾ al-​Sawaʿir.” Quaderni di Studi Arabi 9 (1991): 161–​176. Leder, Stefan, ed. Story-​Telling in the Framework of Non-​Fictional Arabic Literature. Wiesbaden: Harrassowitz Verlag, 1998. Meisami, Julie Scott. “Writing Medieval Women:  Representations and Misrepresentations.” In Writing and Representation in Medieval Islam, edited by Julia Bray, 47–​88. Abingdon, UK: Routledge, 2006. Mottahedeh, Roy. Loyalty and Leadership in an Early Islamic Society. Princeton, NJ: Princeton University Press, 1980. Myrne, Pernilla. Narrative, Gender and Authority in ‘Abbasid Literature on Women. Gothenburg: University of Gothenburg, Orientalia et Africana Gothoburgensia, 2008. Schneider, Irene. “Freedom and Slavery in Early Islamic Time (1st/​7th and 2nd/​8th Centuries).” Al-​Qantara 28, no. 2 (2007): 353–​382. Schoeler, Gregor. The Oral and the Written in Early Islam, translated by Uwe Vagelpohl, edited by James E. Montgomery. London: Routledge, 2006. Sharlet, Jocelyn. “Tokens of Resentment:  Medieval Arabic Stories about Gift Exchange and Social Conflict.” Journal of Arabic and Islamic Studies 11 (2011): 62–​100. Vadet, Jean-​Claude. L’Esprit courtois en orient dans les cinq premiers siecles de l’Hegire. Paris: Maisonneuve et LaRose, 1968. Van Berkel, Maaike, Nadia Maria El Cheikh, Hugh Kennedy, and Letizia Osti. Crisis and Continuity at the Abbasid Court: Formal and Informal Politics in the Caliphate of al-​Muqtadir (295–​320/​908–​932). Leiden: Brill, 2013.

15 Remembering the Umm al-​Walad Ibn Kathir’s Treatise on the Sale of the Concubine

Younus Y. Mirza

i  A cry rang out as the faithful gathered with the second of the Rashidun Caliphs, Umar ibn al-​Khattab (r. 634–​644). Alarmed, Umar sent one of his Companions to investigate. The man returned to report that the mother of a slave girl had been sold. Distraught, Umar called together the Muhajirun (the “emigrants” from Mecca) and Ansar (the residents of Medina). Within an hour, the premises were full. Umar began by praising God then asked if the Prophet Muhammad had advocated the separation of family. They swore that he did not. He then read from the Quran, reciting a verse that supported the bonds of family, then declared, “What greater breaking of kindred could there be than the sale of a mother of one of your children, [especially since] God has advanced [your] fortune!” The crowd replied that Umar should do what he thought was necessary, so he decreed that, throughout the realm, it was forbidden to sell the female slave who gives birth to her master’s child (umm al-​walad).1 The umm al-​walad played an important role in the development of Islamic conceptions of sexuality, religious authority, and humanity. In the treatise considered here, Ibn Kathir, the great hadith scholar and historian discusses the sale of the umm al-​walad at length. The umm al-​walad has often been marginalized, receiving only passive mention in modern studies dealing with slavery. But, as the story of the assembly demonstrates, the umm al-​walad demanded the attention of the second Caliph and played a significant role in the development of early Islam. The wives of many prominent early Muslims were ummahat al-​walad, and a number of caliphs were the offspring of these women. As reliance on slavery waxed and waned throughout Islamic history, jurists agonized over the status of the umm al-​walad, with 297

298 i

Younus Y. Mirza

some emphasizing her status as a mother, others her position as a slave. Ibn Kathir’s treatise is the most comprehensive discussion of the umm al-​walad to be published. The treatise demonstrates that the early Muslim community struggled to define the proper status of the umm al-​walad. The issue would divide Sunnis and Shi‘is; it also highlights Umar’s role in the development of Islamic law. Ibn Kathir’s treatise further represents the rich spectrum of debate that took place in Mamluk society (c. 1250–​1517) over the question of slavery. Muslim jurists of the period debated the umm al-​walad, with some arguing she should be free and others maintaining she was property. Many of the issues relating to the umm al-​walad continue to be relevant today. Muslims continue to debate religious authority, with some stressing the collective intellectual tradition, and others maintaining the absolute authority of the Prophet. The figure of the umm al-​walad is, furthermore, connected intimately to the still-​pertinent issues of polygamy, temporary marriages, and patriarchy. Modern jurists disagree over the number of sexual partners a Muslim may have and whether a relationship can be purely sexual or should contain rights and responsibilities. Examining the historic debates regarding the umm al-​walad can thus help us better understand issues regarding marriage and sexuality in contemporary Muslim societies. The Umm al-​Walad in Western Scholarship: The Significance of Ibn Kathir’s Treatise The figure of the umm al-​walad is overlooked frequently in Islamic studies scholarship although she played an important role in the development of Islam and received the attention of Muslim jurists throughout history. Joseph Schacht presents the most comprehensive discussion of the umm al-​walad by discussing how the concept existed within pre-​Islamic Arabia and how it changed gradually under Islam.2 Schacht deals at length with the various discussions regarding whether the Prophet Muhammad himself forbade the sale of the umm al-​walad, and concludes the traditions prohibiting the sale of the umm al-​walad originate from the second caliph, Umar. However, Umar’s opinion was not accepted universally; it gave “trouble” to his successor, Uthman, and Ali, the fourth of the Rashidun caliphs, eventually would overturn it. As the debate grew, “an attempt was made on the one side to ascribe Umar’s decision to the Prophet and to ascribe the same opinion even to Ali and Ibn Abbas.”3 In his analysis of the umm al-​walad traditions, Schacht relied on sources that drew on Ibn Kathir’s slave treatise. Schacht, for instance, builds on a discussion by al-​San‘ani (d. 1768) regarding transactions, which cites Ibn Kathir’s work.4 Schacht’s skepticism regarding prophetic traditions that deal with the umm al-​walad was connected to his wider skepticism regarding the hadith tradition. Schacht believed that a “growth of legal traditions” occurred between 767 and 867—​that is, the period between the life of Abu Hanifa (d. 767) and the appearance of the canonical hadith collections. Through his engagement with Muslim hadith criticism,5 Schacht contended



Remembering the Umm al-​Walad

j 299

“the best way to know if a tradition did not exist during a certain time is to show that it was not used as a legal argument in a discussion which would have made reference to it imperative, if it had existed.”6 In the case of the umm al-​walad, early authorities reference Umar’s opinion and not that of the Prophet Muhammad. Later traditions in which the Prophet prohibits the sale of the umm al-​walad should therefore be viewed as suspect. Schacht further believed that “countertraditions” were created to buttress specific opinions. These countertraditions would relay an opinion on the authority of the figure, who was, in fact, the opponent of that position. As Schacht hypothesizes, “a favorite device in the creation of counter-​traditions consists of borrowing the name of the main authority for, or transmitter of, the opposite doctrine.”7 In the case of the umm al-​walad, Ali and Ibn Abbas were recognized as early adversaries of manumitting the umm al-​walad, but nonetheless later traditions have them advocating her freedom. These later traditions should not be seen as authentic, but rather as attempts of supporters of the umm al-​walad to promote their position through their opponents—​in this case, Ali and Ibn Abbas. Schacht’s e silencio argument, and his ideas on countertraditions, would lead him eventually to shed doubt on the entire legal hadith corpus: “that every legal tradition from the Prophet, unless the contrary is proven, must not be taken as an authentic or essentially authentic, even if slightly obscured, statement valid for his time or the time of the Prophet, but as the fictitious expression of a legal doctrine formulated at a later date.”8 Schacht was too bold in putting the entire legal hadith tradition into doubt based on his select examples;9 more work needs to be done in examining the authenticity of the traditions.10 Nonetheless, his e silencio framework is helpful in analyzing the traditions of the umm al-​walad.11 Many early traditions, such as that of the umm al-​walad considered here, were statements of the Companions but were later attributed to the Prophet. As Jonathan Brown demonstrates, Muslim hadith scholars were very much aware that some traditions were being ascribed falsely to the Prophet. The backgrowth of isnads was a “recognized problem,” and hadith scholars “were aware that specific transmitters or scholars were elevating these non-​Prophetic reports to Muhammad.”12 Much of Western hadith scholarship has focused exclusively on the early period of Islamic history and, consequently, overlooked later hadith scholarship that identified this challenge.13 During the Mamluk period, scholars like Ibn Kathir argued that it was the opinion of Umar and not the Prophet that the umm al-​walad should not be sold and she was free upon her master’s death. Later traditions stating the Prophet had decreed the freedom of the umm al-​walad were spurious and should not be taken as authentic. The analysis of the various umm al-​walad traditions not only contributes to the study of hadith authenticity but also to early sectarianism. Umar’s opinion was upheld during his rule, was but then overturned by Ali, the first recognized Imam of the Shi‘i tradition. Other supporters of Ali, such as Ibn Abbas and Jabir ibn Abd Allah,14 also rejected Umar’s opinion, believing the umm al-​walad was like any other property, or that Umar acted without sufficient backing from the Prophet. However, Umar’s supporters, who

300 i

Younus Y. Mirza

may be labeled as “proto-​Sunnis,” either ascribed Umar’s opinion to the Prophet or insisted that Umar was acting on the basis of a prophetic injunction. This early sectarian divide is demonstrated by the fact that all mainstream Sunni legal schools adopt the position attributed to Umar whereas the Shi‘i schools accept that of Ali. Speaking about the early emergence of Shi‘ism, Najam Haider states, “while recent scholarship offers valuable insights into the importance of theology in the formative period of Shi‘ism (beginning in the eighth century), it overlooks the concrete manner in which that theology impacted the everyday lives of Muslims.”15 Slavery was a reality of Islamic societies, and Umar’s ruling on the umm al-​walad had tangible effects on the wealth of slave owners. Allegiance to his opinion became a marker of one’s support of his rule and source of religious authority. During the early period, Umar was identified as the advocate for the freedom of the umm al-​walad and, thereafter, Sunnis, who believed his rule and religious judgment to be correct, continued to ascribe to his opinion. Schacht did note that the sale of the umm al-​walad was a legal issue that separated Sunni and Shi‘i law but did not believe that the divide was sectarian. He records, for instance, that the Sunni legal school of the Zahiris also maintained that the umm al-​ walad could be sold. Schacht contended that when Shi‘i scholars began to formulate their own law, they adopted traditions ascribed to Ali that the umm al-​walad could be sold with the full realization that these statements were in opposition to that of Umar and his son, Ibn Umar. They subsequently incorporated Ali’s opinion and “borrowed” from early Sunni discussions regarding her status.16 In making these observations, Schacht did not have access to Ibn Kathir’s treatise (which was still in manuscript form) and was not aware of the full extent of the traditions ascribed to Ali, Jabir ibn Abd Allah, and Ibn Abbas. As Ibn Kathir’s treatise demonstrates, the beginnings of the Shi‘i doctrine can be found in the generation after the Prophet’s death, when Ali and some of his supporters began to challenge Umar’s decision. Nevertheless, more work needs to be done with early Sunni and Shi‘i historical and juridical literature to understand more fully how such a split occurred and in what contexts. Although there has been research on how Umar’s legal decisions caused early dissension,17 there has been only limited study on his ruling on the umm al-​walad. The bulk of the studies have dealt with his prohibition of temporary marriage (mut‘a), a still contentious issue between Sunnis and Shi‘is.18 A mut‘a marriage only required a fixed amount of time and a dowry to the bride. It thus differed from traditional marriages, which required witnesses and gave women the right to inherit from her husband.19 Umar famously forbade temporary marriages, likening them to fornication. His Shi‘i detractors claimed that Umar violated the Quran and the Sunna, whereas his Sunni supporters believed he was carrying out a prophetic order. Yet, scholars have not explored sufficiently the relationship between Umar’s prohibition of mut‘a and selling of the umm al-​walad. As Ingrid Mattson points out, the prohibition of the mut‘a marriage and the sale of the umm al-​ walad were part of Umar’s sexual politics, which sought to define more completely the relationship between partners—​a topic that deserves more research.20



Remembering the Umm al-​Walad

j 301

There are many similarities between the practice of mut‘a and sexual relations with female slaves. Early Shi‘i jurists were unapologetic in comparing the two, noting that one could engage in mut‘a as many times as one wished and take an unlimited number of concubines.21 The mut‘a wife and concubine did not inherit; but, if they had a child, then the owner was obligated to provide for him. Their waiting period before they could get married again or have sexual relations with another man was often identical.22 Both appear in legal discussions of coitus interruptus, in which the man tries intentionally to avoid having the woman become pregnant. In other words, the relationship between the man and his mut‘a wife or concubine was primarily sexual and did not necessarily envision the creation of a family. But, despite the similarities, there are some important differences between the two. A mut‘a wife was usually free and could accept or reject potential suitors. The dowry that came with the mut‘a was exclusively hers and was not shared with any guardian. In contrast, the female slave was considered owned and could not deny sexual access to her master unless she was married.23 Thus, the prohibition of mut‘a and sale of the umm al-​walad could be seen as an early attempt to limit a man’s sexual partners. Although proto-​Sunni scholars agreed that a man could continue to have sexual relations with his umm al-​walad, they contended that he could not sell her or exchange her for another female slave. This made it more difficult for a man to purchase a new concubine because his wealth was now tied up with his umm al-​walad.24 Mattson further argues that Umar’s ruling on the umm al-​walad highlighted the difference between free and slave women, and changed the power dynamics of the master–​ slave relationship. In pre-​Islamic and early Islamic Arabia, a clear distinction between free and slave women did not exist, and it was up to the discretion of the man to determine which of his slaves would become his “wife.”25 But this designation was not always well defined, especially regarding war captives. With regard to the umm al-​walad, the master designated initially whether his female slave could bear the title of the umm al-​walad and thus if she was free after his death. However, after Umar’s ruling, a female slave who bore the child of her master was given the title of the umm al-​walad automatically and the master was bound by law not to sell her. Although this shift of power to the slave woman was obviously limited (she continued to be a slave), it nonetheless triggered protest from male slave owners, who sought the unrestricted right to do as they pleased with their wealth. Nevertheless, the scholarship documents that these early discussions over whether the umm al-​walad could be sold eventually led to a Sunni consensus. Schacht, for instance, believed that “although the four Sunni madhahib [legal “schools”] in the end all declared that the umm al-​walad could not be sold, the existence of ijma [scholarly consensus] on this point is nevertheless sometimes doubted; sometimes, however, also definitely asserted.”26 A Sunni consensus did develop regarding the umm al-​walad, but this consensus was challenged and, at other times, maintained. In his study of early Maliki law, Jonathan Brockopp also observes an early consensus: “by the time Ibn `Abd al-​Hakam (d. 214/​829) wrote his text, consensus on the basic ruling had been reached. His position,

302 i

Younus Y. Mirza

that the umm al-​walad is freed on the death of the master, along with her children, is stated explicitly and without undue polemic.”27 Ibn Abd al-​Hakam did not debate the question of the eventual freedom of the umm al-​walad, but rather details “the rules for the master of an umm al-​walad and certain exceptional cases.” Similarly, Kecia Ali recommends that more work needs to be done on the early status of the umm al-​walad but nonetheless suggests that a consensus did emerge: “the early development of rules about the umm walad has not yet been adequately studied, but it is clear that the protections granted to her by formative-​period jurists were not unanimously upheld by earlier authorities, some of whom allowed her to be sold.”28 During the Mamluk period, the opinion that the umm al-​walad should not be sold and was free upon her master’s death continued to be the norm. Scholarship on the umm al-​walad in Mamluk times unfortunately remains sparse, although female slavery played an important role in society. In a survey article, Yossi Rapaport states, “There are good indications that the number of female slaves in elite households was always at least as high as, and probably much higher than, the number of male slaves, and it would make sense to view Mamluk slavery as a primarily female phenomenon.”29 The only scholar to discuss the umm al-​walad in Mamluk society is Shaun Marmon, who presents a helpful survey of the different types of slavery found in Mamluk period, the jurists’ view of slavery, and its social role within the Empire.30 In terms of the umm al-​walad, Marmon states the standard juridical view that if a man claimed the child of his slave women then she received the title of an umm al-​walad, her children would be considered free, and she would be manumitted upon her master’s death.31 She would, nonetheless, be considered the property of her master until that period and could be rented out for labor.32 Ibn Kathir’s treatise on slavery demonstrates that a group of Shafi‘i hadith scholars challenged this consensus at the beginning of the fourteenth century. As I have pointed out elsewhere,33 a Shafi‘i hadith movement emerged in the middle of the thirteenth century contesting mainstream Shafi‘i opinions based on their critical study of hadith. These hadith scholars maintained that the Prophet held more authority than any of the Companions and that his words and actions did not always substantiate later decisions. They were thus advocates of ijithad (independent reasoning) in that they sought to return to the original sources of the Quran and Sunna, and derive their own legal opinions.34 Although they had tremendous respect for earlier Shafi‘i authorities, such as al-​Rafi’i (d. 1226) and al-​Nawawi (d. 1277), they did not feel limited by them. Rather, they held that they had the authority to engage the primary sources in consultation with the great Shafi‘i jurists of the past.35 The Shafi‘i hadith scholars thus rejected the “judicial pragmatism” that developed in the Shafi‘i school in accepting traditions that were ascribed to the Prophet but had clear signs of being statements of Companions.36 They believed that by simply transmitting a hadith on an issue without properly examining its conflicting variants, jurists were enhancing their legal authority improperly by claiming endorsement by the Prophet. This “judicial pragmatism” is found, for example, in the work of Abu Bakr al-​Bayhaqi



Remembering the Umm al-​Walad

j 303

(d. 1066), who admits that the majority of reports regarding the freedom of the umm al-​walad originate from Umar, but nonetheless suggests there could be prophetic evidence supporting his ruling.37 Al-​Bayhaqi recognized that there were reports in which the Prophet  allowed her sale initially, but then contended that he outlawed the practice eventually. Ibn Kathir’s treatise could be read largely as a refutation of al-​Bayhaqi’s claim, in which he argues there is no evidence from the Prophet that supports Umar’s rulings explicitly. Although Ibn Kathir respected Umar’s decision on the umm al-​walad, he belonged to the “critical rigor” school that believed that no matter how beneficial a report attributed to the Prophet might seem, “It is not possible to say ‘from the Prophet, may the peace and blessings of God be upon him’ for what he did not say.”38 Ibn Kathir most likely felt compelled to respond to al-​Bayhaqi’s claim because the authoritative al-​Shafi‘i jurists of his day simply accepted the freedom of the umm al-​walad as prophetic. In his influential compendium Sharh al-​Muhadhdhab, al-​Nawawi adopts al-​Bayhaqi’s position that the Prophet forbade the practice eventually.39 Al-​Nawawi justifies the freedom of the umm al-​walad through the hadith narrated by Ibn Abbas in which the Prophet declares “whoever’s slave woman gives birth to [her master’s child] then she is free after his death.”40 Al-​Nawawi does not cite where he found this hadith, nor does he examine critically the role of Umar in the development of this ruling.41 By ignoring other nonprophetic traditions, al-​Nawawi was using “juridical pragmatism” in validating a standard ruling through a questionable hadith. In contrast, critical hadith scholar Ibn Kathir contends that prophetic hadiths mandating the freedom of the umm al-​walad were inauthentic and could only be understood as later fabrications. The Editing and Publication of Ibn Kathir’s Treatise One of the challenges of the literature on the umm al-​walad, and Islamic slavery in general, is that much of the material remains in manuscript form.42 The publication of Islamic manuscripts is connected intimately with the goals and objectives of modern Islamic movements. Various Islamic movements vied over religious authority and published different aspects of the Islamic intellectual tradition that was in accord with their intellectual projects. Many of the first publications were thus of legal compendiums, Quranic exegesis, and spiritual manuals that encapsulated the legal, moral, and theological dimensions of the Islamic tradition. The publications often sought to advance the goals of Islamic movements, rather than provide a historical view of the Islamic tradition. In the era of modernity, in which the reality of slavery was declining, many Islamic movements did not publish manuscripts that dealt specifically with slaves. This source problem arises in regard to Ibn Kathir’s slave treatise, which was not published until 2006.43 Ibn Kathir’s Quranic exegesis, in comparison, was first published in 1883 and has been published, abridged, and translated repeatedly ever since.44 The editor of the treatise, Umar ibn Sulayman Hafyan, explains he was not searching intentionally for the work, but rather stumbled over it when he was examining another text. Although

304 i

Younus Y. Mirza

he could not find another manuscript, he nevertheless decided to publish the treatise because of the mounting evidence that it was authentic.45 In his introductory section, Hafyan feels compelled to defend his publication and answers the hypothetical question regarding the wisdom of publishing a slave treatise when the system of slavery has been abolished. He argued the rules relating to slavery were part of the shari‘a of God and that jurists have devoted chapters to the topic in their legal compendiums. The fact that the system of slavery no longer exists does not mean these laws are abrogated; the jurists were known to research topics that did not have direct application in their times. Furthermore, in his view, slavery has been abolished for only half a century, which is not a long period in human history, and there remains the possibility that slavery may return at any time. This is especially true with the tremendous fluctuations in social and economic thought through which the world is going today. Hafyan’s comments here demonstrate the conflicted state of many modern Muslims regarding slavery.46 As Hafyan correctly noted, many legal compendiums devote chapters to slavery leading to the question on whether these rules are now obsolete. Hafyan contended that these rules were not applied because they have been abrogated or proved to be illegitimate, but rather because there is no slavery in the modern world. He thus left the door open that slavery could return and that the rules could potentially be valid. Hafyan proceeded to list other reasons for publishing the treatise, such as learning about Islamic intellectual heritage, Ibn Kathir’s life and thought, and traditionalist jurisprudence in general. By publishing the treatise, Hafyan hoped to give insights into the rich Islamic legal tradition and provide research material to specialists of Ibn Kathir. Hafyan makes no reference to the demands of the book market that frequently drive the editing and publication of Islamic manuscripts; he understood that the task he was undertaking was primarily academic and that he would most likely not reap extensive financial rewards for his project. The fact that Hafyan felt obliged to defend his work demonstrates that many editors and Islamic publishing houses do not have a strong incentive to publish slave treatises. Editors of other classical Islamic works, such as Quranic exegesis or spiritual manuals, begin their introductions speaking about the significance of the work rather than defending their publications. For instance, in the introduction to his abridgement of Ibn Kathir’s Quranic exegesis (tafsir), Ahmad Shakir exclaims that the work is among the best of all tafsirs and emphasizes the necessity for a work that makes the Quran accessible to the average reader.47 Shakir feels no need to defend his abridgement or argue for its relevance.48 There is little doubt that if it were not for the modern popularity of Ibn Kathir and the personal undertaking of Hafyan, the slave treatise would still be in manuscript form. The source problem is compounded further by the fact that Ibn Kathir drew on other works that are either lost or remain in manuscript. He refers directly to other slave treatises in his commentary on the Shafi‘i legal compendium al-​Tanbih, where he devoted an entire chapter to the “freedom of umm al-​walad.”49 He closes the chapter stating that



Remembering the Umm al-​Walad

j 305

“there is a narrated consensus on the prohibition of selling of [the umm al-​walad] by more than one of the later imams, and this is the practice of the people today, while others voice their disagreement [on whether she should be sold].”50 Ibn Kathir explains that the jurists maintain that a consensus exists that the umm al-​walad cannot be sold and this is the general rule of his day. Nevertheless, some postformative jurists, a group that includes his contemporaries, held there was disagreement on whether the umm al-​walad could be sold.51 Ibn Kathir then notes that more than one of the jurists composed treatises on the umm al-​walad and that he devoted a treatise to the subject, in which he summarized four positions of al-​Shafi‘i (d. 820) and eight positions in total.52 Ibn Kathir’s comments here are significant because they suggest that his slave treatise was one of many other works on the subject. Ibn Kathir’s treatise thus gives us a unique window onto Mamluk legal culture and society because it contains many of the views of his contemporaries, many of which are present in works that may never be found. A second source problem is that, although manuscripts of slave treatises do exist, they most likely were not copied as frequently as other works dealing with popular piety. Hafyan, for instance, mentions he was only able to find one copy of Ibn Kathir’s slave treatise whereas multiple copies of Ibn Kathir’s exegesis survived. One unpublished slave treatise that Ibn Kathir mentions explicitly is that of the Shafi‘i jurist and hadith scholar Ibn Sayyid al-​Nas. The biographical dictionaries also speak of Ibn Sayyid’s treatise on the umm al-​walad.53 The Shafi‘i biographer Ibn Qadi al-​Shuhba (d. 1448), for instance, notes that, along with composing a biography of the Prophet Muhammad and a commentary on the hadith collection of al-​Tirmidhi, Ibn Sayyid al-​Nas wrote a significant treatise on the sale of the umm al-​walad. The treatise was gigantic, Ibn Qadi al-​Shuhba exclaimed, and in it, he notes, Ibn Sayyid demonstrated his vast knowledge. Ibn Sayyid’s slave treatise has not been found, however, and the bulk of his published works are those relating to the Prophet Muhammad’s life.54 These writings may have been more popular to both classical and modern readers, who would be more interested in the legacy of the Prophet rather than legal rulings concerning slaves. Nonetheless, more work needs to be done with the manuscript tradition to learn about slavery within Islamic societies. Ibn Kathir’s Slave Treatise: A Summary Ibn Kathir’s treatise is composed of an extended summary treatment of eight questions55 regarding the umm al-​walad and an appendix in which Ibn Kathir discusses miscellaneous issues regarding her status.56 Ibn Kathir’s approach is thus different from that of Ibn Sayyid al-​Nas, who argued for a particular position. Ibn Kathir was more interested in summarizing the different positions, which he subjects to critical evaluation, adding his own opinions frequently. Here, only the debate over the first two questions is discussed. The questions are, first, if the umm al-​walad should be manumitted upon her master’s death and, second, if she can be sold unconditionally.57 Ibn Kathir devoted the great majority of the work to the

306 i

Younus Y. Mirza

debate on these two questions, demonstrating their significance to the study of early Islamic history and its relevance to Mamluk society. Ibn Kathir’s main argument is that there are no Quranic verses, authentic hadiths, and binding consensus that mandate the freedom of the umm al-​walad. The Companions disagreed over her status, with some arguing she was free after her master’s death whereas others contended she was like any other property. The opinion that the umm al-​walad could not be sold and was free upon her master’s death is to be attributed to Umar, and the major Sunni legal schools incorporate it into their authoritative opinions. Ibn Kathir begins the discussion over the prohibition of the sale of the umm al-​walad by explaining that this is the dominant opinion originating from the second caliph, Umar ibn al-​Khattab.58 The third caliph, Uthman ibn Affan, held the same opinion, whereas the fourth caliph, Ali ibn Abi Talib, only did so at the beginning of his rule. This is also the opinion of famous Successors such as Hasan al-​Basri (d. 728), al-​Zuhri (d. 742), and Umar b. Abd al-​Aziz (d. 720); and the formative jurists Abu Hanifa (d. 767), Malik (d. 796), al-​Shafi‘i, and Ibn Hanbal (d. 855). Later jurists also claimed a consensus on the issue such as al-​Bayhaqi (d. 1066), Ibn Abd al-​Barr, al-​Ghazali (d. 1111), and al-​Baghawi (d. 1122).59 Ibn Kathir proceeds to explain the argument that the umm al-​walad cannot be sold. He draws on the Quran, the Sunna, and consensus. As for the evidence from the Quran, Ibn Kathir cites verse 47:22: “If you turn away now, could it be that you will go on to spread corruption all over the land and break your ties of kinship?” It was this verse, as we saw in the opening section of this discussion, that Umar ibn al-​Khattab used to argue against the sale of the umm al-​walad on the grounds that it broke up the slave’s family.60 The incident at the assembly, in which the caliph (Umar) issued his edict, demonstrates that it was Umar, and not the Prophet Muhammad, who explicitly outlawed the sale of the umm al-​walad. He applied a general Quranic principle on the unity of families to the case of umm al-​walad. There was consensus that families should not be broken up, and thus it followed that the umm al-​walad must not be sold.61 Other proponents of the freedom of the umm al-​walad use the Quranic verse (4:59) “Obey God and Obey the messenger and those who have authority among you” as evidence that Umar’s declarations should be followed. Umar was the caliph at the time, and it was incumbent on the believers to follow his order that the umm al-​walad should be freed. This argument demonstrates that many of the advocates of the freedom of the umm al-​walad were also followers of Umar because they believed in his temporal and religious authority. As for the Sunna (prophetic authority), Ibn Kathir cites a series of hadiths in which the Prophet decrees the freedom of the umm al-​walad. Ibn Kathir, however, finds all of these hadiths problematic, believing ultimately that the opinion originated from Umar and not, in other words, the Prophet. The first hadith is narrated by the great early hadith scholars Ibn Hanbal and Ibn Maja on the authority of Ibn Abbas that, “Any man whose slave women bears his offspring, she is free after his death.”62 Ibn Kathir,



Remembering the Umm al-​Walad

j 307

however, finds this hadith inauthentic because one of its narrators, Hasan b. Abd Allah, is unreliable.63 The second hadith is narrated by Ibn Maja, once again on the authority of Ibn Abbas. Ibn Abbas records that the Prophet’s umm al-​walad, Mariya—​the mother of the Prophet’s infant son, Ibrahim64—​was mentioned in the presence of the Prophet and he stated, “Her child freed her.” Ibn Kathir finds this hadith problematic because it also cites Hasan ibn Abd Allah in its chain of transmission and his questionable student Abu Bakr ibn Abi Sabra.65 Nevertheless, Ibn Kathir narrates a second version of this hadith, which does not cite Hasan ibn Abd Allah. Ibn Kathir believes that this second version, without Hasan ibn Abd Allah, is authentic. Later, we will see that many early Muslims did not find this hadith to apply to all umm al-​walads, but rather just to Mariya. Mariya was a “special case,” in other words, and how the Prophet treated her should not be generalized to others. Ibn Kathir closes the discussion on this second hadith by relaying another variant narrated by Umar from Abi Shayba’s Musannaf. In this variant, Mariya is not mentioned and, in it, the Prophet makes a general proclamation regarding the umm al-​walad: “Her child freed her, even if [the child] is stillborn.”66 Ibn Kathir agrees with al-​Bayhaqi that this statement is not that of the Prophet, but rather that of the caliph Umar. As for the third hadith, al-​Daraqutni67 (d. 995) narrates that the Successor, Said ibn Musayyab (d. 94/​712–​713) contended that Umar freed the umm al-​walads and justified the practice by saying, “The Messenger of God freed them.”68 Ibn Kathir finds this hadith problematic because one of its narrators, Abd al-​Rahman al-​An‘am, is considered weak by many hadith authorities, including al-​Bukhari (d. 870). More important, Said al-​ Musayyab never met Umar, which raises further doubt regarding the hadith’s authenticity. Another variant of this tradition states that Said ibn Musayyab was asked about freeing the umm al-​walad, to which he responded, “People say that the first who ordered to free the umm al-​walad was Umar ibn al-​Khattab but that is not true. The first to order the freedom (of the umm al-​walad) was the Messenger of God.”69 Similar to the first variant, Said ibn Musayyab believed that Umar was acting on a prophetic injunction, not only his opinion.70 Ibn Kathir notes that some of the narrators of this tradition are unreliable, but the biggest problem is that the tradition is mursal, or a “report attributed by a Successor to the Prophet directly without the authority of a Companion.”71 There is no link between Said ibn Musayyab and the Prophet himself, shedding doubt on how Said could have known that Umar was acting on a prophetic order.72 The fourth hadith is relayed on the authority of Ibn Umar in which the Prophet prohibited the sale of the umm al-​walad by saying, “Do not sell them, do not gift them, do not make them part of your inheritance; the master can take pleasure from her as long as he is alive and when he dies she is free.”73 However, other traditions reported by al-​ Daraqutni and al-​Bayhaqi have Ibn Umar relay this tradition from his father Umar, not the Prophet Muhammad. Ibn Kathir believes that the traditions ascribed only to Umar are the most correct. To support his argument, Ibn Kathir cites the tradition found in the

308 i

Younus Y. Mirza

Muwatta of Malik in which Ibn Umar reported that his father Umar said, “Any slave girl who bears (a child) from her master, (the master) must not sell her, give her as a gift, not include her as part of his inheritance. He can take pleasure from her and when he dies she is free.”74 This tradition, almost identical to the one relayed earlier, is only narrated from Umar. In addition, similar to Malik, al-​Shafi‘i states that it was the practice of Umar (taqlid Umar) that the umm al-​walad could not be sold or given as a gift and that she is free after her master’s death.75 Thus, the early legal authorities of Malik and al-​Shafi‘i both have only Umar’s decree that the umm al-​walad should not be sold. Ibn Kathir uses al-​Shafi‘i’s statement to argue there are no authentic hadiths that state that the umm al-​walad should not be sold and become free on her master’s death; if al-​Shafi‘i had known of authentic hadiths on the subject, he would have cited them. However, because al-​Shafi‘i cited only the opinion of Umar, then the various traditions that are ascribed to the Prophet Muhammad on the matter must be read with suspicion. Ibn Kathir’s argument here is remarkably similar to Schacht’s e silencio argument—​that is, if earlier authorities do not cite a prophetic hadith on a topic, then later hadiths on the subject must be fabricated.76 Ibn Kathir thus ends the section on the Sunna declaring that that none of the hadiths that he surveyed can be ascribed77 to the Prophet himself (fa-laysa fima taqaddama hadith yasihh raf ‘uhu).78 Ibn Kathir begins a new section regarding hadiths that are not related directly to the question of the umm al-​walad but nevertheless could be used to argue for her freedom. He transmits, for example, a set of hadiths from al-​Bukhari and Muslim (d. 875) that the umm al-​walad should be freed because the Prophet manumitted all his slaves before his death. A hadith in al-​Bukhari states that the “the Messenger of God did not leave behind anything after his death, not a dinar or dirham, no male or female slave.”79 This argument holds that the Prophet’s umm al-​walad, Mariya, was not sold after the Prophet’s death, indicating she was understood to be free. Ibn Kathir further believed the Prophet explicitly granted the freedom of all his slaves, both male and female, before his passing.80 Ibn Kathir makes his argument by emphasizing the last part of the hadith: “The messenger of God did not leave behind upon his death a single dinar or dirham, a male or female slave.” The fact that the hadith refers to his death implies the Prophet decreed their freedom within his lifetime. Thus, if the Prophet had ordered the freedom of his male and female slaves then the umm al-​walad—​who takes priority in being free—​would also be included in this general proclamation. Ibn Kathir buttresses his claim that the Prophet decreed the freedom of Mariya during his lifetime because she was not sold after his death. If the Prophet had not decreed specifically that Mariya should be free, then she would have been sold during the reign of Abu Bakr, who allowed the sale of the umm al-​walad. However, others argued that Mariya held a unique relationship with the Prophet. Just as there are special rules regarding the Prophet’s wives, then there were special rules regarding Mariya.81 Her example of being freed after her master’s death was unique to her and should not be taken as a general rule for all umm al-​walads.



Remembering the Umm al-​Walad

j 309

The fourth group of hadiths contends that believers should follow the normative example of rightly guided caliphs and thus follow Umar’s example in freeing the umm al-​ walad. Hadiths relayed by Abu Dawud (d. 889), Ibn Maja, and al-​Tirmidhi (d. 892) have the Prophet declare, “You have my Sunna, and the Sunna of the rightly guided caliphs; hold on to it tightly.”82 Similarly, Ibn Maja and al-​Tirmidhi report that the Prophet stated, “Follow the two that will come after me, Abu Bakr and Umar.”83 Because the Prophet ordered his Companions to follow Umar, then they should obey his opinion regarding the umm al-​walad. Ibn Kathir, however, does not accept this reasoning. Umar’s practice should not be seen as identical to the Sunna of the Prophet. There are other traditions, which Ibn Kathir relates later, when the Prophet did not prohibit the sale of the umm al-​ walad. In addition, Abu Bakr and Ali allowed for the sale of umm al-​walad, challenging the idea that the rightly guided caliphs were in consensus on the issue. To prove there was no consensus among the caliphs, Ibn Kathir gives the example of how Ali changed his opinion near the end of his life. Ibn Kathir transmits a series of hadiths that indicate that Ali eventually rejected Umar’s opinion and contended the umm al-​ walad could be sold. Traditions narrated by al-​Shafi‘i and Ibn Abi Shayba (d. 849) have Ali explaining, “Umar consulted me in the sale of umm al-​walads so I saw—​him and me—​that she is free. Umar judged upon that in his lifetime and after him. When I was appointed I saw that she was a slave.”84 This tradition demonstrates that Ali changed his original opinion and eventually believed that the umm al-​walad could be sold. Yet, although Ali eventually allowed the sale of the umm al-​walad, some early Companions advised him not to break with his predecessors immediately. When the Companion Ubayda heard that Ali was going to change the ruling, he recommended to him, “Your opinion and the opinion of Umar in consensus is preferable to me than your individual opinion in this civil strife.”85 Ibn Kathir contends that Ubayda was not rejecting the opinion of Ali, but rather counseling him that in a time of conflict, he should not appear to be disagreeing with the popular Umar. In this tradition, Ali took Ubayda’s consul to heart and proclaimed he would continue to stay upon the opinions of Umar and Abu Bakr. Still, the story demonstrates that there was not complete consensus on the issue. According to these traditions, Ali chose to return from his opinion based on political reasons, not because he genuinely believed that she should not be sold. Ibn Kathir further adds that the Companion and anti-caliph Abd Allah ibn al-​Zubayr (d. 692) used to sell the ummahat al-​walad during his brief rule. Approximately 30 years after the death of Ali, Ibn al-​Zubayr was able to rule over a sizable area in the Arabian Peninsula where he allowed the sale of the umm al-​walad. If Ibn al-​Zubayr had permitted the umm al-​walad’s sale, Ibn Kathir asks, “how could there be consensus among the Companions or others [in this issue]?”86 Ibn Kathir then uses this discussion to discredit later claims that a binding consensus emerged. Ibn Kathir was suspicious of claims of “consensus” because they were typically used to stop discussion on a legal issue. He first cites the opinion of Abu al-​Hasan Ali ibn Khalaf ibn Battal (d. 448/​1057) that, “legal scholars throughout the realm are

310 i

Younus Y. Mirza

in agreement that it is not permissible to sell the umm al-​walad”87 and notes that very few disagree with this claim, such as the Zahiris. Ibn Kathir rejects the claim that only the likes of the Zahiris allow for the sale of the umm al-​walad because a group of the Companions believed it to be permissible. According to Ibn Kathir’s earlier analysis, Ali, Ibn al-​Zubayr, and other Companions allowed the umm al-​walad to be sold, discrediting any claims that a unified consensus ever took place. Ibn Kathir then cites a second group of scholars who believed a consensus did occur during the time of Umar, but then recognizes that disagreement arose afterward. Opposed to the first group, which believed that an authoritative consensus emerged, this second group maintained it was better to follow the early consensus rather than the later difference of opinion. Al-​Bayhaqi, for instance, contends that it is possible that Umar and others agreed that the Prophet originally ordered the freedom of the umm al-​walad on the death of their masters. For al-​Bayhaqi, following this early consensus was more preferable than taking sides in the disagreement that arose afterward. Ibn Kathir explains that those who followed al-​Bayhaqi’s argumentation claimed that a silent consensus formed (or an unstated agreement emerged) that the umm al-​walad should not be sold. Umar’s ruling on the umm al-​walad gradually became the norm because there was no explicit opposition to it. For this to have occurred, Ibn Kathir contends that three conditions must have been met: (1) the opinion that the umm al-​walad could not be sold became prevalent, (2) that no one opposed Umar on the issue, and (3) the ruling eventually became authoritative. As for the first opinion, there are many reports that indicate that Umar’s opinion became prevalent, many of which Ibn Kathir has already discussed in the treatise. As for the second issue, Ibn Kathir explains there are many reasons why somebody would not oppose Umar, including the latter’s intimidating manner. Umar was known to issue controversial legal opinions, but few opposed him because of his domineering personality.88 Ibn Kathir did believe that Umar performed his own ijtihad and performed a righteous act in regard to the umm al-​walad, but did not deem that his era ended without any disagreement on the issue, especially from Ali and Ibn al-​Zubayr.89 With regard to whether the opinion became authoritative, Ibn Kathir highlights that Ali eventually differed with Umar, even addressing his followers, saying he and Umar originally believed the umm al-​ walad should not be sold, but he had changed his mind after his appointment as caliph. If Umar’s opinion had become authoritative, then Ali would not have disagreed with him; Ali was known to be one of the rightly guided Imams and caliphs. Ali’s later action also disproves the contention, put forward by al-​Bayhaqi, that there was a clear indication from the Prophet that the umm al-​walad should not be sold. If the order had come directly from the Prophet, then Ali would not have disagreed with Umar. Ali rather saw the decree as an opinion of Umar and one that could be overturned. Ibn Kathir then moves to argue that his own Shafi‘i legal school did not have consensus on the umm al-​walad, despite later claims. Al-​Shafi‘i himself had four opinions on the issue: (1) the umm al-​walad should be freed after her master’s death; (2) she could be



Remembering the Umm al-​Walad

j 311

sold unconditionally; (3) she could be sold during the life of her master, but then only freed after his death; and (4) suspension of judgment on the issue.90 Al-​Shafi‘i thus held a range of issues throughout his lifetime—​from the belief that the umm al-​walad should eventually be freed to that it was not possible to rule decisively on the issue. Ibn Kathir goes on to cite the opinion of the great Shafi‘i legal authority Abu Hamid al-​Ghazali (d. 1111) to strengthen his argument that a Shafi‘i consensus did not emerge by his lifetime. Al-​Ghazali records that the scholars in his time believed the umm al-​walad was free upon her master’s death and it was not permissible to sell her during his lifetime. However, al-​ Ghazali notes that al-​Shafi‘i originally held she could be sold unconditionally and this was the opinion of Ali.91 Ibn Kathir uses al-​Ghazali’s statement here to argue there was no early consensus, as some Shafi‘i scholars claimed. Ibn Kathir concludes the section with two more indirect arguments for the freedom of the umm al-​walad. The first contends there was a consensus that the umm al-​walad’s child was free, so, given that the child was “part” of her, she should also be freed. Supporters of this argument cite the hadith from Ibn Hanbal that a man freed a part of his slave, an act to which the Prophet responded that the slave was free in his entirety. The umm al-​walad should therefore follow her child in freedom. Similarly, a tradition has Umar hear of a master selling his slave woman after she had a stillborn child with him. Umar approaches the man and whips him then exclaims, “After your flesh mixed with her flesh and your blood with her blood, you sell her?”92 Umar condemns the man because the women had become “part” of him and thus should now be free. Ibn Kathir, however, rejects the reasoning that a child is part of his mother as one of her limbs. Just because her child is free does not necessarily follow that she is free as well. Second, Umar may have used the man as an example to prohibit the sale of umm al-​walad, but there are other instances when he did not tolerate the practice. The last argument relied on the prophetic model of good character. It was shameful and against good morals to sell the umm al-​walad. As the Prophet Muhammad famously stated, “Verily God has sent me to perfect character.”93 Separating a mother from her children contradicted true character and relates to the argument that selling the umm al-​ walad was the breaking of kin condemned in verse 47:22: “If you turn away now could it be that you will go on to spread corruption all over the land and break your ties of kinship?” In summary, the various reasons that Ibn Kathir presents for the freedom of the umm al-​walad relate to Umar’s authority and the umm al-​walad’s humanity. Umar’s supporters believed his decrees on the umm al-​walad should be followed because his rulings were akin to the Prophet’s. The Prophet had ordered the Companions to follow Umar and they thus should follow his ruling on the umm al-​walad. Some traditions even suggest, such as Said ibn Musayyab’s, that Umar was acting out the Prophet’s order. The umm al-​walad, furthermore, was a distinct marker of Umar’s rule. Various traditions have Ali both confident and hesitant to change the ruling of the umm al-​walad because it was identified with his predecessors. To argue that the umm al-​walad could be sold

312 i

Younus Y. Mirza

was in effect questioning the legitimacy of Umar’s caliphate. Second, the proponents of the freedom of the umm al-​walad all emphasize her humanity. Traditions imply that selling the umm al-​walad was breaking of kinship ties and that selling her was dishonorable and against true character. As one Umar tradition asks, how could one sell their female slave when one’s flesh mixed with hers? In contrast, the argument for the sale of the umm al-​walad will view her as property and emphasize the authority of the Prophet and Ali. The Unconditional Sale of the Umm al-​Walad Ibn Kathir turns next to the argument that the umm al-​walad can be sold unconditionally and is thus like any other slave. Building on the previous section, Ibn Kathir contends there was no binding consensus that the umm al-​walad should not be sold and there are hadiths that infer that the Prophet endorsed the practice. Ibn Kathir begins by explaining that many of the Companions used to approve of the sale of the umm al-​walad, such as Abu Bakr, Umar (in the beginning of his reign), Ali (at the end of his), and other Companions such as Ibn al-​Zubayr. Ibn Kathir then lists a host of traditions that support his claim. For instance, he cites the tradition from the Musannaf of Abd al-​Razzaq (d. 826)94 that, “Abu Bakr al-​Siddiq used to sell the umm al-​walads in his reign and Umar in half of his reign.”95 Another tradition supports the idea that Umar initially allowed the sale of the umm al-​walad. A man came to Umar and asked, “What do you think about the umm al-​walad?” Umar responds, “[she] is your wealth. If you desire, you can sell; if you want, you can give her as a gift.”96 Another tradition from Ibn al-​Zubayr equates the umm al-​walad to any other form of property: “She has the status of your sheep or camel.” The umm al-​walad was like livestock and did not deserve any special status. Ibn al-​Zubayr is further reported to have said that “we did not free [the umm al-​walad] until there was discussion regarding freeing her,” or during the rule of Umar.97 Umar’s opinion was an innovation and she was previously sold without restrictions. In addition, the umm al-​walad was sold after Umar’s death. Ibn al-​Zubayr used to allow the sale of the umm al-​walad during his reign. Abu Shayba records in his Musannaf that two men narrated to Ibn Umar that they had seen Ibn al-​Zubayr allow for the sale of the umm al-​walad in Mecca. Ibn Umar responds that his father, Abu Hafs, used to free them. The tradition demonstrates that, more than 40 years after Umar’s rule, there still was no consensus on the issue. According to Ibn Kathir, all these traditions establish that some early Companions continued to believe the umm al-​walad could be sold even after Umar’s decree. The fact there was no absolute consensus is demonstrated in that the opinion that the umm al-​ walad could be sold survived marginally within Sunni legal schools. Al-​Shafi‘i subscribed to it initially, it was one of the opinions transmitted from Ibn Hanbal, and it became the majority opinion of the Zahiri school.98



Remembering the Umm al-​Walad

j 313

Even some postformative jurists, such as Ibn Sayyid al-​Nas, believed that the umm al-​ walad could be sold. Ibn Kathir summarizes Ibn Sayyid’s argument, which is derived from the Quran, Sunna, consensus, and analogy (qiyas).99 He cites, for example, the Quranic verse (2:275): “That God has made trading permissible.” This verse is not explicit evidence for the sale of the umm al-​walad, but rather is a general proclamation on the permissibility of commerce. As for 22:5–​6, it states: “Those (believers) who are chaste, except with their spouses or their slaves—​with these they are not to blame.” There is consensus that the master of the umm al-​walad can have intercourse with her and that God describes her as a possession. The verse does not provide, however, any direct guidance on whether the umm al-​walad can be sold. With regard to the Sunna, al-​Shafi‘i cites a hadith in which the Companion Jabir ibn Abd Allah states, “That we used to sell our slaves (the umm al-​walads) and the Prophet was among us, and he did not see it as a problem.”100 The hadith suggests the Prophet implicitly endorsed the sale of the umm al-​walad. Abu Dawud and Muslim also narrate a similar tradition in which Jabir states, “We used to sell the umm al-​walad during the time of the Prophet and Abu Bakr. When Umar came he prevented us [from doing so] so we stopped.”101 Similar to the previous tradition, Jabir contends that the prohibition of sale of the umm al-​walad was an innovation of Umar and it only became law after his order. Other traditions support Jabir’s contention, such as one quoted by al-​Nasa’i (d. 915) that the Companion Abu Said al-​Khudri states that the umm al-​walad was sold during the lifetime of the Prophet.102 Ibn Kathir contends there are no authentic hadiths that abrogate the hadiths of Jabir and Abu Said al-​Khudri in which the umm al-​walad was sold during the lifetime of the Prophet. Ibn Kathir is not convinced of the argument that the Prophet initially allowed the sale of the umm al-​walad and then prohibited it later, at the end of his prophetic career. Rather, the rule of prohibiting sale of the umm al-​walad was a position of the Companions that, by consensus, does not abrogate authentic hadiths. As a staunch hadith scholar, Ibn Kathir did not believe the words and actions of the Companions had the authority to override that of the Prophet.103 To cement his argument further that the Prophet did rule on the umm al-​walad, Ibn Kathir considers critiques of al-​Bayhaqi’s analysis of Jabir’s hadith in which the Prophet himself prohibited the sale of the umm al-​walad. Al-​Bayhaqi claimed that it was possible the Prophet did not realize that the umm al-​walad was being sold so he did not condemn it. In addition, hadiths similar to Jabir’s could originate from a time before the Prophet officially prohibited the umm al-​walad’s sale. The Prophet could have then later forbade her sale, but some Companions did not come to hear about it and continued to sell her. Other scholars, however, refute al-​Bayhaqi’s line of thinking, noting that Islamic law incorporates similar hadiths to that of Jabir. For instance, regarding the alms given in the completion of Ramadan, a hadith states, “We used to give, at the time of the Prophet Muhammad, two handfuls of dates, two handfuls of barley.”104 The hadith begins with the same formula as the umm al-​walad hadith, “We used to,” but nobody rejects its authority

314 i

Younus Y. Mirza

and it is used as evidence in how much alms to give. Similarly, Ibn Umar is reported in al-​ Bukhari to have said, “We used to rank [the Companions] at the time of the Prophet and we said: Abu Bakr, then Umar, then  . . ..”105 This hadith, as well as others, all begin with the phrase, “We used to at the time of the Prophet,” and they are authoritative. Ibn Kathir takes issue with al-​Bayhaqi’s contention that Companions may have sold the umm al-​walad before the Prophet’s prohibition. He asks pointedly, “But where is the prohibition?”106 Ibn Kathir holds that it, the prohibition, goes back to Umar, who believed that forbidding the sale of the umm al-​walad was for the benefit of the community.107 During his reign, the community had greater wealth and could absorb the financial loss of freeing the women.108 Umar reached praiseworthy juridical decisions and correct opinions, but he did not have a definite text from the Prophet prohibiting the practice. The fact that there was no textual support for Umar’s ruling is evident in Ali’s statement: “Umar consulted me on the umm al-​walad and he and I saw that they should not be sold,” but then Ali changed his opinion. Ali felt comfortable disagreeing with Umar, but would not contradict a clear order of the Prophet. The last hadith, in favor of the umm al-​walad’s sale, relates to eschatology. In the famous hadith of Gabriel, in which the Archangel visits the Prophet Muhammad, Gabriel asks the Prophet about the signs of the Day of Judgment. The Prophet responds, in part, that, “The slave woman gives birth to her master.”109 Some scholars note that the slave woman in this hadith is an umm al-​walad because she gives birth to a child. The slave woman, however, is not free but rather under the possession of her child. Because she is possessed by her child, she is also property and could thus be sold.110 In regard to an authoritative consensus on the sale of the umm al-​walad, Ibn Kathir rejects that there ever was agreement that she could not be sold. As he says repeatedly, there was a nominal consensus during the time of Umar against the sale of the umm al-​walad. Others also claim that during the reign of Abu Bakr, there was consensus of selling the umm al-​walad, but Ibn Kathir did not believe that Abu Bakr—​during his brief reign—​made an explicit statement allowing for her sale. If Abu Bakr had ruled on the issue, Umar would not have come afterward and reversed his decision. Rather, it is possible there was a silent consensus on her sale before the time of Umar and that people continued to sell the umm al-​walad until Umar’s proclamation. Ibn Kathir concludes the section explaining that some have argued that the umm al-​ walad should be sold through the Islamic juridical principle of al-​istishab—​that is, the “presumption of continuity.”111 Using this principle, it can be argued that the umm al-​ walad was initially a slave woman who could be sold by consensus. Thus, even if she delivers a child, then she should be governed by the original principle that it is permissible to sell her. Having a child does not change the nature of her being a slave.112 To summarize, this section demonstrates that those who support the sale of the umm al-​walad emphasize the authority of the Prophet and Ali, and believe she was like any other slave. Traditions from the Companions contend the Prophet affirmed their practice of selling the umm al-​walad. Ali would eventually disagree with Umar, giving further



Remembering the Umm al-​Walad

j 315

credence to the argument that the prohibition of her sale was not prophetic; Umar’s ruling was an innovation and not binding. Also, the advocates of the umm al-​walad’s sale believed she was like any other piece of property. This is seen most starkly in the tradition ascribed to Ibn al-​Zubayr: “She has the status of your sheep or camel.” Having her master’s child did not change her status of being a possession, which, in turn, allowed her to be sold. Conclusion Ibn Kathir’s treatise on the umm al-​walad allows for a much better grasp of Islamic conceptions of sexuality, religious authority, and human identity. The traditions discussed here demonstrate that a decisive debate took place over the definitions of marriage and proper sexual relations. The same figures that supported temporary marriage, such as Ali, Jabir ibn Abd Allah, and Ibn al-​Zubayr, also believed the umm al-​walad could be sold. Ali’s opinion would eventually be incorporated within Shi‘ism along with the validity of temporary marriages. In contrast, all major Sunni legal schools of jurisprudence incorporate Umar’s opinion that the umm al-​walad should not be sold and should be free after her master’s death. Early legal manuals are explicit that this ruling on the umm al-​walad was Umar’s, and they show no hesitation in incorporating it within their authoritative opinions. Although a more comprehensive examination of the various traditions needs to be undertaken, Ibn Kathir’s collection does signal a difference of opinion and the potential beginnings of different religious communities. The debate over the umm al-​walad did not end with Umar, but continued throughout the course of Islamic history. Jurists were confronted with outlining the rules and regulations regarding how the umm al-​walad should be treated. During the Mamluk period, leading jurists began to write independent treatises increasingly on the subject. Some argued that the umm al-​walad could be sold based on the fact that the Prophet did not forbid her sale. Others maintained there was a Sunni consensus that prohibited her sale and that Umar was acting out a prophetic injunction. Central to this argument was the question of humanity. Those who saw the umm al-​walad as property advocated for her sale whereas those who viewed her as a human being and mother maintained she should be free.113 Specifically, Ibn Kathir’s treatise demonstrates a move toward prophetic authority. Throughout the treatise, Ibn Kathir seeks to separate clearly between the religious and legal authority of Umar and the Prophet. Although he respected Umar’s opinion and saw it as a valid juridical decision, Ibn Kathir stresses that the sayings and actions of the Prophet are the most authoritative. The debates surrounding the umm al-​walad continue to be relevant today. Contemporary Salafi movements echo Ibn Kathir’s belief that ultimate religious authority resides with the Prophet and that the later intellectual tradition should be read with suspicion. Similarly, claims by Muslim jurists that a man has the indisputable divine right to engage in polygamy or temporary marriages resemble earlier arguments that a Muslim

316 i

Younus Y. Mirza

man has the right to take and exchange concubines. Muslims continue to debate the appropriate number of sexual partners a person may have, with some stressing short-​term sexual fulfillment and others advocating for a longer, more stable relationship. Thus, the historic debates surrounding the umm al-​walad continue to influence modern Muslim discussions of marriage and sexuality. Notes 1. Ibn Kathir, Juz, 49–​50. 2.  Schacht, “Umm al-​Walad.” For a discussion of the legacy of Schacht see Powers, “Wael B. Hallaq on the Origins of Islamic Law.” 3. Schacht, “Umm al-​Walad.” 4. al-​Sanʻani, Subul al-​salam, 3:20. 5. Schacht, Origins, 140. 6. Schacht, Origins, 140. 7. Schacht, Origins, 155. 8. Schacht, Origins, 149. 9.  For a critique of Schacht’s theories see al-​Azami, On Schacht’s Origins; Brown, “Critical Rigor vs. Juridical Pragmatism,” 7–​8. 10. For the most recent work on hadith criticism, see the work of Harald Motzki and his emerging school:  Boekhoff, Verseegh, and Wagemakers, Transmission and Dynamics of the Textual Sources of Islam; Motzki, Boekhoff-​van der Voort, and Anthony, Analysing Muslim Traditions. 11. For a helpful summary of the Western debates over hadith authenticity, see Brown, Hadith, 197–​235. 12. Brown, “Critical Rigor,” 10. 13. Brown, “Critical Rigor,” 7. 14. Jabir ibn Abd Allah was known to be a Shi‘i supporter and appears in many of the chains of transmissions permitting the sale of the umm al-​walad and the practice of temporary marriages (mut‘a); Gribetz, Strange Bedfellows, 53. 15. Haider, “Prayer, Mosque, and Pilgrimage,” 154. 16. Schacht, Origins, 265. 17. Hakim, “Conflicting Images of Lawgivers.” 18. Haeri, Law of Desire. 19. For more on the various rules for mut‘a, see Gribetz, Strange Bedfellows. 20. See the chapter on the umm al-​walad: Mattson, “A Believing Slave,” 126–​182. 21. Gribetz, 92. 22. Gribetz, 102. 23. More work needs to be done on the sexual rights that jurists gave to female slaves. 24. Since mut‘a marriages fell out of favor in Sunnism, the similarities between the mut‘a wife and female slaves have largely been forgotten in Sunni law but nevertheless remain apparent in Shi‘i jurisprudence; Gribetz, 143. 25. Mattson, 133. 26. Schacht, “Umm al-​Walad.” 27. Brockopp, Early Maliki Law, 196.



Remembering the Umm al-​Walad

j 317

28. Ali, Marriage and Slavery, 168. 29. Rapaport, “Women and Gender in Mamluk Society,” 9. 30. Marmon, “Domestic Slavery,” 25. 31. Marmon, “Domestic Slavery,” 6. 32. For a discussion of the umm al-​walad in Ottoman times, see Zilfi, Women and Slavery, 109–​115; Toledano, Slavery and Abolition, 54–​80. For more on the umm al-​walad in nineteenth-​ century Egypt, see Tucker, Women in Nineteenth-​Century Egypt, 164–​193. 33. Mirza, “Ibn Kathir.” 34.  For more on the continuous practice of ijtihad throughout Islamic history, see Hallaq, “Was the Gate of Ijtihad Closed?” For more on the debate between ijtihad and taqlid during Ibn Kathir’s time, see Mustafa, On Taqlid. 35. For more on early traditionist jurisprudence, see Melchert, “Traditionist-​Jurisprudence.” 36. Brown, “Critical Rigor,” 27. 37. al-​Bayhaqi, Maʻrifat al-​sunan wa’l-​athar. For more on al-​Bayhaqi’s methodology, see Brown, “Critical Rigor,” 24. 38. Brown, “Critical Rigor,” 36. Brown is quoting Ibn Taymiyya (d. 728/​1328) here. 39. al-​Nawawi, al-​Majmuʻ, 9:264. 40. al-​Nawawi, al-​Majmuʻ, 15:195. According to Schacht , this tradition would be considered a “countertradition” because it has one of the opponents of the doctrine, Ibn Abbas, as its narrators. 41.  al-​Nawawi discusses only briefly the various umm al-​walad traditions in his chapter on what is not permissible to sell (Al-​Nawawi, al-​Majmuʻ, 9:264). 42. There has been important work on slaves that engages the primary source manuscript and document tradition. See Little, “Purchase Deeds,” and “Court Records; and Brockopp, Early Maliki Law. 43. Ibn Kathir, Juz’ fi bay’. 44. I am currently writing on the modern publications of Ibn Kathir’s Tafsīr. 45. Hafyan argues successfully that the treatise is authentic based on the transmission of the manuscript and its contents. He recollects that he found the treatise at the end of one of Ibn Kathir’s student’s works, Ibn Urwa’s al-​Kawakib al-​darari; Ibn Kathir, Juz’ fi bay’, 11. Hafyan explains that the manuscript he obtained of al-​Kawakib al-​darari was of Ibn Urwa’s himself and that Ibn Urwa subsequently endowed the copy in a library in Damascus, the same city where he studied with Ibn Kathir. Second, Ibn Kathir alludes to the treatise in his own works, such as in al-​Bidaya wa’l-​nihaya and Musnad al-​Faruq. Although Hafyan does not say so, Ibn Kathir also describes the treatise in his Sharh al-​Tanbih; Ibn Kathir, Irshad al-​faqih, 2:120. Third, Ibn Kathir references his primary teacher al-​Mizzi in the treatise and calls him our teacher (shaykhuna); Ibn Kathir, Juz’ fi bay’, 54. In my analysis of Ibn Kathir’s major works, I found that Ibn Kathir constantly cites his primary teacher al-​Mizzi, especially in relation to hadith; Mirza, “Ibn Kathir,” 94. Last, Hafyan notes that later authors, such as al-​San‘ani, reference Ibn Kathir’s treatise and record that he transmitted four opinions from al-​Shafiʻi and eight opinions in total. This description describes the contents of the manuscript that Hafyan obtained. I have found that the jurist al-​Shawkani also cites Ibn Kathir’s treatise and makes a similar description; al-​Shawkani, Nayl al-​awtar min ahadith sayyid al-​akhyar, 5–​6:507. I would further add that the treatise’s legal methodology is consistent with Ibn Kathir’s legal works. Ibn Kathir presents legal issues first by listing the relevant hadiths on the topic. He then ties the hadiths into a larger discussion of the various legal positions, especially that of his own legal school Shafi‘ism; Mirza, “Ibn Kathir,” 109.

318 i

Younus Y. Mirza

46. For more on modern Muslim views of slavery, see Clarence-​Smith, Islam and the Abolition of Slavery. 47. Shakir, ʻUmdat al-​tafsir, 1:5. 48. Shakir, ʻUmdat al-​tafsir, 1:7. 49. Similar to his slave treatise, Ibn Kathir contends in this chapter that the source of the prohibition originated with the caliph Umar. 50. Ibn Kathir, Irshad, 2:120. 51. These unnamed scholars most likely contended there was no prophetic traditions that prohibit the sale of the umm al-​walad. 52. Ibn Kathir, Irshad, 2:120. 53. Ibn Qadi al-​Shuhba, Tabaqat al-​fuqahaʼ al-​shafiʻiyya, 2:80. There is disagreement within the biographical dictionaries over whether the treatise was titled “The Prohibition (mani‘) of Selling the umm al-​walad” or “The Permissibility (jawaz) of Selling the umm al-​walad”; Ibn Kathir, Juz’ fi bay’, 14. Preliminary evidence suggests the treatise was on the permissibility of selling the umm al-​walad because Ibn Kathir cites Ibn Sayyid’s opinion as being in favor of her sale (Ibn Kathir, Juz’ fi bay’, 95). 54. On Ibn Sayyid al-​Nas, see Rosenthal, “Ibn Sayyid al-​Nās.” 55. The eight topics are that (1) she is free on her master’s death; (2) she may be sold without restriction; (3) the master may sell her at any time during his lifetime, but when he dies she becomes free; (4) she may be sold to pay a debt; (5) she may be sold, but if her child is alive at the death of his father (and her master), she is manumitted as part of the child’s inheritance; (6) she can only be sold on the condition she is set free; (7) if she is righteous then it is not permissible to sell her, but if she becomes immoral or an unbeliever then it becomes permissible; and (8) suspension of judgment on the issue. 56. Hafyan explains that the manuscript contains two different versions of the treatise. In his view, the first version is the final version because it is more complete, longer, and contains citations the second does not. This final version, in addition, has Ibn ‘Urwa say that it was related directly to him (anba’ani) by Ibn Kathir himself. Hafyan notes further that Ibn ‘Urwa was born sometime before 760/​1359, and Ibn Kathir passed away in 774/​1373. This would mean the transmission of the work took place toward the end of Ibn Kathir’s life and represents the final version. I agree with Hafyan’s analysis because the descriptions that al-​San‘ani and al-​Shawkani give of Ibn Kathir’s treatise match this final edition (Ibn Kathir, Juz’ fi bay’, 23). In this chapter, I rely on the final version. 57. My methodology of summarizing the treatise is to quote the relevant Quranic verses and hadiths while paraphrasing Ibn Kathir’s argument. I only state the chains of transmission (isnads) if I feel it is pertinent to the argument. Hafyan provides the exact references to the various traditions in his edition of Ibn Kathir’s Juz’ fi bay’. 58. Ibn Kathir, however, notes there is a consensus that she is not unconditionally free after giving birth to her master’s child and he is still alive. The discussion over her freedom is only in regard to when her master dies. 59. Ibn Kathir discusses these scholars’ positions later in the treatise. 60.  Ibn Kathir, Juz’ fi bay’, 50. Ibn Kathir notes the story is buttressed by other chains of transmission. 61. The breaking of slave families was an issue that plagued jurists throughout Islamic history. In his Musannaf, Ibn Abi Shayba devotes a chapter to the “Separation of a [Slave] Child and his



Remembering the Umm al-​Walad

j 319

Parents.” Ibn Abi Shayba lists 18 traditions in this chapter, with the first being the Prophet seeing a slave woman crying. He asks his Companions what was bothering her. The Companion explains that her child was sold. The Prophet responds “return [the boy] or buy him [and return him]”; Ibn Abi Shaybah, al-​Musannaf, 7:567. In his legal work Sharh al-​Tanbih, Ibn Kathir cites two hadiths condemning the separation of a slave mother and her child. The first states “whoever separates between a slave woman and her child then God will separate him from his loved ones on the Day of Judgment” (Ibn Kathir, Irshad, 2:14). Ibn Kathir notes that the hadith is narrated by the hadith scholars of Ibn Hanbal and al-​Tirmidhi. 62.  The hadith scholar al-​Daraqutni (d. 385/​995) also narrates this hadith with a different chain of transmission. 63. Ibn Kathir cites different authorities that all cite the problematic nature of Hasan b. ‘Abd Allah (Ibn Kathir, Juz’ fi bay’, 52–​53). 64. For more on Mariya, see Hidayatullah, “Mariyya the Copt,” 221–​243. Hidayatullah discusses whether Mariya was a wife or umm al-​walad of the Prophet. 65. Ibn Kathir then proceeds into a discussion of whether this particular hadith has a complete chain of transmission back to the Prophet (marfu‘) or has a missing link between the Prophet and a successor (mursal). 66. Ibn Kathir, Juz’ fi bay’, 57–​58. 67. For more on al-​Daraqutni, see Brown, “Criticism of the Proto-​Hadith Canon,” 1–​37. 68. Ibn Kathir, Juz’ fi bay’, 58. 69. Ibn Kathir, Juz’ fi bay’, 60. 70. Ibn Kathir also notes here that Sa‘id b. Musayyab believed the Prophet stated that umm al-​ walad should not be included as part of one’s inheritance or used to pay off a debt. 71. Saleh, “Ibn Taymiyya and the Rise of Radical Hermeneutics,” 138. 72. Ibn Kathir explains here that the issue of mursal hadiths has caused great disagreement among hadith scholars, with competing sides claiming there was a consensus regarding their validity or inauthenticity. Ibn Kathir disagrees that mursal hadiths cannot be used in legal arguments because al-​Shafi‘i, the eponym of Ibn Kathir’s school of law, used mursal traditions from the most famous and well-​respected Successors (kibar). Ibn Kathir also rejects the opposite claim that all mursal traditions should be accepted because this was the practice of many formative jurists. Ibn Kathir does say in his later work Ikhtisar ‘ulum al-​hadith that al-​Shafi‘i accepted the mursal traditions of Said ibn Musayyab. Ibn Kathir, al-​Ba‘ith al-​hathith, 1:156. 73. Ibn Kathir, Juz’ fi bay’, 62. 74. Ibn Kathir, Juz’ fi bay’, 65. 75. Ibn Kathir, Juz’ fi bay’, 65. 76. As Jonathan Brown explains, critical hadith scholars believed that one “must choose the correct version of the report by examining the totality of its existing narrations rather than relying on a fixed rule of acceptance. If the majority of the reliable narrations of the hadith were from a Companion, a Prophetic version was of dubious provenance.” Brown, “Critical Rigor,” 30. 77. Ibn Kathir, Juz’ fi bay’, 66. For more on the hadith category of marfu‘, see Brown, “Critical Rigor,” 8. 78. Ibn Kathir does list a fifth hadith in which the Prophet was asked regarding the umm al-​walad and he stated, “Her master can take pleasure from her during his lifetime but when he dies she is free.” Ibn Kathir also finds this hadith inauthentic, citing the authority Ibn Abi Hatim al-​Razi (d.938).

320 i

Younus Y. Mirza

79. Ibn Kathir, Juz’ fi bay’, 69. 80. Ibn Kathir, Juz’ fi bay’, 71. 81. Ibn Kathir, Juz’ fi bay’, 72. 82. Ibn Kathir, Juz’ fi bay’, 72. 83. Ibn Kathir, Juz’ fi bay’, 73. 84. Ibn Kathir, Juz’ fi bay’, 76. 85. Ibn Kathir, Juz’ fi bay’, 77. 86. Ibn Kathir, Juz’ fi bay’, 78. 87. Ibn Kathir, Juz’ fi bay’, 78. 88. Hakim, “Conflicting Images of Lawgivers,” 163. For instance, the practice of mut’a of hajj or combining the short pilgrimage (‘umra) with the full one (hajj) was not performed during his reign because of his prohibition. However, the mut‘a of hajj is justified in both the Quran and Sunna. 89. Ibn Kathir, Juz’ fi bay’, 80. Ibn Kathir then proceeds into a complex discussion of a third category of ijma‘ sakuti. For the purposes of this chapter, I skip this discussion on the methodology of jurisprudence and focus on the legal discussion over the umm al-​walad. 90. Ibn Kathir, Juz’ fi bay’, 84. 91.  Ibn Kathir then takes the opportunity to question the entire legal category of consensus (ijma‘). He quotes an opinion ascribed to al-​Shafi‘i that simply because there is no known disagreement, it does not mean there is consensus. Similarly, al-​Shafi‘i is quoted as saying that no one had claimed consensus (except for the ritual obligations proscribed to all Muslims) among the Companions and Successors, and nobody from the generation after them, and so forth. Ibn Kathir cites a similar statement, Juz’ fi bay’, 84, from Ibn Hanbal, where he declares “whoever claims consensus is a liar. Perhaps people differ and the (jurist) may not know.” It was more preferable, in Ibn Hanbal’s view, to say “we do not know people differing in this issue” or that “a difference of opinion has not reached me.” Ibn Kathir states these are the opinions of two of the greatest legal and hadith scholars, and they reject claims of consensus even if they have not heard an opinion contrary to it. Ibn Kathir’s legal circle was skeptical of claims of consensus because it was a way to silence one’s opponent and affirm the legal tradition rather than engage in ijtihad. As Hallaq (History of Islamic Legal Theories, 79) states Some Shafi‘ite jurists maintained that if all mujtahids actively participate in forming a consensus then such a consensus is deemed binding before the death of these mujtahids. However, if only some of them openly voice their agreement on an issue, while the rest of them express no opinion, then consensus cannot be considered to be binding until such time as the entire generation of these mujtahids becomes extinct. Obviously, the rationale behind this position is that those who did not express an opinion may do so at a later time in their life, and their opinion may be at variance with that on which the others have agreed. In the case of the umm al-​walad, the fact that Ali changes his mind at the end of his life in favor of her sale means the original consensus is no longer valid. For more on the Sunni consensus, see Hallaq, “Authoritativeness of Sunni Consensus.” 92. Ibn Kathir, Juz’ fi bay’, 86. 93. Ibn Kathir, Juz’ fi bay’, 91. 94. For more on Abd al-​Razzaq’s (d. 211/​826) Musannaf, see Motzki, “Musannaf of ʿAbd al-​ Razzaq al-​Sanʿani.” 95. Ibn Kathir, Juz’ fi bay’, 94.



Remembering the Umm al-​Walad

j 321

96. Ibn Kathir, Juz’ fi bay’, 94. 97. Ibn Kathir, Juz’ fi bay’, 94. 98. Ibn Kathir, Juz’ fi bay’, 95. 99. Ibn Kathir, Juz’ fi bay’, 96. 100. Ibn Kathir, Juz’ fi bay’, 97. Ibn Kathir notes that other hadith scholars such as al-​Nasa’i, Ibn Maja and al-​Bazzar narrate similar traditions. 101. Ibn Kathir, Juz’ fi bay’, 98. 102. Ibn Kathir, Juz’ fi bay’, 99. Ibn Kathir adds that, although some criticize this hadith, it is authentic on the methodology of the hadith scholar Muslim and it is corroborated by other traditions. 103.  It is interesting to note that Umar’s name is not mentioned explicitly in the traditions that state the umm al-​walad was sold during the time of the Prophet. A possible reason for this is that “when these traditions were circulated, Umar’s image as an ideal leader was so deeply rooted that any objection to his authority would have to be raised very carefully” (Hakim, “Conflicting Images of Lawgivers,” 168). Hakim, here, is speaking regarding the early mut‘a traditions, but I feel his observation is applicable to the umm al-​walad traditions as well. 104. Ibn Kathir, Juz’ fi bay’, 101. 105. Ibn Kathir, Juz’ fi bay’, 101. 106. Ibn Kathir, Juz’ fi bay’, 101. 107.  For more on maslaha in Islamic law, see Opwis, Maslaha and the Purpose of the Law. Regarding Umar’s attempts toward maslaha, Opwis states, “The decision of the second caliph Umar (r. 13–​23/​634–​44) to keep the lands of southern Iraq (al-​sawad) under state control, instead of dividing it among conquering tribes, was driven, according to Abu Yusuf (d. 182/​798), by considering the good (khayr) and general benefit (‘umum al-​nafi‘) of the Islamic community” (Opwis, Maslaha and the Purpose of the Law, 9). 108. Ibn Kathir adds here that Umar had an intimidating personality and when he made similar decrees, people simply listened to his command. 109. Ibn Kathir, Juz’ fi bay’, 102. 110. Ibn Rajab al-​Hanbali (d. 795/​1392), however, argues that the hadith of Jabril supports the freedom of the umm al-​walad. The child is the master (rabb) of his mother in that he is the one who frees her. This is similar to the hadith in which the Prophet says regarding Mariya, “her child freed her.” The child is the one who eventually frees his mother. Ibn Rajab, Jamiʻ al-​ʻulum wa al-​hikam, 114. 111. Hallaq, History of Islamic Legal Theories, 113. 112. Ibn Kathir, Juz’ fi bay’, 103. 113. For more on the discussion of “women as property,” see Tucker, Women in Nineteenth-​ Century Egypt.

Bibliography Primary Sources al-​Bayhaqi. Marifat al-​sunan wa’l-​athar. 7 vols. Edited by Sayyid Kasrawi Hasan. Beirut: Dar al-​ Kutub al-​Ilmiyya, 1991. Ibn Abi Shaybah. Al-​Musannaf. 10 vols. Edited by Abu Muhammad Usama b.  Ibrahim b. Muhammad. Cairo: al-​Faruq al-​Haditha lil-​Tibaʻa wa’l-​Nashr, 2008.

322 i

Younus Y. Mirza

Ibn Kathir. Al-​Ba‘ith al-​hathith: Sharh ikhtisar ulum al-​hadith. 2 vols. Edited by Ali Hasan Ali Abd al-​Hamid. Riyadh: Maktabat al-​Ma‘arif li’l-​Nashr wa’l-​Tawzi, 1996. —​—​—​. Irshad al-​faqih ila ma‘rifat adillat al-​Tanbih. 2 vols. Edited by Bahjat Yusuf Hamad Abu al-​Tayyib. Beirut: Muassasat al-​Risala, 1996. —​—​—​. Juz’ fi bay` ummahat al-​awlad. Edited by Umar b. Sulayman Hafyan. Beirut: Muassasat al-​Risala, 2006. Ibn Qadi al-​Shuhba. Tabaqat al-​fuqaha al-​shafiʻiyya. 2 vols. Edited by Ali Muhammad Umar, Cairo: Maktabat al-​Thaqafa al-​Diniyya, 1998. al-​Nawawi. Al-​Majmu, Sharh al-​Muhadhdhab. 16 vols. Edited by Zakariya Ali Yusuf. Cairo: Matba’at al-​Asima, 1969. al-San`ani, Muhammad b. Ismail. Subul al-​salam: sharh Bulugh al-​maram min jamʻ adillat al-​ ahkam lil-​Hafiz Ibn Hajar al-​Asqalani. 4 vols. Edited by Muhammad al-​Dali Balta. Beirut:  al-​Maktaba al-​Asriyya, 2005. al-​Shakir, Ahmad Muhammad. Umdat al-​tafsir an al-​hafiz Ibn Kathir. 5 vols. Egypt:  Dar al-​Ma`arif, 1956. al-​Shawkani. Nayl al-​awtar min ahadith sayyid al-​akhyar. Edited by Wahbah al-​Zuhayli. Beirut: Dar al-​Khayr, 1996.

Secondary Sources Ali, Kecia. Marriage and Slavery in Early Islam. Cambridge, MA: Harvard University Press, 2010. Boekhoff, Nicolet, Kees Verseegh, and Joas Wagemakers, eds. The Transmission and Dynamics of the Textual Sources of Islam: Essays in Honour of Harald Motzki. Leiden; Boston: Brill, 2011. Brockopp, Jonathan. Early Maliki Law:  Ibn ‘Abd al-​Hakam and His Major Compendium of Jurisprudence. Leiden: Brill, 2000. Brown, Jonathan. Hadith:  Muhammad’s Legacy in the Medieval and Modern World. Oxford:  Oneworld, 2009. Clarence-​Smith, William Gervase. Islam and the Abolition of Slavery. London: Hurst, 2006. Gribetz, Arthur. Strange Bedfellows: Mutʻat al-​nisaʼ and mutʻat al-​ḥajj: A Study Based on Sunni and Shiʻi Sources of tafsir, hadith and fiqh. Berlin: K. Schwarz, 1994. Haeri, Shahla. Law of Desire: Temporary Marriage in Shi‘i Iran. Syracuse, NY: Syracuse University Press, 1989. Haider, Najam. “Prayer, Mosque, and Pilgrimage: Mapping Shiʿi Sectarian Identity in 2nd/​8th Century Kufa,” Islamic Law and Society 16, no. 2 (2009): 151–​174. Hakim, Avraham. “Conflicting Images of Lawgivers: The Caliph and the Prophet Sunnat Umar and Sunnat Muhammad,” in Method and Theory in the Study of Islamic Origins, edited by Herbert Berg, 159–​179. Leiden: Brill, 2003. Hallaq, Wael. A History of Islamic Legal Theories:  An Introduction to Sunni Usul al-​fiqh. Cambridge: Cambridge University Press, 1997. Hidayatullah, Aysha. “Mariyya the Copt: Gender, Sex and Heritage in the Legacy of Muhammad’s umm walad.” Islam and Christian–​Muslim Relations 21 (2010): 221–​243. Little, Donald P. “Six Fourteenth Century Purchase Deeds for Slaves from Al-​Ḥaram Aš-​Šarîf,” Zeitschrift der Deutschen Morgenländischen Gesellschaft 131, no. 2 (1981): 297–​337.



Remembering the Umm al-​Walad

j 323

—​—​—​. “Two Fourteenth-​Century Court Records from Jerusalem Concerning the Disposition of Slaves by Minors,” Arabica 29, no. 1 (1982): 16–​49. Marmon, Shaun. “Domestic Slavery in the Mamluk Empire: A Preliminary Sketch.” In Slavery in the Islamic Middle East, edited by Shaun Marmon, 1–​24. Princeton, NJ: Markus Wiener, 1999. Mattson, Ingrid. “A Believing Slave Is Better Than an Unbeliever:  Status and Community in Early Islamic Law.” PhD diss., University of Chicago, 1999. Melchert, Christopher. “Traditionist-​Jurisprudents and the Framing of Islamic Law,” Islamic Law and Society 8, no. 3 (2001): 383–​406. Mirza, Younus Y. “Ibn Kathir (d. 774/​1373): His Intellectual Circle, Major Works and Qur’ānic Exegesis.” PhD diss., Georgetown University, 2012. Motzki, Harald, Nicolet Boekhoff-​van der Voort, and Sean W. Anthony. Analysing Muslim Traditions: Studies in Legal, Exegetical and Maghāzī Ḥadīth. Leiden; Boston: Brill, 2010. Mustafa, Abdul-​Rahman. On Taqlid:  Ibn al Qayyim’s Critique of Authority in Islamic Law. Oxford: Oxford University Press, 2013. Rapaport, Yossef. “Women and Gender in Mamluk Society.” Mamluk Studies Review 11 (2007): 1–​47. Rosenthal, Franz. “Ibn Sayyid al-​Nās.” In Encyclopedia of Islam, 2nd ed., P. Bearman et al., eds. Leiden: Brill Publishing Online, 2012. Schacht, Joseph. The Origins of Muhammadan Jurisprudence. Oxford: Clarendon Press, 1967. —​—​—​. “Umm al-​Walad.” In Encyclopedia of Islam, 2nd ed., P. Bearman et al., eds. Leiden: Brill Publishing Online, 2012. Toledano, Ehud R. Slavery and Abolition in the Ottoman Middle East. Seattle:  University of Washington Press, 1998. Tucker, Judith. Women in Nineteenth-​Century Egypt. Cairo: Cambridge University Press, 1985. Zilfi, Madeline C. Women and Slavery in the Late Ottoman Empire:  The Design of Difference. New York: Cambridge University Press, 2010.

Epilogue Avenues to Social Mobilit y Available to Courtesans and Concubines

Kathryn A. Hain

i  Khayzuran’s manipulation of three generations of Abbasid caliphs and courtiers make her probably the best known concubine of the Abbasid court, a place and time still famous as the backdrop for the stories of The Arabian Nights. As the mother of al-​ Hadi (r. 785–​786) and Harun al-​Rashid (r. 786–​809), she provides us an early example of the social mobility and wealth that an enslaved woman could attain in Islamic society. Nabia Abbott, a pioneer scholar in English on early Muslim women, wrote a biography of Khayzuran. According to her work in the Arabic sources, slavers in Yemen kidnapped this lithe girl, named her Khayzuran (“Slender Reed”), and put her through musical training in Mecca to increase her value before selling her to the caliph on the Hajj. After Khayzuran secured power in the palace, she sent royal envoys to Yemen to search for her family. They found her father to be no more than a roughly dressed freedman working in the fields. This slave concubine who became queen mother influenced royal appointments and dominated the courtiers, her spouse, and her sons, enabling her to funnel incredible wealth to her own treasury. At the time of her death, it was recorded that her yearly income consumed half the land taxes of the empire. Her estate included a huge palace with over 1,000 slaves to serve her, gold, jewels, and 18,000 silk brocade dresses. Not bad for a skinny farm kid from Yemen.1 Khayzuran’s passage through the slave market of Mecca made possible her rise from farm to music school to her own palace. She is only one example of many concubines and singing courtesans across the breadth of the Islamic world who made good or achieved 324

Epilogue

j 325

social mobility because of their slavery. Sexual attraction played a role in their rise. Since royals and elite men, however, had access to hundreds, sometimes thousands of women, the women who rose above the fray used additional means, besides sex, to gain and maintain control over the reins of power in the palace or an elite household. This closing chapter draws on the contributions to this volume to illustrate the means by which these women benefited from Islamic traditions and Muslim culture to achieve social mobility, wealth, and, in rare cases, political power. A stream of foreign slave women flowed into Muslim cities and courts for over 1,200 years, from all directions and from numerous ethnic origins. These women arrived as human booty captured in war, yearly tribute payments from the defeated, gifts sent to rulers from hopeful allies, kidnap victims of pirates and raiders, or commodities transported from outlying slave markets to the higher-​paying urban Middle Eastern, North African, and Andalusian markets. Slaves also birthed children, some of whom were not freed, adding to the native slave population. Demand for slave women came from the highest rulers to poor men who could not afford the dower needed to marry a free woman. Muslim royals, especially from the Abbasid era on, adopted the elaborate court culture traditional to kings of Mesopotamia, and across Eurasia to China and India. These courts required large entourages of richly dressed musicians, entertainers, eunuchs, and concubines to demonstrate power, wealth, and sophistication.2 Rulers added slaves for their own harems and entertainments but also used slave women as gifts for their soldiers and officials, or for exchange in ransom treaty agreements. Elite men and women purchased female slaves to present as gifts to curry favor or to serve as retinue. Literary salons, brothels, and drinking establishments bought women to serve their clientele. And elite slaves could and did own their own slaves. Captive girls’ chances for social mobility depended initially on who was buying and for what purposes on the day they were offered for sale. Ibn Butlan, an eleventh-​century Nestorian Baghdadi physician, suggests that for the great majority of enslaved women, their prospects were bleak. He wrote a medical treatise, which included a section on stereotypes of slave women by ethnicity to guide customers in their purchase. His descriptions reveal the domestic tasks for which most slave women were purchased. He characterizes women from India, for example, as “good for childbirth”; women from Medina as best suited for training as singing girls; and Berber girls, particularly if you could “find a pale one,” as the best choice for motherhood and sexual pleasure. “Zanj,” or, probably, East African slaves, were excellent at dance and rhythm but also could endure hard work. Women from Allan (Azerbaijan) were better for service than pleasure since they were trustworthy; Greek women made the best financial administrators because they were meticulous and not very generous; and Armenians were the “white” slaves best suited for hard work.3 As Muslim households grew in prestige, seclusion, and numbers, the need increased for female slaves who could carry water, cook, serve gracefully, run errands, dress hair, sweep and scrub, wash, spin, weave, and sew clothing. Female administrators were

326 i

Kathryn A. Hain

needed to manage budgets, households, wardrobes, and jails and to train and discipline other slaves.4 Wet nurses and nannies were purchased to care for children. Sexual use was only one of many reasons to purchase a woman. Most faced lives of dull and physically demanding domestic work. For a few of the women, however, opportunities for social mobility eclipsed the usual, nameless servitude. This chapter examines the elements of Islamic society that allowed the talented and lucky to rise. These women hailed from different regimes, geographical regions, and eras over a span of 1,200 years. The contributors to this volume have extracted evidence from diverse texts: the Qur’an and its commentaries, biographies of Muhammad, collections of Hadith, hagiographies, biographical dictionaries, catalogues of singers, poetry and song collections, chronicles, legal treatises, essays, letters, and government archives. These women had in common that, like Khayzuran, the singing farm girl from Yemen who became queen mother, their slave status provided the opportunity to forge new identities. The first section considers how the role of courtesans and concubines in Islamic religious traditions validated their place in Muslim society. A second and longer section demonstrates how Muslim cultural practices provided social mobility to enterprising female slaves. Concubines in Islamic society, with few exceptions, were slaves. Sex with your own property was not considered to be adultery (zina). Owners purchased the sexuality of the enslaved along with their bodies. Sexual relations with a slave remained unencumbered by the legal and financial obligations due with regard to a free wife and her family. Islamic marriage laws obligated a husband to give his wife/​wives sexual satisfaction, an opportunity to conceive, and financial support. For example, birth control, coitus interruptus, was legal by Islamic law. The practice required the permission of wives but not that of concubines.5 Divorcing a wife meant an expensive loss in the dower or divorce settlement while selling a slave woman, in many cases, provided financial gain. Limited rights for slaves made them more useful. Concubinage in Islamic law existed as a type of condoned coupling where the resulting children were born with the religious identity and the free or slave status of their father. In comparison, Christian scriptures promoted celibacy outside monogamous marriage.6 The silence of New Testament writers on sex with slaves contrasts with the Qur’an, which endorses sex with those whom “your right hand possesses.”7 This religious sanction for sexual relations with slave women meant concubines and their children were not stigmatized like their counterparts were in Christendom. Other Islamic religious traditions converged to further legitimize the role of concubines. One set of traditions surrounds the houris. Hadith texts describe paradise as a garden filled with singing houris who functioned similarly to courtesans, that is, trained entertainers (qiyan or jawari) who dominated the cultural scene in Baghdad at the height of the Abbasid caliphate (c. 750–​945), and in tenth-​century Cordoba under the Umayyads. These women appear to have helped shape the Islamic traditions of paradise.8 The minimal description of houris in the Qur’an had exegetes develop the concept of

Epilogue

j 327

these heavenly companions. These descriptions of houris parallel qiyan in many respects, providing pleasure, leisure, and culture (song) to their male (and female) companions. Not only did Islamic theological texts use the qiyan as models for houris in paradise, but Islamic traditions enshrined two concubine mothers. Arab Muslims considered themselves the physical and spiritual heirs of Ishmael (Isma’il), the firstborn son of Abraham (Ibrahim), born to Hagar (Hajar), the slave girl belonging to his wife, Sarah. The foundational stories of the faith immortalize Hajar, the Egyptian concubine, as the matriarch of the Arab people, the mother of a prophet, a pious believer, and “Mother of the People the Water of Heaven.”9 That she was a slave and a woman did not impede Hajar’s status as a forerunner of Muhammad and the Muslims. Muhammad’s relationship with Mariya the Copt, another Egyptian slave girl, provided the basis for the religious sanctioning of sexual enjoyment of slave women. The Prophet Muhammad owned four concubines remembered by name, in addition to his multiple wives. One girl, in particular, captured his attention (and the attention of his biographers). According to the hagiography, the patriarch of Alexander sent an old eunuch and two Coptic sisters to Muhammad as a gift. Muhammad gave one sister to a follower but took Mariya the Copt, with her beautiful curly hair and fair skin, as a concubine (jariya). She bore him a son, Ibrahim, the only child that Muhammad produced after the death of his first wife, Khadija (d. 619). Despite the acceptance of concubines and resulting offspring by the prophet, other sources purporting to be late Umayyad or early Abbasid show that this issue was debated in early Islamic society. Men born of concubines, who aspired to rule, used the concubines Hajar and Mariya the Copt to assert legitimacy and right to rule as Muslims.10 When a rebel castigated the second Abbasid caliph for having a slave mother, Abu Ja`far al-​Mansur invoked the prophet Muhammad who had a child by Mariya the Copt. The invocation of the examples of Hajar and Mariya the Copt show how slave mothers and their children had an impact on Islamic identity, making belonging and authority more inclusive.11 One reason that this argument proved so successful stemmed from the demographic shift brought about by increasing numbers of Muslims born to foreign concubines.12 Equality before Allah as an ideal, if not always as a practice, could be considered yet another religious factor encouraging acceptance of slaves. Theoretically, race and ethnicity were not barriers to acceptance in the Islamic umma. Concubines who made up one part of the large households of the elite were, by and large, foreign: a multi-​ethnic, multi-​ lingual, multi-​religious collection of women from both conquered lands and frontier regions.13 When conquest ended by treaty, many frontier regions sent a yearly tribute that included slaves. Slave traders also brought women long distances to the Muslim markets. The children born to these captive women may often have been accorded the same treatment as children of freeborn wives. Biographical dictionaries show that Muhammad’s Qurayshi tribe gave full family and tribal identity to the sons of concubines. Furthermore, these mixed-​blood sons made a number of marriages to Qurayshi

328 i

Kathryn A. Hain

women, a status marker which even non-​Qurayshi Arabs could access only with difficulty in the post-​conquest era.14 A century after Muhammad, at the end of the Umayyad Dynasty (661–​750), royals began shunning free Muslim wives to reproduce through slave women. In five centuries of the Abbasid Caliphate, perhaps only three caliphs were born to free women. The others were the offspring of foreign concubines: women captured, gifted, or purchased from Afghani, Turkish, Slav, Ethiopian, Greek, and Persian lands.15 Despite this seemingly tidy list of the known ethnicities of the mothers of Abbasid caliphs, it is often impossible to determine the ethnic origins of slaves.16 The detailed descriptions of slaves in Arabic legal documents did use ethnic origin to classify slaves and to prevent fraud, but, even then, ethnic terms tended to be fluid. Added to the linguistic fluidity of a slave’s ethnic description was slave traders’ duplicity. Ibn Butlan, in his treatise, speaks to unscrupulous traders altering the appearance of slave women. “How often they turned the blue eyes kohl-​dark, the sallow cheeks glowing pink . . . turned the blond hair black, the frizzy straight.”17 Social mobility for some slave women may have meant having their backgrounds and appearance fraudulently altered to match whichever ethnicity was desired by that week’s customers in the slave suq. Encouragement to manumit slaves, enshrined in the Qur’an and law, likely contributed to social mobility. Manumitting slaves earned their owner eternal rewards. The Qur’an advocated manumission of slaves as the act of a righteous person or as a religious boon.18 According to the Qur’an, freeing slaves could bring forgiveness for specific sins: killing,19 a futile oath,20 and reneging on a zihar divorce.21 Key hadith also support manumission: “He who has a slave-​girl and teaches her good manners and improves her education then manumits and marries her, will get a double reward.”22 Concubines, however, sometimes had a surer route to manumission than their owners’ desire for spiritual coinage; under certain conditions their wombs could provide escape from slavery. A slave who bore a child to a free man, known as an umm al-​walad, could not be sold, in most circumstances, and at her owner’s death, she was to be freed. Scholars usually assume that Shari`a law mandated freedom for the umm al-​walad. This supposed “freedom via fecundity,” however, was contested between schools of Sunni and Shi‘ite law, and within Sunni law for centuries. This debate is just one example of how Islamic legal history serves as a significant source on the practices of slavery.23 Slave mothers were sold during the prophet’s time and into the time of the second caliph, and, according to early Sunni traditions, after the death of Muhammad, it was Umar, the second caliph, who decreed that slave mothers should not be separated from their children.24 Manumission for slave mothers, even when it was not contested, did not always improve a woman’s position in her later years. For a woman whose children could not provide for her, because they were female or had died, freedom meant that the heirs to her owner no longer had the mandated responsibility to feed, clothe, and care for their mother’s rival.25 Ironically, for at least some women, freedom might not have been an

Epilogue

j 329

improvement over the safety net that slavery provided against hunger and want in their old age. The use of luxury slaves in Islamic society to project wealth and elite status meant, in some sense, that female slaves provided social mobility for those who purchased them. And, as vessels of entertainment and reproduction, slave women were given the opportunity, at least in cases, to become powerful and wealthy in their own right. The following examples distilled from the lives of slave women illustrate the various ways that patterns of Islamic society and culture afforded concubines and courtesans opportunities to move from enslaved outsider to power holders. Slave women and their children gained influence through sheer demographic presence. Biographical dictionaries provide social history referencing the demographic contributions of slave mothers, at least, for the families important enough to be memorialized. Examination of one of the earliest Arabic biographical dictionaries, the ninth-​century Kitab Nasab Quraysh of al-​Zubayri, helps to illuminate the role of feminine slavery during the first century of Islam. A quantitative analysis of the maternal records of 3,000 Qurayshi tribesmen reveals how the influx of war captives absorbed into households changed each generation of the elite.26 Concubinage appears to have been rare in pre-​Islamic Arabia—​a picture that concurs with pre-​Islamic poetry and traditional narratives. From Muhammad’s time, and for the next 10 generations listed in the records of the Quraysh tribe, the percentage increases of offspring from slaves. Twelve percent of the generation of Muhammad’s grandsons were born from slave mothers. That percentage continues to rise sharply for the next three generations until it peaks, five generations after Muhammad, when 42.1% of all Qurashis were being born to concubines. The origins of concubines, elite or imagined, sometimes mattered. Rape, nakedness, long-​distance travel, and multiple exchanges usually erased the identity of all but royal captives. Once acquired, slave owners changed a new slave’s name, language, and place of residence, completing the slave’s “rebirth” into Islamic society. Occasionally, however, Arabic texts emphasized the elite origins of given women before their enslavement. Royal or religious backgrounds of slave women were remembered or even manufactured. The first concubine-​born caliph, the Umayyad Yazid III, praised his mother as a Persian princess from the royal Sasanid line. He bragged of his doubly noble lineage via captives: “I am the son of Khusraw; my father is Marwan. One grandfather is a Caesar, the other a Khaqan.”27 Shi‘ite genealogies of the 12 Imams list the pre-​slavery status of their concubine mothers.28 Most likely, however, their hagiographies created royal lineages for concubines who arrived via the usual anonymity of the slave markets. Members of the early line of Shi‘ite Imams were described as having royal Persian mothers. For instance, Ali ibn Husayn, the fourth Imam, was supposedly the son of Shahrbanu, daughter of the last Sasanian king, Yazdegerd III (r. 632–​651).29 As the Persian Empire faded from memory, so-​called royal Byzantine women took the place of Persian princesses. The concubine mother of the 12th

330 i

Kathryn A. Hain

Iman was portrayed as a descendant of “Peter, delegate of the Messiah” and the granddaughter of a Byzantine emperor.”30 Those who promoted the line of 12 Imams felt the need to elaborate the details of an elite, saintly, and unviolated background for what were perhaps the obscure backgrounds of ordinary concubines. Imams and caliphs were not the only ones who used the status of their concubines to bolster their claims to power. The Almohad ruler Abd al-​Mu’min enslaved the daughters of his defeated rivals to create his own ruling class, a brutal use of sexual politics.31 Despite Shari`a law forbidding the enslavement of fellow Muslims, this conqueror of North Africa and Iberia chose and impregnated daughters of the elite from each defeated city so that their sons by him could govern regions from which each mother originated. Mothers and sons served as human bridges between the conquering Almohads and the diverse sects, tribes, and regions over which they wished to rule. The tactic proved to be politically potent. The Almohad sect had remained small until al-​Mu’min took over the movement. Through his able military leadership, astute political maneuvering, and the targeted use of his seed with elite captives, Abd al-​Mu’min founded the most successful dynasty in the Islamic West in the Middle Period. Extensive education could provide talented slave women with yet more opportunity. Debating the merits of literary or religious texts or expertly handling weapons while on a hunting trip are examples of skills that enhanced a woman’s market value. Promising courtesans were also schooled in the traditional skills of playing instruments, composing poetry, singing, writing love notes, and expertise in the seductive arts including the language of scents, flowers, sexual techniques, and emotional manipulation.32 Slave traders invested in their stock’s intellectual abilities as value-​added upgrades to raise the slave’s price and desirability. An example of the education that qiyan received from slave traders comes from the court of Hudhayl ibn Razin, who founded a short-​lived dynasty in Aragon.33 This minor ruler is remembered in the sources for utilizing talented qiyan to compete in ostentatious displays of court culture. He paid 3,000 dinars for one extraordinary singing girl who could also write, create calligraphy, talk about medicine or natural history, and repair and use a variety of weapons. To showcase her musical abilities, this aspiring ruler purchased another 150 slave girls and 60 Slavic eunuchs to chant the Qur’an and provide her with an orchestra. Profligate spending perhaps explains why Ibn Razin’s dynasty was short-​lived, but it also reveals how competition for prestige via court displays depended on the training and reputation of their courtesans. Sources reveal that some qiyan were not passive in negotiating their ownership. Once slave women were ensconced in households and had learned both the local language and how to wield their charms to influence their owners, their agency increased. Anecdotes about elite slaves reveal their influence over whether they would be “gifted” or sold.”34 For example, Utba, a slave girl of the early Abbasid caliphal household of al-​Mahdi, had an admirer, Abu al-​Atahiya, who tried to win her by plying the caliph with panegyrics. In the triangle of love-​smitten poet, caliph, and slave woman, Utba’s resistance to being gifted

Epilogue

j 331

to someone she considered her inferior triumphed over both the unwanted suitor and the ruler. Clearly, some highly valued slave women were able to influence their own fate. Wealth also signaled social mobility. A very few slaves appear to have become fabulously wealthy. Evidence from the Ottoman archives indicates that women of the court gained wealth that they used, in turn, to gain power. A night of musical or sexual service to the ruler was usually compensated with jewels. When concubines managed to conceive and bear children, they would begin to receive a regular allowance.35 A harem mother’s savings and business investments from her time as a concubine often influenced the decision of which royal son would be crowned. The mother able to provide the most impressive gifts or the largest “donative” to the army and palace staff to secure their loyalty could put her son on the throne against the claims from rival half-​brothers. Once a royal son was enthroned, his mother was made head of the harem and often awarded a sizable portion of the tax revenue. Bribes and gifts for her influence over her son, the sultan, and appointments to lucrative government positions added to her considerable income stream. Palace women in numerous Muslim dynasties hired their own financial administrators and funded their own waqfs (religious trust funds for charities).36 A survey of these waqfiyyat, endowments in eighteenth-​century Mamluk Egypt, reveals the access that slave women had to wealth. Women who entered the society as slaves, in comparison to free women, had “by far, the largest and most lucrative” endowments.37 The wealth of queen mothers and favorite concubines could serve as a dynasty’s reserves or treasury in times of crisis.38 This dependence on the harem for finances became so entrenched in the Abbasid dynasty that a Greek concubine, Shaglub, the mother of al-​Muqtadir (r. 908–​929), financed the defense of Baghdad during the Qaramita attacks and supported the war against Byzantium. After her son was killed, the next caliph faced empty treasuries and feared assassination at the hands of his Mamluk army, who demanded their donative. In desperation, he turned to the former queen mother for finances. When she confessed to having no more than a few trunks of clothes and jewels, he had the elderly, sick woman hung upside down by one foot and scourged a hundred times.39 Status in the royal harem could provide vast wealth, but it was no guarantee of protection. Sources on the qiyan focus on finances: how much these girls cost, the number of background musicians they had, and the expensive gifts they were given. The economics of the courtesan business, including the investment of purchase price, time, and cost of lessons, were all speculative based on how particular girls would turn out.40 The gamble paid richly when accomplished young women sold for 10,000 to 100,000 dirhams compared to common male slave prices of 20–​50 dirhams.41 Once a musician was purchased, the owner incurred further expenses: costly robes, jewelry, perfumes, incense, choice foods, as well as lute strings, writing instruments, and continual training. Some singers required accompanying musicians. Their expenses could be defrayed by hiring the performer out for musical and/​or sexual services.42 Al-​Jahiz decries

332 i

Kathryn A. Hain

the practice of loaning singing girls and warns that the girls were trained to manipulate men’s emotions until the besotted lovers had emptied their purses in purchasing gifts for both the girl and her owner.43 The fame that could come to creators of popular poetry and song provided social mobility to the few who could rise above the fierce competition. Their ultimate goal was purchase by the palace or an elite patron who would free them. Singers purchased by the palace, like Khayzuran, entered the harem competition to produce a son. If this son survived and took the throne, he could give his mother a second career as queen mother. Some singers managed to establish their own literary salons, capitalizing on their fame by buying their own young slaves, training them in their repertoire of musical compositions, and hiring them out. A secondary stream of income existed for popular singers who wrote for caliphs and others of elite circles for which they would receive rich gifts. Failure to compete at the highest levels left probably most female entertainers to entertain in less impressive settings, such as taverns and brothels.44 After the Abbasids adopted the Persian and Byzantine-​style seclusion of the royal family, slave women and the eunuchs who partnered with them gained, in some cases, considerable influence. For example, in the Ottoman court, proximity to the sultan gave women of the harem increased power leading to tense jostling between the harem, the vizier who collected taxes, and the military—​all seeking influence.45 When a ruler was a minor child or was weak, preferring the seclusion provided by the women’s quarters, then military or administrative heads could not enter this inner sanctuary without breaking social codes of decency. Women and eunuchs could rule the realm from the harem by virtue of their ability to restrict access to the ruler and influence his appointments of favorites. Queen mothers of two dynasties of Central Asian origin reveal how machinations of power from the harem operated in fifteenth-​century Timurid Iran and the eighteenth-​ century Ottoman Empire. Turkic custom allowed all male family members to compete for the throne. This dynamic meant that a prince aspiring to rule needed alliances to overcome his male relatives’ claims to the throne. Islamic law prescribed no more than four wives so the Timurids, the Turkish-​Mongol dynasty of intermingled descendants of Tamerlane and Chinggis Khan, allied themselves with powerful families by taking their daughters as concubines. Their steppe culture prevailed over Islamic juridical boundaries that required that concubines could only be non-​Muslim slaves.46 The freeborn women kept their Muslim or Turkic names but would be listed as concubines without the required passage through slavery. One contemporary, Babur, noted that of Sultan Husayn Bayqara’s 14 sons, all but three were “sons of adultery.”47 He implied that the majority of these Timurid sons were illegitimate, because their mothers were free Muslims, and, therefore, their sexual relationship could not be sanctioned by Islamic law. One of these Turkic freeborn consorts, Khadija Begi Agha, was a Muslim amir’s daughter.48 The sultan married her before her second son was born. Thus, her second son was counted as one of the sultan’s three legitimate sons. Khadija Begi Agha pushed her way out of the usual obscurity of Timurid harems by her efforts to ensure the political success

Epilogue

j 333

of this son—​efforts that caused the death of the popular heir and precipitated a major political crisis. Her assets were the backing of a prominent family and the support of the army that reputedly followed her orders.49 One of Khadija’s competitors for influence, Barbar, wrote about the sultan, “By listening to the words of women, he removed himself from the circle of those of good repute.”50 The courtier had been outfoxed by a woman drawing on her prominent contacts. Similarly, in the Ottoman court, queen mothers wielded considerable influence even after the end of the so-​called sultanate of women.51 Rabia Gülnuş Emetullah Valide Sultan (d. 1715) serves as a prime example. Her training and conversion to Islam occurred under the harem rule of the powerful and ruthless Valide Sultan, Hatice Turhan Sultan.52 She rose from the anonymous pool of concubines when she produced two sons, both future sultans. Gülnuş survived in the position of haseki (favorite concubine) for an unusually long period, which meant she acquired numerous gifts and a large income, giving her the financial resources she needed to buy influence.53 She gained an understanding of the structure and power holders of the Ottoman system as a participant in imperial ceremonial processions.54 The greatest weapon in a concubine’s arsenal was finances. Even as early as her haseki years, Gülnuş kept Mehmed IV (r. 1648–​1687) on the throne by bribing the troops when they rose up against the monarch. She later supplied funds for the enthronement of each of her sons. She used her haseki income to build networks of support. She allied herself with the chief eunuch, Yusuf Agha, by appointing him as her representative in the public eye, the administrator of her waqf, which provided a hospital and public kitchen in Mecca and mosques, soup kitchens, public water fountains, libraries, and primary schools elsewhere. These charities provided prestige and public visibility. Gülnuş gained allies among statesmen by interceding for their protection or promoting their careers. Several viziers who did not please her saw their careers—​and sometimes their lives—​cut short. The Grand Vizier, Kara Ibrahim Pasha, was executed because of the failed siege of Vienna in 1683, after “considerable lobbying effort on the part of Gülnuş Sultan and the court eunuchs.” She also played a key part in selecting elite staff for her eldest son, Mustafa, when he came to the throne at 31, from among her own clients. During the revolt that dethroned her first son, the traditional approval of the Valide Sultan was so critical that the rioters sent her a formal letter asking for her permission to dethrone one son and crown another. Like many other Ottoman royal women, she groomed and freed female slaves, in order to marry them to promising courtiers and thereby enlarge the circle of politicians within her network of influence. When her sons came to the throne, Gülnuş’s influence reached beyond the palace. Correspondence in the Ottoman archives reveals Gülnuş’s efforts at intelligence gathering and acting as go-​ between for the King of Sweden and the Crimean Khan. Khadija, the Timurid queen mother, and Gülnuş, the Ottoman Valide Sultan, both demonstrate how concubines could influence the government from behind the walls of the harem. They used a combination of their financial resources; administrative abilities;

334 i

Kathryn A. Hain

ruthless executions of rivals; and networks of patronage with statesmen, powerful families, and the military to put and keep their sons on the throne. These women wielded considerable political clout in their own name, not just through influence over their partners and sons. They turn the Western stereotype of helpless harem women upside down. Military commanders, viziers, and even emperors turned to the most powerful women in the harem for influence and the finances that they needed. Travel also connoted social mobility. Along with the wealth and power that queen mothers possessed came the opportunity to make a magnificent Hajj, the one trip that a sequestered Muslim woman could undertake without loss of respect.55 It is possible that the women who had come to the throne from slavery were more inclined to go on the Hajj than free wives who had spent their lives secluded. Slaves, captured from frontier regions, traveled long distances to urban slave markets, giving them a wider experience of the world. Hajj journeys of former slave concubine queen mothers proved unforgettable for their ostentatious displays of wealth and piety. In addition to the hundreds of retainers, some of these women flaunted their wealth by trailing lengths of expensive silk in the sand behind their camels or carrying baskets of dirt with growing vegetables and herbs for fresh salad in the desert.56 The medieval era produced famous Muslim world travelers. One of these, Ibn Battuta (d. 1368 or 1369), wrote about his traveling companions, slave girls.57 The Moroccan global traveler, with his concubines, traveled an estimated 73,000 miles (three times the distance of Marco Polo). He married and divorced an estimated 20 wives, owned numerous concubines, and fathered uncounted children. Because free wives bolstered by the support of their natal family, regulations of Islamic law, and requirements of modesty could refuse the hazards of travel, Ibn Battuta divorced or deserted them when he moved on. He took concubines in their place for travel companions. Ibn Battuta provides vignettes of slave women from different ethnicities whom he acquired as gifts. Through him, we can witness these women’s lives as they traveled to new lands by caravan or ship and suffered hardships of weather, shipwreck, piracy, kidnapping, robbery, giving birth on the road, and death. Ibn Battuta described one girl as “the one I love” when he put her on a raft to survive a shipwreck while another of his slave girls hung on a rope to keep from drowning. Ibn Battuta’s accounts of these girls’ lives is unusual because they were commoners. Their geographic mobility resulted from their slave status and being purchased by a “traveling man.” It may also have provided them social mobility when, for example, Ibn Battuta’s slaves were “appropriated” by the governor of Sumatra. Entering his palace gave them the rare chance to bear a child who might someday rule. As Ibn Battuta’s desertion of his freeborn wives attests, the abundance of slave women impacted free Muslim women in several ways. Not only men purchased young women for their households.58 Elite women, free and slave, owned jawari as well. Zubayda, the favorite wife of Harun al-​Rashid and one of the only free women to live large in Abbasid politics, is said to have owned a retinue of 1,000 eunuchs and girls. She devoted care to her slave’s training and had 100 girls trained to chant the Qur’an so the sound reverberated

Epilogue

j 335

through her palace all day.59 Zubayda also paid high prices for qiyan. When her son, the Caliph al-​Amin, was assassinated, she fell from power, lamenting that what she missed most was the loss of Arib, one of the great Abbasid singers, who later went on to fame, wealth, and freedom.60 Arib’s confiscation from the household of the former queen mother never slowed the trajectory of her career. This courtesan bragged of passing through the beds of eight caliphs. Eventually, she was freed by al-​Mu`tasim (r. 883–​842) and established her own literary salon featuring a stable of slave girls whom she had trained to sing her compositions.61 She joined elite women, free and slave, who turned their households into finishing schools or music establishments in the profitable business of producing accomplished, fashionable courtesans. Besides profiting from resale, elite women used their slaves in other ways. They could be trained, freed, and married to political aspirants, thus creating patronage networks.62 Free women could not leave the seclusion that befitted their status, so they used their musically talented slaves to perform their compositions. For al-​Harun’s talented royal sister, Ulayya bint al-​Mahdi, this was the only way she could transmit her songs beyond the immediate family.63 Slaves could leave the women’s quarters to carry messages. Beautiful slaves could be presented to men by their female relatives or wife. If a particular slave monopolized a husband’s attentions, his wife might win him back by giving him a new girl. When al-​Harun was smitten by the singing of Dananir, his wife Zubayda distracted him with the gift of 10 slave girls, among whom three became mothers of his sons.64 Al-​ Jahiz explains this strange competition thus, “Should a woman of the royal harem possess a slave girl whom she knows the king desires and rejoices in, then it is her duty to present the king with this girl . . . if she does this, then the king should give her preference over all his women.”65 Despite evidence that elite and royal women used talented slave girls for their own prestige, entertainment, companionship, patronage, profit, and power, some modern scholars have blamed concubines for negatively circumscribing Muslim women’s lives with the need to be submissive to compete with slaves. Free Arab women in pre-​and early Islamic society acted more independently. Several had ruled as queens in the Roman era or led armies like the famed Zenobia (Zaynab) of Palmyra or Aisha, Muhammad’s youngest wife, in the “Battle of the Camel.”66 With conquest, family dynamics changed. Rulers used concubines for reproduction to avoid sharing power with other families. Even if royals married free women for political advantage, the sons who took the thrones were the sons of slaves. This transformation was not limited to the palace. One tenth-​century Muslim father in al-​Andalus complained that he had to pay men to marry his daughters instead of the customary dower paid to fathers because of the abundance of those “cheap, Spanish slave girls.”67 When the same cheap slave girls became the mothers of the next generation, they usurped the power of free women through their sons. Medieval Muslim literature bears witness to the growing disenfranchisement of free women. The majority of early Muslim and Umayyad women who appear in literature are

336 i

Kathryn A. Hain

designated as hara’ir, free. By the Abbasid era, the women mentioned are almost all slaves or freed slaves.68 Reproductive politics favored slave concubines. While slave women in Islamic society could use their wits, charm, talent, and beauty to achieve social mobility, they remained vulnerable. Courtesans were privileged compared to other women, free or slave, yet, even so, a few were remembered for their protests against their enslavement and their treatment by slave merchants and masters. Fadl claimed that her stepbrother sold her illicitly, Shariya that since her Qurayshi father was free, she and her mother should have been freed. Mahbuba, among the best known of the singers, was gifted to al-​Mutawakkil (r. 847–​861) and “she held a lofty place in his affections unrivalled by any other person.”69 But when he was assassinated, she, along with the rest of his harem, became booty distributed to the military elite. When one general who had heard of her fame demanded a performance, she, under great protest, sang not to gain his admiration but to give voice to her broken heart and anguish at the murder of her beloved master. She was imprisoned and never heard from again. These women’s voices, even when mediated by male authors, cry out at the injustice of their situation.70 Some slave women, though relatively few, greatly profited from slavery in Islamic society. The Islamic religious traditions had promoted the status of foreign female slaves by honoring Abraham’s and Muhammad’s concubines as “Mothers of the Faith,” by describing the houris of paradise as heavenly qiyan, by promoting manumission of slaves, and by emphasizing egalitarianism for slaves and their children. Muslim culture accepted concubine mothers because of their demographic weight and real or imagined elite status that they could carry from their previous lives. Slave women occasionally trumped free Muslim women in opportunities. Qiyan could receive excellent educations, negotiate their exchange, and manipulate their patrons. As royal consorts, slave women had finances to fund religious charities and build political networks, becoming extremely wealthy and powerful. Through their sons, some of them even ruled the empires responsible for their enslavement. Notes 1. Abbott, Two Queens. 2. See Duindam, “Royal Courts,” 1–​2. 3. Ibn Butlan’s treatise, Risala fi shira al-​raqiq wa taqlib al-​abid (Epistle on the Purchase and Commerce in Slaves), was later incorporated into manuals (Hisbat al-​suq) for market inspectors who were charged with preventing fraud. For an excerpt, see Lewis, Islam from the Prophet Muhammad, 243–​252. 4. See El-​Cheikh, “The Qahramana,” 41–​55. 5. See Zilfi, Women and Slavery, 112. 6. The New Testament scriptures (I Timothy 3:2, 12) mandate “one wife” for church leaders which is Roman law and a departure from the Jewish scriptures where Abraham, Jacob, David, Solomon, etc., had children by multiple wives and concubines.

Epilogue

j 337

7. Verse 4:24 (Surat al-​Nisa). 8. Rustomji, Chapter 13 in this volume. 9. See Urban, Chapter 11 in this volume. 10. See Urban, Chapter 11 in this volume. 11. See Urban, Chapter 11 in this volume. 12. See Robinson, Chapter 1 in this volume. 13. Cortese and Calderini, Women and the Fatimids, 78. 14. Robinson, Chapter 1 in this volume. 15. Caswell, Slave Girls, Appendix III. 16. De la Puente, Chapter 6 in this volume. 17. De la Puente, Chapter 6 in this volume; see Caswell, Slave Girls, 15–​16. 18. See, for examples, verses 2:177 (Surat al-​Baqqara) and 90:13 (Surat al-​Balad). 19. Verse 4:92 (Surat al-​Nisa’). 20. Verse 5:89 (Surat al-​Ma’ida). 21. Verse 58:3 (Surat al-​Mujadila). A zihar divorce proclaims that your wife is to be as your mother, and therefore Islamic law forbids the reversal of this form of divorce. 22. See al-​Bukhari, Translation, 3:46, no.720. 23. Mirza, Chapter 15 in this volume. 24. Mirza, Chapter 15 in this volume. 25. See al-​Bukhari, Translation, 3:46, no. 721: “Your slaves are your brethren upon whom Allah has given you authority. So, if one has one’s brethren under one’s control, one should feed them with the like of what one eats and clothe them with the like of what one wears.” 26. Robinson, Chapter 1 in this volume. 27. Quoted by Urban, Chapter 11 in this volume. For other examples of imputed imperial status of enslaved women, see Dann, Chapter 12 in this volume. 28. Dann, Chapter 12 in this volume. 29. See Nashat, Women in Iran, 56; Dann, Chapter 12 in this volume. 30. Urban, Chapter 11 and Dann, Chapter 12 in this volume. 31. Empey, Chapter 7 in this volume. 32. Gordon, Chapter 2 and Nielson, Chapter 4 in this volume. 33. Reynolds, Chapter 5 in this volume. 34. Sharlet, Chapter 14 in this volume. 35. See Peirce, Imperial Harem, 214–​216. 36. Argit, Chapter 10 in this volume. 37. See Fay, Middle East Studies, 92. 38. Argit, Chapter 10 in this volume. 39. See El Cheikh, “Gender and Politics,” 159. 40. Nielson, Chapter 4 in this volume. 41. See McCormick, 755–​757. 42. Nielson, Chapter 4 in this volume. 43. As cited by Nielson, Chapter 4 in this volume. 44. See Caswell, Slave Girls, 25–​29, 44, 226. 45. See Peirce, Imperial Harem, 11. 46. As described to me by Usman Hamid (private correspondence), Tamerlane’s descendants, more often than not, kept to the Islamic prohibition against more than four wives and even

338 i

Kathryn A. Hain

sponsored Sunni revivals against drinking and gambling but ignored the laws against using free women as concubines. This illustrates the tension between Turkic-​Mongol customs and Islamic law that defined steppe dynasties. 47. Cited in Hamid, Chapter 9 in this volume. 48. Hamid, Chapter 9 in this volume. 49. Hamid, Chapter 9 in this volume. 50. Hamid, Chapter 9 in this volume, citing Babur, Baburnama. 51. Argit, Chapter 10 in this volume. 52. On Hatice’s own rise to power, see Peirce, Imperial Harem, 252. 53. For the wider context, see Peirce, Imperial Harem, 212–​216. 54. Argit, Chapter 10 in this volume. 55. See Tolmacheva, “Female Piety,” 161. 56. See Tolmacheva, “Female Piety,” 162–​172. 57. Tolmacheva, “Ibn Battuta,” 119–​140. 58. Myrne, Chapter 3 in this volume. 59. See Abbott, Two Queens, 9–​10. 60. Myrne, Chapter 3 in this volume; and see Gordon, “`Arib,” 90, 92. 61. See Gordon, “`Arib,” 92–​93. 62. For an Ottoman-​era example, see Argit, Chapter 10 in this volume. 63. Myrne, Chapter 3 in this volume. 64. See Abbott, Two Queens, 140–​142. 65. As cited by Abbott, Two Queens, 140. 66. For Arab queens in the pre-​Islamic period, see Abbott, “Pre-​Islamic Queens.” 67. See Bennison, “Peoples of the North,” 173. 68. Myrne, Chapter 3 in this volume. 69. See Gordon, “Preliminary Remarks,” 73. 70. Gordon, Chapter 2 in this volume.

Bibliography Abbott, Nabia. “Pre-​Islamic Queens.” American Journal of Semitic Languages and Literatures 58, no. 1 ( January 1941): 1–​22. —​—​—​. Two Queens of Baghdad. London: Al Saqi Books, 1986. Bennison, Amira K. “The Peoples of the North in the Eyes of the Muslims of Umayyad al-​Andalus (711–​1131).” Journal of Global History 2, no.2 (2007): 157–​174. al-​Bukhari, The Translation of the Meanings of Sahih Al-​Bukhari, Arabic-​English Vol. III, translated by Mohammad Muhsin Khan. Lahore, Pakistan: Kazi Publications, 1983. Caswell, Fuad Matthew. The Slave Girls of Baghdad:  The Qiyan in the Early Abbasid Era. New York: I.B. Tauris, 2011. Cortese, Delia, and Simonetta Calderini. Women and the Fatimids in the World of Islam. Edinburgh: Edinburgh University Press, 2006. Duindam, Jeroen. “Royal Courts in Dynastic States and Empires.” In Royal Courts in Dynastic States and Empires, ed. Jeroen Duindam, Tulay Artan, and Metin Kunt, 1–​23. Leiden: Brill, 2011.

Epilogue

j 339

El-​Cheikh, Nadia Maria. “Gender and Politics in the Harem of al-​Muqtadir.” In Gender in the Early Medieval World, ed. Leslie Brubaker and Julia M. H. Smith, 147–​161. Cambridge: Cambridge University Press, 2004. —​—​—. “The Qahramâna in the Abbasid Court, Position and Function.” Studia Islamica 97 (2003): 41–​55. Fay, Mary Ann. Middle East Studies Beyond Dominant Paradigms: Unveiling the Harem: Elite Women and the Paradox of Seclusion in Eighteenth-​Century Cairo. Syracuse, NY:  Syracuse University Press, 2012. Gordon, Matthew S. “‘Arib al-​Ma’muniya:  A  Third/​Ninth Century ‘Abbasid Courtesan.” In Views from the Edge: Essays in Honor of Richard W. Bulliet, ed. Neguin Yavari, Lawrence G. Potter, and Jean-​Marc Ran Oppenheim, 86–​100. New York: Columbia University Press, 2004. —​—​—​. “Preliminary Remarks on Slaves and Slave Labor in the Third/​Ninth Century ‘Abbasid Empire.” Slaves and Households in the Near East, ed. Laura Culbertson, 71–​84. Chicago: The Oriental Institute of the University of Chicago, 2011. Lewis, Bernard. Islam from the Prophet Muhammad to the Capture of Constantinople, Vol. II: Religion and Society. New York: Harper & Row, Publishers, 1974. McCormick, Michael. Origins of the European Economy:  Communications and Commerce AD 300–​900. New York: Cambridge University Press, 2001. Peirce, Leslie P. The Imperial Harem: Women and Sovereignty in the Ottoman Empire. New York: Oxford University Press, 1993. Tolmacheva, Marina. “Female Piety and Patronage in the Medieval ‘Hajj’.” In Women in the Medieval Islamic World, ed. Gavin R. G. Hambly, 161–​180. New York: St. Martin’s Press, 1998. —​—​—​. “Ibn Battuta on Women’s Travel in the Dar al-​Islam.” In Women and the Journey: The Female Travel Experience, ed. Bonnie Frederick and Susan H. McLeod, 119–​140. Pullman: Washington State University Press, 1993. Zilfi, Madeline C. Women and Slavery in the Late Ottoman Empire:  The Design of Difference. New York: Cambridge University Press, 2010.

Acknowledgements

i Matthew S. Gordon I extend deepest thanks to my co-​editor, Kathryn Hain, for the energy and enthusiasm that she brought to this project. I am no less grateful to our contributors for their participation; the quality of their scholarship; and their patience as the volume bumped along to publication. And many thanks to Hiba Turman for her work on the index. I completed much of the initial research that informs my part in this project at the National Humanities Center (Durham, North Carolina) in 2011–​2012. I can think of no better place to think and write. My warmest appreciation to the contributors to the Delta Delta Delta fellowship; the administration and library staff of the NHC for the unfailing quality of their work; and to the colleagues with whom I shared that year. I also thank Nancy Toff and her staff at Oxford University Press. Nancy was a strong advocate for the project from the start, and I know I speak for all of us in expressing gratitude for her firm hand, encouragement and patience. I dedicate this volume to Susan. When I am mindful and aware, it is because of her, an unbounded gift. Kathryn A. Hain My thanks to two men, who saw the possibility of turning me into a scholar. First, ten years ago, my husband, Raymond Hain said, “You have to get a PhD.” He then sacrificed to provide the needed infrastructure underlying my diploma and my contributions to 341

342 i

Acknowledgements

this book. Second, Dr. Peter von Sivers had the courage to accept a white-​haired doctoral student of strong convictions to study the Medieval Middle East. Others have proved crucial to the enterprise. Judy Sarriot is always in my corner, cheering me on and taking me to life changing events like the MESA panel in Washington D.C.  where I  met Lisa Nielson. After Dr.  Nielson finished her panel paper on qiyan, I introduced myself and complained that we needed a panel dedicated to slave women in Islam. She encouraged me by offering the first paper. The unexpected arrival of nine abstracts for that first panel proved that a critical mass of scholarship existed for a book. Dwight Reynolds and Matthew Gordon suggested an edited volume. A growing community of concubine scholars continues to provide numerous panels at MESA, AHA, and other major conferences. I dedicate this book to my three children, Ruth, Anna Marie, and Josiah, their spouses, and their children. They have all cheered on my academic aspirations. Their willingness to forgo my offer to micro-​manage their adult lives, instead of writing, should be acknowledged.

Index

i Aban, Abu al-​Walid, 111 Abbada, 63 Abbas, al-​, 235 Abbas ibn Ahnaf, al-​, 54, 289 Abbasids, 4, 6, 7, 12, 13, 27, 27, 28, 28, 29, 31, 32, 33, 34, 36, 38–​43, 46, 47, 53, 59–​61, 75, 78, 79, 84, 85, 87, 91, 130, 233, 235, 236, 245, 245, 246, 250, 52, 257, 257, 266, 268, 275, 278, 279, 280, 287, 289, 324–​328, 330–​332, 334–​336 Abbott, Nabia, 59, 63, 67, 68, 70, 71, 259, 261, 263, 324 abd (pl. abid), 124, 128, 137 Abd Allah al-​Asghar ibn Wahb, 15, 16 Abd Allah ibn Abd al-​Hakam, 41, 302 Abd Allah ibn al-​Zubayr, 309, 310, 312, 315 Abd Allah ibn Ismail al-​Marakibi, 36, 64 Abd al-​Malik ibn Habib, 24, 241, 267, 274, 276 Abd al-​Malik ibn Marwan, 19 Abd al-​Malik ibn Salih al-​Hashimi, 57 Abd al-​Mu’min ibn Ali al-​Kumi, 144–​159, 330 Abd al-​Rahman al-​An‘am, 307 Abd al-​Rahman al-​Dakhil, 129 Abd al-​Rahman I, 104, 119, 128

343

Abd al-​Rahman II, 109–​111, 118, 128, 130, 138 Abd al-​Rahman III, 120, 129 Abdallah ibn Hasan ibn Hasan, 230 Abid al-​makhzan, 155 abkaran, 269, 270 Abrahamic traditions, 226, 233, 327 Abu Abdallah Muhammad, 148, 152 Abu Adiyy Amir ibn Said, 105 Abu al-​Atahiya, 281–​288, 291, 294, 330 Abu al-​Hasan Ali ibn Khalaf ibn Battal, 309, 310 Abu Ayyub al-​Ansari, 253, 262 Abu Bakr al-​Saddiq, 14, 15, 308, 309, 312–​314 Abu Bakr ibn Abi Sabra Abu Bakr ibn Jabr, 151 Abu Dawud, 309, 313 Abu Dulaf al-​Ijli, 35 Abu Hafs al-​Qu`ayni, 130 Abu Hafs Umar al-​Hintati, 121, 147, 148 Abu Hiffan, 289, 294 Abu Imran Musa ibn Sulayman al-​Kafif, 147 Abu Isa ibn al-​Mutawakkil, 248, 250 Abu Ishaq Ibrahim ibn Ali al-​Qayrawani al-​Husri, 282, 285, 286, 293 Abu Jafar al-​Mansur, 57, 135, 135, 225, 226, 228, 230, 233–​238

344 i

Index

Abu Marwan ibn Hayyan, 101, 112, 113, 115, 119–​121, 138 Abu Muhammad Abdallah, 147, 149 Abu Muhammad ibn Dahhun, 113 Abu Nuwas, 54, 57, 67, 288–​291, 294 Abu Said Uthman, 146–​149 Abu Tammam, Habib ibn Aws, 107, 120 Abu Yaqub Yusuf, 147, 148, 150 Abu Zayd Abd al-​Rahman, 149 Abu Zayd Umar ibn Shabba, 54 Adab, 2, 4, 30, 39, 44, 58, 88, 267, 268, 293 Aden, 165 Adjurri, al-​, 77 Adnan, 229, 233 Afghani Aghacha, 192 Afghanistan, 166, 328 Africa, 34, 159, 164–​167, 17–​180, 185, 260 Ahmad Ataka, 191 Ahmad ibn Khalid Haylawah, 290 Ahmad ibn Ubayd Allah ibn Yahya ibn Khaqan, 248, 260 Ahmed II, 209, 210, 212, 216, 218, 220 Ahmed III, 208–​214, 216 Ahmed, Leila, 39, 47, 59, 61, 68, 69 Ahyaf, 112 Aisha, the wife of the Prophet, 256, 262 Aj, 112 Ajab, 136 Ajami, 131 Akhadha, 147, 148 Alam, 110, 130 Albarracín, 113 al-​Bayan al-​mughrib, 150, 156–​158 Alcazar, 134 al-​Dhakhira fi mahasin ahl al-​jazira, 101, 121 Alexandria, 227 Alfonso X, 166 al-​Hulal al-​mawshiyya, 151, 158 Ali Agha (Yapraksiz), 211, 214 Ali Aghyul, 178 Ali al-​Hadi, 247, 250, 253, 254, 258, 263, 324 Ali al-​Rida, 255, 258, 262 Ali ibn Abu Talib, 15, 256, 258, 298–​300, 306, 309–​312, 314–​316, 320 Ali ibn al-​Ahsan al-​Husayni, 133 Ali ibn al-​Husayn (Zayn al-​Abidin), 22, 233, 234, 237, 240, 245, 254, 256, 258, 329 Ali ibn Hisham, 30, 32, 37, 40, 41, 56, 63, 64, 70 Ali ibn Jahm, 56

Ali Shir Navai, 197, 198, 203 Ali, Kecia, 2, 6, 60, 62, 69, 200, 238, 262, 302 Alid, Alids, 23, 232–​235, 237, 240, 241, 246–​248, 259 Allah, 272, 327 Allan (Azerbaijan), 325 Almohad Revolution, 145, 146, 150, 153, 155 Almohads, 143–​146, 148, 150, 151, 153–​157, 159, 330 Almoravids, 113, 129, 144–​146, 148, 150, 153, 154, 156 al-​Musannaf, 307, 312, 318, 319 al-​Mustazraf min akhbar al-​jawari, 67, 68, 101 al-​Targhib wa’l-​tarhib min al-​hadith al-​sharif, 274, 276 al-​Uyun (river), 114 al-​Waraqa, 54 Al-​Yamama, 28, 54 ama, 29, 66 Amal, 112 Amcazade Hüseyin, 210, 211, 219 Amin, al-​, 36, 47, 64, 335 amir (pl. umara), 128, 176, 177, 190 Amr bi’l-​maruf wa’l-​nahy an al-​munkar, 143 Anatolia, 165, 168, 169 Andalus, al-​5, 35, 101–​104, 108, 109, 111, 113, 114, 116–​119, 126–​131, 134, 135, 137–​139, 148, 152, 153, 156, 325, 335 Antara, 12 Apaq Begim, 191, 194, 201 Apologia, 145 Aq Begim, 195, 196, 201 Aqid, 249, 252, 253, 262 Arabia, 12, 17, 24, 32, 54, 59, 75, 77, 78, 90, 91, 95, 97, 98, 166, 169, 226–​229, 236, 282, 288, 298, 301, 309 Arabian Nights, 324 Arabian Peninsula, 54, 309 Arabs, 1–​4, 6, 12, 13, 16, 20, 21, 28, 34, 35, 62, 78, 81, 90, 101, 124–​126, 289, 293, 324, 327–​329, 335 Aragon, 102, 113, 133, 330 Arib al-​Ma’muniya, 28, 29, 31, 34–​36, 38, 41, 43, 44, 46, 47, 55, 63, 64, 67, 70, 85, 86, 90, 92, 94, 252, 261, 263, 293 Armenia, 34, 60, 325 Ashshar, 129 Asia, 1, 32–​34, 166, 168, 170, 171, 180, 190, 191, 199, 200, 293, 332 Askari, Hasan, al-​, 244–​255, 258, 260

Index Astarabad, 195, 197 aswad, 128 Athamina, Khalil, 12, 13, 18, 22–​24, 43, 45, 69, 174, 175, 177, 183–​185, 239, 261, 292, 293 Athl, 128, 129 Atlantic, 1, 30, 44, 124 Aubry de la Motraye, 213, 221 Azdi, Ali ibn Zafir, al-​, 63, 70 Aziz, 104–​107, 119 Azru, 147, 158 Azza al-​Mayla, 105 Badhl, 41, 105, 112 Badi al-​Zaman Mirza, 193, 197, 198 Baghawi, al-​, 306 Baghdad, 28, 30–​34, 37, 39, 40, 52, 54, 64, 66, 69, 77, 102, 109–​111, 130, 164, 169, 253, 254, 282, 326, 331 Bahja, 104, 107, 108 Bahri Mamluks, 163 Baladhuri, al-​, 234, 236, 239–​241 Baltacı Mehmed Pasha, 213, 214, 216 Banu Awf, 152, 153 Banu Hashim, 234, 248 Banu Hilal, 151, 152, 159 Banu Razin, 113 Barbar, 64, 333 Barmakids, 34, 57, 287, 63 Bas-​Bas, 82 Bashear, Suliman, 12, 13, 22 Basque, 111, 126, 136 Basques (Vascones), 111, 126, 136 Basra, 6, 8, 36, 37, 41, 64, 249, 250, 279 Basri, Hasan, al-​, 306 Battle of the Camel, 335 Battle of Setif, 151, 152 Baydhaq, Abu Bakr Sanhaji, 144–​152, 155–​159 Bayhaqi, Abu Bakr, al-​, 302, 303, 306, 307, 310, 310, 313, 314, 317 Bazi, 112 Bedouin, 121, 151, 152, 227 Bejaia, 143, 149, 151, 179 Bengal, 166, 173 Berbers, 60, 126, 128, 129, 131, 132, 134, 135, 144, 146, 148, 156, 157, 159, 177, 255, 325 Bibi Muhibb-​Jangi, 197, 203 Bida, 112 Bika-​Sultan Begim, 193, 201 Biki Sultan, 194 bikr, 45, 148

j 345

Bint Maksan ibn al-​Mu’izz, 148 Bishr ibn Sulayman, 253, 254, 262 Black Death, 166 Black Sea, 165, 169, 171 Book of Songs. See Kitab al-​aghani Bouhdiba, Abdelwahab, 61, 69, 174, 181, 184, 185 Bray, Julia, 7, 38–​40, 43–​47, 62, 63, 65–​67, 69–​71, 120, 281, 293 Brockopp, Jonathan, 5, 6, 7, 12, 22, 23, 41–​43, 47, 67, 240, 259, 301, 316, 317 Brown, Jonathan, 299, 316, 317, 319 Bugha the Elder, 33 Buhturi, al-​, 107 Bukhait, 175, 184 Bukhara, 170 Bukhari, al-​, 267, 307, 308, 314 Bulgars, 127 Bunan, 290 Caesar, 230, 329 Cairo, 65, 117, 165, 179 Calderini, Simonetta, 39, 47, 70, 71 Calicut, Qaliqut, 171, 173 Candia, 217 Cantemir, Dimitrie, 212, 215, 216, 219 Castile, 133, 166 Caswell, Matthew, 43–​46, 67, 69, 91, 92, 130, 138, 276, 294 Caucasus, 20, 32, 34 Celia, 143, 144, 156 Ceylon, Sri Lanka, 166 Chaharshamba, 191, 194 Charlemagne, 129 Charles XII, 213 China, 44, 81, 82, 91, 92, 166, 171, 173, 177, 182, 293, 325 Chinggis Khan, 332 Chios, 217 Christians/​Christianity, 24, 40, 109, 113, 125–​129, 133–​136, 139, 145, 152, 153, 155, 169, 254, 278, 279, 326 Cilardo, Agostino, 167, 182 Circassians, 192, 213 Coello, Pilar, 134, 139 Companions, 307, 309–​314, 319, 320, 24, 253, 299, 302, 306 Constantine, 153 Constantinople, 13, 165, 170, 183 Conte, Edouard, 236, 241 Copts, 23, 225, 227, 228, 234, 235, 241, 255, 327

346 i

Index

Cordoba, 101, 102, 109, 111–​114, 121, 326 Cordova, 128, 136 Corfield, Penelope, 1 Çorlulu Ali Pasha, 213 Coromandel coast, 172, 175 Cortese, Delia, 39, 47, 70, 71 Crete, 208, 213, 313 Crete War, 208 Crimea(n Khan), 165, 183, 211, 214, 215, 333 Dai, 146, 147, 149, 150 Damascus, 165, 179, 317 Dananir, 82, 335 Dar (al-​) islam, 153, 163, 180, 183 Daraqutni, al-​, 307, 319 Dawlasa, 177 Delhi, 163, 166, 171, 179, 184 Deli Hüseyin Pasha, 208 Dhamm al-​malahi, 88 Dhat al-​Khol, 82, 86 Dhayl, 112 Dhibat al-​Mahal, 176 Dhikr bilad al-​Andalus, 129, 137–​139 Dhimma, dhimmis, 155, 163 Dilaşub Sultan, 217 Dinar, 57, 84, 107, 108, 114, 115, 117, 120, 168, 184, 253, 254, 256, 263, 284 Dirham (derhim), 57, 83, 86, 106, 151, 168, 284, 290, 308 Dubaa bint Zubayr ibn Abd al-​Muttalib, 15 Dukkala, 151 Duqaq, 64 East Africa, 34, 165, 325 Edirne, 212, 213, 216, 217 Egypt, 1, 2, 5, 35, 38, 39, 42, 56, 57, 68, 75, 90, 95, 130, 163–​165, 167, 169, 183, 185, 211, 228, 238, 266, 278, 317, 327, 331 El Cheikh, Nadia, 5, 7, 39, 40, 44, 46, 47, 63, 67, 261, 263 Ennaji, Mohammed, 292, 294 Ethiopia, 24, 60, 134, 159, 168, 328 Euben, Roxanne, 165, 182 Eurasia, 164, 168, 183, 325 Europe, 1, 33, 34, 45, 69, 84, 92, 93, 205, 208, 269 Fabricius, Baron, 213, 220, 221 Fadl, al-​Barmaki, al-​, 57

Fadl ibn Rabi, al-​, 57 Fadl, the Medinese, 28, 29, 31, 35, 36, 42, 68, 82, 110–​112, 130, 336 Farazdaq, al-​, 108 Farba Sulayman, 179 Fasl al-​khitab fi madarik al-​hawass al-​khams li-​uli al-​albab, 101 Fatik, 104 Fatima bint Muhammad, 231, 234, 235, 237, 238, 254, 256 Fatima of Fez, daughter of Yusuf, 148, 149 Fatin, 104 Fawz, 112 Fazil Ahmed Pasha, 210 Fergana, 191 Feyzullah Efendi, 210–​213, 219, 220 Fihr, 16, 18, 23 Fihrist (Ibn al-​Nadim), 22, 53, 67, 77, 92 Fiqh, 131, 164, 280 France, French, 127, 216, 240 Frankish, 129, 132 Frenkel, Yehoshua, 179, 181, 182, 185 Gabriel, 227, 314 Gaiga, 145, 157 Galata, 217 Galicia, 126, 127, 131, 132 Garden (janna), 267, 272–​274, 275 geisha, 43, 81, 83, 92, 100 Geniza, 65 Genoese, 170, 183 Ghalib, 112 Ghazali, Abu Hamid, al-​, 180, 185, 268, 276, 306, 311 Ghiyas al-​Din Khvandamir, 191, 195, 197, 198, 201–​204 Ghizlan, 112 Ghulam, 112 ghulam (pl. ghilman), 112, 151, 268 Gibraltar, 166 God, 54, 57, 88, 105, 108, 114, 115, 129, 131, 132, 143, 153, 181, 185, 225, 227, 231–​235, 250, 252, 254, 269, 283–​285, 287, 293, 297, 303, 304, 306–​308, 311, 313, 319 Goldziher, Ignaç, 12, 13, 17, 20–​22 Gordon, Matthew, 62, 63, 67, 69, 70, 71, 90–​94, 260, 261, 263, 276, 293 Granada, 102, 104, 132, 133, 148, 166 Great Northern War, 208, 214

Index Greece, Byzantine, 1, 7, 24, 40, 43, 44, 49, 78, 81, 82, 97, 165, 168–​172, 183, 208, 227, 254, 257, 329, 330, 332 Gul-​i-​stan, 175 Gulistan, 199 Guo, Li, 165, 182, 262 Habbaba, 105 Habesci, Elias, 212, 220 Hadith, 4, 5, 7, 70, 71, 90, 126, 136, 262, 266, 268, 269, 280, 297–​299, 302, 303, 305–​309, 311–​314, 316–​321, 326 Hajar (Hagar), 225–​228, 230–​241, 327 Hajj, 37, 164, 165, 169, 227, 320, 324, 334 Hakam I, al-​, 104–​111, 118, 120, 129 Hakam II, al-​, 129, 136 Hakima, 252, 254, 262 Halhal, 112 Hallaq, Wael B., 154, 159, 316–​322 Hamduna, 64 Hammadids, 143, 144, 149, 151, 156 Hammawayh, 86 Hanafi, 167 harem, 6, 13, 59, 60, 67, 70, 129, 166, 173, 175, 176, 207, 209–​215, 221, 325, 331–​336 Harra, 15 Harun al-​Rashid, 36, 52, 54, 57, 59, 60, 63–​65, 68, 69, 79, 86, 110, 261, 281, 282, 286, 287–​290, 324, 334 Hasan ibn Abd Allah, 307 Hasan ibn Abi Shawarib, al-​, 249, 250, 260 Hasan ibn Ali, al-​, 232, 234, 246, 258 Hasan-​Shaykh Temur Jalayir, 191, 195, 201 haseki, 208–​211, 213, 216–​220, 333 Hashim ibn Abd al-​Aziz, 112, 128, 129 Hashimite, 29, 37 Hatice Turhan, 207, 209, 217, 218, 333 Hatim (the Generous), 107 Haydar-​Muhammad Mirza, 193, 204 Hayn, 105, 106 Hazm, 129 Herat, 191, 195, 197, 198, 199, 201–​204, 206 hetaera, 81, 82 Heywood, Colin, 169, 183 Hidayatullah, Aisha, 238, 239, 319 High Atlas, 145, 146, 151, 157, 256 Hijaz, 17, 93, 251, 259 Hind, 60, 134

j 347

Hisham I, 104, 119, 114 Hisham II, 135, 136 Hisham ibn Abd al-​Malik ibn Marwan, 13, 231 Holstein Palace, 213 houri (hur'in), 147, 267, 268, 275, 326, 327, 336 Hudayth, 244, 248, 249, 251–​253, 257, 258, 261 Hudhayl ibn Razin, 113–​115, 330 Humam, 112 Humaydi, al-​, 101, 114, 117, 121 Hunayda, 112 Hungary, 211 Hunwick, John, 166 hurra (pl. hara’ir), 53, 60, 147 Husayn ibn Ali, al-​, 232, 234, 246, 255, 258 Husayn ibn Hamdan al-​Khasibi, 250–​252, 261, 262 husn (al-​) taba‘ul, 65, 71 Iberia, 45, 101, 125, 126, 128, 303, 330. See also Spain Ibn Abbas, 298–​300, 303, 306, 307, 317 Ibn Abd al-​Barr, 306 Ibn Abd al-​Hakam, 239, 301, 302 Ibn Abi al-​Dunya, 77, 88, 90, 94, 95 Ibn Abi Shayba, 307, 309, 312, 318, 319 Ibn al-​Ashtar, 230 Ibn al-​Athir, 144, 145, 148, 150–​153, 156, 158, 159 Ibn al-​Attar, 131, 132, 133 Ibn al-​Kalbi, 21 Ibn al-​Kattani, 102, 114, 115, 121 Ibn al-​Mutazz, 29, 68 Ibn al-​Nadim, 22, 53, 77 Ibn al-​Qattan, 149, 158 Ibn al-​Sa`i, 33, 45, 67, 68, 101, 294 Ibn al-​Salim, 129 Ibn al-​Washsha, 62, 70, 77, 81, 82, 88, 90–​93, 101 Ibn Asakir, 236, 239, 240 Ibn Babawayh, 248–​250, 252, 253, 255, 256, 260–​262 Ibn Bajja, 113 Ibn Bashkuwal, 101 Ibn Bassam, 101, 115, 113, 121, 138 Ibn Battuta, Abu Abdallah Muhammad, 163–​185, 334 Ibn Butlan, 40, 47, 134, 139, 274, 325, 328 Ibn Hanbal, 259, 306, 311, 312, 319–​321 Ibn Hawqal, 127, 137 Ibn Jarrah, 54, 55, 57, 67, 68

348 i

Index

Ibn Kathir, 297–​300, 302–​321 Ibn Khaldun, 23, 146, 147, 157 Ibn Maja, 306, 307, 309, 321 Ibn Mughith, 133, 138 Ibn Qadi al-​Shuhba, 305, 318 Ibn Qalqal, 112 Ibn Qutayba, 41, 240 Ibn Sa`d, 21, 67, 236, 239 Ibn Sahib al-​Sala, 144, 149, 150, 155–​158 Ibn Sahl, 135, 139 Ibn Sayyid al-​Nas, 305, 313, 318 Ibn Tashfin, Ali ibn Yusuf, 120, 145, 150 Ibn Tulun, Ahmad, 5 Ibn Umar, 300, 307, 308, 312, 314 Ibn Wafid of Toledo, 135 Ibrahim, 225–​228, 234, 235, 307, 327 Ibrahim (Abraham), 225–​227, 231, 233, 234, 237, 327 Ibrahim al-​Mawsili, 31, 33, 37, 38, 41, 44, 83, 86, 102, 103, 109, 112 Ibrahim I, 209, 212, 213, 218 Ibrahim ibn al-​Mahdi, 29, 37, 38 Ibrahim ibn Hajjaj, 130, 138 Ihya ulum al-​din, 268, 276 Ijithad, 302 Ijma, 301, 320 Ikhtiyar al-​Din Fort, 197, 198 Ima al-​shawa’ir, al-, 30, 53, 67, 69, 101, 288 Imad al-​Din al-​Simnani, 177 Imamate, 255, 231, 232, 239, 257, 260 Imami Shi’is/​Shi’ism, 244–​246, 249, 259 Imams (Twelver Shi’i), 4, 6, 329, 330 Inan al-​Natifi, 28, 31, 35, 43, 45, 52–​58, 64, 66–​68, 278, 280, 281, 288–​292 India, 34, 44, 78, 84, 92, 93, 163, 166–​168, 171, 173, 175–​177, 180, 184, 191, 192, 231, 262, 325 Iran, Persia, Persians, 1, 2, 60, 61, 69, 78, 165, 165, 168, 175, 180, 183, 190, 190, 191, 194, 195, 197, 199, 200, 201, 202, 230, 233, 240, 255, 328, 329, 332 Iraq, 5, 33, 36, 109, 116, 127, 130, 134, 144, 165, 180, 260, 261, 321, 336 Isa ibn Ali, 175 Isbahani, Abu al-​Faraj, al-​, 6, 27, 28–​33, 35, 43–​47, 53–​57, 67, 83, 85, 88, 90–​94, 101, 102, 119, 120, 282, 284, 288, 289, 290, 293, 294 Isbahani, Abu Bakr Muhammad ibn Dawud, al-​, 282, 293 Ishaq, 231, 237

Ishaq al-​Mawsili, 31, 37, 41 Islam, 1–​7, 9, 11–​14, 19–​24, 27, 42, 43, 46, 47, 60, 63, 76, 78, 85, 89, 90, 93, 126, 148, 157, 163, 164, 166, 169, 180, 190, 191, 193, 200, 201, 217, 225, 226, 228, 229, 230, 232, 236–​241, 261, 266–​269, 276, 279–​281, 286, 292, 297–​299, 306, 315–​318, 324–​336 Ismail (Ishmael), 225–​227, 231–​235, 237 isnad, 299, 318 Istanbul, 212, 216, 217 istibra, 135 istishab, 314 Italy, 91, 113, 212, 213 Itr, 112 Iwalatan, 178 Izurba (Souss), 147, 152 Jabir ibn Abd Allah, 299, 300, 313, 315 Jafar al-​Mutawakkil, 32, 33, 35, 36, 45, 56, 68, 94, 247, 248, 336 Jafar al-​Sadiq, 247, 255, 256, 258 Jafar ibn Ali, 247–​251, 253, 258, 261 Jafar ibn Yahya al-​Barmaki, 34, 56, 57, 63, 287 Jahangir Mirza, 199, 203 Jahiz, Amr ibn Bahr, al-​, 30, 31, 35, 37, 38, 42, 44, 46, 62, 77, 81–​83, 87, 88, 90–​94, 101, 103, 119, 268, 270–​273, 276, 331, 335 Jalal al-​Din Davani, 192, 193, 201 Jamharat al-​nasab, 21 Jamila, 82, 83 Jami‘ li-​ahkam al-​Qur’an, 267, 276 janna. See Garden Japan, 43, 81, 82, 83, 84, 92, 93, 100 Japheth, 127 Jaradatan, 78 jariya (pl. jawari), 23, 28, 30, 32, 38, 52, 54–​70, 76, 79–​81, 84, 91, 92, 119, 124, 177, 180, 183, 266, 326, 334 Jerusalem, 168 Jesus, 254 Jews/Judaism, 78, 128, 135, 155, 216, 221, 233, 240, 278, 279 jihad, 5, 71, 126, 153–​155, 166, 169, 183 jinn, 54, 55, 121, 268–​270, 275 Jullanar, 64 Kaaba, 23, 226, 227 Kalaylikoz Ahmed Pasha, 213, 218

Index Kalb, 229, 240 Kamil fi’l-​tarikh, 144, 145, 152, 156, 158, 159 Kara Ibrahim Pasha, 210, 333 Kara Mustafa Pasha, 210, 219 Karima bint al-​Miqdad ibn Amr al-​Bahrani, 15 Karl XII, (the King of Sweden), 214–​216, 333 Karlowitz Treaty, 212 Kethüda, 211, 220 khabar (pl. akhbar), 31, 36, 100 Khadija, 227, 327 Khadija Begi Agha, 191, 195–​198, 202, 203, 332, 333 khadim (pl. khuddam), 175, 260 Khaizuran, 252, 261 Khajal, 112 Khalifa ibn Khayyat, 236, 240 Khaqan, 230, 329 Khayzuran, 63, 324, 326, 332 Khilata, 112 Khumarawayh ibn Ahmad ibn Tulun, 38 Khurasan, 57, 58, 63, 68, 128, 190 Khusraw, 230, 329 Khwaja Muhammad Ataka, 194 Kilpatrick, Hilary, 7, 30, 43–​46, 67, 68, 90–​92, 94, 294 Kitab al-​aghani (The Book of Songs ), 6, 7, 8, 27–​29, 31, 33, 36, 38, 41, 43–​47, 53, 55, 64, 67–​70, 76, 83, 85, 90, 91, 93, 94, 101–​103, 119, 120, 285, 293 Kitab al-​buldan, 32 Kitab al-​muqtabis, 101 Kitab al-​muwashsha, 62, 77, 101, 274 Kitab al-​tawahhum, 267, 268, 276 Kitab al-​Zahra, 282, 284, 285, 293 Kitab fi qiyan al-​Andalus, 101 Köprülü Mehmed Pasha, 207, 208, 210, 218 Kösem Sultan, 209, 217, 218 Kufa, 64, 241 Kulayni, al-​, 71, 248, 250, 254, 255, 260–​263 Kurds, 60 Lamta, 179 Late Antiquity, 2 Latifa-​Sultan Aghacha, 191, 193, 194, 201, 203, 203 Latin, 136, 183 Louis XIV, 215 Lovejoy, Paul E., 164, 181 Lubana, 41

j 349

Lubna, 105 Lutfi, Huda, 168, 181–​183 Ma`bar, 172, 175 madhhab (pl. madhahib), 178, 301 Maghrib, 116, 117, 129, 143, 146, 149–​154, 156, 175 Mahbuba, 32, 33, 35, 36, 38–​40, 56, 70, 82, 336 Mahdi, Abu Abdallah Muhammad ibn al-​ Mansur, al-​, 281–​287, 291, 330 Mahdi ibn Tumart, 155 Mahdi. See Twelfth Imam majlis (pl. majalis), 36, 38, 64, 86, 92, 94, 282 Malabar, 166, 175 Malaga, 134, 166 Maldives, 166, 167, 174–​177 Maliki school, 131, 133, 134, 136, 164, 171, 178, 301, 306, 308 mamluk (pl. mamalik), mamluka, 29, 134, 192 Mamluks, 163, 164, 166, 168, 169, 177, 179, 181, 185, 298, 299, 302, 305, 306, 315, 331 Mann bi-​l’Imama, 144 maqam (pl. maqamat), 87 Maqqari, Ahmad al-​101, 117, 120, 121, 130, 138 Marathi, 175, 184 Marco Polo, 165, 334 Margalita, 168 Marín, Manuela, 43, 67, 135, 137–​139, 157, 159 Mariya the Copt, 23, 225–​228, 230, 234–​239, 241, 252, 255, 307, 308, 319, 321, 327 Marmon, Shaun, 302, 317, 137, 181, 240, 257, 260, 262 Marrakesh, 144, 145, 150–​153, 155 Marriage and Slavery in Early Islam, 2, 60, 62, 204, 262, 276 Marwan I, 19, 109 Marwan II ibn Muhammad, 20, 24, 230, 239, 329 Marwan ibn Abi Hafsa, 54, 55, 289 Marwanid, 109 Masabih, 112 Masalik al-​absar fi mamalik al-​amsar, 101, 104, 116, 118–​121 Mashriq, 101, 117 Maslama ibn Abd al-​Malik, 12, 13, 228, 233, 237 Masmuda, 146–​152 Masrur, al-​, 57, 283, 290 Mas`ud al-​Shamhani, 132 Masudi, al-​, 32, 45, 63, 70, 77, 252, 261 Mattson, Ingrid, 5, 7, 43, 47, 300, 301, 316

350 i mawla (pl. mawali), 21, 78, 245, 248, 259, 262, 285, 294 McCormick, Michael, 7, 33, 45 McLaurin, Melton A., 143, 144, 156 Mecca, 24, 134, 164, 165, 167, 168, 169, 183, 209, 216, 217, 227, 228, 274, 279, 312, 324, 333 Medina, 6, 32, 37, 102, 110, 111, 112, 114, 130, 134, 216, 217, 227, 228, 279, 297, 325 Mediterranean, 4, 6, 29, 44, 101, 114, 124, 170, 180 Mehmed IV, 207–​211, 214, 218, 219, 333 Mehmed Efendi, 211, 212, 215 Mehter Osman Agha, 210 Memoirs of al-​Baydhaq, 144–​146, 148 Mengli Bey Aghacha, 192, 193 Merida, 135 Mernissi, Fatima, 60, 61, 69 Miqdad ibn Abd Allah, al-​, 15 Miranda, Huici, 149, 156–​158 Mirza Yadgar-​Muhammad, 195 Modarressi, Hossein, 251, 259–​262 Molina, Luis, 129, 137, 138 Mongols, 163, 164, 165, 169, 170, 180, 194 Morali Hasan Pasha, 213 Moriscos, 125, 137 Morocco, 166, 176, 177, 334 Mottahedeh, Roy, 292, 294 Muallila, 112 Muawiya ibn Hisham, 19, 111, 229 Muawiya ibn Hisham al-​Shabinisi, 111 Mubarrad, al-​, 54 Mughals, 1, 91, 191, 151, 153–​159, 334 mughaniyya (pl. mughaniyyat), 76, 79, 80, 81, 84, 91, 92 Mughira, al-​, 108, 109, 118 Mughith al-​Din Muḥammad, 177 Muhammad al-​Baqir, 234, 255, 258 Muhammad al-​Jawad, 255, 258, 262 Muhammad al-​Mahdi. See Twelfth Imam Muhammad Hayder Dughlat, 198, 199, 204 Muhammad ibn Abd al-​Mu’min, 151 Muhammad ibn Abdallah ibn Tahir, 32, 33, 45 Muhammad ibn al-​Hanafiyya, 236, 241 Muhammad ibn Marwan, 230, 237 Muhammad ibn Tughluq, 171 Muhammad ibn Tumart, 143–​147, 150, 151, 153–​157, 159 Muhammad ibn Zayd, 255 Muhammad, Imam, 129, 135

Index Muhammad-​Muʾmin Mirza, 197, 198, 204 Muhammad Nafs al-​Zakiyya, 225, 234, 235 Muhammad, the Prophet, 4, 13–​17, 23, 120, 132, 145, 184, 225, 227, 231, 232–​235, 237, 241, 256, 266, 276, 280, 297–​299, 305–​308, 311, 313, 314, 326–​329, 335, 336 Muhammad Sariq, 196, 202 Muhammad Shibani, 199 Muhasibi, al-​, 267, 268 Muhja, 104, 107, 108 Mu‘izz al-​ansab fi shajarat al-​ansab, 191, 193–​196 mukhannath (pl. mukhannathun), 78, 91 Mukhariq, 112 Mukhtala, 112 Multan, 177 Mundhir, al-​, 128, 129, 138 Mundhiri, 274 Muqawqis, al-​, 227 Muqtadir, al-​, 40, 63, 70 Murad, 30, 32, 37, 40, 56, 68 Murjan, 129 Murra, 14 mursal, 307, 319 Musa al-​Kazim, 255, 258 Musab ibn al-​Zubayr, 230 Musannaf, 312, 320 Muslim ibn al-​Hajjaj, 308, 313, 321, 318, 262 Mustafa II, 208–​214, 220, 333 mut‘a, 300, 301, 316 Mutamid, al-​, 116, 120, 248–​250, 260 Mutasim, al-​, 32–​34, 37, 58 Mutawakkil, al-​, 28, 32, 33, 35, 36, 45, 56, 68, 94, 247, 248 Mutayyam, 7, 41, 63, 64, 112 Muwaffaq, al-​, 250 muwallada (pl. muwalladat), 28, 33–​36, 41, 54, 67, 117, 128 Muwashshah, 102, 118 Muwatta, 308 Muzaffar-​Husayn Mirza, 193, 195–​199, 202, 203 Muzayna, 129 Muzna, 129, 138 Myrtale, 82, 92 Nafh al-​tib min ghusn al-​Andalus al-​ratib, 101, 138 Nafza tribe, 128, 137 Najd, 54 Nakhkhas, 40

Index Narjis, 252, 254–​258, 262 Nasa’i, al-​, 313, 321 Nasab, 23, 109, 241 Nasab Quraysh, 11, 14–​18, 21–​24, 329 Nashr, 112 Nasim, 56 Nasim, the slave who saw al-​Mahdi, 252, 261 Nass, 246, 259 Na`t, 5 Natifi, al-​, 54–​57, 67, 288, 289 Nawawi, al-​, 302, 303, 317 Near East, 2, 3, 6, 27, 28, 28, 29, 30, 33, 33, 34, 34, 34, 36, 36, 38, 39, 40, 42, 56, 75, 90, 95, 130, 164, 167, 185 New Testament, 326, 336 Nielson, Lisa, 63, 66, 70, 90, 95, 276 Nikah, 196, 202 Nimr ibn Uthman, al-​, 105 Nisa al-​Khulafa, 33, 51, 67, 101 Non-​Arab, 11, 12, 20, 22, 24, 47, 61, 78, 117, 230, 234–​236, 241, 245, 255, 259, 288, 292 Non-​Muslim/​Islamic, 11, 13, 21, 153, 154, 169, 183, 238, 239, 332 Non-​Qurayshi, 18, 19, 21, 24, 328 North Africa, 1, 33, 102, 104, 118, 122, 128, 131, 135, 137, 138, 144, 153, 154, 156, 168, 282, 325, 330 Nubayka, 38 Nubdhat al-​Asr fi akhbar muluk Bani Nasr, 104 Nubian, 255 Nur, 112 Ottoman/Ottomans, 1, 5, 39, 43, 45, 45, 164, 169, 177, 207, 207, 207, 207, 208, 208, 211, 214–​218, 220, 261, 262, 317, 331–​333 Oud, 6, 36, 82, 85, 87 Oum Rbia River Valley, 149 Ozbek, 165 Palestine, 165 Palmyra, 335 Papa Aghacha, 191, 193, 194, 201, 203, 204 Payanda-​Sultan Begim, 193, 201, 204 Pérès, Henri, 114, 119–​121 Peter, 254, 330 Peter I, 214 Pharaoh, 226 Pierce, Leslie, 5, 6, 7, 219, 259, 262, 249 Polish War, 209, 210

j 351

Poniatowski, General, 216, 22 Powell, Eve Troutt, 166, 182, 185 Prut Treaty, 214, 215 Qabiha, 56 Qahtan, 229 Qalam al-​Bashkunsiyya, 111, 130, 136 Qamar, 130, 136 qanun, 87 Qara Koz Begim, 191 Qaramita, 331 qasida (pl. qasa’id), 91, 283 qayna (pl. qiyan), 43, 46, 52, 62–​64, 66, 68–​70, 75–​81, 84, 85, 88, 89, 90–​92, 100–​106, 108–​120, 121, 134, 245, 252, 257, 266–​268, 270–​275, 326, 327, 330, 331, 335, 336 Qivam al-​Din Nizam al-​Mulk Khvafi, 197, 198, 203 Quma (pl. Qumayan), 194–​196 Qumm, 248, 249 Quran, 2, 4, 5, 11, 63, 67, 77, 113, 114, 115, 136, 143, 149, 180, 184, 228, 229, 231, 235, 266–​269, 271, 275, 280, 297, 300, 302–​304, 306, 313, 318, 320, 326, 328, 330, 334 Quraysh, 11, 14–​16, 18–​20, 23, 24, 29, 64, 112, 229, 327, 329, 336 Qurtubi, al-​, 267, 276 Qusayy ibn Kilab, 14, 16, 23 Rabia Gülnuş Emetullah, 207–​221, 333 Ragib, Yusuf, 33, 35, 44–​46, 68 Rah, 128, 137 Raha, 112 Rapoport, Yossef, 167, 168, 182, 183, 276, 302, 313 Rashidun caliphs, 297, 298 Rayya, 112 Razi, Isa ibn Ahmad, al-​, 110, 319 Rethymno, 208 ribat, 126 Richardson, Kristina, 3, 7, 43, 45–​47, 63, 70, 276 Ridwan, 112 Rightly Guided caliphs, 228, 309, 310 Rihla (Ibn Battuta), 163, 165, 171, 181–​185 Risala fi al-​Ishq wa-​l-​Nisa, 62 Risala fi shira’ al-​raqiq wa-​taqlib al-​abid, 134, 274 Risalat al-​qiyan, 30, 62, 76, 81, 87, 90, 268, 270, 276 Rome, 1, 46, 60, 91, 126, 278, 335

352 i Ruggles, D. Fairchild, 35, 45, 241 Ruhban, 112 Rumford, James, 167, 182 Russia, 163, 200, 208, 214, 215 Russian–​Swedish war, 214 Ruzzini, Carlo, 212, 213 Rycaut, Paul, 211, 212, 218–​220 Saba, 112 Safavids, 1, 192 Saffah, al-​, 57 Saffar, al-​, 250, 254, 261, 262 Safiyya bint Abi Imran, 147, 149, 157 Sahla, 113 Sahrawi, al-​, 151 Said ibn Musayyab, 307, 311, 319 Sakan, 58, 68, 128 Salam, 131, 138 Sallama, 234 Sallamah al-​Qass, 82 Sallamat al-​Qass, 105 sama, 88, 91, 94 Samarkand, Samarqand, 170, 183, 191, 199, 204 Samarra, 28, 32, 34, 39, 44, 45, 50, 77, 247–​251, 260, 261 Sanani, al-​, 298 Sanhaja, 146, 148, 149 Saqati, al-​, 134, 140 saqi, 268 Saqil, 249–​253, 257, 258, 262 Saqlabi (pl. Saqaliba), 127, 134 Sara(h), 226, 227, 327 Saracens, 135 Sarakhsi, al-​, 167 Sasanids, 230, 254, 262, 278, 329 Scent of Ambergris, 175 Schacht, Joseph, 12, 22, 47, 240, 241, 259, 298–​301, 308, 316, 317 Seraglio, 209, 216 Seville, 102, 116, 117, 130, 139, 148 Şeyhülislȃm, 211, 212, 213, 216 Shafi’i, al-​, 302–​306, 308–​313 Shaghab, 40, 63, 252, 261, 331 shaglub Shah-​Gharib Mirza, 196, 197, 202, 203 Shahik, 63 Shahrbanu, 245, 256, 258, 259, 262, 329 Shahr-​Banu Begim, 192, 201, 202 Shah Tahmasp I, 192 Shahzade İbrahim, 212, 213

Index Shakir, Ahmad, 304, 318 Sha’n, 112 Sharaf, 112 Shararah, Abd al-​Latif, 61 Sharh al-​Muhadhdhab, 303 Sharh al-​tanbih, 304, 317, 318 Shari’a, 202, 304, 328, 330 Shariya, 28, 29, 31, 37, 38, 44, 64 Shaybani, al-​, 114, 115 Shaykh Sadi, 199 Shi‘is, Shi‘ites, 70, 244, 260, 262, 298–​301, 315, 328, 329 Shuja al-​Din Muhammad Burunduq Barlas, 198 Shunayf, 112 Silahdar Ali Agha, 213, 214 Sirin the Copt, 227 sitara, 110, 114 Slav, 209, 328, 330 Souss, 143, 145, 149 Spain, 1, 5, 35, 113, 119, 120, 125, 156. See also Iberia Spellberg, Denise, 3, 7 Standen, Naomi, 169, 183 Suada, 116 Subh, 112, 136 sub-​Saharan Africa, 135, 159, 166, 167, 330 Sudan, 60 Sulaym, 109, 120 Süleyman II, 209, 211, 217, 218 Süleyman Pasha, 210 Sultan Abu l-​Hasan ‘Ali, 104 Sultan-​Abu Saʿid Mirza, 191–​193, 195, 196, 201–​203 Sultan-​Ali Mirza, 199 Sultan-​Husayn Bayqara, 191–​198, 200–​203, 332 Sultan-​Khalil, 192 Sultan-​Mahmud Mirza, 196, 199, 202 Sultan Muhammad, 177, 183 Sumatra, 166, 173, 179, 334 Sunna, 300, 302, 306, 308, 309, 313, 320 Sunni, 251, 261, 262, 298, 300, 301, 306, 612, 315, 316, 320, 328 surriya (pl. sarraiy), 91, 194 Sutton, Sir Robert, 211, 219 Suyuti, al-​, 67, 68, 101, 294 Sweden, 208, 213–​216 Syria, 128, 163–​167, 169, 179, 229 Szombathy, Zoltan, 23, 235, 241 Ta’if (Taif ), 6, 32, 36, 126, 127 Tabaqat (Ibn Sa’d), 21, 67, 239

Index Tabari, al-​, 77, 93, 234, 239, 240, 241, 260, 262, 263 Tabaristan, 255 Tadla, 149, 151 Tafsir, 280, 304, 317, 318 Tahir ibn al-​Husayn, 57 Tahr, 129 Taifa, 126, 127 Taj al-​Din Hasan ibn Amir Nizam al-​Din Charkas, 191, 194 Takattum, 254, 258 Takedda, Tagadda, 177–​180 Talal, 112 Talibiyya, al-​, 112, 121 Tamagunt, 145, 146, 148, 155 Tamerlane, 190, 192, 200, 201, 332 Tangier, 184 Tanukhi, al-​, 7, 36, 39, 46, 77, 84, 93 Taqbugha, 170 taqlid, 308, 317 tarab, 95, 111 Tarub, 112 Temur, 190, 191 Thaghr, Thughur, 169, 183 Tifashi, Ahmad, 101, 102, 104, 113, 116–​118 Timurid, 190–​196, 198–​205, 332, 333 Tinmal, 145, 147, 157 Tirmidhi, al-​, 268, 276, 305, 309, 319 Tlemcen, 148, 150 Toledo, 133, 135 Transoxiana, 34, 190 Transylvania, 211 Tucker, Judith, 1, 6, 317, 321 Tuhfat al-​nuzzar fi ghara’ib al-​amsar waaja’ib al-​asfar, 164 Tunis, 101, 117, 143, 152, 156, 168 Turco-​Mongol, 217, 194, 201 Turkestan, 167 Turkey, 2, 33, 38, 44, 45, 71, 134, 169, 183, 218, 328, 332 Tuways, 83 Tuzari, al-​, 168 Twelfth Imam, the, 144, 145, 152, 153, 155, 159, 246, 248, 252–​254, 258–​260 Ubayd Allah, 248–​250 Ulama, ulema, 130, 183, 185, 208–​210, 212 Ulayya bint al-​Mahdi, 54, 79, 91, 264, 290, 335 Umar ibn Abd al-​Aziz, 306

j 353

Umar ibn Abd al-​Rahman, Abu al-​Qasim, 110 Umar ibn al-​Khattab, 15, 255, 297–​301, 303, 306–​315, 318, 321, 328 Umari, ibn Fadl Allah, al-​, 101, 104, 108, 116, 118–​121 Umar ibn Sulayman Hafyan, 303–​305, 317, 318 Umar ibn Yazid, 253 Umar-​Shaykh Mirza, 191 Umayyad/Umayyads, 12, 13, 18–​21, 24, 52, 59, 60, 63, 75, 77, 78, 102, 104, 105, 109, 113, 119, 121, 128–​130, 137, 138, 225, 228–​233, 235, 236, 239–​ 241, 245, 259, 262, 280, 286, 293, 326–​329, 335 Umm al-​Fath bint Jafar, 101 Umm al-​Muqtadir, 40, 63, 261, 331 umm (al-​) walad (pl. ummahat al-​walad), 4, 5, 12, 15, 17, 18, 23, 29, 40, 41, 42, 47, 57, 65, 70, 107, 128, 129, 136, 146, 147, 171, 172, 229, 231, 232, 234, 236–​241, 244, 245, 251, 256, 259, 261, 297, 302, 309, 328 umma(h), 217, 327 Uqba, 252 Üsküdar, 216 Utba, 112, 278–​288, 291, 292, 330 Uthman ibn Affan, 15, 255, 298, 306 Uthman ibn Said al-​Amri, 252, 261 Uzbek, 193, 194, 199, 204 Uzun Hasan, 192 Uzun Süleyman Agha, 213 valad al-​zina (zina), 194, 193, 326 Valencia, 102, 166 Valide Sultan, 207–​221, 333 Vani Efendi, 210 Vanthieghem, Naim, 33, 45 Varisco, Daniel Martin, 236, 241 Venice, 91, 92, 208, 212, 216 Vienna, 209, 210, 220, 333 Volga, 170 Voltaire, 213, 215, 216, 220, 221 Von Hammer-​Purgstal, 210, 219–​221 Wahb ibn Abd Allah, 15 Wahb ibn Zama, 15, 16 waqf, (pl. waqfiyyat), 209, 331, 333 War of Zenta, 211 Ward the Elder, 112 Wasf al-​Firdaws, 267, 276 Wasif, 33, 45, 112 Wasif al-​Turki, 39

354 i wasiyya, 246, 259 Watha’iq, 131, 133 Wazir `Ali, 174 West African kingdom of Mali, 166, 177, 179 Yaʿazza ibn Makhluf, 145, 157 Yahya al-​Barmaki, 56, 57 Yahya ibn Ishaq al-​Masufi, Wanzamar, 150 Yaqub, 15 Yaqubi, al-​, 32, 44, 236, 240 Yaqub al-​Mansur, 150, 152, 153 Yaqub ibn al-​Layth al-​Saffar, 261 Yazdagird III, 254, 240, 329 Yazid III, 228, 230, 239, 262, 329 Yazid al-​Hawra, 284, 285 Yemen, 134, 165, 167, 175, 324, 326 Yintan ibn Umar, 145, 146 Yusuf Agha, 209, 210, 219, 333 Yusufzai chief Malik Shah Mansur, 192 Zabid, 167 Zad al-​Mal, 175, 184 Zafari, 175 Zahir al-​Din Muhammad Babur, 191–​193, 195–​204, 332

Index Zahiris, 300, 310, 312 Zahr al-​adab, 282, 285, 293, 294 Zajal, 102, 118 Zamzam, 277 Zanata, 128, 146, 148 Zanj, 249, 250, 325 Zaydan, Jurji, 61, 69 Zayd ibn Ali, 225, 226, 228, 230–​240 Zaynab Bint Ali ibn Yusuf, 150, 259 Zaynab-​Sultan Aghacha, 194, 201 Zaynab, (Zenobia), 335 zihar, 328 Zilfi, Madeline, 6, 37, 38, 43, 45–​47, 317, 323 Zinkeisen, Johann Wilhelm, 211, 216, 220, 221 Zirids, 143, 144, 148, 151, 156 Ziryab, 111–​113, 120, 121 Zubayda, 54, 59, 63–​65, 290, 334, 335 Zubayda-​Sultan, 191, 194, 195 Zubayr, 15 Zubayri, Mus`ab ibn Abd Allah, al-​, 11, 22, 23, 329 Zuhra Begi Agha, 199, 204 Zuhri, al-​, 306 Zukhruf, 129 Zumurrud, 133