Syria in Crusader Times: Conflict and Co-Existence 9781474429726

Citation preview

Syria in Crusader Times Conflict and Coexistence

Edited by Carole Hillenbrand

Edinburgh University Press is one of the leading university presses in the UK. We publish academic books and journals in our selected subject areas across the humanities and social sciences, combining cutting-edge scholarship with high editorial and production values to produce academic works of lasting importance. For more information visit our website: edinburghuniversitypress.com © editorial mater and organisation Carole Hillenbrand, 2020 © the chapters their several authors, 2020 Edinburgh University Press Ltd The Tun – Holyrood Road 12 (2f) Jackson’s Entry Edinburgh EH8 8PJ Typeset in Adobe Garamond by Servis Filmsetting Ltd, Stockport, Cheshire and printed and bound in Great Britain A CIP record for this book is available from the British Library ISBN 978 1 4744 2970 2 (hardback) ISBN 978 1 4744 2972 6 (webready PDF) ISBN 978 1 4744 2973 3 (epub) The right of the contributors to be identified as authors of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988 and the Copyright and Related Rights Regulations 2003 (SI No. 2498). Published with the support of the University of Edinburgh Scholarly Publishing Initiatives Fund.

Contents

List of Illustrations vi List of Contributors viii Prefacexiii Acknowledgementsxv Part 1  Sources   1 Hamdan al-Atharibi’s History of the Franks Revisited, Again Paul M. Cobb   2 Legitimate Authority in the Kitab al-Jihad of ‘Ali b. Tahir al-Sulami

3

21

Kenneth A. Goudie

  3 Politics, Religion and the Occult in the Works of Kamal al-Din Ibn Talha, a Vizier, ‘Alim and Author in Thirteenthcentury Syria A. C. S. Peacock Part 2  Christians   4 Adapting to Muslim Rule: the Syrian Orthodox Community in Twelfth-century Northern Syria and the Jazira R. Stephen Humphreys   5 The Afterlife of Edessa: Remembering Frankish Rule, 1144 and After Christopher MacEvitt

34

63

86

iv  |  syri a i n crusa d e r time s Part 3  Convivencia   6 Diplomatic Relations and Coinage among the Turcomans, the Ayyubids and the Crusaders: Pragmatism and Change of Identity105 Taef El-Azhari   7 Symbolic Conflict and Cooperation in the Neglected Chronicle of a Syrian Prince Luke Yarbrough

125

  8 A Critique of the Scholarly Outlook of the Crusades: the Case for Tolerance and Coexistence Suleiman A. Mourad

144

Part 4  War and Peace   9 The Portrayal of Violence in Walter the Chancellor’s Bella Antiochena 163 Thomas Asbridge 10 Infernalising the Enemy: Images of Hell in Muslim Descriptions of the Franks during the Crusading Period Alex Mallett Part 5  Cities 11 Sunnites et Chiites à Alep sous le règne d’al-Salih Isma‘il (569–77/1174–81): entre conflits et réconciliations  Anne-Marie Eddé

184

197

12 The War of Towers: Venice and Genoa at War in Crusader Syria, 1256–8 Thomas F. Madden

211

13 Gaza in the Frankish and Ayyubid Periods: the Run-up to 1260 CE  Reuven Amitai

225

Part 6  Saladin’s Men 14 Picture-poems for Saladin: ‘Abd al-Mun‘im al-Jilyani’s Mudabbajat  Julia Bray

247

contents  | v 15 Ayyubid Realpolitik and Political–Military Vicissitudes versus Counter-crusading Ideology in the Memoirist–Chronicler al-Katib al-Isfahani Lutz Richter-Bernburg

265

16 Assessing the Evidence for a Turning Point in Ayyubid– Frankish Relations in a Letter by al-Qadi al-Fadil Bogdan C. Smarandache

285

Part 7  Key Personalities 17 Saladin, Generosity and Gift-giving Jonathan Phillips

307

18 Hülegü: the New Constantine? Angus Stewart

321

Glossary336 Bibliography339 Index of Names 372 Index of Places 380 Index of Terms/Concepts 382

Illustrations

Figures   6.1 Bronze follaro of King William II of Sicily in Arabic – late twelfth century   6.2 Bronze follaro of King Tancred of Sicily in Arabic and Latin – late twelfth century    6.3 Crusader’s imitation of an Ayyubid dirham. Jerusalem mint – CE 1237   6.4 Armenian dirham of the thirteenth century with bilingual Armenian–Arabic writings 16.1 Letter by al-Qadi al-Fadil 18.1 Constantine and Helena with the True Cross (Vatican) 18.2 Constantine and Helena with the True Cross (British Library Board)  Plates Between pages 48 and 49   3.1 Ibn Talha’s da’ira, or circle of letters, in al-Durr al-muntazam    3.2 Ibn Talha’s Matalib al-su’ul fi manaqib Al al-Rasul   3.3 Ibn Talha’s al-‘Iqd al-farid li’l-Malik al-Sa‘id    3.4 Final folio of Ibn Talha’s Nafa’is al-‘anasir, showing the mysterious letters   3.5 Table for calculating days of the months from al-‘Iqd al-farid

119 119 120 121 300 323 326

i llustra ti ons  | vii Between pages 232 and 233 13.1 Detail from ‘Frankish Palestine, South’ 13.2 Further detail from ‘Frankish Palestine, South’ 13.3 Rural Palestine in the Frankish Period  13.4 The ‘Oxford Map of Matthew Paris’  Between pages 248 and 249 14.1 Al-Jilyani, Diwan al-tadbij, Manchester John Rylands MS Ar. 690, ff. 57b–58a: tree finial of Manadih al-mamadih 14.2 Transcription of reading of al-Jilyani, tree finial of Manadih al-mamadih  14.3 Al-Jilyani, Diwan al-Tadbij, Manchester John Rylands MS Ar. 690, f. 79a: the extemporised ‘Seal’ 14.4 Ibn al-Hajj, al-Diwan al-‘amm (Fez, 1995), 550: tree-shaped poem

Contributors

Reuven Amitai is Eliyahu Elath Professor for Muslim History at the Hebrew University of Jerusalem. His areas of research include the Mamluk Sultanate, the Mongols in the Middle East, processes of Islamisation, and medieval Palestine. From 2010 to 2014, he was Dean of the Faculty of Humanities at the Hebrew University, and from 2014 to 2016, he was a Senior Fellow at the Annemarie Schimmel Kolleg in Bonn. His recent publications include: Holy War and Rapprochement: Studies in the Relations between the Mamluk Sultanate and the Mongol Ilkhanate (1260–1335) (2013); co-edited with Michal Biran, Nomads as Agents of Cultural Change: the Mongols and Their Eurasian Predecessors (2015); co-edited with Christoph Cluse, Slavery and the Slave Trade in the Eastern Mediterranean, 11th to 15th Centuries (2017); and co-edited with Stephan Conermann, The Mamluk Sultanate from the Perspective of Regional and World History: Economic, Social and Cultural Developments in an Era of Increasing International Interaction and Competition (2019). In 2018, he received the degree of doctor honoris causa from the National University of Mongolia. Thomas Asbridge is Reader in Medieval History at Queen Mary, University of London. He specialises in the study in the study of the Crusades, knighthood and chivalry, and medieval violence. His major publications include The Greatest Knight (2015), The Crusades – the War for the Holy Land (2010) and The First Crusade: a New History (2004). Julia Bray has taught Arabic language and medieval literature at the universities of Manchester, Edinburgh, St Andrews and Paris 8-Vincennes – Saint Denis. She is currently Abdulaziz Saud AlBabtain Laudian Professor

contri butors  | ix of Arabic Literature at the University of Oxford. She works on medieval to Early Modern literature, life writing and society, and on the Arabic history of emotions and has published Stories of Piety and Prayer (2019), the first part of an edition and translation of the tenth-century al-Tanukhi’s Deliverance Follows Adversity, in the Library of Arabic Literature. Paul M. Cobb is Professor of Islamic History and Chair of the Department of Near Eastern Languages and Civilizations at the University of Pennsylvania in Philadelphia, USA. He is the author of numerous books and articles dealing with the history of the medieval Levant including, most recently, The Race for Paradise: an Islamic History of the Crusades (2014). He is also the translator of Usama b. Munqidh’s Book of Contemplation for Penguin Classics. Anne-Marie Eddé is Emerita Professor in Medieval Islamic History at Panthéon-Sorbonne University in Paris. Her research interests include medieval Arabic sources and the history of Syria from the eleventh to the fourteenth century. She is the author of La principauté ayyoubide d’Alep (579/1183–658/1260), Freiburger Islamstudien, XXI, Stuttgart, 1999, and Saladin, Paris, Flammarion, 2008, repr. 2012 (English translation, by Jane Mary Todd, 2011).  Taef El-Azhari is Professor of Islamic and Middle Eastern History at the University of Helwan, Egypt. He received his doctorate in Middle Eastern history from the University of Manchester. His interests, both in research and teaching, focus principally on Turcoman–Kurdish social–political history and the Crusades. His most recent books include Queens, Eunuchs and Concubines in Islamic History (Edinburgh University Press, 2019) and Zengi and the Muslim Response to the Crusades (2016). Kenneth A. Goudie received his PhD in 2016 from the University of St Andrews for a thesis entitled ‘The Reinvention of Jihad in Twelfth-century al-Sham’. He is now a postdoctoral research fellow at Ghent University. His current research focuses on the historical writings of the fifteenth-century Qur’an exegete and historian, Burhan al-Din al-Biqa‘i.

x  |  syri a i n crusade r time s R. Stephen Humphreys is Professor Emeritus in History and Islamic Studies at the University of California, Santa Barbara. Before coming to Santa Barbara, he taught at the University of Chicago and the University of Wisconsin-Madison.  He is the author of From Saladin to the Mongols: the Ayyubids of Damascus, 1193–1260 (1977), Islamic History: a Framework for Inquiry (1991) and Mu’awiya ibn Abi Sufyan: from Arabia to Empire (2006), in addition to other books and articles. He has been a visiting member of the Institute for Advanced Study in Princeton, New Jersey, and a visiting fellow at All Souls College, Oxford. Christopher MacEvitt is Associate Professor of Religion at Dartmouth College. His first book, The Crusades and the Christian World of the East: Rough Tolerance (2009), examines the ways in which Frankish rule and religious difference gave rise to what he calls ‘rough tolerance’, a mode of coexistence distinctive to the era of the Crusades. He has just completed a book on the narratives about Franciscans who died as martyrs in Islamic lands in the thirteenth and fourteenth centuries, arguing that the stories shape expectations of martyrdom and the image of Islam as the Franciscan Order grappled with accusations of heresy and controversies over poverty. Thomas F. Madden is Professor of History and Director of the Center for Medieval and Renaissance Studies at Saint Louis University. His books include Venice: A New History (2013), Istanbul: City of Majesty at the Crossroads of the World (2016), The Concise History of the Crusades (2014) and Enrico Dandolo and the Rise of Venice (2003). He is a Fellow of the John Simon Guggenheim Foundation, the American Council of Learned Societies and the Medieval Academy of America. Alex Mallett is Assistant Professor at the Waseda Institute for Advanced Study, Waseda University, Tokyo. He is the author of Popular Muslim Reactions to the Franks in the Levant (2014) and editor of Medieval Muslim Historians and the Franks in the Levant (2014). He is currently working on an edition and translation of the Ayyubid period in al-Maqrizi’s chronicle al-Suluk, as part of the Bibliotheca Maqriziana project.

contri butors  | xi Suleiman A. Mourad is a historian of Islam and the Middle East and Professor of Religion at Smith College. He is also Associate Fellow at the Nantes Institute for Advanced Study (France). His publications include The Intensification and Reorientation of Sunni Jihad Ideology in the Crusader Period (2013) and The Mosaic of Islam (2016). A. C. S. Peacock is Professor of Middle Eastern and Islamic History at the University of St Andrews. His research focuses on the history and intellectual history of the premodern Islamic world and Islamic manuscripts. His publications include Early Seljuq History: a new interpretation (2010), The Great Seljuk Empire (2015) and Islam, Literature and Society in Mongol Anatolia (2019). He has also edited Islamisation: Comparative Perspectives from History (2017). Jonathan Phillips is Professor of Crusading History at Royal Holloway, University of London. He is the author of several books including The Life and Legend of the Sultan Saladin (2019) and The Second Crusade: Extending the Frontiers of Christendom (2006) and, with Mike Horswell, he edited Perceptions of the Crusades from the Nineteenth to the Twenty-first Century: Engaging the Crusades I (2018). He is a convener of the MA in Crusader Studies at Royal Holloway and, with Professor Benjamin Kedar, the coeditor of the journal Crusades. Lutz Richter-Bernburg was Chair of Islamic Studies at the University of Tübingen from 2000 until his (mandatory) retirement in 2010. Previously he held a variety of research and teaching positions at Los Angeles (UCLA), Göttingen, Aleppo, Bonn, New York (Columbia), Berlin (Free University), and Leipzig. His publications include studies in Graeco-Islamic medicine and science, Ayyubid historiography and the intellectual history of formative and ­medieval Islam. Bogdan C. Smarandache finished his undergraduate studies at McGill University in 2011 and his MPhil at St Catharine’s College, University of Cambridge, in 2012. Bogdan is now a PhD candidate at the Centre for Medieval Studies in the University of Toronto and a former recipient of

xii  |  syri a i n crusad e r time s the Joseph-Armand Bombardier Graduate Scholarship. He recently finished his thesis on Christian–Muslim relations and the conditions of religious minorities in the Eastern Mediterranean under the supervision of Professor Mark Meyerson, Professor Linda Northrup and Professor Michael Gervers. Angus Stewart is Lecturer in Middle Eastern History at the University of St Andrews, and the author of The Armenian Kingdom and the Mamluks: War and Diplomacy during the Reigns of King Het‘um II (2001). His research focuses on the eastern Mediterranean area in the Ayyubid and early Mamluk era, looking especially at the arrival of the Mongols in the region. Luke Yarbrough is Assistant Professor in the Department of Near Eastern Languages and Cultures, University of California, Los Angeles. He is the author of Friends of the Emir: Non-Muslim State Officials in Premodern Islamic Thought (2019) and the editor and translator of ‘Uthman ibn Ibrahim al-Nabulusi, The Sword of Ambition: Bureaucratic Rivalry in Medieval Egypt (2016).

Preface

S

ince remote antiquity Greater Syria (which extended well beyond the frontiers of the modern state of Syria) has been a conflict zone. Its geographical position – bordering Egypt, the Holy Land, Jordan, Turkey and Iraq – made that well-nigh inevitable. So these lands have endured violent military struggles from all sides for millennia: Greece and Rome against Persia, Byzantines against Sasanians and then against Arabs. However, the coming of Christian European enemies to the Holy Land and Syria under Muslim rule from the 1090s was the beginning of an almost two centuries’ struggle of a rather different kind. Crusaders and Muslims in Syria undoubtedly waged violent war – ­battles, raids and sieges – against each other during the years of the Crusader presence (1099–1291) in the Holy Land, Egypt and Syria. Muslim disunity and apathy after the fall of Jerusalem in 1099 allowed the Crusaders to embed themselves in impressive and well-defended castles in remote rural areas of Syria, and they proved hard to dislodge. However, the Crusaders never succeeded in capturing the two key Syrian cities of Aleppo and Damascus. That meant that their military and political success in Syria was doomed to be partial at best. Moreover, only the First Crusade enjoyed full military success; all the other Crusades which reached the Levant failed, fizzled out or ended in truces. Nor was warfare in Crusader Syria continuous by any means. Truces and trading agreements across the religious divide brought about lengthy periods of peaceful contact between Crusaders and Muslims. Crusaders adopted aspects of the Muslim way of life and Muslims pragmatically developed their own commercial activities with Crusaders because they needed access to the Syrian ports, held by the Crusaders, in order to export their

xiv  |  syri a i n crusad e r time s goods. Coexistence rather than conflict between Crusaders and Muslims was ­therefore much more the norm for lengthy periods. Not surprisingly, then, their cultural interactions often found fascinating and unexpected expression. This book presents little-known and fascinating topics of Crusading and Muslim history. The special focus is on Syria in the twelfth century, though the later periods of the Crusader presence are also represented. There are chapters which deal with little-known Crusader and Muslim sources – ­chronicles, biographies, letters and poems. Other themes discussed include internal relations between Sunni and Shi‘ite Muslims, rivalries among Crusader factions, and contacts between Crusader and Oriental Christians. New insights into the career of Saladin are revealed, as well as unfamiliar aspects of the rule of his family dynasty. In sum, then, this book may justly claim to open a host of new perspectives on war and peace in twelfth- and thirteenth-century Syria, and what it was like to live there in those colourful times. It is also a testament to the vitality of this field which remains undimmed even though it has captured the Western imagination for centuries, and which nowadays has received much fresh impetus from the growing awareness of how Muslims themselves reacted to these events. Carole Hillenbrand

Acknowledgements

I

n April 2016 I organised in St Andrews a symposium entitled The History of Syria, 1099–1250: Conflict and Co-existence. Thanks to the wonderful generosity of the University of St Andrews, the symposium received full financial support and it became possible to invite a very impressive group of ­specialists – from the UK, France, Germany, Belgium, Qatar, Israel, Canada and the USA – all of whom study twelfth-century Syria, whether from Muslim, Crusader or Eastern Christian perspectives. A good number of the scholars who came to the conference had already produced path-breaking research; other younger ones were poised to do so. The beautiful environment of the Fife coastline in Scotland, wonderful accommodation and a most friendly and intimate atmosphere helped to make the symposium every bit as enjoyable and thought-provoking as I had hoped it would be. I should like to thank especially warmly Paul Churchill who, with his exceptional administrative skills, organised all the arrangements for the conference and managed everything so smoothly and amicably. A number of key colleagues at Edinburgh University Press have been most helpful and supportive in the publication of this book and I should like to give special thanks to Nicola Ramsey, Kirsty Woods, Eddie Clark and Lel Gillingwater for all their work in producing this book.

1 Hamdan al-Atharibi’s History of the Franks Revisited, Again Paul M. Cobb

Introduction

T

his chapter has two minor goals: to clarify and to speculate. The subject of both of these goals is the twelfth-century Syrian Muslim historian and man of letters Hamdan b. ‘Abd al-Rahim al-Atharibi (d. 1147). For many scholars of the Crusades in Syria, his name and the few details we have about him will already be familiar: what follows is thus not so much anything that is radically new, as it is a clarification of the details we do have about him (and which, to my mind, have not received adequate scrutiny) and an opportunity to speculate about some details that we, frankly, do not have. But, for all that, I do hope it will at least stimulate some new thought on the worlds – political, economic and literary – that he inhabited. My clarification and speculation will be spread over three broad areas: first, Hamdan’s context and career in northern Syria in the era of the First Crusade; second, his literary production, in particular his History of the Franks, which I feel is a somewhat misunderstood work; and finally, some thoughts about his place in the constellation of Syrian historiography so that we all might gain a better understanding of what Hamdan wrote, and what, therefore, we have lost. The Syrian Context

As no complete works of his own survive, what little we know about Hamdan comes from the biographical entry (tarjama) devoted to him in Ibn al-‘Adim’s biographical dictionary of Aleppo, the Bughyat al-talab fi ta’rikh Halab.1 This entry offers some important details about Hamdan’s career, his prose writings

4  |  syri a i n crusade r time s and samples of his poetry. Moreover, Ibn al-‘Adim (d. 1262) knew some of Hamdan’s descendants personally and they served as direct sources for him for some family lore that he reported in his own historical works. He also had in his possession some sections of Hamdan’s writings, and he quoted them here and there in the Bughya. Ibn al-‘Adim also made use of Hamdan in compiling his chronicle, the Zubdat al-halab min ta’rikh Halab, but – as was usually the case in that work – he is nowhere acknowledged by name.2 Here are the specifics that we know: Hamdan was born around 460/1067 and died in the year 542/1147 in his eighties. Ibn al-‘Adim tells us he spent almost all of his life in the al-Jazr district, that once-fertile land of the country homes, villages, rural monasteries, vineyards and orchards that populate the extant verses of Hamdan’s poetry, and which became a contested border-zone between Muslim Aleppo and Frankish Antioch.3 In this zone, to use Ibn al-‘Adim’s words, Hamdan ‘went back and forth between the Islamic and Frankish regimes’ over the course of his life. And what a life he lived: those eighty years, from around 1067 to 1147, cover one of the most turbulent periods in the history of Syria. It is instructive to link the major events of northern Syrian history at this time with Hamdan’s own lifespan: his childhood saw the coming of the Seljuq Turks to Syria under Alp Arslan (1071), his young adulthood the second reassertion of Seljuq power under Malikshah (1086), and the establishment of the Seljuq prince Ridwan in Aleppo in the 1090s. When the Frankish Crusaders appeared before the walls of Antioch in 1097, Hamdan was an accomplished man of about thirty. When the Turcoman dynasty known as the Artuqids began its unstable, on again–off again rule of Aleppo in the 1110s, he had hit middle age, and he was in his sixties when another Turcoman ruler, ‘Imad al-Din Zengi, gained control of the city more successfully in 1128 and took over most of northern Syria (Frankish or Muslim) as a sequel. Hamdan died without witnessing the failed Second Crusade, which arrived in Syria by late 1147. One wonders what he would have made of it. Geographically speaking, Hamdan was a local talent, spending most of his life in the orbit of Aleppo, albeit with a few journeys to other regional courts. He was born in a village in the vicinity of Aleppo called Ma‘ratha al-Atharib, that is, the modern Ma‘arat al-Atarib, just a few miles north of the larger town of Atarib, medieval al-Atharib, between Aleppo and Antioch. He

history of the franks revi si te d , a ga in  | 5 and his father then moved to al-Atharib proper and settled there, presumably when he was still a child. The outlines of his adult career are rather more complicated. It seems impossible to reconcile the very clear and specific information that we have about the places he lived, the jobs he held and the dates he held them with what is the best known aspect of Hamdan’s life: that over the course of his life he served both Frankish and Muslim lords, and that, as is often remarked, Hamdan stands as one of the few Muslim subalterns mentioned in the sources who worked willingly for a Frankish lord. While there were undoubtedly many more, his case is the best documented, even if it seems impossible to add any detail. What follows is, nevertheless, an attempt to do just that. Since Ibn al-‘Adim’s tarjama of Hamdan mentions Zengi by name (waliya ‘amalan li’l-diwan fi dawlat atabak zanki ibn aq-sunqur), it has always been assumed that Hamdan worked first for the Franks after they took control of the region, even becoming a vassal and landlord under the Frankish lord of al-Atharib, and then shifted his allegiance to Zengi, once he pushed the Franks out around 1130. Indeed, this seems confirmed by the nisba Ibn al-‘Adim grants him in the rubric of his tarjama, where Hamdan is listed as al-atharibi thumma al-halabi (of al-Atharib, latterly of Aleppo), implying that, after living in al-Atharib for some time, he moved to Aleppo to serve his new lord, Zengi. But a close reading of the tarjama suggests there is another line to add to Hamdan’s curriculum vitae: namely, that he had first lived in Aleppo and worked for the Turcoman lords there long before Zengi ever entered the picture. This seems borne out by some of the details of the diplomatic missions he was sent on, all from Aleppo. Ibn al-‘Adim is quite clear about this, although he provides no dates: Hamdan settled in Aleppo and served as a messenger to the Franks, and travelled to Egypt to al-Amir, the Fatimid, and he was also sent to Damascus as a messenger to the atabeg Tughtegin. He also entered Baghdad.5

Based on the vague information that Ibn al-‘Adim provides in this passage we might well assume that these Aleppo-based activities were missions on behalf of Zengi, since he names him explicitly as Hamdan’s patron early on, and the missions to the Fatimid caliph al-Amir and the atabeg Tughtegin in Damascus could well have occurred in Zengi’s early career. However, a

6  |  syri a i n crusade r time s later chronicler, Ibn Muyassar (d. 1278), happens to also mention the arrival of Hamdan in the Fatimid court, bearing a message from Aleppo, and he provides the crucial datum-point that this mission took place in the year 520/1126. At that time, Zengi was not on the scene at all and Aleppo was under the control of the commander Aq-Sunqur al-Bursuqi, or rather one of his deputies.6 What Hamdan’s business was in Cairo is unknown, but a family tale that was recounted to Ibn al-‘Adim says that he was at first denied access to al-Amir, because the caliph had heard Hamdan was a hashishi, that is, a Nizari Isma‘ili opposed to al-Amir’s rule. This might well be true (we know little of Hamdan’s religious beliefs), and it certainly didn’t help Hamdan that he was associated with Aleppo and the district of al-Jazr, which was solid Nizari country. According to the tale, it was only after Hamdan offered some clever verse to reassure the caliph that he was finally allowed an audience.7 Similarly, we can establish Hamdan’s mission to Damascus as occurring before Zengi, too, since Ibn al-‘Adim, quoting Ibn ‘Asakir, specifies that Hamdan was to meet with the atabeg of Damascus, Tughtegin, who ruled from 1104 until his death in 1128. Now, it happens that 1128 is also the year Zengi took control of Aleppo, but as Tughtegin died in February of 1128, and Zengi only took power in June, Hamdan’s mission to Damascus, like his visit to Cairo, must have been for some other lord of Aleppo before Zengi.8 So at least two of the official voyages that Hamdan took from Aleppo that Ibn al-‘Adim mentions took place before Zengi took power, though we cannot be very precise under whom, given the very fluid situation in the city at the time and the absence of any further details in Ibn al-‘Adim’s account. The other voyages that Ibn al-‘Adim mentions, to the Franks and to Baghdad, could have happened at any time, though the Baghdad trip is almost certainly to be identified with a trip he took to Baghdad that Ibn al-‘Adim specifies occurred in 540/1145, two years before Hamdan’s death. Appropriately enough, this is also the last voyage mentioned in Ibn al-‘Adim’s little list, suggesting that the whole passage is arranged chronologically. This, too, accords with the evidence from Cairo and Damascus, which also appear in chronological sequence in this list. It seems, then, that Hamdan made his move to Aleppo before he began working for Zengi. This would make sense: Hamdan would have been among

history of the franks revi si te d , a ga in  | 7 the local bureaucratic talent already working in Aleppo when Zengi arrived and the warlord simply absorbed him into his administration. The many anecdotes that Ibn al-‘Adim recounts about Hamdan’s salons and drinking parties in Aleppo, then, could equally have belonged to this earlier period as to some later time of residence under Zengi. And it was probably in these earlier years in Aleppo and not later under Zengi that he gained the literary skills for which he was famous: studying adab with the Aleppine shaykh Ibn Abi Jarada (d. 1153), and also grammar, history, astronomy, mathematics and medicine. What then of his service to the Franks, that much commented-upon feature of Hamdan’s career track? We have established that Hamdan worked in the service of Muslim lords of Aleppo before Zengi arrived, and this earlier period in Aleppo adds a new colour to how he entered the service of the Franks. It is usually assumed, for example, that Hamdan was already working for the Franks when the Frankish lord of al-Atharib famously granted him lands of his own. The story begins with a simple general statement that when the al-Jazr district was ‘in the hands of the Franks’, they appointed him to administer some lands and then later confiscated his wealth (wa-sadiruhu ba‘da dhalika) – the common fate of many a medieval administrator. However, the al-Jazr district was in the hands of the Franks by c.1111 when the Franks took al-Atharib. So Hamdan’s tax administrative job and subsequent disgrace could have taken place at any time in the period from 1111 to 1126, when, as we’ve seen, he was already working for Aleppo’s Muslim rulers on a mission to Fatimid Cairo. To this general picture, another phrase elsewhere in his tarjama adds the detail that Hamdan was administering the diwan of Ma‘arrat al-Nu‘man when the Frankish lord of al-Atharib granted him lands. Note that this text does not tell us for whom he was administering this diwan. It is usually assumed, as by Cahen, that these two reports refer to one and the same position of service – that the supervision of the diwan of Ma‘arrat al-Nu‘man mentioned in this second account took place as part of the vaguely described administrative work he did for the Franks described in the first, in short, that the diwan in question was a Frankish-controlled office. But, leaving aside matters of administrative terminology, there are a few reasons not to make this assumption. For one, in still a third account in Hamdan’s tarjama, Ibn al-‘Adim specifies that Hamdan received his lands

8  |  syri a i n crusade r time s from the Frankish lord of al-Atharib in late 521/1127. That is, less than a year after he was, as we have seen, working as a messenger to Egypt for the rulers of Aleppo, and before Zengi had taken over. I suggest, therefore, that when Hamdan was administering the diwan of Ma‘arrat al-Nu‘man, he was doing so for Muslim Aleppo and not the Franks. That means that the first short account that describes him administering lands for the Franks before subsequently having his wealth seized must be situated before his move to Aleppo, perhaps in the late 1110s. This also makes narrative sense with that third and most detailed account of this granting of lands. The account is narrated by one of Hamdan’s kinsmen, and it begins as an account about Hamdan’s studies in Aleppo and his relationship with his master, Ibn Abi Jarada. The account then simply says: ‘It then occurred to Hamdan go out to Ma‘aratha al-Atharib [the village near al-Atharib where Hamdan was born], which was his property though it was in the hands of the Franks at that time.’ Hearing that the Frankish lord of al-Atharib was ill, Hamdan stopped to treat him and he cured him. In return, the Frankish lord asks him what he would like as a reward. Hamdan, whom, you will recall, is described as having had his wealth confiscated during his earlier period of employment under the Franks, simply replies that he would like a village. So the lord granted him a village called Ma‘arbuniyya in the vicinity of Ma‘arat Misrin (probably the modern Kafr Buniya), and this village remained in the hands of Hamdan’s kin right up until Ibn al-‘Adim’s day. At the time he received it, the village was a deserted wasteland, but Hamdan developed it and made it his home, bringing his whole household there, building a palace, settling peasants, planting and harvesting. Ibn al‘Adim says he remained there ‘for thirty years’, though this can hardly be accurate since he died within twenty. It is worth pointing out, moreover, that nowhere does it say that Hamdan served the lord of al-Atharib in any capacity, only that his master back in Aleppo, Ibn Abi Jarada, criticised him for dwelling among the Franks in this way. To this he humbly apologised in verse, reminding his master that even perfumed musk has its home in ­stinking glands, and the pearl in the crusty shell of the oyster. Granted, when a Frankish lord gives lands to someone else in return for a service rendered, it certainly looks like feudalism; but the fact is we have no details at all about Hamdan’s relationship with the lord of al-Atharib – so talk

history of the franks revi si te d , a ga in  | 9 of Hamdan’s ‘vassalage’ to the Franks may be a bit premature. Having thus lost his property while first serving the Franks in some vague administrative capacity, during his tenure under the Muslim lords of Aleppo he was able to gain new property at the expense of the Franks because of his kindness and skill as a physician. It is entirely possible that he continued to serve Aleppo while residing at his lands at Ma‘arbuniyya. Whatever the case, when Zengi finally captured Aleppo in 1128, Hamdan continued to administer the lands of the al-Jazr district for his diwan, as he had done for the Franks in his early career, and as he had done for the lords of Aleppo before Zengi. Hamdan’s career path, upon a close-reading, thus turns out to be a bit more complicated than it is usually understood to be.9 Literary Output As we’ve already established, Hamdan was an acknowledged member of the literary circles of Aleppo and northern Syria, a generous host and companion to other literati. He was also something of a hellraiser, famous for his drinking parties at various pastoral settings across the district, some of them raucous. In one such party, he nearly killed himself falling off the roof of a house he had fallen asleep on. He was known for his sharp-tongued satire as much as his social graces. He was a proud member of the venerable Arab tribe of Tamim, and traced his ancestry to the pre-Islamic Tamimi hero and poet Hajib b. Zurara. Indeed, he is said to have compiled a book on the various ayyam of the Banu Tamim, their deeds, heroic battles and their poetry. He also produced a diwan of his own poetry, an autograph manuscript of which Ibn al-‘Adim had consulted, though it had lost a few pages. But certainly his most famous work, the one for which we still write about him, is his now-lost history of the Franks. How much do we really know for certain about this work? It has become a trope in writing about Hamdan that modern authors must mourn the loss of this work, and I can hardly disagree. Any coverage of the Franks and their coming to the Near East by a Muslim observer is a rare thing to be treasured. But what more can we say about it? We might start with its title. Cahen was the first to bring Ibn al-‘Adim’s tarjama of Hamdan to our attention, and he read the title of Hamdan’s history as al-Muwaffaq (‘The Suitable’), a perfectly acceptable title for a book of history, I suppose. More recently, Eddé, following the edition of Zakkar,

10  |  syri a i n crusa d e r time s reads the title as al-Mufawwaf, meaning ‘Thin’ or perhaps ‘White-striped’. Less comprehensible, but entirely feasible titles, too, both plausible readings of the same unpointed text. And this must be the same title behind the book al-Sakhawi calls al-Qut (‘Nourishment’), which I take to be a copyist’s error for al-muwaffaq/al-mufawwaf. Indeed al-Sakhawi’s description of the work is a direct quotation of Ibn al-‘Adim, so I presume this was his source. In other words: we don’t really know what it was called.10 We do, however, have some useful details about its contents. Yaqut only mentions him as the author of a ta’rikh without any details, though he acknowledges him too as a physician and poet and cites some of the same samples of verse that Ibn al-‘Adim preserves. Ibn al-‘Adim, however, has by far the earliest and most detailed account of Hamdan’s historical works, which he describes as follows: He compiled a book on the history of Aleppo from the year 490 [1096] which includes accounts (akhbar] of the Franks and their deeds and their going forth to Syria in the aforementioned year and what happened ­afterward, calling it al-Mufawwaf. 12

Ibn Muyassar, the chronicler of Fatimid Egypt, in his account of Hamdan’s mission to Cairo, describes him as the author of a book with a rather different focus: A History of the Franks Who Went Out to the Lands of Islam in These Years. The conventional wisdom on Hamdan’s output has always been that these two descriptions refer to the same work, what is usually called by way of paraphrase his History of the Franks. But I would like to suggest that this information could be read quite differently, and that what we really have are descriptions of two lost works about the Franks. From the description of the Mufawwaf, the work described by Ibn al-‘Adim, that work sounds like a larger history of Aleppo, with perhaps a separate section devoted to the Frankish material: wada‘a kitaban fi ta’rikh halab min sanati tisa‘in wa-arba‘ami’atin dimnahu akhbar al-firanj wa-ayyamihim wa-khurujihim ila al-Sham min alsanati al-madhkurati wa-ma ba‘dahu (‘He composed a book on the history of Aleppo from the year 490, which includes accounts of the Franks, their deeds, and their coming out to Syria in the aforementioned year and the events that followed [my emphasis].’)

history of the franks revi si t e d , a ga in  | 11 Ibn Muyassar, on the other hand, seems to be describing a monograph on Frankish history, and indeed his description of it, The History of the Franks Who Went Out to the Lands of Islam in These Years, does not sound like its description so much as, given its metre and rhyme in Arabic, its title. Hamdan is called: the composer (musannaf ) of Sirat al-ifranj al-kharijin ila bilad al-islam fi hadhihi’l sinin. Note that this is not called a ta’rikh. This seems to me to be altogether different, and rather more robust, than a history of Aleppo that includes some akhbar of the Franks and their history. It is perhaps telling that, of the quotations from Hamdan that Ibn al-‘Adim uses, most are from a work that he describes as ta’rikhuhu alladhi jama‘ahu (the history that he compiled), and none of these mention the Franks but rather the internal history of Aleppo, and are described as being on loose pages that came into his possession, in Hamdan’s own handwriting – what I am suggesting is the history of Aleppo known to him as al-Mufawwaf/al-Muwaffaq. The one account that does discuss the Franks, however, is said to come from something called the Akhbar al-Ifranj, which Ibn al-‘Adim says he read in a manuscript copy by Yahya b. al-Marawi al-Halabi, so not loose pages in Hamdan’s own hand. The accounts about Aleppo are short, straightforward entries of the sort one expects to find in a chronicle of Aleppo, as in Ibn al‘Adim’s own Zubda or in al-‘Azimi’s Ta’rikh. The account from the Akhbar al-Ifranj, on the other hand, is a much longer, and more literary affair (see the translation in the Appendix). It happened, Hamdan tells us, that some workmen in the employ of Yaghi-Siyan, the ill-fated governor of Antioch, who would lose his city to the Franks in the First Crusade, discovered a covered stone basin. Peeking inside, they discovered a group of brass figurines of mysterious horsemen, ‘each dressed in a long coat of chain mail, grasping a shield and spear’.13 Puzzled, the amir asked a group of the local city elders – native Christians – what they thought these figures could represent. The elders too were at a loss. But the figurines reminded them of something that took place many years before, when the city was still in Byzantine hands. In 1084 the walls of a local monastery had collapsed, and, during the reconstruction work that took place back then, they discovered a similar stone basin, containing brass figurines of horsemen bearing bows and arrows, which they easily identified as Turks. They thought little of the discovery.

12  |  syri a i n crusa d e r time s Little, that is, until a short time later, when an army of precisely these sorts of horsemen – the Seljuq Turks – captured the city and subjected it to decades of Turkish rule. Perhaps, the elders suggested to the amir, these strange new figurines in chain mail represented some other conquering nation still unknown to Antioch? Yaghi-Siyan merely scoffed at them, ignoring their interpretation. Nevertheless, as one might have predicted with a tale like this, word soon arrived that the Franks, undoubtedly dressed in the distinctive armour of these figurines, had encamped before Constantinople, an ironic end to Yaghi-Siyan’s story. But the tale of the figurines doesn’t end there, as Hamdan added a contemporary postscript of his own. In the year 1118, Hamdan tells us, Roger of Salerno, the lord of Antioch, found himself in need of building materials, so he sent some men to a ruined palace in Antioch to strip it of what they could use. It was there that Roger’s workmen found, just as in the past, a covered stone basin containing a figurine of a horseman (this time, only one): ‘except that the mount had features inconsistent with a horse and the horseman wore a head-wrap such that only his eyes showed’. When this figurine was presented to Roger, someone related that story about the Turkish and Frankish figurines, and so he dutifully looked into the matter. One of the priests he questioned begged him to destroy the figure. Roger did as he was told, and at that moment word came that an Egyptian army had arrived before Jerusalem. Roger marched south to the defence of the Holy City, yet to his surprise the Egyptians were routed, and Jerusalem was unharmed. Roger returned in triumph to Antioch, having concluded, no doubt, that the curse of the figurines had been broken. But then, a few days later, Roger decided to attack A‘zaz, and it was only then that Aleppo called upon the group that was truly signified in that mysterious figurine: the Turcoman troops of Il-Ghazi, who swiftly defeated Roger and the flower of Antioch’s fighting men at the Battle of the Field of Blood. Interestingly, this entertaining story was reused by a later Christian source. A version of the tale that closely matches Hamdan’s, but which ends with the defeat of Yaghi-Siyan, appears in the Chronicle of Michael the Syrian (d. 1199) and, with some small changes, in Vardan Arewelc‘i’s later Armenian translation (d. 1271).14 The precise direction of borrowing is impossible to determine; it is just as likely that both Hamdan and Michael

history of the franks revi si t e d , a ga in  | 13 got it from a missing third source than one from the other. Be that as it may, for these Christian sources, as for Hamdan, the tale conveniently explained the fall of Antioch to the Crusaders as fated from the first, and a result of Yaghi-Siyan’s stubbornness.15 For his part, Michael adds a few details to highlight Yaghi-Siyan’s cruelty and, in a rather wistful concluding phrase, states that when the Franks arrived and took Antioch, they ‘promised to the Lord that if they were granted entry to Jerusalem, they would live in peace with all Christians and grant churches and monasteries to every nation that confessed the faith of Christ’. Hamdan, however, was seeking a different moral to this story, and this suggests that the postscript about Il-Ghazi and the more recent battle at the Field of Blood was his own addition. It was important for Hamdan to note that the Frankish ruler Roger never managed to escape his fate, but the Turcomans never conquered Antioch either, as the Seljuqs and the Franks had done. There was a lesson here: ‘If only the army of Il-Ghazi had continued on to Antioch, he would have taken it. But he was over-cautious in the affair. To God alone belongs the Will!’ For Hamdan, the Muslim reconquest of Antioch was simply not to be, no matter what ominous figurines from the past may promise. For God’s will can reverse all fates, even those laid down in remote antiquity and proven time and again to be built into our foundations. This story of talismanic figures, rather like Hamdan himself, then, served both Christians and Muslims in the age of Crusades, in similar form, even if to differing ends. Conclusion It appears then that we can at least provide some greater articulation to both Hamdan’s career and his works. On the one hand, Hamdan worked for the Muslim lords of Aleppo as an administrator well before he joined Zengi’s service, as the chronology of his visits to various regional courts makes clear. Moreover, although he was clearly employed by the Franks before working for Aleppo, the episode in which Hamdan was famously granted lands in gratitude from the Frankish lord of al-Atharib may well have occurred after, rather than during, his period of service to the Franks. Indeed, the account of this land-granting seems to require this sequence of events, as it is part of a narrative in which Hamdan loses his property during his earliest period

14  |  syri a i n crusa d e r time s of service to the Franks, but then, as a satisfactory and ironic testament to Fate, he gains new lands later on. It is, admittedly, unclear whether this granting happened while he was still employed by the Muslim lord of Aleppo, but it makes the resolution and irony all the more powerful if he was. Understanding the sequence of his service in this way thus raises the question about the degree to which Hamdan ever thought of himself as a ‘vassal’ to the lord of al-Atharib, but that remains unanswerable. On the other hand, there is good evidence to suggest, I think, that Hamdan had composed not one, but two, works that involved accounts about the history of the Franks in Syria. One of these seems to have been a chronicle of Aleppo in the manner of many similar chronicles of Aleppo, known as Al-Muwaffaq/Al-Mufawwaf and contained some accounts of Frankish history. This I identify with the work that Ibn al-‘Adim cites as Hamdan’s ta’rikh, of which he has seen only a few sheets in the author’s hand. Another work is described in slightly separate terms as if it were solely devoted to the Franks and their coming to Syria, known as Sirat al-Ifranj and this is what I am identifying with the work Ibn al-‘Adim calls Hamdan’s Akhbar al-Ifranj, which he cites only as a work cited by another companion of his. Granted, titles of medieval Arabic works are famously mutable, and there remains also the possibility that the Sirat al-Ifranj is not so much a separate, independently circulating work of its own so much as a discrete section within Hamdan’s larger history of Aleppo. But by positing two separate works we can at least account for the varying description of Hamdan’s oeuvre that appear across the sources. In providing this articulation, we are now in a better place to assess Hamdan’s legacy. Consider him, for example, in comparison with his younger and better-known near-contemporary, the warrior-poet Usama b. Munqidh (d. 1188). Both were products of the milieu of northern Syria. Both were Arab elites who came under Seljuq rule. Both were udaba’, literati versed in prose and poetry and driven by diverse intellectual interests. Both of their families probably had dealings with one another, given the early history of the Banu Munqidh with the court of Aleppo. But whereas Usama was a product of the ruling military elites at Shayzar, and later the military men that he served, Hamdan was a civilian. His skills marked out him out as an administrator and bureaucrat, not a courtier or warrior. And whereas Usama

history of the franks revi si t e d , a ga in  | 15 spent most of his life wandering the Near East in service to various patrons in Egypt, Syria and northern Mesopotamia, Hamdan remained for almost the entirety of his life in al-Jazr in Syria. As much as Usama embodied the cosmopolitan, transregional elite culture of the military elites of the middle periods of Islamic history, Hamdan was, for his part, profoundly local, a civilian administrator for Franks and Muslims, who hardly spent any time beyond a few kilometres from Aleppo. Whether this would have coloured his works in any distinctive fashion, we can only guess. But the fragments of his writings that do survive thanks to zealous compilers like Ibn al-‘Adim (most of which are translated here; see the appendix) do suggest that it is still appropriate to mourn the loss of his works, maybe now even doubly so. Appendix: Fragments from Hamdan in Ibn al-‘Adim’s Bughyat al-talab 1. Ibn al-‘Adim, 1988, I: 481–483 Ibn al-Muhadhdhab and al-‘Azimi agree that this [Seljuq conquest of Antioch] was in the year 467. But that isn’t how the matter stands; rather, Sulayman b. Qutulmish conquered Antioch in 477, and so it seems that Ibn al-Muhadhdhab was writing that but his pen slipped making 77 into 67, and so he wrote it down in error, and al-‘Azimi simply transcribed that error in his Ta’rikh. The truth is what Hamdan b. ‘Abd al-Rahim al-Atharibi mentioned in Akhbar al-Firanj. I read a copy of it written in the hand of al-ra’is Yahya b. al-Marawi al-Halabi, who mentioned that he transcribed it from an autograph manuscript of Hamdan b. ‘Abd al-Rahim, who said: One of the wonders of the age was that a great earthquake destroyed Antioch four years before its conquest, causing a number of its towers to fall from its walls. The qadi Hasan b. al-Muj al-Faw‘i related the following, saying: I had fled from al-Mijann and arrived at Antioch, entering the service of the Most Venerable Mas‘ud, the vizier of Yaghi-Siyan, and he put me in charge of public works … we rebuilt the portions of the city walls that the earthquake had destroyed, but one of the towers collapsed again and failed. We were advised to rebuild it and to firmly establish its foundations, so we tore it down. We dug down to the lowest course of its foundations, and we discovered there a stone basin, with a great lid that had broken on top of it. We examined it and we found inside it seven brass figurines, sitting astride horses

16  |  syri a i n crusa d e r time s made of brass. Each figure was dressed in a long robe of chain mail, grasping a shield and spear. He said: ‘So I informed the Most Venerable Mas‘ud, and he sent out his assistant, who removed the figurines and examined what lay beneath the basin, but he found nothing else there. So, he carried the figurines to the vizier, who took them in his turn and presented them during the majlis of the amir Yaghi-Siyan. One of those present said, “Perhaps if the amir requests the presence of one of the old men of the city, he could reveal for him the truth of this matter.”’ So, he went ahead and assembled such a group, and I showed them the figurines. It was then said to them, ‘Do you know what these figurines are?’ They replied, ‘We don’t know. But we can relate to the amir something very similar to this affair.’ [I: 482] ‘We have a monastery here known as King’s Monastery (Dayr al-Malik), [built in the form of] one large, wide space. In the year 477, it collapsed upon us, and most of its timbers were broken. So, we rebuilt it and sought out timbers for it of an appropriate span, but we couldn’t find anything in Antioch or its hinterland. As a result, one of the builders advised us to move the outer walls. So, we dug the foundation trenches for the new wall and when we reached the bottom, we discovered figurines of Turks made of brass, with bows and arrows on their torsos, but we didn’t make a fuss over it and built the wall. It was only a small space of time that passed when Sulayman b. Qutulmish seized the city in early Sha‘ban in the year 477, with four hundred or so ghulams, and he ruled us as the amir has heard tell. Perhaps these figurines are from this other nation [we have heard of], whose features are like the Arabs or other Muslim groups.’ And they proceeded to make allusions to the subject of Franks, but only odd tales had reached them, and [stories] that would embolden anyone who heard them against them. And so Yaghi-Siyan cursed them in the harshest terms, saying: ‘What, there are infidels in this world other than Turks?!’ And he ordered them to be taken away. But the affair had barely concluded when it was reported that the Franks had encamped before Constantinople. [Ibn al-‘Adim interjects] This is what the qadi Hasan b. al-Muj related, and all of the dating confirms the fact that Sulayman b. Qutulmish attacked Antioch in the year 477.

history of the franks revi si t e d , a ga in  | 17 Hamdan b. ‘Abd al-Rahim reported after this tale (and I [Ibn al-‘Adim] transcribed it from the manuscript of Ibn al-Marawi): An example similar to this was when Roger, the lord of Antioch, found himself in need of marble to make use of. Someone mentioned to him, ‘As it happens, in such-and-such a place there is a palace that was built by the king who founded Antioch, and there is indeed wonderful sorts of marble to be had there.’ So, he ordered that it be located (this was in the year 512). As it was being explored a marble basin was found, with a horseman astride a horse inside it, except that the mount had features inconsistent with a horse and the horseman wore a headwrap such that only his eyes showed. When this figurine was presented before Roger someone related that story about the Turkish and Frankish figurines, and so he looked into it. One of the priests told him, ‘Dash it on the ground so that it breaks and breaks its evil with it!’ And so Roger dashed it against the ground until it broke. And on that Friday, the watchman arrived from Jerusalem to inform him that the Egyptian army had encamped before them. And so he marched out and until [I: 483] he reached them and made a show of fighting their army and the two sides fought for a few days. Then the Egyptian army retreated, as it had been broken, and Roger returned to Antioch. He did not remain there for ten days before he went out to A‘zaz and besieged it, and so the Aleppines sent word to Il-Ghazi b. Artuq, requesting his aid and offering him control of Aleppo. He roused the Turcomans and they came to him and they assembled at Laylun Notch at a place named Tell ‘Aqbarin. He destroyed the Franks, killed Roger and took his head. Thousands of Franks were killed and if only the army of Il-Ghazi had continued on to Antioch, he would have taken it. But he was over-cautious in the affair. To God alone belongs the Will! 2. Ibn al-‘Adim, 1988, IV: 1969–1970 (s.v. Aq Sunqur b. ‘Abd Allah al-Bursuqi) I read in an autograph manuscript of Abu’l-Fawaris Hamdan b. ‘Abd alRahim of the chronicle (ta’rikh) that he compiled, of which a few leaves came into my possession, and from which I have extracted some of the events of the year 520 (1126): Al-Bursuqi handed over Aleppo and its administration to his son, the commander ‘Izz al-Din Mas‘ud, who entered the city, and improved its affairs and took pleasure in doing good deeds. His father

18  |  syri a i n crusa d e r time s [al-Bursuqi], meanwhile, moved on to Mosul, the two Jaziras, and other of the lands adjoining his realm. On Friday, the ninth of Dhu’l-Qa‘da of that year, al-Bursuqi went out to the mosque in Mosul, intending to perform the Friday prayer and listen to the preacher, as was his custom on most Fridays. He entered the mosque and headed in the direction of the pulpit. When he got close to it, eight individuals dressed as ascetics attacked him, brandishing daggers and making right for him. They passed through the bodyguards who were around him and they struck at him until he was covered in wounds, injuring some of the bodyguards in the process. The guards managed to kill some of the attackers and took the others into custody, and al-Bursuqi was carried back to his home breathing his last breath. Everyone in the mosque fled, and the Friday prayer service was cancelled. Al-Bursuqi died that same day, and his companions killed those Batinis who remained in their custody. Not a single one of the Batinis escaped except for one youth, who was from Kafr Nasih, a village in the district of A‘zaz to the north of Aleppo. Hamdan said, according to what I copied down from his autograph manuscript: A man from that village related to me that that youth had an elderly mother. When she heard about the murder of al-Bursuqi – and she knew that her son was among the group that had been inciting his death – she celebrated the fact and adorned herself in kohl, sitting back in utter delight as if it was for her a feast day. But then, a few days later, her son returned to her safe and sound, and this grieved her. So, she rose and cut her hair and blackened her face. 3. Ibn al-‘Adim, 1988, IV: 1972–1973 (s.v. Alp Arslan) I read in an autograph manuscript of Abu al-Fawaris Hamdan b. ‘Abd alRahim: Mahmud [b. Nasr, the Mirdasid lord of Aleppo] and his mother both came out [of Aleppo] to meet [the sultan Alp Arslan, who was besieging it], and he ceded the city back to them after a siege of thirty-one days. The sultan heard that the king of the Greeks, [Romanos] Diogenes, had left Constantinople on the road to the Syrian passes, so he travelled out from Aleppo five days after meeting with Mahmud. The sultan made directly for [Romanos] until he encountered him at Manzikert, battling against him until he routed him. He captured the king of the Greeks and plundered their camp. The number of the Turks was 700,000 men.

history of the franks revi si t e d , a ga in  | 19 4. Ibn al-‘Adim, 1988, IX: 4158–4159 (s.v. Salim b. Malik b. Badran al-‘Uqayli) I read in an autograph manuscript of Hamdan b. ‘Abd al-Rahim: I saw in some work of commentary that the amir Siraj al-Din Salim b. Malik b. Badran al-‘Uqayli, lord of al-Dawsariyya, which is to say Qal‘at Ja‘bar, died on 20 Sha‘ban of the year 519. Notes   1. Ibn al-‘Adim, 1988, VI: 2926–2932. This is the earliest reference to Hamdan, though he is also cited, largely under the influence of this tarjama, by Yaqut, Ibn Shaddad and others.  2. Cf. the account of the embellished tale of the aftermath of Aq-Sunqur alBursuqi in the Zubda (Ibn al-‘Adim, 1968, II: 235) with that quoted in Bughya (Ibn al-‘Adim, 1988, IV: 1969–1970).   3. Even in Ibn al-‘Adim’s day the al-Jazr district was famous as the favoured locale for the country homes of Aleppo’s elite. Ibn al-‘Adim, 1988, I: 133–134.   5. Ibn al-‘Adim, 1988, VI: 2926.   6. On the career of Aq-Sunqur, see Mallett, 2011, 39–56.   7. Ibn Muyassar, 1919, 70; Ibn al-‘Adim, 1988, VI: 2928; Ibn al-‘Adim, 1968, II: 231–236, which covers the reign of al-Bursuqi, offers no evidence of any gestures toward Egypt. But Hamdan may have been sent, for example, to announce al-Bursuqi’s taking power, to suggest a coordinated attack against the Franks, or to solicit aid during the famine that struck the region that year.   8. Ibn al-‘Adim, 1988, VI: 2927; from Ibn ‘Asakir, 1996, XV: 161. On the events of 1128, see Ibn al-‘Adim, 1968, II: 242 for the date of Zengi’s arrival at Aleppo, and Ibn al-Qalanisi, 1983, 348 for the date of Tughtegin’s death.   9. There is another small chronological problem with this account of his alleged ‘vassalage’ among the Franks, namely just who the Frankish lord of Aleppo was when Hamdan gained his new village late in the year 521/1127. Ibn al‘Adim names him as Manuel, the brother of the sister of the lord of Antioch, that is, the nephew of Bohemond II. But the Frankish Lord of al-Atharib in 1123, Muslim and Christian sources agree, was one Alan, not Manuel, and we know little about him, or if he was the nephew of Bohemond II. It is not known how long this Alan remained as lord of al-Atharib after 1123, except that he held it until his death; perhaps this Manuel, if he indeed ever existed, came after him. Or perhaps these details are hopelessly garbled. I’m grateful to

20  |  syri a i n crusa d e r time s Tom Asbridge for working with me to untangle these details, fruitless though it was. 10. Cahen, 1940, 41–42; Eddé, 1994, 293–308; cf. al-Sakhawi cited in Rosenthal, 1968, 62 and at 466; see also Kedar, 1990, 156–157; Hillenbrand, 2000, 32 and 358. 12. Ibn al-‘Adim, 1988, VI: 2926–2927. Quoted later by al-Sakhawi. 13. Ibn al-‘Adim, 1988, VI: 2926. 14. Many thanks to our colleague Christopher MacEvitt for calling this Syriac version to my attention. See Michael the Syrian, 1910, IV: 586/III: 183, discussed in MacEvitt, 2014, 1–16. 15. See also my discussion of this tale in Cobb, 2014, 271–274.

2 Legitimate Authority in the Kitab al-Jihad of ‘Ali b. Tahir al-Sulami Kenneth A. Goudie

M

odern scholarship on jihad in the Crusading period has tended to focus on the practice of jihad rather than the ideology of jihad. This is, in part, due to the nature of the surviving source material: where there is considerable evidence for the practice of jihad, there is a paucity of sources which reveal what twelfth-century figures meant when they referred to and invoked jihad. The sources which allow us to document and reconstruct the practice of jihad – reports of military victories in chronicles, the poetry and monuments which celebrate these victories, and so on – are ill-suited to helping us recover contemporary understandings of the ideology of jihad. We are not, however, entirely without sources; one such source is the Kitab al-jihad of ‘Ali b. Tahir al-Sulami (d. 1106).1 Composed in a series of majalis over the course of 1105, the Kitab aljihad is one of the earliest surviving Muslim responses to the First Crusade; at its simplest it is a call to jihad which seeks to exhort the residents of al-Sham – in particular Damascus – to unite and fight against the Crusaders. Al-Sulami did not, however, intend his work to be merely an exhortation: in the prefatory descriptions to each part, he states that the work concerns also siyar (the literary genre which focuses on the establishment of rules for proper conduct in war), fada’il al-Sham (the merits of al-Sham), and recent events.2 In short, al-Sulami envisaged his work to be a comprehensive guide to the performance of jihad, from first motivation through to proper conduct in war, all against the background of the historical and political developments which necessitated its performance. Not all of this material survives: the Kitab al-jihad is preserved in a single fragmentary manuscript, which comprises the

22  |  syri a i n crusa d e r time s second, eighth, ninth and twelfth parts of the work.3 The manuscript gives no indication whether the twelfth part was also the final part. The second part of the work has found particular favour amongst modern scholars: when the work was first discussed, Emmanuel Sivan chose to focus only on the second part, which he partially edited and translated into French. His decision is not hard to understand: the bulk of the second part is devoted to exhortation, whilst the rest of the work focuses primarily on siyar. In the broader context, then, the second part is the more strikingly novel. Sivan identified a number of key elements in it. Firstly, it reveals that alSulami considered the First Crusade divine punishment for the failings of the Muslims; secondly, that he considered it part of a pan-Mediterranean offensive by Christians; and thirdly, that it was religiously motivated, as evidenced by his use of the term ‘jihad’ to describe the activities of the Christians.4 Additionally, it is in the second part that al-Sulami proposes a solution: the Muslims must halt their spiritual decline through moral re-armament, and then regroup their forces to launch a counter-offensive.5 The last fifteen years have witnessed an increase in the amount of attention paid to the Kitab al-jihad. On the one hand, attention is paid to al-Sulami’s understanding of the origins of crusading, particularly how it can be explained and what it means for the development of crusading more broadly.6 On the other hand, al-Sulami’s hortatory techniques have been examined: the stress al-Sulami places on the importance of the duty of jihad, the fearful consequences which he states follow its neglect, and in particular the importance of the reconquest of Jerusalem as a step towards the conquest of Constantinople.7 There remains, however, one major lacuna in modern scholarship: how did al-Sulami actually understand jihad? In general, he is regarded as espousing the ‘standard jihad doctrine of the period’, at which point attention turns to the more attractive issues of his understanding of crusading or his hortatory techniques.8 Such a superficial reading of the central concern of the Kitab al-jihad is hardly a desideratum, and it is this issue which will be addressed in this chapter. Given limitations of space, attention will be paid primarily to the issue of legitimate authority in the Kitab al-jihad: that is, according to al-Sulami, who could legitimately call for and lead jihad? Before doing so, however, it is necessary to explore in more detail what is meant by ‘standard jihad doctrine’.

l e git im ate a uthori ty i n the kitab al-jihad  | 23 Jihad at the Turn of the Twelfth Century Insofar as we can describe any jihad doctrine as ‘standard’, by the twelfth century there were a number of elements within the broader tradition of juristic thought on jihad which had gained significant widespread acceptance such that it is possible to describe a juristic model of jihad. This model neither refers to martyrdom nor invokes the language of ascetical piety, but rather devotes considerable attention to siyar. It considers jihad an aggressive act whose purpose was to preserve the dar al-islam through its (theoretical) extension to encompass the entire world. Central to this is the subordination of jihad to political authority, as vested in the figure of the caliph. Jihad is defined not as a fard ‘ayn (an obligation of the individual), but as a fard kifaya (an obligation of sufficiency) wherein the obligation to perform it is considered lifted from the Muslim community at large when there were sufficient numbers to ensure success: the responsibility for mandating sufficient numbers lay with the caliph. The reformulation of jihad as a fard kifaya owes itself to al-Shafi‘i, and was accepted by the four Sunni madhahib.9 Majid Khadduri makes the invaluable observation that this transition from fard ‘ayn to fard kifaya ensured that jihad would not be included amongst the pillars of Islam, and that consequently its being enjoined not on the individual but the community would curtail jihad and theoretically strengthen caliphal authority over it.10 Two mid-eleventh-century works, both of which are entitled al-Ahkam al-sultaniyya, indicate how the juristic model had crystallised before the arrival of the First Crusade. These works were compiled by the Shafi‘ite Abu al-Hasan ‘Ali b. Muhammad al-Mawardi (d. 1058) and the Hanbalite Abu Ya‘la Muhammad b. al-Husayn b. Muhammad b. Khalaf b. Ahmad b. alFarra’ (d. 1066) respectively.11 Although the works are markedly similar – at some points even identical – al-Mawardi’s work has received the most attention, and underpins the concept of the ‘classical theory of the caliphate’.12 The similarity between the two works has led to much speculation about their relative chronology, but what is pertinent here is that the two works emerged almost simultaneously and reached identical conclusions, despite being compiled by authors who held diametrically opposed views on government service: where al‑Mawardi accepted it, Abu Ya‘la displayed the typical

24  |  syri a i n crusa d e r time s reluctance of the ‘ulama’.13 Even if one copied from the other, this nevertheless points towards the ubiquity of the ideas contained in their works. Both al-Mawardi and Abu Ya‘la focus on the proper delegation of caliphal authority, in which context they discuss jihad. Both works open with the appointment of the caliph (who is referred to as the imam. After discussing the various physical and personal requirements, al-Mawardi and Abu Ya‘la outline the ten public duties which are incumbent upon the caliph: these range from administrative (appointing counsellors, collecting taxes, making legal judgements, and so on) to religious (upholding the tenets of Islam, debating with dissenters, and so on). The sixth duty pertains to jihad: ‘Sixth, jihad against those who deny Islam after being invited to embrace it until they convert or enter into protection, so that the truth of God Almighty prevails over all religion.’14 In this way, al-Mawardi and Abu Ya‘la affirm the subordination of jihad to caliphal authority. At the same time, they conform to the juristic model’s understanding of jihad as inherently aggressive: the defence of Muslim lands is covered by the third and fifth duties of the imam: Third, protecting the territory and defending the sacred precinct so that the people can move freely in life and spread out in travel safe from peril to their life or property … Fifth, the strengthening of the thughur with deterring equipment and defensive strength until the enemy cannot triumph over what is unguarded, violate what is sacred or shed Muslim or mu‘ahid blood.15

Crucially, neither al-Mawardi nor Abu Ya‘la use any derivations of the root j‑h‑d to denote defence of Muslim territory. The caliph’s authority over jihad is reiterated when al-Mawardi and Abu Ya‘la move to discuss the duties of the amir: they both state that if the amir is given command of a border province, he must wage jihad against the neighbouring infidel.16 They specify, however, that if the country of this amir is adjacent to a border, then he is not allowed to start the jihad of his people except with the permission of the caliph. He needs to wage war against them and repel them without permission if they attack because repelling them is part of the laws of protecting and in accordance with defending the sacred precincts.17

l e git im ate a uthori ty i n the kitab al-jihad  | 25 How the amir should wage jihad is discussed in its own section, which focuses more on the proper conduct in war and the organisation of the army than it does jihad itself.18 These works therefore espouse the juristic model of jihad in its normative form: not only do they frame jihad as an element of statecraft, the declaration of which is the sole prerogative of the caliph, they further reduce the religiosity of jihad by focusing primarily on the establishment of rules for conduct in war instead of traditions pertaining to the merit of performing jihad and achieving martyrdom. Legitimate Authority in the Kitab al-Jihad It is hardly surprising that modern scholarship has concluded that al-Sulami espoused the standard jihad doctrine. Although the state of the manuscript does make it is impossible to reconstruct al-Sulami’s understanding of jihad in its entirety, the surviving fragments do suggest that he was heavily influenced by the juristic model: this is apparent not only due to the quantity of siyar within his work, but also by his quotation of legal opinions from adherents of the four Sunni madhahib. He does, however, display a marked preference for the opinions of Shafi‘ites, the bulk of which are attributed to al-Shafi‘i himself, as mediated through the work of al-Shafi‘i’s student, al-Muzani (d. 878). Al-Muzani’s Mukhtasar, an abridgement of al-Shafi‘i’s Kitab al-umm, is al-Sulami’s most important source for legal opinions: alSulami quotes from it verbatim at numerous points. Nevertheless, whilst it is clear that the juristic model of jihad underpins al-Sulami’s understanding of jihad, it is equally clear that he also found parts of the model problematic. Not only did al-Sulami find the Muslim desire to reconquer the lost lands incompatible with the juristic theory of jihad as offensive warfare, he was writing at a time when al-Sham was divided between various competing powers, both internal and external to the region: there was a lack of powerful or motivated leadership, be it caliphal or otherwise, under whose authority jihad against the Crusaders could operate. Al‑Sulami thus found himself writing in a situation which militated against one of the key tenets of the juristic model of jihad: the subordination of jihad to caliphal authority. To assuage the problem of caliphal disinterest, al-Sulami constructs an alternative basis of legitimate authority by focusing on when jihad becomes a fard ‘ayn and stresses the importance of Islamic unity.

26  |  syri a i n crusa d e r time s Al-Sulami opens the second part of his Kitab al-jihad by arguing that after the death of the Prophet, the rashidun caliphs and companions agreed that jihad was obligatory for all, and that: None of them abandoned it during his caliphate, and they were followed in that by those who were appointed successors and ruled in their own days, one after another; the sultan carried out the ghazw each year, either himself or by appointing a deputy to carry it out. It continued in this way until the time when one of the caliphs abandoned it because of his weakness and the inadequacy of his hand for it. Then others followed him in that, because of the above‑mentioned or similar reasons.19

Al-Sulami does not indicate which caliph he believes first abandoned jihad, but nevertheless argues that by doing so the Muslim community had fallen into error: this ‘made it necessary for God to disperse their unity, fracture their ascendancy, cast amongst them animosity and hatred, and tempt their enemies to wrest their country from their grasp, thereby healing their hearts from their faults’.20 For al-Sulami, the lack of leadership and the lack of Muslim unity are two sides of the same coin, and both were caused by failure to perform jihad. Yet whilst al-Sulami considers proper authority essential to jihad, he presents the ruler as a figure who has essentially abdicated any right to authorise or command jihad by virtue of their failure to do precisely that: The most astonishing is the sultan who takes pleasure in life or remains where he is, despite the ruins of this calamity, of which the consequence is conquest by these kuffar, expulsion from the country by subjugation and force, or staying with them in shame and servility, with the consequent daily and nightly killing, captivity, abuse and torture.21

He goes further, quoting the Prophet: ‘Whosoever guards a flock and yet does not attend them with sincere advice, God has forbidden him the Garden.’22 Al-Sulami explains that ‘attending them with sincere advice’ means watching over them, protecting them and driving their enemies from them.23 Even al-Sulami’s plea to the imam to unite with the sultans of al-Sham, the Jazira, and Egypt – setting aside their differences – serves only to underline the failure of the ruling elites to do precisely that.24

l e git im ate a uthori ty i n the kitab al-jihad  | 27 To deal with this, al-Sulami focuses on when jihad is a fard kifaya and when it is a fard ‘ayn. Al-Sulami quotes a number of legal opinions from all but the Hanafi madhhab to reinforce the same basic message that certain contexts obligate the individual to perform jihad. According to al-Sulami, whilst jihad was a fard kifaya in most situations, the Crusaders posed a threat of such magnitude that the obligation of jihad had devolved upon the individual: it was a fard ‘ayn. Al-Sulami’s justification of this rests upon the juxtaposition of two legal opinions from the Mukhtasar. These opinions are as follows: Al-Shafi‘i said: The least of which is to the imam is that no year passes without his organising a ghazw by himself or by his raiding parties, in accordance with the consideration of the Muslims, so that the jihad is not stopped in any year without a valid excuse. He said: ‘If enough do not depart with the raiding party, those left behind must go and consider necessary that which God, be He praised, said.’25

These legal opinions are found verbatim in the Mukhtasar but are juxtaposed by al-Sulami: they occur in separate sections in the Mukhtasar, though both are drawn from the chapter concerning siyar, and then only from a small section of it.26 In keeping with the Kitab al‑umm and the juristic model more generally, al-Muzani does not afford jihad its own distinct chapter. Al-Sulami uses the second of these legal opinions to qualify the first: whilst he does hold that the ruler has the duty of organising jihad and launching annual raids into the lands of the unbeliever, if the ruler fails to do this or if too few depart on the raid, those remaining behind must go out to perform jihad. Essentially, al-Sulami uses the second opinion to free the Muslim community from obedience to their rulers when it comes to successfully performing jihad. He argues that: It has become evident from what he mentioned that, if a group were needed to carry out a ghazw, the duty was incumbent upon all of them. That was concerning a similar situation to that in which we now are, with this band attacking the land of Islam.27

28  |  syri a i n crusa d e r time s The additional legal opinions which al-Sulami cites are all deployed to reinforce this conclusion. Thus for example, Ibn Hanbal is quoted as saying: ‘The obligation is on the people that if the enemy comes, the poor and the rich amongst them go forth to fight them,’28 and that ‘[t]hey do not go out to the enemy except with the permission of the amir, unless the enemy surprises them, all of them are afraid, and it is impossible for them to seek permission.’29 In the middle of these legal opinions is a clear explanation of al-Sulami’s reasoning, albeit mediated through the words of another. Essentially, alSulami stresses that jihad is only a fard kifaya when there are sufficient numbers to ensure success; if not, then the duty of jihad devolves upon neighbouring people. Al-Sham is given as an example: if the enemy were to make for one of its cities, and there were not within it enough people to fight and repel them, it is incumbent upon the rest of the cities of al-Sham to send groups until there were sufficient numbers. At that point, the duty falls from the others because the lands of al-Sham are as one city. If the capable amongst them hasten to the enemy, and they are insufficient, it is incumbent furthermore upon all those who are close to al-Sham to send groups to them and join them until there are sufficient numbers. At that point, the duty would also fall from the others.30 For al-Sulami, then, the jihad against the Crusaders is in the first instance a fard ‘ayn before transitioning into a fard kifaya: it is incumbent upon the individual to go out to it until there are sufficient numbers to ensure success, at which point it becomes a fard kifaya. Because it begins as a fard ‘ayn, the initiative lies solely with the individual and not with the ruler. He supports his contention that the ruler can be ignored by noting that the prosecution of jihad was an obligation for the ruler, and that that which was incumbent upon him was likewise incumbent upon the Muslim community.31 At the same time, al-Sulami focuses on the importance of unity: he conceives jihad as a duty which devolves on all Muslims in the vicinity, who are obligated to struggle against the invading enemy. According to al-Sulami, it was the lack of such unity which encouraged and facilitated the Christian conquest.32 He calls upon the imam to unite with the sultans of al-Sham, the Jazira and Egypt, citing in support of this the examples of the Persian kings and the pre-Islamic Bedouin. Al-Sulami notes that both would set

l e git im ate a uthori ty i n the kitab al-jihad  | 29 aside their differences and unite to fight against common enemies; if they were victorious they would either part ways or stay united, with the latter being preferred.33 Al-Sulami argues that this is incumbent upon the current ruling elite, who must follow the example of their predecessors and act in accordance with the prophetic command: ‘Do not snub each other; do not turn your back on each other; do not be envious each other. Be servants of God as a brotherhood, as God (be He exalted) commanded you.’34 He then continues that it is incumbent upon everyone, no matter their station in life, to support those with authority over them in order to ensure the success of the jihad. Al-Sulami considers the unity of the Muslim community essential to the successful prosecution of jihad, and whilst he considers it preferable that jihad be directed by a ruler, it can be undertaken without their permission because the severity of the situation has rendered it an individual obligation. The latter is at odds with the juristic formulation of jihad as a fard kifaya, which was – as mentioned above – described by Khadduri as a means whereby jihad could be brought under caliphal control. Asserting that jihad was enjoined not on the individual but the community at large and subject to caliphal authority was one way in which early ‘Abbasid jurists could check the frontier culture of ascetic warriors and, perhaps more importantly, the Kharijites. The latter’s unbridled use of violence against the Muslim community had brought into sharp relief the danger of leaving the impulse to perform jihad outside of the control of both the jurists and the caliphs. Al-Sulami’s reassertion of the role of the individual, whilst perhaps not a direct refutation of the juristic model’s positioning of the caliph as the sole legitimate authority, certainly recalls this competition for control over the conduct of jihad in the early ‘Abbasid period, between what Paul L. Heck calls state and non-state actors.35 Thus for example, we find in the eighthcentury Kitab al-jihad of Ibn al-Mubarak (d. 797) an understanding of jihad which presents jihad as the means by which one demonstrates devotion to God, and which culminates with martyrdom: martyrdom was the confirmation of one’s righteousness before God.36 It was very much a personal endeavour, and worldly authority was de‑emphasised: with the exception of the Prophet, the only ruler named in Ibn al-Mubarak’s Kitab al‑jihad was mocked for cowardice.37 For Ibn al-Mubarak, then, there is no place for

30  |  syri a i n crusa d e r time s worldly authority within the practice of jihad, for – axiomatically – only God can determine if one is performing jihad. It is tempting to see some sort of continuity between Ibn al-Mubarak and al-Sulami, some sort of popular understanding of jihad which existed outwith caliphal authority and which survived the intervening centuries. Yet the fact that al-Sulami operates from an assumption that the caliph is the sole holder of legitimate authority and goes to such lengths to justify and define an alternative source of legitimate authority suggests the opposite. Furthermore, there is little evidence in al-Sulami’s Kitab al-jihad that the eighth-century understanding of jihad expressed by Ibn al-Mubarak exerted any real influence on al-Sulami: al-Sulami demonstrates no awareness of Ibn al-Mubarak’s Kitab al-jihad, and whilst there are five traditions which occur in both works, Ibn al-Mubarak himself appears in only a single isnad.38 The lack of textual overlap does not of course mean that al-Sulami was uninfluenced by Ibn al-Mubarak. Niall Christie argues that al-Sulami drew upon the legacy of eighth- and ninth-century scholars who were involved in jihad on the Muslim–Byzantine frontier to galvanise his audience. These scholars were Makhul al-Dimashqi, Ibn al-Mubarak, al-Fazari, Hajjaj b. Muḥammad al‑Massisi (d. c.730–7) and Sunayd b. Dawud al‑Massisi al‑Tarsusi (d. 821). Christie suggests that al-Sulami held up these figures as examples whom his audience should follow.39 The evidence from the Kitab al‑jihad, however, does not support this. Only Makhul al-Dimashqi is held up as an example, and even then al-Sulami places the emphasis on his credentials as one of the tabi‘un rather than as a mujahid.40 The others appear only as names in isnads, and none more than four times. In any case, Ibn al-Mubarak’s reputation as a mujahid only became emphasised in the biographical literature with Ibn ‘Asakir’s Ta’rikh madinat Dimashq, and so it is unlikely that a scholar otherwise famed as an ascetic would have been useful for al-Sulami.41 This is not to argue against any continuous Syrian discourse of jihad, merely to caution that the similarity between al-Sulami and Ibn al-Mubarak regarding legitimate authority was not due to any direct continuity between the two writers, but rather due to their similar political circumstances. Ibn al-Mubarak compiled his text in al-Sham while it was still outwith the firm control of the ‘Abbasids, and al-Sulami wrote his text in al-Sham whilst

l e git im ate a uthori ty i n the kitab al-jihad  | 31 it was similarly outwith the control of any single authority. Their attitude towards the issue of legitimate authority could not, however, have been more different. Where Ibn al-Mubarak considered the ruler and his authorisation inessential, the absence of legitimate authority looms large over al-Sulami’s understanding of jihad. This forced him, through creative legal argument, to construct an alternative source of legitimate authority by arguing that in certain contexts the individual’s responsibility to the community at large supersedes any caliphal authority. Notes  1. Al-Sulami’s Kitab al-jihad has been edited three times. Emmanuel Sivan prepared a partial edition, which was published in 1966 alongside a French translation; see Sivan, 1966, 197–224. The first full edition was prepared by Suhayl Zakkar and published in 2007. See Arba‘at kutub fi al‑jihad, 2007, 43–165. Niall Christie has recently prepared a new edition, which was published alongside an English translation; see Christie, 2015, 39–197. Unless otherwise stated, all references are to Christie’s edition, which is preferred because he provides a fuller scholarly apparatus and indicates the format and appearance of the manuscript. All translations of al-Sulami’s Kitab al-jihad are, however, my own.   2. Christie, 2015, 39, 77, 117, 149.  3. The manuscript, for which there is no cataloguing information available, is held in the Asad National Library in Damascus; previously, it was held in the Zahiriyya Library in Damascus under the class marks 3769 and 4511.   4. Sivan, 1966, 199–201.   5. Ibid., 201–203.  6. See in particular Chevedden, 2013, 27–53; and Christie and Yazigi, 2006, 57–72. See also Chevedden, 2006, 90–136; Chevedden, 2008, 181–200; Chevedden, 2010, 191–225; and Christie and Gerish, 2003, 139–148.   7. See in particular Christie, 2007b, 1–14; Christie, 2007a, 209–221; Christie and Gerish, 2003, 139–148.   8. Christie, 2015, 12.   9. See in particular al-Shafi‘i, 1987, 83–85. 10. Khadduri, 1955, 59–62. 11. Al‑Mawardi, 1989; Abu Ya‘la, 2000. 12. The term was coined by H. A. R. Gibb. See Gibb, 1937, 291–302; and Gibb, 1939, 401–410. Claude Cahen was the first to state that Abu Ya‘la copied

32  |  syri a i n crusa d e r time s al‑Mawardi’s work verbatim save for replacing references to Abu Hanifa, Abu Yusuf and al‑Shafi‘i with quotations from Ibn Hanbal. See Cahen, 1955, 151. Cahen’s assumption has generally been accepted as fact, with scholars such as Hanna Mikhail stating that Abu Ya‘la was prompted to write his book to make up for al‑Mawardi’s neglect of the Hanbalites. See Mikhail, 1995, 59. As both Donald P. Little and Eric J. Hanne have noted, however, the general conclusion rests primarily on the fact that Abu Ya‘la was the younger of the two, and therefore must have copied his older contemporary. See Little, 1974, 6–8; Hanne, 2004, 55–57. 13. Hanne, 2004, 57. 14. Al‑Mawardi, 1989, 22–23; Abu Ya‘la, 2000, 27–28. 15. Al‑Mawardi, 1989, 22–23; Abu Ya‘la, 2000, 27–28. 16. Al‑Mawardi, 1989, 40–41; Abu Ya‘la, 2000, 34. 17. Al‑Mawardi, 1989, 44; Abu Ya‘la, 2000, 37. 18. Al‑Mawardi, 1989, 47–73; Abu Ya‘la, 2000, 39–51. 19. Al-Sulami, Kitab al‑jihad, 42. 20. Ibid., 42. 21. Ibid., 48. 22. Ibid., 71. 23. Ibid., 71. 24. Ibid., 71. 25. Ibid., 43. 26. Al-Muzani, 1998, 353. 27. Al-Sulami, Kitab al-jihad, 43. 28. Ibid., 46. 29. Ibid., 46. 30. Ibid., 44–45. 31. Ibid., 47. 32. Ibid., 42–43. 33. Ibid., 71–72. 34. Ibid., 72. 35. Heck, 2004, 109. 36. Goudie, 2016, 50–69. 37. Ibn al-Mubarak, Kitab al-jihad, n.d., 128, no. 131. 38. Al-Sulami, Kitab al‑jihad, 166–168; cf. Ibn al-Mubarak, Kitab al-jihad, ­140–142, no. 145. Of the other four traditions, two assert that jihad is the most meritorious act (al-Sulami, Kitab al‑jihad, 118–119 and 121–122), whilst the

l e git im ate a uthori ty i n the kitab al-jihad  | 33 other two concern the Qur’anic reduction in the number of the enemy each Muslim must face (al-Sulami, Kitab al‑jihad, 86–87 and 152). In connection with these traditions, however, no reference is made to Ibn al‑Mubarak or his Kitab al‑jihad. 39. Christie, 2015, 25. 40. Al-Sulami, Kitab al‑jihad, 41. 41. Goudie, 2016, 177–204.

3 Politics, Religion and the Occult in the Works of Kamal al-Din Ibn Talha, a Vizier, ‘Alim and Author in Thirteenth-century Syria A. C. S. Peacock1

U

nder the year 652 /1254, several medieval Arabic chronicles mention the death of a certain Kamal al-Din Ibn Talha, the former khatib (chief mosque preacher) of Damascus, who is uniformly described as virtuous, learned and ascetic. Ibn Talha’s claim to fame was that he was appointed vizier by the Ayyubid ruler al-Nasir Yusuf in 648/1250, but resigned the post after two days, instead embracing a life of asceticism; according to some accounts, even rejecting the position after the document of appointment had been formally drawn up.2 We are given, then, an idealised picture of an ‘alim, who if briefly tempted by the vanities of this world soon rejected them. This picture is, however, undermined by references elsewhere in the chronicles to Ibn Talha’s earlier involvement in Ayyubid politics, serving as an emissary in the 636/1238 peace negotiations between the Ayyubid princes al-Jawad and al-Salih.3 Moreover, the notices in al-Safadi and al-Dhahabi conclude with an intriguing and critical comment: ‘He entered into perdition and error and made a circle of letters, claiming he could interpret knowledge of the unknown and of the final hour.’4 This alludes to the ‘science of letters’ (‘ilm al-huruf ), the proponents of which believed that the letters of the Arabic script and the Qur’an encoded occult knowledge in their numerical values.5 The most detailed account of Ibn Talha’s career comes in Ibn ‘Imad’s Shadharat al-dhahab, and refers to the same allegations, which, however, Ibn ‘Imad seeks to defuse. He calls Ibn Talha ‘one of the great preeminent men and leaders’ (ahad al-sudur wa’l-ru’asa’ al-mu‘azzamin), and reports that

po l iti cs, reli g i on and the occul t  | 35 he was born in 582/1186, and studied hadith in Nishapur with the famous transmitters al-Mu’ayyad and Zaynab al-Sha‘riyya.6 He became a faqih, ‘and excelled at jurisprudence, usul and disputation’. Ibn Talha, he says, also wrote diplomatic correspondence for kings (tarassala ‘an al-muluk), and served as qadi in Nusaybin, his home town, before becoming khatib of Damascus. Ibn ‘Imad mentions Ibn Talha’s famous two-day vizierate, and also reports that ‘he is attributed with a preoccupation with the science of letters (‘ilm al-huruf    ) and [future] times, and predicting things from the unknown (al-mughibat). But it is said that he repented.’ Some verses attributed to Ibn Talha are quoted in this connection: Do not give yourself to the talk of the astrologer, all affairs belong to God, that is it. Know that if you ascribe to a star control (tadbir) of an event you are not a Muslim.7

Ibn ‘Imad is also one of the few chroniclers to refer to Ibn Talha’s literary works, mentioning two of his books, al-‘Iqd al-farid and al-Durr al-­ muntazam; the latter contained the notorious circle of letters. Ibn Talha was also the author of several other books, not all of which have survived.8 However, al-‘Iqd al-farid and al-Durr al-muntazam, a ‘Mirror for Princes’ and a work on letters respectively, enjoyed considerable popularity and influence down into Ottoman times and arguably even beyond, and his Matalib al-su’ul fi manaqib Al al-Rasul, a work on the Twelve Imams, was also widely read. None of these been the subject of any detailed study in the West of which I am aware, with the exception of a useful chapter in the unpublished PhD dissertation by Masad which does, however, need correcting in several respects and which concentrates on his lettristic works.9 Although editions of sorts exist of his three major works, these are sufficiently poor to make recourse to the manuscripts absolutely imperative, and the textual history of all these works is complex, all having been renamed and rewritten at various stages. These reasons would be enough to merit a study, but Ibn Talha is also of particular interest as his literary oeuvre gives an insight into some crucial but neglected currents in the intellectual history of the thirteenth-century Islamic world. His works may at first seem to comprise extremely diverse if not wholly separate fields of endeavour; in fact they share more than immediately meets the eye. Lettrism, prayer, asceticism and

36  |  syri a i n crusa d e r time s contemporary political concerns all combined in the works of Ibn Talha, which we will survey here. Kitab al-jafr/al-Durr al-muntazam fi’l-sirr al-a‘zam/al-Durr al-munazzam The most famous of Ibn Talha’s works, known under several variant titles which we will refer to here simply as al-Durr al-muntazam, is a short treatise and is one of the earliest surviving works on the science of jafr – divination from the numerical value of letters. It was incorporated wholesale into another famous treatise by the Ottoman specialist in the science of letters ‘Abd alRahman al-Bistami (d. 858/1454), the Miftah al-jafr al-jami‘, together with which it is usually found in manuscripts, meaning there is a distinct lack of early versions of the text.10 It was also included in al-Bistami’s Shams al-afaq.11 Ibn Talha’s famous da’ira, or circle (Plate 3.1), which showed how to calculate the value of letters and thus prognosticate the future, was adopted by numerous other writers on jafr, and seems to be found in the Shams al-ma‘arif al-kubra, the most famous Islamic magical text attributed to his contemporary al-Buni (d. c.622/1225), which may, however, be a composite work of much later date or dates.12 While similar circles purporting to demonstrate the power of letters are mentioned or depicted in earlier magical treatises, sometimes attributed to Hermes Trismegistos, the version in al-Durr al-muntazam was perhaps much more broadly influential for being more thoroughly Islamic in inspiration, claiming ‘Ali b. Abi Talib as its source.13 Popular editions of Ibn Talha’s work still circulate in the Middle East, attributed to ‘Ali b. Abi Talib.14 The work starts with a description of how an unnamed friend of Ibn Talha, described as ‘one of the righteous’ (ba‘d al-salihin) was at his devotions, when a tablet was revealed to him. A voice told him to ‘Take it and make use of it’, and he found that it was a circle with lines and letters. He saw ‘Ali b. Abi Talib, who reveals that he is the one who has given him this tablet and tells him things that he does not understand. Ibn Talha’s friend tells ‘Ali he does not understand it, to which ‘Ali replies that Ibn Talha will explain it. And when it was completely daylight he came to me and told me about this event in full and recited to me the verses of its sura, and drew a picture of

po l iti cs, reli g i on and the occul t  | 37 the circle and what there was outside and inside and within it. And I studied it and contemplated it and saw that its writing comprised remarkable fates, and its root and branches comprised strange secrets, and I looked into its letters composed both singly and in pairs, and its names both individually and together, and I realised that it was not possible to understand its full meaning nor to solve it nor to understand its aim and the secrets of its purpose except by divine aid.15

Ibn Talha therefore prayed and received the necessary divine assistance. After this introduction, Ibn Talha moves on to the famous circle itself, which he explains covers time from the prophethood of Muhammad to the last day. The next section contains examples of calculations from the letters to generate the dates of historical events and prophecies for the future. The last concrete event mentioned is the Ayyubid defeat of the Khwarazmians at Qasab in 644/1246, on which Ibn Talha remarks, ‘It is understood from the symbols of the circle, and God knows best, that their state shall not return to them, nor shall fortune.’16 Whereas the earlier calculations were for historical events such as the beginning of the Seljuq state or the end of Frankish rule over Jerusalem, after the mention of the Khwarazmian defeat at Qasab, the other prophecies which continue up to ah 71817 all seem very vague, without any specifics. Masad is thus probably right to argue that this shows the work was composed very shortly after Qasab, which he argues had an apocalyptic significance to contemporaries and Ibn Talha in particular.18 After 718, various vague prophecies of bloodshed and destruction suggest the end of time and calculations for establishing the date of the emergence of the Mahdi and the Antichrist (Dajjal), and the Day of Resurrection, are hinted at, although no date is given. Finally, the work concludes with some instructions on how to make the necessary calculations for jafr, although these are extremely brief and allusive. Despite Ibn Talha’s claim to have explained the circle given to him by ‘Ali, it must be said that al-Durr al-muntazam, at least in the forms that have come down to us, is extremely allusive. It is hard to see it as a practical do-it-yourself manual on how to calculate jafr. Indeed, Bistami’s Miftah al-jafr al-jami‘, with which it survives in most manuscripts, is intended to explain and expound Ibn Talha’s work and is many times its length – al-Durr al-muntazam taking up a

38  |  syri a i n crusa d e r time s mere twelve pages in the printed edition to just over 200 for Bistami’s work. In a sense, al-Durr al-muntazam seems to aim to defuse claims of the apocalypse rather than promoting them. While the reign of the Dajjal is certainly to come, it will not occur until after 718: imminent enough, but not enough to worry any Ayyubid reader during his lifetime. Further light on the text is shed by al-Safadi and al-Yunini, who record the identity of Ibn Talha’s pious friend. They write of the well-known Sufi ascetic Sharaf al-Din al-Ikhmimi, another resident of Damascus, who related hadith on the authority of Ibn Talha that, He is the one mentioned by Kamal al-Din Ibn Talha in his composition on the science of letters. He [Ibn Talha] said, ‘Shaykh Muhammad [that is, al-Ikhmimi] saw ‘Ali b. Abi Talib in a dream, and the latter showed him the circle of letters.’19

Al-Ikhmimi, who was noted for his miracles (karamat) and claimed that the hidden essence of prayers had been revealed to him, said, ‘God has revealed to me the [hidden] truths of the prayers of things, so that I have seen the trees and stones with their different prayers.’20 This passage underlines the strong connection between asceticism and occult interests which can be observed throughout Ibn Talha’s career and writings. Such an association should not surprise us. Al-Ghazali himself had been involved with the ‘ilm al-huruf,21 while it seems that in his own time al-Buni was best known as a Sufi ascetic, rather than a magician. Later lettristic writers like ‘Abd al-Rahman al-Bistami also sought to legitimise the science of letters by associating it with prominent Sufis.22 Matalib al-su’ul fi manaqib Al al-Rasul Ibn Talha’s claim in al-Durr al-muntazam to being in direct communication with ‘Ali b. Abi Talib, whose instructions he interprets for al-Ikhmimi, is repeated in his next work to be considered, the Matalib al-su’ul fi manaqib Al al-Rasul. The only one of Ibn Talha’s compositions to have been published in a remotely adequate edition (Beirut, 1419), the editor lists nineteen manuscripts and three previous editions (two lithographs, Tehran (1275) and Lucknow (1302), and a commercial edition from Najaf in 1371).23 Ibn Talha refers in his preface to having compiled an earlier version called the Zubdat

po l iti cs, reli g i on and the occul t  | 39 al-maqal, the manuscript of which he lost. Some library catalogues list the Zubdat al-maqal, but all surviving manuscripts appear to be of the later version, the Matalib al-su’ul. Although the work has no date of composition in the printed text used here or the manuscripts I have seen, a late source, Muhammad Baqir al-Khwansari’s (d. 1313/1895–6) Rawdat al-jannat, gives it as Rajab 650/September–October 1252.24 Presumably this information derives from a statement in a manuscript. The Matalib al-su’ul is dedicated to the Twelve Imams that succeeded Muhammad – ‘Ali b. Abi Talib, Hasan, Husayn, Zayn al-‘Abidin, Muhammad al-Baqir, Ja‘far al-Sadiq, Musa al-Kazim, ‘Ali al-Rida, Muhammad al-Qani‘, ‘Ali al-Mutawakkil, Hasan al-‘Askari (called here Hasan al-Khalis) and Muhammad al-Hujja al-Mahdi. Despite this ostensibly Shi‘ite topic, Ibn Talha is careful to rely exclusively on Sunni sources, in particular the classical hadith collections such as the Sahihs of Muslim and Bukhari, al-Tirmidhi and Abu Da’ud. The work is divided into twelve chapters (babs), one for each imam, most of which are then each divided into twelve sub-chapters (fasl) which treat each imam’s birth, genealogy, names, appearance, piety and asceticism, courage, eloquence, children, age and death. However, in the penultimate chapter dealing with the obscure Eleventh Imam, about whom even Twelver tradition had little to say, these fasls become abridged to a single sentence or so, and in the final chapter, to which we shall return shortly, the structure is abandoned entirely. In his introduction, Ibn Talha describes how ‘Ali has ‘mandated me to elucidate the occult secrets (asrar min al-ghayb), which God only allots to one of his chosen servants’.25 The occult meaning of letters is alluded in Ibn Talha’s discussion of the symbolism of the phrases la ilah illa allah and Muhammad rasul allah, which in the Arabic script are both composed of twelve letters, signifying the Twelve Imams. Similarly, Ibn Talha notes the significance of the fact that both day and night are composed of twelve hours each, and that twelve generations separate Muhammad from his ancestor Nadr, the founder of the tribe of Quraysh, just as twelve generations separate him from the Last Imam.26 It is perhaps not coincidental that the number twelve also commonly appears in the apocalyptic literature, given its association with the Twelfth Imam or Mahdi.27

40  |  syri a i n crusa d e r time s The chapters are also exceedingly unequal in their treatment of the Imams. The overwhelming bulk of the Matalib al-su’ul is devoted to ‘Ali b. Abi Talib,28 with much shorter but still substantial chapters on his sons Hasan and Husayn.29 The treatment of the other Imams is much more cursory. Ibn Talha tells us that he had been ordered to write the book by ‘Ali b. Abi Talib himself: One of the righteous (ba‘d al-salihin) saw ‘Ali [in a dream] and asked him questions connected with sacred knowledge (al-ma‘arif al-qudsiyya) and divinity. ‘Ali replied with words, and he said, ‘O Commander of the Faithful, I do not have the knowledge to understand them.’ So [‘Ali] handed over to me the task of explaining to him that, to give details of everything and to expound the details of his speech and sentences.30

This passage is reminiscent of that at the start of al-Durr al-muntazam, where Ibn Talha stakes a similar claim to be the interpreter of ‘Ali b. Talib’s words to ‘one of the righteous’; it is possible that here, too, the allusion is to the Sufi al-Ikhmimi. The extensive passages in the Matalib al-su’ul discussing ‘Ali highlight certain aspects that may have been particularly significant to Ibn Talha with his interests in the occult and asceticism. Much emphasis is given to ‘Ali’s promotion of asceticism (zuhd) and warnings against the vanity of this world.31 ‘Ali’s prominent role in Ibn Talha’s occult writings is also suggested by the passages which describe ‘Ali’s complete mastery of all forms of knowledge, in particular of the unity of God (‘ilm al-tawhid), of fate and predestination, of prophecy, of the Day of Judgement and of the nature of the afterlife.32 A hadith is cited in which ‘Ali says, ‘Ask me about the ways of the heavens, for I know more of them than I do of those of earth. If I wanted, I could load a mule with the interpretation of bismillah al-rahman al-rahim.’33 Nonetheless, this occult knowledge does not feature particularly prominently in most of the Matalib al-su’ul, which concentrates on the characteristics of the Twelve Imams. The Matalib al-su’ul generally avoids promoting a sectarian agenda, beyond its philo-‘Alidism. The great dispute at Siffin is mentioned, although not in great detail, but there is no particular effort to condemn the Umayyads. At the same time, the work’s broadly pro-Twelver agenda is introduced

po l iti cs, reli g i on and the occul t  | 41 ­ ithout polemics against, or indeed much reference to, any of the opponents w of ‘Ali or the ‘Alids, although the Kharijites are condemned.34 Despite treating the most sensitive questions in Islamic history, Ibn Talha seems intent on stripping these of their polemical potency. Of ‘Ali b. Abi Talib, for instance, we are told that he is the source of knowledge (marja‘ al-‘ilm) for Mu‘tazilites, Ash‘arites, the Shi‘a and Khawarij.35 Ash‘arism was most likely the theology to which Ibn Talha himself adhered, it being typically associated with the Shafi‘i madhhab to which he belonged. Thus, despite the differences of these groups, ‘Ali is portrayed as a common, unifying element, and no mention is made of other legal or theological schools. A similar approach can be seen in a hadith which Ibn ‘Adim reports Ibn Talha as relating. A group of Shi‘ites from Kufa complain to Hasan b. ‘Ali of their mistreatment by the Umayyad governor Ziyad b. Abihi, saying, ‘We ask God that He let us kill him with our own hands!’ Hasan is reported to have condemned the idea of killing him, saying instead, ‘Let us ask God to make him die in his bed.’36 In the final chapter, Ibn Talha turns to discuss the Twelfth Imam, whom, in conformity with at least one strand of Twelver thinking, he identifies as Hasan al-‘Askari’s son Muhammad, born in 258/871.37 Ibn Talha describes the Twelfth Imam as the Mahdi, and spends some time justifying this identification on the basis of hadith transmitted from Abu Da’ud and al-Tirmidhi. The Mahdi is described by the Prophet as ‘with a fairer forehead than me and a crooked nose, who will fill the world with justice just as it was [previously] filled with injustice and oppression and will rule for seven years’. This will happen at the end of time. However, as Ibn Talha says, also in accordance with Twelver tradition, that although the Twelfth Imam has been identified – for Muhammad b. Hasan is the only candidate who meets the description – he has gone into occultation. He is still alive, for God has prolonged his life. Ibn Talha is careful not to specify a date for the Mahdi’s return after which his reign of universal justice will be instituted.38 This is perhaps in keeping with Shi‘ite traditions that tended to be rather more cautious than Sunni ones about predicting the imminence of the apocalypse.39 With its use of Sunni sources to present a basically Twelver view of the Imams, Ibn Talha’s Matalib al-su’ul represents part of a more general intellectual tendency to the Shi‘itisation of Sunnism in the pre-Mongol period, of which Eddé has provided a number of examples from Aleppo.40 Nor was

42  |  syri a i n crusa d e r time s it unique. A comparable work written at roughly the same time is al-Ghanji’s al-Bayan fi akhbar Sahib al-Zaman, composed in 648/1250–1, who similarly identified the Mahdi with the Twelfth Imam, but also drew on Sunni sources to support an essentially Twelver approach.41 In Ayyubid Damascus, there were a number of prominent Shi‘ite-leaning Sunnis, such as the Shafi‘i qadi Ibn al-Zaki, who purported to be a descendant of the Caliph ‘Uthman, yet preferred ‘Ali to his own ancestor.42 It is interesting to note that Matalib al-su’ul may also have had some appeal to a Shi‘ite constituency, as a copy was found in the library of Ibn Tawus, Ibn Talha’s Shi‘ite contemporary from Baghdad.43 Ibn Tawus himself wrote an apocalyptic work, the Kitab almalahim wa’l-fitan, which seems to have drawn exclusively on Sunni sources despite the author’s Twelver affiliation.44 Thus, the Ayyubid period saw something of a collapse in the boundaries between Shi‘ite and Sunni, as can also be seen in Ayyubid patronage of Shi‘ite shrines. Al-Nasir Yusuf is said to have reconstructed a large part of the Mashhad al-Muhassin in Aleppo, a major shrine commemorating a child of the Imam al-Husayn in which one inscription mentions the name of al-Nasir’s father the Ayyubid sultan al-Malik al-‘Aziz Muhammad (r. 613/1213–634/1236) alongside the names of the Twelve Imams.45 Mulder has argued this patronage was inspired by a variety of factors: as such shrines were visited by Muslims of sectarian divisions, given a popular Sunni regard for the ahl al-bayt, they served both to ingratiate the Sunni Ayyubid rulers with their Shi‘ite subjects,46 as well as legitimising them as Sunni rulers by ‘bringing Shi‘i popular practice within the realms of official Sunni religious policy’.47 Given the lack of a dedicatee for the Matalib al-su’ul it is perhaps going a step too far to claim that the work might have been written specifically to support such a putative state policy, although such an interpretation has undeniable attractions. Although no patron of the Matalib al-su’ul is given, al-Nasir Yusuf was the dedicatee of another of Ibn Talha’s works, his Mirror for Princes entitled the Nafa’is al-‘anasir, discussed below, and one early copy of the Matalib al-su’ul seems to come from a royal or elite library, given its elaborate frontispiece (Plate 3.2). The Matalib shows that this ecumenical approach could also be supported and indeed promoted by the Sunni ‘ulama’ themselves, who sometimes are portrayed as the villains whose rigidity could result in intercommunal

po l iti cs, reli g i on and the occul t  | 43 v­ iolence.48 Rather, Twelver piety within the framework of Sunnism was not simply a matter of popular religiosity or of political necessity but part of a more general trend toward conciliation, which we can also see in the prominent role of Shi‘ite scholars like Ibn Tawus at the ‘Abbasid court in Baghdad, and which affected learned, scholarly forms of Islam as well as popular ones – if indeed such a differentiation has any validity.49 This Shi‘itisation of Sunnism is also reflected, albeit in different forms, through the Ismaili influences that are so evident in the works of Ibn ‘Arabi, in the general rise of apocalyptic thought in the thirteenth-century eastern Mediterranean, and in the importance of ‘Ali b. Abi Talib as a key figure in the ‘ilm al-huruf and other forms of occult knowledge that were becoming increasingly widespread in period.50 Nafa’is al-‘anasir/al-‘Iqd al-farid Ibn Talha’s political connections are demonstrated by his works entitled Nafa’is al-‘anasir li’l-Malik al-Nasir dedicated to the Ayyubid al-Nasir Yusuf and al-‘Iqd al-farid dedicated to the Artuqid ruler of Mardin, Najm al-Din Ghazi (r. 637/1239–658/1260). Despite their different titles, and dedicatees, the content of these two works is virtually word for word identical beyond the mention of the dedicatee in the preface. The only significant exception is a long passage celebrating al-Nasir Yusuf’s defeat of the Khwarazmians, which is included in the Nafa’is al-‘anasir,51 indicating that work must have been composed shortly after the Ayyubids’ defeat of the Khwarazmians under Berke Khan at al-Qasab near Homs in 644/1246.52 It is clear that this version dedicated to al-Nasir Yusuf predates al-‘Iqd al-farid dedicated to Najm al-Din Ghazi, for in the later parts of al-‘Iqd al-farid the name of al-Nasir is still found – for example on the table to calculate the first days of the months, discussed below. One must wonder what the Artuqid dedicatee would have made of it had he seen the names of his Ayyubid overlord so prominently displayed on a work purportedly dedicated to him, especially given that, shortly before, the Ayyubids had sought to extirpate the Artuqids after the latter had allied themselves with the Anatolian Seljuqs.53 It was as al-‘Iqd al-farid that the work became popular, surviving in numerous manuscripts, and it is also under this title that an unsatisfactory edition has been published in Cairo,54 whereas the Nafa’is al-‘anasir to my k­ nowledge

44  |  syri a i n crusa d e r time s is only extant in three manuscripts.55 Al-‘Iqd al-farid certainly enjoyed an appeal to later rulers; a fourteenth-century copy comes from an unidentified royal library (MS Süleymaniye Library, Yeni Cami 985, Plate 3.3), as did a fifteenth-century copy now held in the Landesbibliothek Gotha.56 Moreover, parts of the text of al-‘Iqd al-farid were incorporated into a ‘Mirror for Princes’ written for the Mamluk sultan Barquq (d. 801/1399), al-Durr al-nadid fi manaqib al-Malik al-Zahir Abi Sa‘id, offering further testimony to Ibn Talha’s influence.57 The Nafa’is al-anasir/al-‘Iqd al-farid also appeared under a third title as Miftah al-falah fi i‘tiqad ahl al-salah which is preserved in Cairo (Dar al-Kutub MS Tasawwuf M ‘Arabi 201, comprising 239 folios, copied in Ramadan 1114/January 1703 by Muhammad b. ‘Abdallah b. Muhammad b. Jum‘a al-Hindi). The Miftah al-falah differs from the other versions in its preface, which emphasises that the text was composed as a precis (zubda) of earlier works dealing with correct behaviour and belief and how to emulate the Companions of the Prophet and the righteous forebears (al-salaf al-sahih). From the microfilm I have seen, it appears that the first folio, which presumably contained details of the dedicatee, is missing, and there is no reference to the Khwarazmian defeat. In other respects, the work seems to contain the same text as the other two versions, right down to the tables including the name of al-Nasir Yusuf. The late date of this copy and its variant title and preface confirm the text’s lasting appeal in a variety of incarnations. At first sight the work appears to be a conventional ‘Mirror for Princes’, urging the ruler to justice. However, its contents are rather more complex than the standard ten-chapter format for the genre popularised by al-Ghazali. The work is divided into four qa‘idas, which are then further subdivided. The work is thus structured as follows: i. The first qa‘ida, on important morals and attributes, divided into ten babs: 1. On the intellect (‘aql) and the necessary belief in the unity of God (tawhid) and the required rituals (‘ibadat) which are based on it 2. In praise of patience and acting carefully, and in condemnation of weakness and haste 3. On the character of thankfulness and in praise of it, and condemnation of ingratitude

po l iti cs, reli g i on and the occul t  | 45

4. On consultation and its blessings and in condemnation of failing to do it 5. On justice, and in condemnation of injustice and oppression 6. On agreement and alliance and in condemnation of division and dissent 7. In praise of loyalty and in condemnation of treachery 8. On being alert and seizing the opportunity, and in condemnation of dilatoriness and inattention 9. On forgiveness and doing good 10. In praise of honesty and condemnation of lying

ii. The second qa‘ida, on the sultanate and dominion, divided into two babs: 1. On the sultanate, and the attributes of the sultan 2. On the officials upon whom the kingdom relies and in whose hands are the reins of public interest, in five tabaqat: a.  The vizierate b.  The chancery c.  The secretariat of the army d.  The secretariat of the treasury e.  The remaining retinue (hashiya) iii.

The third qa‘ida, on sharia and religion, in four rukns: 1. The office of mufti 2. The office of qadi 3. The office of muhtasib 4. The awqaf

iv. The fourth qa‘ida, on perfecting the goal [of the book] through various additions, consisting of diverse questions: 1. On questions of ritual (‘ibadat) 2. Questions on purchases (muba‘ayat), 3. Questions on marriage law (munakahat), 4. Questions on crimes (jinayat) 5. Mathematical questions

46  |  syri a i n crusa d e r time s Conclusion of the book: on prayers which are answered Admonition: Stories of the righteous. This division into four qa‘idas is most immediately reminiscent of al-­Ghazali’s works. His Kimiya-yi sa‘adat is divided into four quarters which are then subdivided into ten sections, as is the Ihya ‘ulum al-Din, which shares many of the Kimiya’s concerns and its structure (with some variation in the arrangement of the sections). Neither of these works – both of which are considerably larger than al-‘Iqd al-farid – are mirrors, but rather were more generally concerned with expounding Sufism, although the Persian Kimiya was written for the Seljuq political elite. The Kimiya was known in Syria and influenced Mirrors for Princes, for it was a major influence on the Persian Bahr alfawa’id composed in Aleppo in the mid-twelfth century.58 The first qa‘ida of the ‘Iqd, in ten chapters, offering moral advice, forms a self-contained homiletic ‘Mirror for Princes’ in its own right. This ten-part structure, ultimately derived from the idea of the ten branches of the tree of faith popularised by [pseudo?]-Ghazali in his Nasihat al-muluk, which had become the standard format for Mirrors for Princes by the thirteenth century,59 and can be observed, for instance, in the near-contemporary Arabic mirror from Anatolia, al-Zanjani’s al-Lata’if al-‘ala’iyya.60 Unlike al-Zanjani’s work, however, the debt to al-Ghazali for structure or contents is not explicitly acknowledged by Ibn Talha, although like al-Zanjani and al-Ghazali (at least in the first part of the Nasihat al-muluk), Ibn Talha draws heavily the Qur’an, hadith (with highly abbreviated isnads) and stories of early caliphs (down to the ‘Abbasid era),61 to illustrate anecdotes pointing to correct behaviour, although on occasion there is also reference to Isra’iliyyat.62 However, there are very few Iranian elements compared to al-Ghazali or al-Zanjani’s work, with just one passing mention of the Sasanian ruler Ardashir b. Babak in the text.63 This emphasis on Arab and Islamic models is reminiscent of another near-contemporary work, Ibn Haddad’s al-Jawhar al-nafis dedicated to Badr al-Din Lu’lu’ of Mosul.64 The second qa‘ida is also fairly typical of other examples of the genre which seek to offer both ethical and practical advice, such as the Bahr al-fawa’id, composed a century or so earlier in Aleppo. Unlike the Bahr al-fawa’id (or indeed its Arabic counterpart, the Mufid al-‘ulum),65 Ibn Talha’s work has

po l iti cs, reli g i on and the occul t  | 47 a much stricter and more logical structure. The practical advice on ruling, if idealised, is at least to some degree adjusted to the circumstances of the Ayyubid state, again in contrast to many works in which the prescriptions are so general that they could apply almost anywhere. Thus the sultan is lauded in conventional terms as the one who ‘upholds the protection of God’s servants, land, religion and upholds his prescribed punishments (hudud) and preserves God’s rules,’66 while being warned against the five negative characteristics of arrogance, conceit (‘ujb), pride, meanness and lying.67 The sultan is also advised to delegate his authority to his vizier, whose role is also described in largely formulaic and conventional terms as the king’s ‘eye and hand, the writer of his speech, the chamberlain of his nature, the messenger of his tongue’.68 The vizier is described as the pole of the state and its axis, the forearm of the kingdom and its knight who enlightens the sultan in the darkness of important affairs by the lights of his administration, and relieves him of the burdens of events, great or small, noble or low …69

The importance of the office of vizier is sanctioned through reference to the Qur’an and hadith.70 Ibn Talha distinguishes two different types of vizier, the wazir tafwid (vizier of delegation) and the wazir tanfidh (executive vizier). The first of these, the wazir tafwid, is the most important, for the sultan invests his authority in him. The wazir tafwid has complete authority over the running of affairs, with only two exceptions: he cannot choose the heir apparent (wali al-‘ahd), nor can he depose one whom the sultan has appointed. The wazir tanfidh is an intermediary between the sultan and his people, responsible for executing the sultan’s orders, but he does not have the authority to intervene in decisions, to investigate abuses (mazalim) or to make senior appointments. Unlike the wazir tafwid, however, the wazir tanfidh may be a slave and may not know the rules of sharia, may be ignorant of the conduct of war and, according to al-Mawardi, may be a dhimmi, although Imam al-Haramayn al-Juwayni is quoted as forbidding this.71 The section on the vizierate derives from Chapter 2 of al-Mawardi’s al-Ahkam al-sultaniyya.72 The discussion of administration also draws on al-Mawardi but is adapted in places to suit Ayyubid circumstances. For instance, in his description of army administration and pay, Ibn Talha seems

48  |  syri a i n crusa d e r time s to be indebted to al-Mawardi (albeit in abridged form), describing the tribal system of the Arabs and their precedence for pay according to their proximity to the Prophet. However, when it comes to discussing non-Arab peoples, al-Mawardi had mentioned Turks, Indians, Daylamis and Jibalis, who he recommended should be paid according to their precedence in converting to Islam. In al-‘Iqd al-farid, however, Indians and Jibalis are omitted, and rather ‘Turks, Kurds and Daylamis’ are listed, in recognition of the realities of the composition of the Ayyubid forces.73 Despite the substantial borrowings from al-Mawardi, much of the text is independent. The organisation of the work owes nothing to al-Ahkam al-sultaniyya, which had little to say about the moral qualities required of the ruler. Even in chapters which share concerns with al-Mawardi, there can be substantial differences of content and approach. The section in the ‘Iqd on the office of qadi, for instance comprises ten exemplary anecdotes about ‘Abbasid qadis that are not paralleled in al-Mawardi or al-Ghazali.74 It is interesting that Ibn Talha offers no discussion whatsoever of the role of the caliph: there is not even al-Mawardi’s famous attempt to justify the imarat al-istila’. Rather the sultan is conceived of as God’s representative on earth, and Ibn Talha specifically refers to the ‘kings’ as doing the duties which jurists had previously allotted to the caliphs, for example mentioning, ‘kings whom God had established to protect religion and protect the people (milla) and the sharia’.75 Thus even before the Mongol conquest of Baghdad and extirpation of the ‘Abbasid line, at least for this author the caliphate simply does not exist, although tales of the praiseworthy conduct of ‘Abbasid caliphs of the eighth and ninth centuries abound.76 This is thus not a case of Shi‘ite anti-‘Abbasid prejudice but rather a reflection of the irrelevance of the caliphate even to theoretical discussions of Islamic governance in the years before its demise. In other passages, Ibn Talha seems to allude to his own distinctive concern with ‘the science of letters’. In his discussion of the necessary qualifications of a secretary (katib), he stresses that the secretary must know ‘the state of the letters’ ‘in order that he can open locks thereby, and put their correct state and explain their forms’, and says he has written in detail on letters in a treatise entitled al-Kawkab al-najim fi ma‘rifat al-tarajim.77 However, it is far from certain that this was a numerological, occult treatise. Another work possibly suggestive of lettristic interests which Ibn Talha mentions in al-‘Iqd

Plate 3.1  Ibn Talha’s da’ira, or circle of letters, in al-Durr al-muntazam. MS Istanbul, Süleymaniye Library, Hafid Efendi 204.

Plate 3.2  Ibn Talha’s Matalib al-su’ul fi manaqib Al al-Rasul. MS Istanbul, Süleymaniye Library, Damad İbrahim 303.

Plate 3.3  Ibn Talha’s al-‘Iqd al-farid li’l-Malik al-Sa‘id. MS Istanbul, Süleymaniye Library, Yeni Cami 985, dated 767/1365.

Plate 3.4  Final folio of Ibn Talha’s Nafa’is al-‘anasir, showing (bottom left) the mysterious letters. MS Istanbul, Beyazit State Library, Veliyüddin 2548.

Plate 3.5  Table for calculating days of the months from al-‘Iqd al-farid. MS Istanbul, Süleymaniye Library, Yeni Cami 985.

po l iti cs, reli g i on and the occul t  | 49 al-farid as one of his compositions is the Zubdat al-musannafat fi al-asma’ wa’l-sifat,78 although again its nature is not wholly clear and no manuscripts have come to light. Apart from al-Durr al-muntazam, one other such lettristic work by Ibn Talha has come down to us, the Inas al-hikam min anfas Abi’lHakam, a short treatise preserved in an apparently unique manuscript, today in Leiden.79 The association of Ibn Talha’s mirror with lettrism for at least some of its audience is suggested by the final folio of the oldest surviving manuscript of the Nafa’is al-‘anasir, dated 777/1375, which was copied from the autograph; the manuscript concludes with mysterious letters (laghz alhuruf    ) apparently copied from the original (Plate 3.4). The most distinctive section of al-‘Iqd al-farid is the fourth qa‘ida, with its problems of fiqh and mathematics. Ibn Talha remarks that he has included these various topics for the benefit of men of knowledge, and action, for those who come and go, and I started with that which is a means to the knowledge of scholars who strive to bear knowledge in their breasts and went deep to obtain it until they reached the limit of their abilities.80

The fiqh questions are further subdivided into questions on rituals (‘ibadat),81 questions on purchases (muba‘ayat),82 questions on marriage law (munakahat)83 and questions on crimes (jinayat).84 These comprise sixty questions in total. Here is an example of the type of question and solution: Problem: A man rented a house from a man to store in it one kurr of wheat, but he stores there two kurrs of wheat. Should the tenant pay an excess on the agreed rent because of his extra wheat or not? Neither an absolute yes or no is correct. Answer: If the rented house is on the ground, there is no need to pay extra for the excess wheat, but if the house is a room on the roof, he must pay rent representing the extra wheat, because the extra wheat can cause extra damage to the roof.85

Another example is as follows: Problem: A man has two wives, one Muslim, one Christian. He says to the Christian one, ‘You have apostacised and become a Muslim.’ He says to

50  |  syri a i n crusa d e r time s the Muslim one, ‘You have apostacised and become a Christian.’ Both of them call him a liar, but one of them is not telling the truth. Are both of the marriages void or not, or is just one void and the other still valid? Neither of these answers is correct. Answer: If it is before consummation, both marriages are void owing to the presence of a factor which voids them in his claim, on account of which he criticises them: in the case of the Muslim woman because of his declaration that she has apostacised and the Christian woman because of her denial of Islam has apostacised according to his claim. If it is after consummation, the marriage of the Muslim woman remains valid, but the marriage of the Christian woman is suspended until the end of the ‘idda, when her marriage is cancelled.86

Ibn Talha then moves on to a second type of question which demands considerably more complex answers, which, he tells us at some length, were extremely popular at the Ayyubid court: The second type [of problem] is more perfectly beautiful, more comprehensive in meaning, and only one who has devoted himself heart and soul to learning can give the correct answer. And the sultan al-Malik al-Kamil – may God sanctify his soul and bestow blessings upon our generous87 lord the sultan al-Malik al-Nasir – used to make particular use of it and devote his attention to it on occasions when the most excellent men of the land cane to him, and when great men came to him on days of [great] occasions, festivals and gatherings of the the pilgrimage caravan (mahfal). He would ask them these questions which would inform him of their level of virtue so that he would look after them to his [best] ability, and he would settle each one according to the station he deserved in his generosity and charity, and those who were conciliatory and those who were contrary would thereby become evident by their report and expertise. By my life, my noble lord al-Malik al-Nasir Salah al-Din, may God pour upon him the lights of certainty and make him one of his pious servants, by the purity of his essence, the intelligence of his mind, the perfection of his understanding, the light of his perception and the perfect awareness with which God has singled him out, the strength of his nature, the excellence of his talent,

po l iti cs, reli g i on and the occul t  | 51 and the cleverness of his character, does not need to mention problems to distinguish between one who gives voice to delusions, for he wears the robes of falsehood, and one whom God has singled out from the lamp of lights with light upon light. However, emulating the excellent habits of bygone sultans is counted a custom (sunna), and following their praiseworthy tracks is a good deed. Therefore, I include a glimpse of this type of problem in this book so that it may serve the sultan and he can use it as a pretext for testing [people], even if, with his noble gaze, he has no need of it. I have made it as brief as possible, mentioning the form of the question, the way of answering, and some analysis. Problem: Two men go out hunting and they find a prey. They both go after it and shoot at it with their arrows successively, one after the other. They wound it and it later dies. What is the judgement? This is the form of the question upon which many judgements depend, despite its brevity and the simplicity of its form.88

The answer to this extends over two and a half printed pages, assessing the various possibilities for rights of ownership of the prey, whether it was licit to eat it, and the circumstances under which one of the men might have to compensate the other for it, and the rate that compensation should be set at. Of similar type are the mathematical puzzles which conclude this section of the book. These represent fairly basic examples of algebraic problems and their solutions involving linear equations. They are not, however, set out as algebraic problems with equations and explicit variables as one would see in a formal mathematical work such as those of al-Khwarazmi,89 just as the fiqhtype problems and their solutions are not set out as they would be in a formal fiqh work. To give one example of these algebraic problems: Problem: A man owns a horse which three people come to him to buy. They ask him about the price, and he tells them. The elder one tells the middle one, ‘If you give me three fifths of what you have with you in dinars, I’ll have the price of the horse.’ The middle one says to the youngest, ‘If you give me four sevenths of the dinars you have, I’ll have the price of the horse.’ How much is the price of the horse? And how many dinars did each one of them have?90

52  |  syri a i n crusa d e r time s While there is some evidence for Ayyubid patronage of mathematicians and interest in mathematics, mainly for the purposes of astrology and the construction of automata,91 these sorts of basic mathematical problems would have contributed little or nothing to such fields. Yet to be cultivated in the Ayyubid age, to be an adib, was to have a broad knowledge of all fields of learning, which perhaps explains why these problems are presented in laymen’s terms. Just as the first qa‘ida seeks to address the sultan’s personal morals and conduct, the fourth qa‘ida aims to give him the tools to appear as a cultivated layman, to hold his own with the scholars who frequented his court. Ibn Talha’s description of al-Kamil’s questioning of scholars suggests that the questions and answers in the fourth qa‘ida were intended as a sort of cheat-sheet for the sultan, whatever his polite disclaimers about al-Nasir Yusuf having no need of it. Equally, however, it might be valid to understand them as another expression of the love for puzzles and problems that ­characterised court life.92 Finally, we have the diagram for calculating the first days of months through the name of al-Nasir Yusuf (Plate 3.5). The interpretation of this is somewhat problematic. In apocalyptic malhama literature, the day on which a given month began is vested with a divinatory significance.93 However, the dates in our table are all historical rather than in the future, so it can have had no use practical use for determining on what day to undertake a task. The table may point, like the questions, to an Ayyubid court culture which loved brain-teasers and puzzles, but it may have other implications too. A parallel to the table of dates can be found in Ahmad al-Buni’s al-Lum‘a al-nuraniyya, a roughly contemporary work dealing with the magical properties of the divine names. Al-Buni’s table is to determine what prayers should be said at what hour of what day, the efficacy of the prayers depending on the correct choice, and has very similar instructions for use to that in al-‘Iqd al-farid.94 The correspondence between the two is intriguing, and suggests that a proper understanding of the far from apparent use of Ibn Talha’s table may be found through looking at comparable magical works. Moreover, it must be significant that the table in al-‘Iqd al-farid goes up to the year 644/1246 – the very year of the Khwarazmian invasion, which seems to have had such importance for Ibn Talha; indeed such importance that even after the rededication of the work to Najm al-Din Ghazi there was no effort to update or change this.

po l iti cs, reli g i on and the occul t  | 53 Certainly, the presence of occult elements in a ‘Mirror for Princes’ would not be unprecedented: the Bahr al-fawa’id contains a chapter to devoted to ikhtilaj (palmomancy).95 The final sections of the work, after the conclusion of the fourth qa‘ida and the table of dates, superficially have a very different tone, being devoted to prayer and homily. After the table in al-‘Iqd al-farid, a passage largely based on hadith emphasises the necessity of prayer,96 then a new section opens under the heading tanbih (warning). This section is devoted to the necessity of piety towards God given the proximity of death and the next world, based on hadith and Qur’anic quotations such as al-Infitar 13–14: ‘The virtuous are in heaven and the wrongdoers in hell’, and al-Baqara 281: ‘Fear a day when you return to God, every soul will be compensated for what they have gained, they will not be oppressed.’ There follow anecdotes on the theme of rulers facing death, for instance: Hasan al-Basri wrote to ‘Umar b. ‘Abd al-Aziz when he sent to him saying, ‘Tell me something useful, but be brief.’ He wrote to him, ‘O Commander of the Faithful, even if you had the life of Noah, the wealth of Solomon, the certainty of Abraham and the wisdom of Luqman, there would still be fear of death before you, and behind it are two doors, one heaven, the other hell and if you make a mistake you will go to that one. Behave accordingly.’97

Similar stories are related of other caliphs. Ibn Talha then cites al-Ghazali at length on the necessity of preparing oneself for death.98 Finally, there appears a long anecdote set in Umayyad times about three brothers, one an ascetic (zahid ), one a great amir associated with ‘Abd al-Malik b. Marwan, and the other a respected merchant. The anecdote shows how the amir and the merchant learn from their ascetic brother to divest themselves of wealth in order to prepare for death.99 Then at the very conclusion of the work, Ibn Talha draws these threads together. Embracing a lifestyle of asceticism, he says, is the way to ensure one can enter paradise; but God has also created another way, which is through just rule.100 Such homiletic passages are found elsewhere in advice literature. PseudoGhazali’s Nasihat al-muluk, for instance, also stresses the necessity of preparing oneself for death, and in the near-contemporary works of Najm al-Din Razi, the ruler is urged to embrace an ascetic Sufi lifestyle, but they contain none of

54  |  syri a i n crusa d e r time s the practical reflections on how the diwan should be organised.101 In the Bahr al-fawa’id too, a whole chapter (number 33) is devoted to admonishing the ruler to remember the Hereafter.102 However, such concerns are much less evident in the more legalistically minded normative advice literature such as al-Mawardi’s al-Ahkam al-sultaniyya, which as we have seen is a major influence on Ibn Talha. To a degree then, one may see Ibn Talha’s project as successfully weaving together two disparate strands in the ‘Mirrors for Princes’ tradition, the more homiletic, religiously motivated works with the legalistic and normative. To this is added Ibn Talha’s own unique contribution, the fourth qa‘ida with its problems and solutions. At the same time, given Ibn Talha’s commitment to asceticism – the factor which as the medieval chronicles tell us, led to him abandoning the vizierate – it is also tempting to see this work as reflecting his own personal agenda. Yet in the first half of the thirteenth century, enthusiasm for asceticism was becoming widely spread among not just the piety-minded but broader swathes of society, including the elite. Another Syrian ‘Mirror for Princes’ of the period to emphasise the need for asceticism was Sibt b. al-Jawzi’s al-Jalis al-salih wa’l-anis al-nasih, dedicated to the Ayyubid al-Ashraf Musa in 613/1216.103 Ayyubid interest in piety is reflected in another short treatise by Ibn Talha, on the benefits of prayer over fasting (Tahsil al-maram fi tafsil al-salawat ‘ala al-siyam),104 which was written at the request of a learned friend to resolve a debate raging at the court of al-Nasir Yusuf over the relative virtues of these two forms of piety.105 Conclusion Ibn Talha’s works show that the biographical dictionaries mislead: dedicated to Ayyubid and Artuqid rulers, they indicate a much closer involvement in court and political life than his two-day vizierate implies. Asceticism, occultism and contemporary political concerns all blend together in his various works in different ways. Despite the very different contents of al-Durr al-muntazam and al-‘Iqd al-farid, both share an interest in asceticism, and both allude to the fateful year of 644/1246, the date of the Khwarazmian defeat. Courtly concerns about piety are reflected in the Tahsil al-maram, and possibly even in the Matalib al-su’ul, which seems to be an expression of the Sunni–Shi‘ite conciliation that was a hallmark of the late Ayyubid period and was encouraged by members of the dynasty such as al-Nasir

po l iti cs, reli g i on and the occul t  | 55 Yusuf, the dedicatee of the Nafa’is al-‘anasir. Ibn Talha thus serves as a prominent representative of the Shi‘itisation of Sunnism in the period, and his court connections suggest his part in giving intellectual credibility, as a leading ‘alim, to this process. However, it would be wrong to see this process of Shi‘itisation as purely a matter of practical politics, an effort to conciliate rival communities. It may also reflect the growing interest in the occult exhibited in the works of Ibn Talha and his contemporaries like al-Buni, in which ‘Ali b. Abi Talib plays a crucial part alongside Sufism. Although the critical comments of the biographical dictionaries indicate that his occult interests certainly provoked disapproval among some, Ibn Talha seems an important representative of the process by which the occult entered the intellectual mainstream in Islamic society, as it did increasingly from the thirteenth century. Indeed, Ibn Talha’s own claim to be able to convey and interpret ‘Ali’s hidden words and the mystical secrets of the letters of the Arabic script was sufficiently potent to ensure the continued circulation of his works in later generations. Notes 1. The research leading to these results has received funding from the European Research Council under the European Union’s Seventh Framework Programme (FP/2007–2013)/ERC Grant Agreement n. 208476, ‘The Islamisation of Anatolia, c.1100–1500’. 2. Al-Dhahabi, 1960–66, V: 213; a short reference in al-Yafi‘i, 1997, IV: 99–100; Ibn Kathir, 1966, XIII: 186; Abu Shama, 1947, 188; al-‘Ayni, 1987, I: 64; Sibt b. al-Jawzi puts his death under ah 653, but gives no information about him; Sibt b. al-Jawzi, 1343, XV: 92. 3. Humphreys, 1977, 247; Ibn Wasil, 1977, V: 200. 4. Al-Safadi, 1974, III: 176; al-Dhahabi, 1960–66, V: 213. 5. For an overview, see Lory, 2004. 6. Further evidence for Ibn Talha’s interest in hadith comes from an ijaza of a manuscript where he is cited as auditor of Majd al-Din Ibn al-Athir’s Jami‘ al-usul in 604; Ritter, 1953, 72. 7. Ibn ‘Imad al-Hanbali, 1998, V: 389–390. Ibn ‘Imad’s entry seems to rely partly on al-Safadi’s, but contains additional information. 8. For an overview of manuscripts of his surviving works, see Brockelmann, 1943–1947, I: 463–464, S, I: 838–839.

56  |  syri a i n crusa d e r time s 9. Masad, 1998. A few passing references to Ibn Talha appear in Pouzet, 1988, 134, 231, 258, 419, and there is a short discussion of his circle in Coulon, 2016b, 525–529. 10. A rare and early independent manuscript is Leiden University Library, MS Or 2832, which may be thirteenth century, although while the famous da’ira is referred to here it is not actually represented in this manuscript. In this study I rely largely on the printed edition by ‘Atiyya (n. 15 below) and MS Istanbul, Süleymaniye Library, Hafid Efendi 204, a fifteenth-century manuscript that also includes al-Bistami’s Miftah al-jafr al-jami‘, and which does illustrate the circle. See also my remarks on the text in note 19 below. On ‘Abd al-Rahman alBistami, an extremely important but neglected figure, see Gril, 2005, 183–195. 11. On this see Coulon, ‘L’ésotérisme shiite’, 525–528. 12. It seems that in fact some versions of the Shams al-ma‘arif attributed to al-Buni drew on one of Bistami’s works, meaning it can be no earlier than the fifteenth century. On this see Coulon, 2016a, 1–26, esp. 13; Gardiner, 2012, 81–143, esp. 102–103, 123–129. Further research on the relationship of Ibn Talha’s al-Durr al-muntazam with the Corpus bunianum is required. 13. Compare, for instance, another recently published circle, which shows remarkably little Islamic influence and shows how the letters are served by angels who all have Hebrew names. Bonmariage and Moureau, 2016, esp. 11–37 for a discussion. 14. Al-Imam ‘Ali b. Abi Talib, n.d. [c.2014]. 15. Ibn Talha, 2004, 17. 16. Ibid., 25. 17. Masad wrongly says 728; 718 is confirmed by both the MS Hafid Efendi 204 and the edition (Ibn Talha, al-Durr al-muntazam, 26). 18. Masad, 1998, 72. 19. Safadi,  al-Wafi bi’l-Wafayat, II, 353; cf. Yunini, Dhayl Mir’at al-Zaman fi Ta’rikh al-A ‘yan, ed. ‘Abbas Hani al-Jarrah, XII:  al-juz’ al-rabi‘ min aldhayl  (Beirut, 1434), 298 (under the notice for al-Ikhmimi, ah 684). This passage may suggest that the text seen by the medieval biographers was rather different from the manuscripts of al-Durr al-muntazam I have seen, for alIkhmimi is not mentioned. 20. Cited in Masad, 1998, 85–86. However, Masad was unaware of the references in al-Yunini and al-Safadi, which were later copied by ‘Abd al-Rahman al-Bistami. Masad therefore argued that Bistami’s identification of Ibn Talha’s pious friend as al-Ikmimi was not credible. See also Coulon, ‘L’ésotérisme shiite’.

po l iti cs, reli g i on and the occul t  | 57 21. The authenticity of these works needs examining. See Hamdan, 1985. 22. Coulon, 2016a, 8–13; Gardiner, 2012, 91–94. 23. Ibn Talha, 1419, 19–22. 24. Al-Khwansari, 1390–1392, V: 259, unseen by me, cited by Kohlberg, 1992, 264. 25. Ibn Talha, 1419, 34. 26. Ibid., 42–44. 27. Cook, 2002, 36–38. 28. Ibn Talha, 1419, 61–224. 29. Ibid., 225–245 deal with Hasan; Husayn is treated in ibid., 246–266. 30. Ibid., 33. 31. Ibid., 119, esp. 185–195. 32. Ibid., 111, cf. 178–179. 33. Ibid., 111. 34. Ibid., 166. 35. Ibid., 110. 36. Ibn al-‘Adim, 1988, V: 2127. 37. Ibn Talha, 1419, 211–213; cf. Newman, 2000, 150. 38. Ibn Talha, 1419, 219–220. 39. Cook, 2002, 195. 40. Eddé, 1999, 443–445. 41. Masad, 1998, 153–155. 42. Pouzet, 1988, 248–250. 43. Kohlberg, 1992, 264. 44. Cook, 2002, 28. 45. Mulder, 2014, 76–80, 103–105; on the Shi‘ite population of Aleppo see also Eddé, 1999, 436 ff. 46. Mulder, 2014, 103–105. 47. Ibid., 105. 48. Ibid., 1–2. 49. Cf. the comments of Pouzet, 1988, 260. 50. On apocalyticism, see Masad, 1998; on Ibn ‘Arabi and Ismailism, see Ebstein, 2014; for ‘Ali’s role in the occult, see Coulon, 2016b. 51. Ibn Talha, Nafa’is al-‘anasir, fol. 82b–83a. 52. Humphreys, 1977, 286–287. 53. However, for other examples of cultural exchange between Artuqids and Ayyubids in this period, see Taragan, 2012, 79–88.

58  |  syri a i n crusa d e r time s 54. Ibn Talha, 2000. I have also consulted MS Istanbul, Süleymaniye Library, Yeni Cami 985, which is dated to 767/1365. 55. For a preliminary survey of the manuscripts of these works see Brockelmann, 1942–1947, I: 463–464, S, I: 838–839. Note that Brockelmann is mistaken in his reference to a London manuscript of the Nafa’is al-‘anasir, which he gives as Br. Mus. 1530; in fact this is entry no 1513 in Cureton’s catalogue, corresponding to British Library MS Add 25,753 (dated ah 1166). 56. This is Gotha, Landesbibliothek, MS orient A. 1882. Korn describes this manuscript as belonging to ‘a later phase of Artuqid cultural patronage’, but this seems uncertain to me as the manuscript is dated 856/1452, after the conventional dates for the end of the dynasty. Unfortunately, the inscription on the shamsa has been erased, but it clearly is from a royal library. See Korn, 2011, 385–407, at 397–398 (with a reproduction of the title folio). I am grateful to Bruno De Nicola for supplying me with images from this manuscript. 57. These are the fiqh questions discussed on p. 49 above. See Ibn Talha, 2000, 214 fols, which is incorporated word for word into al-Durr al-nadid fi manaqib al-Malik al-Zahir Abi Sa‘id, Staatsbibliothek zu Berlin, MS Wetzst. I. 133, fol. 24aff. I am very grateful to Anne-Marie Eddé for drawing my attention to this text, of which I understand she is currently preparing an edition. 58. On these works and their relationship, see Carole Hillenbrand, 2013, 59–69. See also Anon., The Sea of Precious Virtues, 1991, xv–xvi. 59. Al-Ghazali 1971, 4; cf. Marlow, 2004, 180. 60. Peacock, 2016, 285. 61. Ibn Talha, 2000, 81 on al-Mu‘tasim; ibid., 90 on al-Rashid billah; ibid., 134 on al-Ma’mun; ibid., 110 on al-Mansur; ibid., 126 on al-Muntasir b. al-Mutawakkil, ibid., 147 on al-Mahdi. 62. Ibid., 79. 63. Ibid., 119. 64. On whom see Stefan Leder, 2015, 106–107. 65. On the relationship between these works see van Gelder, 2001, 313–338. 66. Ibn Talha, 2000, 158. 67. Ibid., 161–163. 68. Ibid., 169. 69. Ibid., 170. 70. In particular Suras Taha 29, al-Furqan 34 and al-Qiyama 11. 71. Ibn Talha, 2000, 173. 72. Cf. al-Mawardi, 2013, 39–49; tr. Yate, 1996, 37–47. Ibn Talha does briefly

po l iti cs, reli g i on and the occul t  | 59 mention al-Mawardi (Ibn Talha, 2000, 214), although the book cited here is al Mawardi’s al-Hawi fi’l fatawa. 73. Ibn Talha, 2000, 183; cf. al-Mawardi, 2013, 334–335; tr. Yate, 1996, 289. The following passage dealing with the rate of pay is more or less identical in the ‘Iqd (184) and al-Mawardi, 2013, 336, tr. Yate, 1996, 291). 74. Ibn Talha, 2000, 196–206. 75. Ibid., 191. Leder, 2015, 99 suggests a distinction was made in this sort of literature between sultans (who were reconcilable with sharia, even if the institution was influenced by Iranian royal traditions) and kings, who were not. However, in this text Ibn Talha clearly uses malik and sultan as synonyms. 76. See note 61 above. 77. Ibid., 181. 78. Ibid., 247. 79. Leiden University Library, MS Or 2833. The manuscript consists of twelve pages and is dated by Voorhoeve to the thirteenth century (Voorhoeve, 1957, 129). However, readers’ notes on the final folio dated ah 923 and 919 indicate it continued to circulate in the sixteenth century, perhaps in conjunction with al-Bistami’s works – a note in a later hand on the cover seems to confuse the two authors (hadhihi rumuz Ja‘far b. Talha al-Bistami). The Inas al-hikam deals with the values and meaning of letters in the Qur’an. According to a note at the end of Leiden MS Or 2832, fol. 19a (al-Durr al-muntazam), with which this manuscript was probably bound, the works contain the answers given by Ibn Talha to questions about the tafsir of al-Imam Abu’l-Hakam b. Abi al-Tarkhan. This copy of the Inas al-hikam was evidently originally bound together with another short work by Ibn Talha, the Tahsil al-maram fi tafsil al-salwat ‘ala al-siyam, now held in Staatsbibliothek zu Berlin, MS Ldbg. 681, which on palaeographic grounds was copied by the same scribe and was evidently part of the same collection of Medinan manuscripts purchased by Landberg. 80. Ibn Talha, 2000, 213. 81. Ibid., 214. 82. Ibid., 218; title from MS Yeni Cami 985, fol. 214a. 83. Ibn Talha, 2000, 222. 84. Ibid., 227, title from MS Yeni Cami 985, fol 224a. 85. Ibn Talha, 2000, 221. 86. Ibid., 224. 87. Read ghamr with Yeni Cami 985, not ‘umr as the printed text has. 88. Ibn Talha, 2000, 228–229.

60  |  syri a i n crusa d e r time s 89. I am very grateful to Scott Trigg for examining these problems and offering these comments. 90. Ibn Talha, 2000, 228–229, 238. 91. Brentjes, 2011, 326–365, esp. 342–344. 92. Cf. the chapter by Julia Bray in this volume. 93. See for example the richly illuminated malhama composed for an unidentified Mamluk patron, Istanbul, Süleymaniye Library, MS Hekimoğlu 592, fol. 294b–295a: if the Syrian month of Kanun al-akhir begins on a Sunday, it means the winter will be mild and rainy, the summer hot and travel dangerous; if it begins on a Monday, it means there will be high prices, fighting and bloodshed in the land of Rum, and so on. 94. Coulon, 2013, III, edited text of al-Lum‘a, 36; analysis in ibid., I: 454–456. 95. Anon., Sea of Precious Virtues, 1991, Chapter 31. On ikhtilaj, see Fahd, 1966, 397–429. 96. Ibn Talha, 2000, 245–248. 97. Ibid., 251. 98. Ibid., 255–257. 99. Ibid., 259–261. 100. Ibid., 261. 101. See the discussion of Razi’s Mirsad al-‘ibad and marmuzat-i Da’udi in Peacock, 2016, 289–295. 102. Anon., Sea of Precious Virtues, 191, Chapter 31, 301–306. 103. Sibt b. al-Jawzi, 1989. The ninth chapter of this work is devoted to accounts of ‘the righteous and ascetics’ (al-salihin wa’l-zuhhad). 104. MS Berlin, Staatsbibliothek, Ldbg. 681. 105. Wa dhakara innahu jara fi al-maqam al-‘ali al-mawlawi al-sultani al-maliki al-nasiri al-salahi … tashajur fi al-tafdil bayna al-‘ibadatayn salatan wa-sawman (MS Berlin, Staatsbibliothek, Ldbg. 681. fol. 1b). The work draws on the hadith collections of Muslim and Bukhari, but work also reflects contemporary intellectual trends in its division into an argument based on transmitted sciences (naqli, that is, the hadith collections), and one based on reason, ‘aqli).

4 Adapting to Muslim Rule: the Syrian Orthodox Community in Twelfthcentury Northern Syria and the Jazira R. Stephen Humphreys

D

espite the mass of scholarship on twelfth- and thirteenth-century Syria, impressive both in size and quality, our understanding of many major problems remains very unsatisfactory. Foremost among these is surely the social and political life of the Christian communities in the Muslim-ruled parts of Syria and the Jazira. This situation stems in the first instance from our documentation, or more properly the lack of it. Very few directly relevant original documents survive, while medieval Muslim writers seem to work from inside the bubble of their own community, quite oblivious to their indigenous Christian neighbours. We do have the works of a number of well-informed and articulate Eastern Christian chroniclers, in both Syriac and Armenian. However, they are clerics, with the predictable perspectives and biases of their class. They have much to tell us about political and military events as well as ecclesiastical matters, but they devoted little attention to the secular institutions and everyday life of the laity. One might expect some interesting observations from the Latin writers residing in the Crusader principalities, but even a figure as deeply grounded in ‘Oriental’ affairs as William of Tyre has strikingly little to say about the Christian communities in Muslim territories, even when they are (as in Damascus) almost next door. The lack of attention to the social life of ordinary Christians living under Muslim rule is not only an artifact of our sources, however. It also reflects the long-established preoccupations of Western scholarship on Eastern

64  |  syri a i n crusa d e r time s Christianity, going back to the seventeenth century, which have focused on ecclesiastical institutions, religious doctrine and practice, conflicts within and between the various Christians sects, and occasionally the broader currents of intellectual life.1 Despite these problems, however, the age of the Crusades offers significant prospects for productive research, at least on the Christian populations of northern Syria and especially the Jazira. Despite their limitations, our sources for this region are more plentiful and represent a wider range of perspectives than at any time since the late tenth century.2 Moreover, both the Frankish settlements and (more recently) the Muslim polities of this area have been intensively studied, giving us a richly layered context in which to situate the information about Christian communities that we can glean or infer from our texts. This situation contrasts markedly with that for central Syria, between Damascus and Homs, where data on Christian life are scarce and unrevealing. Ibn al-Qalanisi’s (d. 1160) closely observed account of Damascene and regional politics hardly hints that Christians are even to be found in his city. On the other hand, he evinces no hostility toward them, in stark contrast to the Franks and especially the Isma‘ilis. Perhaps one must simply infer from his silence that he did not view the Christians of Damascus and its hinterland as political actors in any meaningful way. They were, so to speak, just part of the landscape, and there was no real reason to mention them. It is important to recall that Syria is not one place but many. Northern Syria and the Jazira constitute an extremely complex terrain, but southern and central Syria are a simpler matter. Here we have two sets of actors – Franks on the west and Muslims on the east, with relatively clear-cut and roughly stable borders, at least between the Frankish capture of Tyre in 1124 and the conquests of Saladin in 1187–8. Disruptions during those six decades were occasional and largely the work of outsiders – Crusader expeditions from Europe in the first instance, and secondarily Muslim forces based in Mosul, Aleppo and (after 1174) Cairo. The north in contrast represents a classic borderland: Franks, Byzantines, Armenians, Turcomans, Kurds, Arab tribesmen and town-dwellers, Aramaic-speaking native Syrians, all intermingled in constantly shifting patterns of settlement and political formation. These disparate groups were governed (in a manner of speaking)

ada pti ng to musli m rul e  | 65 by local warlords under a variety of titles. Some of these were content to carve out petty domains comprising a few towns and castles, others had dreams of grandeur – mostly frustrated, but in two cases (the Rum Seljuqs in Anatolia, and the Zengids and their Ayyubid rivals in Syria and the Jazira) quite handsomely realised. In this madcap political theatre, the indigenous Syrians seem a striking anomaly, for they were not engaged in the struggle for power and ­territory – or at least, none of the region’s countless warlords emerged from that population. And that is the case even though they likely constituted the largest single religio-ethnic group in many parts of the region. In some areas (for example, the district around Melitene/Malatya, or the Tur ‘Abdin lying east of Amida and Mardin) they were surely the majority. Politically almost powerless, they are only half-visible to us, but their demographic importance and deep social and religious roots in the region demand serious attention. The indigenous Syrian Christians were not a unified community. Among the several things that set them apart from one another (for example, every­ day spoken language, local customs, scattered settlement clusters across a broad terrain) the most salient was no doubt religious confession. In the western and central Jazira – Diyar Mudar and Diyar Bakr – the great majority of native Christians belonged to the miaphysite Syrian Orthodox Church.3 Northern Syria also had a substantial Syrian Orthodox community, but especially in Aleppo and the surrounding districts there were many Melkites, in principle members of the official Byzantine Church, however tenuous their concrete institutional ties might be. Finally, in the eastern Jazira (Diyar Rabi‘a), centered on Mosul, the largest denomination was almost certainly the Church of the East, the so-called Nestorians. Intermingled among them, however, were numerous Syrian Orthodox villages and monasteries. The three churches had emerged out of the bitter Christological struggles of the fifth and sixth centuries. These quarrels were not forgotten by the twelfth century (nor are they today). Hostility and suspicion between the Melkites and Syrian Orthodox seem to have been especially acute, perhaps because the Byzantine Empire was still a real if sporadic actor in northern Syrian politics down to the death of Manuel Comnenus in 1180. The Byzantines, when they were in a position to do so, demanded conformity to the doctrines and hierarchy of the official church, while the Syrian Orthodox utterly rejected

66  |  syri a i n crusa d e r time s both the orthodoxy and the ecclesiastical authority of Constantinople. On the other hand, relations between the Church of the East and the Syrian Orthodox seem to have been more relaxed, even if there was some degree of mutual antagonism. Since the eastern Jazira was entirely under the control of Muslim warlords (no Franks, no Byzantines), each of the two churches in that region was compelled to focus first and foremost on its own place in a Muslim-dominated landscape. We can only speculate about why Syrian Christians, whatever their religious confession, did not join the struggle for power. The answer may lie in part in the circumstances of the Arab-Muslim conquest in the mid-seventh century. The Melkites and Syrian Orthodox fell under Muslim rule very early and abruptly, when Heraclius withdrew his battered armies from northern Syria in 637 and ‘Iyad b. Ghanm occupied the Jazira the following year almost without resistance. As to the Church of the East, it had long since adapted to living under non-Christian rule by the Sasanian Empire. From the beginning, then, the Syrian Christians had had to learn how to deal with the reality of Muslim rule. In contrast to the fractious and warlike Armenians to their north and east, they had no tradition of a military class and no opportunity to develop one. Whether we think of them as passive or as supremely adaptable, they had no choice but to come to terms with a changing cast of Frankish, Armenian and Turkish rulers. They suffered periodic bouts of violence and oppression from all three. On the other hand, all three did concede them a degree of religious liberty.4 The Melkites were deeply entangled in Frankish and Byzantine politics, and dwelt in a zone where towns and fortresses seemed to change hands constantly, under Frankish control this year and Muslim rule the next. The Church of the East in turn was concentrated in territories which lie beyond the geographical scope of this chapter. For those reasons we will focus our attention on the Syrian Orthodox community, in the region where it was clearly the largest and most deeply rooted non-Muslim population, the western and northern Jazira (Diyar Mudar and Diyar Bakr). So far as we can tell, this community’s only overarching public institution was the Syrian Orthodox Church, and that church – quite apart from its difficult relations with its Melkite, Roman and Armenian cousins – was very much at war with itself throughout the twelfth century. In the twelfth century its struggles revolved

ada pti ng to musli m rul e  | 67 mostly around factions struggling to control the major bishoprics rather than questions of theology and liturgy, and around the authority of the patriarch to impose discipline on his bishops. There were even a few cases of clergy converting to Islam (at least temporarily) and moving to Muslim-dominated towns. Such conversions were rarely a matter of deep spiritual conviction, and sometimes Islam was only a way station on a more complex confessional path. Some of these converts moved on to Christian-ruled cities – Antioch or Jerusalem – and there adopted the Chalcedonian creed of the Byzantine and Roman churches. On rare occasion such a soi-disant convert might even return to the Syrian Orthodox community. Under such convoluted circumstances the Patriarch of Antioch and all the East might strive to be a symbol of identity and unity within his community, but it was a unity that he found very difficult to enforce. Very often he could validate his title, let alone enforce his authority, only by appealing to a Muslim ruler – typically in Amida or Mardin. These amirs were often quite willing to lend their support, for due consideration, but they governed very circumscribed domains and hence their support could have only limited effect.5 Though there is no evidence for a secular aristocracy among the Syrian Orthodox, our Syriac sources do at least imply that there was some sort of lay leadership on the village level and in the various cities – most importantly Edessa (under Frankish rule from 1098 until 1144), Melitene, Mardin and Amida – where they constituted a high proportion and perhaps a majority of the population. Admittedly this leadership is barely visible even in the Syriac sources. To some degree it would have consisted of well-to-do merchants, village headmen and the like, but the influence of such persons could only have been exercised within their villages or urban neighbourhoods, though they might have had the social standing to speak with regime officials about some matters. At what we might call the managerial level, some of these officials (but a declining number) were themselves Christians, but persons of this rank had to be cautious about flaunting even their limited influence in a highly visible manner. They certainly enjoyed little of the status or notoriety of the high-ranking Christian officials in the early Umayyad Caliphate, or even those of the tenth and early eleventh centuries.6 A sign of their limited standing is reflected in the fact that – to all appearances at least – bishops or archimandrites could and did deal directly with Muslim, Armenian or

68  |  syri a i n crusa d e r time s Frankish rulers without needing to rely on lay notables or bureaucrats as intermediaries. Why, then, do the lay ‘notables’ among the Syrian Orthodox, such as they were, merit our attention? In part, because of the material resources they could bring to bear in rebuilding neighbourhoods, churches and monasteries, which were constantly assailed by earthquakes, fire and pillage. The material infrastructure of the Syrian community – not only churches and monasteries but also marketplaces and homes – clearly suffered severe damage from the constant raiding and counter-raiding of the twelfth century, inflicted by Franks and Armenians as well as by ambitious Turkish dynasts. Despite repeated destruction and pillage, however, efforts to restore and rebuild were unremitting.7 No doubt there was an overall arc of loss and degradation, but we rarely read that damaged structures were left derelict, with the major exception of stricken Edessa after Nur al-Din’s violent reconquest of the city in 1146. The larger churches and monasteries had resources of their own (for example, rents from peasant tenants and shopkeepers), but these seem to have been barely adequate for ordinary maintenance and operations. In any case, such income-producing properties were often subject to financial penalties or outright confiscation by local rulers. Repairs and rebuilding required something extra, and such funds could only come from the wealthier laity. At one point Michael comments briefly and ­without explanation that his flock had no rich or powerful members. If that was actually the case, we must surely think of middling merchants and bureaucrats – men with something to spare but certainly no vast fortunes. Modest as their wealth may have been, however, it was crucial to maintaining the infrastructure of the Syrian community in a turbulent age. No less important – possibly more so – would have been the moral example of the lay notables. The very presence of men of standing assured the poorer laity (peasants, artisans, bakers, and so on) of the Syrian community’s continuing vitality. In contrast, the conversion of these men to Islam (a particular danger in the case of Christian regime officials), or the degrading of their economic and social circumstances, was surely demoralising. Most people then as now find it hard to stick with a community in visibly rapid decline. They instinctively look for a group that is strong enough to protect

ada pti ng to musli m rul e  | 69 them or least advocate for them, and cohesive enough to provide structure and order in their everyday social relationships. With the above comments as background, let us look more closely at the single most important issue confronting the Syrian Orthodox ­community – to wit, the fact that the great bulk of its people lived under Muslim rule. To begin with, the main urban centres of Syrian Orthodox life were all held by Muslim (usually Turkish) amirs. In Diyar Bakr, east of the Khabur River, the fortress cities of Amida, Mardin and Mayyafariqin had been under Muslim rule since the early Arab conquests. Further west, Melitene had been conquered by the Turks (the emergent Danishmendid confederacy) in 1101. Danishmendid rule did not bring peace and stability, however. Melitene’s strategic location as a major gateway between Anatolia and the western Jazira (Diyar Mudar) ensured that it would constantly be the target of Armenian, Frankish, Byzantine, Danishmendid, and eventually Rum Seljuq ambitions, and the city was besieged and sacked more than once. The great city of Edessa suffered a unique fate. It was for many centuries the largest and most important Christian centre in northern Syria and the Jazira. During the Crusade era it lay progressively under Byzantine, Armenian and Frankish rule until it was overwhelmed in 1144 by the atabeg Zengi, the most fearsome Turkish warlord of the early twelfth century. Upon his death in 1146, Count Joscelin II’s bold but disastrous bid to reclaim his capital left the city half in ruins at the hands of Nur al-Din Mahmud. Thereafter it remained at peace for the balance of the century, but it only began to be restored (and then only partially) a half century later under Ayyubid rule.8 What then was the character of Muslim domination? As the above sketch suggests, simple answers are not to be hoped for. First of all, Muslim rule in the Jazira was highly fragmented and constantly changing, due largely to unremitting conflict among a rotating cast of Turkish amirs as each struggled to carve out and hold on to a domain of his own. Second, and perhaps more important, we do not have the sort of detailed fiscal and administrative data required to develop concrete and reasonably objective conclusions. What we do have are the narratives of our Syriac chroniclers, especially the Syrian Orthodox patriarch Michael and the 1234 Anonymous Chronicle. In this way we know how the situation was perceived by a high-ranking churchman, and by a well-informed clerical chronicler – or at least how they wanted their

70  |  syri a i n crusa d e r time s readers to perceive it. As befits his office, Michael in particular is writing sacred history, in which the events of his age embody the fearful judgement of God against an erring community. His information is invaluable (and often all we have), but his chronicle is anything but a disinterested register of events. The language of the 1234 Anonymous is drier and more objective, but the underlying agenda of his text is not notably different. On the other hand, these comments do not mean that the statements of our two chroniclers have no basis in fact. In the first place, their statements about the impact of Muslim rule on Christians (Syrian and otherwise) are important facts in themselves, for those statements are the embodiment of their understanding and interpretation of that impact. Likewise, they have much to say (much of it far from complimentary) about their fellow Christians – Armenians, Franks and Byzantines – who were (unlike the Syrian Orthodox) organised for war and were thus an integral part of the regional struggle for power. All this gives us whatever insight we can have into their subjective experience as members of a numerous but powerless community. On another level, the Syriac chroniclers’ accounts of rulers and political or military events can be tested against our numerous Frankish, Muslim, Armenian and Byzantine sources. There is certainly some misdating and confusion, especially regarding far-off events – for example, in Egypt, Jerusalem or Constantinople. Likewise, their descriptions of battles or the conquests of cities are highly stereotyped, and the numbers of dead and captives that they put forward are simply not usable. As a general narrative of events, however, their testimony stands up fairly well, and they display a far better comprehension of Muslim politics than do most Frankish or Armenian writers, with the partial exception of William of Tyre. The Danishmendids of eastern Anatolia are probably more fully represented in the Syriac chronicles than in Muslim sources, and the Artuqid principalities of Diyar Bakr at least as well. Michael and the 1234 Anonymous are independent of each other, but they clearly share some common sources. The degree to which they might have drawn on Muslim Arabic or Crusader historical writing demands further investigation. Both writers would have known Arabic on some level, though probably not with the mastery displayed by their successor Bar Hebraeus. Michael certainly did not have access to the main synthetic works in Arabic by Ibn al-Athir, Ibn al-‘Adim or Abu Shama, since they were composed only

ada pti ng to musli m rul e  | 71 after his lifetime. However, the 1234 Anonymous may well have consulted Ibn al-Athir’s al-Kamil fi’l-ta’rikh, which was completed and in circulation while he was still active. Moreover, Michael clearly had close contacts with the Franks of Antioch as well as refugees from Edessa. Although it seems unlikely that he could read Latin or Old French, he was certainly able (either directly or through translators) to gather oral testimony from Frankish informants. The 1234 Anonymous seems also to have had good sources of information on Frankish affairs, though he is not as revealing as Michael in this matter. But whatever information Michael and the 1234 Anonymous may have gathered from Arabic and Frankish sources, they would have been able to test it against their own experience, and they deployed it for their own purposes. The Syriac accounts permit some general remarks about the relationships between the Syrian Orthodox population and those who dominated them. The Byzantines are uniformly portrayed with loathing and contempt. They are first of all despicable heretics, intent on oppressing the Syrian Orthodox whenever opportunity offers. Almost as bad, their occasional intrusions (for example, in the expeditions led by the emperors John II and Manuel) are disruptive at best, and Byzantine armies in northern Syria never achieve any important objective.9 In contrast, the chroniclers’ attitudes toward the Franks are ambivalent. By Michael’s adult lifetime these had become as much a part of the political and social landscape as the Turks – they had after all arrived only three decades later – and he never questions their settlement in Syria and their right to rule the lands they held. The Franks were famously courageous if often reckless warriors, and at least down to the crisis years between 1144 and 1149 (between the fall of Edessa and the disastrous battle of ‘Inab), they provided the towns and monasteries in the western Jazira and northern Syria some measure of protection from Turkish raids.10 Finally, though the Franks were themselves heretics, they seldom interfered with the Syrian Orthodox Church and in fact took some care to support its hierarchy, especially in Edessa. On the other hand, the Franks were far from reliable allies and protectors. They occasionally attacked Syrian villages or seized the treasures of the churches and monasteries in order to finance their ongoing wars against the Turks, or simply to enrich themselves. They were, in short, a force of disruption as much as stability.

72  |  syri a i n crusa d e r time s Much the same was true of the Armenian warlords – variously allies and rivals of the Franks – who dominated Cilicia and the mountains to the north and east. But here at least there was no theological animus, since the Armenians and Syrian Orthodox shared broadly the same miaphysite creed. Because the churches of the two communities were by now defined largely by ethnic-linguistic boundaries, the fact of parallel hierarchies was not a point of contention. Finally, there were the Turkish dynasts and warlords, who dominated most of the Syrian Orthodox heartland. Their territories were highly fragmented and localised, though ‘Imad al-Din Zengi gradually emerged as the paramount power among them as he built a powerful state based on Mosul and Aleppo in the 1130s and 1140s. However, only the consolidation of power under his son Nur al-Din in the lands south of the Taurus in the 1150s and 1160s, and under the Rum Seljuq sultan Kilich-Arslan in central and eastern Anatolia at more or less the same time, brought a degree of equilibrium. In the first decades of the twelfth century, every amir had followed his own policy vis-à-vis his Christian (mostly Syrian) subjects, but there are some relatively consistent patterns. In the chaos of the early twelfth century, Syrian and Armenian towns and monasteries were of course constantly subject to raiding, pillage and occasional massacres. In Michael’s presentation, however, the Danishmend Ghazi of Melitene (1101–4) and some of the Artuqids in Diyar Bakr are praised for the relative security they provided, especially their protection of churches and monasteries, and their light hand fiscally.11 The Syrian Orthodox had never been a privileged people under any regime, and they did not expect much. Steady taxation without undue harshness, allowing the church to conduct its own internal affairs without interference (unless the Patriarch or one of his rivals sought it) and to maintain its physical infrastructure, permitting people at large to go about their everyday lives – that was enough. Apart from a few mid-rank bureaucrats (as mentioned previously), the Syrians were not part of the power structure and had no desire to be. For many years, the Patriarch in fact preferred to reside in Amida, under the reliable protection of a Muslim prince, rather than to jostle for status and autonomy in the Christian centres of Edessa and Antioch ruled by the Franks. Eventually, even Amida no longer seemed viable, and when Michael

ada pti ng to musli m rul e  | 73 became Patriarch he moved his residence to the great monastery of Mar Bar Sauma near Melitene, under Danishmendid and then Rum Seljuq rule. Overall, at least for our two chroniclers, if not for every member of their community, Muslim rule was not only acceptable but at times might be preferable to that by Christians, especially those who professed the wrong creed. Michael makes this point with exemplary clarity in referring to an earlier episode: After a certain time [1031], the Byzantines reigned again in [Edessa]. Then when the Muslims (Syriac, Tayyaye) who were there took their children to flee, the Christians also took their own children and fled along with the Muslims, because they had adopted Arab ways in their speech and writing, and felt horror toward the Byzantines because of their heresy and ill-conduct.12

Admittedly, this passage refers to a time some forty years before the Turcoman invasions, when the Jazira was under Arab and Kurdish rule, but it is clear that his sentiments remained unchanged. These are the generalisations we can draw from Michael and the 1234 Anonymous, though of course their narratives are sufficiently dense and detailed to require far more nuance than we can offer here. However, certain passages in both chronicles diverge from a linear narrative of political and military vicissitudes and point to an acute and deep-seated sense of anxiety and loss. Two topics in particular seem paradigmatic: the fall of Edessa in 1144/1146, and the character of Nur al-Din Mahmud. The fall of Edessa in 1144 of course had an enormous impact on Christian writers in both the East and West. Both Michael and the 1234 Anonymous give lengthy accounts of Zengi’s conquest that are among the most vivid narratives of a major urban siege that we possess from this period. Most Muslim accounts of this event are rather concise and matter of fact; only Ibn al-Athir’s lengthy and highly rhetorical narrative in the Ta’rikh al-bahir provides a real counterpart to the Syriac narratives, though of course in a triumphalist register.13 All three accounts concur in emphasising the abrupt contrast between the destruction and rapine of the first two days after the fall of the city, and Zengi’s sudden decision to halt the violence, repair the damage, liberate most of the captives and establish security for its

74  |  syri a i n crusa d e r time s inhabitants, who were at this time almost entirely Christian. Zengi’s decision came as a surprise both to his troops and the despairing Edessenes, since he was notoriously capable of cold-blooded brutality, as when he massacred the Muslim garrison of Baalbek after it had surrendered to him on terms in 1138. Ibn al-Athir ascribes it both to his clemency and to his admiration for the city; a place so wealthy and handsome, an ornament to his empire, could not be consigned to destruction. The Syriac chroniclers emphasise a more personal element in Zengi’s decision – the crucial intervention of the Syrian Orthodox Bishop Basilius, who artfully appealed to his clemency and justice. In general, the Syriac chroniclers viewed Zengi with fear and anxiety (a point of view they share with William of Tyre) but also with considerable respect and occasional admiration. In their accounts he quite consistently treated his Christian subjects with moderation and the clergy with great respect unless they got caught in the crossfire of his wars. In contrast, both chroniclers excoriate Edessa’s abandonment and betrayal (in 1146 even more shamefully than in 1144) by the Franks, chief among them the reckless Count Joscelin II, but also by the Antiochenes.14 Both portray the events of 1146 as far more catastrophic and permanent than the initial conquest two years earlier, since this time a vengeful Nur al-Din put no restraint on the rampaging Turkish soldiers, nor did he make any effort to rebuild the city after the destruction and massacres ended. But in looking back on these sad events, the Syriac chroniclers describe their significance in quite different ways. The 1234 Anonymous takes the opportunity not simply to grieve, but to make an invidious comparison between two segments of the larger Christian community: At this time the Christians who dwelt east of the Euphrates, especially the inhabitants of Mardin, Shabakhtan and Siverek, distinguished themselves by their mercy and benevolence . . among them the venerable Mar Yokhannan, Bishop of Mardin, whose original home was Edessa. One cannot give him the praise due him; his name is inscribed in the heavenly Jerusalem. As to the lands west of the Euphrates, their inhabitants showed no mercy [towards the victims of Edessa], but only harshness and severity, hardness of heart, and perversity of thought – especially the priests, monks and bishops.15

ada pti ng to musli m rul e  | 75 The author’s personal agenda, which contrasts the piety of the Tur ‘Abdin (a wholly Syrian Orthodox region) with the crass indifference of the mixed Christian communities (Latin, Greek, Armenian, as well as Syrian Orthodox) lying between Antioch and Samosata, might be worth exploring. In any case, it is noteworthy that much of his lengthy account is devoted to a bitter and angry critique of the conduct of his co-religionists. The ruin of Edessa was one thing, but the way in which men who professed themselves Christians had brought it about or responded to it was a far more important issue. Michael is a much more subjective and emotionally charged writer than the 1234 Anonymous. His account of the fall of Edessa is infused by a profound sense of loss, a feeling that here Christianity had suffered a painful wound that would never heal – an elegiac tone that continues to infuse his work long after his narrative has turned to other events. He conveys all this in his own way, of course. He is a churchman, and for him the fate of Edessa can only be understood as a richly merited punishment for sin. The fate of Edessa indeed provides the paradigm for his sense that he lives in, and is himself part of, a Christian society so deeply steeped in sin, neglect and turpitude that it is perpetually subject to divine judgement. To Michael’s own narrative, he appends two disquisitions by the theologian Dionysius of Amida, composed precisely to justify the ways of God to man.16 Interestingly, Michael never portrays Muslims as punished for their sins, even when bad things happen to them. Apparently, they lie outside the covenant of divine grace and chastisement, and are more the instrument than the object of God’s unfathomable justice. The punishment meted out to Edessa was so painful not only because it was the largest and most prestigious Christian centre in the Jazira, but even more because it was literally a sacred city. According to an ancient legend recounted by Eusebius, and thus dating at least to the end of the third century, Edessa’s unique link to Christianity went back to the early first century, when King Abgar sent a letter to Jesus imploring him to come and heal the incurable disease from which he suffered. Jesus demurred, saying he had other work to accomplish at the moment, but promised to send a disciple in due time. After his accession he fulfilled that promise by sending Thaddeus, through whom Abgar’s illness was indeed healed. And so, Michael laments,

76  |  syri a i n crusa d e r time s The Assyrian boar prevailed and devoured the delicious grapes. Ah, what a bitter tale! The city of Abgar, the friend of Christ, has been trampled underfoot by our iniquities – the priests massacred, the deacons sacrificed, the sub-deacons burned, the temples pillaged, the altars overthrown.17

Michael’s concluding description of Edessa, truly nightmarish, speaks for itself: Edessa remained deserted: a vision of terror, wrapped in a black cloak, soaked (ivre) with blood, infected by the cadavers of its own sons and daughters. Vampires and other savage beasts ran into the city to feast during the night on the flesh of the massacred. It became the abode of jackals, for no one entered there save those who were digging for treasure. The people of Harran [a predominantly Muslim town] and the rest of its enemies tunnelled into the churches and the houses of notables, saying, ‘Aha! Aha! Our eyes have seen it!’18

Michael closes his lengthy discourse on Edessa with a capsule history of the city, from its founding by Nimrod, the son of Canaan, down to what was for him the bitter end: ‘After the murder of Zengi, it was totally ruined in the year 1458 (1146).’19 It is as if the enormity of what had happened could only be understood in the light of the city’s long, unbroken history, stretching back into the dimmest reaches of antiquity. His portrait is plainly a rhetorical construct and may well exaggerate the true extent of desolation. No matter – for him Edessa is a symbol, Israel fallen yet again into the vile hands of the Assyrians. His purpose is moral truth, not literal reportage. A second theme – namely the character of Nur al-Din and his impact on the Christian communities of his realm – is intimately linked to the fate of Edessa because of his role in the destruction of that city, but it is also a major topic in itself. For medieval Muslim chroniclers, as indeed for most modern scholars, Nur al-Din has seemed almost the ideal Muslim ruler – pious, austere, just, devoted to the welfare of his subjects, deeply committed to the task of defending Muslim territories against the Franks and, if possible, expunging them from the land. Certainly, he built his empire largely by subduing rival Muslim princes, but he could hardly have pursued his larger goals if he had been content with the small patrimony – northern Syria and

ada pti ng to musli m rul e  | 77 Diyar Mudar – that he controlled after his father’s assassination. Nur al-Din often comes across as a rather dour, boring fellow – as a personality he is no match for the far more colourful Saladin – but as a ruler he has always seemed the superior figure. That portrait is shared even by William of Tyre, who understood as well as anyone how relentless an enemy Nur al-Din was, and how irreparable the damage he had wreaked on the Frankish settlements in Syria: ‘a mighty persecutor of the Christian name and faith … He was a just prince, valiant and wise, and, according to the traditions of his race, a religious man.’20 The 1234 Anonymous gives us a portrait generally similar to William of Tyre’s. However, he departs from William in pointing to Nur al-Din’s severity toward the Christians of his own territories – something that passes unnoticed by William and barely emphasised by the Muslim chroniclers. The latter do mention a few specific cases, but they do not explicitly connect these to any general policy of strict adherence to the severe provisions of the Pact of ‘Umar or more generally the normative ordinances of sharia. The Syriac chronicler, in contrast, not only makes general assertions about this issue but gives quite detailed accounts of specific cases. Thus, when Nur al-Din seized control of Mosul and its hinterland upon the death of his younger brother Qutb al-Din Mawdud in 1170, he took immediate steps to rectify the latter’s lax policy: When [Nur al-Din] arrived in Nisibin, the Muslims were in a tumult over the Christians, claiming that they were restoring their churches. Then Nur al-Din commanded that all new construction be destroyed. Thus they demolished certain parts of the sanctuary of Mar Ya‘qub, which belonged to the Nestorians, and likewise parts of a sanctuary belonging to the Jacobites [that is, the Syrian Orthodox] … Once in Mosul … he imposed new and painful rules on the Christians. He ordered that where a new church had been constructed, it should be destroyed. Likewise, he commanded Christians to wear a girdle [around their vestments]. He excluded Christian clerks from state offices and decreed that no Christian should ride a horse or a mule with a saddle. All the [Christian] secretaries were expelled from the palace, except the Shammas ‘Abdun, who was an old Syrian, intelligent, wise, and rich both in gold and knowledge.21

78  |  syri a i n crusa d e r time s In this case at least, Nur al-Din’s reforms were short-lived. As our chronicler notes with satisfaction, as soon as he left the city, the new atabeg (his nephew Sayf al-Din Ghazi II) lifted all the new impositions and things went right back to the way they had been in the days of his father Qutb al-Din. Sayf al-Din obviously thought that contented tax-paying subjects would serve him better than strict religious principle. But overall the 1234 Anonymous does not dwell too much on Nur al-Din’s rigour towards his Christian subjects, and in summing up his reign he portrays him with a cool respect, in a manner that even Nur al-Din’s great champion Ibn al-Athir could hardly have objected to: He was hypocritical and cunning. However, he observed the laws [of his religion] such that he never drank wine nor permitted others to drink it. He contemptuously did away with the songs, games and dances which are customarily performed for kings. They say that no one ever saw him laugh. Nor did he eat with anyone else, but always alone and only once per day. He did not debauch himself in marriage, or multiple marriages, in contrast to the usual excess of Turkish kings. He wore simple attire, and was serious and calm. He always observed the [Ramadan] fast, and studied the books [of his religion]. He practised justice, gave alms and showed mercy to poor Muslims and pious Christians [!]. He strove to strengthen everywhere Muslim laws and practices. He suppressed every sort of unlawful tax and injustice throughout his domains … Whenever some injustice was committed, he repaired it and made compensation as soon as he learned about it. He did not punish the guilty without due process and certain witness. His military encampments were without noise, games and shouting; silence and tranquility held sway there.22

Michael confronts us with a strikingly different portrait – striking because even as he mimics in many respects both the usual Muslim narrative and that of his anonymous colleague, he gives it a dark, ironic turn. In Michael’s eyes, Nur al-Din was no doubt austere and pious, just towards his Muslim subjects, and a relentless warrior against the Franks. But he could be as brutal and bloodthirsty as his father ever was, most notoriously when his forces overwhelmed Edessa in 1146. And whereas Zengi could display not only formal tolerance but even a rather relaxed attitude toward his

ada pti ng to musli m rul e  | 79 Armenian and Syrian subjects, Nur al-Din was uniformly severe. Michael parallels the 1234 Anonymous in stressing his strict application of the most demeaning rules of the so-called Pact of ‘Umar and his efforts to drive Christians out of public life. He develops this theme at far greater length and with a more heated rhetoric than the later chronicler, however. Sometimes Michael repeats the same motifs, but he adds interesting detail and some new anecdotes along the way. For example, Nur al-Din’s agent during his occupation of Mosul in 1170 was a favourite jurist, Ibn Abi ‘Asrun (Syriac, Bar ‘Azroun), who was supposed to survey all the churches in Mosul and demolish any that were recently constructed or repaired. But it turned out that the learned and pious Ibn Abi ‘Asrun was quite amenable to pecuniary conversations; any plausible claim of antiquity would do in return for an appropriate consideration.23 More striking than Michael’s description of Nur al-Din’s policy or any additional details, however, is his portrait of the man. Nur al-Din was eager to link his regime to the four rashidun caliphs and the pious Umayyad ‘Umar II b. ‘Abd al-‘Aziz (r. 717–20). Michael not only notes this linkage, but arguably models his account of Nur al-Din on the portrait of ‘Umar II which had been articulated in the Syriac historiographic tradition in the eighth and ninth centuries, and which was fully preserved in Michael’s own chronicle.24 Both Umar II and Nur al-Din are presented as men of undoubted piety, probity and austerity; both exhibited strict justice towards their Christian subjects, but both also subjected them to legal inferiority and social humiliation. However, Michael’s portrait does not stop with this comparison. In fact, his Nur al-Din is both a hypocrite and, in his own eyes, a kind of messianic figure, divinely chosen to usher in the final triumph of Islam. So far as I can determine, this portrayal is unique to Michael; I find no echoes elsewhere, not in the 1234 Anonymous or William of Tyre, and certainly not in the Muslim Arabic tradition. Michael’s first allusion to Nur al-Din’s grandiose self-image occurs in his narrative of the expedition to Mosul in 1170, where he comments that he ‘assiduously observed the prohibition against wine, nor did he ever miss the time for prayer. The Muslims indeed called him “prophet”. That is why he was so hard on the Christians and agreeable to the Muslims (Syriac, Tayyaye).’25

80  |  syri a i n crusa d e r time s That terse if suggestive comment is developed at length a bit further on, and the passage in question requires to be cited in full: As our discourse has partially made clear, he allowed himself to exploit the vainglory of the Muslims, who considered him even to be a ‘prophet’. Thus, he made every effort to molest the Christians, in order to pass among the Muslims as one who was dedicated to observing their laws. Thus when, beyond Syria and Egypt [hence at the very end of his reign, post-1170] he also dominated Assyria [the region of Mosul], he rose up in his arrogance as though he already ruled all the earth. Then Satan excited him to destroy all the Christians. To this end he wrote letters and sent messengers to the caliph [al-Mustanjid, r. 1160–70], saying: ‘This is written in the Qur’an, where we find everything that Muhammad said when he uttered this prophecy to them: “the Muslims will reign for five hundred years, during which they will not maltreat the Christians.” But these years are now concluded. Henceforth it is proper that the Christians should disappear entirely from the Muslim realm, so that anyone who does not become a Muslim should be put to death.’ In his letter to the caliph, he stated further that he was minded to come before his presence. This frightened the caliph; he understood that [Nur al-Din] cunningly intended to come and drive him out, as he had already done to the ruler of Egypt, and to become caliph himself. And for that the caliph despised him. [When al-Mustanjid died and was succeeded by his son al-Mustadi’ (r. 1170–80), the latter] wrote a response to [Nur al-Din]: ‘It does not belong to you to be called ‘prophet’ nor to establish laws like God. You do not even comprehend correctly Muhammad’s statement about the passage of years. God does not prescribe that we should kill men who have done no wrong.’ Nur al-Din was covered in confusion; he sent messengers and gifts, asking to come venerate the tomb of the late caliph [al-Mustanjid]. All this profited the Christians [of Baghdad], for the new caliph was confirmed in his belief that Nur al-Din perfidiously sought to enter the city in order to reign there. For that reason the caliph answered him with menaces and forbade him to come.26

This is altogether a most remarkable narrative. To the best of my knowledge it has no parallel in any of the Muslim sources, nor is it picked up by the 1234

ada pti ng to musli m rul e  | 81 Anonymous or by William of Tyre. Possibly it was based a popular rumour circulating among Syrian Christians, generated out of their growing anxiety in the face of Nur al-Din’s victories over the Frankish states, his power as the master of a vast and expanding empire, and his severity towards them. Or just possibly Michael simply invented it, as a kind of parable or cautionary tale. There may even be some small germ of truth in it, for Nur al-Din was indeed widely revered among many Muslims, for his personal austerity, his commitment to justice, and his vindication of the cause of Islam against its enemies. And he might well have come to regard himself as one whom God had especially called to the service of Islam in a time of crisis (though in his public propaganda he presented himself simply as one who strove to meet the obligations of justice and struggle incumbent on every Muslim ruler). Yet in Michael’s eyes even that reverence of Nur al-Din’s subjects for him is not wholly innocent and unambiguous, for he ends his account of Nur al-Din with the following comment: He considered himself to be like Muhammad and expected that the Lord would speak with him as with Moses. That is why the Muslims, who had grasped the vanity of his desires, called him ‘prophet.’ They repeated constantly, ‘Today or yesterday, he was seen in Mecca, or at some other mosque’. And he accepted that [that is, naively and at face value].27

So, in the end, Michael’s Nur al-Din is anything but a heroic figure, or at least he is a gravely flawed one. On the contrary, he seems deluded, the victim and tool of Satan, led astray by his grandiose dreams of power. He is even the butt of jokes among his subjects, but fails to notice. In conclusion, both Michael and the 1234 Anonymous view the fall of Edessa and the rise of Nur al-Din as the two closely linked crises that defined the fate of the Syrian Orthodox community (and Christianity generally) in Syria and the Jazira. Both were crucial events in their own right, but they were even more important in what they presaged for the future – the loss or degradation of the key centres of Christian life, and the threat to Christian autonomy and vitality posed by the emergence of a politically powerful and ideologically mobilised Islam. The two chroniclers convey this message in quite distinct ways – the 1234 Anonymous relatively restrained and objective, Michael highly rhetorical, intense and emotional. This difference in

82  |  syri a i n crusa d e r time s part reflects contrasting authorial strategies. The 1234 Anonymous presents himself as a recorder of fact, Michael is always the preacher. However, the difference may also be grounded in the very different moments at which the two men were composed their works. Michael concluded his chronicle in the wake of Saladin’s near-destruction of the Frankish states and the highly uncertain future which now faced Syrian Christians. In contrast, the 1234 Anonymous wrote four decades later, after the partial recovery and apparent stabilisation of the Frankish states, and under the relatively lenient rule of Saladin’s successors. Facing (so far as he knew) no imminent catastrophe, he could afford a more Olympian perspective. Notes   1. An important exception is Eddé, 2006, 153–180, which presents in a precise and critical manner most of the scattered information that can be gleaned from our sources. I am grateful to Prof. Eddé for bringing this article to my attention. On a more general level, see Hillenbrand, 1999, esp. 407–420, with numerous citations from Arabic texts. For Syrian Christians under Frankish rule, MacEvitt, 2008 is indispensable, but he deals only incidentally with those subject to Muslim regimes. For the twelfth century, Elisséeff, 1967 has numerous comments here and there on non-Muslims but no dedicated chapter; Mouton, 1994 does have a brief chapter (345–350) which emphasises how little is actually known. For the thirteenth century, we have more information, but this can be applied to the very different circumstances of the twelfth century only with great caution: Eddé, 1999, 452–472; Pouzet, 1988, 305–338.   2. It is impossible to review here all the Muslim Arabic sources for twelfth-century Syria. However, only a few of these are of real value for the subject of this chapter. By far the most important source for political life in Muslim Syria down to 1160 is Ibn al-Qalanisi, 1908; partial translations by Gibb, 1932 and Le Tourneau, 1954. The Jazira is covered in a slightly later work, the fascinating but rather chaotic work of Ibn al-Azraq al-Fariqi, 1990. In contrast to Ibn al-Qalanisi, Ibn al-Azraq does contain some important nuggets of information about native Christians in the Jazira. The famous raconteur Usama b. Munqidh has much to say about the Franks but very little about Syrian Christians, though he was plainly aware of them. They are rarely part of the stories he wants to tell: Usama b. Munqidh, 1929. The best English translation is Cobb, 2008. Among thirteenth-century Muslim historians, our best information (such as it is) is

ada pti ng to musli m rul e  | 83 found in Ibn al-Athir, 1965–7 and Ibn al-Athir, 1963. The Kamil for the period 1097–1233 has been translated by Richards, 2006–8. Among Christian sources, two Syriac texts are fundamental: Michael the Syrian, 1899–1914 and Anonymi auctoris chronicon ad a.d. 1234 pert inens, 1974. Down to 1130, the Armenian chronicle – Matthew of Edessa, 1993 – is distinctly inferior to those by his Syriac counterparts but does contain valuable information. Finally, William of Tyre brings a well-informed Frankish perspective to the period, despite his lack of interest in the Syrian Christians living under Muslim rule: William of Tyre, 1943.   3. Commonly called Jacobites, after the sixth-century bishop and preacher Jacob Baradaeus.   4. Michael the Syrian, 1899–1914, 3: 222: ‘The Franks … had bishops in their states. The priests of our church were in their midst, without being persecuted or molested, for though the Franks concurred with the Greeks as to the duality of natures [in Christ], they nevertheless differed from them in faith on many points, and they were far apart from them in their usages … We speak of this here in order to show that the Franks … never raised any difficulty on the subject of faith, nor in order to reach a single formula among all the peoples and languages of the Christians. Rather, they regarded as a Christian anyone who adored the cross … For their part, the Turks, who had no grasp of the sacred mysteries and indeed considered Christianity as an error, did not have the habit of informing themselves about professions of faith, nor of persecuting anyone for his creed, as do the Greeks, a vicious and heretical people.’   5. Things had actually been much the same in the eighth and ninth centuries. The chronicler and patriarch Dionysius of Tell-Mahre (d. 843), whose work is transmitted to us via Michael and the 1234 Anonymous, is extremely eloquent on this matter. On dissident bishops, the ‘ecclesiastical chronicle’ appended to the end of 1234 Anonymous (182–260] is an almost unbroken catalogue of dissension and schism. Michael, vol. 3, is likewise a bottomless well of such matters; see 243–244, 246–248, 251–252 for the machinations of one Basilius, Bishop of Jayhan, ‘a most skilful and cunning man’. Michael also mentions several ‘converts’, of whom the most interesting is one Aharon of Sijistan, who was at one point Bishop of Haditha, and successively passed through Syrian Orthodox, Muslim, Chalcedonian and Maronite phases in his spiritual journey. See Michael the Syrian, 1899–1914, 3: 291–292.   6. Ibn al-Athir names a few Christian officials at the caliphal court, though those mentioned in fact were converts to Islam, presumably in furtherance of their

84  |  syri a i n crusa d e r time s careers: Amin al-Dawla Abu Sa’d, the highly regarded chief of the chancery (diwan al-insha’ ), who died in 497/1104–5 and had converted to Islam in 484/1090 (Ibn al-Athir, 2006, 10: 377–378, tr. Richards, 2006–8, 1: 82); and his nephew Abu Nasr, died 498/1105, whose family remained Christian and thus could not inherit from him (Ibn al-Athir, 1965–7, 10: 397; Richards, 2006–8, 1: 94). Saladin’s brother al-‘Adil Sayf al-Din appointed a Christian secretary, Sani‘at al-Mulk b. al-Nahhal, when he was made governor of Aleppo in 579/1183, but on condition that he accept Islam: Ibn al-‘Adim, 1951–1968, 3: 75.   7. Among many such cases, the most striking is the systematic pillage and vandalism inflicted on the great monastery of Mar Bar Sauma near Melitene by the erstwhile Count of Edessa, Joscelin II, in 1148. Having lost his capital and much else besides, he no doubt hoped to restore his fortunes from the treasures held there. Adding insult to injury, he also extorted a payment of 10,000 dinars from the monks and their tenants. After Joscelin’s departure, the Metropolitan of Kaysum collected 5,000 dinars from the Syrian Orthodox faithful to restore the monastery. Joscelin II is universally portrayed by the Frankish, Muslim and Syrian sources as a reckless and unprincipled scoundrel: 1234 Anonymous, 113–115; Michael the Syrian, 1899–1914, 3: 283–288.   8. After 1146 Edessa was replaced as the chief city of Diyar Mudar by the predominantly Muslim town of Harran, and in the thirteenth century Harran was clearly the administrative centre of the Ayyubid domains east of the Euphrates, the so-called al-Bilad al-Sharqiyya. The fate of Edessa is detailed later in this chapter.  9. The ambitious expedition of John II Comnenus in 1137 is paradigmatic: Michael the Syrian, 1899–1914, 3: 245; 1234 Anonymous, 81–86. 10. Michael is highly critical of many Frankish chiefs, but this is balanced by his admiration for their bravery and daring. His paean to the Knights Templar (whom he calls ‘Phrer/frères) for their charity and discipline as well as their military prowess is well known. On the founding of the order and its institutions: Michael the Syrian, 1899–1914, 3: 201–203; on their charity in a famine in Jerusalem in 1120, 207–208. 11. On Danishmend Ghazi in Melitene, see Michael the Syrian, 1899–1914, 3: 188. 12. Ibid., 3: 280. 13. Ibid., 3: 260–264; 1234 Anonymous, 89–95, 99–102 (perhaps the most detailed account); Ibn al-Athir, 1963, 66–70. Cf. the narratives of Ibn al-Qalanisi, 1908, 279–280 (Gibb, 1932, 266–268; Le Tourneau, 1954, 266–267); and Ibn

ada pti ng to musli m rul e  | 85 ­al-Athir, 1965–7, 11: 98–99, tr. Richards 2006–8, 1: 372–373). Ibn al-Athir was, of course, fully aware of the political and strategic importance of this victory. 14. Michael the Syrian, 1899–1914, 3: 270–272; 1234 Anonymous, 104–111. 15. 1234 Anonymous, 110–111. 16. Michael the Syrian, 1899–1914, 3: 265–267, 272–274. 17. Ibid., 3: 261–262. 18. Ibid., 3, 272; Ps. 35:21. Psalm 35 is a plea for deliverance; the immediate context is given in vv. 19–21: ‘Let not those rejoice over me who are wrongfully my enemies, and let not those wink their eye who hate me without cause. / For they do not speak peace, but against those who are quiet in the land they conceive words of deceit. / They open wide their mouths against me, saying, “Aha! Aha! Our eyes have seen it!”’ (cited from New Oxford Annotated Bible [1977], slightly altered). 19. Michael the Syrian, 1899–1914, 3: 281. 20. William of Tyre, 1943, 2: 394. 21. 1234 Anonymous, 126–127. 22. Ibid., 127–128. 23. Michael the Syrian, 1899–1914, 3, 340. 24. For the eighth-century Syrian Christian portrayal of ‘Umar II, see Hoyland, 2011, 215–217. 25. Michael the Syrian, 1899–1914, 3: 340. 26. Ibid., 3: 344–345. 27. Ibid., 3: 353.

5 The Afterlife of Edessa: Remembering Frankish Rule, 1144 and After Christopher MacEvitt

T

he Frankish County of Edessa had a brief life; established in 1098 when Baldwin of Boulogne was acclaimed the ruler by a rebellious faction of the population, it had vanished by 1150, when its remaining lands passed to the Byzantine emperor Manuel I Comnenus. The brevity of the county’s existence, however, is belied by its long afterlife among Eastern Christian communities. Religious Practices, Cultural Values and Historical Perspectives

The Frankish principalities have long been portrayed as alien to Syria, and scholars are now beginning to see them as an integrated part of a complex Levantine world. Little attention has been paid, however, to what their heritage was after their political power was extinguished. When this topic is considered, the conclusion has generally been negative. The condemnation of Steven Runciman, while dated, still resounds in scholarship: ‘When they were ejected, they left the local Christians to bear the wrath of the Moslem conquerors.’1 While in many senses this dire portrait may be true, scholars fail to acknowledge the myriad ways in which the Franks and their influence have lingered. Of all the communities of Syria, the Frankish encounter with Armenians was the most influential on both communities. Two events loom large in Armenian historical memory concerning the Frankish County of Edessa: the conquest of Edessa by Zengi in 1144 and its subsequent recapture by his son Nur al-Din in 1146, and the transfer of the castle of Hromklay to the Armenian kat‘olikos Grigor III around 1150. These two events have marked

the af terli f e of edessa  | 87 the beginning and the end of the collapse of the county in historical narrative: the city of Edessa was the first great loss that signalled (in hindsight at least) the doom facing the remaining lands, and the castle of Hromklay was one of the last remaining strongholds of the county in 1150. Each event was prominent in Armenian historical narratives at different moments; the conquest of Edessa was seared into the memory of the Armenian generation that lived through the events in the twelfth century, but by the thirteenth century, its importance faded. In contrast, twelfth-century Armenian sources have almost nothing to say about Hromklay; it is only in the thirteenth century that the transfer of the castle was deemed significant. The passing of Hromklay from the county to the kat‘olikos in particular tied the memory of Frankish rule to one of the enduring Armenian institutions in Syria. Armenians in the twelfth and early thirteenth centuries were searching to find their footing, both in the political world and historiographically. Armenians had a long and distinctive historiographic tradition, stretching back to the invention of the Armenian alphabet in the fifth century, and continuing throughout the Middle Ages. That tradition was rooted in the Armenian homeland in the Caucasus highlands and in eastern Anatolia, and focused on aristocratic and royal dynasties with ancient roots.2 The twin calamities of Turkish invasions and Byzantine annexation in the eleventh century destroyed both the independent kingdoms and the dynasties that ruled them, but the historiographic tradition survived, supported by new patrons. Armenian chronicles continued to be written in Syria and Cilicia, as well as in historical Armenia, as the locus of Armenian power shifted. How should Armenian authority in this new land be understood? Were the new principalities in Syria and Cilicia a part of the older tradition of Armenian rule, or were they a new beginning more closely tied to neighbouring Frankish principalities? The Frankish county of Edessa was viewed as a cousin of the Armenian principalities which it neighboured and in part subsumed; it was built upon long-existing local political structures and reliant particularly on the Armenian community throughout its existence. Baldwin of Boulogne came to power not through conquest or siege, as was the case in the establishment of Frankish control of Antioch, Jerusalem and Tripoli, but was elevated by the largely Armenian population in a coup against the ruling dux, a Greek

88  |  syri a i n crusa d e r time s named T‘oros (probably Theodore in Greek). He and his successors placed fellow Franks in position of power throughout the county, but Armenian lords also continued to rule important strongholds. In the cities as well, Armenians held local civic offices, and wielded significant authority.3 The Franks intermarried with the Armenian elite; each of the ruling counts of Edessa (except the last) took an Armenian wife, and also married their sisters and daughters to Armenians. The collapse of the county of Edessa extended over a six-year period, and we are fortunate to have two Armenian texts written in the midst of the process. The Lament on Edessa was written around 1145 after the first conquest of Edessa by Zengi but before its brief liberation by the Franks and final conquest by Nur al-Din in 1146, and was composed by Nerses Shnorhali, the brother and successor of kat‘olikos Grigor III.4 The funeral oration of Count Baldwin of Marash was written shortly after the second conquest.5 Nerses’s poem assumed the voice of Edessa herself, addressing in turn Rome, Alexandria, Antioch, Jerusalem and Constantinople, as well as Armenia, lamenting Edessa’s destruction and the dispersal of her citizens. The poem ended with a hopeful if bloody prediction: Edessa will be saved by a new crusade, which will annihilate Muslim armies and conquer all Islamic territory, destroying the Ka‘ba in Mecca and ushering in an age of peace and piety. The poem is well known among scholars of both Armenian literature and historians of the Frankish East. In contrast, the ‘Oration’ has received less attention, particularly in the context of Edessa’s fall. The subject of the oration was a Frank, Baldwin of Marash, who died during the attempt to recapture Edessa from Nur al-Din in 1146.6 He was the ruler of a broad swathe of territory from near Melitene in the north to Marash in the south, and his lands served as a sort of a marcher lordship between the Frankish principalities of Edessa and Antioch. Baldwin himself was probably the brother of Raymond of Antioch, and a son of William IX of Poitou and Aquitaine, an identity we only learn about through the ‘Oration’ itself. 7 The oration was composed by Baldwin’s confessor, an Armenian vardapet (scholar–priest) named Barsegh, about whom we know little else. Barsegh’s oration was preserved in the chronicle of his contemporary Gregory the Priest (Grigor Erets‘), who wrote a continuation of Matthew of Edessa’s chronicle.8

the af terli f e of edessa  | 89 Baldwin’s death came at a pivotal moment; the county of Edessa still existed, and its count, Joscelin II, continued to rule from Tell Bashir, his castle west of the Euphrates, yet the eponymous heart of the county (particularly for Armenians), the city of Edessa, had just been conquered for the second time in two years, and by all accounts, the second conquest was devastating. Whereas the brunt of the first conquest fell upon the Frankish population of the city, the second decimated the indigenous Christian communities as well. Baldwin died in the disastrous retreat from the city, when the Frankish and Armenian forces realised that they could not hold the city against Nur al-Din’s advancing forces. The oration itself is unique; we have no similar eulogies delivered across religious and ethnic lines from anywhere else in Armenian, Latin or Syriac sources in this period that I am aware of.9 The oration, however, was not entirely about Baldwin; it was also a critique of Frankish rule in northern Syria in the wake of Edessa’s loss. The eulogy may not even have been delivered at the funeral: Gregory the Priest explained that ‘Barsegh wrote (greats‘ ) a funeral oration on the occasion of Baldwin’s death as an admonition to those reading (ēnt‘erts‘ozats‘) it and as a record for future generations’.10 The emphasis on writing and reading suggests that this was not delivered orally, and that Gregory saw it primarily not as a eulogy of the dead but as a warning to those who were still living. Indeed, Barsegh addressed himself to ‘all the Christian faithful’ in the territory of Baldwin, ‘whatever their nationality or language might be’,11 and also to ‘kings, princes, all you judges of the earth’.12 The ‘Oration’ was of course written in Armenian, and was intended for Armenian lords as much as for Frankish or Byzantine ones, perhaps even more so.13 Barsegh used Baldwin to exemplify both the best and worst aspects of Frankish rule, and he was an ideal figure to do so. He had no immediate family to claim him or to object to Barsegh’s characterisation of him; according to the vardapet, his wife and children pre-deceased him, and he was alienated from his parents and siblings. Furthermore, Baldwin’s body was never recovered, making the count an abstracted figure even at his own funeral. Baldwin was also ideal because his qualities (at least in Barsegh’s view) exemplified the complexity of Frankish rule in Syria, which was often violently exploitative as well as tightly interwoven with local elites and systems of power. This

90  |  syri a i n crusa d e r time s dichotomy is best captured in Barsegh’s rhetorical question: ‘Who is it that would ask Baldwin who he was and where he came from without feeling his ferocious nature mollified by his modest, sensible and prudent answer?’14 Thus, asking about Baldwin’s identity (Frankish or Armenian? From Syria or France?) produced both evidence of his aggressive, threatening manner, and disarmed the questioner with his reasonable and self-effacing response, which as the next line made clear, was made in fluent Armenian. Nevertheless, Barsegh called Baldwin ‘an example and an admonition to the unrepentant, arrogant and wicked leaders of the Roman [Frankish] forces’.15 Barsegh particularly singled out covetousness as Baldwin’s downfall, accusing him of seizing the lands and villages of his neighbours, and building ‘sumptuous edifices and city walls built at the expense of the sweat of strangers, orphans, widows and unfortunate peasants’.16 Michael the Great, patriarch of the (‘Jacobite’) Syrian Orthodox Church and writer of one of the great chronicles of the twelfth century, similarly portrayed Baldwin as oppressing his own people, saying that ‘he made them like slaves’ in his attempt to rebuild the walls of Kesun in stone.17 Barsegh furthermore accused Baldwin of ‘aggrandisement through extortion, rapine and treacherous behaviour’.18 At the same time, Baldwin’s confessor also displayed considerable warmth towards the prince, calling him a ‘mighty champion and famous soldier of Christ, my beloved Baldwin’,19 characterising him as ‘a handsome young man’ and a ‘wise and prudent prince’.20 The vardapet praised his fluency in Armenian, and averred that ‘he was known by everyone to possess sagacity, will power, maturity, outward attractiveness and all those characteristics of this life that show the magnanimity and excellence of princes’.21 He was the very model of a medieval Syrian prince. Having drawn a complex portrait of Baldwin, Barsegh impersonated him in order that he might confess the sins of the count – particularly his political transgressions. In effect, Barsegh had the dead count address himself ‘to all you kings, princes, all you judges of the earth, and you who are in charge of important affairs and who have control over the management of temporal affairs’ – that is, his colleagues and rivals.22 His crimes, enumerated above, were buttressed with biblical citations, ‘which were taught [to Baldwin] by the vardapet in two languages and were ceaselessly pronounced in an unfaltering, changeless and suitable manner’.23 Barsegh cited Genesis (9:6)

the af terli f e of edessa  | 91 and the Psalms (55:23), condemning perfidy, bloodthirstiness and greed, as well as describing the punishment meted out to sinners – death. Clearly, the Armenian believed that other princes were guilty of similar sins, and Baldwin’s fate and posthumous confession might serve as a warning to them, though Barsegh also suggested that Baldwin’s soldiers might have been his victims rather than perpetrators. Barsegh made clear that the count sinned not only in his action, but also in what he failed to do, which was ‘saving and protecting unfortunate Christians’.24 In making the count’s posthumous confession, Barsegh hoped to gain salvation for his prince. His willingness to name Baldwin’s worst sins was paradoxically a sign of his deep friendship and abiding love. Yet the vardapet’s efforts were apparently in vain, since Baldwin did not repent whilst alive. Barsegh described the count as a ghost: neither dead nor alive, but floating in a liminal state, an ambiguity exacerbated by the loss of his body and the subsequent failure to conduct a proper funeral for him. Yet the vardapet also suggested that Baldwin did manage to redeem himself, earning redemption on the battlefield on which he died. The timing of Baldwin’s death is crucial. He did not die during the seizure of the city by the Frankish forces, nor defending the city from Muslim forces in the five days that followed. He died during the chaotic retreat from the city, when the Franks were followed by a portion of the city’s surviving Christian population. Barsegh praised the compassion Baldwin showed for his soldiers, and his words urging them: ‘On the day of battle, do not avoid seeking a blessed death for the sake of Christ and dying in the defence of Christians.’25 Not only was Baldwin willing to die, he wanted to die, ‘because he wished to be with those [who perished miserably] he also perished’.26 He was, in a word, a martyr. Thus, in death, Baldwin got right what he consistently got wrong in life. He devoted himself to ‘protecting innocent Christians’ and was willing to die to be with them. What does Barsegh’s ‘Oration’ tell us about the last days of Edessa and perception of the Frankish rule in Syria? First of all, let us note that events from Barsegh’s perspective were quite different from ours. While we consider the fall of the city of Edessa as the start of the collapse of the whole county, Barsegh expressed little concern about this. Even though he presented Baldwin as a holy warrior–martyr who died fighting against a detested Muslim enemy,

92  |  syri a i n crusa d e r time s he was not seeking to warn others of the threat that Nur al-Din posed. Nor was the battle over the city tinged with the same apocalyptic doom as Nerses’s poem, though Nerses was writing at the same time in the nearby castle of Tsovk‘. The story of Baldwin was held up as a threat to other rulers; the count died because of his sins as a ruler, and so will other Franks if they do not abandon their grasping greedy ways. But Baldwin’s sins were not just about oppressing his subjects. Barsegh linked the count’s avariciousness to failure to submit to authority, both human and divine. Baldwin failed to keep ‘obedience in the heart’ in respect to those superior to lords such as himself.27 This, then, is a particularly interesting critique: to avoid Baldwin’s fate, other princes needed to submit to their superiors – namely the Franks, most prominently Joscelin II of Edessa and Raymond of Antioch. When we remember that the ‘Oration’ was written in Armenian and must have been directed at Armenian lords as much as at Frankish ones, the ‘Oration’ becomes a surprising endorsement of Frankish rule. The lesson of the calamitous loss of Edessa was greater Frankish–Armenian cooperation, not less. The vardapet’s focus on Baldwin and his claim for his martyrdom is also striking in the context of Edessa’s conquest. Instead of focusing on the great suffering of Edessa’s Armenian population in the wake of the city’s double conquest (as Nerses did), Barsegh instead focused attention on the suffering and death of a single Frankish nobleman. Barsegh’s insistence that the count died for the citizens of Edessa allowed Baldwin to stand for both groups – the covetous Frankish rulers and the dead martyred citizens of Edessa. The pain of the city’s conquest lingered long after; it took decades for Edessa to recover from it, and the city never regained its centrality for Eastern Christians, particularly Armenians. Yet by the thirteenth century, the power of the narrative of its conquest had faded. Samuel of Ani, for example, writing at the same time as William of Tyre, did not even mention it. This was in large part due to a geographical shift in Armenian power. In the late eleventh and early twelfth centuries, northern Syria appeared to be the locus of Armenian autonomy. By the late twelfth century, that had shifted to Cilicia and to the Armenian highlands under Georgian leadership. The city of Edessa under Frankish rule and the story of its conquest fell out of Armenian historical narrative, as northern Syria was no longer a hub of Armenian culture and political power. Yet the triumph of Nur al-Din (and the sultans of

the af terli f e of edessa  | 93 Rum) did not entirely extinguish Armenian authority in Syria. It continued not through secular lordship, but through the ecclesiastical authority of the Armenian catholicate, planted firmly in Hromklay from c.1150 to 1292. The establishment of the catholicate at Hromklay was inextricably tied to two brothers, Grigor III and Nerses Shnorhali, two brilliant scions of the Pahlavuni dynasty.28 The Pahlavunis were one of the few families that managed to survive the collapse of the older kingdoms of historical Armenia in the eleventh century and maintain their status in the new centre of Armenian power in Syria and Cilicia in the twelfth. Their power was at first both secular and ecclesiastical; their great-uncle Vasak had been the Byzantine doukos of Antioch in the late eleventh century, their uncle was vizier for the Fatimid caliph al-Hafiz (c.1135), and their father was lord of the castle of Tsovk‘ near Dülük (Latin Tuluppa, ancient Doliche). Another of their greatuncles was Grigor II Vykaser (the Martyrophile), kat‘olikos of the Armenian Church through the difficult years of schism and invasion (1066 to 1105). While several other kat‘olikoi also claimed authority in this period,29 Grigor II established a Pahlavuni prerogative over the catholicate that lasted nearly a century and a half; the last Pahlavuni kat‘olikos was Grigor VI, who died in 1203. How the castle of Hromklay, situated in the midst of the Euphrates east and a little north of Aintab (modern Gaziantep), came under Pahlavuni control was a story that changed over time, charting the varied ways in which Armenians positioned themselves in relation to Frankish Syria. The standard explanation in secondary literature is that the castle was sold or donated to Grigor III by Beatrice, Countess of Edessa.30 Her husband Joscelin II had been captured by Nur al-Din in 1150, and she sold the remainder of the county to the Byzantine emperor Manuel I Comnenus for a stipend, fearing that she would be unable to defend what was left. However, the accounts about the transfer of Hromklay date to the thirteenth century, nearly a century after the transfer supposedly took place. Contemporary accounts are silent or contradictory. Gregory the Priest, alive in 1150 and writing until 1162, never mentioned the castle though living nearby – instead, he referred to Tsovk‘ as the patriarchal residence in an entry about 1161, suggesting the patriarch (then Nerses IV Shnorhali) still lived there at that time.31 William of Tyre indicated that Hromklay was one of the fortresses handed

94  |  syri a i n crusa d e r time s over to Manuel I Comnenus, and did not associate it with the Armenian Church.32 At some point in the later twelfth century, the castle did come under Pahlavuni control, but we cannot be certain when or how. Even Armenians did not see the transfer as significant until the thirteenth century. The earliest mention of the transfer that I am aware of in any source appeared in the Syriac Chronicle of Michael the Great, who told the story with a distinctly anti-Armenian slant. According to him, Joscelin had established an Armenian named Michael at Hromklay.33 Michael asked for help from the kat‘olikos Grigor, but the Pahlavuni patriarch instead attacked Michael, seized all his possessions and claimed the castle as his own takeover by trickery from another Armenian rather than peaceful transfer from the Franks.34 Michael’s account was translated into Armenian in 1248 under the direction of the kat‘olikos Constantine I Bardzrberdts‘i at Hromklay; perhaps unsurprisingly, the Armenian version narrated the transfer quite differently.35 Two versions of the translation survive, only one of which mentioned the transfer. ‘Version 1’ put the initiative in the hands of the countess Beatrice.36 According to this version of the story, Beatrice sent for Grigor, residing at the ancestral Pahlavuni castle of Tsovk‘, for she ‘wished to cross the sea and return to her parents’. The text seems to presume that her son Joscelin III was not with her, for she insisted that ‘if he lived and returned, it [Hromklay] would be given to him’. Joscelin did eventually show up, and the kat‘olikos paid him for the castle, the younger Joscelin being unsure whether he would be able to rule it in the midst of the Turks. Thus, ‘through the grace of God, [Hromklay] became their eternal see’.37 This, then, is the account that has become normative, though quite different from Michael’s twelfth-century account. The kat‘olikos received the castle through donation, but also paid for it, gaining double rights. The account’s prominence is likely due to the translators, one of whom was Vardan Arevelts‘i.38 Vardan later wrote his own chronicle, which is likely the source of the story in later historiography. This was not, however, the earliest Armenian version. The first such account I have found appeared in the anonymous biography of Nerses Shnorhali, written in 1240, also composed at Hromklay. Much of the account is taken up with the deeds of Nerses’s Pahlavuni relatives, which

the af terli f e of edessa  | 95 depict them beset in a world dominated by the Byzantines. The Franks, in contrast, hardly appear. Following Matthew of Edessa, the author depicted Nerses and his brother growing up under Kogh Vasil, the Armenian warlord who ruled Kesun and Raban until 1113. Even when the anonymous author described the conquest of Edessa, the Franks were curiously erased. ‘At this time,’ the account recorded, ‘the Ismailite Mahmetans had become very powerful, ruling the East as far as the boundary of the Euphrates.’39 Edessa is, of course, east of the Euphrates, and thus the author already imagined the city as surrounded by Muslim-controlled territory. The triumph of the Muslims was due to ‘our sins’, by which the reader would understand ‘the Armenians’. Nevertheless, the account recorded that the ‘the great city of Edessa’ was ruled by the ‘God-loving count Joscelin’.40 While his name clearly betrays his non-Armenian background, the author at no point made any reference to Franks or Latins or any marker of ethnic, linguistic or religious difference that would suggest that Joscelin was not Armenian, or that would separate his authority from the community to whom the ‘Life’ was addressed. And indeed, in many ways he was Armenian. Joscelin’s mother was a sister of Leo of Cilicia, and in some Arabic sources the count was known as ‘Joscelin the Armenian’.41 Whereas Vardan had put the initiative in the hands of the countess Beatrice, the anonymous hagiographer suggested that Nerses and Grigor sought to move to the castle of their own volition. ‘Seeing the weakness of our nation,’ Nerses and his brother ‘feared to remain at the fortress of Tsovk‘ because it was not strong enough to save them from the world-destroying apocalypse.’ They sought support from the ruler of Georgia (at the time Demetrius I), but God instead pointed them in another direction: ‘Through the divine plan, God gave to them the inapproachable castle of Hromklay, on the banks of the Euphrates originating in Eden, by the hand of the blessed wife of the pious Count of Edessa, Joscelin.’42 The story of the transfer of the castle to the Pahlavuni patriarchs was thus a legend, intended to give a satisfactory origin story. Given the twelfthcentury evidence of Gregory the Priest and William of Tyre, it was more likely that the kat‘olikos got Hromklay from Manuel I Comnenus, perhaps during the time that Nerses was negotiating with the Byzantine emperor concerning church union.43 A gift from a pious Frankish countess was much

96  |  syri a i n crusa d e r time s preferable to receiving it from the Byzantines, against whom the Armenians of Cilicia in the thirteenth century defined themselves. Such a switch was particularly ironic given the name of the castle, which meant ‘castle of the Romans (meaning Byzantines)’ in Armenian, Turkish (Rumkale), Arabic (Qal‘at al-Rum), and Syriac (Kal‘ah Rhomaita).44 For thirteenth-century Armenians, however, Hromklay could also be understood as ‘castle of the Franks’, due to the switch in the meaning of the term ‘Roman’ to describe the Franks (from Rome) rather than the Byzantines.45 The story of the Frankish donation may in part reflect that switch in meaning. Furthermore, the story may reflect the ambiguous political position of Hromklay. Perched on a steep headland on the western bank of the Euphrates, the fortress was on the frontiers of competing centres of power: the Armenians in Cilicia, Ayyubid princes in Aleppo and elsewhere, and the Seljuqs of Rum. The Franks still ruled Antioch, but their authority never extended anywhere near the Euphrates. Until it was conquered by the Mamluks in 1292, the patriarchs, and Hromklay, managed for the most part to stay independent of Syrian power struggles. The story of the transfer of the castle from Beatrice of Edessa made this independence clear: the kat‘olikos was not indebted to or reliant upon any political entity, but stood free of all political entanglements. Its only political debt had been to the county of Edessa, which was now defunct. Hromklay’s neutrality was guaranteed by the political irrelevance of the Franks in the thirteenth-century Euphrates region.46 The catholicate itself was transformed by the Pahlavunis in the course of the twelfth century. Traditionally, the kat‘olikos had been patronised by one of the aristocratic dynasties that ruled Armenia. The Turkish invasions and Byzantine annexations of the eleventh century destroyed those dynasties, and the kat‘olikoi were set adrift. At one point in the late eleventh century, there were six rival patriarchs, reflecting the dispersed nature of Armenian political power. The Pahlavunis, beginning with Grigor II Vykaser, consolidated their control of the institution even as the family re-established themselves as local lords in Syria, abandoning their ancestral lands near Ani. As the pillars of traditional Armenian society crumbled, the Pahlavunis sought a rebirth of the church. Born with the name Vahram, the first Pahlavuni kat‘olikos took the name Grigor as his patriarchal name, in homage of the first Armenian

the af terli f e of edessa  | 97 patriarch and the founder of Armenian Christianity, Grigor Lusavorits‘ (Gregory the Enlightener). Not only was Grigor Lusavorits‘ the founder of Armenian Christianity, he was also claimed by the Pahlavunis as an ancestor. Vahram was the first to use the name Grigor since the fourth century, and his Pahlavuni successors relished taking names associated with the early history of the church, racking up five Grigors and one Nerses (grandson of Grigor Lusavorits‘ and his successor). The Pahlavunis also established new means of demonstrating their authority: for the first time since the Grigorids of the fourth century, the basis for patriarchal authority was grounded in dynastic descent. They also elevated their stature through theological and literary production. Grigor II earned the name ‘Vykaser’ for his love of the martyrs, which he demonstrated through translating passiones from Greek into Armenian. Grigor III was a poet and a translator, as was his brother, Nerses IV Shnorhali. Nerses was particularly well known for his hymns, and of course the lament on the fall of Edessa; his grand-nephew and successor Grigor IV Tgha wrote a similar poem on the conquest of Jerusalem by Saladin. Their authority became rooted in Armenian monasteries and in the culture of the vardapets, scholar-priests who were distinctive to the Armenian tradition. Hromklay gave the Pahlavunis the opportunity to retain control of the catholicate by remaining outside of Armenian secular control. The Pahlavunis lost control of the patriarchate when a new Armenian kingdom arose in Cilicia. Grigor VI crowned Leo as king in Sis in 1198, the first anointed Armenian king in over a century. Control over the catholicate was clearly a priority for the new king; Grigor VI was as a consequence the last Pahlavuni patriarch, and the last dynast to sit on the patriarchal throne. Henceforth, the kat‘olikos was selected by the king, not on the basis of dynastic descent. It is thus perhaps appropriate that the last account of the transfer of Hromklay should be written by a member of the ruling class of Armenian Cilicia, Smbat Sparapet (1208–76).47 Smbat was a brother of King Het‘um I (1226–70), and the title of sparapet indicated his position as constable of the Armenian kingdom in Cilicia; his great-grandmother was the sister of Grigor VI Apirat, the last Pahlavuni patriarch. His chronicle gave the longest account of the transfer of Hromklay to the kat‘olikos. For Smbat, the Franks served an even grander purpose than as proof of the patriarchate’s ­independence and

98  |  syri a i n crusa d e r time s as convenient replacements for the Byzantines: they were a proxy for divine benediction, serving even to inspire God. According to him, Grigor grew concerned following the conquest of the city of Edessa and of other cities even closer, such as Kesun, worried how ‘weakened the power of the Christians had become’, and feeling that his current home of Tsovk‘ was no longer secure.48 He desired to return to Ani, where his ancestors had once ruled a more than century earlier. At the beginning of his journey, he stopped at Tell Bashir, then still home to the young Joscelin III and his mother, Countess Beatrice. Beatrice gave Hromklay to Grigor, who, delighted, abandoned his migration to Ani. He later gave Joscelin 15,000 dahekans for the castle, and received from Joscelin in return ‘a document written in his own hand that Hromklay would remain the seat of the Armenian patriarchate in perpetuity’. The charter was placed in a chest in the castle, ‘as a reminder for the future, so that no one would come seeking after that fortress’.49 It is unlikely that Smbat was concerned that a count of Edessa was going to try to reclaim Hromklay – although there were descendants of Joscelin still in the Latin East.50 Rather, the story of the charter served as a guarantee of the patriarchate’s independence. The grant was not just a transfer of title; it also functioned as a source of blessing. Joscelin could not of course guarantee that the castle would remain the seat of the catholicate. So Smbat added a small postscript: ‘May the Lord grant that what was established in that physical document regarding that place be confirmed spiritually as well.’ In other words, Smbat prayed that God might be as generous as the Frankish countess. Thus, what was one of the last legal acts of the counts of Edessa within their own lands became the foundational document for the catholicate. The role of the Franks for Smbat was reversed; their authority, rather than being a guarantee of neutrality through its absence, became an eternal seal on patriarchal independence through assimilation to the divine. The Frankish county in Armenian historical consciousness was both fleetingly ephemeral and yet etched into the foundation stones of enduring Armenian institutions. Armenians in the 1140s were doubling down on Frankish authority even as it was slipping away. Baldwin of Marash was a cypher for the whole complex project of the Franks, embodying both their aggressive cupidity and the virtues of the martyr. For Nerses, the conquest of

the af terli f e of edessa  | 99 Frankish Edessa would provoke a new crusade, which would usher in the end of time. By the thirteenth century, memories of the county of Edessa served a much different purpose – the Franks themselves began to disappear from Armenian chronicles, being subsumed into a general category of ‘Christians’, which generally meant ‘Armenians’. The Franks thus became a naturalised part of Armenian history. Their utility came in serving as a transition for the catholicate, which had been wandering from place to place for nearly a century. The disappearance of the county paradoxically allowed the patriarchate to exist in a single location and became one of the most vital centres of Armenian culture and authority in the medieval period. Notes   1. Runciman, 1955, 3: 474.   2. Thomson, 2005, 35–44; Mahé, 1992, 121–153.   3. MacEvitt, 2008, 74–99.   4. Nerses Klaiets‘i [Shnorhali], 1878. For translation and commentary, see van Lint, 1999, 29–47. Nerses also commissioned a similar poem for the fall of Marash, but the poet Grigor Marashets‘i wrote it as a personal lament; see Alishan, 1873, 216–221. For a translation, see ‘Invasion of the City of Marash’, 2002, 446–448.   5. Grigor Erits‘ [Gregory the Priest], 1898, i: 375–395; translations adapted in consultation with Gregory the Priest, ‘Continuation’, 1993, 245–257.   6. For more on Baldwin and the county of Marash, see Beech, 1996, 35–52. For one of the few other references to Baldwin, see Paterson, 2001, 133–149.   7. We know very little about Baldwin from western sources. He was mentioned only fleetingly in Matthew of Paris’s Chronica, while William of Tyre referred to Baldwin only as ‘noble and powerful’, but did not tell us anything about his family or background (Willelmus Tyrensis, 1986, 734). In this sense, Barsegh’s warnings were curiously prescient. At the end of the oration, he mourned that Baldwin was forgotten: ‘he is not remembered among the dead, nor does he appear among the living. No one rings the bells for him; his name is not remembered in any church; divine liturgy is not said for his soul, nor is there even a simple commemoration of mercy on his behalf in this land.’ (Grigor Erits‘, 1898, 393; Gregory the Priest, 1993, 256.) It is only through Armenian sources that we know that Baldwin was the brother of Raymond of Antioch. Barsegh (in Baldwin’s voice) specifically referred to his brother, Raymond of Antioch,

100  |  syri a i n crusad e r time s who followed ‘my same path’ (Grigor Erits‘, 1898, 387; Gregory the Priest, 1993, 252). This association grows stronger when we notice that the oration was followed immediately in Gregory’s chronicle by the report of the death of ‘the prince of Antioch, the abandoned brother of Baldwin’ in 1149 (Grigor Erits‘, 1898, 395; Gregory the Priest, 1993, 257).   8. Very little has been written on Gregory the Priest as well; for a concise overview, see Beech, 2004, 119–143.  9. There was, however, a long tradition of such eulogies in Armenian; see for example the eulogy of Grigor II Vykaser by Vardan Haykazn, 1911, 7–10, tr. Hacikyan, 2000, II: 362–364. 10. Grigor Erits‘, 1898, 375; Gregory the Priest, 1993, 245. 11. Grigor Erits‘, 1898, 375; Gregory the Priest, 1993, 245. 12. Grigor Erits‘, 1898, 380–381; Gregory the Priest, 1993, 248. 13. Barsegh claimed to preach in two languages, so one wonders if a translation was ever produced. Grigor Erits‘, 1898, 382; ‘Gregory the Priest, 1993, 249. In the next century, the chronicler Smbat Sparapet labelled the oration a letter, and claimed that Barsegh had it sent around to the districts under Baldwin’s rule. Smbat Sparapet, 1956, 165. 14. Grigor Erits‘, 1898, 376; Gregory the Priest, 1993, 246. 15. Grigor Erits‘, 1898, 375; Gregory the Priest, 1993, 245. As was common in the twelfth century, Grigor called the Franks ‘Romans,’ a term which used to be reserved for the Byzantines; see MacEvitt, 2014, 260–275. 16. Grigor Erits‘, 1898, 383; Gregory the Priest, 1993, 249–250. 17. Michael here may be biased, perceiving Baldwin to have favoured the Armenians over the Jacobites. Michael the Great, 2009, 636–637; translation adapted from Moosa, 2014, 670. 18. Grigor Erits‘, 1898, 382; Gregory the Priest, 1993, 249. 19. Grigor Erits‘, 1898, 376; Gregory the Priest, 1993, 245. 20. Grigor Erits‘, 1898, 379; Gregory the Priest, 1993, 247. 21. Grigor Erits‘, 1898, 376; Gregory the Priest, 1993, 246. 22. Grigor Erits‘, 1898, 380–1; Gregory the Priest, 1993, 248. 23. Grigor Erits‘, 1898, 382; Gregory the Priest, 1993, 249. 24. Grigor Erits‘, 1898, 377; Gregory the Priest, 1993, 246. 25. Grigor Erits‘, 1898, 390; Gregory the Priest, 1993, 254. 26. Grigor Erits‘, 1898, 391; Gregory the Priest, 1993, 255. 27. Grigor Erits‘, 1898, 381; Gregory the Priest, 1993, 249. 28. Dédéyan, 1992, 237–252.

the a f terli f e of edessa  | 101 29. For more on the fragmentation of religious authority at this time, see MacEvitt, 2008, 157–181. 30. See, for example, Sirapie Der Nersessian, 1962, 641–642. 31. Grigor Erits‘, 1898, 427; Gregory the Priest, 1993, 279. 32. William of Tyre, 1986, 782. 33. An Armenian lord named Michael appeared in other accounts of this period. According to Matthew of Edessa, Michael recaptured Gargar from the Turks in 1124–5; he was the son of Constantine, presumably the Constantine who was ruling in Gargar at the time of the First Crusade. Michael traded Gargar for Dülük, which was in the general vicinity of Hromklay. Subsequently, Gargar was ceded by Joscelin II to Vasil, the brother of Nerses and Grigor. See MacEvitt, 2008, 74–99. 34. Michael the Great, 2009, 653–654. 35. The translators were also not particularly interested in the conquest of the city of Edessa in 1144 and 1146. Only brief references to both appeared in the translation, and the reader was referred to other sources for more details, even though Michael himself gave a long account. 36. For more on the two versions, see Schmidt, 93–128. While only one Syriac manuscript of Michael’s chronicle has survived, about forty manuscripts of the Armenian translation are extant. 37. Tearn Mikhayēli Patriark‘i Asorwots‘ Zhamanakagrut‘iwn, 1870, 430. Later the account referred to Joscelin as the heir to Hromklay. My translation is adapted from Bedrosian, 2013, 176. 38. In his chronicle some years later, he recorded a similar version of the story. Beatrice was living at Hromklay, and Vardan placed the story in the space allotted to the Armenian year 612 (1163–4). She mentioned that she left her son in Europe with relatives. Patmut‘iwn (Venice), 1854, 128. A slightly different version appeared in the history of Vardan’s contemporary, Kirakos Gandzakets‘i. In his account Beatrice has been widowed, and Grigor requested the castle as a donation, which she agreed to, ‘eagerly’. Grigor then sent the countess to Cilicia, where T‘oros gave her villages and fields and other goods. She then returned to the country where she came from. Here T‘oros replaced Manuel I Comnenus who did give Beatrice as a stipend the Edessan lands, including Hromklay. Kirakos Gandzakets‘i, 1961, 108. 39. Srboyn Nersesi Shnorhalwoy Patmut‘iwn varuts’, 1954, 14: 45; translation in consultation with Isabelle Augé; see Augé, 2011, 286. 40. Patmut‘iwn varuts, 1954, 47; tr. Augé, 2011, 286.

102  |  syri a i n crusad e r time s 41. Most commonly in Arabic sources. See, for example, in the biography of Zengi: Ibn Khallikan, 1843–71, tr. de Slane, I: 540. 42. Patmut‘iwn varuts’, 1954, 47; tr. Augé, 2011, 287. The Life was written for the vardapet Hovhannes – it was not commissioned by the kat‘olikos at the time, Constantine. Patmut‘iwn varuts’, 1954, 82–83; tr. Augé, 2011, 302–303. 43. It is a little difficult to imagine that Manuel retained control of Hromklay much after 1150, however. For the negotiations, see Augé, 2011; Dédéyan, 1992; Darrouzès, 1990, 153. 44. The castle was Runculat in Frankish sources, a name which clearly derived from the same nomenclature as all the others, but did not translate its meaning into French (or Latin). 45. MacEvitt, 2014, 260–275. 46. For more on Hromklay’s geographic location, physical remains, and later history, see Stewart, 2006. 47. Smbat was also confused about the history of the county of Edessa, but gave more detail than his compatriots, again mostly taken from Matthew of Edessa. He identified Baldwin of Boulogne as the brother of Raymond of St Gilles (though he is correctly identified later, when he becomes king of Jerusalem), and told the story of Baldwin’s connivance in the murder of T‘oros. But he did not follow Matthew’s apocalyptic framework, which made his description of Frankish violence more damning. He also left out events that depicted the Franks’ action in a different light. When discussing the taking over of al-Bira from the Armenian lord Ablgharib by Galeran du Puiset, he did not include Galeran’s marriage to Ablgharib’s daughter, and instead recorded that Ablgharib fled al-Bira for Cilicia. 48. Smbat Sparapet, 1956, 168. 49. Ibid., 168–169. 50. A descendant of the counts of Edessa, Joscelin de la Mandalee, was living in Acre in 1280; Röhricht, 1893, #1435, 374. 

6 Diplomatic Relations and Coinage among the Turcomans, the Ayyubids and the Crusaders: Pragmatism and Change of Identity Taef El-Azhari

D

uring the two centuries of the Crusading era in the Levant (1095–1291), numerous Christian European and Eastern Muslim powers clashed with each other; however, at the same time, several rulers and dynasties had established different forms of accords and coexistence, many of which had developed into various levels of collaboration, even including the creation of military pacts. The most common reason for coexistence in medieval times was concerned with trade and commercial activity, resulting from the economic needs of the different sides, despite the political–religious differences. Another important factor was diplomacy, especially covert diplomacy, restricted to rulers and their immediate entourage only, so they had more freedom to express their true needs, thoughts and intentions without worrying about the public’s reaction, as Saladin once expressed to King Richard.1 To support this theory, documents of the time reflected the extent of these contacts, providing more valuable information than accounts in contemporary chronicles, which provided only select information and the views of various historians. Before I discuss the Turcoman–Ayyubid accord with the Crusaders, I will present a brief anatomy of the ethnic–religious powers in the Levant to provide a greater understanding of such coexistence and cooperation.

106  |  syri a i n crusad e r time s When the Crusaders arrived to Syria in 1097, the Eastern and Muslim powers were divided as follows: Fatimid Berber or Musta‘li Isma‘ili Shi‘ites, controlling the Syrian coast with their power base in Cairo; the Nizari Isma‘ili Shi‘ites, who had rejected Cairo authority in 1094, and who were forming their power base in the mountains of central Syria (later using Masyaf during the 1130s as their power base);2 and the Druze, who believed in the divinity of the Fatimid caliph al-Hakim (d. 1021), and who actively preached in southern Syria. In addition, there were the Nusayris (now known as the ‘Alawites), from northern Syria and south-east Anatolia, emerging from the ninth century as a radical Shi‘ite sect, elevating ‘Ali, cousin of the Prophet Muhammad, to the level of a god. Added to this were local Arab principalities, such as Tripoli, ruled by the Banu Ammar, Shayzar by the Banu Munqidh, and Tyre by the Banu ‘Uqayl, who changed their political loyalty between the Shi‘ite Fatimids and the Sunni powers as dictated by the need for survival.3 Sunni powers were largely represented by the Seljuq Turcomans, who had their power base in Iran and Iraq, with branches in Syria and Anatolia. The dramatic arrival of the Seljuqs, several decades before the Crusaders came to the Middle East, altered the Middle East both politically and ethnically, especially after the Battle of Manzikert in 1071. Many Turcoman powers established their own political careers outside Seljuq authority, such as the Artuqids in Anatolia, northern Syria and the Jazira, the Zengids in Syria in the twelfth century, and the Ayyubid Kurds, who were originally mamluks of the Zengid house, but established their own realm under Saladin between 1171 and 1260,4 to be followed by the Mamluk Turks. In addition to that, there were the Cilician territories of south-east Anatolia, ruled mainly by Armenians, especially from 1198 to 1375.5 Armenian powers could also be found in Edessa, which had fallen to the Crusaders in 1098, and there were dominant Armenian populations in Antioch and the surrounding areas.6 Syria, at that time, consisted of a great ethnic mosaic of Syriacs, Arabs, Armenians and Turcomans, who followed Judaism, Christianity and Islam with its various sects, as well as the Yazidis in Syria, Iraq and Anatolia, and a number of other local sects and religions. As can be seen from the above, these diverse ethnic–religious powers and followers needed to coexist, challenging as it was, in order to reach a modus

d ipl o ma ti c relati ons a nd co in a ge  | 107 operandi with the Crusaders, who represented more than fifteen different nationalities from northern and western Europe. In this chapter I will seek to examine the collaborations and pacts that were established between the Turcoman forces and the Crusaders, and why these took place. I will also explore the Ayyubids from Saladin onwards, who had accords with the Crusader powers and the military and the cultural impact of that on the Levant. Finally, I will examine some rare coins produced by the Crusaders and Armenians in the Levant in the thirteenth century, which serve as a strong sign of the change of identity arising from the long coexistence with Muslim powers in the East. The Seljuqs and the Crusaders Almost five years after the capture of Jerusalem by the Crusaders in 1099 (it had been a Turkish Seljuq city since 1070), the Crusaders tried to expand along the coast at the expense of the Fatimids, and inland to the north, to share some of the countryside ruled from Damascus. In June 1104 Duqaq, the ruler of Damascus, died, leaving his infant son and successor, Tutush II, in the care of his capable atabeg Tughtegin.7 For political reasons that are unclear, Tughtegin overthrew Tutush II, and installed the young Artash, uncle of the toppled prince, as the new lord of Damascus on 10 October 1104.8 In December 1104 Artash, who feared the influence of Tutush’s grandmother, Safwat al-Mulk (who was angry that the throne had been taken from her grandson), decided to flee to King Baldwin I of Jerusalem, after coordinating with the Damascene deputy of Busra, Aytakin al-Halabi. Artash had contacted King Baldwin and took refuge in Jerusalem for several weeks.9 During his stay in Crusader Jerusalem, Prince Artash tried hard to convince King Baldwin to march with him to invade Damascus, but the latter declined. Here we have, for the first time, a Muslim ruler travelling in person to the newly founded Crusader kingdom, seeking collaboration and offering information about the domestic situation in his kingdom. I strongly believe that the Seljuqs, who also appeared in Syria for the first time in 1070, were as alien as the Crusaders. They came from Central Asia, ­speaking the Turcoman language, and they had a heavy Shamanist influence to their Islam, and so because they had not been an

108  |  syri a i n crusad e r time s organic part of that society, their political survival took priority over any notion of jihad. The second case of contact between a Seljuq ruler and the Crusaders, that we know of, is that of Prince Ridwan of Aleppo, who was suffering from having to pay heavy taxes to the expansionist Prince Bohemond of Antioch in Summer 1103, upon the latter’s orders in return for not looting the Aleppan countryside.10 When Prince Tancred took over the rule of Antioch, he inflicted a crippling defeat on Ridwan in April 1105 at Artah, north of Aleppo. The result was that the Crusaders now had control of most of the countryside between them and the Seljuqs.11 As Thomas Asbridge states: ‘The battle of Artah marked a watershed in the history of northern Crusader states.’12 Ridwan tried to present himself to the Crusaders of Antioch as an ally. This new shift in relations between the Turcomans of Aleppo and the Crusaders of Antioch took on a new level of modus vivendi through a full military pact in the summer of 1108. When the Turcoman lord of Mosul, Chavli, attacked the north-eastern dominions of Aleppo, the king of Aleppo turned to Tancred for help. After a decade in the Levant, Tancred had an understanding of the geopolitical influence and danger of Mosul. After all, it was Kerbogha of Mosul who was the only Muslim lord who had come to save Antioch in 1098. Tancred arrived at Aleppo and joined forces with Ridwan, while Baldwin of Bourg, lord of Edessa, had surprisingly formed a pact with Chavli, who was pleased with this arrangement. Near Tell Bashir there was a confrontation between the two sides, and the Tancred–Ridwan pact emerged victorious.13 Although no documents have survived from this period, it is surprising that the contemporary and post-contemporary Muslim chroniclers did not condemn such collaborations between Muslim and Crusader forces, nor did they evaluate them negatively; however, it is noteworthy that Latin sources, such as Fulcher of Chartres, ignored this shift of ideology completely. If we move to examine the Seljuqs of Damascus, who ruled the city and most of southern and central Syria from 1075 to 1154, we will see that the priority for them à propos the Crusaders was how the revenues of the countryside should be shared, especially between Damascus and Jerusalem on one

d ipl o ma ti c relati ons a nd co in a ge  | 109 front, and Tripoli and Hims on the other, as well as continuing, undisturbed, their commercial activities. A four-year peace treaty between Damascus and Jerusalem was concluded in 1108 after three years of arduous warfare, which was less about eliminating one another as much as to gain more lands for agriculture.14 Damascus gave additional concessions to the Crusaders, especially after the foundation of the principality of Tripoli – thus threatening Seljuq Hims – and the seizure of Sidon, which negatively affected Damascene trade.15 When King Baldwin of Jerusalem increased his economic pressure on Damascus by besieging Fatimid Tyre, the main commercial outlet for Damascus, in 1111, we find that Tughtegin of Damascus, who ruled without any sultanate diploma until 1115, asked the lord of Mosul, Mawdud, for military help. At the same time, Tughtegin was forced to have an exceptional collaboration with the Fatimids by ruling Tyre in their name, as they could not afford to defend it: Damascus guarded Tyre from 1112 to 1124. Mawdud came to the aid of Damascus at the famous Battle of Sinnabra, which took place in northern Palestine in late June 1113. Fulcher of Chartres and William of Tyre describe the Muslim victory over the Crusaders of Jerusalem as a disaster; yet, we see at the end of the campaign Tughtegin and Mawdud retiring to Damascus, where Tughtegin had arranged for Assassins to kill him inside the Umayyad Mosque.16 The question is why? Tughtegin had been ruling without recognition from the Seljuq sultan between 1104 and 1115. In addition, he feared the superior military ­influence of the Seljuqs of Iraq in his dominions, so he chose to form an accord with the Crusaders, rather than join forces with fellow Muslims to perform jihad. Tughtegin negotiated a peace treaty with Baldwin I the following year to bring about economic stability in the countryside, which was very successful. Tughtegin also formed a military pact with Jerusalem and Antioch in order to confront the Muslim armies dispatched by the sultan from Iran and the East for jihad and to punish Tughtegin for murdering Mawdud.17 This spirit of and need for coexistence and full military cooperation against other Turcoman powers continued to the end of the Seljuq Turkcoman rule of Damascus. Between 1131 and 1140 Damascus suffered regular attacks by ‘Imad al-Din Zengi. On 6 December 1139, in the heart of

110  |  syri a i n crusad e r time s winter, Zengi launched an unprecedented siege of the Seljuqs in Damascus. The iron siege was too much for Damascus and its atabeg, Unur, who resisted Zengi until June 1140. Once he had lost all hope of holding out against Zengi, and the city was about to be captured, Unur sent a message to King Fulk of Jerusalem, seeking urgent military help. William of Tyre furnishes us with details about Unur, who surrendered members of his own family to the Crusaders as a sign of goodwill, in addition to paying 20,000 dinars to cover the expenses of King Fulk. Receiving reports of the approach of the Crusader army, Zengi lifted his siege on 19 June 1140.18 The Crusaders as much as the Turcoman Seljuqs of Damascus needed this relationship in order to preserve their mutual interests. If we think that this Seljuq behaviour by Unur reflects only the ruling class, we are ­underestimating the situation, as the contemporary Damascene chronicler Ibn al-Qalanisi furnishes us with details of the ahdath and local Syrians who, together with the rulers, put up stiff resistance to Zengi.19 Here I have to mention Usama b. Munqidh, who reports several stories about Unur and Fulk and their hunting together in the Kingdom of Jerusalem, as well as their exchanges of gifts, which reflects the human dimension and not only political interests.20 Paul Cobb states that: ‘Unur of Damascus ­continued to enjoy the fruits of his treaty with Jerusalem, periodically ­sending the suave Usama to the court of King Fulk to keep relations smooth.’21 Most of these contacts and diplomatic relations were reported by ­medieval sources; unfortunately, no documents have survived to shed further light on the details of the nature and mentality of the negotiations themselves. Before moving on to the Ayyubid model, I would like to mention that the Zengid realm in Syria did not launch a serious attack on the Crusaders, with the goal of eliminating them, before 1144. For example, when Emperor John Comnenus invaded northern Syria in April 1138 and formed a pact with the Crusaders, Zengi called on his sultan, Mas‘ud, to perform jihad. However, when the sultan surprised him with a swift reaction, dispatching 20,000 Turcoman knights, the same Zengi urged the sultan to withdraw them.22 Zengi preferred to coexist with the Crusaders rather than being uprooted by his sultan, even under the banner of jihad.

d ipl o ma ti c relati ons a nd co in a ge  | 111 The same pattern and mentality was deployed by Zengi after he captured Edessa in 1144, when he did not move forward with his claims of jihad, but rather spent the last two years of his life fighting Kurdish and Arab Sunni Muslims (for reasons that are unclear); meanwhile, he remained satisfied with his Crusader neighbours in Syria.23 The Ayyubids and the Crusaders The Ayyubids, who ruled between 1171 and 1260, were ethnic Kurds with a Turcoman influence passed down from their Zengid lords; one could describe them as ‘Turkicised Kurds’. They descended from the Rawaddiyya clan, part of the Hadhbaniyya tribe, from the suburbs of the Georgian capital Tiflis in the northern Caucasus.24 Saladin had been suffering a crisis of confidence over the situation in Egypt when he started his rebellion against his Turcoman sultan, Nur al-Din of Syria, in 1169. Saladin came from a Kurdish minority that was part of a larger diverse Kurdish population of twenty-two major tribes.25 These tribes were a minority in an even larger community of Oghuz Turcomans, within a larger still circle of millions of Christian Egyptians. Saladin simply did not have the troop numbers to count on, and he had many enemies to confront at the same time, such as the Shi‘ite Isma‘ili Fatimid loyalists, the Zengids, the Crusaders and the Egyptian natives. Saladin had attempted to escape from his master by dispatching his brother Turan Shah to invade Yemen in March 1174;26 he intended to invade Syria after the death of his sultan, Nur al-Din, in May of the same year, in order to fight the Nurid house, but had to reconcile with the Crusaders of Jerusalem, especially as the danger from supporters of the Fatimid dynasty, brought to an end by Saladin in 1171, was still imminent. Saladin sent a message to the new king of Jerusalem, Baldwin IV, expressing his sincere condolences for the passing of his father, King Amalric, on 11 July 1174. The long eloquent letter contains the following lines: God bless the glorious king, protector of Jerusalem. My earnest congratulations to you for ascending the throne of the kingdom, and my sincere condolences for the passing of your father, King Amalric. I would like to let Your Majesty know that we hold nothing but friendship, loyalty and

112  |  syri a i n crusad e r time s amicable ties to your kingdom, as we did before with the late king, because of our common interests, and despite our different religions. So, you can count on us, like a son who counts on his father, and hesitate not to consult or contact us on any matter. God bless your rule, keep and protect your realm, and guide you to believe in this friendly and sincere message.27

We have to question not only why Saladin took such a daring step, but also why it is not commented on or criticised by contemporary or post-­contemporary Muslim chroniclers who had liberty to use their own judgement. It forms part of the Egyptian archive, passed down by al-Qalqashandi. Baha’ al-Din b. Shaddad, who mentions more than twenty-five letters exchanged between Saladin and the Crusaders, ignores this message, and we do not see William of Tyre or any of his successors shed any light on Saladin’s approach. I believe that Saladin, who was a complete outsider to the area, had to adopt realpolitik and pursue his own interests. During the long negotiations between Saladin and King Richard the Lionheart of England between 1191 and 1192, which concluded with a peace deal at the end of the Third Crusade, we see very frank language being used by the two lords, who avoided meeting each other. King Richard wrote: The Muslims and Franks have been devastated, and their dominions completely ruined. Wealth and souls have been lost from both sides, and this matter has consumed more than it merited. We should have back Jerusalem and the Cross, and we should reach an accord to end this suffering.28

Saladin replies: ‘Jerusalem is ours as it is yours, and we cannot surrender it or even talk about that among Muslims.’29 Here we see King Richard reflected by Muslim chroniclers as a tired warrior who is longing to return to his realm. Saladin was concerned about his image, and not the principle of returning Jerusalem, or jihad. Other letters and correspondence reflect very amicable feelings and trust, including a few daring suggestions, such as King Richard’s idea of appointing one of his nephews (Henry II), as a vassal to Saladin in order to govern Jerusalem; there were a number of other bold and challenging ideas, such as a marriage proposal between Richard’s sister or his niece and al-‘Adil.30 Anne-Marie Eddé states that Saladin had accepted the proposal,

d ipl o ma ti c relati ons a nd co in a ge  | 113 but Richard failed to persuade al-‘Adil to become a Christian.31 In addition to this, Saladin sent fruit and other foods to King Richard when the latter fell ill, and Ibn al-Furat reports Richard describing Saladin as his ‘brother’ in these letters. Saladin, too, described Henry of Champagne ‘as one of his own sons’.32 Saladin died the year after the truce of 1192 between him and Richard. In 1200 the Ayyubid house came under the rule of Saladin’s brother, al-‘Adil, and his descendants, and remained so up until the fall of the dynasty in 1260. Two different and pragmatic approaches were taken by two different Ayyubid sultans: the first being al-‘Adil (r. 1200–18). As Stephen Humphreys states: Perhaps the dominant theme of al-‘Adil’s reign is the Ayyubid expansion in the Jazira and Armenia. The struggle against the infidels is not really a salient feature of his reign. He seems to have played it down as much as possible; throughout his entire reign he undertook but one major campaign against the Latin states.33

The major event of al-‘Adil’s reign was the Fourth Crusade, and its echoes in the Levant. While the Crusaders were besieging Constantinople in the summer of 1203, Ibn Wasil reports a dialogue that took place around June of that year between an envoy of the Templars in Syria and the Ayyubid lord of Hama, al-Mansur. The latter, a vassal of al-‘Adil, wrote to al-‘Adil, seeking his advice about how to deal with the Crusaders. After the Hospitallers had failed several times to attack the Ayyubid countryside of Hama in the spring of 1203, they sought the help of the Templars to mediate with the Ayyubids in order to reach a peace settlement.34 Although the envoy of the Templars used harsh language initially, threatening the Ayyubids with the new European forces arriving from overseas, al-Mansur agreed to a truce, and the Crusaders were happy with this outcome.35 Al-‘Adil was asked for advice by al-Mansur as to how to deal with the Crusaders. He replied: ‘Whatever action your council considers advantageous should be taken regarding the Franks. Their supplies are scarce, their support is delayed and messengers come from every corner carrying the news of their weakness.’36 I believe that some of the information received by al-‘Adil was from Muslim and Italian merchants. Al-‘Adil was keen to avoid a major

114  |  syri a i n crusad e r time s c­ onfrontation, residing in Damascus throughout 1203 and 1204. Even when a small Frankish fleet from Acre raided the Nile Delta in May 1204 he did not launch a counter-attack on the Franks in Syria.37 There were several reasons behind al-‘Adil’s non-aggressive response. He was fighting the Zengid house in Mosul, he was repairing the many fortifications damaged by the strong earthquake that struck the Middle East in 1202, and, since 1201, Ayyubid Yemen had been in a political–military crisis following the accession to power of al-Nasir b. Tughtegin, al-‘Adil’s nephew.38 Thus, al-‘Adil was over-stretched both militarily and economically, and he therefore opted for cooperation with the Crusaders as far as possible for most of his career. Al-Maqrizi, who evaluated his career, wrote: ‘He believed in not initiating wars against his enemies, but instead used methods of conspiracy and manoeuvres. As a result, the Franks made a truce with him.’39 The second Ayyubid ruler in this category is Sultan al-Kamil b. al-‘Adil (r. 1218–38). In the course of his long reign we see the politics of jihad and the art of realpolitik reach their peak. During the strong and long attack of the Fifth Crusade in the Nile Delta of Ayyubid Egypt, al-Kamil agreed in 1221 to meet the Crusaders’ demands to achieve peace. They demanded the surrender of Jerusalem, among other territories seized by his uncle Saladin in 1187. Al-Kamil, who had come to power during the Fifth Crusade, agreed.40 The question here is why Muslim historians like Ibn al-Athir, Ibn Wasil and al-Maqrizi, among others, did not condemn or criticise al-Kamil for such a deed. They also did not report the reaction to this deal of al-Kamil’s brothers (al-Mu‘azzam ‘Isa and al-Ashraf Musa), who had come to his rescue. During the civil wars between al-Kamil and his brother, al-Mu‘azzam of Damascus (1225–7), we see the latter seeking the help of Sultan Jalal al-Din b. Khwarazm Shah, and cancelling the khutba for al-Kamil. This resulted in al-Kamil contacting the German emperor, Frederick II, in 1127, seeking a military pact against Damascus in return for the surrender of some Syrian territories.41 When Frederick II arrived in Acre during the Sixth Crusade, he found that al-Mu‘azzam had died. After long negotiations between him and his ally, al-Kamil, led by Fakhr al-Din b. Shaykh al-Shuyukh and the qadi of the army, al-Irmawi, he demanded the surrender of Jerusalem, to which al-Kamil agreed.

d ipl o ma ti c relati ons a nd co in a ge  | 115 There are many details about how the Crusaders were going to rule Jerusalem, without denying the Muslims their religious rights. The sanctuary of the Dome of the Rock and the Aqsa Mosque and its vicinities would remain in Ayyubid hands, and praying ceremonies would not be interrupted in any way. Here we see an exceptional level of collaboration and understanding, never seen before during the Crusading era. Frankish forces would guard Jerusalem, but all fortifications would remain in ruins, and Frederick would retain control of all the countryside around Jaffa, Acre and Jerusalem.42 The question which should be asked here is why we do not see Muslim chroniclers, especially those who were at liberty of doing so, criticise al-Kamil politically or religiously for surrendering a holy place such as Jerusalem. Instead, medieval historians reported only a few lines about how the Muslim locals of Jerusalem were evacuated in gloomy conditions; no one dared to rebel or resist the sultanate decree. I believe that those historians and scholars themselves became ideologically prepared for coexistence with the Crusaders. They reported how the Damascenes gathered at the Umayyad Mosque and wept over their loss, as encouraged and orchestrated by al-Kamil’s nephew, al-Nasir Da’ud, and the famous historian and scholar Sibt b. al-Jawzi.43 Yet we do not see major public actions, such as that of the chief qadi of Damascus, al-Harawi, when he gave a fiery speech in 1099 in the same mosque about the loss of Jerusalem to the First Crusade, and his march to Baghdad, leading an embassy to the caliph asking for help, which attracted the sympathies of many local Iraqis to their cause.44 The ‘Abbasid Caliphate did not react to the surrender of Jerusalem in 1229, as it had not offered any tangible help in 1099. If we look at the Crusader side, we see that many letters were exchanged between al-Kamil and Frederick about the surrender of Jerusalem. Gerald of Lausanne, Patriarch of Jerusalem, wrote to Pope Gregory IX from Acre on 26 March 1229: The emperor revealed [to his advisors] that the sultan was offering him the Holy City, except for the Temple of the Lord which would be maintained by the Saracens, so that the Saracen pilgrims could enter it freely. The emperor could strengthen the city of Bethlehem.

116  |  syri a i n crusad e r time s It is important to mention that the Patriarch had reported some political jealousy among the Franks in Jerusalem, especially by non-German subjects.45 While the diplomacy between Saladin and Richard was focused on political matters, we have very few details regarding the era of al-‘Adil. On the other hand, we see Frederick at the end of the Sixth Crusade, exchanging several letters with al-Kamil after the surrender of Jerusalem. These letters were concerned with theology, philosophy, mathematics and engineering matters and theorems, and requested al-Kamil to reply to them. Al-Kamil commissioned Shaykh ‘Alam al-Din to look into these theorems, and provided his answers, which al-Kamil sent back to Frederick.46 Al-Maqrizi reported how Frederick visited the Aqsa Mosque in admiration and climbed the pulpit, and was keen to hear the Muslim prayer call and the night hymns.47 Such behaviour reflects a complete change from the mentality of Pope Urban II in 1095, and how he perceived the Saracens (as he referred to them at the beginning of the Crusading movement). Through the history of the Crusading era, no one represents the spirit of pragmatism and coexistence more than al-Kamil. His uncle recaptured Jerusalem in the name of jihad, and he surrendered it in order to keep his throne. Although the Fatimid relationship is not part of this study, we might draw a comparison with the Fatimid vizier, Shawar, who allowed King Amalric to visit Giza and Cairo in 1167 and gave the Crusaders part of northern Cairo’s wall to establish a camp for around six months.48 If we look panoramically, we will see that the Italian republics in particular were following their own interests for much of the time, putting commercial interests before political causes. The annals of Caffaro of Genoa, produced during the twelfth century and recently translated by Jonathan Phillips, reflects a consistent and flourishing trade between the Genoese and the Saracens (as he refers to them), although he himself took part in the attacks on the Syrian coast after 1100. During the twelfth century, dozens of commercial contracts with Syria, Fatimid Egypt and the Almoravids of North Africa were undertaken. Although Genoa, like many other Italian cities, took part with their navies in attacking the Syrian coast,49 Jonathan Phillips reports on Saladin, who mentioned in 1175 how Genoa, Pisa and

d ipl o ma ti c relati ons a nd co in a ge  | 117 Venice were delivering arms to him.50 The same activity could be identified in correspondence between the Italian republic, for example, Pisa, and al-‘Adil and his son al-Kamil, seeking their protection, wishing them prosperity and showing great respect.51 During the reign of al-‘Adil I the Ayyubid, his diwan of correspondence with the Italian ambassadors within the Ayyubid realm contained friendly and diplomatic language. One reason for this was to guard his commercial economic interests, but it was also to keep open a political channel with a Crusader side who never regarded the Crusading movement as a true religious movement. As a result, we see a letter sent from al-‘Adil (no date is given) to the Italian ambassadors saying: ‘To the glorious rectors and ambassadors, heroes and lions of Christianity, and leaders of the Franks.’52 In another illuminating letter, we find several Italian merchants from Pisa, Venice, Frankish Beirut and Crete writing a warm petition to al-‘Adil, asking his mercy after they had been seized in Alexandria on their arrival from Crusader Beirut, on a charge of mistaken identity – the Ayyubid authorities thought they were from Cyprus.53 What took place under the reign of al-‘Adil, also extended to his son, al-Kamil: not only did merchants seek protection, but we find Italian leaders such as Lottario, Archbishop of Pisa, and a group of Italian ambassadors, renewing their oaths of allegiance to al-Kamil.54 From what is outlined above, we can see how the Italian powers saw in the Ayyubid Muslim side, among others, a respected partner or even an ally during the heat of the Crusading warfare in the Middle East. The long coexistence between the two sides influenced the language of the Muslim Mamluks in Egypt and Syria, and unravelled the political entity of the Crusaders to the new rulers of Egypt and Syria. In a letter sent from the Mamluk sultan Baybars to Bohemond VI (the last prince of Antioch), after Baybars had seized the city in 1268, we see him writing a long letter, which starts with the following lines: ‘The venerable Count, the lion Bohemond VI, glory of the Christian umma (nation) – ra’is al-ta’ifa al-Salibiyya – head of the Crusader denomination, who is now de-ranked from prince to count, after losing Antioch.’55 To the best of my knowledge, the term ‘al-Salibiyya’ had not been used or written in Arabic documents or in other sources. It appeared only during

118  |  syri a i n crusad e r time s the nineteenth century, in the translation of J. François Michaud’s Histoire des Croisades. The significance of this letter is that it was written originally in the Arabic language by the Mamluk diwan, which reflects a new development in the understanding of the Crusading movement by the Mamluks, nearly two decades before the end of this age, by distinguishing them as Christians, or Franks, as they were known from the eleventh century. It is of note that Ibn Jubayr, in his Rihla of 1182, wrote, while travelling from Damascus to Acre and before returning to Spain: ‘The Franks are using a ­certain term for sailing in the autumn, and they call it, “al-Salibiyya”.’56 It is not certain how he came to know the original term, nor how he translated it. Numismatic Evidence Moving to the last section of this study, I will examine some coins from the age to ascertain how they reflect certain developments between the two sides in respect of the coexistence and acceptance of one another. I have to emphasise here the importance of coins as a type of document, noting that they have not enjoyed much interest by researchers in recent decades. First, there are two coins from Norman Sicily from the late twelfth ­century. The first of these was minted by King William II of Sicily and Southern Italy (r. 1166–89). He was surprisingly proud of the Arabic ­language and Muslim heritage, probably through his contacts with North Africa. He minted a bronze follaro in Messina, on the reverse of which is his title in Arabic with nothing else: ‘al-Malik William al-Thani’, while on the obverse is a lion head, the emblem of William (see Figure 6.1). Ibn Jubayr, who stopped in Sicily for three months during 1182–3, gave a detailed account of William’s court, which was heavily influenced by the Arabic language and Muslim scholars. William II’s insignia was ­‘al-hamdu-lillah haqq hamdahu’, whilst the insignia of his father, William I (r. 1154–66), was ‘al-hamdu-lillah, shukran li-anamih’.57 The question is why William II used this mint at the same time as dispatching his fleets to attack the Ayyubids in Egypt, such as in 1174, in collaboration with Fatimid loyalists.58 We have to take into consideration that the king might have been an admirer of the Arabic heritage on a personal level inside his palace, but

d ipl o ma ti c relati ons a nd co in a ge  | 119

Figure 6.1  Bronze follaro of King William II of Sicily in Arabic – late twelfth century. Collection of the author.

in minting such coins that were widely circulated, mainly among Italian Christian citizens, it demonstrated a strong will in imposing some elements of a different culture. Unfortunately, neither the rare book Coinage in the Latin East nor The Meeting of the Two Worlds: the Crusades and the Mediterranean Context have addressed or focused on these coins.59 King Tancred, who ruled from 1189 to 1194, minted a bronze follaro in Messina, on the obverse ‘Rex’ in the centre and ‘Rogerivs’ around it, and on the reverse ‘al-Malik Tancred’ in Arabic, in the centre (see Figure 6.2). So we have a bilingual Norman coin, inscribed with both Latin and Arabic, as a new dimension of cultural fiscal coexistence; Bohemond I of Antioch in 1098 would probably have been highly disturbed by

Figure 6.2  Bronze follaro of King Tancred of Sicily in Arabic and Latin – late twelfth century. Collection of the author.

120  |  syri a i n crusad e r time s this ­ development between his descendants and the language of the Saracens. The second category of coins comes from Crusader Jerusalem. While the twelfth-century denier minted in Crusader Jerusalem had portrayed the Holy Sepulchre, with the name of different kings, such as Amalric60 or Baldwin IV, we see, in 1237, a new development. While the Ayyubid brothers of al-‘Adil I quarrelled among themselves over the rule of Syria and Damascus after the passing of al-Ashraf Musa, we find that the Crusaders of Jerusalem took the opportunity to mint an imitation of an Ayyubid dirham, in the name of the new Ayyubid king of Damascus, al-Salih Isma‘il b. al-‘Adil I, who came to the throne in 1237 for one year.61 The obverse of the dirham carries the title of the ‘Abbasid caliph, al-Mustansir, while the reverse has the name of al-Malik al-Salih (see Figure 6.3). The quality of the imitation is not as good as the original Ayyubid dirhams of the same period, but by the time it was discovered it would have already entered the commercial circles of the Levant. We know that Crusader Acre became a centre for coin forgery62 and that the Crusaders probably tried to influence negatively the Ayyubid economy with poor-quality ­dirhams made with a lesser weight of silver. We also know of the strong ties and various treaties between the Ayyubids and the Venetians who had an inn in Aleppo, early in the thirteenth century, and who were encouraged by alSalih Isma‘il during his second reign to import arms from his realm.63 Still, the question is why the Arabic–Muslim i­ nfluence? Also, why did the Muslim

Figure 6.3  Crusader’s imitation of an Ayyubid dirham. Jerusalem mint – ce 1237. Collection of the author.

d ipl o ma ti c relati ons a nd co in a ge  | 121

Figure 6.4  Armenian dirham of the thirteenth century with bilingual Armenian– Arabic writings. Collection of the author.

powers not counter the imitations? It seems that in the thirteenth century the Arabic language had become the lingua franca across the Mediterranean. We have the anonymous Christian silver millares, imitating the Almohad mint, while the Reconquista was at its peak. In addition, Queen Tamar of Georgia (d. 1213), had minted her full name in Arabic, for no clear reason. Chingiz Khan had also done the same. The last coin which reflects a high level of cultural coexistence is the Armenian dirham of the period 1230s–40s. Despite their pride in their long heritage, and the preservation of their distinct language, we see here a bilingual Armenian–Arabic coin. It belongs to Armenia’s King Hethum I (r. 1226–70). On one side the cross and crescent are displayed together. On the reverse is the name of the Seljuq sultan of Anatolia, Kaykhusraw b. Kaykobad (r. 1237–46) (see Figure 6.4). I believe that the Armenian kingdom of Cilicia had submitted themselves either partially or fully to the Seljuq Turcomans in the first half of the thirteenth century. This is a significant progression of political relations between the Armenians and the Seljuqs, and other Muslim powers in the area, since the start of the Crusading movement. Notes   1. Saladin replied during negotiations about Frankish rights in restoring Jerusalem to King Richard: ‘We dare not say that among the Muslims’, Ibn Shaddad, 2003, 194.

122  |  syri a i n crusad e r time s   2. Bosworth, 2004, 68.   3. Ibn al-Athir, 1997, X: 65, 310.   4. Bosworth, 70, 190, 194.   5. MacEvitt, 2006.   6. Ibn al-Athir, 1997, X: 274.   7. Ibn al-Qalanisi, 1983, 233.   8. Ibn ‘Asakir, 1979, V: 250.   9. Ibid., III: 368; Ibn al-Qalanisi, 1983, 235; Ibn al-Athir, 1997, X: 376. 10. Ibn al-‘Adim, 1951–68, II: 511. 11. Ibid., II: 515; Fulcher of Chartres, 1990, 135. 12. Asbridge, 2010, 142. 13. Ibn al-Qalanisi, 1983, 270; Ibn al-Athir, 1997, X: 465. 14. El-Azhari, 1997, 187–188. 15. Ibid., 190–195. 16. Fulcher of Chartres, 1990, 151; William of Tyre, 1976, I, 494–495; Ibn alAthir, 1997, X: 497. 17. Fulcher of Chartres, 1990, 155; Ibn al-‘Adim, 1951–68, II: 538. 18. William of Tyre, 1986, II: 105–106; Ibn al-Athir,1997, XI: 74–75. 19. Ibn al-Qalanisi, 1983, 427. 20. Usama b. Munqidh, 2008, 300–301. From the same period come a number of stories about the conversion from Islam to Christianity in large numbers, such as in April 1138 when Emperor John Comnenus invaded Syria, the qadi of Buza (part of Zengid Aleppo), converted to Christianity (Ibn al-Qalanisi, 1983, 416). So, we see that, after forty years of coexistence, some Muslims were influenced by the political presence of the Crusaders, although Syriac Christians, together with Armenians, were in the majority in these areas before 1097. 21. Cobb, 2014, 134. 22. El-Azhari, 2016, 35. 23. Ibid., 105–106. 24. Al-Maqrizi, 1973, III: 305; Ibn al-Athir, 1997, XI: 341. 25. Al-Qalqashandi, 1987, IV: 373–379. 26. Ibn al-Athir,1997, XI: 396. 27. Al-Qalqashandi, VII: 115–116. 28. Baha’ al-Din b. Shaddad, 2003, 193. 29. Ibid., 194. 30. Ibid., 218, 203.

d ipl o ma ti c relati ons a nd co in a ge  | 123 31. Eddé, 2011, 264. 32. Ibn al-Furat, 1942, IV: section 2, 83–84; Abu Shama, 1288 (1871), II: 199. 33. Humphreys, 1977 127. 34. Hardwicke, ed. Setton, 1962, II: 535; Ibn Wasil, 1977, III: 146. 35. Ibn Wasil, 1977, III: 147. 36. Ibid., III: 141. 37. Ibn al-Athir, 1997, XII: 198. 38. Ibn Wasil, 1977, III: 136–139. 39. Al-Maqrizi, 1997, I: 312. 40. Ibid., I: 327. 41. Ibid., I: 345–346. 42. Ibid., I: 353. This is very similar to the principle negotiated between the Palestinian Authority and the Israeli government in 1993. From 1948 until the present day, the Aqsa Mosque and its holy Muslim vicinities come under the Jordanian administration. 43. Ibid., I: 354. The same ruler, al-Nasir Da’ud, changed his political stance in 1240 while fighting al-Salih Ayyub of Egypt. He contacted the Crusaders, offering to form a pact with them against Egypt and to surrender all of what Saladin had regained after 1187, in return for their military aid to Da’ud against Egypt: al-Maqrizi, 1997, I: 407. 44. Ibn al-Athir, 1997, X: 500. 45. Letters from the East, 2013, 129–130. 46. Al-Maqrizi, 1997, I: 354. 47. Ibid. 48. Al-Maqrizi, 1973, III: 286–287. 49. Caffaro, Genoa and the Twelfth-century Crusades, 2003, 17. 50. Ibid., 143. 51. I diplomi Arabi del R. Archivo Fiorentino, 1885, 79–80. 52. Ibid., Letter XXII, 69. 53. Ibid., Letter XXIII, 70. 54. Ibid., Letter XXVII, 81–82. Such amicable relations continued within the Mamluk state. 55. Al-Qalqashandi, 1987, VIII: 299–302. 56. Ibn Jubayr, Rihla, Beirut, 1988, 244. 57. Ibid., 267–268. 58. Ibn al-Athir, 1997, XI: 399.

124  |  syri a i n crusad e r time s 59. 60. 61. 62. 63.

Edbury, 1980, 126; Bornstein and Soucek, 1981. Edbury, 1980, 106. Humphreys, 1977, 232. Edbury, 1980, 20. Ibid., 74, 77.

7 Symbolic Conflict and Cooperation in the Neglected Chronicle of a Syrian Prince Luke Yarbrough

W

hat part did texts play in the political and military struggles that convulsed Syria during the Crusades? And what difference did it make when a text that narrated those struggles was authored by a prince and warrior – a leading agent in the grand events of the age – rather than by a scholar? This essay will explore these questions through the lens of a chronicle written by an Ayyubid prince of Hamah, al-Malik al-Mansur Muhammad (d. 617/1220): Midmar al-haqa’iq wa-sirr al-khala’iq (‘The arena of truths and secret of created things’). The Midmar has received only sporadic attention since 1968, when the only known section of it first appeared in print, treating, in lacunose fashion, the years 575–82/1179–86.1 Most of the foregoing work on the text has been done by just two scholars, the Egyptian historian Hasan Habashi, its editor, and the German Islamicist Angelika Hartmann.2 But like other Arabic historiography of the era, the chronicle bristles with just the sort of conflict and cooperation with which the present volume is concerned. It describes events not only in Syria, where Saladin’s campaigns take centre stage, but also in North Africa and Iraq, for both of which it is a uniquely precious source.3 What is more, the unusual perspective of its author, the learned prince al-Mansur, might be expected to lend special interest to its version of events, much as the better-known work of alMansur’s great-grandson Abu’l-Fida owes some of its interest to its author’s position as ruler of Hamah.4 The continued neglect of the Midmar in modern scholarship makes it difficult to say quite what the work adds to our knowledge of events in Syria. Recently, however, an additional manuscript section of it has been

126  |  syri a i n crusad e r time s discovered in Tunis. Although only the North African portions of this new section have been edited, it is clear that its contents will profoundly change our understanding of the work, and perhaps also our narrative of the years 582–6/1186–91, in Syria, North Africa and (in particular) Iraq.5 Our limited knowledge of the sources, methods and agendas of the Midmar, especially in its unpublished portion, would render hazardous any effort to analyse historical events in Syria as the work represents them. Instead, the focus of this essay is limited to the work itself. I shall argue that the Midmar served al-Mansur as a textual tool of self-interested symbolic competition. When compared closely to its sources, notably al-Barq al-shami of ‘Imad al-Din al-Isfahani and Ta’rikh Mayyafariqin of Ibn al-Azraq, the Midmar, notwithstanding its many merits, emerges as a curiously derivative and subtly misleading chronicle that lionises its author and seeks to forge an implicit connection between his rule and exemplary earlier models of righteous kingship, notably that of Saladin. In order to see how the Midmar functioned as an instrument of learned propaganda, we shall review what is known concerning work and author; examine the use of sources in the Syrian sections of the Midmar, showing that the author used al-Barq al-shami and Ta’rikh Mayyafariqin to a more pronounced degree than has been recognised whilst suppressing evidence that he had done so; briefly consider how a text like the Midmar could serve as a tool of in symbolic competition and cooperation among the Muslims of Syria in the age of the Crusades; and provide, in an appendix, a précis of the Syrian material in the newly discovered portion of the work. Al-Mansur and Midmar al-haqa’iq Al-Mansur, whose fuller name was Muhammad b. Taqi al-Din ‘Umar b. Shahanshah, was the son of Saladin’s nephew and trusted lieutenant, al-Malik al-Muzaffar (d. 587/1191). He was born as his renowned greatuncle was first consolidating power in Egypt, c.567/1171–2.6 He may have accompanied his father al-Muzaffar on at least some of the latter’s frequent missions in Syria and Egypt, where in 579–82/1183–6, al-Muzaffar acted as Saladin’s vice-regent. During this posting the young Muhammad had the opportunity to study in the Sunni traditionist hotbed of Alexandria. This experience must have contributed to his lifelong affection for scholarship.7 By 585/1189, when he was in his late teens, as we learn from the unpublished

sy m b oli c conf li ct and coope r a tio n  | 127 portion of the Midmar, he was able to fight at the Battle of Acre, unhorsing and killing Frankish opponents and receiving praise poetry for his geste. Soon thereafter he appears in attendance on Saladin along with his father.8 When his brother al-Malik al-Fa’iz Shahanshah died, he immediately took his place as governor in Mayyafariqin, arriving there with much fanfare in Sha‘ban 586/September 1190.9 His father al-Muzaffar died slightly more than a year later while on campaign in the Artuqid vassal regions of the Jazira. Al-Mansur succeeded him as ruler of Hamah and the nearby towns, some of which he was compelled to hand back to Saladin. An attempt to expand his own territories in 595/1199 was blocked by Saladin’s brother and successor, al-‘Adil. From this date until the end of his life, al-Mansur settled down to rule Hamah, which he built into a bulwark against the Franks, a stable island in a sea of Ayyubid dynastic squabbles, and a haven for scholars. The Midmar is one of three literary works that medieval authors attributed to al- Mansur, all of which survive only in part.10 We are fortunate that the available evidence surrounding the work has been gathered and analysed by Angelika Hartmann, though her findings can now be supplemented with additional details. It was originally quite substantial – nearly twenty volumes according to Ibn Wasil – and was praised by later historians, though their testimonies do not reveal the original extent of its coverage.11 The surviving portion, which derives from the eleventh volume, is a chronicle arranged by year.12 For each year that this portion covers, it alternates among three narrative ‘tracks’: events surrounding the caliph in Baghdad; Saladin’s exploits; and the Ayyubid campaign in North Africa. The third track ends in AH 584, when the North Africa campaign was brought to a close and its general, Sharaf al-Din Qaraqush, recalled to Syria.13 Each track has a rather different tone and set of concerns, an effect that owes mainly to the author’s sources. Those he used for North African events are unknown. For Baghdad, he may have relied upon a lost chronicle by the Hanbali scholar Ibn al-Maristaniyya.14 For events in Syria, as Hartmann has noted and we shall see further below, he made very heavy use of al-Barq al-shami. While the importance of the Midmar for events in North Africa and Bagdad is beyond dispute, largely because the sources for these sections are lost, its Syrian portions, for which the author was often ostensibly present and supplies ‘numerous autobiographical data as well as reports of eye-witnesses’ and ‘numerous official documents, several

128  |  syri a i n crusad e r time s of which can only be found here’, can be scrutinised more closely because several related sources survive, not only the Barq but also works derived from it, such as al-Bundari’s abridgement, Sana al-barq al-shami, and Abu Shama’s al-Rawdatayn fi akhbar al-dawlatayn.15 The newly discovered manuscript of the work permits us to specify the date of its composition with greater precision.16 Here Tughril III (d. 24 Rabi‘ I 590/4 March 1194) is identified as ‘the last of the Seljuq sultans’ (fol. 97b). Henry II of Champagne (al-Kundharri) is said to have ‘lived until the end of 594[/1198]’ (fol. 108b). A still later indication is the blessing invoked upon the memory of Muhyi al-Din b. al-Zaki, who gave the first Friday sermon in Jerusalem following its capture in 583/1187. Muhyi al-Din died on 7 Sha‘ban 598/2 May 1202.17 While further study of the manuscript may provide additional clues, we can now be assured that al-Mansur composed the work well after he had come to rule Hamah, and most probably around the same time that he wrote his other two books – 600/1203–4 and 602/1205–6 – though not later than the latter, which refers to the Midmar.18 Midmar al-haqa’iq, al-Barq al-shami and Events in Syria The newly discovered manuscript of the Midmar preserves some narrative that overlaps with the period of al-Mansur’s adulthood. The tone of that narrative is encomiastic towards its author, as we shall see. But what of the extant passages, published and unpublished, that narrate earlier events in Syria? In these passages, as Hartmann remarked, one frequently encounters first-person testimony that implicitly positions the author as an eyewitness. What does that testimony reveal about the author’s goals, sympathies and methods of arranging historical information? In this section I will propose that few if any of the eyewitness observations are al-Mansur’s own, that most apparent indications of his involvement are misleading, and that the author of the Midmar suppressed evidence that he was using earlier sources. Hartmann, who considered some of the same questions three decades ago, found that the Syrian portions of the work are based heavily on al-Barq al-shami, the memoir–chronicle of Saladin’s career by one of his leading advisors and officials, ‘Imad al-Din al-Katib al-Isfahani (d. 597/1201).19 Only the third and fifth volumes of the Barq survive, and we are fortunate that these

sy m b oli c conf li ct and coope r a tio n  | 129 overlap substantially with the published sections of the Midmar, though not with the newly discovered portion.20 The Midmar shares its dependency on ‘Imad al-Din’s works with most of the Arabic historiography for this portion of Saladin’s career.21 Hartmann made exploratory comparisons of the published Midmar with the corresponding sections of the unique, then-unpublished manuscripts of the Barq, as well as with its derivatives, particularly the Sana of al-Bundari and the Rawdatayn of Abu Shama. She concluded that despite al-Mansur’s heavy use of the Barq in composing the Midmar, he still adds unique and valuable information, much of it based on his personal experience or connections. His primary modus operandi was to mine ‘Imad al-Din’s notoriously ornate rhymed prose for its narrative content, which he reproduced in an unadorned manner along with selections from the Barq’s numerous official documents and poems and occasional additional material of unknown provenance. Now that the surviving portions of the Barq have been edited and published, it is a much simpler matter to compare the two works where they overlap. This comparison confirms Hartmann’s finding that al-Mansur does indeed add material that is not found in the Barq. However, these additions are fairly few. Of the sixty printed pages of the Midmar that can be compared directly with the Barq, only about ten consist of material that is not derived wholly from it. Five of the ten are occupied by a single panegyric to Saladin.22 The structure of the narrative and the vast bulk of its substance are taken straight from the Barq. More interestingly, the author of the Midmar subtly distorts the material from the Barq in a way that the other historians who made heavy use of that work, such as Abu Shama, al-Bundari and Ibn al-Athir, do not. This distortion, which Hartmann did not discuss, takes two principal forms. The first is the abundant use of what we might call the ‘first-person vicarious’. The Midmar, that is, repeatedly copies ‘Imad al-Din’s first-person-plural narrative unaltered. This gives the reader to believe that al-Mansur was party to the events thus described. It is not inconceivable that he, too, took part in some of the same events that ‘Imad al-Din did, and even that he occasionally did so in the same manner, but this cannot have been the case in most instances. The Midmar also makes liberal use of the first-person plural in describing any event in which al-Mansur’s father, al-Muzaffar, was reportedly involved. Again, it cannot be ruled out that the young al-Mansur

130  |  syri a i n crusad e r time s was with his father when some of these events took place, but it is difficult to accept that he was on hand for most of them. Secondly, the Midmar almost completely removes the character ‘Imad al-Din from the narrative that it borrows from the Barq, attributing his actions and words to others or omitting them entirely. Given that the Barq has been justly described as ‘Imad al-Din’s memoirs, the effect of this removal on the narrative is quite substantial.23 A few illustrations of these subtle but pervasive distortions will suffice. Regarding the siege of Nisibis in 578/1182, al-Mansur writes that ‘we avoided shedding blood and violating women, we appointed in it men who would prevent any to enter or leave it, and we authorised people to oversee the citadel and its superintendent’. These words are taken virtually verbatim from the Barq.24 It is perfectly plausible that ‘Imad al-Din helped in formulating and implementing these decisions, but it is quite unlikely that the teenage al-Mansur was involved, since there is no evidence that he took part in Saladin’s administration at this tender age, or any other. Yet because the Syrian sections of the Midmar are riddled with other first-person references that can only refer to al-Mansur – notably the repetitious ‘my father’ that is appended to all mentions of al-Muzaffar – the reader of this passage in the Midmar naturally envisions al-Mansur at Saladin’s elbow in Nisibis. Shortly thereafter, before the walls of Mosul, the Midmar depicts Saladin alone as he ‘flatly refuses a truce, or to accept any mediation’. In the Barq, ‘Imad al-Din claims some part of this intransigence; for him it is ‘we’ who refuse, though in precisely the same words.25 The accuracy of the Midmar suffers more grievously a few lines further on, when, as the Barq has it, ‘Imad al-Din is appointed as one of Saladin’s three representatives at the peace talks convened between his army and the Mosulites. In the Barq, ‘Imad al-Din is the last negotiator standing when the other two representatives tire of Mosul’s obstructionist tactics and depart. In the Midmar, however, ‘Imad al-Din is not so much as mentioned in connection with the talks, so that they effectively break down when Saladin’s second emissary leaves. The dual verbs used to refer to the two emissaries in the Midmar leave no doubt that ‘Imad al-Din has been intentionally excised from his own story.26 Later in the narrative, the Barq relates how al-Muzaffar hastens to join Saladin when, after the failure of the Mosul siege, Saladin departs and the

sy m b oli c conf li ct and coope r a tio n  | 131 Zengids of Mosul muster forces to oppose him. ‘His arrival’ in the Barq is ‘our arrival’ in the Midmar, an elision that would be less suspicious if all the language describing the arrival were not borrowed from the Barq, and if a hash had not been made of the speakers in the exchange that next follows. In the Barq, the impetuous Muzaffar says, ‘Let us go out to them right away’, to which the circumspect ‘Imad al-Din replies, ‘They are many, and there is no harm in caution.’ In the Midmar, it is Saladin indeed who advises restraint, using the same words.27 Later, when the comically greedy figure Abu’l-Hayja’ the Gross surrenders Amid after a siege, the Midmar editorialises that ‘he would have been well advised to take refuge in the sultan’s shade’. This view is in fact owed to the Barq.28 Similarly, the Midmar intimates that ‘we have only mentioned [a certain generous act of Saladin’s] that it might be known that the whole material world had no value for the sultan’. This is the same reason that the Barq gives for including the episode, in the same words.29 When Saladin returns to confronting the Franks, at Baysan in 579/1183, one pictures al-Mansur in the camp when he writes, ‘We expected every day from the Franks their customary frontal assault.’ This expectation, however, is ‘Imad al-Din’s, and in the absence of other evidence it cannot be certain that al-Mansur shared it at the time.30 Even transcripts of those documents that would seem most likely to have been pulled from al-Mansur’s patrimonial archive often turn out to be derived entirely from the Barq, as in the case of the letter that ‘reached us in Egypt, urging my father al-Malik al-Muzaffar … to send the Egyptian armies to the jihad’. This letter is found in the Barq in more complete form, where we learn that it was in fact composed by ‘Imad al-Din himself.31 The broad concern in the Midmar to stress the heroic part al-Muzaffar played in Saladin’s wars, too, leads to selective treatments of the Barq’s material. For example, when al-Muzaffar is dispatched to govern Egypt in 579/1183, the Barq emphasises how al-Qadi al-Fadil was sent with him to advise and indeed to rein him in, since al-Muzaffar ‘had a certain impetuosity (hidda) that [the previous governor] al-‘Adil did not’. The Midmar omits all mention of al-Qadi al-Fadil here, though it does supply some information not found in the Barq: the revenues of certain districts of Egypt. Instead it stresses al-Muzaffar’s unfettered independence, in direct contradiction to the testimony of the Barq.32

132  |  syri a i n crusad e r time s These are only a few examples of widespread tendencies within the Midmar. It should be stressed that this comparison, like Hartmann’s, is exploratory rather than exhaustive. It has not attempted to compare passages in the Midmar that do not overlap with the surviving portions of the Barq, but that do find parallels in other works derived from the Barq, such as al-Bundari’s Sana or Abu Shama’s Rawdatayn. That important labour remains to be done, and there is every indication that the Midmar preserves information from the lost portions of the Barq that these sources omit. Neither should it be forgotten that the Midmar does add unique material, even for the period before al-Mansur’s adulthood, much of it having to do with his father al-Muzaffar. For this reason alone, historians of the period should consult the Midmar more regularly than they have done. Nevertheless, the Midmar must be used with caution as a source of detailed information about events in Syria. It cannot be assumed that the information it contains about Saladin’s campaigns is either a reliable condensation of the Barq or al-Mansur’s own ‘numerous biographical data as well as reports of eyewitnesses’. Indeed, one may doubt whether the published portion of the work provides a single autobiographical datum or eyewitness account that is truly al-Mansur’s own. What appears to be his first-person testimony turns out in virtually every verifiable instance to be either ‘Imad al-Din’s, or else al-Muzaffar’s, as narrated in the third person in the Barq and mechanically converted to the first person in the Midmar. Statements about the agency of major figures surrounding Saladin in the Midmar must be used with special caution, since its author has systematically excised mentions of ‘Imad al-Din, the principal informant for him and a key actor in the Barq. The chief exception are a few instances in which ‘Imad al-Din is acknowledged as the author of a document that the Midmar reproduces. These instances aside, it is clear that the author of the Midmar intentionally suppressed evidence that he had used another source. Thanks to the newly discovered portion of the work, however, we can now study portions of the Midmar that include more reliably autobiographical material, notably the description of al-Mansur’s first leadership role, in Mayyafariqin. We shall find that even this portion of the work is heavily dependent on unacknowledged borrowings from an earlier one – the Ta’rikh

sy m b oli c conf li ct and coope r a tio n  | 133 Mayyafariqin of Ibn al-Azraq – and that most of its unique material consists of panegyric poetry addressed to al-Mansur. Al-Mansur in Mayyafariqin We have seen that al-Mansur travelled to Mayyafariqin from Manbij immediately upon his appointment as governor of the city, arriving in Sha‘ban 586/ September 1190. Mayyafariqin (Greek Martyropolis; Syriac Maypherqat) – was a sizeable and well-fortified ancient town of north-eastern Diyar Bakr, surrounded by well-watered mountains and pasturelands.33 The young prince reports that he was quite taken with it when he arrived: ‘I found it the noblest place in Diyar Bakr and the finest town, with the highest and strongest walls.’ There follow a few lines describing the basic layout of the city, its fortification, and its springs and rivers. The author then relates how he cancelled the unlawful taxes that had previously been levied, evidently by his late brother. Young and old thronged to meet him, and the poets proceeded ‘to praise us … for our justice, which we re-established’.34 This praise goes on for 105 couplets, taken from two panegyrics that were composed for the occasion. After a month of unabated just rule, we are told, the poets trooped out again on ‘Id al-Fitr (1 November 1190) to acclaim their teenage lord. The reader is permitted to revel in thirty more lines of panegyric. The remainder of the volume – three and a half folios – consists of a detailed description of the founding of Mayyafariqin and of its early history, urban topography and inscriptions, and conquest by the Muslim armies in 18/639. This disquisition is introduced with the statement: ‘Those with learning and knowledge of the early campaigns have noted (dhakara ahl al-‘ilm wa-l-siyar) that the site of Mayyafariqin was once covered in thickets, brambles, and reeds.’35 Yet the account that follows evinces a contemporary authorial presence; the distant history of various sites is consistently paired with their condition at the time of writing. On the same folio, for instance, it is noted that a wall from a chapel dating to the time of Christ ‘stands until our time’. A different chapel ‘is now a ruin, though its site remains visible’.36 Another structure is rumoured to have been built in twenty-eight years, but the author declares, ‘I think that this is sheer fantasy on the part of the transmitter, because it is exceedingly improbable that such a great building was erected in such a short time.’37 The author reveals himself perhaps a

134  |  syri a i n crusad e r time s half-dozen other times in the passage, which includes a textual walking tour of the city. One reflexively imagines the young al-Mansur strolling the city’s byways, notebook in hand. This would be a mistake, however, for all of the authorial interjections are in fact in the ‘first-person vicarious’; they are taken straight from the Midmar’s unacknowledged source for the history of the town: Ta’rikh Mayyafariqin of Ibn al-Azraq (d. after 577/1181). Relatively little is known about Ibn al-Azraq, a minor official and scholar of Diyar Bakr whose Ta’rikh Mayyafariqin survives in two incomplete manuscripts in the British Library – Or. 5803 and 6310 – which represent rather different recensions. The early sections of the work that were used in the Midmar have attracted interest in recent years, particularly from Chase Robinson and Harry Munt, who have shed light on the work as a whole and especially on its use of a translated Syriac hagiography of the fourth-century Christian holy man Marutha.38 The Midmar quotes much of the section on Marutha, which had been translated from Syriac to Arabic by a Christian at the request of Ibn al-Azraq. The only significant portions of Ta’rikh Mayyafariqin that have been edited are those which treat the Marwanids and the early Artuqids, which B. A. L. ‘Awad and Carole Hillenbrand published in 1959 and 1990, respectively.39 While many later authors used the work, only three others quoted at length from the section that is found in the Midmar: Yaqut al-Hamawi (d. 626/1229), al-Qazwini (d. 682/1283) and Ibn Shaddad (d. 684/1285).40 Comparison of Or. 5803 to the section on Mayyafariqin in the Midmar – a section that does not survive in Or. 6310 – reveals that Ibn al-Azraq’s work was doubtless the Midmar’s source, making the Midmar a significant early witness to the text of Ta’rikh Mayyafariqin. Little to none of the Mayyafariqin material in the Midmar fails to appear in Or. 5803, and the wording is often identical in the two manuscripts.41 There is, however, some variation in wording, in the ordering of episodes, and in certain details, much as one finds in comparing the Midmar to corresponding sections of the Barq. But there is a better explanation for these discrepancies than to attribute them to the author of the Midmar. Hillenbrand has shown that Ibn Shaddad used the recension of Ta’rikh Mayyafariqin that is represented in Or. 6310 when he composed al-A‘laq al-khatira fi dhikr umara’ al-Sham wa’l-Jazira, not the one represented in Or. 5803, and the text in the Midmar is in fact

sy m b oli c conf li ct and coope r a tio n  | 135 much closer to the corresponding passages in the A‘laq than it is to Or. 5803.42 For example, on fol. 116a of the Midmar, where the narrative corresponds in rapid succession to passages on fols 12b, 1a–b and 5b of Or. 5803, it matches quite closely the order in the A‘laq. It thus seems that the author of the Midmar used the earlier recension of Ta’rikh Mayyafariqin that is represented in Or. 6310. This is not absolutely certain in all instances, however, for while Ibn Shaddad names his source for one of the relevant passages as Ibn al-Azraq’s work, in an earlier case he simply cites ‘a certain history’ (ba‘d al-tawarikh), and it cannot presently be ruled out that his source for this passage was some derivative of Ta’rikh Mayyafariqin, perhaps even the Midmar itself.43 The salient point here, however, is that the content of the Midmar – here in its treatment of Mayyafariqin as in those instances where it corresponds to the Barq – turns out again to comprise two kinds of material. Firstly, there is material derived from an earlier source with no explicit acknowledgement of that source and with the implication that first-person references copied from it are in al-Mansur’s voice. Secondly, there is a smaller quantity of material that is connected to the author himself or his family, and which invariably serves to lionise them in some way. In the final section of this essay, we will propose an explanation for this distribution of material. Symbolic Labour in Midmar al-haqa’iq During Saladin’s siege of Amid in 579/1183, ‘Imad al-Din al-Isfahani ­composed an open letter to the city’s defenders. This document, according to his own subsequent description in al-Barq al-shami, served ‘to entice and terrify, now promising them things and now menacing them’. It had its effect. The Amidites promptly abandoned their miserly master Abu’l-Hayja’ the Gross, who sued for terms. When the author of the Midmar availed himself of ‘Imad al-Din’s account, he suppressed ‘Imad al-Din’s role in his usual fashion, asserting merely that Saladin had ordered ‘one of his entourage’ (ba‘d ashabihi) to compose the document.44 The point in mentioning the episode here, however, is to illustrate that texts could do serious political work in Syria during the era of the Crusades. Although much textual p ­ roduction was addressed to scholarly coteries, and the absence of mass media limited the reach of propaganda, nevertheless poems, epistles and literary works had

136  |  syri a i n crusad e r time s important social and political functions, which they discharged not despite or alongside their literary qualities, but by virtue of them. On the basis of the foregoing analysis, I would propose that the chronicle Midmar al-haqa’iq wa-sirr al-khala’iq, like the epistles and poems of ‘Imad al-Din on which a significant portion of its surviving sections were based, served to do what the sociologist Pierre Bourdieu called ‘symbolic labour’, by which ‘the [self-]interested dimension of human action become[s] transformed into disinterested ideology’.45 As Saladin’s armies conquered Syria, ‘Imad al-Din scratched out a steady stream of texts that glorified his master and demonised the perfidy and infidelity of his rivals, particularly the Zengid lords of Mosul, to say nothing of the Franks. The propagandistic nature of this textual production takes nothing away from its literary artistry, which enabled and enhanced the effect of the propaganda, and there is every indication that ‘Imad al-Din sincerely believed his master’s cause to be just. The Midmar, which we now know to have been composed well after al-Mansur came to rule in Hamah, likewise served to valorise the Ayyubid ruler who stands at the focus of its narrative more and more as it progresses: al-Malik al-Mansur himself, the work’s ostensible author. Indeed, the only two incontrovertibly autobiographical passages in the surviving portions of the work – the brief accounts of al-Mansur’s exploits at Acre and of his arrival in Mayyafariqin – serve to set up transcriptions of panegyrics that were composed to honour him. The reader who comes to the Midmar in search of a fresh, princely perspective may thus be disappointed. Given this curious shortage of personal observations concerning scenes at which the author was, after all, present, coupled with the pervasive implicit co-option of other authors’ first-person remarks (that is, those of ‘Imad al-Din and Ibn al-Azraq), one might speculatively question the extent to which the Midmar was really the work of al-Mansur. In Hamah, al-Mansur surrounded himself with a notably large group of scholars, whose interests ran the gamut that one could expect to find at an Ayyubid court.46 They will have provided a pool of research assistants as well as an appreciative audience for any works their patron issued, which works might augment his prestige, as well as their allegiance to this prince-aftera-scholar’s-own-heart. In this light it is interesting to consider the choice of sources for the surviving portions of the Midmar. The Barq and Ta’rikh

sy m b oli c conf li ct and coope r a tio n  | 137 Mayyafariqin shared a potentially convenient quality, in addition to their recent vintage and limited circulation: they included their own authors as participants and/or commentators regarding events and places at which alMansur had also been present. It was a simple matter to use their texts, without explicit acknowledgement and including their first-person references, while effacing instances in which they themselves appear. It is obvious that humans represent and conduct conflict and cooperation in intangible, symbolic forms. These were no less ubiquitous in Syria of the Crusading era than in other societies, and thus too in the work of modern historians.47 Ayyubid princes competed and allied to obtain titles and renown for the sake of these intangible things in themselves, as well as for the material benefits to which they were tied. Texts like the Midmar, themselves made of symbolic stuff, bore a complex relationship to the pursuit of symbolic desiderata; they were at once its products, records and instruments. Close study of the Midmar suggests that its author chose and arranged his sources, sometimes taking considerable liberties even by the historiographical standards of the day, with the goal of elevating al-Mansur’s prestige by lionising him and his family directly and by linking them to the paradigmatic example of Saladin.48 This observation casts a rather different light on Shihab al-Din al-Qusi’s observation about the Midmar and its author. ‘It is large and valuable,’ he said, according to al-Safadi. ‘It indicates his excellence, and no one had ever written anything quite like it.’49 Conclusions The chronicle Midmar al-haqa’iq wa-sirr al-khala’iq, attributed to an Ayyubid ruler of Hamah, al-Malik al-Mansur, survives at about twice the length that has generally been thought. A newly discovered Tunisian manuscript allows more precise conclusions to be drawn about the extent and date of the work. Comparing surviving portions on Syria to their unacknowledged sources – notably al-Barq al-shami of ‘Imad al-Din al-Isfahani and Ta’rikh Mayyafariqin of Ibn al-Azraq – reveals that the Midmar depended more heavily on its sources than has been recognised. Its author suppressed evidence of that dependence and transcribed his sources’ first-person testimony in such a way as to imply that al-Mansur was narrating the events. Thus, while the Midmar retains a great deal of importance as a source and stands in need of

138  |  syri a i n crusad e r time s an extended and updated edition, it must be used with considerable caution for events in Syria and the Jazira. Like any chronicle, it was an instrument of symbolic competition as well as an arrangement of historical data. As a work that bore the name of a Syrian prince, more a grandee who traded in power than a scholar whose business was knowledge, it served to extol its putative author in a particularly powerful manner.

sy m b oli c conf li ct and coope r a tio n  | 139 Appendix: The Contents of BNT 13597 as They Pertain to Syria Fols 1a–22a:  [events in Iraq] Fols 22a–23b:  Saladin’s movements in AH 582: travel to Hims, Hamah and Damascus – various appointments – al-Muzaffar in Egypt feigns a move toward the Maghrib, but Saladin prioritises Jerusalem and summons alMuzaffar, leaving al-Mansur in charge of the forces in Egypt – al-Muzaffar is granted Hamah and several smaller towns – al-Mansur is summoned from Egypt – a rumour of the apocalypse in Egypt – the death of the atabeg Pahlawan – an alliance among the Franks – Reynald of Châtillon’s breaking of the truce. Fols 23b–26a:  [campaigns of Qaraqush in the Maghrib]; al-Mansur, Midmar (ed. ‘Ar‘ar), 75–80. Fols 26a–38b:  the year AH 583 [events in Iraq] Fols 38b–55a:  Saladin’s deeds in AH 583: departure from Damascus – Battle of Hattin and aftermath – surrender of Tiberias – killing of Templars and Hospitallers – move to Acre – taking of numerous fortresses – capture of Sidon, Tyre, Jubayl – transcript of letter by al-Qadi al-Fadil to caliph detailing towns and fortresses captured – affairs of the Franks in Tyre and Acre – siege of Ascalon and capture of surrounding towns and fortresses – siege of Jerusalem – Balian negotiates with Saladin – terms of surrender – entry to the city – letter by ‘Imad al-Din to the caliph concerning the victory – the first Friday communal prayers in Jerusalem, with transcript of the sermon given – arguments about payment of the agreed ransom for the lives of the city’s defenders – armies move to Tyre and Acre – praise poetry for Saladin. Fols 55a–58b: [campaigns of Qaraqush in the Maghrib]; al-Mansur, Midmar (ed. ‘Ar‘ar), 80–89. Fols 58b–77b:  the year AH 584 [events in Iraq] Fols 77b–82a:  Saladin’s deeds in AH 584: movement from Acre to Kawkab – Qaraqush returns from North Africa and is appointed at Kawkab – Saladin goes to Damascus, summons armies of Syria and the Jazira to him at Jerusalem – al-Muzaffar complies – sieges of Tartus, Jabla, Latakia (which is granted to al-Muzaffar), Sayhun, and so on – the sultan’s chivalrous deeds – truce with Antioch – move to Aleppo – conquests of Karak, Kawkab, Safad – dispatch of al-‘Adil to Egypt.

140  |  syri a i n crusad e r time s Fols 82a–87b:  the year AH 585 [events in Iraq] Fols 87b–99b:  Saladin’s deeds in AH 585: summons Qaraqush to Acre – proceeds to Damascus – the naming of al-Nasir’s son as heir apparent to the caliphate – copy of congratulatory response – Diya’ al-Din al-Shahrazuri dispatched to Baghdad with the cross of the Holy Sepulchre – Saladin moves to Marj ‘Uyun – skirmishing – movement of Franks from Tyre to Acre – Muslim armies summoned – fighting before Acre [fol. 92 belongs to the Iraq portion of the narrative] – Battle of Acre, with first-person account of al-Mansur’s participation – letter recounting the victory – congratulatory letters arrive – panegyrics to leading amirs, including al-Mansur himself – Saladin moves to Kharruba, described as ill-advised – Franks dig a defensive ditch – al-‘Adil arrives from Egypt – emissaries of the Seljuq sultan Tughril III arrive – Egyptian fleet arrives – al-Muzaffar intercedes repeatedly with Saladin on behalf of Muhyi al-Din b. ‘Asrun, the qadi of Damascus – Saladin remains at Kharruba. Fols 100a–105b:  the year AH 586 [events in Iraq] Fols 105b–116b:  campaigns in Syria in AH 586 – Saladin at Kharruba, keeping watch on Acre and occasionally hunting – al-Mansur attends on him with al-Muzaffar – a Frankish sally at Acre (Waq‘at al-raml) [fols 106 and 107 are bound out of order] – Saladin moves to Tell Kaysan – emissaries of the caliph arrive – failed Frankish attempt to break the siege using towers (when these are burned, al-Mansur describes riding out with Saladin to take in the sight) – news of the victory sent abroad – armies arrive from the Jazira – news arrives of Frederick Barbarossa’s departure for Syria, difficult passage through Anatolia – German dealings with Qilij Arslan and Leo II of Cilician Armenia – drowning of Frederick and aftermath – assassination of Conrad of Montferrat – news of the death of al-Mansur’s brother, governor of Mayyafariqin, reaches him while he is Manbij and al-Muzaffar is in Hamah – al-Mansur proceeds to Mayyafariqin – description of Mayyafariqin – copious encomia for al-Mansur – al-Mansur sets to work ruling the city justly – more encomia – how Mayyafariqin was founded – a description of the city – the Sasanian conquest thereof – the Muslim conquest – colophon for eleventh volume, lacking date and scribe.

sy m b oli c conf li ct and coope r a tio n  | 141 Notes   1. Al-Mansur, 1968, repr. 2005.   2. Hartmann, 1986a, 571–607; Hartmann, 1984/1986b, 336–367; Hartmann, EI 2, s.n. ‘al-Mansur, al-Malik Muhammad b. ‘Umar’, VI: 429–430. See also Kalbhenn, 1975. (I owe my copy of this thesis to the kindness of Hannah-Lena Hagemann).   3. In this essay, ‘Syria’ is used as an expansive shorthand for the Bilad al-Sham, with the addition of the Jazira. For the importance of the Midmar with respect to North Africa and Iraq, see respectively, for example, Baadj, 2013, 267–295, Hartmann, 1975, 14–17.   4. Abu’l-Fida’, 1907, tr. P. M. Holt, 1983, 9. In this respect the Midmar would also appear to bear comparison to the work of history by ‘Abdallah b. Buluggin, the last Zirid ruler of Granada, the Kitab al-tibyan.   5. The undated manuscript, Bibliothèque Nationale de Tunisie 13597 (116 fols), henceforth BNT, was identified by Murad ‘Ar‘ar (see al-Mansur, 2012). Though bound separately, it was originally an unbroken continuation of the Tunis manuscript used by Hasan Habashi for the 1968 edition; it picks up precisely where that manuscript ends. Thus Habashi’s description applies to it, too, in many respects. It is unclear why Habashi overlooked this second manuscript. Both were wrongly catalogued as Ta’rikh al-Badr fi awsaf ahl al-‘asr by al-‘Ayni, Badr al-Din (Mansur, 1969, 366, nos 4938–4939).   6. On al-Muzaffar, see Hartmann, EI 2, VI: 816–18. Unless otherwise noted, the narrative here relies upon the studies cited in n. 2.   7. On Alexandria in this era, see Walker, 2014, 36–48.   8. BNT 13597, fols 93a, 96a, 105b.   9. BNT 13597, fols 108b–109a. 10. The extant section of his biographical dictionary of poets has now been published from the Leiden manuscript (al-Malik al-Mansur Muhammad b. ‘Umar, Akhbar al-muluk. His unpublished literary miscellany, Durar al-adab wa-mahasin dhawi’l-albab, survives in three manuscripts, in Leipzig, Rabat and San‘a’; see Yarbrough, 2016, 139–169, 166. 11. Ibn Wasil, 1977, IV: 78; cf. Hartmann, 1986a, 581; Hartmann, EI 2, VI: 429. 12. On fol. 116b of BNT 13597, the author promises that ‘the year 597 will follow in the twelfth …’ 13. BNT 13597, fol. 77b. 14. Hartmann, 1986a, 590 and studies cited there.

142  |  syri a i n crusad e r time s 15. Hartmann, EI 2, VI: 430; Hartmann, 1986a, 571. 16. Hartmann previously identified ah 583, 587 and 589 as possible termini post quos, based on the conquest of Jerusalem heralded in the edited portion, blessings invoked on the memory of al-Muzaffar, and the general attitude toward Saladin, and 602 as a terminus ante quem, based on a citation in another of al-Mansur’s works (1986a, 584–585, 604; EI 2, VI: 429–430). 17. The blessing, which could of course be the addition of a scribe, is ridwan Allah ‘alayhi. For the death date, see al-Safadi, 1974, 169–171, no. 1706. 18. Hartmann, 1986a, 581, 585, 603–604. 19. Cf. Kalbhenn, 1975, 91–92. On ‘Imad al-Din, see Richter-Bernburg, 2014, 29–51 with extensive bibliography. 20. ‘Imad al-Din al-Isfahani, 1987, III: 161–181 overlap with al-Mansur, 1968, 15–30. ‘Imad al-Din al-Isfahani, 1987, V: 17–172 overlap with the Syrian portions of al-Mansur, 1968, 96–164 (that is, 96–115, 138–164). Thus, certain hypotheses proposed cautiously by Hartmann, notably that the manuscript (which is undated but doubtless early) ‘was probably written shortly after Ramadan 589/September 1193’, and that the Midmar could thus be of importance for dating the composition of the Barq, may be set aside for the time being. 21. Gibb, 1950, 58–72; Richards, 1980, 46–65. 22. Al-Mansur, 1968, 20–24. Other unique reports are not insignificant, however, including information about al-Muzaffar’s campaign against Qilij Arslan at the time of Marj ‘Uyun (45–46), the siege of Bayt al-Ahzan (33–34), al-Muzaffar’s role in the surrender of Aleppo (142–143) and details related to the surrender of Harim (145–146). 23. Gibb, 1953/55, 93–115, 98. 24. Al-Mansur, 1968, 106 (emphasis mine); ‘Imad al-Din al-Isfahani, 1987, V: 30. 25. Al-Mansur, 1968, 108; ‘Imad al-Din al-Isfahani, 1987, V: 37. 26. Al-Mansur, 1968, 108–109; ‘Imad al-Din al-Isfahani 1987, V: 38–39. For the Jazira campaign of this year as narrated in the Barq, see Gibb, 1953/55, 102–106, and more generally Ibn al-Athir, Part 2, tr. Richards, 2006–8, 284–295; Hirschler, 2006, 86–92; Durand-Guédy, 2008. 27. Al-Mansur, 1968, 113; ‘Imad al-Din al-Isfahani, 1987, V: 81–82. 28. Al-Mansur, 1968, 138; ‘Imad al-Din al-Isfahani, 1987, V: 95. 29. Al-Mansur, 1968, 139; ‘Imad al-Din al-Isfahani, 1987, V: 91. 30. Al-Mansur, 1968, 153; ‘Imad al-Din al-Isfahani, 1987, V: 148. 31. Al-Mansur, 1968, 163; ‘Imad al-Din al-Isfahani, 1987, V: 171–172. 32. Al-Mansur, 1968, 154; ‘Imad al-Din al-Isfahani, 1987, V: 156.

sy m b oli c conf li ct and coope r a tio n  | 143 33. See Minorsky and Hillenbrand, EI 2, VI: 928–932. 34. BNT 13597, fol. 109a. 35. BNT 13597, fol. 113a. 36. BNT 13597, fol. 115a. 37. BNT 13597, fol. 115b. 38. This particular hagiography of Marutha survives only in Ibn al-Azraq’s work and its derivatives, though separate ones exist in other languages. See Robinson, 1996, 7–27; Munt, 2010, 149–175, and studies cited in both essays, particularly those by Amedroz and Fiey. 39. Ibn al-Azraq, 1959; Hillenbrand, 1990, with exhaustive treatment of earlier literature. 40. Munt, 2010, 150 n. 4. 41. BNT 13597, fols 113a–116a correspond, with some internal rearrangement, to BL Or. 5803, fols 8b–12b, and BNT 13597, fols 116a–116b to text from BL Or. 5803, fols 1a–1b and 5b–6a. 42. Hillenbrand, 1990, 18, 23. 43. Ibn Shaddad, 1978, III/1: 278–279, 260. 44. Al-Mansur, 1968, 136; ‘Imad al-Din al-Isfahani, 1987, V: 91. 45. Swartz, 1997, 93. 46. Al-Safadi, 1974, 260. See more generally Hirschler, 2008, 95–132; Brentjes, 2011. 47. For example, Choksy, 2007; Berger, 2014; Kadi, 2012. For ecological analogues, see Levin et al., 2009, 186–219, 219–238, 253–422. 48. In addition to stressing Saladin’s virtue, al-Muzaffar’s importance to Saladin, and al-Muzaffar’s paternal connection to the author, one could also note that one of the few extraneous details that the Midmar regularly adds to its borrowings from the Barq is the cancellation of unlawful taxes (mukus and dara’ib) upon the conquest of each stronghold (for example, al-Mansur, 1968, 106, 111). This is, naturally, the very thing that al-Mansur does upon taking over from his brother in Mayyafariqin. 49. Al-Safadi, 1974, 260; Ibn Shaddad, 1978, III/1: 260–268, 278–282. Acknowledgements  I would like to thank the conference participants, especially Carole Hillenbrand, as well as Harry Munt for his comments on a draft of this essay. For research support and assistance I am grateful to the Dean’s Travel Fund of Saint Louis University, the New York University Abu Dhabi Institute, the Bibliothèque Nationale de Tunisie and particularly Dr Rachida Smine, and the British Library.

8 A Critique of the Scholarly Outlook of the Crusades: the Case for Tolerance and Coexistence Suleiman A. Mourad

T

he Crusader period is defined in scholarly research as one of the most brutal and bloody chapters of Christian–Muslim relations in the Middle Ages. No doubt, the period witnessed a heightened level of religious militancy and propaganda between Muslims and European Christians in the medieval period. However, this was not the only reality at the time. There were countless instances of tolerance (religious and social), alliances (political and military) and exchange (commercial, cultural and scientific) between Muslims and Crusaders. Scholars have acknowledged some of them, but only as cases of marginal historical curiosity or opportunism and realpolitik. In their totality, such instances have not been studied or conceptualised as possibly shaped and informed by other factors including religion, or as forming a pattern that reflected the different agendas of the various actors during the Crusader period. There is another problematic aspect of the Crusader period, namely the way it is often perceived as an epoch stretching from 1095 to 1291. When the Crusader period is treated as such, it exacerbates a tendency – that many have already challenged and criticised (for example, by Housley)1 to think of it as separate from what came before 1095 and what came after 1291. The argument of this chapter is that the dominant discourse of the clash between the worlds of Christendom and Islamdom has determined the contours of scholarly conventions on the Crusades in the Middle East. Thus, it has undermined the scholars’ ability to understand the period as

t h e c ase f or tolera nce a nd coex is te nce  | 145 involving many actors on each side with conflicting agendas, and a complex web of exposures, cooperation and animosities that not only unfolded during the Crusader period, but also preceded it and endured after it. When the Crusader period is re-examined with an awareness to these problematic issues, the exceptions to the dominant outlook of warfare – such as trade, political and military cooperation, religious tolerance, scientific exchange, and so on – cease to be isolated cases and become patterns, and the military clash ceases to be the dominant discourse and is reduced to another pattern. The Problem with the Emphasis on al-Sulami and his Book of Jihad In two separate articles, which he later incorporated into his book L’Islam et la croisade, Emmanuel Sivan advanced two theories that have had tremendous impact on the fields of Jerusalem studies and Crusader studies. In the first theory, Sivan postulated that the religious and political significance of Jerusalem to the Muslims only became an important matter because of the Crusades.2 Despite its popularity, especially among non-specialists in Islamic history, this theory has already been proven wrong by several scholars on the basis of textual and archeological evidence.3 In the second theory, Sivan introduced al-Sulami (d. 1106) and his Book of Jihad as reflecting the first ‘Islamic’ reaction to the Crusades,4 ushering a previously unknown figure into the centre of scholarship on Islam and the Crusades. The appeal and acceptance of the theory about al-Sulami has been sustained by scholarly conventions that gave it potency and persuasiveness.5 I am not contesting that al-Sulami wrote his Book of Jihad in reaction to the Crusader invasion and what he considered as the Muslims’ religious lapses that allowed the invaders to accomplish their objectives. What I am contesting is the significance accorded to him in modern scholarship as the herald of the first Islamic response to the Crusade, which in my opinion reflects the scholarly convention that for a reaction to be labelled ‘Islamic’ it must be inherently militant and religiously fanatical, and so other reactions that include cooperation with the Crusaders or indifference to them cannot be deemed Islamic. Moreover, the evidence at our disposal shows that al-Sulami was a marginal religious scholar and does not represent the views of the religious

146  |  syri a i n crusad e r time s establishment in Damascus or Syria. Irrespective of what his book conveys, his peers and later scholars did not consider him worthy of mention, and in all medieval prosopographies, we only have one rather very short biography of him.6 Additionally, al-Sulami preached his book outside Damascus in the mosque of Bayt Lihya, which at the time played no known role of any scholarly or social significance. Except for two figures who attended the reading of one section of the book, those who listened to al-Sulami’s preaching (ranging between three in most cases and up to eight people in just one instance) were not scholars and seemed to occupy minor functions. Most importantly, al-Sulami’s book does not have a single demonstrable case of impact. In other words, even though al-Sulami wrote a book on jihad, as far as we can tell no one who wrote or preached on jihad during the Crusader period read it or mentioned it.7 Al-Sulami’s sentiments surely reflected the views of some people in Damascus, including some religious scholars. However, given the indifference shown to him and his book, it is not an exaggeration to say that the majority of the Damascene society and its political and religious elite did not share his enthusiasm. Therefore, when his book is presented as the first Islamic response to the Crusader onslaught, we are imposing a specific conceptualisation that for a reaction to be considered Islamic it has to fit the parameters of what al-Sulami expressed. Other responses – including indifference to the Franks and cooperation with the Franks – are, according to such a convention, non-Islamic. A similar case to al-Sulami’s is the significance accorded to al-Findalawi, a Maliki scholar who came to Damascus from Muslim Iberia, who rushed out of Damascus to wage jihad against the Crusaders attacking Damascus in 1148.8 Again, the mention of al-Findalawi is repeated in modern scholarship as another example of the Islamic reaction to the Crusades.9 Yet, the fact that our sources only name him shows that his behaviour stood in contradistinction to that of the overwhelming majority of local Damascene religious scholars who, as far as we can tell, did not move a finger.10 Why their collective reaction (that is, abstaining from fighting) is not seen as Islamic is precisely the point that this chapter is criticising: scholarly convention on the Crusades does not consider ‘Islamic’ anything other than militancy and fanaticism. There is no doubt that the reactions of al-Sulami and al-Findalawi

t h e c ase f or tolera nce a nd coex is te nce  | 147 are one type of Islamic responses to the Crusades, and they must always be presented as such, not as the only type. Evidence of Tolerance and Coexistence Stories of non-violent interactions between local Muslims and Franks are encountered everywhere in the sources. They reveal a significant degree of tolerance and cooperation (obviously, I am not intending these concepts to be understood according to modern standards and usages). They problematise the dominant scholarly outlook and undermine any simplified presentation of the period as governed by animosity, war and religious fanaticism. For instance, Ibn Wasil (d. 1298) describes the relinquishing of Jerusalem to the Franks in 1244 by al-Malik al-Salih of Damascus and al-Nasir Da’ud of Karak: The Franks entered Jerusalem and took control of the sacred Rock, the Aqsa and all the sacred sites on the Noble Haram … At the end of this year, I travelled to Egypt, and stopped on my way in Jerusalem. I saw the monks and priests over the sacred Rock, and wine jars on it intended for the holy communion (bi-rasm al-qurban). I went to the Aqsa Mosque and saw there a suspended bell. The call for prayer and dwelling in the Noble Haram were abolished, and heresy was declared.11

Ibn Wasil’s story is plagiarised by al-Maqrizi (d. 1442), with some minor variations that have major implications to the central point of this chapter: Al-Nasir Da’ud and al-Salih Isma‘il reached out to the Franks and enlisted them against al-Malik al-Salih Najm al-Din. They promised them Jerusalem and gave them Tiberius and Ascalon, and the Franks rebuilt their castles and fortifications. They also took hold of the Rock in Jerusalem and sat on it drinking wine, and even hung a bell in the Aqsa Mosque.12

There are a few points in this event as reported by the two historians that speak to the complex web of animosity and cooperation between Muslims and Crusaders, as well as to the manipulation of history on the part of medieval chroniclers. To start with al-Maqrizi, it seems rather evident that his paraphrasing of Ibn Wasil’s language was intended to cast the Crusaders’­ capture of Jerusalem in 1244 as an intentional desecration of Islamic sacred

148  |  syri a i n crusad e r time s sites: ‘The Franks took hold of the Rock in Jerusalem and sat on it drinking wine.’ But Ibn Wasil’s version does not show a process of desecration. No doubt, his words echo a sense of humiliation or disappointment for losing control of the Haram of Jerusalem (Temple Mount). Yet, Ibn Wasil’s words clearly show that the Crusaders venerated the Rock, and he understood and described the presence of wine there as specifically intended for the service of the Eucharist. Thus, what al-Maqrizi described as deliberate desecration, the eyewitness Ibn Wasil saw as veneration. Moreover, Muslims were still permitted to visit and pray, though they were not allowed to dwell in the Haram. The two narratives, therefore, reflect different Muslim outlooks about the period and about interactions with and perceptions of the Crusaders. Al-Maqrizi’s is rooted in animosity and peculiar negative historical memory, whereas Ibn Wasil’s is reflective of dynamics that involved tolerance and coexistence. This example raises a bigger issue that has not been given due attention in modern scholarship, namely authors as actors and also readers as actors in relation to the transmission/reading of history and how they establish conventions that last for centuries. To return to this episode of handing over Jerusalem to the Crusaders in 1244, the decision was made by al-Nasir Da’ud and his uncle al-Salih Isma‘il, and was part of a broader complex web of alliances and competitions among Crusader and Muslim rulers at the time. It was not the first time Jerusalem was turned over or shared with the Franks. One particular case occurred fifteen years earlier when the Ayyubid sultan al-Malik al-Kamil handed over Christian Jerusalem to Frederick II when the two rulers signed a treaty. At that time, al-Nasir Da’ud was ruling Damascus and opposed the agreement, despite the fact that al-Kamil was able to maintain Muslim control over the Haram and most of the countryside around Jerusalem. Al-Nasir Da’ud enlisted Sibt b. al-Jawzi (d. 1256) to preach against the deal in the Great Umayyad Mosque in Damascus. The celebrated scholar obliged and delivered a very emotional sermon on Jerusalem and its religious merits that left everyone who crammed into the Umayyad Mosque weeping for the loss of the Holy City.13 Interestingly, the handing over of Jerusalem in 1244, according to terms much worse than the treaty of 1229, was not even mentioned by Sibt b. al-Jawzi. So, was al-Nasir Da’ud in 1244 an opportunist by handing Jerusalem back to the Crusaders? Or was he an opportunist to reject the agreement

t h e c ase f or tolera nce a nd coex is te nce  | 149 between his uncle al-Kamil and Frederick II in 1229 and enlist members of the religious establishment to preach against it? To ask the question differently, which of his actions can we describe as ‘Islamic’? The answers generally proposed reflect stereotypes, whereby religion is invariably used to justify fanaticism and war, and realpolitik and pragmatism are used to explain peace overtures and cooperation. Religious tolerance, and sharing sacred religious sites in particular, was a pattern for which we find many examples during the Crusader period. One example, rather overused in modern scholarship, relates to the famous medieval Iberian traveller Ibn Jubayr (d. 1217). In late summer 1184, Ibn Jubayr journeyed from Damascus to Acre in the hope of catching a ship to take him back home after a year and a half in the Levant. He observed on his way countless farming villages inhabited by Muslims who seemed to him to live in complete harmony with the Crusaders. What irritated Ibn Jubayr the most was not only that the Franks were not harming those Muslims. He bemoaned the fact that those Muslims did not seem to be bothered by their mingling with what he described as ‘Christian pigs and filth’, and even though they were still practising Islam openly, he considered living under those circumstances does not allow for the proper practice of Islam.14 In another instance, while in Acre, Ibn Jubayr visited a sacred site called the Spring of the Cattle – the spot where, according to religious belief, God caused a spring to gush out so that Adam could provide water for his cattle, and where he was buried. There, Ibn Jubayr saw the remains of a mosque, and described the following: The Franks had installed for themselves a prayer niche on its eastern side. The Muslims and the heretics gather in it, each turning towards their prayer direction. It is under the control of the Christians; they venerate it and protect it. God saved a prayer spot in it for the Muslims.15

It seems rather evident that it was the Crusaders, not God, who allowed the Muslims to keep coming to that spot and did not mind the two communities having their respective religious services side by side. Ibn Jubayr’s anger and disgust at the sight of fellow Muslims living in harmony with Christian Crusaders, and, even worse, having their religious practices respected by

150  |  syri a i n crusad e r time s the Crusaders, are the kind of reactions that speak of a mindset, not unique to Ibn Jubayr. This mindset reflects the agendas and biases of medieval historians who saw the Crusader period as one of animosity and war. Yet, it is precisely because his accounts offer a treasure trove of evidence about tolerance and shared religious heritage that to accept his ideological stance as normative disregards the complex reality of the period. What Ibn Jubayr saw was very common. When Jerusalem was under Crusader control and prior to its recapture by Saladin in 1187, Muslims regularly visited it to worship at the Haram. John of Würzburg, who came on pilgrimage in the 1160s, described how the Muslims were allowed to visit and worship outside the Dome of Rock.16 This is also corroborated by contemporary Muslim chroniclers, such as Abu Shama, who spoke of group of Sufi mystics having made a pilgrimage to Jerusalem while under Crusader control.17 One can also add the case of the famous poet/diplomat Usama b. Munqidh (d. 1188), who described his own visits to Jerusalem whilst under Frankish control as frequent, and had friends among the Templars who would clear up a space for him to pray inside the Aqsa Mosque. In one instance, Usama was harassed by a newly arrived Frank who took offence at the sight of a heretical Saracen praying in the direction of the south (that is, Mecca). The novice Frank stormed to interrupt Usama’s prayer, holding Usama’s head and turning it toward the east. Usama’s Templar friends rushed to his defence and pulled away the new Frank.18 What is interesting in this story is not only the respect some members of the Templar order had for some Muslims and their right to pray in the Aq‚a Mosque. What is stunning for us is Usama’s own admission that his visits to Jerusalem were frequent, and that he had many Crusaders (in Jerusalem, and elsewhere) as close friends. We also have the example of the scholar al-Harawi (d. 1215) who related that he took up residency in Jerusalem for a few weeks in 1173.19 He also went constantly to the Haram to pray, even though the Dome of the Rock was being used by the Crusaders as a church, but they, according to alHarawi, did not make any serious alteration to it or to the Aqsa Mosque.20 In his Kitab al-isharat, al-Harawi criticised and deconstructed many of the popular customs of making pilgrimage to particular shrines and religious

t h e c ase f or tolera nce a nd coex is te nce  | 151 sanctuaries in Palestine and the surrounding regions as reflective of popular superstitions and false associations. Yet, his book attests to countless cases of Muslims, Christians and Jews converging on the same spots to worship. Some of these locations were under Crusader rule and others were under Muslim rule. Sites as far away from Jerusalem as the Church of Our Lady of Saidnaya, near Damascus, witnessed Muslims and Templars converging on a Christian Orthodox shrine. Also, when al-Kamil gave back Bethlehem and the Church of the Nativity to Frederick II, the sultan insisted that Muslims have a right to visit it and worship there.21 The fascinating thing about these examples is that the reality at the time featured systematic tolerance in relation to religious observance. We are not looking here at examples of Muslims or Crusaders concealing their religious beliefs and identities in order to avoid persecution or mistreatment when they were travelling through or living in each other’s territory. Most of the examples discussed above point to Muslims who were determined to publicly display their adherence to Islam in Crusadercontrolled areas. Muslims also secured and protected Crusaders’ right to sacred sites and safe conduct to them. When Saladin captured Jerusalem in 1187, he and his advisors debated the fate of the Church of the Holy Sepulchre. The words of Saladin’s secretary ‘Imad al-Din al-Isfahani (d. 1201) leave no doubt that the sultan and most of his entourage, including religious scholars, were cognisant of their religious obligation to protect the site for the Christians: The majority of [Saladin’s] advisors declared that [the Church] should not be demolished or razed, and its gates should not be locked barring the infidels from making the pilgrimage to it. Their target of worship is the spot of the Cross and [Jesus’s] grave, not the building itself. Even if it were to be shattered to pieces, the Christians in all their diversity will keep coming to the site. When Commander of the Faithful ‘Umar, may God be pleased with him, conquered Jerusalem in the early years of Islam, he confirmed their right to the place and did not order the structure to be demolished.22

Saladin’s triumphal conquest of Jerusalem came at a time when the Crusaders were on the brink of complete elimination from Palestine. The sultan could have confiscated or destroyed the Church of the Holy Sepulchre as some

152  |  syri a i n crusad e r time s in his court urged him to do. However, the voice of the majority of commanders and religious scholars prevailed. Their awareness of their religious obligation to protect the Christians’ right to come on pilgrimage and worship in Jerusalem was not a matter of political opportunism but rather an Islamic precedent set by the caliph ‘Umar (r. 634–4). A few years after this conversation in Saladin’s court, the events of the Third Crusade unfolded. The final settlement between Saladin and Richard the Lionheart correlate with the argument made above. Saladin opened the city to Crusader pilgrims who came in large numbers and he even held a feast for them in Jerusalem.23 There is one additional interesting detail that was under consideration between Saladin and Richard: the proposal to have Saladin’s brother al-‘Adil marry Joan, the sister of Richard, and have the married couple jointly rule over Jerusalem and Palestine. This proposal shows that the presentation of the period as governed by a religious clash is all the more illogical. Saladin and al-‘Adil could only advance such a proposal after it was vetted and approved by religious scholars, and the idea, even though it appealed to Richard and some in his court, was rejected by certain Christian clerics in Joan’s entourage.24 These examples highlight the widespread cases of cooperation, coexistence and religious tolerance between Muslims and Crusaders. Many of them were informed by factors other than realpolitik and opportunism. Crusader and Muslim rulers did not only defend the religious rights of their respective communities; they also acknowledged in no passive terms the religious rights of the other group. The Peace of al-Kamil and Frederick The circumstances of al-Kamil turning over Christian Jerusalem to Frederick II, as part of a peace the two monarchs forged in 1229, is another case that speaks to alliances and tolerance where religion was a factor. The dominant view in Crusader scholarship is that al-Kamil did so out of political opportunism and realpolitik.25 The evidence, however, points otherwise. When Frederick arrived in Acre in the autumn of 1228, al-Kamil was at the pinnacle of his power. Internally, the sultan had no serious challengers, and was reorganising the Ayyubid realm along terms very favourable to him. The major

t h e c ase f or tolera nce a nd coex is te nce  | 153 headache was the potential alliance between his brother al-Mu‘azzam and the Khwarazmians,26 but that fell apart with al-Mu‘azzam’s death in 1227. Al-Mu‘azzam’s son al-Nasir Da’ud was too young and immature to be able to secure his own transition to power in Damascus, let alone represent any threat to al-Kamil. With respect to Frederick, there is every reason to believe that even though he was keen on fulfilling his vow and lead a crusade, he did not come to fight. Because of Rome’s intrigues, the emperor was facing a precarious situation in Italy and could not afford a prolonged stay in Palestine. He was also excommunicated, which caused most of the local Franks to snub him or conspire against him when he arrived in Palestine, a matter that was known to the Muslims.27 More importantly, he came invited, for al-Kamil in 1226 had dispatched his confidant, Amir Fakhr al-Din, to Sicily to try to hammer out a deal. The idea appealed to Frederick and he followed up on al-Kamil’s overture and sent Thomas of Acerra and Bishop Berard of Palermo to Cairo. Frederick even went further and ordered Thomas and Berard to proceed to al-Mu‘azzam and try to secure an agreement with the strong man of Damascus, under whose jurisdiction Palestine fell. Al-Mu‘azzam’s infamous words to Frederick’s envoys – ‘Tell your master I only have for him the sword’28 – are nothing more than posturing. For, according to Sibt b. al-Jawzi and al-Nuwayri, al-Kamil complained that he did not have much room to negotiate the terms of the deal over Jerusalem with Frederick in 1229, because al-Mu‘azzam had already concluded a treaty with the emperor, giving him ‘the area between the Jordan River and the sea, including the villages from the gate of Jerusalem to Jaffa as well as other places’.29 So, all in all, before he even set sail, Frederick had negotiated the broader outline of a peace treaty with the Muslims. Al-Kamil and Frederick stood to lose a lot of their religious and political capital. They were vehemently attacked by groups within their respective communities for not pursuing war. In a letter, Patriarch Gerald of Jerusalem condemned Frederick precisely because his enterprise did not feature war and his treaty with al-Kamil was an affront to Christianity.30 Several contemporary Muslim chroniclers echoed similar resentments towards al-Kamil for ‘losing’ Jerusalem, and how that led to the evacuation of the Muslim population from the city and its surrounding area.31 It did not matter to them that,

154  |  syri a i n crusad e r time s in reality, the treaty secured the Muslims’ presence and control over parts of Jerusalem and most of its countryside.32 To return to the central issue of this chapter, can one consider al-Kamil’s deal with Frederick a case of opportunism and realpolitik and therefore unIslamic? If we look at al-Kamil’s career, we realise that opportunism was not his political philosophy. A great example that speaks to this point unfolded during the Fifth Crusade (1218–21). When al-Kamil was on the verge of annihilating the Crusader army near al-Mansura in 1221, he halted his attack and reached out with a peace proposal instead. He even sent the Crusaders food and supplies that saved them from assured starvation.33 Al-Kamil decided to do so against the advice of some of his generals who urged him to destroy the Crusader army. Here again, historians have explained the sultan’s overture as a case of opportunism and realpolitik,34 thus eliminating any possibility that he could have done so out of religious or moral considerations. I am not completely dismissing political opportunism and realpolitik as factors in as much as pointing to other factors shaped and informed by religion and a philosophy of rule. Indeed, throughout the Fifth Crusade, al-Kamil proposed one truth after another (giving Jerusalem and its territory back, paying for the rebuilding of its fortification and returning the relic of the True Cross), only to see them rejected by the dominant faction among the Crusaders. With respect to the relic of the True Cross, it is thought to refer to the one taken by Saladin during the Battle of Hattin in 1187.35 However, this particular relic of the True Cross came from the Fatimid treasury in Cairo, and al-Kamil ended up sending it to the Crusaders in Damietta with an unidentified monk,36 which is to be seen as an act of tremendous religious deference. One might explain that al-Kamil’s earlier peace overtures were made under duress when his forces were in retreat and when his position as sultan was in danger. This, however, does not explain him pausing the attack against the Crusaders when fortunes turned in his favour and could have wiped them out. It is precisely this last peace overture that shows realpolitik and opportunism were not always the factors that determined cooperation and tolerance. Relinquishing part of Jerusalem was not a light matter to al-Kamil, and he must have struggled with its religious implication. The ruler of

t h e c ase f or tolera nce a nd coex is te nce  | 155 Irbil urged him to inform the caliph in Baghdad. Al-Kamil relented and  dispatched  Amir  Fakhr al-Din, but only after he expressed his resentment toward the ­insinuation in such advice: ‘We are the slaves of that sacred house. Our fathers’ and our own services to it are very well known. We are neither deceitful nor hypocrites in that.’37 That al-Kamil was advised to inform the caliph shows the religious significance of the deal, for the caliph at the time did not muster any real power on the scene of Syria and Egypt. Thus, his angry outburst was about religious pride that his ­ family’s legacy derived from their liberation and protection of Jerusalem, given its religious symbolism to Islam, and that his treaty with Frederick did not ­compromise that legacy or his religious responsibility towards the city. Moreover, the negotiations between al-Kamil and Frederick were not done in secrecy. They involved a large apparatus of religious scholars and military advisors. There was al-Kamil’s most trusted advisor Amir Fakhr al-Din, who was knighted by Frederick and carried the emperor’s emblem on his flag until his death in the battle against the Seventh Crusade in 1250. There were also two other key figures in al-Kamil’s court who were involved, both religious scholars: al-Salah al-Irbili (d. 1234), who was al-Kamil’s chamberlain, and Shams al-Din al-Urmawi (d. 1252), who was chief–religious jurist of al-Kamil’s army and professor of Shafi‘i law at the Salihiyya School in Cairo. Some Muslim historians even speculated that al-Kamil’s friendship with Frederick and his earlier promises to the emperor forced his hand during the negotiation.38 Indeed, the peace between al-Kamil and Frederick laid down a lasting friendship between the two dynasties, and it was renewed between Frederick and al-Kamil’s son al-Salih (r. 1240–9), and then between Frederick’s son Manfred (r. 1254–66) and the Mamluk sultan Baybars (r. 1260–77).39 Thus, the deal to share Jerusalem was done despite the problems both monarchs were facing to their legitimacy and popularity. Each was aware of the right of his community to religiously significant sites in the sacred city, while respecting the right of the other community to their sacred sites. What al-Kamil considered important for the Muslims he did not relinquish, and what he considered significant to the Christians, he turned over. Similarly,

156  |  syri a i n crusad e r time s Frederick’s visit to the Haram displays his deference to the Muslims and Islam,40 and his words to Amir Fakhr al-Din underline what he could and could not do: If I did not fear for my reputation among the Franks, I would not have bothered the sultan with anything. I do not have any need for Jerusalem or other than it. I only sought to preserve my rule in their midst.41

In these words, Frederick is essentially emphasising the religious significance of Jerusalem to the Christians, an issue he saw bigger than himself and could not therefore relinquish. He repeated similar sentiments in his letters to King Henry III of England, in which he attributed the achievement to Jesus’s intervention.42 Back to the Outlook: Concluding Remarks When modern historians present the Crusader period as governed by religiously motivated war and violence, they inadvertently endorse and promote particular agendas and biases, thus undermining and obliterating the complex historical reality of that time. The corollary is that the entire period is conceived and assessed as belonging to a barbaric age. Accordingly, the evidence for tolerance and exchange is not recognised, even though it is staring us in the eye every time we read the sources, because the conventions that shape the field of crusader studies tell us this evidence is not a competing pattern to violence but rather a list of non-representative cases that have no religious legitimacy. It is not that modern scholarship has not discussed cases of exchange, alliances and tolerance between Muslims and Crusaders. Some scholars have written extensively on these issues.43 My argument is that such studies did not realise that the cases of tolerance, alliances, cooperation and exchange form a pattern, and more importantly, aside from a passing reference, their conclusions have not altered the dominant scholarly outlook of the Crusades. For instance, in a recent study on Muslims’ reaction to the Crusades, we are told that ‘the Muslims who helped the Franks almost exclusively came from groups outside the elites of society’.44 Those who did so ‘were willing to sacrifice their religious ideals for their own personal safety, which suggests that their faith was not as strong as it was for some’.45 Even in a study dedicated

t h e c ase f or tolera nce a nd coex is te nce  | 157 to tolerance and peace, the following words of Friedman sum up what I am describing as the problem: Even if the prevailing perception of the 1096 to 1291 era is one of perennial holy war, this epoch of holy war was in fact punctuated by interludes of peace. These treatises were accepted by both sides as legally binding and, although in principle a temporary remedy for inability to continue to fight a holy war, in fact both the Muslim and the Christian political entities could legitimately enjoy periods of peace.46

Why are cases of cooperation and tolerance between Muslims and Crusaders seen as lacking religious legitimacy and those involved in them as having questionable religious beliefs? Why could tolerance and peace only be opportunistic pauses for warriors otherwise keen on pursuing their holy wars? Why cannot we look at the era as one dotted with occasional wars instead of one dotted with occasional peace? These questions touch the core problem that even studies about tolerance cannot escape the grip of the dominant scholarly outlook of the Crusader period as one of war and violence and of true religion as only informing fanaticism and the clash between Islamdom and Christendom. As seen in the cases of alliances, they were never instances of a monarch acting individually. Each court had an apparatus of scholars, religious jurists and military advisors. At times they lobbied for war, and at other times they were split between groups who promoted tolerance, cooperation and exchange, and other groups in the same court who advocated for war and violence. In the particular case of Islam, which lacks a church institution to impose uniformity and be the sole voice (if one were to argue that at the time Rome was such a case), what we commonly call the teachings of Islam are actually the views of individual religious scholars. So, when al-Sulami and Ibn ‘Asakir wrote and preached on jihad they were expressing their individual voices, not the voice of all of Islam. Many, no doubt, would have agreed with them, but many others did not. Similarly, when al-Urmawi and alIrbili helped negotiate and draft the treaty between al-Kamil and Frederick, they were also expressing their voices as religious scholars. By so doing, the agreement became Islamically sanctioned, even though many other Muslims rejected it. When Saladin presented to the religious scholars in his court the

158  |  syri a i n crusad e r time s proposal to marry his brother al-‘Adil to Richard’s sister Joan, their vetting and approval of the proposal made it an Islamically sanctioned marriage proposal. So, we do not need to go searching for theological treatises that champion peace between Muslims and Crusaders. Those treatises are there in front of us, but because of the dominant scholarly discourse, that evidence is not acknowledged. The cases of alliances, tolerance and cooperation formed a pattern and were not isolated cases. The Crusader period was in some sense a clash between European Christians and Oriental Muslims, but in another sense it was an opportunity for alliances between them against others (be they other European Christians or Muslim foes). In yet another sense, it was also a period that continued and ushered additional forms of cooperation, exchange of learning and science, trade, sharing religious and cultural heritage and sites, and other common interests. Those who pursued cooperation and alliances represented the entire strata of society and were not always opportunists. It is judicious to listen to the evidence that speaks of religious factors, friendships and philosophy of rule that could have informed and shaped this reality of coexistence. Notes  1. Housley, 2006.   2. Sivan, 1967; Sivan, 1968, 115–120.   3. Elad, 1995; Mourad, 2008; Marsham, 2013; Avni, 2014.   4. Sivan, 1966; Sivan, 1968, 23–37.   5. For instance, Christie, 2007b; Chevedden, 2011.   6. Ibn ‘Asakir, 1996, 4. The very few other biographies of al-Sulami in later biographical dictionaries were copied from Ibn ‘Asakir.  7. Al-Sulami’s Book of Jihad was read in 1113 at the Umayyad Mosque of Damascus to three boys and the reader’s assistant, who could have been Ibn ‘Asakir’s older brother. Yet, there is no indication that any one of them copied it or had a role in transmitting it. In this respect, the view that Ibn ‘Asakir could have been aware of it – as argued in Mourad and Lindsay, 2013, 43 – stands corrected.   8. Ibn al-Qalanisi, 1908, 298.   9. Mouton, 1983; Mourad and Lindsay, 2013, 36–37. 10. Actually, another person is mentioned by the contemporary Ibn al-Qalanisi who

t h e c ase f or tolera nce a nd coex is te nce  | 159 also went to fight the Franks and was killed. He was a recent immigrant from Iraq, a mystic named ‘Abd al-Rahman al-Halhuli. 11. Ibn Wasil, 1977, 332–333. 12. Al-Maqrizi, 1956, 314–315. 13. Sibt b. al-Jawzi, 1951, 654; Ibn Wasil 1977, 332–333. 14. Ibn Jubayr, 1988, 273–282; Ibn Jubayr, 2008, 315–325. 15. Ibn Jubayr, 1988; Ibn Jubayr, 2008, 318–319. 16. Kedar and Pringle, 2009, 135. 17. Abu Shama, 1997, 63. 18. Usama b. Munqidh, 2008, 147. 19. Al-Harawi 2004, 70–76. 20. Ibid., 70–72. 21. Hamilton, 2000b, 210. 22. ‘Imad al-Din al-Isfahani, 1965, 146. 23. Al-Maqrizi, 1956, 110. 24. ‘Imad al-Din al-Isfahani, 1965, 555–557; Ibn Shaddad, 1964, 293; and Ibn Shaddad, tr. Richards, 2001, 187–188. 25. Van Cleve, 1972, 213–220; Madden, 2005, 146–153; Cobb, 2014, 210–212. 26. Ibn Wasil, 1977, 233. According to Sibt b. al-Jawzi, al-Mu‘azzam was only reacting to the possible alliance of al-Kamil and Frederick; Sibt b. al-Jawzi, 1951, 647–648. 27. Sibt b. al-Jawzi 1951, 657; Kantorowicz, 1931, 184. 28. Sibt b. al-Jawzi,1951, 643; Mourad and Lindsay, 2013, 96. 29. Sibt b. al-Jawzi, 1951, 654; al-Nuwayri, 2002, 151–152. 30. Munroe, 1896, 27–31. 31. Sibt b. al-Jawzi, 1951, 654; Ibn al-Athir, tr. Richards, 2010, 481; Ibn al-Athir, 2008, 294; Ibn Wasil, 1977, 243; al-Maqrizi, 1956, 23. 32. Ibn Wasil, 1977, 241–242; Abu’l-Fida, 1907, 141–142. 33. Al-Dhahabi, 1996, 29–30. 34. Powell, 1986, 175–194; Madden, 2005, 145. 35. Oliver of Paderborn, 1948, 45; Riley-Smith, 2005, 148. 36. Al-Dhahabi, 1997, 30. 37. Al-Hamawi, 1981, 183. 38. Ibn Wasil, 1977, 234; al-Hamawi, 1981, 176–177; al-Maqrizi, 1956, 229–230. 39. Ibn Wasil,1977, 234–235; Abu’l-Fida, 1907, 141. 40. Sibt b. al-Jawzi, 1951, 655–657; Ibn Wasil,1977, 244–255.

160  |  syri a i n crusad e r time s 41. Ibn Wasil, 1977, 243. 42. Roger of Wendover, 1899, 522–524. 43. For example, Draelants et al., 2000; Köhler, 2013; Kedar, 2008; Burnett, 2009; Friedman, 2011. 44. Mallett, 2014b, 138. 45. Ibid., 139. 46. Friedman, 2016, 98.

9 The Portrayal of Violence in Walter the Chancellor’s Bella Antiochena Thomas Asbridge

O

n 28 June 1119, Roger of Salerno – the southern Italian Norman ruler of the Latin principality of Antioch – was defeated in the Battle of the Field of Blood (Ager Sanguinis), in northern Syria, by a Muslim army from Aleppo led by the Turcoman warlord Najm al-Din Il-Ghazi b. Artuq of Mardin. The Field of Blood was a catastrophe for the Franks who had settled in Syria in the wake of the First Crusade. Roger of Salerno himself was slain, the Antiochene relic of the True Cross was lost and the bulk of the Latin principality’s army – estimated at 700 knights and 3,000 infantrymen – were either killed or taken into captivity.1 One of the most important contemporary accounts of these events was preserved in Walter the Chancellor’s Bella Antiochena (The Antiochene Wars).2 Walter appears to have been a highly placed Antiochene official who served as the Latin principality’s chancellor between c.1114 and c.1122.3 He authored a Latin narrative of the Battle of the Field of Blood and its aftermath, as well as describing an earlier campaign fought in 1115, and his chronicle has long been regarded as a valuable source for the early history of the northern Crusader states – offering a detailed record of events that certainly could be characterised as well informed and that often seems to draw from Walter’s own eyewitness experiences and memories.4 The Nature of Walter the Chancellor’s Bella Antiochena Bella Antiochena appears to have been composed in at least two distinct phases. Book One, covering the events of 1115 and Roger of Salerno’s victory at the first battle of Tell Danith against Bursuq of Hamadan on 14 September 1115,

164  |  syri a i n crusad e r time s was probably completed before 1119. Book Two recounted the Frankish disaster at the Field of Blood, the subsequent arrival of King Baldwin II of Jerusalem in northern Syria and the partial recovery of Frankish fortunes heralded by Baldwin’s victory in the second battle of Tell Danith on 14 August 1119. Walter the Chancellor was almost certainly present at the Field of Blood and appears to have been among those taken prisoner after the battle. According to his own account, Walter then suffered through a period of imprisonment (and various forms of ill-treatment), including a phase of incarceration in Aleppo itself, until he seemingly was ransomed, though the precise date and circumstances of his release remain unclear. It is assumed that, once freed from captivity, Walter began writing Book Two of his account, and this would date the bulk of this section of Bella Antiochena to the early 1120s.5 Heinrich Hagenmeyer identified seven extant manuscripts of Walter the Chancellor’s text and, in the critical edition of Bella Antiochena published in 1896, concluded that the oldest and best of these (A) was the exemplar for the other six.6 This codex, now held as BN. 14378 in the Bibliothèque Nationale in Paris, was prepared by William of Grassegals as a gift for King Louis VII of France somewhere between 1137 and 1146 and comprises the work of two notable chroniclers of the First Crusade – Fulcher of Chartres and Raymond of Aguilers – alongside that of Walter the Chancellor. In studying the Grassegals codex, Jay Rubenstein demonstrated that Walter’s Bella Antiochena was copied and read alongside narratives of the First Crusade, but there is nothing to suggest that Walter himself was directly influenced by either Fulcher’s or Raymond’s narratives, and arguably it has been proven that Walter did not draw upon the section of Fulcher’s text covering 1119 when composing his own account of the Field of Blood.7 In fact, beyond some evident scriptural and classical allusions, little is known of Walter’s own sources of knowledge and literary influences, particularly with regard to the course of the First Crusade and the initial foundation of the Latin settlements in the East.8 Given its subject matter, it is no surprise that Bella Antiochena has much to reveal about the conduct of warfare in the nascent Crusader states. I have argued elsewhere that, when recording the early military history of the principality of Antioch, Walter the Chancellor conveyed the sense that the first Frankish settlers in northern Syria were engaged in an ongoing holy

po rt ray al of vi olence i n bella antiochena  | 165 war – or perhaps even an extension of the First Crusade – particularly with regard to the Latins’ apparent motives, the spiritual preparations for battle they undertook and the reward of Christian martyrdom supposedly enjoyed by those who fell in combat.9 Bella Antiochena should also figure in any study of the emergence of the knightly class and the formation of the shared cultural sensibility known as chivalry, given that Walter’s account contains relatively early references to knights as a clearly distinct martial grouping and also emphasises that these warriors were driven by a desire to earn honour and thus burnish their reputations.10 A Question of Violence At present, however, I wish to consider what the close study of Walter the Chancellor’s Bella Antiochena can contribute to another significant field of inquiry. Though the discipline of Crusader studies undoubtedly has proven fertile in terms of scholarship published in recent decades, relatively little focused research has been conducted on the precise nature and extent of the violence enacted during the Crusades.11 I would contend that the question of whether the Crusades were more violent and savage than other wars is of the first order of importance. A tacit assumption endures in both scholarly discourse and popular imagination that these medieval conflagrations were waged in a manner that was more extreme, barbaric and destructive than other forms of ‘normative’ medieval warfare. Academic specialists, such as Jay Rubenstein, have continued to present a blood-soaked image of the Crusades, emphasising their supposedly apocalyptic nature.12 Meanwhile, scholars seeking to offer a synoptic overview insist on particularising these wars. Drawing upon a limited range of speculative research, Steven Pinker characterised the Crusades as a form of genocide in his broad 2011 study of the history of violence and humanity, concluding that these medieval conflicts were ‘equivalent to the Nazis’ genocide of the Jews’.13 In 2014 Karen Armstrong sought to judge the First Crusaders by modern standards in her examination of the relationship between religion and violence, and this questionable approach (alongside her uncritical reading of the primary evidence) prompted her to brand the Crusaders as ‘psychotic’ savages whose deeds were inherently immoral and un-Christian.14 Work of this type has the capacity to influence the general perception of the Crusades

166  |  syri a i n crusad e r time s as a historical phenomenon. It is worth noting that, in February 2015, US President Barack Obama was even moved, when pondering current affairs in the Middle East, to observe that the medieval Crusaders ‘committed terrible deeds in the name of Christ’.15 Elsewhere, entrenched views about Crusading warfare have had significant bearing upon our understanding of the history of Latin Christian settlement in the Levant after the First Crusade. Brian Catlos recently argued that the unusually violent and ‘interruptive’ nature of the Crusader conquests conditioned the nature of Latin–Muslim contacts in the Near East for the next two centuries.16 Assumptions about violence have also shaped analysis of the Crusades’ supposed position in the wider meta-history of relations between the broadly defined Western and Muslim worlds, playing in to Samuel Huntingdon’s influential thesis regarding the inevitability of a ‘clash of civilisations’ between peoples of different faiths.17 Crusade violence is undoubtedly a sensitive topic, and any attempts to downplay the brutality and savagery of the Crusades, or to recalibrate our understanding of these holy wars, might be cast as efforts to diminish the inherent horror of such events or even to legitimate the Crusaders’ actions. Of course, there can be no doubt that the Crusades waged in the Holy Land in the course of the Middle Ages were violent and, from a twenty-first-century perspective, barbaric. Nonetheless, as historians, our task is to understand and then assess the Crusades in their proper medieval context. On this basis, our primary objective must be to determine – to the fullest extent that the evidence will allow – whether the violence witnessed during the Crusades was extreme by the standards of the day. When seeking to study Crusade violence, historians turn first and foremost to the surviving contemporary written sources. A conspicuous feature of this material is that works composed from the perspective of the Latin West, either in medieval Latin or vernacular European languages such as Old French, tended to exaggerate rather than downplay the violence enacted by Crusaders in the Holy Land. Benjamin Kedar traced the process by which Crusade brutality during the sack of Jerusalem in mid-July 1099 was heightened and amplified by successive waves of Western historians. As Kedar noted, the desire by contemporary and near-contemporary chroniclers to stress the ‘exceptionality’ and ‘extraordinary nature’ of the First Crusaders’

po rt ray al of vi olence i n bella antiochena  | 167 conquest of the Holy City, alongside their willingness to draw upon literary models and Old Testament imagery, could impact significantly upon the version of events they recorded: 3,000 dead could become 10,000 (or more) and the quantity of Muslim blood spilled in Jerusalem’s Aqsa Mosque might rise from ankle-deep to reach the Crusaders’ knees and even the bridles of their horses.18 Robert the Monk’s Historia Iherosolimitana, which drew on earlier eyewitness accounts such as the Gesta Francorum and which was composed around 1108, became the most widely circulated Latin account of the First Crusade. Robert’s text exemplarises the tendency of Western narratives to revel in opportunities to present the Franks shedding blood and butchering their Muslim enemies. Both in recording the Crusader sack of Jerusalem and elsewhere in his account, Robert deployed a range of techniques to intensify his portrayal of Crusader violence. Perhaps the most important of these were: first, his extensive recourse to what might be termed ‘blood imagery’ – including an intense focus on the quantity of blood spilled and the sensory qualities of blood, such as its colour, smell, warmth and viscosity; and second, Robert’s repeated attempts to emphasise the human response to violence, particularly the capacity for Crusader brutality to inspire sensations of fear, dread and even horror in its Muslim victims.19 It is the case that the Crusaders’ Muslim opponents were presented as both violent and barbaric in texts such as Robert the Monk’s Historia, but such depictions almost always were limited to atrocities supposedly committed by Muslims in the East prior to the Crusading expedition, and usually catalogued in the context of the expedition’s promulgation by Pope Urban II at Clermont in November 1095 and through the course of the Crusade’s subsequent preaching.20 In contemporary Latin narratives, a heightened sense of violence and atrocity was rarely associated with Muslim protagonists during the First Crusade itself – rather it was the victorious and triumphant Crusaders who were presented as the shedders of blood and the sources of terror. Academic debate continues regarding the potential motives, conscious or otherwise, behind this heightening in the representation of Crusader violence. Recently, historians have argued variously that Robert the Monk and his ­contemporaries may have been: attempting to emphasise, and thus

168  |  syri a i n crusad e r time s celebrate, the First Crusaders’ fearsome martial qualities; seeking to present the Crusade as a just act of retribution and vengeance; presenting blood-letting as a purgative or purificational process, particularly in the context of Jerusalem itself; or reflecting contemporary apocalyptic sensibilities, while also drawing upon scriptural allusion and precedent to validate the Crusaders’ endeavours. While all of the above deserve careful consideration, it also seems evident that Robert the Monk’s approach to the portrayal of Crusade violence was part of, or adhered to, a broader narrative strategy. Believing himself to be engaged in recording one of the most momentous events in Christian history, Robert set out to captivate his audience – to move beyond the strictures of the relatively sterile chronicle form by crafting a narratio, or story, replete with the colour, imagery and suggestion of sensation that might allow a reader to glimpse the experience of the Crusade.21 Walter the Chancellor’s Depiction of Violence How then might Walter the Chancellor’s portrayal of violence be characterised, particularly in comparison to other near-contemporary chroniclers such as Robert the Monk? In Book One of Bella Antiochena, covering the campaigns of 1115 and the Frankish victory in the first Battle of Tell Danith, Walter adopted a relatively orthodox approach to the representation of warfare and violence. That is to say, he related the details of a brutal battle, glorying in some elements of the Latin Christian triumph, but he showed little or no interest in using specific narrative strategies to heighten his depiction of bloodshed. When describing Roger of Salerno’s defeat of the Muslim forces led by Bursuq of Hamadan on 14 September 1115, Walter asserted that the Antiochenes sought the ‘annihilation’ (interitus) of the enemy.22 Once battle was joined, the Franks were said to have ‘put the enemy to flight, hacked them to pieces and killed them’.23 As a result, ‘innumerable’ Muslims apparently were slain, such that the battlefield was left ‘partly covered by dead bodies’.24 Latin losses seem to have been limited, though Walter did record that the ‘distinguished knights’ Robert Sourdeval and Bochard were ‘slaughtered and became martyrs of Christ’.25 In this section of his account, Walter paused only once to experiment with blood imagery, noting that, in the thick of the fray: ‘from our side and theirs, in turn, there was immense bloodshed’.26

po rt ray al of vi olence i n bella antiochena  | 169 It has long been recognised that the nature and character of Walter the Chancellor’s account shifted dramatically in Book Two of Bella Antiochena.27 In chronicling the events of 1119 – including the fateful Battle of the Field of Blood and its shocking aftermath – Walter was no longer in a position to celebrate a Latin victory. Instead, as he noted in the prologue to this section of his work, Walter apparently felt ‘compelled’ to record the ‘grief of griefs’ and to seek an explanation for the ‘sudden disaster’ that befell the Principality of Antioch. In a remarkable passage that seems to suggest a degree of selfconscious authorial intent that is unusual in Latin narratives of this period, Walter actually acknowledged that he would have to adjust his approach in Book Two, writing that, ‘having changed the style I adopted for the first war’ he would now proceed to record ‘the second detestable war’.28 Even so, Walter’s description of the Battle of the Field of Blood itself was not particularly distinctive and made relatively sparing use of close narrative depiction or heightened imagery. Seizing upon a rare opportunity to highlight Frankish success in the lead-up to the battle, Walter noted that a group of knights, including the Norman warrior Robert of Vieux-Pont, prevailed in an initial skirmish near al-Atharib on 27 June. These Franks were shown ‘dealing violent blows in a knightly fashion’ such that ‘with blood pouring forth, they drove [the enemy] down to hell with fatal wounds’.29 Perhaps unsurprisingly, Walter’s account of the crushing Frankish defeat at the Field of Blood on 28 June was quite brief and might be characterised as little more than an attenuated sketch of the ‘action’ that played out. Only the death of Roger of Salerno – Walter’s central protagonist – received focused attention. In the midst of the conflagration, Roger was shown ‘obeying the command of God … by fighting energetically’ beside the Antiochene relic of the True Cross, but ‘he was struck by a warrior’s sword through the middle of his nose and right into his brain’ and thus ‘gave up his body to the earth and his soul to heaven’. Walter went on to note that some Muslim troops then ‘came to swift blows amongst themselves’ because they were ‘possessed by greed for the gold and precious stones’ covering the reliquary of the Antiochene True Cross. As a result, they were ‘made the fodder of death [and] passed away to the … fires of hell’. On the fate of the bulk of the Latin army, Walter concluded that some ‘were wounded, some were killed, and some who were mutilated with different injuries were brought out

170  |  syri a i n crusad e r time s for destruction on that same battlefield … while some were put in the bonds of most wretched captivity’.30 In fact, it was only once Walter turned to recording the aftermath of the Field of Blood – when he appears to have been among those taken prisoner by Il-Ghazi’s forces – that his focus upon, and representation of, violence changed significantly. In two sections of his Bella Antiochena, covering Chapters 6 to 7 and 13 to 15 of Book Two, he presented a detailed and sustained account of what became of these Frankish captives in the summer of 1119, himself included – cataloguing a grim trail of violent abuse, torture and execution.31 In the immediate wake of the battle, some 500 Franks were captured and placed in fetters. Thereafter, around 200 wounded prisoners were examined in the ‘death tent’ (letali tenta) and those who were already severely injured were dragged out by their hair and summarily executed.32 Walter went on to describe how the total number of captives dwindled in the days that followed such that, eventually, only a small group who were deemed worthy of ransom were spared death. These survivors were taken to Aleppo, where they continued to be subjected to physical and psychological abuse, and some were killed.33 Around this same time, in the late summer of 1119, the Frankish nobleman Robert fitz-Fulk the Leper was taken captive and subsequently beheaded in Damascus.34 Il-Ghazi then ordered the summary execution of a further thirty-seven prisoners in Aleppo, but later agreed to ransom an unspecified number of surviving Antiochene captives.35 There can be little doubt that the bulk of these passages were designed to be read and understood as emotive representations of Walter’s own lived experience. As such, they have been valued as a rare example of eyewitness testimony regarding the treatment of Franks held captive by Muslims in the Holy Land during the Crusader period. Yvonne Friedman asserted that ‘[Walter] is unique in telling the story from the captive’s side’, and also observed that ‘being an eyewitness and one of the terrorised victims himself, the need to recount the horrors of fright and uncertainty is understandable’.36 In her work, Friedman was primarily concerned to mine Bella Antiochena for details of the practical mechanics surrounding the treatment of Latin prisoners. The concern in this present study is to examine and assess the methods used by an early-twelfth-century chronicler to record and depict the

po rt ray al of vi olence i n bella antiochena  | 171 e­ xperience of captivity, with particular attention paid to issues of representation, imagery and narrativity. Walter the Chancellor as a Captive and a Witness A number of important points of analysis emerge from a close reading of the sections of Walter’s Bella Antiochena that record the treatment of captives. First, there is a significant change of narrative tone and textual construction in these passages. Chapters 6 to 7 and 13 to 15 of Book Two appear to have been much more actively and intensively narrativised and, as such, were shaped to convey a clearer picture of both the story-world in which events played out and the experiences endured by the narrative’s actors. Narrativisation One marker of this process of narrativisation was Walter’s close focus on human sensation. In a long passage describing how the surviving captives were led bound and naked into a vineyard near Sarmada on 29 June 1119, he sought to convey the experience of extreme thirst. So intense was the summer heat, Walter claimed, that ‘not a single one of the prisoners would have refused to offer either a limb or even life itself for a cup of cold water’. The Latin captives were driven into a frenzied state, such that they became obsessed with the idea of slaking their terrible thirst by any means possible. This included gathering pitiful quantities of juice from ‘dirt-covered grapes’ that had fallen into the soil during a recent harvest, ‘bit[ing] their own tongues with their teeth’ and drinking urine. Walter also seemed at pains to stress the sheer abasement of this episode. Not only were the grapes covered in soil, they had also been ‘trampled by feet’ and were ‘sticking to the earth’, and the tumult all but transformed the captives into animals willing to fight one another for ‘a little drop of juice’.37 Walter’s sudden recourse to the extensive use of reported speech is a second indicative feature of the move to a more purposeful narrativisation of his account in these passages. Muslim protagonists, both named and unnamed, were made to speak and sometimes deliver extensive orations. There is no evidence to indicate that Walter the Chancellor could speak either Arabic or Turkish, nor is there any suggestion that Walter actually heard and understood these words through the mouth of an interpreter. Instead, one

172  |  syri a i n crusad e r time s of the primary functions of this reported (and presumably imagined) speech seems to be to accentuate the tension and action within the narrative. For example, after a lengthy debate about the fate of the captives, later on 29 June Il-Ghazi ordered the immediate execution of a large group of Franks, supposedly shouting: ‘Come on soldiers … Here is a sacrifice of dogs ready for your swords … Run at them readily and quickly!’38 As Damien Kempf has noted, the study of narrativisation can contribute to our understanding of a medieval source’s ‘textual archaeology’.39 In the case of Bella Antiochena, this approach can help to inform and refine thinking about the work’s date of composition. In 2000 I argued that Chapters 13 to 15 of Book Two ‘describing the fate of Latin captives … sit rather uncomfortably with the rest of the text’ because they were less concerned with relating the details of military and political events, and I suggested, therefore, that they may have been added at a later date.40 It now seems clear that, given the continuity in narrative voice, these passages are actually closely linked to Chapters 6 to 7. This means that Walter either composed Book Two from his prologue to Chapter 15 as a whole, or added the material covering the fate of the captives at a later date, interpolating Chapters 6 and 7 into his original draft of the text. Decapitation Another important feature of this material is the contribution it makes to our understanding of the incidence and function of decapitation in the context of the Crusades and Crusader states. The targeted abuse and mutilation of enemies by both Latin Christians and Near Eastern Muslims was a distinctive, and arguably extreme, feature of Crusading warfare in the late eleventh and early twelfth centuries. One notable feature of this practice involved beheading. A range of contemporary and near-contemporary Western European accounts testify to the fact that, almost as soon as the First Crusaders crossed over into Asia Minor, they adopted the gruesome practice of decapitating Muslim corpses and then using these severed heads to intimidate their opponents. This might be achieved by catapulting these heads into a fortress the Crusaders were investing, as at the siege of Nicaea in the summer of 1097, or by displaying severed heads as visible trophies of war affixed to lances or spears, as was the case when the Franks returned

po rt ray al of vi olence i n bella antiochena  | 173 victorious to Antioch after defeating Ridwan of Aleppo in the so-called ‘Lake Battle’ in February 1098.41 In his Bella Antiochena, Walter the Chancellor made a number of references to the process of decapitation, but in the context of the Antiochene Wars it was the Latin Christians who were beheaded and whose corpses sometimes were then abused post-mortem. In the immediate aftermath of the Battle of the Field of Blood, for example, a large number of wounded captives were ‘led to execution [and] they fell at the hands of the heathen not only with their heads cut off but they even suffered agonising death with the skin flayed from the living and half-severed head’.42 Walter made no reference to Roger of Salerno being decapitated post-mortem, even though the Muslim chronicler Ibn al-Athir later claimed that Roger’s corpse was beheaded, writing that the ‘Lord of Antioch was killed and his head taken’. It is not clear whether Walter deliberately avoided any description of this mutilation. As already noted, he did provide a precise account of Roger’s death, so appears to be indicating that he possessed a detailed knowledge of these events, but it is possible, of course, that he simply did not know what had happened to Roger’s body. Questions remain regarding the reliability of Ibn al-Athir’s account, given that he was writing in the early thirteenth century and that he also appears to be contradicted by the Anonymous Syriac Chronicle of 1234, which stated that ‘Roger, who was killed in this battle … was never found, neither among the dead nor among the prisoners’.43 Perhaps the most significant example of decapitation in Bella Antiochena related to the fate of the Frankish knight and Antiochene landholder, Robert fitz-Fulk the Leper. Robert held a significant lordship within the Principality of Antioch, which included the frontier settlement of Zardana, just ten miles from the site of the Field of Blood.44 He does not appear to have fought in the battle on 28 June (or if he did, he managed to escape the disaster), but Robert was subsequently taken captive by Muslim troops near Zardana later that summer and was sent as a prisoner to Il-Ghazi in Aleppo. Walter devoted all of Chapter 14 of Book Two to a close narration of Robert’s treatment, and some of the details of his account are corroborated by independent strands of near-contemporary testimony in Arabic sources, including Usama b. Munqidh’s Kitab al-i‘tibar (‘Book of contemplation’) and Ibn al-‘Adim’s chronicle of Aleppan history.

174  |  syri a i n crusad e r time s After an initial interview with Il-Ghazi in Aleppo, Robert was sent to one of Il-Ghazi’s Muslim allies, Tughtegin, the Turkish atabeg of Damascus, seemingly to finalise satisfactory terms of ransom. Robert had evidently had earlier dealings with Tughtegin and a degree of friendship may even have developed between the pair. Nonetheless, Tughtegin summarily executed Robert, beheading him with his own hand. In Walter’s account, Robert was only killed after a protracted series of discussions designed to mock and publicly humiliate him. Tughtegin – whom Walter branded as ‘the investigator and discoverer of various tortures’ – finally decapitated Robert ‘with a stroke of his sword’ and then handed the head to his ‘judiciary herald’ (juidiciario praeconi) so that it could be put on public show. The trophy was then ‘paraded through the squares and the districts’ of Damascus in what was essentially a gruesome fund-raising exercise, and this supposedly led to Tughtegin receiving 500 bezants in donations.45 According to Walter, Tughtegin subsequently had Robert’s skull made into a ceremonial drinking cup ‘fashioned from purest gold and precious jewels … which he exhibited as a symbol of his courage and victory for him and his descendants to drink from during festive celebrations’.46 Walter appears to have intended this whole episode to be read as a clear example of excessive and savage brutality – one of what he categorised as a range of ‘perverse torments’ dreamed up by Il-Ghazi and Tughtegin. He cannot have witnessed most of the events that led to Robert’s death in person, but he nonetheless peppered his account with narrative elements that signposted the transgressive nature of this violence, not least by imputing animalistic qualities to Tughtegin when noting how, at one point, the atabeg’s face was lit up by ‘a bestial grin’.47 Even so, it remains open to question whether such deeds – macabre as they may now seem – can accurately be categorised as being particularly extreme or unusual forms of violence. Carole Hillenbrand has examined another earlier example of similar behaviour by Tughtegin of Damascus, when he beheaded and scalped the Frankish lord of Tiberias, Gervase of Bazoches, in 1108. According to the Latin chronicler Albert of Aachen, Tughtegin hung Gervase’s scalp from his spear ‘as a token and memorial … to stir up grief among the Christians’, while other sources indicate that Tughtegin used his Frankish victim’s skull as a drinking-cup.48

po rt ray al of vi olence i n bella antiochena  | 175 Contemporaries may have suggested that such behaviour was, by its nature, designed to shock and horrify, and thus intended to have a n ­ egative impact on enemy morale, but acts of this type were not unique to the Crusades or the Near East. While recognising that decapitation could have varied symbolic meanings in different cultures, historians have catalogued the widespread practice of beheading during conflict and war across the ancient and medieval world. Maribel Fierro has shown that decapitation was frequently used as a method of execution and dismemberment by both Muslims and Latin Christians in medieval Iberia.49 It is worth noting that, elsewhere in Europe, post-mortem beheading was often, though not always, associated with the treatment of enemies deemed to be members of ‘alien’ cultures – so, for example, during struggles with the Anglo-Norman and Angevin rulers of England, Scottish, Welsh and Irish opponents were said to have been decapitated after death.50 The grim practice of fashioning a drinking-cup from an enemy’s skull also had a long and varied history. This custom can be traced back at least to the fifth century BCE and has been variously associated with Eurasian groups including the Scythians and the Xiongnu Huns, the ‘barbarian’ Gauls and the Amerindian Incas.51 While we might conclude, therefore, that neither purposeful post-mortem decapitation nor the ritualised abuse of human skulls were the sole preserve of the Crusades, these practices still appear to have been regarded as relatively extreme and unusual expressions of violence in the twelfth century and thus worthy of being recorded and commented upon. Blood and horror Perhaps the most remarkable feature of these sections of Bella Antiochena relates to the narrative strategies Walter employed when depicting the treatment of Latin captives. After the First Crusade, Bella Antiochena is, in fact, the first surviving Latin narrative to present a close and detailed portrayal of violence enacted by victorious Muslims against defeated Franks. The Battle of the Field of Blood was not the first reversal experienced by the Franks in the Levant. In 1102, for example, King Baldwin I of Jerusalem had been defeated in the second Battle of Ramla, while a combined force of Antiochene and Edessene troops were bested at the Battle of Harran in 1104. But the contemporary Latin authors who recorded these events and

176  |  syri a i n crusad e r time s others showed little interest in describing acts of Muslim violence in any depth. As has already been noted, even Walter the Chancellor recorded only a passing summary of Il-Ghazi’s triumph at the Field of Blood. But once he moved on to recount the treatment of the Latin prisoners after the battle, Walter adjusted the focus of his account and reshaped his narrative palette – arguably doing so to lend his extended description of the torment and abuse he and his fellow captives suffered a greater sense of immediacy and emotive force. In two sustained sections of Bella Antiochena, Muslims are shown – in lavish and chilling detail – carrying out appalling acts of cruelty and brutality. Walter’s authorial approach in these passages is striking. Whether consciously or unconsciously, he employed the same techniques to heighten his narrative of Muslim violence – namely blood imagery and the sensation of horror – that had been widely utilised by First Crusade chroniclers, including Robert the Monk, when portraying violence enacted by Latin Christians. Thus, in Bella Antiochena, the images of extreme violence are noticeably similar, but the identities of the perpetrators and victims have been reversed. Contemporary authors recounting the violent deeds of the First Crusaders often used human blood as an amplifying narrative effect. Emphasis might be placed on the visceral quality and sensory nature of blood, including its colour and scent, alongside references to the historical and literary topos of a great ‘effusion’, or overwhelming quantity, of blood. In recording the sack of Jerusalem on 15 July 1099, for example, Robert the Monk wrote of an ‘indescribable slaughter’ within the Aqsa Mosque: So much human blood was spilt there that the bodies of the slain were revolving on the floor on a current of blood; arms and hands which had been cut off floated on the blood … Even the soldiers carrying out the massacre could hardly bear the vapours rising from the warm blood.52

Elsewhere, Robert made frequent references to the fact that the Crusaders’ swords supposedly possessed an insatiable appetite for Muslim blood, at one point asserting that Frankish swords ‘could not be blunted and could not tire of blood.’53 In Bella Antiochena, Walter described how Muslims shed blood in this same manner. On 29 June 1119, the surviving Latin prisoners were divided

po rt ray al of vi olence i n bella antiochena  | 177 into two groups, those considered worthy for ransom and those who were to be butchered there and then at Il-Ghazi’s urging. According to Walter: In one attack the impious soldiers struck the condemned men with deathdealing swords, and they did not cease from killing until there was no part of the prisoners’ bodies left whole; and there they rolled in spilt blood as a pig wallows in mud.

Here the grisly image of corpses and dismembered body parts rolling ‘in spilt blood’ seems to echo Robert the Monk’s references to ‘revolving’ and ‘floating’ in blood. Walter added that, because the Muslim troops were ‘not yet sated with the slaughter of men’, they begged to be allowed to execute the remaining captives, and this also appears to resonate with Robert’s references to the First Crusaders’ unquenchable thirst for bloodshed.54 Later, in Chapter 15 of Book Two, Walter recorded another mass execution, this time enacted outside a ‘royal palace’ in Aleppo, presumably at some point in the summer of 1119. On this occasion, Il-Ghazi executed thirtyseven of the remaining prisoners in front of a large crowd of Aleppan citizens, and Walter stated that ‘their bodies were beheaded in a moment, and a wave of blood flowed forth and spattered all the pavement of the entrance hall to the royal palace’.55 This description highlighted the quantity of Latin Christian blood shed by stressing that it ‘flowed forth’ to cover the surrounding space. Interestingly, Walter only made limited use of this other narrative effect associated with blood – its distinctive colour and consequent ability to transform the appearance of its immediate environment. Latin First Crusade narratives were replete with references to the earth of battlefields or the waters of rivers being turned red by Muslim blood. The distinctive name associated with the fateful battle fought on 28 June 1119 – that of Ager Sanguinis or the Field of Blood – would seem to offer an obvious opportunity for Walter to employ blood imagery, but in fact he only made one passing reference to the name, observing that the site of the battle was known as the ‘Field of Blood … both in reality and in name’.56 Walter also heightened his portrayal of violence by stressing its capacity to inspire fear, dread and horror. This narrative effect was used by Robert the Monk when describing how 5,000 Muslims were supposedly butchered

178  |  syri a i n crusad e r time s by the First Crusaders on Antioch’s Bridge Gate in early 1098 – a spectacle that, according to Robert, ‘filled all those watching with great horror’. The theme of horror was further developed, alongside blood imagery, by William of Tyre, the famous later-twelfth-century Latin chronicler who lived in the Crusader Kingdom of Jerusalem. Writing of the Jerusalem massacre of 1099, William stated that: It was impossible to look upon the vast numbers of the slain without horror; everywhere lay fragments of human bodies, and the very ground was covered with the blood of the slain. It was not only the spectacle of headless bodies and mutilated limbs strewn in all directions that roused anguish in all who looked upon them. More dreadful still was it to gaze upon the victors themselves dripping with blood from head to foot, a sight which shocked all those who met them.57

In Bella Antiochena this same narrative effect was used to accentuate Muslim, rather than Latin, violence. Throughout Chapters 6 to 7 and 13 to 15 of Book Two, Walter made repeated references to the abject sense of torment and horror endured by the Frankish prisoners. He described the experience of their first night in captivity (28–9 June) in a particularly evocative passage, using the literal darkness of night as a counterpoint to the ‘dark’ nature of their suffering: Knowing they were to be tortured, [the survivors] spent that dark night in outrage and dread, desiring death with their minds, they [nonetheless] raged at death … After a night squashed together in fetters … the sun rose on the following day – sun, I say, which was blacker for them than the dark of night.58

In a similar manner, he recorded that sometime later in Aleppo a mixed group of Latin fighting-men and nobles (presumably including Walter himself ) were paraded, bound and naked, in front of Il-Ghazi. There they were repeatedly mocked, insulted and beaten, before the Muslim amir offered them ‘through an interpreter’s words’ a choice between ‘foreswear[ing] the Christians’ law’ or immediate beheading ‘by a blow of his own sword’. Walter highlighted the physical and psychological effects of this abusive treatment, noting that:

po rt ray al of vi olence i n bella antiochena  | 179 Embarrassed by their shameful nakedness, their teeth chattered, their bowels trembled and they gaped heavenwards; in place of heaven, on earth nothing was more certain for them than the threatening blows of the sword, and they were tormented in mind and body.59

Walter also suggested that their Muslim captors deliberately sought to increase or inflame this anxiety through intimidation and cruelty. In describing the events that led to Robert fitz-Fulk’s execution, Walter suggested that Il-Ghazi and Tughtegin repeatedly postponed any decision of Robert’s fate so that he would ‘feel that he was being mocked by both of them in turn, and so that the dread and might of both of them could work upon him’.60 Walter the Chancellor’s Influences and Purpose It is not possible to determine the full scope and nature of Walter the Chancellor’s intentions when composing Bella Antiochena, but he does appear to have made a sustained and purposeful attempt to present the Muslim treatment of Latin prisoners after the Field of Blood as a grave atrocity, worthy of condemnation and censure. To do so, he employed many of the same narrative strategies and amplifying effects used by Latin chroniclers of the First Crusade, but reversed the identities of his actors such that the triumphant Muslims became the shedders of blood and sources of dread, and the captive Franks became their victims. It should be recognised that, as a result, Bella Antiochena became the first Latin narrative of the Crusading era to present a sustained image of Muslim protagonists as ‘victors’ perpetrating acts of extreme brutality. The evident similarities in imagery and textual construction might be thought to suggest an otherwise undetected connection between Bella Antiochena and works such as Robert the Monk’s Historia Iherosolimitana. It is perhaps equally likely that these narratives shared a common array of historical and literary influences, ranging from the Old and New Testament, to classical works that explored the depiction of violence, such as Virgil’s Aeneid, Lucan’s Pharsalia and Ovid’s Metamorphoses.61 Nonetheless, an intriguing possibility remains. Walter the Chancellor may have made a conscious decision to mimic and invert the portrayal of barbarity witnessed in the Latin First Crusade narratives, perhaps as an oblique indication of

180  |  syri a i n crusad e r time s his awareness that both Near Eastern Muslims and Latin Christians were capable of savagery. A remarkable passage at the end of the fifteenth chapter of Bella Antiochena’s second book would seem to indicate that Walter was considering this issue. Here, Walter conceded that there was a marked tendency among men – regardless of class or creed – to ‘inflict many different punishments on their prisoners [and] to extort money when they capture them justly and when they capture them unjustly’. Apparently fearing that his account might inspire imitation, Walter stated that he would offer no further descriptions of the atrocities committed by Muslims, adding that: I think it is better for me to keep quiet about the kind and quantity of their tortures than to express them, lest Christians bring the same to bear on Christians and turn them into accustomed practice.62

Given the nature of the narrative he had already composed, this statement seems to come rather late in the day, but it still stands as a remarkable acknowledgement that cruelty and violence were not the sole preserve of Muslims. Notes   1. For the narrative and analysis of the Battle of the Field of Blood, see Cahen, 1940, 283–286; Smail, 1956, 179–180; Asbridge, 1997, 301–316; Hillenbrand, 1999, 108–110; Richard, 1999, 136–138; Asbridge, 2010, 163–167; Barber, 2012, 121–128; Cobb, 2014, 122–123. For the career of Il-Ghazi, see Hillenbrand, 1981, 250–292.   2. Walter the Chancellor, 1896 (Hagenmeyer). For the English translation with extended commentary, see Walter the Chancellor’s The Antiochene Wars, 1999. To avoid confusion, subsequent references to the introduction and commentary is this work will be cited as Asbridge and Edgington, The Antiochene Wars.  3. Asbridge and Edgington, The Antiochene Wars, 1999, 5–7. No reference to Walter the Chancellor has been found outside of his own account and the first external reference to the Antiochene office of chancellor related to a different holder, Ralph, in an 1127 charter issued by Prince Bohemond II.   4. Asbridge and Edgington, The Antiochene Wars, 1999, 43–72. Other primary sources for the Battle of the Field of Blood and its aftermath include: Fulcher

po rt ray al of vi olence i n bella antiochena  | 181 of Chartres, 1913, 621–624; Orderic Vitalis, 1978, 6: 104–108; William of Tyre, 1986, 556–562; Ibn al-Qalanisi, 1932, 159–161; Usama b. Munqidh, 2008, 131–132; Kamal al-Din b. al-Adim, La Chronique d’Alep, RHC Or. III: 617–622; Ibn al-Athir, 2006, I: 203–205; Matthew of Edessa, 1993, 223–224; Anonymous Syriac Chronicle, 1933, 87–88.   5. Asbridge and Edgington, The Antiochene Wars, 1999, 8.   6. Walter the Chancellor, 1896, 52–55; Asbridge and Edgington, The Antiochene Wars, 1999, 8–9.   7. Rubenstein, 2004, 131–168; Raymond of Aguilers, 1969.   8. Asbridge and Edgington, The Antiochene Wars, 1999, 7–10.   9. Asbridge, 1997, 308–309. On this debate see also: Riley-Smith, 1978, 87–102. 10. Asbridge and Edgington, The Antiochene Wars, 1999, 52–53. 11. A selection of recent works on Crusade violence includes: France, 1994; Barber, 2001; Mastnak, 2002; Kedar, 2004; Auffarth, 2005; Kangas, 2005; Housley, 2008; Abels, 2014; Smith, 2017. A very short selection of notable works on the study violence in the Middle Ages and beyond includes: Strickland, 1996; Kaeuper, 1999; Carroll, 2007; Brown, 2011; Buc, 2015). 12. Rubenstein, 2011. 13. Pinker, 2011, 169. 14. Armstrong, 2015, 214–215. 15. President Barack Obama, National Prayer Breakfast, 5 February 2015. A transcript is available at: (last accessed 25 June 2019). 16. Catlos, 2014, 161–162. 17. Huntingdon, Samuel, The Clash of Civilizations (New York, 1996). 18. Kedar, 2004, 15–75. 19. Robert the Monk, 2013; Gesta Francorum et aliorum Hierosolimitanorum, 1962; see also Sweetenham, 2005, 1–68; Bull, 2014, 127–139. 20. Asbridge, 2004, 33–36. 21. Bull, 2014, 127–139; Purkis, 2014. 22. Walter the Chancellor, 1896, 73. 23. Walter the Chancellor, ‘hostes eliminant, eviscerant et obtruncant’, 1896, 74. 24. Walter the Chancellor, 1896, 75–76. 25. Ibid., 75. 26. Walter the Chancellor, ‘viccisim tamen hinc et inde sanguis immensus ­funditur’, 75.

182  |  syri a i n crusad e r time s 27. Asbridge and Edgington, The Antiochene Wars, 11–42. 28. Walter the Chancellor, ‘necessario’, ‘doloris dolorum’, ‘repentina calamitate’, ‘mutato stilo’, ‘secondo detestabili’, 1896, 78–79. 29. Walter the Chancellor, ‘viccism effuso sanguine’, 1896, 81–82. 30. Walter the Chancellor, 1896, 87–89. 31. Ibid., 91–94, 105–12. See also Asbridge and Edgington, The Antiochene Wars, 1999, 5–6. At the start of Book Two, Walter also noted that his memory of events might have been ‘dulled by the experience of the prison-cell’ [Walter the Chancellor, 1896, 79]. 32. Walter the Chancellor, 1896, 91. 33. Ibid., 91–94, 105–107. 34. Ibid., 107–109. 35. Ibid., 110–112. 36. Friedman, 2002, 122–124. Friedman rightly observed that Walter’s narrative could be usefully compared to John of Joinville’s account of his experiences during the King Louis IX of France’s Egyptian crusade. 37. Walter the Chancellor, 1896, 92. 38. Ibid., 93. Elsewhere in these passages Walter seems to have used reported speech to explore and even rationalise his captors’ motives, with these explanations and justifications for their deeds varying from issues of religious difference and the Latin Christians’ supposed paganism, to the potential for prisoners to be ransomed either for money or strategic advantage, such as gaining control of the significant fortress at Azaz, north-west of Aleppo. 39. Kempf, 2014. 40. Asbridge and Edgington, The Antiochene Wars, 1999, 8. 41. Asbridge, 2004, 126, 167–168. 42. Walter the Chancellor, 1896, 91. 43. Ibid., 88; Ibn al-Athir, 2006, I: 204; Anonymous Syriac Chronicle, 1933, 88. 44. Asbridge, 2000, 159–161. 45. Walter the Chancellor, 1896, 107–109; Usama b. Munqidh, tr. Cobb, 2008, 131–132; Ibn al-‘Adim, 1988, 621–622. 46. Walter the Chancellor, 1896, 108. 47. Ibid., 108. 48. Hillenbrand, 2011. Hillenbrand also observed that Tughtegin’s name itself might have derived from his reputation for displaying the scalps of his enemies. 49. Fierro, 2008.

po rt ray al of vi olence i n bella antiochena  | 183 50. Gillingham, 2000; Suppe, 1989. 51. On the widespread use of skull drinking-cups, see Kim, 2015, 23; Fields, 2005, 55–59; Ogburn, 2007, 505–522. 52. Robert the Monk, 2013, 99. 53. Ibid., 44. 54. Walter the Chancellor, ‘ibique in effuso sanguine provoluti, ut sus in volutabro luti, necdum caede virorum satiati’, 1896, 93. 55. Walter the Chancellor, ‘Horum in momento corposibus detruncatis, totus pavimentum porticus aulae regiae profluens adspergit unda sanguinis’, 1896, 110. Walter also reported that Tughtegin harboured a desire to bathe in the blood of the slain Franks ‘and be restored to youth like an eagle’, an allusion to Psalms, 102 (103):5. 56. Walter the Chancellor, 1896, 81; Asbridge and Edgington, The Antiochene Wars, 1999, 114, n. 28. 57. William of Tyre, 1986, 412. On William of Tyre’s use of the Bella Antiochena, see Asbridge and Edgington, The Antiochene Wars, 1999, 9–10. 58. Walter the Chancellor, 1896, 92. 59. Ibid., 110. 60. Ibid., 108. 61. Asbridge and Edgington, The Antiochene Wars, 1999, 7–8. 62. Walter the Chancellor, 1896, 111–112.

10 Infernalising the Enemy: Images of Hell in Muslim Descriptions of the Franks during the Crusading Period1 Alex Mallett

I

n recent years, there has been considerable scholarly interest in perceptions of ‘the other’ between the Islamic world and the West, from the beginnings of Islam up to the present day, due in no small part to recent political events. For the medieval period, while scholarship in this field has forged ahead strongly in the area of Latin European perceptions of the Islamic world, questions of how Muslims saw Latin Europeans – the Franks – and why, have been comparatively ignored, with a few notable exceptions.2 This chapter represents a modest attempt to add to current knowledge in this field by studying one aspect of Muslim perspectives on the Franks during the Crusading period. In Hillenbrand’s ground-breaking study on Islamic perspectives on the Crusades, two of the chapters, ‘How the Muslims saw the Franks’ and ‘Aspects of Life in the Levant, 1099–1291’, examine Muslim perceptions of the Franks. These highlight that there is little surviving evidence to suggest that there was any attempt to understand the Franks and that the Muslim writers from the period tended to simply use pre-existing negative stereotypes. As such, the Franks as a whole are seen as stupid, uncouth, ill-educated, religiously gullible, unhygienic, and with little idea how to comport themselves properly, although there is some praise for their fighting qualities.3 Yet within one subsection of the latter of the two chapters, which has the rubric ‘Muslim views of the Frankish leadership’, Hillenbrand highlighted

i nf erna li si ng the enem y  | 185 that Muslim writers presented individual Franks, and specifically the Frankish leaders with whom they were most familiar, rather differently. Despite there being harsh invective aimed at Reynald of Châtillon (d. 583/1187) and Conrad of Montferrat (d. 588/1192), it is shown that the rest of the Frankish leaders are not seen similarly, and are instead either lauded or presented neutrally.4 Given the intensity of ill-feeling often displayed in the nearly two hundred-year period of encounters between Latins and Muslims in the Levant that constituted the Crusades in that region, this comparative lack of criticism of the Frankish leadership may seem, at first glance, rather surprising.5 However, such a lack of overt criticism of Christians, of any rite, by Muslim writers is fairly standard in medieval Arabic historical texts. This has been highlighted recently by David Thomas in an assessment of the ­presentation of Christians in Muslim chronicles from across the Islamic world in the period between the ninth and thirteenth centuries. Thomas wrote that in such works Christians are generally presented neutrally, when they are not being ignored, and no special attention is given to them, despite the fact that many were, at this time, highly placed within the administration of many Muslim governments.6 As such, it seems that the views of the Franks found in medieval Islamic writings from and about the Crusading period follows a wider pattern. However, this is not all that Thomas wrote on the subject. For he continues ‘it is possible that in some histories at least, Christians are singled out as alien and hostile by sophisticated ­structural elements and indirect portrayals … rather than descriptions’.7 The aim of this chapter to explore one specific aspect of medieval Islamic ­writings on the Franks in the Crusading period that fits neatly into that theory: how and to what extent Muslim writers used images of Hell within their texts to convey their negative opinion of the Franks to their readers. The Concept of Hell in Islamic Theology Before moving on to examining the images themselves, it will be useful to highlight a few of the main aspects of the concept of Hell in Islamic thought. As a subject, the Muslim idea of Hell remains little studied, and many basic questions surrounding its nature, purpose and inhabitants remain, although two recent contributions by Christian Lange have moved some way towards

186  |  syri a i n crusad e r time s doing so.8 The Islamic Hell is often described as al-nar (the fire), summoning up images of the Christian view of Hell with its scorching flames, but it is much more complicated than that. There are all manner of tortures including garrotting, impalement, the smashing of faces, and being made to stand up to the neck in fire; it is pitch black, so impossible to see; on occasion it is freezing cold rather than scorching hot; and the only relief offered to those suffering are ironic tortures – to those suffering from intense hunger, for example, only the disgusting fruits of the Zaqqum tree are given to them to eat, fruits with human faces that themselves try to bite back at those attempting to eat them. Another important aspect is that Hell is not outside God’s ‘remit’, as it seems to be in the (popular idea of the) Christian tradition; sinners are not thrown by God into it to be tortured by Satan’s evil minions outside God’s care or jurisdiction. In the Islamic idea of Hell, the torturers, labelled ‘punisher-angels’, are God’s own agents of justice, directed by him, and Satan and all the other evil spirits are also tormented there. A final important point that needs to be stressed is that, unlike in the Christian worldview, Hell in Islamic thought is not something that is only to be encountered by humans in the future – it exists here and now, synchronically with this world, and ‘the otherworld is in a continuous and intimate conversation with the world of the here-and-now’.9 This leads to the idea of the ‘eternal now’, in which an individual’s fate post-death is reflected in aspects of their life, appearance, or fate on earth.10 The aim of this chapter is to highlight ‘hidden messages’ that Muslim writers gave their audiences regarding their belief that the Franks were destined for Hell. However, before doing so, it will be useful to highlight briefly some examples where Muslim writers explicitly did so. Examples of such are most clearly seen in the writings of Saladin’s two close associates and apologists, Baha’ al-Din b. Shaddad and ‘Imad al-Din al-Isfahani. For example, in his account of the death of the German emperor Frederick Barbarossa, ‘Imad al-Din writes simply that he suffered an illness that ‘led him to Hell’.11 The same author described how the Frankish soldiers at Hattin were ‘the people of the Fire (al-nar)’.12 Perhaps the most vivid example of this comes in ‘Imad al-Din’s typically overwrought description of the death of Conrad of Montferrat:

i nf erna li si ng the enem y  | 187 The terminator of his [Conrad’s] hope came to the door. He was summoned to Hell, [the chief angel of Hell] Malik anticipated his entry, and Hell awaited him; the very deepest level of Hell burned for him, the fire blazed for him and the flames roared as they awaited him.13

Similarly, in a letter sent by al-Qadi al-Fadil to Baghdad, in which he reported the triumph of Saladin’s forces over Reynald of Châtillon’s Red Sea raid in 578/1183, it is written: ‘They [the Muslims] pursued them for five days on horseback and on foot, by day and by night, until they did not leave a trace of them: “And those who disbelieved will be driven to Hell in groups” [Q 39:71]’.14 Such overt references, which occur fairly regularly in the texts written during or about Saladin’s reign, help to provide some context for the main area of investigation of this chapter, which will be divided into two main sections: the infernalisation of individual Franks, and the infernalisation of the Franks as a group. The Infernalisation of Individual Franks This section will begin with perhaps the most obvious example of the infernalisation of an individual, that of Reynald of Châtillon, who was criticised as, among other things, ‘one of the most devilish of the Franks, and [one of] the most demonic’.15 Yet, in addition, some rather subtler devices are employed to link this Frank with Hell. For example, during the reported (and invented) debate between Reynald and Raymond of Tripoli just before the Battle of Hattin, Ibn al-Athir puts into Reynald’s mouth the following statement regarding the enormous size of the Muslim army: ‘As for what you [that is, Raymond] say, that they are great in number, a large load of fuel for Hell will not harm it’.16 The author here uses a form of dramatic irony, as his Muslim readership would know very well both the outcome of the battle and the fact that it was the Franks, and Reynald foremost among them, rather than the Muslims who were soon to provide Hell with its fuel. Similarly, following his execution, Reynald’s body is described by Abu Shama as having been dragged out of Saladin’s tent.17 This act prefigures the current state of Reynald’s soul, as one of the punishments for sinners in Hell

188  |  syri a i n crusad e r time s is to be dragged around by their forelocks by the punisher-angels, as indicated by Q 96:15–18: Let him beware! If he desists not, We will drag him by the forelock, a lying, sinful forelock! Then, let him call (for help) to his council (of comrades); We will call on the angels of punishment (to deal with him)!18

As such, the ‘eternal now’ is demonstrated through the fate of the Franks, but also via Saladin’s forces, who, through their dragging of Reynald’s body, can be seen as the instrument of divine justice, earthly versions of the punisher-angels. Another example is evident in Sibt b. al-Jawzi’s description of the German emperor Frederick II.19 While modern scholars have noted the intensity of feeling expressed by the Muslim author, and the ridiculousness of the description, there is more going on than simple insults, as the description summons up images that would remind the Muslim readers of the time of the denizens of Hell.20 First, Frederick is described as bald and short-sighted. This, while pandering to the pre-existing prejudices Muslims had against the Franks, also suggests the infernal end of Frederick, through the Islamic notion of ‘ilm al-firasa (the science of physiognomy). This was essentially a form of divination in which a person’s physical blemishes or the shape of their body was used to gain an insight into their character; the more deformed a person was, the worse a person they supposedly were.21 Thus, by highlighting his facial ‘deformities’, Sibt b. al-Jawzi also implies that he had spiritual ones too, and thereby suggests his ultimate destination, particularly as ‘maimed faces were the sign of sinners in Hell’.22 The description of Frederick as having a red, and therefore darkened, face may also be significant. For in Islamic eschatology the inhabitants of Paradise have white faces, while those of Hell have black ones, and the reddening of his face could be read as a prelude to his inevitable fate, as if it is in the process of turning black.23 Certainly it would have made little sense for him to describe Frederick’s face as white in such a context.24 Another Frankish leader described in such a way is Baldwin IV, who is described by al-Qadi al-Fadil as a ‘freckled and leprous malefactor’ in al-Barq al-shami.25 Even if the hadith that states ‘one should run away from the leper as one runs away from a lion’26 is ignored, to the Muslim reader it

i nf erna li si ng the enem y  | 189 would also suggest God’s punishment on Baldwin, the ruler of the Franks, for immorality – either his own or that of his subjects – as leprosy was seen as a punishment for that sin.27 Furthermore, freckles are seen as being caused by leprosy and sometimes presage death in Islamic thought.28 Thus, like Frederick, ‘ilm al-firasa suggests his extreme physical deformity reflects his extreme moral deformity, and by extension that of the Franks in general, and their ultimate destination. The final example of the Frankish leaders comes in the description of the incarceration of King Baldwin II of Jerusalem, Joscelin I of Edessa, and other Frankish leaders, in Khartbirt in the mid-510s/early 1120s. The text, found in the writings of Ibn al-‘Adim, reads, very simply: ‘they were imprisoned in the pit of Khartbirt’.29 At first glance, this may seem like a fairly restrained description of the imprisonment of some of the Frankish leadership. Yet what is important in this context is the Arabic verb used to describe their state of imprisonment. Rather than use a term such as asara, as is often found elsewhere, here the word employed is masjunin (they were imprisoned). The root letters of this word, sin-jim-nun (‫)سجن‬, bring to mind the word Sijjin, which is found in Q 83:7–8: ‘Surely the record of the wicked is (preserved) in Sijjin. And what will explain to thee what Sijjin is?’ A number of medieval Qur’anic exegetes, such as Fakhr al-Din al-Razi (d. 606/1209), interpreted this verse as meaning that Sijjin was a prison in the very lowest circle of Hell, where Satan himself is chained.30 Thus, through his use of this specific term it seems that Ibn al-‘Adim (or his source) regarded the fate of the two Frankish leaders, again by use of the concept of the ‘eternal now’; as they had been imprisoned in that Sijjin on earth, so they are, at the time of writing, in the eternal Sijjin.31 The Infernalisation of the Franks as a Group One of the aspects of Muslim presentations of the Franks that has been noted most often in modern examinations of the subject are the various derogatory epithets given to them, such as ‘polytheists’ (mushrikun, also translatable as ‘idolaters’).32 While these terms do, as has been widely noted, mark out the religious ‘otherness’ of the Franks, these would not have been simply insults to the Muslim writers who employed them or their audiences; instead, they would have highlighted that they would lead the Franks, as a group, to Hell.

190  |  syri a i n crusad e r time s This is because the sin of shirk, of associating other things with God, whence comes the term mushrikun, is said to be the only sin that will be eternally punished.33 Yet, in addition to these overt statements, there are other means by which Muslim writers linked the Franks as a group with Hell. One of these is the employment of animal imagery. For example, they are called pigs by Ibn Jubayr and Ibn al-Furat.34 Not only would such epithets have reflected the ritual impurity of the Franks, as Hillenbrand has noted,35 they would also have served to underscore the eternal fate of the Franks in Hell, as one of the punishments for those in Hell is to be turned into pigs, along with other animals such as dogs and apes.36 In a somewhat similar vein, ‘Imad al-Din alIsfahani describes the Franks as being like a swarm of flies and grasshoppers.37 Just as, to this writer, the insects and the Franks share similar characteristics, so is their eternal fate seemingly the same, a fate which can be seen in the hadith reported by al-Suyuti that all flying insects (except bees) are in Hell.38 Another means is the ascription to them of vices associated with the people of Hell, of which two are greatly significant. The first of these is arrogance. Of particular importance in this context are the words of Q 40:70–5, which declare that the arrogant will have their eternal abode in Hell, and which specifically highlights the fundamental relationship between polytheism (shirk), unbelief and arrogance. As the two former are sins that will lead the perpetrators straight to Hell, thus, too, will the linked sin of arrogance. Furthermore, arrogance is at its greatest when against God; the two best examples of this in the Qur’an are the Devil (Iblis) and Pharaoh, both of whom failed to act as they should towards God because of their arrogance (Q 2:34 and 10:75). As arrogance is demonstrated regularly by the Franks because of their resistance to the Muslims, and by extension God, the medieval Muslim hearing or reading of their arrogance would immediately understand the place that was reserved for them in Hell; as Q 40:46 suggests, the arrogant are sent to the lowest depths of Hell. There are a number of specific examples in which such can be inferred. One is in Sibt b. al-Jawzi’s account of the siege of Damascus by the Second Crusade in 543/1148. Here, the writer details how a Frankish priest, who was holding a cross and the gospels in his hands, rode in front of the Frankish army and proclaimed to them: ‘The Messiah has promised that today this

i nf erna li si ng the enem y  | 191 city shall be destroyed.’39 However, his hubris is misplaced, as, at that very moment, the true followers of the Messiah (the Muslims), made an attack on the assembled Frankish army and routed them, and the author is sure to note that the arrogant priest was killed by one of the Muslim soldiers. Another example of this is Reynald of Châtillon’s comments to Raymond of Tripoli, seen earlier; he arrogantly presumes that the Franks would triumph over the Muslims at Hattin, whereas he, like the Frankish priest at Damascus, received his comeuppance and was killed.40 The other Frankish vice leading to Hell is greed. This can be related in a number of hadith, including the following: ‘[Muhammad] said: The inmates of Hell are five groups: … [one of whom is] those dishonest [people] whose greed cannot be concealed even in the case of minor things.’41 Frankish greed is seen in many passages, such as Ibn al-Athir’s hyperbolic and religiously charged description of the Franks’ looting following their capture of Jerusalem,42 and in their capture of Tripoli in 503/1109–10, when the same writer comments that ‘they plundered from its population money, goods and books beyond computation’.43 While there is little to suggest that such did not occur, it is the reporting of it serves the purpose of highlighting the Franks’ attitude and consequent fate. Conclusion It was the aim of this chapter to explore the images of, or allusions to, the Islamic Hell and its denizens that can be seen in medieval Muslim writers’ descriptions of the Franks, and thereby assess whether their written works contain a wider general level of hostility to the Franks and their leaders, manifested through subtle allusions, than has previously been thought, as Thomas posited. The above discussion has demonstrated that the Franks as a group and their individual leaders were described, both overtly and rather more subtly, in ways that would bring to mind the image of Hell in the minds of medieval Muslim readers. Thus, Thomas’s thesis would seem to have been vindicated. However, it would be too hasty to concurrently accept the idea that the Franks were wholly reviled by the Muslim writers, and thereby reject Hillenbrand’s assessment. For the infernal images highlighted above are found primarily in the works of writers such as ‘Imad al-Din al-Isfahani,

192  |  syri a i n crusad e r time s Ibn Jubayr and Sibt b. al-Jawzi; these are writers who also overtly criticise the Franks as being devilish, demonic and untrustworthy, among various other negative descriptions. Thus, this article has suggested that there are no examples of new writings in which Muslims criticise the Franks, merely that those writings in which such was already known have been shown to have another method of so doing. The writers who seem to have held a generally more neutral view, or even one approaching positivity, do not seem to have employed infernal imagery in their descriptions of the Franks. Notes   1. I would like to thank Christian Lange for reading a draft version of the conference paper on which this chapter is based, and for his useful comments and suggestions.   2. There have been many studies published recently on Latin Christian perceptions of Muslims in writings from the Middle Ages. These include (but are in no way limited to): Daniel, 1960; Southern, 1962; Tolan, 2000; Tolan, 2002; Tolan, 2008; Bly Calkin, 2005; Conklin Akbari, 2009; and Pagès, 2014). Studies of Muslim perceptions of Latin Europeans include: Lewis, 1982; alAzmeh, 1992; Hillenbrand, 1999; and Mallett, 2014a. An excellent overview of Christian–Muslim relations and perspectives, from both sides and across the whole of the medieval period, is Thomas and Mallett, 2009–13 (available at: http://referenceworks.brillonline.com/browse/christian-muslim-relations-i (last accessed 20 November 2018)), which contains a number of useful introductory essays and a brief summary of and full bibliography for, as far as was possible to ascertain, every surviving piece of source material in which someone from one religion wrote about or against the other religion.   3. Hillenbrand, 1999, 257–427.  4. Those lauded are Baldwin II, Richard the Lionheart, Louis IX of France, Frederick II and Manfred, while those who are presented neutrally are Baldwin I, Bohemond VI, Guy II, Raymond of Tripoli, and Reynald of Sidon; see Hillenbrand, 1999, 336–345.   5. Even the negative stereotypes of the Franks in general alluded to above must be seen in the same way; these were, as stated earlier, pre-existing images of the European ‘other’, and were not, in general, intensified by the conflict, although they do, unsurprisingly, appear more often; see Hillenbrand, 1999, 267–282.

i nf erna li si ng the enem y  | 193   6. Thomas, 2011, 15–27. This lack of interest in the subject by medieval Muslim chroniclers is reflected by modern scholars of medieval Islamic historiography, who have almost completely ignored how Christians are presented; see, for example, Khalidi, 1994; Duri, 1983.   7. Thomas, 2011, 16.   8. Lange, 2015c; Lange, 2015b; Lange, ‘Hell’, EI 3.   9. Lange, 2015c, 12. 10. These ideas are explored in detail in Lange, 2015c, and Lange, 2015b. 11. Abu Shama, 1898, 4: 458. 12. ‘Imad al-Din al-Isfahani, 1972, 25. 13. Ibid., 420. 14. Abu Shama, 1898, 4: 235. 15. Ibn al-Athir, 2006, X: 18. 16. Ibid., X: 23. 17. Ibid., X: 26. 18. For the punisher-angels and their dragging of sinners around by their forelocks, see Lange 2015d, 74–99, at 83; Lange, 2015c, 145. 19. Sibt b. al-Jawzi, 1951, II: 656. 20. Sibt b. al-Jawzi never laid eyes on Frederick, and his presentation is merely a method by which he can ridicule the Frankish rulers, which in turn was primarily based on the rivalry between his own political master al-Nasir and the Egyptian sultan al-Malik al-Kamil, as the latter signed the very unpopular Treaty of Jaffa (1229) with Frederick. Thus, the image is ultimately based on intra-Muslim political rivalries, although in order to be effective it has a strong religious basis; see Hirschler, 2014, 157–159. 21. See Lange, 2007, 437–438. 22. Ibid., 440. 23. It may also be of significance that, according to some medieval Muslim writers, God used his breath to fan Hell’s flames so that they started white, turned red, and ended up black; perhaps the same process was regarded as happening to Frederick’s face to show his infernal end. See Lange 2015c, 134. 24. Lange, 2007, 437; Lange, ‘Hell’, EI 3. 25. Cited in Hillenbrand, 1999, 295, from an unpublished edition by Lutz Richter-Bernburg. 26. Bukhari, Sahih, 5707. Available at (last accessed 30 May 2019). 27. Dols, ‘Djudham’, EI 2.

194  |  syri a i n crusad e r time s 28. Fahd, T., ‘Shama’, EI 2. 29. Ibn al-‘Adim, 1997, 1: 410. For the use of the ‘pit’ as a prison see Hillenbrand, 2003, 86. 30. Razi, 1934–62, XXX, 84; cf. Lange, 2015a, 1–28. See also Hillenbrand, 2003, 86–87. 31. Ibn al-‘Adim was writing in the mid-thirteenth century, although he had access to contemporaneous sources that are no longer extant. It is likely to have been they, rather than he himself, who introduced this term, as he seems generally to have held a neutral position towards the Franks. Baldwin II and Joscelin I are the only Frankish leaders who are heavily criticised by him (or his source), because of their very damaging plundering of Aleppo and its hinterlands for decades; see Eddé, 2014, 128. 32. See, for example, Hillenbrand, 1999, 299, 301, 303, 304. Related to this is a tradition in which the specific fashioner of images is sent to Hell; see Asad b. Musa, 1976, #78. 33. See Lange, 2015c, 155. 34. Ibn Jubayr, 1907, 307; Ibn al-Furat, 1986, 206. 35. Hillenbrand, 1999, 284–303, passim. 36. Lange, 2015c, 150. 37. Cited in Hillenbrand, 1999, 295, from an unpublished edition by Lutz Richter-Bernburg. 38. See Lange, 2015c, 139. 39. Sibt b. al-Jawzi, 1951, VIII/1: 198. 40. See above, XXX. For further comments on Frankish arrogance, see Christie, 2014, 17–18. 41. Sahih Muslim, 2865. Available at (last accessed 30 May 2019). 42. Ibn al-Athir, 2006, VIII: 425. 43. Ibid., VIII, 579.

11 Sunnites et Chiites à Alep sous le règne d’al-Salih Isma‘il (569–77/1174–81): entre conflits et réconciliations Anne-Marie Eddé

L

’étude des relations entre sunnites et chiites à Alep au vie/xiie siècle n’est pas nouvelle. Dès 1935, dans un article pionnier, Claude Cahen avait insisté sur la situation des chiites en Syrie du Nord à travers le témoi­gnage de l’historien chiite Ibn Abi Tayyi’ (m. 627/1230).1 Nikita Elisséeff et Henri Khayyat abordèrent aussi le sujet, l’un dans sa monographie sur Nur al-Din et l’autre dans un article sur les révoltes chiites.2 Plus récemment, Carole Hillenbrand a consacré un article aux chiites alépins sous le règne des Zenguides et Stephennie Mulder a, de son côté, renouvelé le sujet avec son étude des sanctuaires alides de Syrie.3 Tous ces travaux ont souligné la forte implantation du chiisme à Alep, depuis l’époque hamdanide au IVe/Xe siècle. Contrairement à ce que l’on pourrait penser, l’arrivée des Seljoukides ne modifia pas vraiment la situation, car jusqu’aux mesures prises par Nur al-Din contre les chiites, au milieu du xiie siècle, la population resta majoritairement chiite duodécimaine. Tous les dirigeants zenguides, que ce soit Aqsunqur, gouverneur d’Alep de 479/1086 à 487/1094, son fils Zengi (522–41/1128–46) ou son petit-fils Nur al-Din (541–69/1146–74), prirent grand soin des sanctuaires alides d’Alep, en particulier du Mashhad al-Muhassin (ou al-Dikka), situé à l’ouest de la ville.4 Même Nur al-Din, souvent décrit comme un défenseur du sunnisme, ne fut pas hostile, au début de son règne, aux chiites alépins, comme l’atteste l’historien chiite Ibn Abi Tayyi’.5 Ce n’est qu’en 543/1148–9 qu’il prit des mesures contre eux,6 pour des raisons essentiellement politiques (obtenir

198  |  syri a i n crusad e r time s l’alliance des Bourides de Damas) et sous la pression de quelques oulémas sunnites. Le chiisme duodécimain ne fut pas éradiqué pour autant comme le montrent les émeutes chiites de 552/1157, lors de la grave maladie de Nur al-Din, et les troubles au lendemain de sa mort, en 569/1174. Plus tard, à l’époque ayyoubide, même si les duodécimains n’étaient plus en position majoritaire, ils ne conservaient pas moins encore à Alep une certaine liberté d’action et de parole.7 Dans cette historiographie du chiisme alépin, on notera que le court règne d’al-Salih Isma‘il (569–77/1174–81) a été peu étudié. Pourtant, les relations entre sunnites et chiites à cette époque fournissent un certain nombre d’informations intéressantes non seulement sur les rapports qu’entretenaient les deux communautés, mais aussi sur les luttes sociales et politiques qui agitaient alors la ville d’Alep. À la mort de son père, al-Salih Isma‘il n’avait que 11 ans.8 Il en ­résulta une certaine confusion dans les anciens territoires de Nur al-Din: à Mossoul, son neveu Sayf al-Din s’empara des territoires situés en Jéziré, entre le Tigre et l’Euphrate; à Damas, où se trouvait al-Salih, c’est l’émir Ibn al-Muqaddam qui fut désigné par ses pairs commandant de l’armée et chef de l’administration; à Alep, Nur al-Din avait prévu, avant de mourir, de confier la régence à l’eunuque Gümüshtegin, alors gouverneur de la citadelle de Mossoul, mais quand celui-ci arriva pour prendre ses fonctions, trois autres émirs s’étaient déjà emparés du pouvoir: deux frères de la puissante famille des Banu’l-Daya se partageaient l’administration et la police, tandis que Shadhbakht, eunuque d’origine indienne, contrôlait la citadelle.9 Pendant ce temps, au Caire, Saladin exerçait le pouvoir. Il reconnut officiellement la suzeraineté d’al-Salih en faisant frapper au Caire des dinars égyptiens au nom du jeune souverain, ce qui ne l’empêcha pas de s’emparer de Damas dès le mois de rabi‘ I 570/octobre 1174. Le début du règne d’al-Salih à Alep fut dès lors marqué par une lutte entre les Zenguides d’Alep et de Mossoul d’une part, les troupes de Saladin d’autre part. Après deux tentatives infructueuses de Saladin pour s’emparer d’Alep (en 570/1174 et en 571/1176) et de longues négociations, une paix fut finalement conclue, en 571/1176, aux termes de laquelle Saladin renonçait une fois pour toutes à exercer la régence et reconnaissait le pouvoir d’al-Salih

sunni tes et chi i tes à a lep  | 199 sur tout le nord de la Syrie. Ce traité fut respecté de part et d’autre jusqu’à la mort d’al-Salih, à l’âge de 19 ans, en 577/1181.10 Le sujet qui nous intéresse ici est celui de la nature et des enjeux des conflits entre sunnites et chiites dans la ville d’Alep entre 1174 et 1181. Ces oppositions étaient-elles uniquement religieuses ou recouvraient-elles des fractures plus politiques et sociales ? Face aux conflits quelles furent les stratégies mises en œuvre par le pouvoir pour calmer les esprits et renforcer les solidarités entre sunnites et chiites ? Nous essaierons de répondre à ces questions en trois temps: en présen­ tant d’abord le contexte politique à Alep au lendemain de la mort de Nur al-Din, en étudiant ensuite les concessions faites aux chiites duodécimains et en analysant, enfin, la politique des dirigeants alépins à l’égard des chiites ismaïliens. Les enjeux politiques à Alep au lendemain de la mort de Nur al-Din Dès que la mort de Nur al-Din fut connue, à Alep, une violente crise politique éclata doublée d’un conflit entre chiites et sunnites.11 D’un côté, les frères Banu’l-Daya, qui contrôlaient la citadelle et la police, auraient souhaité exercer la régence au nom d’al-Salih; ces émirs turcs avaient acquis une grande puissance sous le règne de Nur al-Din, l’un des leurs, Majd al-Din, mort quelques années plus tôt en 565/1170, ayant été frère de lait et proche conseiller de Nur al-Din.12 D’un autre côté, les habitants chiites d’Alep, sous la conduite de leur chef Ibn al-Khashshab,13 craignaient leurs ambitions et surtout leurs positions très hostiles au chiisme, et se déclaraient ouvertement en faveur du jeune prince al-Salih. Les ambitions de cette famille à exercer le pouvoir à Alep n’étaient pas nouvelles. Dès la maladie de Nur al-Din en 552/1157, les Banu’l-Daya avaient eu des prétentions à succéder à Nur al-Din, ce qui avait entraîné une révolte des chiites qui avaient tenté de proclamer comme successeur de Nur al-Din, l’un de ses frères, Nusrat al-Din, connu pour sa tolérance à l’égard du chiisme.14 Déjà à cette époque, les chiites s’en étaient pris aux grandes personnalités sunnites et à leurs biens,15 trouvant dans la milice urbaine des ahdath un appui de poids. Bien que l’appartenance religieuse de ces derniers ne soit pas précisée, le fait que Nusrat al-Din rétablît l’appel à la prière chiite sous leur pression16 laisse penser qu’une majorité d’entre eux au moins était

200  |  syri a i n crusad e r time s chiite. Ce qui se jouait, toutefois, dans ces conflits, c’était moins une lutte pour le pouvoir entre chiites et sunnites – puisque tous les prétendants au trône étaient sunnites – que la tentative des chiites de maintenir sur le trône un dirigeant qui les laisserait exercer librement leur culte, comme Nur al-Din l’avait fait au début de son règne. La guérison de Nur al-Din avait suffi à ramener l’ordre, mais sa mort, en 569/1174, réveilla les querelles. Les chiites se rangèrent derrière Ibn alKhashshab, proclamant qu’ils voulaient défendre les intérêts d’al-Salih contre les ambitions des Banu’l-Daya. Plusieurs grandes familles sunnites comme les Banu’l-‘Ajami, et les Banu Amin al-Dawla, se rangèrent derrière les Banu’lDaya.17 D’autres comme les Banu’l-Tarsusi18 ou les Banu’l-‘Adim refusèrent de les rejoindre, ce qui montre bien que le conflit qui allait suivre n’était pas une simple opposition entre sunnites et chiites, mais aussi une lutte entre grandes familles pour le contrôle du pouvoir. Un auteur sunnite comme Ibn al-‘Adim accuse d’ailleurs sans détours les Banu’l-Daya d’avoir attisé le conflit en provoquant ‘une fitna à Alep entre sunnites et chiites’.19 Il affirme aussi que Shams al-Din b. al-Daya avait l’intention d’installer al-Salih sous sa coupe dans la citadelle afin de gouverner seul, sans l’aide des autres émirs. L’attitude de cet historien et notable alépin, s’explique sans doute par sa volonté de défendre lui aussi les intérêts de la dynastie zenguide que son père et son oncle avaient fidèlement servie, l’un comme qadi et l’autre comme khatib.20 Elle démontre surtout que la solidarité politique pouvait, dans certains cas, dépasser les ­divisions religieuses. Les Banu’l-Daya firent croire à Ibn al-Khashshab que certains l’accusaient de vouloir s’emparer du pouvoir, ce qui eut pour effet immédiat de provoquer une violente réaction de ses partisans. Les maisons de plusieurs notables sunnites furent pillées, en particulier celles du chafiite Qutb al-Din b. al-‘Ajami et du hanafite Baha’ al-Din b. Amin al-Dawla dans laquelle était aussi conservé l’argent servant à l’entretien des orphelins.21 Ce même jour, le cheikh sunnite Abu’l-‘Abbas al-Maghribi fut tué dans la madrasa des Verriers où il exerçait les fonctions de lecteur de Coran et traditionniste. En 552/1157, déjà, les chiites s’en étaient pris aux madrasas, preuve qu’elles étaient bien perçues comme des institutions hostiles au chiisme.22 L’émeute entre les deux partis s’étendit dans la ville et les plus démunis se mirent à piller les maisons des plus riches. La révolte politico-religieuse se transforma ainsi en révolte sociale.

sunni tes et chi i tes à a lep  | 201 Les Banu’ l-Daya soutenus par des combattants et des habitants sunnites de la ville, accusèrent Ibn al-Khashshab de vouloir s’emparer du pouvoir ‘avec l’appui des chiites et d’une partie des sunnites’, selon les propres termes d’Ibn al-‘Adim.23 Ils obtinrent des renforts de l’émir Shadhbakht qui dépêcha en ville des mineurs et des lanceurs de naphte (zarraqin), c’est-à-dire de gros moyens, pour arrêter l’émeute, en attendant l’arrivée à Alep d’al-Salih. Les combattants hostiles aux chiites élevèrent des barricades, se glissèrent par les rues et par les toits jusqu’à la maison d’Ibn al-Khashshab qu’ils pillèrent puis incendièrent. Des maisons voisines subirent le même sort. Le cadi Ibn al-Khashshab s’enfuit et se cacha dans une maison d’un quartier voisin où il demeura jusqu’à l’arrivée d’al-Salih à Alep au mois de muharram 570/début août 1174. Pendant ce temps, le convoi ramenant al-Salih de Damas se rapprochait d’Alep. Sa mère, qui semble avoir joué un rôle dans la régence, l’accompagnait.24 Shadhbakht et Gümüshtegin, avec la complicité de quelques émirs de Damas, décidèrent alors de se débarrasser des Banu’ l-Daya qui furent arrêtés et jetés en prison.25 Une fois installé dans la citadelle, al-Salih accorda un sauf-conduit à Ibn al-Khashshab. Celui-ci se rendit sans se méfier à la citadelle, en compagnie d’un certain nombre de ses compagnons, mais tomba dans un guet-apens qui lui coûta la vie. Sa tête fut exhibée d’abord à la porte de la citadelle, puis sur l’une de ses tours. Ibn Abi Tayyi’26 explique ce retournement de situation par la pression qu’exercèrent les Banu’l-‘Ajami27 et les Banu Amin al-Dawla, dont les maisons avaient été pillées lors de l’émeute chiite, sur l’émir ‘Izz al-Din Jurdik un ancien mamelouk de Nur al-Din. Alléché par l’idée de récupérer une partie de ses biens perdus lors de l’émeute, Jurdik obtint d’al-Salih le sauf-conduit qui attira Ibn al-Khashshab dans un piège.28 Même si le récit est rapporté par un auteur chiite, il apparaît clairement qu’Ibn al-Khashshab fut victime de la vengeance des Banu l-Ajami et Banu Amin al-Dawla. Al-Salih et son proche entourage n’avaient aucune véritable raison de le faire exécuter puisqu’il avait pris le parti du jeune prince et l’avait déjà chèrement payé. Cette victoire que le parti sunnite d’Alep remportait grâce à l’inexpérience du jeune prince, relevait donc au mieux d’une lutte de pouvoir entre grandes familles alépines et au pire du règlement de compte personnel. Rien n’indique, en tout cas, qu’il s’agissait d’un véritable conflit entre sunnites et chiites.

202  |  syri a i n crusad e r time s Ces émeutes que l’on a toujours considérées comme des conflits s­ unnites– chiites, furent donc en réalité beaucoup plus politico-sociales que religieuses, la question centrale étant de savoir à qui devait revenir le pouvoir à Alep au lendemain de la mort de Nur al-Din alors que sa succession était encore mal assurée. Le clivage – on l’a vu – ne se réduisait pas à une opposition entre sunnites et chiites, car si les Banu’l-Daya réussirent à rallier une partie de la population sunnite, ils ne parvinrent pas à convaincre certaines autres grandes familles sunnites et ils furent eux-mêmes finalement victimes des eunuques sunnites (Shadhbakht et Gümüshtegin) et non des chiites. En outre, un certain nombre de sunnites suivirent le chef chiite Ibn al-Khashshab. Bien qu’il ne le dise pas ouvertement, Ibn al-‘Adim laisse penser, dans sa façon de présenter les événements, que sa propre famille fut de ceux-là. L’historien, en effet, non seulement appartenait à l’une des anciennes familles alépines, installées dans la région depuis au moins deux siècles,29 mais défendait aussi la dynastie des Zenguides que les Banu’l-‘Adim avaient toujours fidèlement soutenue. Notons, enfin que ces émeutes s’achevaient sans véritable vainqueur, du côté sunnite comme du côté chiite, puisque les Banu’l-Daya croupissaient en prison tandis qu’Ibn al-Khashshab subissait un châtiment ignominieux.30 S’il existait incontestablement une opposition entre sunnites et chiites, celle-ci était donc loin d’être déterminante. Les conflits qui opposaient d’une part les grande familles alépines entre elles pour le contrôle du pouvoir, et d’autre part les notables locaux – qu’ils soient d’installation ancienne ou plus récente comme les Banu’l-Daya – aux eunuques récemment arrivés au pouvoir, ont joué un rôle bien plus important. Al-Salih et le chiisme duodécimain La position clairement pro-zenguide des chiites duodécimains, au lendemain de la mort de Nur al-Din, explique sans aucun doute la politique très tolérante d’al-Salih à leur égard, tandis qu’eux-mêmes lui pardonnèrent assez vite, semble-t-il, le meurtre de leur chef, préférant en attribuer la responsabilité à son émir ‘Izz al-Din Jurdik. Comme son père, son grand-père et son arrière-grand-père, al-Salih mani­festa un grand intérêt pour les sanctuaires alides d’Alep. C’est sous son règne, en effet, que fut construit le Mashhad al-Husayn à quelque trois

sunni tes et chi i tes à a lep  | 203 cents mètres du Mashhad al-Muhassin, à l’ouest d’Alep. La légende voulait qu’un berger vit ‘Ali en songe lui indiquer l’endroit où construire un sanctuaire dédié à son fils al-Husayn. Al-Salih Isma‘il, lui-même, disait-on, avait encouragé la population à se mettre au travail pour édifier le monument, une façon pour lui d’utiliser une forme de religiosité populaire pour renforcer l’unité de sa population, sunnites et chiites confondus. Les successeurs d’alSalih, ayyoubides et mamelouks (Baybars en particulier), prirent eux aussi grand soin de l’édifice. Plus significative encore de l’esprit conciliant d’al-Salih à l’égard des chiites, fut son attitude lors du second siège de la ville par Saladin au mois de dhu l-hijja 571/fin juin 1176. Toute la population, sunnites et chiites confondus, se mobilisa, nous dit-on, contre Saladin. Les habitants sortaient régulièrement de la ville pour attaquer les assiégeants et même les menaces des pires sévices infligés par leurs adversaires à leurs prisonniers – comme d’avoir les pieds coupés – ne les dissuadaient en rien. La population dans son ensemble encouragea le jeune al-Salih à résister et accueillit l’émissaire de Saladin à coups de pierres, même si les négociations finirent malgré tout par aboutir le 16 muharram 572/25 juillet 1176.31 C’est au cours de ces événements qu’al-Salih, pour obtenir le soutien des duodécimains, leur rendit la plupart des droits que Nur al-Din leur avait enlevés: le droit de disposer de la partie orientale de la grande mosquée, de lancer l’appel à la prière chiite (‘Venez à la meilleure des œuvres’), d’évoquer dans les cortèges funèbres et dans les marchés les noms de leurs douze imams, de prononcer la prière chiite des morts32 et d’être sous la juridiction de leur cadi, le sharif Abu’lMakarim Hamza b. Zuhra al-Husayni (m. 585/1189),33 en particulier pour les mariages; ils obtinrent aussi l’application de la loi à tout fauteur de trouble qui provoquerait un conflit (fitna) entre sunnites et chiites,34 et le père d’Ibn Abi Tayyi’ fut autorisé à diriger la prière dans la partie orientale de la grande mosquée. Ces quelques indices montrent qu’al-Salih, tout aussi sunnite que son père – comme en témoigne son respect envers le célèbre juriste hanafite alKasani (m. 587/1191) installé à Alep sous le règne de Nur al-Din – n’eut pas moins, par conviction ou par calcul politique, une attitude très conciliante envers les duodécimains d’Alep. Ce faisant, il obtint leur appui et évita toute nouvelle émeute entre chiites et sunnites.

204  |  syri a i n crusad e r time s La politique des dirigeants alépins à l’égard des ismaïliens La politique d’al-Salih à l’égard des chiites ismaïliens, membres de la fameuse secte des Assassins, fut bien différente. L’hostilité des Alépins envers ces chiites extrémistes datait du début du VIe/XIIe siècle. Il suffit, pour s’en convaincre, de rappeler leur violente réaction à leur encontre, en 507/1113, au lendemain de la mort de Ridwan accusé de les favoriser.35 Tous, sunnites et chiites duodécimains, redoutaient en effet leurs actions et les condamnaient sans aucune ambiguïté.36 Sous le règne d’al-Salih Isma‘il, il faut distinguer néanmoins deux sortes d’attitudes parmi les dirigeants. Gümüshtegin, en effet, rechercha leur ­alliance pour mieux éliminer ses rivaux: par deux fois, il demanda l’aide des Assassins pour se débarrasser de Saladin qui assiégeait Alep, en 570/1174 puis en 571/1176.37 Quelque temps plus tard, il semble avoir été également à l’origine de l’assassinat de l’influent vizir alépin Shihab al-Din b. al-‘Ajami, le 4 rabi‘ I 573/31 août 1177. Gümüshtegin usa, semble-t-il, d’un stratagème en donnant à signer au jeune al-Salih un document en blanc qu’il envoya ensuite à Sinan pour lui demander de faire assassiner le vizir. Sinan ne se douta pas qu’il s’agissait d’un faux et envoya ses adeptes tuer le vizir à la sortie de la mosquée. Les deux assaillants furent également tués lors de cet assaut. Est-ce pour dédouaner al-Salih, qui n’avait alors que quatorze ans, qu’Ibn al-‘Adim rejette la responsabilité de ce meurtre sur le seul eunuque ? Ou bien Gümüshtegin, qui avait déjà sollicité, trois ans plus tôt, l’aide des Assassins pour tuer Saladin, figurait-il naturellement au premier rang des suspects ? ‘Imad al-Din al-Isfahani, contemporain de ces événements et repris plus tard par Ibn al-Athir et Abu Shama, est beaucoup plus prudent lorsqu’il dit que la puissance acquise par Gümüshtegin suscita la jalousie de certains émirs qui réussirent à convaincre al-Salih de la trahison de son émir.38 Selon Ibn al-‘Adim, le plus disert sur ces événements, d’autres personnalités haut placées de l’entourage d’al-Salih furent elles aussi visées peu de temps après. Si les Assassins ne purent atteindre Shadhbakht, trop bien gardé à l’intérieur de la citadelle, ils s’en prirent en revanche au tuteur du jeune souverain, l’eunuque al-Mujahid Yaqut, qui fut attaqué par trois d’entre eux mais réussit à leur échapper.39 On découvrit alors que deux de ses assaillants étaient des familiers de l’un de ses écuyers (rikabdar), ce qui témoigne de

sunni tes et chi i tes à a lep  | 205 leur art bien connu de l’infiltration. L’un d’entre eux fut tué puis crucifié. L’écuyer subit le même sort et un panneau lui fut attaché sur lequel il était écrit: ‘Voici la récompense de celui qui renforce les hérétiques’. Le second assassin capturé fut transporté à la citadelle où on lui arracha sous la torture les raisons de son action. Il dit à al-Salih n’avoir agi que sur son ordre et Sinan confirma ses aveux. Les ennemis de Gümüshtegin trouvèrent là l’occasion de se débarrasser de lui et firent valoir à al-Salih que Gümüshtegin ne cherchait qu’à l’écarter du pouvoir. Al-Salih le fit donc arrêter le 9 rabi‘ I 573/5 septembre 1177. L’émir, ayant refusé de rendre la ville de Harim dont il était le gouverneur, fut torturé sous ses murs puis étranglé. Son corps fut jeté dans le fossé.40 Les raisons de l’alliance de Gümüshtegin avec les Ismaïliens, certaine dans le cas des attaques contre Saladin et très probable dans le cas du vizir et de l’eunuque al-Mujahid Yaqut, furent par conséquent clairement politiques. La dimension religieuse tenait peu de place dans la stratégie de cet eunuque d’origine franque, converti à l’islam, qui utilisait les conflits entre chiites ismaïliens, chiites duodécimains et sunnites pour défendre ses propres intérêts. Le vizir Ibn al-‘Ajami l’avait bien compris, lui qui, dès sa nomination comme vizir, en 570/1174, avait des craintes face au gouvernement des trois eunuques pourtant sunnites comme lui.41 Al-Salih, soutenu par la population sunnite et duodécimaine d’Alep, eut quant à lui une politique clairement hostile aux ismaïliens. Dès 572/1176–7 (il n’avait alors que 14 ou 15 ans) il envoya une expédition combattre une branche de la secte réfugiée dans le Jabal Summaq, au sud-ouest d’Alep, après que certains se furent livrés, d’après nos auteurs sunnites, à toutes sortes d’actes outranciers (orgies, relations incestueuses, femmes habillées en hommes, etc.). L’un d’entre eux déclara même que Sinan était son Dieu (rabbihi).42 Toutefois Sinan les désavoua et fit même tuer leurs chefs, montrant ainsi qu’à l’intérieur même du mouvement il pouvait y avoir d’importantes dissensions.43 La même année dans la région d’al-Bab, au nord d’Alep, d’autres ismaïliens extrémistes furent exterminés par des sunnites adhérents de la futuwwa, appelés nabawiyya.44 Les ismaïliens se réfugièrent dans des grottes, mais y moururent enfumés. Leurs maisons furent pillées et leurs femmes humiliées.

206  |  syri a i n crusad e r time s Quelques années plus tard, en 575/1179–80, pour une raison inconnue, al-Salih enleva aux ismaïliens un village nommé Hajira, au sud-est d’Alep. Il fit la sourde oreille aux protestations de Sinan qui envoya alors des ismaïliens mettre le feu aux souks d’Alep. Les dégâts furent considérables entraînant la ruine de nombreux commerçants, mais aucun ismaïlien ne fut capturé.45 Conclusion Al-Salih tomba malade le 9 rajab 577/18 novembre 1181 et mourut le 25 rajab/4 décembre. Il fut enterré dans la citadelle avant d’être transféré par Saladin, en 579/1183, dans une khanqah, située à l’ouest de la citadelle, qu’avait fait construire sa mère.46 Il est décrit dans la plupart des sources comme un souverain aimé de son peuple, pieux, juste, chaste et sage malgré son jeune âge. L’historien Ibn al-‘Adim – dont on a déjà souligné l’attachement aux Zenguides – assure même, d’après le témoignage de son père, que non seulement les habitants d’Alep le pleurèrent chaudement mais qu’une odeur de musc se dégagea de son cercueil le jour de son transfert vers la khanqah, lui attribuant ainsi implicitement les attributs d’une quasi-sainteté.47 Sous son règne, les oppositions entre sunnites et chiites furent certes alimentées par des divergences religieuses, mais celles-ci ne furent pas les seules causes des conflits. Bien souvent les causes politiques et sociales furent prédominantes, en raison des luttes de pouvoir entre émirs (notamment entre anciens et nouveaux émirs) et entre grandes familles alépines. Au lendemain de la mort de Nur al-Din on assista ainsi à un affrontement entre partisans des Zenguides et partisans de Saladin, mais une fois les partisans de Saladin (Banu’l -Daya) neutralisés, sunnites et chiites s’unirent pour repousser Saladin et défendre la dynastie zenguide. Les rapports entre sunnites et chiites duodécimains n’ont donc pas tou­ jours été conflictuels. Al-Salih prit soin de sa population chiite, en lui accordant une grande liberté de culte et en développant les sanctuaires alides, là aussi pour des raisons essentiellement politiques, son appui étant indispensable pour résister à Saladin. La communauté chiite d’Alep continuait donc, à cette époque, de représenter une force sociale et politique non négligeable. Les successeurs zenguides d’al-Salih, ‘Izz al-Din Mas‘ud et son frère ‘Imad al-Din Zengi, poursuivirent cette politique tolérante à l’égard des chiites jusqu’à la prise de la ville par Saladin en 579/1183.48

sunni tes et chi i tes à a lep  | 207 Dans ce contexte, les rapports avec les ismaïliens occupèrent une place particulière car ces derniers étaient détestés par les duodécimains autant que par les sunnites. Quand al-Salih n’était encore qu’un enfant les ismaïliens trouvèrent un allié auprès de Gümüshtegin. Mais une fois débarrassé de la tutelle de Gümüshtegin, al-Salih eut une politique plus énergique à l’encontre des ismaïliens, ce qui entraîna leur réaction violente et l’incendie des souks d’Alep. Aucune conciliation ne pouvait être, dans ce cas, envisageable. Notes   1. Cahen, 1935, 258–269.   2. Elisséeff, 1967; Khayat, 1971, 167–195.   3. Hillenbrand, 2012, 163–179; Mulder, 2014, 66–68.   4. Sur ce sanctuaire, voir Tabbaa, 1997, 109–110; Mulder, 2014, 68–82; Eddé, 2018, 51.   5. Ibn al-Furat, Ta’rikh, ms Vienne Arabe 814 (AF 118), II, fol. 159r; Hillenbrand, 2012, 170–171.   6. Telles que l’interdiction de l’appel à la prière chiite et des malédictions lancées publiquement contre les premiers califes ou bien l’exil forcé de quelques notables chiites.   7. Eddé, 1999, 436–452.   8. Les principales sources sur le règne d’al-Salih Isma‘il sont: Ibn al-‘Adim, 1988, IV: 1822–1826; Ibn al-‘Adim, 1951–68, III: 9–42; Ibn Abi Tayyi’ dont des extraits sont conservés dans Abu Shama, éd. I. Shams al-Din, 2002 et dans Ibn al-Furat, éd. H. al-Shamma‘, IV, part. 1, 1967; al-Bundari, 1979. Cf. aussi Lyons et Jackson, 1982, 72, 78–80.   9. Ibn al-‘Adim, 1951–68, III: 11. Pour éviter tout problème, Shadhbakht avait tenu secrète la mort de Nur al-Din et avait fait venir les notables et les émirs pour leur demander de prêter serment à al-Salih que son père venait de faire circoncire et de désigner comme son fils héritier. 10. Sur ces événements, cf. Lyons et Jackson, 1982, 78–80, 86–95; Eddé, 2008, 88–95, trad. anglaise, 2011, 70–74. 11. Cf. Khayat, 1971, 191. 12. Sur cette puissante famille d’émirs, voir Elisséeff, 1967, index (Majd ad-Din, Sabiq ad-Din, Shams ad-Din et Badr ad-Din); Humphreys, 1977, 32–33; Eddé, 1999, 265–266. 13. Sur la famille des Banu’l-Khashshab à Alep, voir Eddé, 1999, 440–442.

208  |  syri a i n crusad e r time s 14. Cf. Ibn al-Furat, ms Vienne, Arabe 814 (AF 119), III, (an 552/1157), fol. 110r–111r; Ibn al-Qalanisi, 1908, 349; trad. partielle Gibb, 1932, 341–342; Ibn al-‘Adim, 1951–68, II: 308–310; Elisséeff, 1967, I: 518–519; Khayat, 1971 184–185. 15. Ainsi ‘Ali b. al-‘Ajami, membre d’une grande famille shafiite d’Alep, et la madrasa des Banu Bani ‘Asrun avaient souffert de leurs attaques. 16. Ibn al-Qalanisi, 1908, éd. p. 349, trad. Gibb, 1932, 341–342. 17. Sur les Banu’l-‘Ajami, cf. Eddé, 1991, 61–71. 18. Cette famille était sunnite comme le dit Ibn Abi Tayyi’ lui-même dans Ibn al-Furat, Ta’rikh, ms AF 117, I, 189v et II, 159v; Eddé, 1999, 370–371. 19. Ibn al-‘Adim, 1951–68, III: 16. 20. Ibid., III: 38, 48. 21. La maison du premier se trouvait dans le quartier des Verriers (al-Zajjajin), dans la partie sud d’Alep, et celle du second était à proximité, dans le quartier de la Cuve Jaune (Ğurn al-Asfar). À la suite de ces pillages, Ibn al-‘Ajami déménagea dans le quartier d’al-Balat (ou al-Ballat), à l’ouest de la citadelle, tandis qu’Amin al-Dawla s’installait sous la citadelle. Cf. Ibn al-‘Adim, 1988, V: 2431 et Ibn al-‘Adim, 1951–68, III: 16. Sur ces quartiers d’Alep, voir Gaube et Wirth, 1984, 99 (fig. 19), 105, 361 (n° 142). 22. C’est d’ailleurs à Majd al-Din b. al-Daya qu’Abu Shama attribuait cette phrase: ‘En fondant des madrasas, notre seul but a été répandre la science, de chasser les innovations (bida‘) hors de ce pays et de manifester au grand jour la foi religieuse’. Cf. Abu Shama, 2002, I: 117, trad., Recueil des Historiens des Croisades, Historiens orientaux, 1898, IV: 28. 23. Ibn al-‘Adim, 1951–68, III: 16. 24. Cf. Ibn al-‘Adim, 1988, IV: 1825. 25. Sabiq al-Din fut arrêté à Qinnasrin, alors qu’il rentrait de Damas, avant même d’arriver à Alep. Badr al-Din Hasan fut arrêté lorsqu’il sortit de la ville pour aller à la rencontre d’al-Salih. Shams al-Din ‘Ali, atteint de la goutte, fut amené devant al-Salih sur une litière, puis emprisonné comme ses frères. 26. Cité par Abu Shama, 2002, II: 217–218. 27. En particulier Qutb al-Din al-Hasan b. ‘Abd Allah (m. 588/1192), traditionniste et homme de lettres respecté, chargé des waqfs de la grande mosquée d’Alep. Ibn al-‘Adim, 1988, V: 2430–2431. 28. Une partie de l’or pillé dans la maison d’Ibn al-‘Ajami fut en effet récupérée. Quand le gouverneur d’Alep voulut le lui rendre, Qutb al-Din préféra l’envoyer à Ibn Amin al-Dawla pour l’affecter aux orphelins dont l’argent avait aussi été pillé.

sunni tes et chi i tes à a lep  | 209 29. C’était, entre autres, le cas de la famille des Banu’l-‘Adim installés à Alep dès le début du IXe siècle et des Banu’l-Khashshab établis à Alep depuis le règne de Sayf al-Dawla au Xe. Cf. Eddé, 1999, 362, 440. 30. Il fallut attendre le règne de Saladin pour que la famille chiite des Banu’lKhashshab retrouve une place importante à Alep. Majd al-Din Ibrahim (m. 589/1193), neveu du cadi assassiné, retrouva alors une fonction de cadi. Cf. al-Tabbakh, 1923, IV: 310 et al-Dhahabi, 1996, 320. 31. Ibn al-‘Adim, 1951–68, III: 29–30. Récit d’après son père comme il est dit dans Ibn al-‘Adim, 1988, IV: 1825. 32. La tradition sunnite était de prononcer quatre fois le takbir (prononciation de la formule Allah akbar, ‘Dieu est le plus grand’) sur la dépouille de tout musulman, la tradition chiite était de le prononcer cinq fois sur celle d’un simple fidèle et sept fois sur celle d’un sharif (pl. ashraf  ) descendant du Prophète. Cf. Ibn al‘Adim, 1951–68, III: 54; Ibn Abi Tayyi‘ dans Abu Shama, 2002, II: 228. 33. Membre d’une grande famille d’ashraf chiites d’Alep. 34. Abu Shama, 2002, II: 228. Sur b. Zuhra, cf. Ibn al-‘Adim, 1988, VI: 2946–2947 et Eddé, 1999, 438–439. 35. Cf. Eddé, 1986, 101–125; El-Azhari, 1997, 114–146. 36. Ibn Abi Tayyi’ les juge très sévèrement, que ce soit en Iran-Irak ou en Syrie (cf. par exemple son récit des événements de 502/1108 dans Ibn al-Furat, Ta’rikh, I: fol. 28r°–28v°, 152 v°–153 r°). 37. Toutes les sources sont d’accord pour penser que Gümüshtegin fut le commanditaire de la première tentative de meurtre. Cf. Ibn al-‘Adim, 1951–68, III: 21–22; Ibn al-Athir, 1965–7, 13 vols, XI: 419, tr. Richards, part 2, 2007, 234; Abu Shama, 2002, II: 228–231. La première tentative des Assassins contre Saladin ayant échoué, Gümüshtegin se tourna vers le comte de Tripoli pour obtenir qu’il détourne l’attention de Saladin d’Alep, ce qu’il fit en se dirigeant vers Homs. Quant au second attentat contre Saladin en 571/1176 devant ‘Azaz, Ibn Abi Tayyi’ le met en relation avec la peur des Alépins de se trouver dépossédés par Saladin, raison pour laquelle ils demandèrent, d’après lui, l’aide des Assassins (d’après Abu Shama, 2002, II: 269). Cf. aussi al-Bundari, 1979, 100–101. 38. Al-Bundari, 1979, 135; Ibn al-Athir, 1965–7, XI: 445, tr. Richards, 2007, 255; Ibn al-‘Adim, 1951–68, III: 32–34. 39. Ibn al-‘Adim, 1951–68, III: 33. 40. Óarim ne se rendit véritablement à al-Salih que l’année suivante, en 574/1178, après avoir subi un siège des Francs durant quatre mois. Cf. Ibn al-‘Adim, 1951–68, III: 32–38; Abu Shama, 2002, II: 308.

210  |  syri a i n crusad e r time s 41. Ibn al-‘Adim, 1951–68, III: 18; Eddé, 1991, 64–66. 42. Ibn al-‘Adim, 1951–68, III: 31; Lewis, 1966, 225–267, notamment 230, 239–242; Plusieurs mouvements dissidents chiites semblent s’être répandus à cette époque en Syrie du Nord. Ainsi en 571/1175–6, un homme prétendit être le prophète attendu (al-muntazar) c’est-à-dire le Mahdi pour les chiites, dans la région du Jabal Laylun, entre Alep et Antioche. Il parvint à séduire une partie de la population et prétendait repousser ses ennemis en leur jetant de la terre. Gümüshtegin prit la tête d’une armée pour le combattre: les hommes furent tués, les femmes enlevées. Certains se réfugièrent dans des grottes où ils furent soit capturés soit enfumés. Cf. Ibn al-Adim, 1951–68, III: 25–26. 43. Cf. Lewis, 1967, 111–112. 44. Ibn al-‘Adim, 1951–68, III: 32. Sur la futuwwa et sa branche nabawiyya, cf. Cahen, 1959, 25–56, 233–265, notamment 38, 49, 56, 238. 45. Ibn al-‘Adim, 1951–68, III: 38–39. 46. Cf. Eddé, 1999, 651, fig. 52. 47. Ibn al-‘Adim, 1951–68, III: 40–41 et Ibn al-‘Adim, 1988, IV: 1822–1826; Ibn al-Athir, 1965–7, XI: 472–473, tr. Richards, 2007, 277–278. 48. Ibn al-‘Adim, 1951–68, III: 54.

12 The War of Towers: Venice and Genoa at War in Crusader Syria, 1256–8 Thomas F. Madden

O

n an otherwise unremarkable day in 1256 a Genoese merchant, Barozio Malono, sailed his newly purchased vessel into the harbour of Acre and docked it at the wharfs. It caused quite a stir, for the Venetian residents of the city recognised the vessel as one that had recently been captured by pirates. The pirates had subsequently sold their prize to Malono. Unsurprisingly, the Venetians took some offence at Malono’s choice in business partners. Heated words were exchanged, and the disagreement soon became violent. A group of Venetians attacked Malono and took the vessel by force. News of the altercation quickly spread to the Genoese quarter, setting off a tinderbox of anger and resentment. Across the neighbourhood, men took up arms, while the leaders sent word to the Pisans in their quarter that, given an alliance recently sealed between their respective governments, they were bound to assist them. The Pisans complied. The Venetians did not have long to savour their victory at the harbour. Soon hundreds of Genoese and Pisans poured into the area, first retaking the disputed vessel, and then capturing all the Venetian ships docked there. The victors then rushed to the Venetian quarter, where they caused significant property damage before at last being ejected.1 Tense days followed, about which we know little, although it is certain that an uneasy truce was eventually concluded. According to the sixteenth-century Venetian secretary, ambassador and historian, Giovanni Giacomo Caroldo, who had access to materials now lost, it was the Venetian consul (bailo) of Acre, Nicolò Michiel,

212  |  syri a i n crusad e r time s who ‘knowing that from such quarrels could arise a major war, procured with his authority a truce, which was signed by the Venetians with pure motives, but by the Genoese with much artifice and dissimulation’.2 According to the contemporary Venetian chronicler, Martino da Canal, who frequently invented dialogues to enrich his narrative, some of the Genoese who seized the Venetian vessels returned to Genoa ‘with so much jubilation and great celebration as if they had conquered the whole world’.3 When they told their story to a local wise man, he responded, ‘Sirs, know for certain that never since Genoa was founded have goods been brought from across the sea that will be costlier for us. You have dishonoured Genoa!’4 As is so often the case, the wise man was correct, for this altercation was the spark that would set off the War of St Sabas, waged primarily in Acre and Tyre between 1256 and 1259. This war, fought completely between Christians in the twilight age of the Crusader states, is often cited as an example of the divisions that weakened them at a time when the Muslim power was ascending. And, certainly, that is correct. But given the place it is afforded in most histories of the Crusades, it is surprising that more attention has not been paid to it. The most complete treatment remains that of Georg Caro in his 1895 work Genua und die Mächte am Mittlemeer.5 More recently, Antonio Musarra has brought additional Genoese materials to bear on the event.6 One purpose of this essay is to employ additional Venetian sources for the event, while seeking to untangle, at least partially, the often-contradictory thicket of other sources. The other purpose is to contextualise the war within broader dynamics in Italy and the eastern Mediterranean. The War of St Sabas between Venice and Genoa is often said to have begun in Acre as a result of a long-disputed ownership of a monastery (St Sabas), which was on the border of their two quarters.7 Yet, as David Jacoby demonstrated, the dispute actually revolved around a house owned by the monastery of St Sabas that was situated at a strategic location overlooking Acre’s harbour in the royal ‘District of the Chain’. It was, therefore, a struggle over control of the harbour, not a neighbourhood border dispute.8 That quarrel dates to at least 1251, when two letters from the Genoese pope, Innocent IV, refer to the property in question.9 The first ordered the abbot and monks of St Sabas either to sell the property (domum quam apud Acon, in vico qui catena dicitur) at a just price to the Genoese of Acre, or to transfer

the wa r of towers  | 213 it to them with a ten-year renewable lease. The second letter, addressed to the Archbishop of Tyre, Nicolò Lercari, also Genoese, ordered him to oversee the matter, ensuring that the Greek monastery complied.10 Both letters threatened the abbot with excommunication should he refuse to obey. Apparently, the abbot did not obey, although there is no evidence of any subsequent repercussions under Innocent’s pontificate. On 22 April 1255, a few months after becoming pope, Alexander IV renewed the call for St Sabas to sell, rent or trade the property at the port to the Genoese. Alexander’s letter to the abbot, however, was much more conciliatory, asking rather than commanding the abbot to transfer the property. In another letter sent the same day, Alexander ordered the Prior of St Lawrence of Acre and the ‘cantori Tyrensi’ to oversee the matter. Here again, though, the Pope chose a much less rigid approach. Should the abbot refuse to let the property go, he stated, the prior could employ ‘ecclesiastical censure’ against him, yet only if the monastery would be unharmed in any way by the transfer (proviso quod monasterium ipsum dampnum aliquod non incurrat).11 Given that this letter to the abbot follows the form of Innocent IV’s previous letter, citing the devotion to Rome of the beloved sons of Genoa, this was likely Alexander’s attempt to tie up a loose end from the previous pontificate, yet without the zeal for the matter of his Genoese predecessor. Over the next several months, though, Alexander learned, probably from the Venetians, that the transfer to the Genoese would cause grave harm to the monastery. In a letter of 22 July 1255 to the Prior of the Hospital and the Venetian parish priest of San Marco at Acre, Alexander nullified his previous letter and forbade the transfer to take place.12 According to the Venetian sources, in 1256, probably before the violence broke out at the harbour, contradictory letters regarding the property were in circulation in Acre. The newly arrived Venetian consul, Marco Giustiniani, presented to the Patriarch of Jerusalem a letter stating that the house in question should at once be given to the Venetians. The other letter, presented by the Genoese to the Master of the Hospitallers, said just the opposite, that the property should be rendered to them.13 According to Andrea Dandolo and Marino Sanudo, both writing in the fourteenth century, the letter favouring the Venetians came from the Pope. According to Dandolo, the Genoese letter, on the other hand, was ‘similar’, thus leaving open the question as to whether it too came from Rome.14 With the agreement of Marino Sanudo and Andrea

214  |  syri a i n crusad e r time s Dandolo on the matter, this became the standard narrative in Venice and beyond. However, Caroldo did not accept it. He reported instead that the papal letters enjoined peace, and commanded that the disputed property should be held in common by both parties. The letter that Marco Giustiniani brought to Acre, Caroldo maintained, was from the doge, asking the Patriarch in Acre to help the Venetians acquire the disputed property. Likewise, the Genoese letter to the Hospitallers was from their communal government, making a similar request.15 Caroldo’s description of the competing letters to the patriarch and the Hospitallers seems plausible, since it obviates the need for contradictory missives written by the same pope and sent to the same city.16 But the papal call for peace is not confirmed by any other sources, and seems unlikely given that war had not yet broken out. What is certain is that both sides were seeking support for their opposing claims. After the tumult at the harbour there remained significant street violence between Venetians and Genoese, both in Acre and Tyre.17 This was not simply a product of the decline of the Latin East. Venetians and Genoese had been attacking each other’s quarters in Constantinople since 1171, and preying upon each other’s shipping well before that. The two states had fought a costly war over Crete between 1206 and 1211, which ultimately led to Venetian sovereignty over the island. And since Venice at this time controlled three-eighths of Constantinople, the Genoese were largely barred from trading there as well. The Venetians and Genoese were fierce competitors, freely spilling each other’s blood across the eastern Mediterranean. What made Acre and Tyre unusual is that the Italians there held their quarters as largely exempt states-within-states, where they enjoyed their own law courts, taxes and militia. Venice had first acquired this with the Pactum Warmundi of 1123, which handed over one-third of Tyre to them, as well as a quarter in all other cities of the Kingdom of Jerusalem.18 By the mid-thirteenth century Acre and Tyre were a patchwork of competing authorities, directly answerable to their home governments, and with little interest in the king or his representatives.19 Nowhere is this more evident than in the skyline of thirteenth-century Acre. Most sources for this conflict describe how the massive towers of the Italian quarters play a central role. Indeed, were this war not so firmly associated with the tangential St Sabas, it would much more accurately be termed

the wa r of towers  | 215 the War of Towers. As in most medieval Italian towns, towers were not just fortified structures designed to defend the quarter and attack its neighbours, but a visible sign of the power of their owners in a city in which power was always in flux. Just as families and factions built towers in medieval Italian cities, so their national groups built them in faraway Syria. The Genoese tower, for example, was especially famed. Da Canal, no friend of the Genoese, described it as ‘the most beautiful and most defensible in all the world’.20 According to Caroldo, the Genoese tower at Acre cost 200,000 bezants to construct and stood 200 feet high with a circumference of 200 feet.21 In Acre, the Italian towers were also a symptom of a weak royal government, that ceded extraordinary rights to these groups, and thereby could not prevent them from fortifying their autonomous quarters. This was not so elsewhere. Although warfare occasionally erupted between Constantinople’s Italian quarters, it was quickly put down by the Byzantine authorities and penalties exacted. No tower could have arisen in those carefully controlled areas. Only in the fourteenth century, when the Byzantine Empire was greatly weakened, did the Genoese erect their tower in Galata. John of Ibelin, nominally in charge in Acre, was unable to quell the daily violence. If, as seems likely, the Pisans continued to support the Genoese, then the Venetians must have had the worst of it. This may have played a role in the subsequent decision of Philip of Montfort to expel all the Venetians from Tyre and confiscate their quarter in late 1256.22 Philip had been the bailiff or regent of Tyre since 1246 and was a long-time supporter of the Genoese. Given the number of surviving archival documents and the report of Marsilio Zorzi, a Venetian consul in Acre, the Venetian community in Tyre must have been large.23 According to Andrea Dandolo, the Venetians had planned to raise troops in Tyre for the war they believed was looming in Acre.24 After their expulsion, many of the Venetian refugees undoubtedly went to Acre, but many others likely sailed to Constantinople, Crete or Venice. The expulsion from Tyre, one of Venice’s oldest and most profitable foreign quarters, was naturally seen as an act of war. According to Caroldo, Marco Giustinian sent an armed fleet from Acre to Venice with news of the expulsion. There appears to have been no doubt on the Venetians’ part that the Genoese were behind the action.25 The doge sent an embassy to Genoa to discuss peace, but it returned empty-handed. Shortly thereafter the Genoese

216  |  syri a i n crusad e r time s sent their own envoys to Venice, but talks stalled on the question of Genoese reparations for the damage done to Venetian ships and homes in Acre.26 The expulsion of the Venetians from Tyre transformed the simmering civil unrest in Acre into a full-scale war between Venice and Genoa. Given the number of towers in Acre, and the stone-hurling devices mounted on them, there was no doubt that it would be a destructive and bloody war. Both sides stepped up preparations, seeking additional allies wherever they could find them. Venetian successes in that would ultimately decide the war. Two Venetian envoys, Pietro Barossi and Giovanni Ferro, were sent to Modena, where they negotiated a peace with their Pisan counterparts. Their efforts culminated on 18 July 1257 with a still-extant treaty between the two powers. It bound both parties to a ten-year alliance against Genoa and specifically nullified Pisa’s previous treaty with the Genoese. In a reference to the situation in the East, the treaty declared that neither side was bound to send forces against the enemies of the other, yet if present they were required to aid the other against any Genoese attacks. Throughout the duration of the alliance all Venetian and Pisan vessels were to fly both flags.27 Beginning with Caro, scholars have traditionally seen this as a poor deal for Pisa. Although they were not bound to send anyone to fight the Genoese in Acre, the Pisans already had a large community there that would be forced to switch sides in the war, turning against their former allies. However, there were other elements to consider. The previous treaty between Pisa and Genoa was less than a year old, and it had been entered into reluctantly. Pisa had been forced into an unfavourable peace by a Tuscan confederation when their war in support of Conrad of Hohenstaufen went badly. (Conrad had died in 1254.) A Pisan official in Cagliari, Sardinia, took advantage of the situation by declaring himself independent. When the Pisans attacked him, he requested aid from the Genoese, who were happy to gain a foothold on the island. In the end that is all they got before Alexander IV put a stop to the war and forced Pisa and Genoa into the treaty.28 The new alliance between Pisa and Venice was clearly aimed at prosecuting a war against Genoa in Acre. Indeed, the doge’s agents must have made the Venetian plans in that regard very clear. By switching sides, Pisa would force Genoa to shift resources and men away from Sardinia to defend its lucrative position in the East. This, however, would not come without cost to the Pisans. According to Da Canal,

the wa r of towers  | 217 the Genoese in Acre responded to the news of the betrayal by invading the Pisan quarter, capturing their tower and the harbour, which they closed off with a chain.29 According to Caroldo, the Genoese simply held the Pisan tower that they had occupied when their alliance was still in effect.30 In August 1257 a Venetian fleet of fourteen galleys (or perhaps thirteen galleys and one naves) sailed out of Venice bound for Acre.31 Its captain was Lorenzo Tiepolo, who would later become doge (1268–75). The arrival of the war fleet at Acre dramatically changed the situation in Acre. Casting anchor outside the harbour, Tiepolo came ashore to discuss the situation with the consul, Marco Giustiniani. With the Pisans now on their side, they agreed that it was time to act. The alliance and the fleet also galvanised the attention of Philip of Montfort, who came to Acre and offered to host a conference at a house he owned there to avoid war. Lorenzo Tiepolo and the Genoese consul, Leo Grimaldi, attended. The contemporary ‘Templar of Tyre’, whose information for events in Acre is detailed and usually accurate, provides an account of the meeting. The Teutonic Knights began by stating their desire to broker an agreement that would be favourable to all. Tiepolo, who had no reason to trust Philip of Montfort or the Genoese, began by proclaiming that he would not depart from Acre until he could take one of the stones from the foundation of the Genoese tower to Venice … Grimaldi put his hand to his sword and drew it and ran at Sir Lorenzo, and they came very near killing each other, but the lords of Acre and the Templars and Hospitallers separated them. And from that moment the war between them, which was most bloody, commenced.32

A few days later Tiepolo’s fleet broke the chain, occupied the port and destroyed all the Genoese vessels there.33 According to Da Canal and Dandolo, the Genoese had built a fortification on the roof of the building in dispute. The Venetians promptly burned it to the ground. As Jacoby has shown, it was the strategic location, not the building itself, that made it so valuable. Indeed, after the war the Venetians paid 7,000 bezants for the property and ruins – more than five times the average price for a palace in the city.34 The Venetians and Pisans then invaded the Genoese quarter up to Montmusard before concluding a two-month truce.35

218  |  syri a i n crusad e r time s The truce, however, did nothing to calm passions. When it had expired the War of St Sabas became a highly destructive urban struggle. The towers of the city were filled with archers and topped with siege machines that hurled heavy stones across the urban landscape. According to the ‘Templar of Tyre’, these machines threw rocks day and night ‘and they did a lot of harm to each other, and knocked down a number of houses’.36 Each of the engines had its own name. One was called Boverel, which could mean ‘stables’ or could refer to the neighbourhood of Acre by that name.37 Others were called Vincheguerre, Peretin and Marquemose. Reportedly, one of the engines could hurl stones weighing more than 280 pounds.38 The Master of the Temple, Thomas Bérard, even moved his household to the order of St Lazarus ‘in order to get away from the battle of the siege engines, which were firing at each other, for the Templar establishment was very near the Pisan quarter’.39 The Annales Ianuenses claimed that there were more than fifty stone throwers in the city.40 Although they had a large tower and many siege engines, the Genoese were in a difficult position. They had only 800 armed men to hold their quarter and could not venture much beyond it. The Venetians allowed no Genoese vessels into the harbour, hoping to starve them out. But Philip of Montfort, who obviously did not want to see the Venetians win in Acre, sent provisions which were smuggled through the Hospitallers’ port in Acre.41 With various breaks for other events, this urban war continued for the next fourteen months. As the Pisans hoped, the Genoese sent additional warships from home to Tyre to organise an attack on Acre. Led by the Genoese captain Pasquetto Malono, the fleet would frequently sail to Acre, but when the Venetians boarded their vessels and prepared to repel it, they retreated.42 At some point in late 1257 Tiepolo brought the Venetian fleet to Tyre and engaged the Genoese. It was a resounding victory for the Venetians. Tiepolo captured Malono himself, three galleys and 300 Genoese.43 According to Da Canal: The Genoese who were in Acre were so sad and so angered when they saw the arrival of their relatives, friends and neighbours who had been captured, as I have told you, that they believed they would never be happy again. And to tell you the truth, they have not been happy since the time that Mesire Lorenzo defeated them at Tyre.44

the wa r of towers  | 219 The Venetians and Genoese sent word home to send more galleys. For some time John of Ibelin and the Master of the Temple had been seeking a way to end the war. They convinced Bohemond VI of Antioch and Tripoli to come to Acre with his sister, Plaisance of Cyprus, the regent for her infant son, Hugh II, who was regent of Jerusalem for the six-year-old Conradin of Hohenstaufen. John and the Master clearly believed that a strong hand was needed to stop the chaos. They convinced Plaisance to ask the barons of the kingdom to accept Bohemond as the regent for her son, which would allow him to use his position and power to bring the two sides to terms. But the Genoese did not trust Bohemond. Although they had helped with the original conquest of Tripoli, their properties and rights had been stripped from them there many years earlier. According to the ‘Templar of Tyre’, Bohemond feared that if the Genoese should emerge victorious they could force that issue, winning back their position in Tripoli. In other words, Bohemond’s attitude toward the Genoese was very like Philip’s toward the Venetians. The Genoese urged Philip and their other supporters among the barons to reject Plaisance’s proposal, saying that she did not have the authority to shift the regency, and professing loyalty only to far-away Conradin. Plaisance, who had attempted some neutrality, responded to this challenge by backing Venice and Pisa and ordering all the barons to do the same. Genoa was left with only Philip of Montfort and the Hospitallers.45 The last act in this drama came later that year. Ten additional warships were sent from Venice under the command of Paolo Falier and fifteen from Crete captained by Andrea Zeno.46 The Genoese sent some twenty-five vessels from Genoa and Sardinia under the command of Rosso della Turca, thus bringing both sides in the east to around forty galleys.47 According to Da Canal, when news of the arrival of the Genoese fleet came to Acre, the Genoese celebrated loudly from their tower and hung the banners of their allies down its sides. They shouted to their enemies, ‘Flee, flee from the city, for you are all dead! Behold, the flower of Christianity has arrived! Tomorrow you will all be killed, on the sea and on the land!’48 Genoa’s remaining allies shared this enthusiasm. Philip of Montfort brought an army of some eighty horsemen and thirty archers outside Acre, which was supplemented by a force of Hospitallers. Meanwhile the Genoese fleet sailed to Acre and cast anchor before the chained harbour. According

220  |  syri a i n crusad e r time s to Da Canal and the ‘Templar of Tyre’, the Genoese fleet outnumbered the Venetian. The latter described it as forty-eight galleys and four nefs, each with a siege engine. The plan was for the Genoese to secure the harbour while the land forces overran the Venetian and Genoese quarters. In the judgement of the ‘Templar of Tyre’, had the Genoese run the chain and attacked immediately they would have won the day.49 But they may have feared the many siege engines in the city, as well as an attack from the Tower of Flies at the harbour entrance, which the Venetians controlled. The Venetians delayed their attack, but it is not certain why. According to the ‘Templar of Tyre’, they feared that if they manned their vessels the Genoese would storm their quarters. They therefore asked Thomas Bérard, the Master of the Temple, to provide brothers and turcopoles to protect them in their absence. He agreed. The brothers on armoured horses with banners flying were so imposing that the Genoese in Acre fled to their quarter, fearing they would have to defend it.50 According to Da Canal, the Venetians were not worried, for they were ‘strong as lions’.51 They also had a lot of money. According to the ‘Templar of Tyre’, the Venetian commander offered ten Saracen bezants for anyone willing to serve in the fleet. There were so many takers that they filled not only their forty galleys, but an additional seventy vessels, many with archers.52 They quickly decided that the consul, Giustiniani, would see to the defence of the Venetian and Pisan ­quarters, while Zeno and Tiepolo would command the fleet. ‘But the wind blew so fiercely that the Venetians were unable to reach them in order to suppress their arrogance that day.’53 Taking this as cowardice, the Genoese lit torches and yelled insults and threats all night to the Venetians on the other side of the chain. The next morning after Mass and a small breakfast, the chain was lowered and the Venetian galleys rowed out.54 At the end of the day the Venetians decisively defeated the Genoese. The ‘Templar of Tyre’ put the number of Genoese killed or captured at 1,700, which he based on a roll call in Tyre. They also lost twenty-four galleys.55 Da Canal counted twenty-five galleys, with 666 captured, 400 drowned and 600 ‘hacked to pieces’.56 Accounting for the poetic licence of the first number, the figures are the same. Philip of Montfort and the Hospitallers returned home. According to Da Canal, Philip exclaimed, ‘The Genoese are men who only

the wa r of towers  | 221 brag, and are like seagulls who eat fish, hurling themselves into the water and drowning there! Their arrogance is crushed!’57 The war of towers, which had brought such ruin to Acre, ended at the great Genoese tower. Seeing their fleet destroyed the Genoese removed their banners and fled to the refuge of the Hospitallers. They were subsequently allowed to move permanently to Tyre. Lorenzo Tiepolo then ordered the Genoese tower completely demolished, so that, as Da Canal reports, ‘not one stone was left upon another’.58 As he had promised, the ‘Templar of Tyre’ tells us, Tiepolo took a stone from the foundation back to Venice with him. The rest of the stones were used by the Venetians and Pisans to build walls around their respective quarters.59 The War of St Sabas had the immediate effect of separating the Genoese and Venetians, with the former settling in Tyre and the latter in Acre. But the war between the two would continue throughout the eastern Mediterranean and Black Sea unabated. Acre was only one facet of that larger struggle. It served as a prominent battlefield because it lacked a government with the authority, means and will to impose order. The result was a geography of devastation at the heart of the Latin East. As the Rothelin continuation of William of Tyre reported ‘nearly all the towers and the fortified houses of Acre were completely destroyed, with the exception of religious houses, and there were killed in that war 20,000 men on one side or the other’.60 Notes   1. The Malono episode is described in the continuation of Caffaro’s Annales Ianuenses, MGH, SS 19:238. According to Andrea Dandolo, the Genoese and Pisans pressed into the Venetian Quarter as far as the church of San Marco: Dandolo, 1938, 12: 307. Given the destruction of Acre after 1291, much of the topography of the Crusader city is speculative. Nonetheless, Giovanni Giacomo Caroldo claimed to have seen the church in Acre in 1495: Caroldo, 2008–11, 2: 32.  2. ‘… conoscendo che da tali differenze si potrebbe accender maggior guerra, procacciò, con ogni suo potere, una tregua, la qual fù fermata et da Venetiani puramente servata, ma da Genovesi con molto artificio et simulation’: Caroldo, 2008–11, 2: 33. Caroldo’s history is seldom used for medieval events, but it is a generally sober source drawing almost exclusively from Venetian records and early chronicles.

222  |  syri a i n crusad e r time s  3. ‘… et fesoient li Jenoés si grant joie et si grant feste con s’il eüssent vencu trestos li mondes …’: da Canal, 1972, 156, 158.  4. ‘Seignors, sachés certainement que vos avés aportés tes merchandise dela la mer, que, puisque Jene fu faite, ne fu aportee nule merchandie que plus en fust perdux con se perdra de ceste. Vos avés honis Jene! ’ Ibid., 158.   5. Caro, 1895, 1: 28–43.   6. Musarra, 2017, 443.  7. See, for example, Runciman 1962, 2: 568; Richard, 1999, 389; Goldsmith, 2006, 1059.   8. Jacoby, 1977, 227–228; Jacoby, 2003, 2: 246.  9. I Libri iurium della Repubblica di Genova, 1998, 1: 69–72. The strategic location is also referred to in the Rothelin Continuation of William of Tyre as ‘une maison qui seoit seur le mer entre la terre des Venicienz et des Genevoiz.’: RHC, Occ., 1859, 2, 634. 10. Liber iurium reipublicae genuensis, 1854, 1, coll. 1097–1099, nos 818–819. 11. Les registres d’Alexandre IV, 1895, 1: 116–117, nos 390–391. 12. Ibid., 185, no. 606. 13. Torsello, 1611, 220. 14. ‘Eodem anno, Marcus Iustiniano venit Achon, consul Venetorum, et presentavit literas pape patriarche, ut Venetos poneret in possesione sancti Sabe; Ianuenses autem portaverunt similes literas priori Hospitalis pro hac causa; et ob hoc cepit discordia inter eos.’: Dandolo, 1938, 307. 15. Caroldo, 2008–11, 2: 32–33. 16. Musarra plausibly suggests that the Genoese letter could have been the original from Innocent IV, subsequently nullified by Alexander IV: Musarra, 2017, 444. 17. See ‘Templar of Tyre’, 1906, 2: 742–743, no. 269. 18. Urkunden zur älteren Handels- und Staatsgeschichte der Republik Venedig, 1856–7 (henceforth, T. Th.), 1: 78, no. 38. 19. Caroldo, 2008–11, 2: 32. 20. ‘… la plus bele et la plus defensable que le’en seüst en tos li mondes’: da Canal, 1972, 170. 21. Caroldo, 2008–11, 2: 32. Caroldo reckoned the tower as forty passi high and round. The passo was approximately five feet. Krüger, 1830, 228–229. This compares favourably with other contemporaneous Italian towers. The Tower of Pisa, under construction at this time, is slightly shorter and with a circumference at the base of 160 feet. The Torazzo in Cremona, completed in 1309, is nearly 350 feet tall. The campanile of San Marco in Venice was 323 feet tall.

the wa r of towers  | 223 22. Dandolo, 1938, 307; Caroldo, 2008–11, 2: 33. The last Venetian document from Tyre is dated September 1256, the will of Andrea Dolfin the bailo, Archivio di Stato di Venezia, Procuratori di San Marco, Misti, b. 120. 23. Marino Zorzi in T. Th., 1856–7, 2: 351–389, no. 299; Jacoby, 2003, 243–245. 24. Dandolo, 1938, 307. 25. Dandolo reports that Philip of Montfort’s expulsion order against the Venetians was ‘ad Ianuensium requisicionem’, Ibid. 26. Caroldo, 2008–11, 2: 33. 27. Müller, 1879, 447–450; Dandolo, 1938, 307; L’Estoire de Eraclès empereur et la conquest de la terre d’Outremer, in RHC, Occ. 2: 443; Caroldo, 2008–11, 2: 34. 28. Pryor, 1999, 5: 435. 29. Da Canal, 1972, 160. 30. Caroldo, 2008–11, 3: 34. 31. ‘Templar of Tyre’, 1906, 2: 742, no. 268; L’Estoire de Eraclès, RHC, Occ. 2: 443; Dandolo, 1938, 308; da Canal, 1972, 160. 32. ‘Templar of Tyre’, 1906, 2: 743, no. 270; tr. Crawford as The ‘Templar of Tyre’, 2003, 24. 33. According to Dandolo the burned vessels numbered twenty-eight naves and two galleys: 1938, 308. Da Canal puts the number at thirty naves and two galleys: 1972, 160. Caroldo explains the discrepancy, saying that Tiepolo salvaged two vessels, keeping them for his fleet: 2008–11, 2: 35. 34. Jacoby, 2003, 247. 35. Dandolo, 1938, 308; da Canal, 1972, 160. 36. Crawford, tr. Templar, 2003, 24. 37. On the district, see Jacoby, 2005, 73–105, esp. 77. 38. ‘Templar of Tyre’,’ 1906, 2: 743, no. 270; Crawford, Templar, 2003, 24, n. 4. 39. ‘Templar of Tyre’, 1906, 2: 746, no. 281; Crawford, Templar, 2003, 27. 40. Carrafo, Annales, 239. 41. ‘Templar of Tyre’, 1906, 2: 745, nos 274–276. 42. Da Canal, 1972, 162; Dandolo, 1938, 308. 43. Da Canal, 1972, 162; Torsello, 1611, 220. 44. Da Canal, 1972, 162; tr. Morreale as Martin da Canal, 2009, 62. 45. ‘Templar of Tyre’, 1906, 2: 743, no. 271; Dandolo, 1938, 308; Richard, 1999, 390. 46. Da Canal, 1972, 166, 168. 47. Ibid., 168; Dandolo, 1938, 309; Caroldo, 2008–11, 2: 36–37.

224  |  syri a i n crusad e r time s 48. ‘Fuiés, fuiés de la vile, que vos estes mors, que ja s’en vient la flor de Crestientés! A l’endemain serés trestuit ocis, que en mer que en terre!’: da Canal, 1972, 170. 49. ‘Templar of Tyre’, 1906, 2: 745–46, no. 281; da Canal, 1972, 170. 50. ‘Templar of Tyre’, 1906, 2: 745–46, nos 281–282. 51. ‘hardis come lions’, da Canal, 1972, 170. 52. ‘Templar of Tyre’, 1906, 2: 746, no. 284. 53. Da Canal, 1972, 170; Morreale, Martin da Canal, 2009, 65. Conversely, Caroldo claimed that the Venetian fleet was becalmed: 2008–11, 2: 37. 54. Da Canal, 1972, 170, 174. 55. ‘Templar of Tyre’, 1906, 2: 746, no. 284. 56. Da Canal, 1972, 174. 57. ‘Les Jenoés ne sont pas homes se de borde non et resenblent osiaus que manjuent poison, que il se jetent en mer et se neient dedens: lor orgueill est abatus!’: Ibid., 172. 58. Ibid. 59. ‘Templar of Tyre’, 1906, 2: 747, no. 286. 60. ‘Don’t il avint que prez que toutes les torz et les forz maissonz d’Acre furent toutes abatues forz que tant seulement les maissonz de religion, et I ot bien morz de’ cele guerre .xx. .m. homes que d’une part que d’autre.’: Rothelin Continuation, RHC, Occ., 1859, 2: 634.

13 Gaza in the Frankish and Ayyubid Periods: the Run-up to 1260 ce Reuven Amitai

T

his present chapter is part of an ongoing larger study on the history of Gaza and the surrounding countryside in the late Middle Ages, with an emphasis on the Mamluk period.1 My interest in late medieval Gaza was piqued while leafing through the then recently published Volume 4 of Moshe Sharon’s Corpus Inscriptionum Arabicarum Palestinae (CIAP),2 which is largely devoted to Gaza City. I was struck by the large number of Mamluk-era inscriptions, almost seventy, which is similar to the quantity that we find in Jerusalem from the same period. The corpus of Mamluk epigraphy of Gaza is a little different from that of Jerusalem, mainly in the fact that many of the inscriptions in the former are devoted to a few important projects (particularly thirteen in the Great Mosque, originally a Crusader church), while in Jerusalem, the inscriptions are generally each devoted to one construction project.3 This large corpus of extant Mamluk inscriptions in Gaza, early on the provincial capital for south-west Palestine, along with much of the coast, is clear evidence of large-scale patronage in the city by the military–­political elite, hinting at not insignificant cultural and intellectual activity there. This impression has been confirmed by my subsequent researches into the quarter millennium of the city’s history under the Mamluks, showing also demographic and economic prosperity throughout this period. This view has been further strengthened by the examination of the history of the city’s agricultural hinterland, where a few interesting Mamluk (and Ayyubid; see below) inscriptions have been found, and we can also postulate a thriving rural sector.4 In short, Gaza and its environs under the Mamluks seem to have been a vibrant place in many respects.

226  |  syri a i n crusad e r time s While some important preliminary studies on Mamluk-era Gaza and its immediate hinterland have been published, notably by ‘Ata Allah and Sadek,5 we lack a comprehensive and up-to-date monographic study. It is thus with some relish that I have embarked on my own study of Mamlukera Gaza. To this end, one must look back over the earlier centuries to the region’s history, in order to provide background, to examine long-term trends and to have a point of reference for the Mamluk era. For the time since the advent of Muslim rule in the 630s up to the end of the eleventh century, we have a long section of the still useful classic work by ‘Arif al‘Arif’s from 1943, although this surely needs to be updated and expanded.6 For the early decades of the twelfth century, there are just brief mentions (see below). However, matters improve for the almost four and a half decades of Frankish rule (from 1149 to 1193, with a brief hiatus towards the end), as we have some information about construction in the city of Gaza, presented and analysed by Meron Benvenisti7 and Denys Pringle,8 both adding concise historical backgrounds; these are very good starts. Nothing comparable is found in the research literature for the time of Ayyubid rule (1193–1260), although the CIAP contains a few rural and urban inscriptions. Of course, Gaza frequently appears in passing in the many modern histories of the Crusades9 and the few about the Ayyubids,10 but from these one does not get a coherent picture of the city and its surroundings during these times. The lack of a benchmark for later developments in the rural hinterland is especially acute for the Frankish period. This is clearly seen from a quick look at a couple of modern maps. The first is taken from the 1956 Atlas of Israel, which contains a significant historical section.11 Not surprisingly, this volume has a very detailed sheet devoted to the Crusaders, prepared under the direction of the late Joshua Prawer and his then assistant Meron Benvenisti. As can be easily seen in Map 1, there is not much in the immediate region of Gaza, with the exception of the city itself and Darom (Darum, today Dayr al-Balah), about 15km to its south. Look at the section of the map to the more distant north and east (Map 2): lots of settlements, big and small, are marked. These include six settlements that can be considered to the very north of Gaza’s countryside in the Mamluk period, in the area of Wadi al-Hasi, but were actually part of the seigneury of Ascalon in the time

gaz a in t he f ra nk i sh a nd a yyubid pe r io d s  | 227 of Frankish occupation.12 We have no specific information of anything closer to Gaza City in the time of its occupation by the Franks. Some forty years later Ronnie Ellenblum published a map of the cities, towns and villages in his pioneering book on rural settlement in Palestine under Frankish rule (Map 3).13 Ellenblum’s attention in that study was drawn mainly to the area north of Jerusalem and the Galilee, and thus it should not surprise us to see that the region of Gaza still appears to be sparsely settled; in fact, he omitted some of the villages noted in the Atlas of Israel. I am, of course, not suggesting on the basis of the maps in these two volumes that the immediate rural hinterland of Gaza did not contain villages during the time of Frankish rule, but that rather we have yet to find information about them. This makes a useful comparison with the Ayyubid–Mamluk periods, for which we have the names of some dozen settlements, impossible at this time. Our knowledge of the city of Gaza in the Ayyubid period is generally not much better. Gaza and its environs come up frequently in the sources, mostly as a stopover for armies moving between Egypt and Syria, and as the occasional battlefield. Meron Benvenisti has written regarding the period after Frankish rule: ‘The city was restored by the Moslems and became an administrative, military and commercial centre.’14 This seems true, but it is very general, and perhaps refers more to the Mamluk than the Ayyubid period. This present chapter – on the immediate pre-Mamluk history of Gaza and its environs, that is, for the Frankish and Ayyubid periods – thus comes to fill in some of this lacuna in modern scholarly literature. It is intended as a prelude to the study of the history of Mamluk Gaza, but also has some significance in its own right, as the city and its environs were an important frontier area over more than a century and a half, from about 1100 to 1260. The last year, by the way, is significant for Gaza and the larger region: it is the end of formal Ayyubid rule, there was a short Mongol occupation, and it is the commencement of a new era, when Gaza becomes part of the expanding Mamluk state. From the middle of the second millennium BCE until the time of late Roman (or Byzantine) rule, Gaza was significant as a centre of regional and even international trade. In late Roman times, Gaza flourished economically and

228  |  syri a i n crusad e r time s was also an important cultural centre, famous for its school of rhetoric. By the end of the fifth century, Christianity was clearly the leading religion of the city and region, but there was also a significant Jewish community. The agricultural richness of the area – wheat, grapes and other fruits – p ­ resages the 15 Mamluk period. In the Muslim tradition, pre-Islamic Mecca had conducted trade with Roman Gaza: the great-grandfather of the Prophet, Hashim b. ‘Abd al-Manaf, is said to have been buried there, and the future caliph ‘Umar b. al-Khattab (r. 634–44) ‘grew rich at the time of the Jahiliyya, for this place was a highway (mustatriq) for the people of the Hijaz’.16 After the Muslim Arab occupation of the mid-630s, it is only rarely mentioned in the Arabic sources, perhaps reflecting declining fortunes and prominence; it seems to have suffered from fighting among Arab tribes at the end of the second hijri century (that is, early in the ninth century CE).17 While Gaza and the surrounding area appear to have become heavily Muslim relatively early in the Muslim period, a substantial Christian community continued, and it was the seat of a bishopric (as was nearby Ascalon). A Jewish community still existed and there were Samaritans too.18 The Arabic geographers writing in the later tenth century describe a city of some grandeur (or at least size). Thus, al-Muqaddasi (d. 991), a native of Jerusalem writes: Gaza is a large town lying on the main road (jadda) into Egypt, on the border of the desert. The city stands not far from the sea. In it is a beautiful mosque, and there ‘Umar b. al-Khattab grew rich. [Further, this city was] the birthplace of al-Shafi‘i, and [there is] the tomb of Hashim b. ‘Abd Manaf.19

It may be that this and the parallel passages in other contemporary geographical works reflect an earlier period. On the other hand, it may be that Gaza City and its countryside – in spite of not infrequent fighting in the area and the recurrent movement of troops – enjoyed some prosperity. Be that as it may, the Fatimid conquest of the country from about 970 was not a smooth process, having to deal with the Qarmatis on the one hand and local Bedouin tribal groups on the other.20 Things apparently got worse with the influx of Turcomans into the country in the 1070s, more-or-less under Seljuq authority: around 1076, there is a report that the Seljuq chieftain Atsiz massacred all the population of Gaza, after putting down a rebellion in Jerusalem.21 It is doubtful that indeed everyone was killed (and it is unclear how this was

gaz a in t he f ra nk i sh a nd a yyubid pe r io d s  | 229 connected to the trouble in Jerusalem), but surely there was much disruption at this time, extensive damage, and many lost their lives. The coming of the Crusaders to the country at the end of the eleventh century probably did not lead to an improvement, especially since over the decades there were to be much fighting and movement of troops in the area as the Franks tried to take Ascalon, still controlled by the Fatimids. Meron Benvenisti correctly described Gaza as a key link in the ongoing Fatimid effort to keep supplying Ascalon.22 These were not the circumstances for a thriving city and countryside. Yet, we can assume that at least some of the city’s inhabitants as well as many of the local peasants remained. A modicum of regular agriculture must have been maintained in the area in spite of the unsettled conditions.23 This would belie the evidence of William of Tyre (d. 1186), who describes that in 1149, the Franks took a city that was ‘in ruins and entirely uninhabited’.24 Benvenisti also had his doubts, since within a few years the city was populated and these were probably not all new inhabitants.25 We might assume that indeed that the inhabitants got out of the way before the entrance of the conquering Franks, but many soon returned when things calmed down. The population was surely augmented by newcomers from various communities – Franks, Arabic-speaking Christians and perhaps even Muslims. The later Arab bureaucrat and writer Ibn Shaddad al-Halabi (d. 1285) notes that in the year AH 544 (1149–50) the Franks ‘rebuilt (‘ammaru, implying also repopulation) Gaza near [Ascalon], settling people in it, and providing foot soldiers and horsemen for it’.26 After taking Gaza in 1149 the Franks set about fortifying the city (or perhaps, refortifying it). William of Tyre says that this was ‘so that Ascalon might be hemmed in on the south just as it was on the north and east by the fortresses they had built there’.27 Some historical memory was also preserved – or invented – for the city. Thus, William writes: Gaza was a very ancient city and had been one of the five cities of Philistia. It was celebrated for its building, and abundant proof of its ancient excellence was manifest in the churches and spacious dwelling houses, albeit destroyed, in the marble and large stones, and in the multitude of cisterns, wells and running waters.28

230  |  syri a i n crusad e r time s Whatever the situation, in the winter of 1149–50, the Franks under King Baldwin III set about building a castle on a hill. It should be remembered that the use of the word ‘hill’ is relative: the hills in the Gaza area are low affairs, rising above a generally flat or gently rolling plain. However, William of Tyre then tells us that the king did not have enough men to fortify the entire hilltop and thus had to be content with a smaller castle; no attempt was made to repair the town walls. The castle, along with the town, was handed over to the Templars, along ‘with all the adjacent district’.29 This aspect will be discussed later in the chapter. The Fatimid defenders of Ascalon clearly understood the dangerous implications of this new Frankish stronghold to their south. Thus, they soon launched an attack against it, perhaps with reinforcements from Nur al-Din Mahmud b. Zengi, ruler of Damascus and Aleppo. This Muslim force was beaten back, and we learn that the brother of the famous Usama b. Munqidh was killed in the fighting.30 In 1154 Ascalon was finally taken by the Franks, and Gaza would serve as a jumping-off point for campaigns to Egypt, especially in the 1160s.31 The town was soon enjoying some prosperity and growth. The Muslim geographer al-Idrisi, writing in Sicily as early as 1154, informs us that: Between Ascalon and Gaza there is a distance of some twenty miles, [Gaza] is today populous [and] in the hands of the Greeks [al-Rum!, referring actually to the Franks]. The port [of Gaza] is called Tida (or Tayda).32

By 1170, according to William of Tyre, civilian residents had settled in the area around the hill with the castle, and these quarters, a veritable faubourg, was fortified with some walls and gates. In the mid-1170s it is described as large as Aascalon, with walls to match. This description is probably exaggerated, but it gives some sense of the town’s expansion and overall prosperity.33 According to John of Ibelin (d. 1266), Gaza (Gaudres) also hosted a burgess court, another sign of a relatively large population; however, some scholars take this statement with a grain of salt, not the least since this author also claims that Gaza was subservient in some way to Blanchegarde (Tell al-Safi), about 25km east of Ascalon.34 Benvenisti suggests that the inhabitants of the new quarters were mostly Muslims and eastern Christians, and just a few Franks,35 a reasonable supposition, but not more than that. Gaza was the home to at least one Greek church, named after St Porhyrius, and was the

gaz a in t he f ra nk i sh a nd a yyubid pe r io d s  | 231 centre of a Greek bishopric that extended at least to Beit Guvrin (Bayt Jubril/ Bayt Jubrin).36 We know hardly anything of the rural area near the town itself, although as noted above, to the north of Gaza, in what was then the seigneury of Ascalon, we have the names of six villages, as follows (from west to east):37 Table 13.1  Six villages in the seigneury of Ascalon. Frankish name

Arabic name

First evidence for existence in the post-Frankish period

Algie Semsem Amouhade Heleiqat

al-Jiyya Simsim/Sumsum [Khirbat] al-‘Amuda Hulayqat

Saarethe Malaques

[Khirbat] Sha‘ariyya Malakis / Umm Lakis

In sixteenth-century Ottoman registersa In sixteenth-century Ottoman registers In sixteenth-century Ottoman registersb On Palestine Exploration Fund map from 1880c In sixteenth-century Ottoman registers Mentioned in Mamluk periodd

a See Hütteroth and Abdulfattah, 1977. The section on the region of Gaza is found on 142–151. b This site is about 4km east of Bayt Jirja (itself some 15km north of Gaza City, for which there is evidence of existence in the Ayyubid and Mamluk periods; Amitai, 2017, 20–21; see Khalidi, 1992, 88. c Condor, et al., 1881–9. For the map see: (PEF layer) (last accessed 25 September 2017). d See al-‘Umari, 1998, 247. For the usage of Umm Lakis, see Blakely and Huster, 2016, 39. From the current transcription by the Franks, the form Malaqis must have been in use in Frankish times.

Important as this evidence is, it only tells us something about an area that later might be considered the northern reaches of the rural hinterland of Mamluk Gaza. What about further to the south? The late Jonathan RileySmith had stated in a general way that there were some Templar estates around Gaza City. He did not provide details, since he hardly had any (and neither do we).38 The only settlement, really a town of some size, around Gaza was Darom (Arabic Darum, later Dayr al-Balah), some 15km south of the city. It was there at Darom that King Almaric built a castle with four towers (one especially large) in 1170.39 This same year it was besieged by Saladin, still just the Fatimid vizier of Egypt, and not yet an independent ruler. A Frankish relieving force under the king arrived from the direction of

232  |  syri a i n crusad e r time s Gaza, but Saladin did not succeed in drawing the Franks, who maintained formation, into a battle, so he turned and set upon Gaza. The population of the faubourg tried to flee to the citadel, but they were not allowed in by its commander; having no alternative, they tried their best to defend the town. After fierce fighting, the Muslims broke in, and there followed much killing of townspeople, destruction and looting. Among the defenders were sixty-five young Franks from al-Bira (Magna Mahomeria), north of Jerusalem. These reportedly fought courageously, and most were killed.40 While we have no direct evidence of this, we can suppose that after the withdrawal of the Muslim army, repairs were made to the fortifications and the local population achieved some return to normalcy. Saladin was back in force in 1177, but his initial attacks on the town were repulsed by the Templars, reinforced by the Frankish king. From here the Muslim army went on to Gezer (Montgisart), near Ramla, where they were defeated.41 In 1182 Muslim forces from Egypt attacked Darom and Gaza, as part of a wide-ranging offensive in Syria. Most of the Frankish armies were in the north, but the local Templars put up a good show, r­ epulsing the Muslims.42 For their part, the Franks did not just sit in south-west Palestine, waiting to be attacked yet again by Saladin or one of his commanders. Thus, in 1181 an army of unknown provenance launched a raid against al-‘Arish further south, and then continued all the way to Tinnis (or Tanis) in the eastern Nile Delta, where they seized some merchant ships.43 This was a dynamic situation and neither side had primacy yet in this specific region nor in the southern Levant as a whole. Nobody could have predicted the big changes in store after 4 July 1187. In mid-August 1187, after his great victory at Hittin, Saladin arrived in south-west Palestine, soon taking Ascalon and Darom. Gaza initially resisted, but in September surrendered in order to obtain the release of the Templar Grand Master, Gerard of Rideford. During the Third Crusade, as the battling armies of Richard and Saladin drew towards the south, the latter ordered in September 1191 the destruction of the fortifications of several cities, most famously Ascalon, but also Gaza (but not Darom). Six months later Richard arrived at the scene, repairing the castle of Gaza and returning it to the Templars. This was, however, not a long-term arrangement: Gaza remained outside Frankish control in the peace treaty of 1192 between the Sultan and

Plate 13.1  Detail from ‘Frankish Palestine, South’, prepared by J. Prawer and M. Benvenisti, from Atlas of Israel (Jerusalem, 1960), showing a lack of evidence regarding rural settlement in the Gaza (‘Gadres’) region.

Plate 13.2  Further detail from ‘Frankish Palestine, South’, prepared by J. Prawer and M. Benvenisti, from Atlas of Israel (Jerusalem, 1960), showing the state of knowledge then of rural settlement in the areas to the north and north-east of Gaza.

Plate 13.3  Rural Palestine in the Frankish Period, from Ronnie Ellenblum, Frankish Rural Settlement in the Latin Kingdom of Jerusalem (Cambridge, 1998). Note that no. 102 in the south-east is Gaza and 82 is Darom.

Plate 13.4  The ‘Oxford Map of Matthew Paris’, MS Corpus Christi College 2*, fols 1v–2r. Gaza (‘Gazre’) is encircled towards the lower left corner.

gaz a in t he f ra nk i sh a nd a yyubid pe r io d s  | 233 the English king, and this was the end of Frankish control of Gaza the town and its region.44 How might we sum up this period of forty plus years of Frankish rule of Gaza? We see that in spite of frequent fighting in the area and its position as a frontier town, some prosperity is indicated by immigration to the city. This shows the potential for economic growth when the situation would become more stable, as we will see in the Mamluk era. The Frankish Church of St John (note that this was a church, not a cathedral) would survive the change of regime, eventually becoming the Great Mosque of Gaza under the Mamluks (and most probably already before that, under the Ayyubids). Although many changes and additions were made to it over the generations by Muslim patrons, much of its original character was left intact. In the fighting between British and Ottoman troops in World War I, it was severally damaged.45 Regarding the archeological legacy of Gaza in the Frankish period, Denys Pringle writes: Little is known of the topography of Crusader Gaza, apart from the location of two surviving churches … Walls surrounding the tell were excavated in the 1920s, but provided no evidence of medieval construction … The location of the Templars’ castle is unknown. Most of the surviving medieval buildings date from the Mamluk and Ottoman periods, when Gaza rose to prominence as an important staging post on the main road between Cairo and Damascus.46

What can we say about the new Ayyubid order in Gaza and its environs? Benvenisti tersely writes: ‘The city was restored by the Moslems and became an administrative, military and commercial centre.’47 This is most probably true as far as it goes, but certainly we would want more details; as suggested above, perhaps this statement is more appropriate for the Mamluk than the Ayyubid period. The sources that we have put little social and economic meat onto the bones of the political and administrative history of Gaza and its hinterland under Ayyubid rule. Once Muslim rule was finalised over the area and Richard and his cohorts left the country, Ayyubid administration could begin in earnest. Saladin managed to appoint his mamluk ‘Alam al-Din Qaysar as the governor

234  |  syri a i n crusad e r time s (wali) of southern Palestine (from Gaza, Darom and Ascalon in the west to Hebron in the east), as part of his reorganisation of much of Syria.48 This was an important precedent; both the larger swathe of territory and the more prescribed area around Gaza were to be ruled by trusted commanders, not members of the Ayyubid family, as in Baniyas, Baalbek and Bosra, let alone Karak. It is not that Gaza and the adjacent lands were not significant enough to warrant an Ayyubid prince, but rather perhaps the sultan thought that it would be best if this strategically important area would be run by someone personally loyal to him. The relative quiet and security that characterised the region in the years after Saladin’s death, along with a renewed regular trade between Egypt and Syria through Gaza, surely contributed to some prosperity in country and town. However, Gaza seems to have been a modest place and did not figure prominently in larger concerns. Thus Yaqut (d. 1229), writing in the early thirteenth century, describes the city thus: [Gaza] is a city on the edge of Syria on the way to Egypt. Between it and Ascalon there is a difference of two farsakhs or less. It is part of the Filastin district, west [sic, should be ‘south’ or ‘south-west’] of Ascalon.

This short depiction does not mitigate against a view of growth and prosperity, but it suggests nothing in its favour either. The fact that the we have only two extant construction inscriptions from Gaza City (1247 and 1249) and two from nearby villages (Bayt Hanun, 1239; Ni‘ilya, evidently from 1247–8), indicates limited (and relatively late) patronage, and thus modest prosperity.49 One notes their relatively late dates. The Syrian historical geographer Ibn Shaddad al-Halabi, writing in early Mamluk times, provides a quick rendition of the Ayyubid princes who ruled this city. Almost all did so from far away, mainly from Cairo and Damascus. The governors were invariably commanders, whose names are not given.50 We have no indication that the fortifications of Gaza and Darom were repaired in any substantial way during these decades. Yet, no matter what the state of Gaza City and its rural hinterland in the first third or so of the thirteenth century, it is clear that the situation deteriorated in subsequent decades. In a nutshell, Gaza and its surroundings were the focus of much movement of armies, their frequent long-term stays

gaz a in t he f ra nk i sh a nd a yyubid pe r io d s  | 235 and some fighting on a big scale. This situation would continue until 1260, and certainly had a negative impact on the region’s economy, agriculture, demography and overall quality of life. Already in August 1228 al-Kamil Muhammad, ruler of Egypt, led his army towards Palestine, camping at Tall al-‘Ajul, south of Gaza, remaining there most of the time for many months (into 1229) and conducting the business of state.51 There were other movements of troops through and in the area, but this really begins in earnest during the second reign of al-Salih Ayyub (1240–9). Even before, in November 1239, a Crusader force under Count Henry II of Bar was ‘cut to pieces’ near Gaza by an advanced force sent by al-‘Adil II b. al-Kamil, then ruler of Egypt (although not for long; by June 1240 he was replaced by his older brother al-Salih Ayyub).52 In the summer of 1240, the leading figure among the Franks, Thibault of Champagne, concluded a treaty with al-Salih Isma‘il, Ayyubid ruler of Damascus, resulting inter alia in Ascalon and the district of Gaza being ceded to the Franks. Frankish forces advanced to the environs of Gaza, to counter al-Salih Ayyub’s army that had moved south of the city. No fighting, however, occurred at this time.53 The next event of relevance was the famous battle of 17–18 October 1244, where Frankish and Damascene forces were defeated at La Forbie, or Hiribiyya (sometimes mistakenly vocalised as Harbiyya). This is not the place to discuss this battle,54 but it should be noted that a large number of troops with their horses and baggage trains were in the area for some time. Most important may have been the Khwarazmian mercenaries, numbering some 10,000 horsemen, along with families, horses (surely more than one per soldier) and maybe herds. Even without the livestock, this was a tremendous logistical burden on the region, and certainly did no good for local agriculture. The Khwarazmian troops would have had no compunction in grazing their horses on farmland. These were surely tough times for the peasants. Two years later, in 1246, another army of al-Salih Ayyub was in Gaza, this time in preparation for a conflict with the now rebellious Khwarazmians.55 With this sultan’s death in 1249, and the subsequent coup d’état by the officers leading to the establishment of a Mamluk state in Egypt, a new phase commenced in Gaza’s history. The town became the southern boundary of the Syrian Ayyubid state, ruled by al-Nasir Yusuf, and he soon used it as a

236  |  syri a i n crusad e r time s jumping-off point for a campaign to eliminate Mamluk rule in Cairo and return Egypt to Ayyubid rule.56 We need not trace all the military comings and goings through Gaza in the 1250s, let alone long stays of armies there, in the framework of the struggles between the nascent Mamluk state and the Ayyubids, as well as inner-Mamluk struggles and internal Ayyubid ones. This has been handily rendered and analysed by Stephen Humphreys in his book From Saladin to the Mongols.57 I would like, however, to retell some of the highlights of the annus mirabilis of 1260 for Gaza in more detail. The year begins with a growing concentration of various refugees from Syria – and perhaps beyond – fleeing from the Mongols who invaded northern Syria in force at the beginning of the year – although preliminary raids had commenced before. Many of these refugees were soldiers, with or without families, but certainly with horses and beasts of burden. One thinks of the Shahrazuriyya Kurds, assorted Ayyubid officers and troopers, the occasional Ayyubid prince with his entourage and then Baybars, probably with several hundred comrades from the Bahriyya regiment. Probably by mid-spring most of these had moved on to Egypt. By then, Mongol raiders were in the neighbourhood. In fact, through the summer an advanced Mongol force under an officer named Baydar (or Baydara) was there to keep an eye on developments in Egypt. It was this force that fought briefly with the Mamluk advanced guard under Baybars in August 1260, before withdrawing or being forced back. This was followed by the main Mamluk army’s arrival under Sultan Quṭuz, which after a short stop moved north, leading to the victory over the Mongols at ‘Ayn Jalut in northern Palestine on 3 September 1260.58 The events of 1260 were the crescendo of more than two decades of military activities in the region of Gaza; this surely had a decidedly detrimental effect on the economy and demography of this region. Perhaps an indication of the poor state of the area and its lack of importance, at least from a Frankish point of view, is this map from one of the manuscripts of Matthew of Paris’s History from around 1250 (see Map 4). From Damietta in the Nile Delta to Ascalon we have no settlements marked on the coast. Gaza (here ‘Gazre’) is found, but it is far inland. The location of Gaza so distant from the coast seems to indicate that it was also far from the consciousness of the Franks at this time (or at least this particular Frank), perhaps also indicating its relative unimportance in the political and economic scheme of things.59

gaz a in t he f ra nk i sh a nd a yyubid pe r io d s  | 237 It is not a surprise that after driving the Mongols from Syria after the victory at ‘Ayn Jalut, the Mamluk authorities found Gaza a desolate city and region. Matters, however, were about to change: thus, Ibn Shaddad al-Halabi, writing in the 1270s, reports: When al-Malik al-Muzaffar Sayf al-Din Qutuz al-Mu‘izzi al-Turki defeated the Mongols near ‘Ayn Jalut, and the country was taken back, the inhabitants [of Gaza] returned to [the city] and it was built anew. In our time – when this book was composed – there are in it governors of our lord, the Sultan, al-Malik al-Zahir Rukn al-Dunya wa’l-Din Baybars al-Salihi – may God make his reign last forever, and bring his rule over the entire land!60

Some seventy years later we have a description of a thriving city and countryside, from the encyclopedia Masalik al-absar by Ibn Fadl Allah al-‘Umari (d. 1349): Gaza is a city between Egypt and Damascus. Hashim b. ‘Abd al-Manaf was buried there, and there al-Shafi‘i was born. It is built of stone and plaster, its buildings are solid, on a high spot, at a distance of a mile from the Mediterranean Sea. It has good and pure water that is easy to digest, but it is not considered tasty. The drinking water of its inhabitants [originates] in wells, and it has reservoirs for rainwater, into the winter rains run, although these are considered small. It has many fruits, of which grapes and figs are the best. It has colleges (madaris) and grave sites that adorn it. It is a respectable district, in which there are army units, Bedouins and Turcomans. It borders on its two sides the land and the sea, and it is near the Sinai desert. To its south are agricultural and pasture lands, and it is a place of meeting between settled and nomadic people.61

In short, we see that the thriving city and countryside in the early fourteenth century contrasts with the apparently much less auspicious situation in the early thirteenth century, which only got worse by the mid-century before the Mamluks decisively took over. What can we draw from all this? Firstly, research into the history of a particular region contributes to the better understanding of a larger territory, in

238  |  syri a i n crusad e r time s this case the southern Levant and perhaps even Syria as a whole. More specifically, the examination of the more restricted area helps us to see the wider dynamics of Muslim–Frankish relations in a larger region. While there were times of quiet in the almost forty-five years of Frankish domination, there is no evidence of an ongoing modus vivendi between Franks and Muslims, certainly among the military–political élites, which has been suggested as a leitmotif for political and military relations for the Levant in this period.62 In this particular corner of Palestine, it was an ongoing contest lasting for almost half a century, not an example of a de facto rapprochement. Secondly, I think that we can say that for the common people, certainly in the countryside, it was not business as usual as the princes and kings fought their wars with (mostly) professional soldiers. The movement of armies and their clashes certainly impacted on the countryside and its inhabitants, and it was clearly not for the better. From the arrival of the Franks in the country until the establishment of stable Mamluk rule more than 150 years later, the local inhabitants surely paid the price for the frontier status of Gaza and its hinterland. Thirdly, a study of Gaza in the Frankish and Ayyubid periods helps us better understand the subsequent quarter millennium of Mamluk rule. While we unfortunately lack many details (especially regarding the rural sector), we now have at least a partial basis of comparison to discuss more effectively the development of the city and countryside under the relatively firm hand of the Mamluk sultans and their governors. Finally, I can suggest that we should aim to put Gaza and its hinterland into the mainstream of regional and wider history in the Middle Ages. While relatively small and unimportant at times, it was never off (literally) the beaten track. On occasion, it plays a role of some significance in the larger scheme of things, and certainly can serve as an illuminating example for larger trends, as I hope to show for the Mamluk period.63 By studying places like Gaza, we will better understand matters in the centre, as well as the larger picture. Notes   1. The research and writing of this chapter was sponsored by the Israel Science Foundation (ISF, grant no. 1827/16), but the groundwork for it was laid while I was previously a fellow at the Anne-Marie Schimmel Kolleg at the University

gaz a in t he f ra nk i sh a nd a yyubid pe r io d s  | 239 of Bonn in Germany. This is an opportunity to thank the directors, staff and fellows of the Kolleg for their support and encouragement. Preliminary results of the study have now been published as Reuven Amitai, ‘The Development of a Muslim City in Palestine: Gaza under the Mamluks’, ASK Working Paper 28, Bonn: Anna-Marie Schimmel Kolleg, August 2017, available at (last accessed 26 June 2019). I would like also to thank Kate Raphael, with whom I have been discussing matters related to late medieval Gaza since 2014.   2. This volume was published at Leiden 2009. Since the publication of Volume One in 1997, seven volumes of this series have appeared and several more are now in preparation.   3. For an initial analysis of the epigraphy of Mamluk Gaza, see Amitai, 2017, 12–14. The spread and nature of inscriptions in Mamluk Jerusalem will become clear with the forthcoming publication of relevant volumes of the Corpus Inscriptionum Arabicarum Palestinae (CIAP).   4. Amitai, 2017, 15–17. The countryside around Gaza is now the subject of a study by Kate Raphael, under the aegis of the ISF funded project noted above in note 1.  5. See ‘Ata Allah, 1986; Sadek, 1991. See also the preliminary notes in the ­section on Gaza in CIAP, and throughout the section, as well as al-‘Arif, 1943, 140–167.  6. Al-‘Arif, 1943, 112–127. One can note the general comments in both the second and third editions of the Encyclopedia of Islam, by Dominique Sourdel and Johann Büssow respectively (see below for details), and also the many en passant remarks by Gil, 1992; ‘Athamina, 2000.   7. Benevisti, 1970, 189–194.   8. Pringle, 2008–10, 1: 194–196 (‘Dair al-Balah’), 208–219 (‘Gaza’).   9. For example, Prawer,1970. 10. Most importantly, Humphreys, 1977. 11. Atlas of Israel, 1956. The map of the Crusader period is 12/IX. 12. These are villages are (from west to east): Algie, Semsem, Amouhade, Heleiqat, Saarethe and Malaques; see Map 2 for their locations. The basis for their inclusion in this map is Prawer, 1958, 234–237 [Hebrew] (see especially the map on 236. See also the map in Riley-Smith, 1967, 481. I will discuss these settlements in further detail (with their Arabic names) below. 13. Ellenblum, 1998, xviii. 14. Benvenisti, 1970, 191. 15. See Leisten, 1996–2003, 4: 815; Avi-Yonah and Gibson, ‘Gaza,’ Encyclopaedia

240  |  syri a i n crusad e r time s Judaica, 7: 398–399. See Abulafia, 2012, 218–221 for some discussion on Jews and Christians, as well as the decline of the pagan community, in Gaza during the last centuries of Roman rule. 16. Al-Istakhri, 1927, 58; Ibn Hawqal, 1938, 172. The text in these two works is almost the same, and has much similarity to that of al-Muqaddasi, cited below. Cf. Le Strange, 1890, 442. 17. Sourdel, ‘Ghazza’, EI 2. Sourdel writes that ‘its port was Mimas, the ancient Maioumas mentioned as early as the third century bc, the site of which corresponds with the modern al-Mina’. 18. Levy-Rubin, 2011, 164–171. 19. Al-Muqaddasi, 1906, 174. André Miquel was correct in reading ‫( ايسر‬became rich) instead of ‫( اثر‬monument), which was read by the editor. This is clearly seen when comparing this text to that of Ibn Hawqal and al-Istakhri cited above; tr. Miquel, 1963, 203 and n. 236. The mistaken reading is the basis for the erroneous translations in Le Strange, 1890, 442; and Collins, 1994, 157–158. 20. Gil, 1992, 316–429, reviews events in Palestine and neighbouring areas, putting them into the wider regional context. Gaza is hardly mentioned at all in these sections. On p. 395, he writes: ‘Apparently, one of the results of the uprising [of Bedouins against the Fatimids] was that the Jews who lived in villages or small towns were forced to find refuge in the cities, as is evident from the version: “from your brethren, the community of Gaza and those who fled to it” found in the signature of a letter from this community to the Fustat court.’ This is interesting evidence, but frustratingly brief and unique. 21. Ibid., 412 (section 605). I am grateful to Dr Shimon Gat for bringing this incident to my attention. 22. Benvenisti, 1970, 190. For ongoing Frankish efforts to take Ashkelon, see: Prawer, 1970, 1: 328–330, 406–411. Büssow (‘Gaza’, EI 3, online edition, available at: (last accessed 31 August 2017)) writes: ‘When the Crusaders conquered Gaza in 493–4/1100, they reportedly entered an empty town; its entire population appears to have fled after receiving news of the massacres perpetrated during the conquest of Jerusalem.’ This is not correct: as we will see, the Franks took over Gaza only several decades later, and the matter of the massacre in Jerusalem in July 1099 would thus not have been relevant. See below for the question whether the city was indeed empty. 23. There is hardly any mention of the city in the Arabic sources for this first half    

gaz a in t he f ra nk i sh a nd a yyubid pe r io d s  | 241 of the twelfth century, that is, before Frankish rule was established in this part of the country. Thus, Ibn al-Athir only notes Gaza once in passing during these years; Ibn al-Athir, 1965–7, X: 666; tr. Richards, 2006–8, 1: 285. The chronicle by Ibn al-Qalanisi, 1908 only mentions some events concerning the city in the late 1140s and 1150s; cf. tr. Gibb, 1932. 24. William of Tyre, 1986, LIII and LIIIa, Book 17, ch. 12; William of Tyre, tr. Babcock and Krey, 1943, II: 202; see also Pringle 2008–10, 1: 208. 25. Benvenisti, 1970, 190. 26. ‘Izz al-Din b. Shaddad, 1963, 260. 27. William of Tyre, loc. cit. (original text and translation as in note 24 above). 28. Ibid.; cited in Pringle 2008–10, 1: 208. 29. William of Tyre, loc. cit; cited in Pringle 2008–10, 1: 208, along with other sources; see also Kennedy, 1994, 31; Prawer, 1980, 106–107. This hill would be the higher ground in the centre of the pre-modern city, where the Ottoman Serai (‘palace’ or administrative centre) was later erected. 30. Usama b. Munqidh, 1424/2003, 72–73. Usama b. Munqidh, tr. Cobb, 2008, 26; cf. tr. Hitti, 1987, 42. Usama himself had been dispatched from Egypt to Nur al-Din to bring help to Ashkelon in the aftermath of the Frankish activities in Gaza. Returning from Damascus with a not insignificant force (860 horsemen), Usama saw much fighting in the area. After four months he returned to Egypt; presumably the Zengid expeditionary force remained and was part of the above described attack in which Usama’s brother, ‘Izz al-Dawla Abu’ l-Hasan ‘Ali, was killed. Usama b. Munqidh, 1424/2003, 63–73; tr. Cobb, 2008, 18–26; tr. Hitti, 1987, 34–42. This account is summarised by ‘Izz al-Din b. Shaddad, 1963, 260–261. 31. Beyer, 1950, 175: ‘Nach dem Falle des lezteren [that is, Ashkelon, R.A.] (1153) hat Gaza dem König Amalrich I. von Jerusalem als Stützpunkt für seine ägyptischen Unternehmungen gedient.’ For more on these campaigns, see: Prawer, 1970, 1: 427–441. 32. Al-Idrisi, 1970–84, 4: 356; cf. translation in Le Strange, 1890, 442. Benvenisti, 1970, 190, notes that the name Tida/Taida is derived from the Greek Anthredon. 33. ‘The construction of the fortress and the capture of Ascalon four years later seem to have brought about a revival in Gaza’s fortunes.’ Pringle, 2008–10, 1: 208; see also Benvenisti, 1970, 190. Prawer, 1980, 107, describes this process as similar to other developing towns in Europe at this time: ‘Around the fortress a new settlement blossoms. Desirable permanent elements are attracted: ordinary people seeking a measure of security, farmers possibly from the vicinity, but

242  |  syri a i n crusad e r time s also traders and artisans to supply the needs of the castle, the new settlement and neighboring villages. To protect themselves, the newcomers built, by their own effort, a second wall around the fortress, which enclosed the castle and the homes of the new settlement. This wall, equipped with gates and towers, was low and weak, but succeeded affording a measure of security from assault.’ All of these modern accounts go back to William of Tyre, 17: 12 (cited above) and 20: 20 (ed. Huygens, 1986, 937–939; tr. Babcock and Krey, 1943, 373–375). His contemporary, Theodoric (fl. 1169–74), mentions Gaza twice in passing without too much comment: (1) ‘Ten miles [sic] from Hebron to the north, and on the shore of the Great Sea is Gaza, now called ‘Gazara’, in which Samson did many miracles, and one night took its doors away. Eight miles from Gaza the well-defended city of Ascalon is situated on the short of the Great Sea. These cities had been located in Palestine or the Land of the Philistines.’ (2) ‘After this you reach first Gaza or Gazara, and the well-defended city of Ascalon, all of which have been mentioned above. These are all the cities along the coast, and all of them are large and have walls.’ Theodoric, 1976, 40–41, 51 (chs XXXVI and LI); tr. Wilkinson, with Hill and Ryan, 1988, 307, 314. 34. John of Ibelin, tr. Edbury, 1997, 194. The problematic nature of this evidence is dealt with in ibid., 159–161. See also Ellenblum, 1989, 104–105. 35. Benvenisti, 1970, 190. 36. The Greek bishopric centred in Gaza is noted in ibid., 194; however, Pringle, 2008–10, 1: 216, appears not to accept this, hinting that Beit Guvrin was the more important centre. Cf. al-‘Arif, 1943, 132, for a confused rendition of ecclesiastical events in Beit Guvrin, mistakenly placed in Gaza. See Pringle, 2008–10, 1: 216–219, for a detailed description of the church. 37. The basis for this evidence is a list of villages (which extends further east) provided in a contract in 1256–7, between John of Ibelin, count of Jaffa and Ashkelon, and the Hospitallers, in which these thirteen villages were transferred to the John to the Order. Of course, this was almost completely theoretical, since the villages in question were at the time most probably not under the control of this or that Frankish authority. One also wonders if their mention here means their actual existence in the 1250s or harks back to a period of firmer Frankish rule. Paoli, 1733, reprint 2011, 150. This part of the document was analyzed by Prawer, 1958, 234–237; and exhaustively by Blakely and Huster, 2016, 35–53. I am not convinced their suggestion that Elroeiheib is Hiibiyya (site of the famous battle in 1244, known also as La Forbie), to west of Amouhade. The phonetics of this identification are very far-fetched, and this was

gaz a in t he f ra nk i sh a nd a yyubid pe r io d s  | 243 a well-known site, which surely would have received a better transliteration than this. I agree, however, that previous attempts to identify the site with an Arab name have been unsuccessful. 38. Riley-Smith, 2012, 34; Riley-Smith, 1967, 73. Cf. Beyer, 1950, 175: ‘Die Stadt und das umliegende Land war also keine Herrschaft, auch kein Teil einer solchen, sondern ein den Templern anvertrautes königliches Gebiet, das wohl auch in kirchlicher Hinsicht dem Templern unterstand.’ Cf. William of Tyre’s statement, cited in note 29. 39. Kennedy, 1994, 31; Prawer, 1980, 107–109, provides a detailed discussion, and notes that part of the purpose was ‘to establish political and economic hegemony which would carry with it the right to collect taxes from the agricultural settlements in the area [about which we have neither details nor names] and duties from caravans plodding northward from Egypt on the Via maris, which crossed the Frankish border.’ Al-‘Arif, 1943, 130, suggests that Darom was militarily more important than Gaza itself, at least as a forward position vis-à-vis Egypt. 40. Lyons and Jackson, 1982, 42–43; Eddé, 2011, 45–46, 187; Benvenisti, 1970, 191; Pringle, 2008–10, 1: 208. 41. Lyons and Jackson, 1982, 122–123; Benvenisti, 1970, 191; Prawer 1970, 1: 550–553. 42. Lyons and Jackson, 1982, 170; Benvenisti, 1970, 191; Prawer, 1970, 1: 604–605. 43. Ehrenkreutz, 1972; Eddé, 2011, 188; cf. Prawer, 1970, 1: 611. 44. Benvenisti, 1970, 191; Pringle, 2008–10, 1: 208; Lyons and Jackson, 1982, 272–273; Eddé, 2011, 268–269; cf. Prawer, 1970, 1: 668; 2: 83, 93. On the continued standing of Darom as a fortified town see Stubbs, 1864, 280–281, 330, 352–356; see also tr. Nicholson 1997, 261. For its resistance to Richard and its later capture by him, see ibid., 298–299, 316–319. 45. Benvenisti, 1970, 194; for details of the Church, see Pringle, 2008–10, 1: 208–216. 46. Pringle, 2008–10, 1: 208. From another study by this author, we see that there are no extant secular buildings in Gaza or in its immediate rural hinterland; Pringle, 1997. 47. Benvenisti, 1970, 191. 48. Humphreys, 1977, 79. 49. CIAP, 1: 189 (Ni‘ilya, where it is mentioned in passing; a separate entry will appear in a later volume of CIAP); 2: 98–104 (Bayt Hanun; see below, note 57); 4: 52–55 (nos 6–7, for Gaza City). There are also two Muslim epitaphs from

244  |  syri a i n crusad e r time s the city from around this time; ibid., 4: 55–57. Yaqut notes three villages in the vicinity of Gaza City; Amitai, 2017, 15–16. 50. Ibn Shaddad, 1963, 265–266. For more on this, see Humphreys, 1997, 206, 241, 244, 254. 51. Humphreys, 1997, 195–206. Tell al-‘Ajul is an archeological mound, located at the mouth of Wadi Ghazza, just south of the city of Gaza. 52. Ibid., 261 (from which this quote is taken); Prawer, 1970, 2: 272–274. Lower, 2005, ch. 9, esp. 167–171; Jackson, 1987, 37–40. A senior Muslim commander had a mosque built in Bayt Hanun after the battle, called Jami‘ al-Nasr (Mosque of Victory), with an informative inscription, now lost; CIAP, 2: 98–104, including an account of the battle and its aftermath. See also al-‘Arif, 1943, 133, n. 1. 53. Humphreys 1997, 268–269; Lower, 2005, 173–174. There is some question of Gaza City’s fate at this time: Humphreys notes that the city itself was not turned over to the Franks, while Jackson, 1987, 58, suggests otherwise. In any case, there is no indication that the Franks actually occupied the city and may have only had a symbolic presence to its north. Jackson, 1987, 46–56 provides information on the frequent presence and movement of armies in the area until the battle of La Forbie in 1244. 54. Humphreys 1997, 274–276; Prawer 1970, 312–313; Berkovich, 2011, 9–44. For more on the Khwarazmians in the environs of Gaza, see Humphreys, 1997, 284. 55. Humphreys, 2007, 285. 56. Ibid., 311, 315. 57. Ibid., 311–358. See also Irwin, 1986, 26–30. 58. For these events, see Humphreys 1997, 347–358; Amitai-Preiss 1995, ch. 2. 59. I am grateful to Dr Mordechai Lewy, who brought this map to my attention, explaining its cartographic peculiarities. He notes (in a personal communication on 15 September 2017) that ‘Gazre’ appears a little far from the coast because Matthew erased the original coastline, putting it closer to the edge of the parchment, in order to gain more space for copying the map (from an anonymous source); he forgot to reposition Gaza on the new coastline. To my mind, this confirms my suggestion that in the earlier part of the thirteenth century, Gaza did not figure highly in the consciousness of the Franks. 60. ‘Izz al-Din b. Shaddad, 1963, 266. 61. Al-‘Umari, 1984, 142–143. 62. See the interesting and detailed work by Köhler, 2013. I hope to prepare in the near future a more considered response to this thought-provoking study. 63. Some comments to this effect are already made in Amitai, 2017, 19–20.

14 Picture-poems for Saladin: ‘Abd al-Mun‘im al-Jilyani’s Mudabbajat Julia Bray

I

n her life of Saladin, Anne-Marie Eddé notes that Saladin’s physicians were often highly cultivated men of letters, and mentions in particular the Andalusian emigré ‘Abd al-Mun‘im al-Jilyani (531–602 or 3/1136–1206), regretting the loss of his ‘book on Saladin’s conquests’.1 We do not know much about al-Jilyani as a person, even though his extant writings are to some extent ego-documents in which his self-evaluation looms large; and in spite of his auto-bibliographies and Ibn Abi Usaybi‘a’s list of works based on them,2 our picture of him as an author and cultural actor lacks focus. To touch selectively on his output, throughout most of his adult life he composed mystical sayings, which he collected in Kitab adab al-suluk. They are unpublished and have not been studied, and we do not know how they relate to his other writings.3 He wrote poetic descriptions of Saladin’s battles, dated from 565/1169–704 to 587/1191,5 which are reasonably well known because they are quoted by the historian of medicine Ibn Abi Usaybi‘a and the political historian Abu Shama. They are cited in twentieth-century Arabic scholarship on the poetry of the Counter-Crusade,6 and in a new English study of anti-Crusader poetry.7 He was also known for picture-poems, mudabbajat, literally ‘brocaded’ or ‘figured’ pieces – the term is apparently his own invention. In Arabic literary criticism, verse was often compared to woven stuffs, but al-Jilyani’s mudabbajat were a new departure. He dedicated mudabbajat to Saladin, and other Ayyubids, in celebration of his achievements and to glorify his new dynasty. An untapped source for the ideology and propaganda of Saladin’s regime, they survive in manuscripts whose contents and titles vary.

248  |  syri a i n crusad e r time s Brief extracts from the Uppsala and Paris MSS of mudabbajat were published by K. V. Zetterstéen in 1927. Further extracts based on these plus the Manchester John Rylands and Damascus Zahiriyya MSS8 were incorporated into ‘Abd al-Jalil ‘Abd al-Mahdi’s reconstruction of al-Jilyani’s collected ‘Poems of good tidings and of Jerusalem’ (Diwan al-mubashshirat wa-lQudsiyyat) – the book on Saladin’s conquests that Eddé had in mind. A full edition of the Manchester MS by Kamal Abu Deeb, which will be described presently, appeared in 2010 under the title Diwan al-tadbij, prefaced by a preliminary literary and codicological study. Mudabbajat manuscripts had previously been described in some detail in Arabic and European bibliographies and library catalogues, often with aesthetic admiration9 (though the Paris cataloguer declared them ‘of no interest either historical or literary’);10 but otherwise they attracted no scholarly attention until Konrad Hirschler noted how the title of the mudabbaja called Manadih al-mamadih (‘Prairies of praiseworthiness’) chimes with the ideological nature imagery of counterCrusading historians.11 Hirschler, however, had not seen the piece and did not know about the literary specifics and visual form of the mudabbajat, which are remarkable for three main characteristics: textual interweaving, and the use of colour and of geometric or natural shapes. The text of the mudabbajat may be poetry, poetic prose (which rhymes and is known as saj ‘), or a combination of the two. In addition, the text is layered and interwoven: it contains one or more secondary texts, which the reader can derive from the primary text by following colour coding in which words, syllables or letters common to the primary and secondary texts are picked out in red, green, blue, yellow or gold. All poems, primary or secondary, conform strictly to the classical rules of Arabic metre and rhyme, a considerable technical feat given that they are written inside a wide range of geometric and vegetal shapes – stars, circles, polygons, columnar tables, chequer-boards, trees – combined so as to lead in and out of each other. The overall effect – which, for reasons which will be clear presently, cannot be reproduced here in plates – is so striking that the mudabbajat have been hailed as unique;12 and so they are in their time. There is nothing to compare with them in the Arabic art and literature of the Counter-Crusade and the culture of Saladin’s court; and given that al-Jilyani chose, as an

Plate 14.1  Al-Jilyani, Diwan al-tadbij, Manchester John Rylands MS Ar. 690, fols 57b–58a: tree finial of Manadih al-mamadih.

Plate 14.2  Transcription of reading of al-Jilyani, tree finial of Manadih al-mamadih.

Plate 14.3  Al-Jilyani, Diwan al-Tadbij, Manchester John Rylands MS Ar. 690, fol. 79a: the extemporised ‘Seal’.

Plate 14.4  Ibn al-Hajj, al-Diwan al-‘amm (Fez, 1995), 550: tree-shaped poem.

pi cture-poems f or sala d in  | 249 exile from Spain, to attach himself to Saladin, they are clearly important for understanding Saladin’s prestige. Here, though, because there is still so much we do not know about the mudabbajat, I will consider them only in formal literary and visual terms. The literary origins of the mudabbaja form have yet to be determined, but newly published research by Lara Harb suggests that the underlying idea of deriving one poem from another was already several centuries old by al-Jilyani’s time.13 Art historians have not noticed the mudabbajat, and there is no scholarship on them as a visual art form. As for their reception in their own time, the literary historian Yaqut, who met al-Jilyani in Damascus in around the year of his death, gives a description of his picture-poems’ appearance and how they work visually which makes it clear that he had never come across anything similar. This is our only contemporary description.14 Slightly later, Ibn Abi Usaybi‘a, whose father had known al-Jilyani, is the only person other than al-Jilyani himself to use the technical term tadbij to refer to this part of his output; he seems to use it as a familiar literary term but does not explain its visual dimension.15 At around the same date, Manadih al-mamadih is listed by name, with no description, in the catalogue of the Damascus Ashrafiyya Library.16 How does this compare with the surviving evidence? All the manuscripts of mudabbajat that I have seen post-date al-Jilyani. The undated, stylistically Mamluk Jerusalem MS and the early Mamluk Manchester MS dated to 735/1334 are not very much later. Nevertheless, we do not know how true they are visually to what al-Jilyani drew. For example, the complicated naturalistic tree finial with inscribed wording that occurs on fols 28b–29a of the Jerusalem MS and fols 66b–67a of the Manchester MS is dealt with confidently by these scribes, perhaps indicating a continuous tradition, but is less crisp in the Gotha MS (fols 43b–44a). In a very recent article, Murad Tadghut says that the Jerusalem MS, catalogued as undated, is in fact dated 598/1201 and is in al-Jilyani’s own hand,17 but I have not been able to examine images closely enough to confirm this. All the MSS display considerable consistency as regards the body of their texts, however. They are all carefully copied in spacious formats using both black and coloured inks, in the same script style, naskh, and in similar styles as regards the figures, with exceptions only in the Paris MS. Title pages

250  |  syri a i n crusad e r time s where present are less consistent. The Manchester MS, Diwan al-tadbij, has an elaborate frontispiece which misattributes its authorship,18 whereas the Uppsala MS, with a similar title (Kitab diwan al-tadbij), but shorter and containing fewer poems, has a frontispiece grandly tailored to celebrate alJilyani. The Jerusalem MS of Manadih al-mamadih has a modest frontispiece; the Gotha MS of Manadih al-mamadih has an unadorned title page; the Paris MS has no title page. Happily, just when the ongoing development of Arabic codicology and book studies has opened up new lines of enquiry, the Gotha, Paris and Jerusalem MSS have been made fully publicly accessible by digitisation.19 Within this framework, we can make better use of two bodies of older Arabic scholarship, one documenting copies of al-Jilyani’s mudabbajat, the other the occurrence of later, non-mudabbaja Arabic compositions that have a significant visual dimension involving text derivation or interweaving and which may help to understand the mudabbajat.20 In a companion article to this,21 I have given an overview of current knowledge of mudabbaja and nonmudabbaja visual verse, which is already slightly out of date in the light of Lara Harb’s findings, of new interpretations suggested by the physical examination I have since been able to make of the Paris, Uppsala and Gotha MSS, and of additional evidence about the long-term influence of the mudabbajat. Also to be taken into account are recent developments in the study of Arabic esoteric word-and-figure texts22 and in the understanding of Fatimid uses of paper. Advances in paper studies provide a new context for investigating the physical production of al-Jilyani’s mudabbajat, while the concept, newly applied to Islamic esoteric texts, of ergodic literature, suggests an appropriate way to understand mudabbajat simultaneously as visual art and as poetry. Here I shall try to draw some of these strands together. I shall ask how, physically, the mudabbajat were composed and performed and what they contributed to Saladin’s court. To end with, I will look briefly at new evidence about their survival and influence. Unlike Zetterstéen and ‘Abd al-Mahdi’s publications, Abu Deeb’s edition of the Manchester MS shows the pictorial elements of the mudabbajat, using colour in the text where it occurs in the original and reproducing pages of figures from the Manchester, Paris and Uppsala MSS. For the first time, it conveys the visual impact and codicological features common to all the

pi cture-poems f or sala d in  | 251 MSS that I have so far examined (Uppsala, Paris, Gotha) or seen in complete digital reproductions (Jerusalem and Manchester). All of them are codices, whose pages carry either what I call ‘plain’ text, set out in the normal way in horizontal lines, or text written inside figures based on geometric and vegetal shapes, the majority of which are interconnected in sequences. In none of the MSS is it possible to read the longer sequences as wholes, because they run over several folios, recto and verso. To understand the poems they contain, the reader must instead rely on the plain text transcripts which alternate with the figure sequences. The transcripts also contain instructions on how to decipher the majority of the figures and sequences by reading up, down, across and between them. A handful of figures are without transcripts. Later, this chapter discusses two of them. A brief word about the MSS to supplement Abu Deeb’s data: the Manchester MS contains the most figure sequences, but the order of the figures’ components is sometimes muddled, for although there are catchwords on every verso folio of plain text, there are none on the folios inscribed with figures. With 117 folios excluding the guard pages, the Manchester MS has sixty-six pages of figures excluding the frontispiece. The insecurely dated Gotha MS,23 with seventy-three folios, has twenty-six pages of figures and contains a total of thirteen picture-poems. I will confine my discussion here to these two manuscripts. My limited time with the undated Uppsala MS24 was mostly spent on its beautiful, stylistically Mamluk binding which, if integral, would place it among the earlier group of manuscripts. The digital version of the Jerusalem MS came to my attention too late to be included. It has seventy-two folios in a modern binding, no catchwords, and has been rebound with the components of some figures out of sequence. The undated Paris MS will also be left out of the discussion because its text has large gaps and is interrupted by intrusions modelled on the mudabbajat but different in content.25 Obtaining a digital file of the Manchester MS enabled me to print off the figures in it and reconstruct them in sequence. Not all are complete and some passages of reconstruction are speculative. The result, for the time being, is as follows: there is one figure of one page: fol. 79a; four figures that occupy two pages each: fols 19b–20a, 55b–56a, 111b–112a, 114b–115a; one asymmetrical two-part figure on three pages: fols 110a/b–111a; one

252  |  syri a i n crusad e r time s three-part symmetrical figure on three pages: fols 108a/b–109a; one two-part figure on four pages: fols 34b–35a/b–36a; one sequence of three figures on six pages, which seems to be missing a component of one or more pages: fols 25b–26a– [missing] –26b–27a[?missing]–27b–28a. In each of the following sequences, each figure occupies two pages: one sequence of eight pages/four figures: fols 103b–104a/b–105a/b–106a/b–107a; two sequences of ten pages/ five figures: fols 6b–7a/b–8a/b–9a/b–10a/b–11a and fols 66b–67a/b–68a/b– 69a/b–70a/b–71a; one sequence of twelve pages/six figures: fols 93b–94a/b– 95a/b–96a/b–97a/b–98a/b–99a; and one sequences of fourteen pages/seven figures: fols 57b–58a/b–59a/b–60a/b–61b–66a–65b–65a–64b–64a–63b. The sequences contained in the Manchester MS are the components of several named mudabbajat, of which the twenty-six pages of 55b–56a + 57b– […] –63b + 66b–71a form Manadih al-mamadih, dedicated to Saladin. The Gotha MS, which is freely accessible online, contains only Manadih al-mamadih, and yields three sets of figures, which I will describe in detail: one of two pages containing a single figure, a table in columns (fols 31b–32a, equivalent to fols 55b–56a in the Manchester MS); one of fourteen pages containing seven sequential figures, each occupying two pages and connected to the following figure by a duct or ducts leading to the left in the direction of the book openings; the figures are: a straight-branched tree with a central trunk; an eight-pointed star inscribed in a circle; a table in columns; concentric circles; a second table in columns; a circle inscribed with an eightpointed star; a third table in columns (fols 33b–40a, equivalent with minor variants to the fourteen-page sequence in the Manchester MS); and one sequence of ten pages containing five figures, each occupying two pages and connected with each other as above; the figures are: a curvilinear tree with interlacing branches and a double stem; a rectangular table with an inscribed diamond and on each long side six angled branches leading off the top and bottom of the page and not connected with the preceding and following figures; a circle inscribed with an eight-pointed star; a circle inscribed with a hollow cruciform design made up of cusps and four small circles; concentric circles inscribed with an eight-pointed star (fols 43b–48a, equivalent to fols 66b–71a in the Manchester MS). In both MSS, with the exception of the single-page figure in the Manchester MS, which will be discussed below, the text inscribed in the

pi cture-poems f or sala d in  | 253 figures makes sense only when it is followed across rectos and versos along the connectors. It seems clear, therefore, that the figures must originally have been inscribed not on the folios of a codex, but on a continuous surface – that is, on scrolls, a conclusion reached independently by myself and Salam Rassi in his Hill Museum and Manuscript Library catalogue entry on the Jerusalem MS. When printouts of the figures are stuck together in this form, they are most beautiful and impressive. The fact that the Gotha MS’s three sequences are uninterrupted and the connectors between the figures very accurately aligned suggests that they were copied directly from scrolls, or from a prototype that had scrolls as its guide. So, despite its probably earlier date, the Manchester MS, with its occasional interruptions and losses and often poorly aligned connectors, is more likely to have been copied from another codex with similar faults, the ultimate prototypes for its figures having perhaps been scrolls folded accordion-like into codex format. This would help to account for discontinuities when the figures were transcribed on to folios. The question of the source of the transcripts and instructions in plain text which intersperse the picture-poems in their present codex format will be addressed when we discuss how the mudabbajat might have been used. The discovery that the mudabbajat were originally scrolls immediately leads to two sets of questions: how, physically, were they produced? What were their probable dimensions? And how were they read and used? In the companion article to this, I drew attention to various scroll formats and uses that were current in Egypt and Syria at the time of al-Jilyani’s residence there from around 564/1168–9, when he started to chronicle Saladin’s deeds in verse.26 From Egypt, small, block-printed amuletic scrolls (tarsh) offer good visual parallels with some of the motifs and layouts found in al-Jilyani’s figures.27 Scrolls were also in use as ‘sleeve books’.28 Scrolls in use in Fatimid and early Ayyubid Syria and of a scale comparable to alJilyani’s (the dimensions of the mudabbajat MSS will be given presently) also provide a possible context. These include parchment Qur’an scrolls and paper pilgrimage certificates. One particular sort of scroll seems especially relevant. In a keynote speech given at the 2017 meeting of the Society for the Medieval Mediterranean in Ghent, Marina Rustow presented her latest findings on

254  |  syri a i n crusad e r time s the Fatimid decrees sent from the Egyptian capital to the provinces.29 She emphasised their extravagant size: with a roughly standard width of 45cm, they could reach up to 8m in length. They were, she argues, performative, designed for public display and to impress (whereas, as shown by Geoffrey Khan, copies intended for archiving were written on bi-folios). As for their medium, paper, Rustow argued that, in the Islamic world, from the time it was adopted, although cheaper than parchment, paper had been a prestige medium associated with imperium. Since the greatest attested size of paper produced in the Islamic world at this period is about 1m in length,30 the scrolls on which the display versions of Fatimid decrees were written had to be made by gluing several sheets together. In technique and visual impact, Fatimid decrees seem to offer a useful parallel with al-Jilyani’s scrolls. We do not know what al-Jilyani’s originals were written on, but, in terms of cost of production, paper seems more likely than parchment. For their original dimensions, we might suppose a folio breadth of up to 45cm, as for Fatimid decrees. The surviving copies are not uniform. The Paris MS is very large: 40.5cm × 28cm. The Gotha MS is much smaller: 26cm × 18cm, and the Jerusalem and Manchester MSS lie in between at 30 × 21cm and 30.5cm × 21cm respectively. When the Manchester scrolls are printed on slightly smaller A4 paper, the longest measures about 2.75m. Fatimid decrees also offer a parallel with the mudabbajat’s purpose of glorifying the ruler and his dynasty, and bearing in mind that Saladin had been a Fatimid vizier, we may suppose that al-Jilyani drew on Fatimid art forms and courtly practice. Apart from display, what might have been the uses of the mudabbaja scrolls? At this point we must turn to the texts of the picture-poems and how we are instructed to decipher them. Physically, this raises a problem. Decipherment of a whole figure sequence requires manipulation of the scrolls, turning them this way and that, as practical experiment with printouts has demonstrated. The longer the scrolls, the harder they are to manipulate. Perhaps this was circumvented by the gradual unscrolling only of the sections being deciphered, with the complete sequence being revealed only when the puzzles posed by individual parts had all been solved. The scroll format would then have had two purposes: to give readers the satisfaction of cumulative

pi cture-poems f or sala d in  | 255 puzzle-solving, and to cap this with the surprise of finally seeing the whole sequence displayed. The scrolls were the primary product, or performance scripts, of alJilyani’s mudabbaja-making activities. The codices that have come down to us bookend them with plain text transcripts and guides to decipherment. Originally, I would guess, these would have taken the form of separate booklets. In other the words, the mudabbajat as we now have them are in fact a compound of scroll and codex. Let us now look in detail at the texts and at some of the readable effects produced by deciphering them, taking two small examples. The first example is the figure of a tree,31 which forms the finial of a sequence of seven figures (fols 33b–34a in the Gotha MS, equivalent to fols 57b–58a in the Manchester MS; see Plate 14.1). In the Manchester MS, from which I originally deciphered it, and also in the Uppsala and Paris MSS, this figure has no transcript and only fragmentary instructions, but in the Gotha MS full instructions are written around the figure, whose key line is transcribed. Gotha fol. 33b says: This tree contains twenty-one lines of verse in the tawil metre which branch out from a common stem. The branches of the tree are [or rather: contain] the second hemistich of each line. [Part of] the first hemistich of each line is [contained in] the trunk, and each line begins with the word it shares with the ‘seal’ [the preceding figure].

This word runs through the connector. Fol. 34a continues: ‘The word [actually two syllables] shared by all of them is mawa which runs out of the end of the eighth line of the “seal”, [being the last syllable], mu, of its [last] word.’32 The remaining instructions tell the reader to add to each of the branches of the tree first the syllables mawa, then mawa plus the next, colour-coded, syllable/word in the line which runs down the trunk of the tree. The symmetrical branches are of diminishing length: on each side of the trunk, the two longest, with which the reader should begin (and which are labelled as such in the Gotha MS) lack one syllable to make them metrical; the next two branches lack two syllables, and so on. The metrically complete line that runs down the trunk and has been cannibalised to make up the other lines should then be read in its entirety as the last line of the poem.

256  |  syri a i n crusad e r time s The result of this exercise is shown in transliteration in Plate 14.2. Here is a translation:   1.a Scales of justice – orthodoxy and the common good – are held aloft in the hands of Saladin;   1.b His pledges are all fulfilled; his onslaught is destruction; his greatness is doubly assured by his ferocity and his benefactions.   2.a His are gifts before which the Nile in flood is nothing; they belong to him alone, whence they spread among all people.   2.b By virtue of these gifts, may you enjoy victory and a return on them heralded by the proclamation of God’s glory!   3.a They are the gifts of one whose deeds are indefatigable, whose generosity is incessant,  3.b The gifts of one whose loftiness is assured, whose beneficence is a natural attribute,   4.a The gifts of a man of firm purpose, vigilant, in war both bold and staunch;   4.b The gifts of one wise and politic, a winner of hearts and minds.   5.a These are gifts that the Lord of the Throne has composed: what God has joined together cannot be put asunder;   5.b Gifts bestowed by the Lord of the Throne, neither niggardly nor won by chariness;  6.a Gifts whose protocol has been ordained by the Lord of the Throne who bestowed them, they are branded on the forehead of Time;   6.b Gifts against whose ruling, ordained by the Lord of the Throne, there is no appeal of any kind.   7.a They are gifts whose stock has been set by the Lord of the Throne, as firmly as the blessings in [Saladin’s] hand;   7.b They are gifts whose stock has been rooted by the Lord of the Throne in clemency and undeflected ardour;   8.a Gifts from whose stock, set by the Lord of the Throne, wafts sweetest musk;   8.b Gifts whose law, fixed by the Lord of the Throne, commands friends and enemies alike.

pi cture-poems f or sala d in  | 257   9.a Gifts whose law was so fixed by the Lord of the Throne that its benefits will surely endure;   9.b Gifts of stock planted by the Lord of the Throne, they have grown, and [Saladin’s] labours have grown still more. 10.a Gifts of stock planted by the Lord of the Throne, long will be the fragrant litany of their remembrance! 10.b Gifts of stock planted by the Lord of the Throne, long is the reach of their sheltering green branches! 11. Gifts of stock planted by the Lord of the Throne, they have grown into trees towering over mankind. This is performative poetry in several ways. In deciphering the figure, the reader not only finds out what the poem says and solves the riddle of the tree as a well-known symbol of noble lineage, but by completing the poem, if he writes down the answer as we have done in Plate 14.2, he creates a second figure, symmetrical with the first, making himself a poet, even something of a magician, in the mirror-image he has created. If the reader is Saladin, by reading the figure, he fashions his own dynastic tree, and conjures up his own praises, beholding in them his own ideal likeness.33 This is strong panegyric. When the poem is read aloud, the incremental anaphora has two effects: the solemn effect of a litany, linking Saladin, as ruler, to God, and the ludic effect of a memory-game, in which the section to be remembered and repeated grows longer and longer until all the combinations have been played out. These features are reminiscent of a practice which al-Jilyani, hyperbolically, attributes to the Franks in a poem which he sent to Saladin in his camp outside Acre in Safar 587/February–March 1191. The Franks, he declares, place in their churches Saladin’s image (sura), which they believe in as if it were one of the persons of the Trinity, and the priest uses its ‘description’ (wasf   ) in charms and writes Saladin’s [name] on healing talismans.34 The ludic aspect is to the fore in the second figure, fol. 79a of the Manchester MS (Plate 14.3), whose message, when deciphered, is on one level simply a robust lampoon on a detractor.35 This eight-pointed star-ina-circle, or ‘seal’, as al-Jilyani calls the figure, is a singleton, unconnected to other figures and lacking a transcript. It is not present in the Gotha MS, and was deciphered ad hoc by Peter Webb,36 whose translation I have slightly

258  |  syri a i n crusad e r time s polished. It is easier to read than the tree figure: its starting point is clearly marked, and all the reader has to do is proceed in straight lines, turning the figure, or walking round it, until, at the end of line eight of the poem, they find themselves back at the starting point. The red syllables or disyllables, of which there are four in each line at each intersection with another line, are sounds shared between words belonging to different lines. Introducing the piece, al-Jilyani says he composed it in 569/1173–4, when he heard that someone at Saladin’s court had been showing off his own eloquence and claiming that al-Jilyani’s kind of poetry was a one-off that he would be unable to repeat unless given ample notice. To refute him, in the presence of Saladin, al-Jilyani ‘extemporised’ this ‘Seal’ (ansha’tu hadha al-khatam … irtijalan),37 which is addressed to himself in praise of himself:38 1. Should anyone hasten to cast doubts on you, let him be; rather, pity him! for his own foolishness will strike the flint whose flame will consume him 2. Like the loser he is. It makes him quake to behold the celestial bodies of my stars; 3. Figure and circle, the Seal – the epigram our wit produces – is beyond him. 4. What is joining or even doubling figures to one whose fancy roams freely? 5. Be not envious of one noble by nature, for his talents are God-given. 6. Knowledge is a light in the heart which grows and increases; whoever impugns it merely disgraces himself. 7. Truth is revealed by those whose wishes are granted easily, by the superior person who is comfortable in the orbit of the arcane (majal al-sirr), 8. And the true elite are those whose deeds are their dignity, who preserve their courtesy even if their own inventions are disputed. Al-Jilyani’s, or his Seal’s, self-referential allusions to arcane truths and Godgiven talents might be taken as purely playful, since all we have here is a public conjuring trick. But al-Jilyani follows his Seal with an account of its reception in which self-praise is justified as praise of the patron, and which is full of Qur’anic quotations and echoes (a popular literary device known as

pi cture-poems f or sala d in  | 259 iqtibas, here highlighted in bold), by means of which the poet casts himself in the role of Moses confounding Pharaoh’s sorcerers and the Egyptian elders: When the elders examined this mudabbaja they were impressed and said: ‘By God, were you to cast down the rod of what you have said before the magicians of prose and verse, all that they profess would be swallowed up, and all their deeds would be in vain!’ Nay, this is a natural endowment, a strength, that I owe to the felicity bestowed on me by the Victorious King39 [Saladin]. Its purpose is to display to the world noble arcana, paragons that nevertheless fall far short of the deserts of the Victorious King. ‘Your Lord creates and chooses what He wills’ and ‘He has power over all things.’40

Is this playful, serious, or both? On one level, al-Jilyani’s utterances of this kind need to be understood within the emotional culture of his period and its ideas of selfhood and acceptable self-praise, something which would call for a separate study. But they also belong to the intellectual culture of the period, and it is in this light that I shall try to understand the mudabbajat and their ‘noble arcana’ as visual art and intellectual performance. Before I do so, a fresh query about physical performance arises. How could the Seal (however much it had probably been rehearsed beforehand) have been ‘improvised’? For its full effect to be admired, it requires not just reading aloud but drawing, and not just line (the outlines of the figure) but colour, to point up its clever use of shared sounds. It would also have had to be drawn on a large scale to make it visible to Saladin and his court. Al-Jilyani must have come equipped with a giant piece of paper, the means of tracing a circle and drawing straight lines, outsize pens or brushes, and coloured inks – all of which would have added to the anticipation and drama of the occasion. He must have been an excellent calligrapher to carry off such a performance: the copyist of the figure in the Manchester MS had some difficulty squeezing all the words legibly into their proper places. Al-Jilyani’s remarks about the last mudabbaja he composed for the dying Saladin stress that planning and execution were a huge and prolonged mental and physical strain.41 What was the point of so much effort in composing ‘arcana’, entailing no little effort for readers too, even with

260  |  syri a i n crusad e r time s instruction booklets (and perhaps al-Jilyani himself on hand) to help them? And what is the substance of the ‘arcana’? It is true that the mudabbajat enact a sort of ritual when they call down blessings on the dedicatee through litany-like anaphora, or other forms of sustained repetition or rich rhyme. But they are not, apparently, mystical. (Their content awaits detailed study, however.) Neither are they esoteric: they are not secret, since al-Jilyani shows how to read them, nor do they contain restricted knowledge. They are in fact an open book, even though at first glance they look like a closed one. But they can only be deciphered if the reader collaborates with the rules of the game they play; and reading them produces surprise effects through a kind of playful magic. The notion that best seems to fit here is that of ergodic literature, a term used by Espen Aarseth of works of verbal art in any medium that set out the rules for their own use, require a substantial effort of readers, and involve them actively in the construction of the text/artwork.42 First used to describe cybertexts, the term has been found to be a good fit with early modern Islamic esoteric, or superficially esoteric, texts that are visually patterned and coded and depend on rules and play.43 Ergodic writing seems to have been something of a trend in al-Jilyani’s period and part of the world. An example which may owe something to al-Jilyani is the table of recurrent dates and corresponding princely ­attributes, the latter apparently derived from those of Saladin, which Abu Salim al-Nasibi (582–652/1186–1254) included, with a key to its use, in his offering to the Artuqid ruler of Mardin, al-‘Iqd al-farid li-l-Malik al-Sa‘id.44 Al-Jilyani, as we have seen, says his mudabbajat are inspired by Saladin’s qualities and reflect them. Yet the sumptuousness of the mudabbajat, and the fun and aesthetic pleasure that one imagines would have come from their performance, are not what some modern scholars and medieval hagiographers have associated with the court of the devout and austere Saladin, of which, however, Eddé gives a livelier picture.45 The weight to be given to al-Jilyani’s contribution to Saladin’s court culture cannot be explored here; but in a wider perspective, the mudabbajat should encourage us to rethink the Arabic literature, and indeed art, of the Crusades. But the madabbajat are more than period pieces. The evidence is

pi cture-poems f or sala d in  | 261 accumulating not just that they fed into a widespread esoteric ergodic trend that is now attracting serious attention, but that they contributed over the centuries to a princely ergodic literature that has not been noticed in current scholarship. As I show in the companion piece to this, like al-Jilyani’s mudabbajat, later Arabic visual literature is strongly associated with offerings to princes.46 But the trajectory is not straightforward as concerns the mudabbajat themselves. Our mudabbajat MSS were recopied in the Mamluk and, probably, Ottoman periods in codex formats. This made the figures difficult, it some cases impossible, to read properly, obscured their function and impact, and suggests that they were not fully understood. Now, however, we have evidence that the mudabbajat were not merely recopied, but imitated, and that their ideological purpose was appreciated. A tutor and court companion to the future Moroccan ruler Sulayman b. Muhammad III (r. 1793–1822), Ibn al-Hajj went on pilgrimage at the end of the eighteenth century, taking time to meet men of letters and read manuscripts in the countries he passed through. On his return, he wrote several ergodic poems in praise of his sovereign. They directly imitate figures found in the picture-poems of al-Jilyani, and one, consisting of nineteen lines in the shape of a tree and in the metre tawil, (Plate 14.4)47 is calqued structurally on al-Jilyani’s tree-poem to Saladin. Here is evidence that some at least of the Arabic art and literature of the Crusades outlived them. Ibn al-Hajj’s tree-poem, which celebrates the patron’s beauty, ostensibly has nothing in common with its model, but when juxtaposed with it can be recognised for what it is: political poetry. Al-Jilyani’s mudabbajat are openly political. Now that we have seen how they might have attracted their dedicatees’ attention as works of art, their propaganda content can be studied in depth. Notes   1. Eddé, 2011, 356.   2. Al-Jilyani, 2010, 73–74 (see also the opening folios of his unpublished Kitab adab al-suluk); Ibn Abi Usaybi‘a, 1965, 635.  3. They are dated from 562/1166–7 to 599/1202–3. Of the numerous manuscripts, the four in Berlin are described in Ahlwardt, 1891, 184–185.

262  |  syri a i n crusad e r time s   4. Abu Shama, 1288 (1871), 2: 115.   5. Ibn Abi Usaybi‘a, 1965, 630.   6. For example, Badawi, 1954, 434, 443.   7. Latiff, 2018.   8. See Bray, 2019, text to notes 49, 50 and Murad and al-Sawwas, 1402 (1982), 298–301 for this nineteenth-century copy of Manadih al-mamadih.   9. Tornberg, 1849, 88; Pertsch, 1881, 270; Zirikli, 1992, 4: 167. Mingana, 1934, 940–941 is ambivalent, and Zetterstéen, 1927, 563 admires the pictures but not the poems. 10. De Slane, 1883–95, 554. 11. Hirschler, 2006, 68. He states, inaccurately, that it is included in Abu Shama’s Rawdatayn. 12. Al-Jilyani, 2010, Introduction, 9. 13. Harb, 2017, 136–141. 14. Yaqut, 1999, 2: 182, s.v. Jilyana. 15. Ibn Abi Usaybi‘a, 1965, 635. 16. Hirschler, 2016, 296. 17. Tadghut, 2018, 50, without indication of which folio bears this information. I am grateful to Maribel Fierro for drawing my attention to this article. 18. Mingana, 1934, 939–941. 19. The Paris manuscript can be accessed in black and white at (last accessed 2 July 2019); the Gotha manuscript has been digitised in colour at: (last accessed 2 July 2019); folios from the Jerusalem manuscript can be viewed in colour at (last accessed 2 July 2019) and the complete manuscript has been digitised and put online by the Hill Museum and Manuscript Library at (last accessed 2 July 2019). I am most grateful to Salam Rassi, who catalogued it, for this information. 20. See Bray, 2019. To the references given there should be added García Gómez, 1941, 409: a reference for which I thank Maribel Fierro. 21. Bray, 2019. 22. See the forthcoming papers of the conference ‘Islamic Occultism in Theory and Practice’, organised at the Ashmolean Museum, Oxford, 6–8 January 2017, by Liana Saif, Matthew Melvin-Koushki, Francesca Leoni and Farouk Yahya.

pi cture-poems f or sala d in  | 263 23. The final folio bears the date 745 (1344), but this may be the date of the MS from which it was copied. 24. I am grateful to the staff of the Carolina Rediviva for their helpfulness while the library was undergoing reorganisation. 25. See al-Jilyani, 2010, Introduction, 53–54, and fols 40a–46b of the Paris MS. I am most grateful to the staff of the BnF for their kindness and help. 26. Al-Jilyani, 2010, 75. 27. See Schaefer, 2006 and Alsaleh, 2014. 28. Olszowy-Schlanger, 2016. 29. Marina Rustow, ‘State institutions and daily life in the Mediterranean: evidence from the Cairo Geniza’, University of Ghent, 10 July 2017. 30. The study of Islamic paper at this period is still largely based on books and does not, Rustow says, reflect the range of types and formats found in documents. 31. The figure is illustrated in al-Jilyani, 2010, from the Manchester and Paris MSS, in unnumbered plates preceding 9, and again from the Manchester MS in unnumbered plates following 169, but the poem does not appear in the edited text. 32. Mawa and mu are spelled identically in Arabic. 33. As remarked by Abu Deeb, al-Jilyani, 2010, Introduction, 26, 29. 34. Ibn Abi Usaybi‘a, 1965, 631. 35. The figure is illustrated in al-Jilyani, 2010, 211, but the poem does not appear in the edited text. 36. With Tobias Nünlich and Julia Bray at the workshop ‘Picture-poems and the Esoteric in Sixth/Twelfth-century Syria: Working Out How to Read al-Jilyani’s Mudabbajat’, Oxford, 23–4 July 2015. 37. Al-Jilyani, 2010, 208–209. 38. The metre is basit. 39. Saladin’s Fatimid vizieral title. 40. Al-Jilyani, 2010, 212. Compare Qur’an 7:117, Q 26:40–45, and Q 5:17, Q 24:45, Q 28:68. 41. Al-Jilyani, 2010, 76. 42. Aarseth, 1997, 2, 179. 43. I am grateful to Nicholas Harris, PhD candidate in the University of Pennsylvania’s Department of Religious Studies, for telling me about his application of the concept. 44. Abu Salim al-Nasibi, 1310 (1892), 208, 210–211. The same table and key can be viewed online in an early fifteenth-century manuscript on fols 125a–126b,

264  |  syri a i n crusad e r time s 127b–128a of BnF arabe 2440. Available at: (last accessed 2 July 2019). 45. Eddé, 2011, 344–346, 352. 46. Bray, 2019, text to notes 33, 34 and 38. 47. Ibn al-Hajj, 1995, 549–550.

15 Ayyubid Realpolitik and Political–Military Vicissitudes versus Counter-crusading Ideology in the Memoirist–Chronicler al-Katib al-Isfahani* Lutz Richter-Bernburg

P

ractically from the moment ‘Imad al-Din al-Katib al-Isfahani ­(519–7/ 1125–1201)1 put his chronicles into circulation – even though at first only among his intended audience, the cognoscenti – he did not quite meet with the hoped-for reception. While his command of Arabic, emulating and striving to out-Hariri al-Hariri, elicited a certain admiration, it caused his readers embarrassment at the same time, an embarrassment apparently not felt vis-à-vis the not precisely plain prose of the Katib’s superior al-Qadi al-Fadil.2 A scant generation after his death, his fellow townsman al-Fath b. ‘Ali al-Bundari undertook, by cutting back the Katib’s effusive rhetoric, to make his work palatable to an educated but not equally obsessive readership.3 He achieved a modicum of success, but it was primarily through Abu Shama’s compilation Kitab al-rawdatayn (‘Book of the Two Meadows’) that the Katib’s historiographic afterlife was ensured.4 The notable attrition of transmission, although by no means unique but shared by other unusually voluminous works, is still to be regretted. Al-Bundari generally succeeded in preserving the factual core, but inevitably slipped here and there, and in any case, modern readers would like to decide for themselves what to focus on and what not.5 As it is, of the seven-part memoir–chronicle al-Barq al-shami (‘The Syrian Lightning’) only two sections are extant in a unique manuscript each.6 Even of al-Bundari’s epitome, Sana l-barq al-shami (‘The shine of the

266  |  syri a i n crusad e r time s Syrian Lightning’), only the first part has survived in a single copy.7 The Katib’s Seljuq history, which does not concern us here, has fared better; alBundari’s abridged version, Zubdat al-nusra (‘The choicest part of Succour’) is fortunately supplemented by a manuscript of the original, Nusrat al-fatra (‘Succour against the interval of lassitude’).8 The Katib’s works in political history mirror his and his forebears’ involvement with the powers that were; his Seljuq history was originally designed as a memorial to his uncle ‘Aziz al-Din, whom he idolised as a martyr. After the Katib went in 562/1167 to Syria into exile, he recorded the deeds of his two masters, Nur al-Din Mahmud b. Zengi and Salah al-Din b. Ayyub, as well as his own literary activity, in ‘Lightning’, which title incidentally does not refer to military exploits or other swift actions, but appropriates from Abu Tammam a metaphor of short-lived brilliance.9 Saladin’s final years, beginning with the triumphal campaign of 583/1187 and continuing through the peace of 588/1192 to beyond his death the following year, are the subject of al-Fath al-Qussi fi ’l-fath al-Qudsi (‘The inspiration of a Quss concerning the conquest of Jerusalem’). Actually, both titles proclaim the author’s celebratory attitude, the ‘Lightning’ flash as well as the reference to Quss, the semi-legendary embodiment of pre-Islamic Arabic eloquence. But even more fundamentally, al-Fath al-Qudsi is by itself suggestive of divine support and illumination; on the one hand, it alludes to the early Islamic conquests, but in addition to the reference to Jerusalem, qudsi may intimate a supernal emanation illuminating the author’s mind.10 According to him, it was his benevolent colleague and superior al-Qadi al-Fadil who suggested adding the first colon al-Fath al-Qussi to the original formulation;11 in this way the title displays a variety of the Katib’s preferred figure of speech, paronomasia, along with tarsi‘, ‘isocolon’; al-fath occurs twice, but with different meaning, ‘inspiration’ and ‘godly conquest’, and the two nearly identical nisba forms qussi and qudsi are tied together by partial paronomasia. Further, the pair Quss and al-Quds allude to another of the Katib’s pet conceits, that it takes the pen for the sword’s exploits to survive or in other words, that the pen is more powerful than the blade.12 Now, the Katib was too much of a professional, to follow Donald Richards,13 or differently put, his self-image was too lofty, for him merely to compose a panegyric, which could then be passed off as literary hyperbole. He

re a lpoli ti k and vi ci ssi tud e s  | 267 did hold the grand style to which he was accustomed and on which he prided himself to be the indispensable, indeed the only adequate, medium for what he set out to record, namely history, his two masters’ noble deeds; however, the writing of history entailed a commitment to veracity. By harnessing the two together, ornate rhetoric and historiographic truthfulness, he strove to create a new genre in Arabic, prose epic. In the event, the Katib’s endeavour was not crowned by success. Historiography generally remained rather more sober or pedestrian in Arabic, and even when, at the hands of chancery officials, it did not, the stylistic model to emulate was the Qadi al-Fadil.14 But apart from literary aspects, the question of the Katib’s dependability as a chronicler remains. Elsewhere I have argued that his was a conflicted personality; on the one hand, his outlook was from early on informed by a certain Manichaeism; the side he took was always in the right and well-nigh infallible, whereas opponents were in error and corrupt to the core.15 People who did not concern him or he did not allow himself to be concerned about are treated with callous indifference – and be assured that the standard to which the Katib is held here is by no means anachronistic but amply documented in contemporaneous sources.16 Yet it goes without saying that there was another side, or more than one other side, to the Katib; he was too perceptive not to notice common human frailty or simply the refractoriness of reality, even if on occasion he could just barely bring himself to admit as much. Here a distinction has to be made between the more tightly structured and more official work al-Fath and the more loosely organised memoir-chronicle al-Barq.17 The latter, ‘Lightning’, was only completed about five years after Saladin’s death, when the author’s relations with the new regime of al-Malik al-Afdal and especially his vizier had soured and he lived in Damascus in forced retirement.18 The greater liberty he allowed himself in ‘Lightning’ to voice criticism expresses itself in a few choice observations which considerably shade the account given of the same time span in ‘Inspiration’. Further, and more than ‘Inspiration’, where the inserted insha’, chancery documents, came all from the Katib’s own pen, ‘Lightning’ provides intertextuality through the regular inclusion of correspondence by the Katib’s great professional ‘other’, al-Qadi al-Fadil. While their literary interests and ideals may have been similar, they evidently were of different temperament; other than the driven Katib, the Qadi appears to have been at peace with himself,

268  |  syri a i n crusad e r time s self-assured in action and serenely resigned to life’s vicissitudes.19 Some of the most incisive testimonies on the Katib as well as Saladin are owed to the Qadi, as we will shortly see. But even apart from concrete detail, the mere time frame of ‘Lightning’, covering Saladin’s entire career, automatically casts its subject in a less flattering light than ‘Inspiration’, which conveniently only begins in 583. Thus, Saladin’s long, drawn-out campaigns against his Muslim rivals, primarily the Zengids of Aleppo and Mosul, are relegated to the sidelines – or so the Katib may have tried to convince himself. However, memories were not in all quarters as short as Ayyubid propaganda might have hoped for. Later in the year the caliph, unfazed by Saladin’s successes, made this abundantly clear, as ‘Inspiration’ cannot avoid recording; in his usual fashion, though, the Katib at the same time is trying to obfuscate the issue.20 While the proximate cause of caliphal indignation may have been a blunder of protocol – sending a junior envoy of dubious standing – it was only the last drop to make the cup run over, or it was merely adding insult to injury. As the Katib’s defence amply proves – never mind his insinuation of intrigues at the caliphal court – the core issue was Saladin’s expansionist policy, of which the Counter-Crusade was but part. Naturally, in Saladin’s fictional apology in ‘Inspiration’, this is only acknowledged indirectly by way of implicit denial; but the length to which the Katib goes points to the gravity of the risk he and Saladin saw in an open rift with the caliphate.21 The Katib credits himself with restoring relations through the deft diplomacy he lent to Saladin’s response.22 In ‘Lightning’, as far as extant in ‘Shine’, an additional perspective is opened through quotations from al-Fadil,23 but it is Abu Shama, who provides a fuller view of the issue, quoting at length from the caliphal major domo’s missive to Saladin and calling on Ibn al-Qadisi for the controversial envoy’s background.24 Complex and contradictory political or military reality, of which the heightened tension between the caliphate and Saladin in 1187 is but one example, is narratively represented to noticeably different degrees of comprehensiveness even in the pro-Ayyubid sources here used, the Katib, with ‘Inspiration’ and ‘Lightning’, and Abu Shama. Usually avoiding outright denial, the Katib addressed unpleasantness or embarrassment in an indirect, allusive or evasive manner. However, his account of the mentioned issue, relations with the caliphate, also shows that even in ‘Inspiration’, his most

re a lpoli ti k and vi ci ssi tud e s  | 269 panegyric prose work, shadows are not completely whitened out. ‘Lightning’ achieves, as observed, a more nuanced and, as it were, polyphonic presentation by including other voices; al-Qadi al-Fadil’s, whether on Saladin’s behalf or in his own name, features most prominently among these, but on occasion other literary correspondents of the Katib’s are also given the rostrum.25 The Katib’s presentation of the Battle of Montgisard in 573/1177 illustrates his actual if limited critical and self-critical capacity as well as Saladin’s and his courtiers’ propagandistic and diplomatic, mildly put, re-casting of the events.26 The distance of approximately two decades still left him defensive about his own less than heroic desertion, which stood in marked contrast to al-Qadi al-Fadil’s – by the Katib readily acknowledged – valiant efforts.27 His express criticism of military leadership before battle is narrowly limited to pointing out over-confidence in numerical superiority, while he indeed does mention the troops’ dispersal on plundering raids. As for an appraisal of the consequences, even towards the middle of the 590s the Katib is not able to draw the obvious conclusion, succinctly formulated by Ibn Shaddad, that the defeat at al-Ramla in 573/1177 was only made up for ten years later by the victory at Hattin.28 When dealing with adversity, other than success, the Katib’s temporal compass frequently narrows as if to contain negative consequences.29 Thus in ‘Lightning’, his implicitly final assessment of al-Ramla dates from a few months later, not from the vantage of twenty years; in a letter to Saladin, al-Qadi al-Fadil claims the consequences of al-Ramla were much harder to bear for the Franks than such a minor temporary setback was for the Ayyubid side.30 The fragmentary transmission of both ‘Lightning’ and al-Bundari’s abridgement, ‘Shine’, permits comparison with ‘Inspiration’ primarily to the extent of their overlap, the respective accounts of the year 583, and beyond that indirectly, by way of Abu Shama. The notorious foray into the Red Sea by a detachment of Reynald of Châtillon’s troops ended in a massacre of the captured force in Cairo, while two men were slain as sacrificial victims in Mina. Abu Shama does reflect, if not without distortion, the legal dispute between Saladin and his brother and vicegerent in Egypt, al-‘Adil, about the killing of prisoners apparently promised quarter; the internal correspondence he quotes – possibly from ‘Lightning’, if not directly from al-Qadi al-Fadil – provides inestimable witness.31 In ‘Shine’, only the all-too-brief statement

270  |  syri a i n crusad e r time s that al-‘Adil acted at Saladin’s command alludes to the preceding dispute. Perhaps comparable as an instance of conflict between the noble norms of jihad and political–military expediency are the repeated massacres of ‘regular’ Frankish prisoners, which were even staged as entertainment; pious men of religion were invited to act as executioners. In 574/1179 the Qadi al-Fadil objected, whereas the Katib demurred, pleading his civil calling – which meant he did not want to cut a laughable figure.32 In 583/1187, after Hattin, Saladin had all the Knights Templar and Hospitaller butchered; on this occasion, the Katib appears as an enthusiastic spectator and reporter in ‘Inspiration’;33 in ‘Lightning’ – via ‘Shine’ – he stresses Saladin’s utilitarian, military motives on the one hand – the military orders were the backbone of the Crusading fighting forces; their refusal to have funds allocated to their ransom meant that to provide for them was not lucrative, nor could they be put to work. On the other hand, the Katib does feel obliged to add, nearly as an afterthought, that all the victims were given the choice of conversion or death, thus putting a veneer of legality on the killing.34 It is a truism that the reality of jihad at times rather clashed with than followed the norms ideally governing it. Nor can it come as a surprise that the Katib had a hard time admitting obstreperous reality into the harmonious, as if stencilled, flattened-out two-dimensional picture he may have wished to present. Deferral to God’s ineluctable pre-ordainment was always, if on occasion all too conveniently, an option in order to attenuate misadventures or failures, which it did not take as ill will, but merely precipitate or shortsighted negligence or misjudgement to produce. A case in point is the Katib’s obfuscating evasiveness concerning the ultimately elusive capture of Tyre in the summer of 1187.35 While a straightforward, thus implicitly critical account of the sequence of (in-)actions and events may not have been easily squared with the encomiastic intent of ‘Inspiration’, in ‘Lightning’ the author’s greater temporal and mental distance from the subject at hand – about half a decade after Saladin’s death – might have produced a more sharply drawn, thus more convincing outline rather than the sanctimonious resignation to divine decree. Still, his triple invocation of God’s pre-ordinance in ‘Lightning’ is a clear enough sign that even then the issue of Tyre did not leave him indifferent.36 Yet as far as can be gathered from ‘Shine’ and perhaps for sheer carelessness, he by and large adhered to the lacunary chronology

re a lpoli ti k and vi ci ssi tud e s  | 271 of ‘Inspiration’, which is to say, a mention of Tyre as an object of attack is postponed, actually relegated to a flashback, until after the occupation of Beirut and Saladin’s turning back from his northward campaign along the coast.37 At that date, after the first week of August (Beirut was occupied Thursday, 29 Jumada I/6 August), when Conrad of Montferrat’s efforts at strengthening the town’s defences – on which more below – had borne some fruit, Saladin’s decision to spare Tyre until later was no longer an embarrassment, but seemed plausible in view of the greater prize to strive for – as the Katib has him put it – Jerusalem.38 Only at this point, in retrospect, does the Katib acknowledge an opportunity apparently missed earlier on, in his words during the interval between Raymond III’s departure and Conrad’s arrival; a few sentences earlier he had still alluded to this opportunity as ‘seemingly seized’.39 The obvious question here concerns the dating of such a ‘window of opportunity’. After the occupation of Toron (Sunday, 18 Jumada I/26 July), Saladin marched on Sidon by way of Sarepta, but since the Katib had him escort the garrison of Toron to Tyre, he apparently did not then consider it his next target. In any case, the Katib ignores Tyre completely at this point, pretending, as it were, it did not exist, not wasting a word of comment on Saladin’s choice of itinerary and bypassing Tyre.40 Fortunately, the ‘economy’ of the Katib’s narrative can to some extent be supplemented by Crusader sources on the one hand, but ‘closer to home’, by the Katib himself;41 at one remove, the Katib’s account is also accessible through Ibn al-Athir’s abrégé in al-Kamil (‘Chronicon … p ­ erfectissimum’). At the very least, it represents Ibn al-Athir’s reading of the Katib’s studied ambiguity and blurredness.42 Further, in a despatch from Saladin at Acre, right after its occupation, to the caliphal diwan in Baghdad the Katib as actual drafter announced that Taqi al-Din ‘Umar, Saladin’s nephew, was going to invest Tyre and Tibnin (Toron), whereas al-Malik al-‘Adil was at Saladin’s behest on the march up the coast from Egypt and going to invest Gaza and Ascalon; Saladin himself was setting out for Jerusalem.43 The fact that there is as yet no mention of al-‘Adil’s – imminent – successes in Palestine, the capitulation of Majdal Yaba (Mirabel) and the taking of Jaffa by storm, confirms the date indicated in the opening sentence.44 In addition to transmitting, with fulsome self-praise, the news of the victory at Hattin and the ensuing conquests, the letter draws an outline of Saladin’s and

272  |  syri a i n crusad e r time s al-‘Adil’s forthcoming campaigns which was soon to be rendered obsolete; neither was Saladin immediately headed for Jerusalem, nor al-‘Adil for Gaza and Ascalon, nor did Taqi al-Din besiege Tyre – or if he did, then unsuccessfully, which explains the Katib’s evasiveness. In the event, Taqi al-Din’s problems at Toron led to Saladin’s intervention there and his further march north, sidestepping Tyre, while al-‘Adil secured central Palestine without attacking the two Philistian ports. In mid-August Saladin joined forces with al-‘Adil; together they reduced Ascalon and occupied all those places which might otherwise have threatened supply lines along the route to Jerusalem. Thus, in early autumn, all the targets mentioned in the above-cited letter to Baghdad were accounted for – with the exception of Tyre. We saw earlier that Tyre became an uncomfortable issue; thus, it was glossed over in the Katib’s subsequent chancery correspondence and narrative accounts. In Saladin’s well-known year-end survey letter to his brother and deputy Sayf al-Islam in the Yemen, the Katib attributes Tyre’s successful defiance to seizing the opportunity of Saladin’s preoccupation with ‘her sisters’.45 Compared to this facile self-exculpation without any further explanation, the Katib’s later, thrice-reiterated resignation to God’s ineluctable decree in ‘Lightning’, appears to reflect, even if indirectly, a measure of self-criticism. Saladin’s opportunity to occupy Tyre during the ‘interregnum’ between Raymond III of Tripoli and Conrad of Montferrat, obliquely alluded to by the Katib, also figures, with much embellishment – and vital distortions – in Ernoul46 and subsequent compilers of the continuations of William of Tyre. Without delving into the problems of this text bundle, it agrees with or at least does not contradict, the Katib’s account on a few essential points: on the one hand, stress is laid on the strength of Tyre’s fortifications and the manpower flocked there, consisting of rank-and-file fighters as well as barons. On the other hand, after Raymond’s flight to Tripoli, a lack of leadership and resolve – against the wishes of the burghers – is betrayed by negotiations with Saladin about terms of surrender, which were at the very last moment thwarted by the arrival of Conrad of Montferrat. The narrative takes as a given that apart from this fleeting moment of weakness Tyre remained too strong to become easy prey for Saladin. The overlap between the post-William corpus and the Katib’s vague allusions to opportunities ‘seemingly captured’ is too marked to be easily

re a lpoli ti k and vi ci ssi tud e s  | 273 dismissed; it is further corroborated by Ibn al-Athir, regardless of whether he fitted this piece of information as an extraneous addition into his abridgement of the Katib. Additional evidence, even if only extant in a later revision, is furnished by the Templar preceptor Terricus, who speaks of the town’s siege by the ‘Turks’ in his letters of appeal for help to a variety of European addressees.47 The question which still awaits an answer concerns the dates of the Tyrian interval between Raymond III and Conrad.48 A date as early as mid-July would seem to be suggested by the Katib himself and Ernoul, not to mention, from a different filiation of reports, William of Newburgh.49 However, the archival evidence is so scanty as not to permit any dependable conclusion either way. Conrad’s first Tyrian charter only dates from October, which means an obviously inconvenient, if possibly not lethal, hiatus of nearly two months since Raymond’s latest acts at Tyre.50 To sum up, according to the Katib and in his wake, Ibn al-Athir, Tyre was ‘destined’ to remain unconquered in the summer of 1187, even though it was within easy reach or even as good as occupied, had not the unforeseen arrival of Conrad of Montferrat turned things around at the very last moment.51 In both ‘Inspiration’ and ‘Lightning’, Raymond (III) of Tripoli’s escape from Hattin to Tyre is reported, and his further flight to Tripoli, when he learnt of Saladin’s approach. Raymond had good reason to expect an attack, and, in the Katib’s words in ‘Lightning’, judiciously decided to head for safety. Here it may be in order to counter a possible objection to our reading of the Katib’s dual account of Saladin’s campaign northwards along the coast of Palestine and ‘Phoenicia’. After the capitulation of Beirut, of which the Katib drafted the contract, he went on sick leave to Damascus and only returned the day of the capture of Jerusalem, nearly two months later. Thus, he depended on reports from participants in the campaign for his own record. However, given his position as one of Saladin’s closest and most trusted courtiers, it would seem inconceivable that he should not have received all the information necessary in order to continue his account at the previous level of detail. But actually, as we have tried to demonstrate, the crucial omission within a wealth of information occurred earlier, at a time around the rendition of Tibnin when the Katib was still in attendance. Certainly, in addressing the caliphal administration in Baghdad, on Saladin’s behalf the Katib speaks of the sultan’s nephew Taqi al-Din ‘Umar as besieging both Tibnin and

274  |  syri a i n crusad e r time s Tyre – the implication being that victory is expected within a short time. In the event, it took considerably more effort than expected to reduce Tibnin, not to mention Tyre, but that alone may not exhaustively explain the Katib’s sudden vagueness about Tyre. By the time of writing ‘Inspiration’, it had become more important to embroider, with some glee, on the ‘count’s’, that is, al-qumis Raymond’s, unfortunate end than to report straightforwardly on Saladin’s decisions, or indecisions, concerning Tyre. The Katib’s wording would suggest that the town was thought to be taken – had it not been for the sudden appearance of that other devil, the ‘marquis’ (al-markis) Conrad of Montferrat. In ‘Lightning’, at least as abridged by al-Bundari, the Katib invokes the divine decree three times,52 first concerning Conrad’s arrival at Tyre, then Saladin’s ‘passing by’ it, and finally at the chapter end, in summing up the consequences of the neglect of Tyre in favour of Jerusalem. In both ‘Inspiration’ and ‘Lightning’ the Katib stresses Saladin’s expectation that the town would fall to him as easily as had all the other places he had invested or made a show of investing after Hattin. In ‘Inspiration’, Saladin’s lackadaisical attitude is twice alluded to as an impersonal ‘it was opined’ (zunna), whereas in ‘Lightning’, his agency is glanced over first as missed opportunities and second as unconcern and indifference in favour of the coveted goal, Jerusalem. However, two qualifications have to be made: in ‘Inspiration’, the end-of-year survey letter to Sayf al-Islam in the Yemen provides an additional perspective, and of ‘Lightning’, it cannot be ruled out that its original version was the source from which Ibn al-Athir and Abu Shama lifted some of their precious excerpts. Saladin’s expansionist policy could not but cause tension with the ‘Abbasid caliphate. If in ‘Inspiration’ the Katib could accept the substance of the caliph’s censure only in a backhanded way, in ‘Lightning’ he was more forthright, even if so to speak, at a remove. He resorted to the technique of hiding behind an unimpeachable witness, and which witness could have commanded more respect than the Qadi? The occasion is Saladin’s killing of Reynald of Châtillon with his own hand or rather, the sultan’s vow to kill him. In the Katib’s account, he could not fathom Saladin’s motives and apparently did not feel he had the requisite rank to broach such a sensitive question to the Qadi, who as Saladin’s closest political adviser could have

re a lpoli ti k and vi ci ssi tud e s  | 275 been expected to know.53 In the event, an old literary acquaintance of the Katib’s, ‘Abd al-‘Aziz b. Shaddad b. Tamim ‘b. Badis’, was obliging enough; the difference of their backgrounds, Ibn Shaddad descending from, as it were, a princely family of Ifriqiya and the Katib from the Arabicised Iranian notability, does not seem to have precluded them achieving – over time? – fairly intimate terms.54 Ibn Shaddad thus passed on to the Katib what the Qadi had told him: he himself had exacted the vow from Saladin the previous year when slow recovery from grave illness made him feel vulnerable and amenable to admonition. The pledge was to abstain from fighting other Muslims but instead to devote himself exclusively to fighting the infidel and in the case of success, personally do away with Reynald55 and Raymond (III) of Tripoli.56 The Qadi’s and doubtless his peers’ hatred can be read as a tribute of sorts to the two influential antagonists in the Latin kingdom – regardless of how their perceptions would be assessed. But what is really worth noting is the Katib’s ready agreement to the distanced view the Qadi takes of Saladin’s course of action up to that point in time. It is no surprise that such fundamental criticism is not found in ‘Inspiration’, which, after all, was directly presented to the sultan. A good example of more narrowly circumscribed, financial rather than political critique is furnished by the Katib’s report on the embezzlement of funds after the capture of Acre in 1187;57 in ‘Lightning’, the sultan, upon hearing the pertinent section of ‘Inspiration’ read to him in Jerusalem in 1192, is said to have remarked that responsibility lay with three untouchable individuals, of whom two had been enfolded in God’s mercy in the beyond and one enjoyed immunity in this life. This was al-Malik al-Afdal, whom the author then exonerates, shifting blame on his entourage. In ‘Lightning’, the author reports the plundering of Acre to the detriment of the sultan’s coffers in substantially greater detail than in ‘Inspiration’; in addition, he mentions his own considerable loss at the hands of marauding troops. Generally, he is much concerned about his own emolument and easily upset at interruptions of cash flow; actually, in the following year, 1188, an acute shortage of funds momentarily has him entertain the thought of quitting Saladin’s service altogether.58 Whether it was on the strength of his own better sense or the Qadi’s gentle admonition or a combination of the two, he did not act on the impulse; but so much for the principledness of Saladin’s eulogist-in-chief, to quote Lyons and Jackson – and reserving judgement on

276  |  syri a i n crusad e r time s the overall applicability of this moniker.59 However, thanks are again due to the Qadi; his correspondence, which, as we have seen, the Katib also put to good use, preserves more of the nitty-gritty of private and public life than any other contemporaneous source. Needless to add that in ‘Inspiration’ the author exclusively cites health reasons for his extended absence from service. Saladin’s own spendthrift habits, repeatedly justified both as pious and politic, are again presented from at greater distance in ‘Lightning’, where the sultan’s brother al-Malik al-‘Adil is called in as witness for the author’s side. Material as well as symbolic gains were upmost on the Katib’s mind when evaluating the outcome of battles and conquests. Generally, he disapproves of looting and destruction, for they mean losses to the treasury as well as to Islam’s symbolic capital; occupied churches should not be despoiled of precious materials, because rededicated to Muslim use, such buildings would contribute to the glory of Islam.60 While the Katib’s celebration of jihad is an obvious given, he himself would rather stay out of the heat of battle; either he finds procurement too costly or he takes to flight precipitously – to his own subsequent embarrassment.61 He also stays out of the macabre butchering of captured Knights Templar and Hospitaller, which Saladin staged both after Montgisard in 1177 and after Hattin ten years later.62 However, the Katib’s deportment is motivated primarily by an awareness of his own limitations; not being a swordsman, he does not want to cut a poor figure as executioner. A bit like an afterthought, he adds that killing is not his job. To raise questions of principle is left to the Qadi. Of the massacre of 1187, in the wake of Hattin, the Katib gives two differing accounts; in ‘Inspiration’, he hails it as a noble deed in the service of Islam. In ‘Lightning’, which was completed nearly a decade later, a note of defensiveness may sound in the closing remark that the Templars and Hospitallers were not killed before having been offered and refused conversion to Islam. Furthermore, he puts a mundane consideration into Saladin’s mouth: their refusal to be ransomed meant that keeping them in captivity did not promise any profit. In 1191, after the fall of Acre, Richard the Lionheart had the captured townspeople massacred wholesale, which Muslim witnesses unhesitatingly construed as breach of contract.63 While the Katib’s judgement of the

re a lpoli ti k and vi ci ssi tud e s  | 277 ‘accursed’ Franks’ treacherous action is unequivocal, the event does not move him to an extended flourish of evocation (wasf   ). Nor does he deign to speculate about the king’s motives, as Baha’ al-Din b. Shaddad did.64 Rather, he expresses ready assurance of the victims’ reception into Paradise, which appears to have settled the matter for him. Perhaps there was an element of cynical calculation involved in the Katib’s equanimity in that he did not see the military balance measurably affected, once the city had been lost anyway. In a diplomatic context, though, the Katib utilised the death toll exacted to lend urgency to the sultan’s appeal for support.65 The carnage of the Battle of Hattin inspired the Katib to one of his virtuoso set pieces of wasf, as it were ululating the triumph of the faithful over the infidel.66 In ‘Lightning’ this is more muted and summed up in a Qur’anic verse the author claims to have recited on the spot: ‘And the unbeliever says were I but dust.’67 Perhaps the experience of horror toned down the strident triumphalism a bit. A similarly extreme coloratura insert is devoted to women among the Frankish forces before Acre in 1189.68 Not surprisingly, their presence fundamentally challenged all norms of gender roles and relations as known to and upheld by the Katib. Also, it gave him the welcome opportunity to demonstrate Frankish depravity in a particularly telling instance. Actually, it would be cause for wonder if he had not willingly responded to this occasion. Nevertheless, the well-nigh pornographic lengths the Katib goes to have an obsessive ring to them – and the question is, is it only obsession with his favourite figure of speech, paronomasia (tajnis), or also with women as constructed in the Islamic imaginaire. The latter would seem more likely since he displays a similar attitude to the unredeemed women captives in Jerusalem. As for the Frankish women of pleasure during the siege of Acre, he does allow that not all of Saladin’s fighters were able to resist their lure, but, of course, he has no problem relegating this regrettable transgression to the level of deluded simpletons. Unfortunately, other than regarding his evocation of the battlefield of Hattin, here later authorial revisions in ‘Lightning’ cannot be ascertained because of loss of the corresponding section and Abu Shama’s exclusive reliance on ‘Inspiration’. Abu Shama’s extensive quotations show that the subject of the Frankish women was of much interest to him as well, but at the same time, his pared-down version is devoid of the Katib’s prurience.69

278  |  syri a i n crusad e r time s Above, internal ‘intertextuality’ was mentioned as a feature of ‘Inspiration’ and even more so of ‘Lightning’, where al-Fadil’s official and private correspondence is regularly quoted. However, the time-honoured device of having the dramatis personae take a fictive rostrum also allows the Katib to express diverse views without clearly taking position himself. Especially with respect to sensitive, controversial issues this technique commends itself. Illustrative examples are the failed siege of Tyre in the fall and early winter of 1187 and the debate about concluding a truce with Richard the Lionheart in 1192. In either case, the Katib avails himself of the possibilities of a multi-actor, ‘dramatic’ representation in order to give voice to a variety of points of view while himself maintaining a certain reticence. As concerns the failure before Tyre, he does not hold back his criticism of subordinate agents, which spares him the articulation of a more fundamental political–military self-examination.70 The conclusion of a ‘truce’ – legally, no ‘peace’ treaty was permissible with a non-Muslim dominion – with the hated enemy, Richard the Lionheart, instead of achieving decisive victory, was a bitter pill to swallow for the Katib as the Franks’ implacable enemy, and a fortiori his master, Saladin. While nobody and least of all the Katib himself can have been blind to military and economic realities, he delegates their acknowledgement to Saladin’s amirs, in this way seeking to preserve his master’s and his own pride and unyielding devotion to jihad.71 However, for a change he does rein in his habitual scorn for the Franks and matter-of-factly reports on their joyful pilgrimage to Jerusalem after the treaty took effect.72 In comparison with other contemporaneous writers, the Katib may be the most staunchly hostile toward the Crusaders, which is in no way to suggest that authors such as Usama b. Munqidh,73 ‘Ali al-Harawi,74 Ibn Jubayr75 or Baha’ al-Din b. Shaddad76 were more accommodating on principle; they were just ever so much more prepared to perceive individual variations, the well-nigh infinite multifariousness of humanity. The Katib’s tendency toward a stereotyped, quasi-Manichaean perspective constrained the scope of his historiographic outlook, regardless of the vantage his administrative position afforded him. The best he could achieve in acknowledging the limits and fallibility of even the best-intentioned human actions was grudging resignation. The very fact that his writing permits of such observations is a tribute to its quality.

re a lpoli ti k and vi ci ssi tud e s  | 279

Notes   * Since no proper transcription of Arabic is permitted in this volume, I will limit myself to the barest minimum of Arabic references and quotations and regularly resort to English translations instead. Regrettably, readers will have to check, on the basis of the precise references provided, the corresponding Arabic for themselves.   1. Abu ‘Abdallah/Hamid Muhammad b. Safi al-Din al-Faraj Muhammad b. Nafis al-Din al-Raja’ Hamid (519–97/1125–1201) will, for simplicity’s sake, here mostly be referred to by his professional moniker, al-Katib, (‘the chancery official’, ‘secretary’); for his biography see Richter-Bernburg, 2014, esp. 29–41; Richter-Bernburg, 1998, 25–136. Cf. for a brief outline, Eddé, 2008, 162f., 611; also, for the Katib’s views of the Franks, Kedar, 1997a, esp. 137 and 140; Kedar, 1997b, esp. 120; Kedar, 1998, esp. 131; Kedar, 1996, esp. 347–352.   2. As illustrated by the solid manuscript transmission of his chancery prose (insha’  ) as well as verse; regrettably his ‘journals’ (mutajaddidat), possibly for their prolixity, did not fare as well. Al-Qadi al-Fadil, a Fatimid style for a higher echelon administrator, has remained ‘Abd al-Rahim b. ‘Ali al-Baysani’s (539–96/ 1135–1200) moniker (shuhra) to this day. For brief outlines, with refs, see Eddé, 2008, 160ff., and Smarandache 2015.  3. See here instantly and Durand-Guédy, 2013; cf. Richter-Bernburg, 2014, esp. 43f.   4. Cf. Antrim, 2009.   5. Richter-Bernburg, 1998, 176–82; cf. note 8 below.   6. Cf. Richter-Bernburg, 1998, 259–383, arab. 5–99 [German tr. and Arabic ed. of section of year 573].   7. The second section, which commences with the turn of 584/March 1188, only survives fragmentarily – if indeed it does – by way of quotations.   8. Durand-Guédy, 2005, 2006; Peacock, 2013.   9. Richter-Bernburg, 1998, 133f, 194; cf., critically, Eddé, 2008, 296. 10. For its denotations and connotations in Avicennian philosophy, see, for example, Gutas, 2014, 179–201, esp. 182ff. 11. ‘Inspiration’, 12:1–4. Be that as it may, the Qadi certainly cannot be taken to represent even a readership of above-average education – which one would be inclined to identify as al-Bundari’s target audience; ‘Shine’, 3:9–13, 4:3–17. 12. Obviously, the conceit did not originate with the Katib (merely, for example,

280  |  syri a i n crusad e r time s see Richter-Bernburg, 2014, 42, n. 63). If subsequent transmission be trusted, it found a telling contemporaneous application, namely to al-Qadi al-Fadil, by none other than Saladin himself. Reportedly he used to say, ‘don’t think that I came to rule the land by your swords, but by the pen of al-Fadil’ (Sibt b. alJawzi, ‘Mirror’, VIII: 472; Ibn Taghribirdi, 1963–71, VI: 157:1ff.; cf. Musawi, 2006, esp. p. 107f.; Gully, 2005). 13. Richards, 1993. 14. See Musawi, 2006, and Bauden, 2017, esp. 126b, 127a–b. 15. Richter-Bernburg, 2014, esp. 42f. 16. Richter-Bernburg, 1998, esp. 30ff., 81–84, 110f., 179f., 240–243. 17. Criticism in ‘Inspiration’, where he does allow himself to voice it, is more narrowly circumscribed; see below, with n. 21. 18. Richter-Bernburg, 2014. 19. Eddé, 2008, 160ff., 610f.; cf. Richter-Bernburg, 1998, esp. p. 224f. 20. Eddé, 2008, 122f. 21. ‘Inspiration’, 95:18 – 101:14. 22. Internal policy debate in Saladin’s circle is briefly mentioned in ‘Inspiration’, but without naming names; these are supplied – from ‘Lightning’? – by Abu Shama, ‘Two Meadows’, III: 420:16ff. 23. See ‘Shine’, 424:5 – 425:6; cf. 413:20 – 414:3, which barely touches on the contentious issues. 24. ‘Two Meadows’, III: 421–423; Muhammad b. Ahmad b. Muhammad ‘Ibn ­ al-Qadisi’ (d. 632/1235) reportedly continued a consecutive series of ­ chronicles, in succession to al-Tabari, down to AH 616 (cf. Rosenthal 1968, 73f.). 25. For example, ‘Lightning’, III, fols 9b–11a; see Richter-Bernburg, 1998, 212f., arab. 18–20. 26. Lightning’, III, fols 12a–19b, 22b–23a; see Richter-Bernburg, 1998, 213–24 and arab. 22–32, 37. 27. ‘Lightning’, III, fols 8b–12a, 15b–16a; see Richter-Bernburg, 1998, 210ff. and arab. 16–18, 26:17–27:5. 28. apud Abū Shāma, ‘Two Meadows’, II: 466:3–13; cf. Richter-Bernburg, 1998, 213f. 29. See Richter-Bernburg, 1998, 207ff. 30. ‘Lightning’, III, fols 39b–40a; see Richter-Bernburg, 1998, 224f. and arab. 59:13–60:2. 31. See Lyons and Jackson, 1982, 186–188, 411f.; cf. ‘Shine’, 280:1f.; Abu

re a lpoli ti k and vi ci ssi tud e s  | 281 Shama, ‘Two Meadows’, III: 135:1ff., 137:6–10 (from al-Qadi al-Fadil [via ‘Lightning’?]), 137:11–17, 137:18–138:2. 32. ‘Lightning’, III, fols 97r–100v; ‘Shine’, 192f.; Lyons and Jackson, 1982, 132 with fn. 59 (al-Qadi al-Fadil, MS BritLib Add. 7307, fol. 65). 33. ‘Inspiration’/1888, 28f.; ‘Shine’, 379:5–22; Lyons and Jackson, 1982, 265. 34. Cf. Eddé, 2008, 251. 35. The failure of its siege in late fall is a different matter; cf. Eddé, 2008, 256–258; Möhring, 1980, esp. 58–61 (his reconstruction of the situation of Tyre between July and October 1187 does not impinge on his account of subsequent developments). 36. ‘Shine’, 389:13f., 390:4, 8f. 37. At that point, he mentions a previous passing by Tyre, ‘as narrated’ (wa-kanat [!] lamma ‘abarna biha ‘ala ma hukiya amruha sahlan, ‘Shine’, 390:8), but to date, the alleged ‘narration’ has eluded me. 38. ‘Shine’, 390:1–4; cf. ‘Inspiration’, 44:6–10, 17–22. 39. The loss of the corresponding volume of ‘Lightning’ precludes a close comparison with ‘Inspiration’, where according to an ‘it was opined’, ‘the opportunity had been seized’ (‘Inspiration’, 42:17ff.); still, the wording as preserved in ‘Shine’, 390:1, ‘its [that is, Tyre’s] opportunities had lit up’, is in substantial agreement. The passive voice, here and more than once further on, appears to be designed to deflect possible criticism of the perfectly identifiable actors in the drama; see ‘Inspiration’, 44:6–9. 40. See ‘Inspiration’/1888, 37, and ‘Shine’, 387: 13–20. 41. Not, however, by the letter by al-Qadi al-Fadil which Möhring cites as evidence of an all but completed occupation of Tyre in July 1187 (Möhring, 1980, 43f. with n. 33); dating from 593/1197, it was written in reply to a victory message by al-Malik al-‘Adil after taking Jaffa – emphatically put, ­ at a second instance. Further, the writer refers to Saladin as resting in his turba and to comte Henri’s death, which occurred in Acre on 10 September 1197; see Muslim b. Mahmud al-Shayzari, 483f. The writer, al-Fadil, boldly outlines a well-nigh millenarian progress toward Constantinople – and even beyond – for al-‘Adil’s – not Saladin’s! – current campaign. A full scrutiny of Möhring’s reconstruction of the course of events (Möhring, 1980, 36–58) will not be undertaken here; Lilie’s critical observations (Lilie, 1984) have by no means settled all debatable points; cf. Jacoby, 1993; Mayer, 1996, II: 879f. 42. Ibn al-Athir, Perfectissimum, 1851–76, XI: 358:-4–359: ult.; while it cannot

282  |  syri a i n crusad e r time s be ruled out that Ibn al-Athir integrated other authors into his account as well, his narrative substantially reproduces the Katib’s, as a comparison with ‘Shine’ readily confirms. Thus, the following summary from Ibn al-Athir will here serve as supportive evidence of ‘Lightning’: Raymond of Tripoli, for fear of Saladin’s attack on well-nigh defenseless Tyre upon returning from Beirut, flees to Tripoli; when Conrad arrives at Tyre, he finds the town crowded with evacuees from various places taken by Saladin, including Beirut. Here a cautionary note may be sounded regarding the implied date of Conrad’s arrival (see below). Ibn al-Athir next suggests that had Saladin ‘begun with’ Tyre instead of Toron, he would easily have taken it, but he found it daunting because of its impregnable fortifications and as long as the neighboring strongholds had not been conquered. Only then did he think he could comfortably win it (cf. below with n. 45). That was the reason of Tyre’s survival – and here Ibn al-Athir even echoes the Katib’s invocation of divine preordination: God’s command was a peremptory decree (Q 33:38 [al-Ahzab]), in effect largely exonerating Saladin. The question of whether Ibn al-Athir’s translation of the Katib’s narrative into chronological sequence corresponds to the actual course of events will be taken up again below. 43. Although not extant in ‘Shine’, it cannot be ruled out that Abu Shama took it from ‘Lightning’ (‘Two Meadows’, III: 318:-5–321:5, esp. 320:-4f.). 44. Al-‘Adil’s victory message about Jaffa and Mirabel must have arrived just a few days later; see ‘Inspiration’, 32:3–5 and 111:21–112:2, the first passage from the sequential narration and the latter from the comprehensive letter to Sayf al-Islam late in the year, see below and n. 45. 45. ‘Inspiration’, 108:4–109:1. 46. ‘Ernoul’/Morgan, 1982, 57, 59, 60–62/chs 45, 46, 48, 49; Ernoul/de Mas Latrie, 1871, 174, 178f., 182f.; cf., critically, Möhring, 1980, 36–58 47. Pryor, 2010. 48. Admittedly, the Genoese Regni Iherosolymitani Brevis Historia lists Raymond among the lords who welcomed Conrad to Tyre, but for internal chronological confusion and want of external corroboration, its witness will be left out of consideration here (note the seeming synchronicity of Conrad’s landing at Tyre and Raymond’s arrival at Tripoli, with his death telescoped to follow in short order. According to Kamal al-Din b. al-‘Adim, it only occurred during Rajab (6 September to 5 October; Mayer, 1996, I: 205, n. 35; Annali, 145:11–24), whereas Raymond most likely issued charters in Tripoli as early as after the first week of August (Mayer, 1996). 49. Möhring, 1980, esp. 40f.

re a lpoli ti k and vi ci ssi tud e s  | 283 50. Mayer, 2010, II: 859–877, nos 519–522, III: 1339–45f., nos 769–771; cf. Revised Regesta … Database, nos 1231, 1232, 1249–1252. 51. See above, with nos 39, 46. 52. The expressions used are murad Allah (twice) and qada’ Allah; see ‘Shine’, 389:13, 390:8f., 390:4, resp. 53. See ‘Shine’, 380:1–381:2. 54. To the best of my knowledge, ‘Izz al-Din Abu Muhammad – laqab and kunya are owed to other, later transmitters, for example al-Nuwayri – appears twice in the Katib’s works, here and more than a decade earlier, in 571, in al-Kharida; in Damascus, where he lived by that time, he lent the Katib the collected verse – the diwan – of his grandfather, the amir Tamim. The Katib did not then provide any additional information on his interlocutor, only in ‘Lightning’ does he pay his respects to the amir ‘Abd al-‘Aziz’s gracious sociability and his family’s exalted status; see ‘Shine’, 380:2–5, and cf. Talbi, 1969. 55. He still elicits quite diverse reactions and assessments; recent studies on him include: Crawford, 2017; Mallett, 2008; Milwright, 2006; Hillenbrand, 2003; Hamilton, 2000a, esp. 170ff., 179–185 (cf. 132–185 passim), and 286b (index), s.v. Reynald of Châtillon; Leiser, 1977, 87–100 (corresponding quotations from newer editions: ‘Lightning’, V: 69–75; ‘Shine’, 268:13–270:2 [cf. 233:10ff., 234:16–235:2]; ‘Two Meadows’, III: 133:9–141:1 [cf. 82:9–12, 83:1–4]). 56. Obviously the Qadi still considered Raymond, whose dealings with Saladin some observers qualified as nothing short of treason, a formidable foe; see Lewis, 2017, esp. 233–284 and 336a–b (index), s.v. Raymond III; Hamilton, 2000a, esp. 223, 227ff., and 286b (index), s.v. Raymond III; Edbury, 1993. 57. ‘Shine’, 385, with reference to ‘Inspiration’, 31:19–23; cf. Eddé, 2008, 192. 58. Eddé, 2008, 496; Lyons and Jackson, 1982, 293f., 425 (with ref. to MS BrLib, Add. 7465, fols 69r:11–70r:5, of 12 al-Muharram [584]/13 March [1188]). 59. Emphasis is on ‘overall’, precisely because of the nuances in the Katib’s detailed and not automatically glowing accounts – as I hope to show here; cf. Lyons and Jackson, 1982, 371 (‘paints a consistent picture of a hero …’). 60. In 584, he deplores the destruction wrought at Laodicea, apud Abu Shama, ‘Two Meadows’, IV: 22:12–23:15. 61. Lyons and Jackson, 1982, 303 (cf. ‘Inspiration’, 198:11–199:16). 62. Eddé, 2008, 165, 611; Lyons and Jackson, 1982, 132; ‘Shine’, 192f. 63. See Lyons and Jackson, 1982, 333; Eddé, 2008, 250f., 353, 364. Eddé qualifies the reaction of Saladin’s entourage as profound shock – which may be open to debate, Eddé, 2008, 353; cf. Richter-Bernburg, 2014, esp. 49f. For contempo-

284  |  syri a i n crusad e r time s raneous Western assessments see Gillingham, 1989, 181–184 (the discussion is not advanced by Gust, 2012). The only dissenting voice not approving if not outright celebrating Richard’s massacre was Sicard’s of Cremona, who called the action ‘illegal and illicit’ (contra fas et licitum). Instead, he advocated enslaving them and profiting from their labour (apud Salimbene de Adam, Cronica/1966, I: 15:17–20 [=Monumenta Germaniae Historica, SS XXXII: 16:26–29]). 64. Yusuf b. Rafi‘, ‘Rarities’, 174f. (= Richards, 2001, 164f. [improving on alShayyal, Richards also uses MS Berlin 9811(Wetzstein 1893; Ahlwardt, IX: 302a–303a)]). 65. He deploys the massacre in a plea by Saladin for military support, apud Abu Shama, ‘Two Meadows’, IV: 270:8–16. 66. ‘Inspiration’, 26:10–27:6; ‘Shine’, 378:11–15; Richter-Bernburg, 2014, esp. 48. 67. ‘Shine’, 378:11–15; Q 78:40 (al-Naba’  ). 68. Richter-Bernburg, 2014, esp. 47f. 69. ‘Two Meadows’, IV: 105:13–106:5; cf. ibid., 106:6–10 (a Frankish noblewoman with her own retinue of fighting men) and 106:11–107:5 (female knights). 70. Richter-Bernburg, 2014, esp. 46f. 71. Ibid., 49. 72. Ibid., 49. 73. Kitab al-I‘tibar; cf., on the perception and ‘construction’ of ‘the other’ by Usama and his contemporaries, Eddé, 2008, 368–387, 653–657 (notes), and RichterBernburg, 2004, esp. 79, 82, 86. 74. Kitab al-Isharat (= Guide); see Richter-Bernburg, 2009, esp. 118–121. 75. Tadhkira; see Richter-Bernburg, 2009, esp. 118, 121–124. 76. As in n. 64.

16 Assessing the Evidence for a Turning Point in Ayyubid–Frankish Relations in a Letter by al-Qadi al-Fadil1 Bogdan C. Smarandache

Introduction

T

he focus of this chapter is a letter in a manuscript of documents attributed to al-Qadi al-Fadil, the chief secretary of the Ayyubid sultan, Yusuf b. Ayyub ‘Salah al-Din’ (570–89/1174–93), who is best known for having unified Egypt and Syria, for conquering Jerusalem in 583/1187, and for campaigning against King Richard I during the Third Crusade. The thirty-twofolia manuscript containing the letter is housed in the Cambridge University Library under the classmark Qq. 232.2 The letter that is the focus of my study contains no explicit date and the Ottoman-era manuscript remains unedited. In the following, I first discuss the importance of the chancery for the Ayyubid confederacy. I then examine the diplomatic and historical content of the letter, arguing for an earlier date for its composition than hitherto proposed by other historians. If my dating of the letter to the year 575/1179 is correct, then the information which it contains points to a shift in the political situation of the Levant, namely the decline of the Latin Kingdom of Jerusalem, the ascendancy of the Ayyubid Sultanate, and the reconfiguration of regional alliances in that year. In addition, the letter contains some rare details on a partition agreement in effect. The Frankish–Muslim partitions were agreements whereby the revenues of a specified territory or town were shared between the two parties to the agreement. Other aspects of this letter also merit attention but will require separate studies.

286  |  syri a i n crusad e r time s Until new evidence surfaces, I proceed under the assumption that the letter in question indeed came from the pen of al-Qadi al-Fadil. The contents of the letter eliminate the possibility of a non-Ayyubid context. The only known contemporary of al-Qadi al-Fadil capable of such ornate style and aware of the politic scene in such depth was Saladin’s other chief secretary, al-Katib al-Isfahani ‘Imad al-Din (519–97/1125–1201). At present I cannot prove definitively that all the documents in Qq. 232 originate with al-Qadi al-Fadil, and not ‘Imad al-Din, but I have no evidence to refute the claim of the manuscript’s copyist, who labelled this manuscript Insha’at al-Qadi alFadil. Furthermore, several of the stylistic features and imagery that Ibrahim al-Hafsi observed while preparing his thesis on al-Qadi al-Fadil’s letters appear in the letter under examination here. Al-Qadi al-Fadil and the Ayyubid Chancery The chancery came to play a crucial role in the jihad al-Ifranj (the jihad against the Franks) under the Zengid ruler Nur al-Din (541–69/1146–74), who is credited with having carried out the first orchestrated and unified response to the Crusaders and the Frankish kingdoms of the Levant.3 When Saladin succeeded Nur al-Din as sultan of Egypt and Syria, he too advanced his policies of Sunni unification and spread anti-Frankish propaganda. But Saladin had a pressing need for legitimacy during his early years of consolidation since he had created an empire at the expense of local Muslim rulers, including the Zengid princes. His reign thus depended even more heavily on official letters as a medium for propaganda and communication. The Ayyubids and their successors came to rely heavily on professional scribes and secretaries for the diffusion of legitimising propaganda and for efficient communication in a vast and complex political framework.4 As noted above, Saladin employed two official secretaries, and in this chapter I will focus on ‘Abd al-Rahim al-Baysani al-‘Asqalani al-Qadi alFadil (‘the Excellent Judge’) (529–96/1135–1200). ‘Abd al-Rahim had first served the Fatimid Caliphate of Egypt and he had to navigate through precarious circumstances in the decade of instability leading up to the end of the Shi‘a caliphate in 567/1171. Indeed, many Fatimid officials and courtiers suffered ­imprisonment, execution, or exile even before the arrival of Nur al-Din’s expeditionary forces in Egypt. ‘Abd al-Rahim himself had

a t u rning poi nt i n a yyubi d–f ra nk is h r e l a tio ns  | 287 been imprisoned and was released by Nur al-Din’s deputy Asad al-Din Shirkuh (Saladin’s uncle) when he assumed control over Egypt on behalf of the Zengids in 564/1169. Under Shirkuh’s supervision the last Fatimid caliph, al-‘Adid (555–67/ 1160–71), re-appointed the vizier Shawar, who in turn honoured his son’s fellow prisoner, ‘Abd al-Rahim, with the title of al-Qadi al-Fadil. Al-Qadi alFadil continued to serve Shawar and his son. Surprisingly, he survived their demise and continued service under both the Fatimid House and Shirkuh. He drafted Shirkuh’s diploma of investiture on behalf of the Fatimid caliph as well as his letters to Nur al-Din. After the death of Shirkuh, al-Qadi al-Fadil composed an investiture diploma for Saladin. He remained in the young commander’s employ following the abolition of the caliphate he once served and continued to draft letters for him for the remainder of his reign.5 Since Adolph Helbig’s pioneering work, only one dissertation and one monograph on the life of al-Qadi al-Fadil have appeared, both written by Hadia Dajani-Shakeel.6 The first volume of Ibrahim al-Hafsi’s 1,849-page dissertation on al-Qadi al-Fadil’s letters also contains a detailed biography of the secretary and an analysis of his writing style and his social network.7 But the main focus of al-Hafsi’s thesis is al-Qadi al-Fadil’s letters. Al-Hafsi begins by adding eleven additional sources for fragments of al-Qadi al-Fadil’s letters to Helbig’s list of twenty-one manuscripts. Unfortunately, al-Hafsi’s imposing dissertation, containing some 474 documents along with a critical apparatus, remains unpublished.8 In addition, a new dissertation on al-Qadi al-Fadil’s letters is being prepared by Kalthoum Haddad at the École Pratique des Hautes Études.9 In the following section, I examine one of al-Qadi al-Fadil’s letters in detail, exploring first his epistolary style as well as the communicative aspects of the letter. Diplomatic Commentary The copying of Qq. 232 was completed by an unnamed scribe on 3 Jumada al-awwal 1171/13 January 1758. The codex binding measures approximately 12cm × 16cm and the text-block measures 11.5cm × 15.5cm; it is a small manuscript, almost pocket-size. The codex contains only thirty-two folia and most of the folia contain only the introductions or addresses/dedications of letters attributed by the copyist to al-Qadi al-Fadil. The script is a simple

288  |  syri a i n crusad e r time s Ottoman book hand that is easily legible in most cases. The use of occasional ligatures makes some words challenging to read. All of the documents in Qq. 232 are separated by rubrications. For example, the letter under examination here is introduced as Kitab ila’l-Malik al-Muzaffar Taqi al-Din (‘Letter to al-Malik al-Muzaffar Taqi al-Din’) written in red ink. Excerpts from the matn (main text) of letters are introduced by the rubricated word fasl (section) or mufrad (self-standing) and poems are introduced by the word shi‘r (poem). This letter appears to be one of the only intact or nearly intact letters in the manuscript. It extends from line 16 of folium 22v to line 4 of 24r and totals fifty-two lines of text. The word mufrad, also rubricated, then begins a new letter on line 4. The address that follows confirms that it is a separate letter. In this letter al-Qadi al-Fadil begins with a formula of address. He emphasises the role of Taqi al-Din in the jihad al-Ifranj by calling him the Border-Warrior (al-Murabit); the Jihad warrior (al-Mujahid); ‘Victor of the Mujahidin’; and ‘Repressor of the Unbelievers and Associators’ (Qami‘ al-kafara wa’l-mushrikin). This address also reveals al-Qadi al-Fadil’s role of circulating Ayyubid jihad propaganda. Although the letter was composed for an individual, we may be certain that official letters were read out loud and were therefore accessible to the courts and retinues of rulers.10 Ayyubid jihad propaganda thus involved the members of Saladin’s family in an effort to legitimise the pious activities of the clan as a whole. The entire address is composed in rhyming prose (saj‘  ), thereby also exhibiting the skill of its composer. Addresses as well as dedications expounded the addressee’s persona and al-Qadi al-Fadil’s address is replete with invocations pertinent to Taqi al-Din’s rank, role and identity.11 Historical Commentary – Dating the Letter This specific letter appears in M. C. Lyons and D. E. P. Jackson’s monograph Saladin: the Politics of the Holy War and Michael Köhler’s Allianzen und Verträge.12 Lyons and Jackson have taken it as part of the letter that follows it, which deals with Taqi al-Din’s intention to undertake an expedition to the Maghrib. Thus, Lyons and Jackson date the letter to 581/1185, probably based on Ibn al-Athir, who discusses disturbances in the Maghrib in AH 581 involving a mamluk of Taqi al-Din. But the scribe separated the two letters

a t u rning poi nt i n a yyubi d–f ra nk is h r e l a tio ns  | 289 with the word mufrad, marking the beginning of a new letter fragment, as mentioned above. Köhler refers to fol. 23v of the letter for evidence on condominium agreements in the Hawran region south-west of Damascus. He also dates the letter to around 581/1185 based on references to negotiations with Raymond III and diplomatic contact with the Armenians. However, he provides no analysis of its contents. As mentioned previously, the same letter contains no exact date and one must rely on clues from its contents to reconstruct a chronology and determine a context. Let us begin with the individuals whom al-Qadi al-Fadil names in the letter: 1. The addressee, al-Malik al-Muzaffar Taqi al-Din, Salah al-Din’s nephew and ruler of Damascus from c.570–4/1175–9 and Hama from 574–87/1179–9113 2. Ja‘d al‐Sahriyy, a messenger or envoy 3. Al-‘Adi, Salah al-Din’s najib (envoy)14 4. Al-qumis (‘the Count’) = Count Raymond III of Tripoli (c.546–83/1152– 87) and bailli (regent) for Baldwin IV for the period 570–2/1174–6 and Baldwin V (c.579–82/118–36) or his successor Bohemond IV of Antioch (c.567–630/1172–1233)15 5. Ibn Lawun = Roupen III, grandson of Lewon I and ‘Lord of the Mountains’ (c.570–c. 583/1175–87) or Lewon II, ‘Lord of the Mountains’ (c.583–c.615/1187–1219) and king of Armenian Cilicia (c.594–c.615/1198–1219)16 6. Qilij = Qilij Arslan II, Seljuq Sultan of Rum from c.550/1156 until his death in 588/1192 7. Shams al-Din Taj al-Dawla b. Munqidh (523–600/1129–1204), nephew of the courtier and diplomat, Usama b. Munqidh It is clear from the absence of post-mortem invocations (for example, rahamahu Allah) that the Muslims mentioned above were alive at the time of the letter’s composition. Considering the above dates for the persons named in the letter gives us a terminus post quem of Shawwal 570/April 1175, when Saladin became sultan of Egypt and Greater Syria.17 Our terminus ante quem must be 588/1192, the year of Qilij Arslan’s death. Considering that the

290  |  syri a i n crusad e r time s Arabic sources refer consistently to Count Raymond as the qumis it is more likely that the dating of the letter falls within his lifetime, thus narrowing the parameters to 571–83/1175–87.18 The scribe also provides a crucial hint by describing an incidence of inflation that coincided with a drought (lines 23–25): And the prices in the country of the Franks rose up. And whatever the country had already grown withered. And most of the crop around Damascus dried up … And its people abandoned the greater part of it, and they vacated their villages, despairing of it, and abandoned it.19

The sequence of events in this account can be summarised as follows: (1) a drought occurred in the environs of Damascus or perhaps in the whole of Greater Syria; (2) the prices of commodities in the Frankish domains increased; and (3) panic ensued and the villagers and pastoralists left their farms and orchards and headed to Damascus. It is odd that al-Qadi al-Fadil provides no reason for why the prices of goods in the Frankish domains increased. It is possible that agriculture in the neighbouring Kingdom of Jerusalem or County of Tripoli also suffered drought or that the Franks deliberately increased the prices of their goods or restricted their exports. In any case the two simultaneous occurrences were disastrous for the Muslim farmers and caused prices to double. Al-Qadi al-Fadil also recorded changes in the price of wheat, noting that ‘the prices in Damascus then reached a dinar from having been half for a sack of grain’.20 Ibn al-Athir likewise recorded the high price of wheat in the year 574/1178–9 and lived through a drought and famine in the same year. According to his account: This year the rains failed completely in all the lands of Syria, the Jazira, Iraq, Diyar Bakr, Mosul and the Uplands, Khilat and elsewhere. Famine became severe and widespread in all the lands. A ghirara of wheat, which is [fourteen] makkuks in Mosul measures, was sold in Damascus for twenty old Tyrian dinars.21

But the dating of the letter still depends on the dating of the drought. The only date that appears in al-Qadi al-Fadil’s letter is the third of Muharram, and we know from Ibn al-Athir that the situation lasted until Dhu’l-hijja 575/May

a t u rning poi nt i n a yyubi d–f ra nk is h r e l a tio ns  | 291 1180.22 It is important to note that al-Qadi al-Fadil does not explicitly mention the extremely low levels of precipitation in his letter whereas Ibn al-Athir mentions it first, thus placing it as the root cause of the increasing prices and subsequent famine. For the past century, the rainy season in Syria has usually commenced in October and extended into May of the Gregorian calendar.23 Thus for the solar year 1179, the corresponding period of precipitation in the Islamic calendar would be Jumada al-awwal to Dhu’l-hijja. Since al-Qadi al-Fadil mentions his assignment on the third of Muharram, he most likely composed the letter in the year AH 575, otherwise the time lapse between his report and the initial assignment would be too great, or his observation on the rising prices would be anachronistic. Furthermore, al-Qadi al-Fadil actually mentions the long-term effects only of failing precipitation. In other words, it takes a while for harvests to fail and market prices to change following a weather anomaly. That is to say that the weather conditions of AH 574 affected the crop cycle of the following year. This suggests the more precise termini of Muharram 575 – Dhu’l-hijja 575/June 1179 – May 1180 for the letter. Considerable time elapsed after al-Qadi al-Fadil’s composition of the letter during which the prices may have increased to the levels Ibn al-Athir describes. We thus have in the year 575/1179 a very close match for the economic hardship recorded in the letter. The only thing left to determine is where ­al-Qadi al-Fadil was at the time. Unfortunately, this is not a well-documented period in the literature on the Crusades and on al-Qadi al-Fadil.24 He travelled between Syria and Egypt several times in the period 570–84/1174–88 and went on pilgrimage to Mecca in 573/1178 and again in 574/1179.25 The contents of the letter certainly leave out the possibility that he was writing from Egypt since he describes the famine (line 23), rising prices (lines 23, 26), and a parade of Frankish prisoners (lines 17–18) as if he was a witness to these events. It is highly likely that he stayed in Syria for the intervening period between his two pilgrimages. He probably served as the chief scribe of Saladin’s deputy in Damascus, while residing in a military camp nearby.26 We can now begin to analyse the political context for the events alluded to in the letter. Several passages in the letter remain enigmatic and I suspect that these passages will only become clear once the letter is analysed

292  |  syri a i n crusad e r time s alongside other letters of the Qadian diwan. At present I will only focus on two related details in the letter: (1) the broader shift in the power structure and alliance network of Greater Syria and the Sahil, and (2) the possible consequences of the drought of 574–5/1179 on a Frankish–Muslim stalemate, tenuously resting on a partition agreement between Antioch and the Ayyubid confederacy. First of all, the main focus of Saladin’s military activity in Syria after the spring of 574/1179 was investing the new Frankish fortress of Vadum Iacob north of the Sea of Galilee.27 If the letter dates to the year 575/1179, al-Qadi al-Fadil describes the aftermath of the sultan’s successful campaigning in the Hawran and the arrival of Frankish prisoners (lines 17–18, 38–40) prior to Saladin’s siege of Vadum Iacob. The sultan conquered the fortress by Rabi’ al-awwal 575/August 1179 and had it demolished. Second, al-Qadi al-Fadil describes the overtures of the Armenian king Roupen III (lines 20–21), suggesting as well that the Ayyubids were in an advantageous position. When Roupen III came to power in 570–1/1175 he pursued a policy of cooperation with the Franks and campaigned against the Seljuq sultan Qilij Arslan II until they concluded a peace around 576– 7/1181. In 576/1180, however, Saladin finally entered Cilician Armenia to intervene in the Armenian–Seljuq wars and forced Roupen to release the Seljuq prisoners that he had captured while on campaign.28 These prisoners might be the same prisoners referred to in line 21 prior to their liberation. Roupen’s request for a gift (line 21) would thus signal his attempt to profit from his victories over Saladin’s northern ally. It remains unclear why the Armenians would contact the Ayyubids about prisoners as early as 575/1179 and this question requires further research. At present, I can only suggest the possibility of the incipient stages of Armenian–Ayyubid rapprochement. Finally, the letter describes how the drought led to the breach of a Frankish–Muslim truce. Count Raymond III and the Ayyubids had held territory conjointly, probably around Banyas, which meant that both parties were to cease fighting in the aforementioned territory and divide the profits from its harvests. But the arrangement ceased to function due to the evacuation of peasants to Damascus. Naturally, the count complained and al-Qadi al-Fadil was at a loss in his capacity as deputy:

a t u rning poi nt i n a yyubi d–f ra nk is h r e l a tio ns  | 293 And the count (al-qumis) wrote that whoever was settled in the partitions (al-munasafat) must pay the taxes. I answered that this is not normal and that the mamluk [al-Qadi al-Fadil] expects injunctions and instructions …29

It appears that the count tried to collect his share of the taxes from the peasants and pastoralists (lines 44–45). But most of the peasants and pastoralists were fleeing the drought and famine and making their way to Damascus (lines 23–26), which in the end rendered the agreement null. This sequence of events shows that, whereas in normal circumstances partitions could have functioned despite military campaigns taking place elsewhere, the drought and famine of 574–5/1179–80 and the shortage of affordable imports – which might have been partly caused by the ­campaigning of Saladin – meant that the peasants and pastoralists could no longer sustain themselves in the territory held conjointly and therefore fled, which in turn prevented the Franks from obtaining their share of revenues. The breach of the truce, though accidental, allowed the Muslims full leeway to conduct further campaigns of conquest, including the complete annexation of the Hawran. Conclusions This period was marked by successful Ayyubid campaigns on multiple fronts, which resulted in the reconfiguration of the political map. The loss of revenue from the partitioned territory further isolated the Frankish footholds in the Hawran, such as the newly constructed fortress of Vadum Iacob. As Malcolm Barber notes, this period thus witnessed the beginning of the decline of the Latin Kingdom. The success of the Ayyubids elicited diplomatic overtures from the Armenians and desperate complaints from the Franks. It is likely that the sultan’s administration was over-extended and short-staffed in these circumstances and that he sent for al-Qadi al-Fadil, who remained in Syria during this period and only later returned to Egypt. Al-Qadi al-Fadil now arrived to serve Saladin as chancery scribe and intermediary between the sultan and his deputies in Damascus and Hama, and the sultan entrusted him with the delivery of important instructions and information. Al-Qadi al-Fadil composed this letter as a report in early 575/mid-1179 and sent it to Taqi al-Din in Hama.

294  |  syri a i n crusad e r time s At first sight, it appears that Qq. 232 is a typical didactic insha’ collection, containing a sample of exemplary letters by al-Qadi al-Fadil. These letters could serve as models for a variety of situations necessitating tactful communication.30 Indeed the identifiable addressees of the letter include a wide range of persons, namely six Ayyubid magnates, one Zengid prince, the governor of Yemen, the governor of Mecca, a vizier, one Damascene qadi, one Aleppan qadi, the ruler of Mecca, the ‘Abbasid caliph, and ‘Imad al-Din, which surely invited an eclectic collection of epistolary topics. However, the fact that only four letters in the collection appear to be intact, or nearly intact, out of more than forty address fragments and twenty-three ‘sections’ (fusul  ) suggests on the other hand a more random process of selection. A document like the congratulatory letter on the conquest of Jerusalem (fols 12v–16r) exists in many copies and such a letter would have been readily available, whereas many of the other documents in the collection, which are not attested in other manuscripts, might have come from locally available sources, perhaps even a government archive that no longer exists. Why the scribe considered these letters and fragments important enough to locate and copy invites an analysis of social networks and communication in the Levant of the Ottoman period. Although such letters became separated from their original context, they could still inspire and inform modes of communication shared among individuals belonging to, or entering, a social network. Such individuals could use letters, to quote Thomas Bauer, as ‘a communicative strategy […] used by both professional and non‐professional poets and prose writers to establish, strengthen and improve their social position’.31 It is conceivable that in the context of an Ottoman Empire challenged by emerging colonial powers, a local Arabic-speaker sought out whatever traces he/she could find of the works of the famous al-Qadi al-Fadil’s for ideas or inspiration, or as a way to explore, commemorate, and perhaps even revive Ayyubid high literary culture.

 2  3  4  5  6

‫ ظهير‬،‫ والمسلمين‬،‫ عماد االسالم‬،‫ المرتضى تقى الدين‬،‫ الظهير‬،‫المنصور‬

‫ عمدة‬،‫ جمال االمه‬،‫ عضد الدولة جالل الملة‬،‫االمام فخر االنام‬

‫ مصطفى‬،‫ قامع الكفرة والمشركين‬،‫ ناصر المجاهدين‬،‫الملوك والسالطين‬

،‫ وانفذ في االعداء سطوته‬،‫ ومكن صولته‬،‫ واعز دولته‬،‫امير المؤمنين‬

 7  8  9 10 11 12 13

‫ وخلد سلطانه‬،‫ ووفر من التوفيق حُظوته‬،‫وقرن بالتأييد خطوته‬

‫ وجال دهره وزينه‬،‫ واوضح برهانه وبينه‬،‫ و قناه وحصنه‬،‫ومكنه‬

‫ والثغر من رايه السديد‬،‫وال خال االسالم منه حمادا والملك منه سنادا‬

‫ تكون ال منها رماحه او [ذ]اتا‬،‫ وال برحت االرض به مهادا‬،‫سدادا‬

‫ ويعتقدونها‬،‫فال زالت فرا[ﺋ]ﺾ الحمد تقوم بها السنة الخلق له وظا[ﺋ]ﻒ‬

ّ ‫ صدرت خدمة المملوك على يد جعد الصحر‬،‫اورادا‬ ‫ي في ثالث المحرم مقترنة‬

‫ ثم خدمته االن مقرنة‬،‫بالجواب السلطاني عن احدى نوبتين كانتا وصلتا‬

f. 23r

 1

‫كتاب إلى الملك المظفر تقى الدين‬

Fol. 22v

،‫ المرابط المجاهد‬،‫ العادل‬،‫ العالم‬،‫أدام هللا سلطان موالنا الملك المظفر الكبير‬

Appendix: Text and Translation

May He meet his action with approval and make his favour abundant from His bounty. May He preserve his power and strengthen it. May He guard his spear and fortify it.a May He clarify his proof and manifest it. May He elucidate his epoch and adorn it. Islam was not deprived of praise nor did authority become devoid of a supporter nor did the frontier become devoid of His rightly apposite watchfulness [through Taqi al-Din]. The land remains serene through him. And there are no spears in it, or [rather] none other than his own. And there are still obligations of praising him, which the tongues of creation undertake as services. And they consider them to be private prayers.b The service of the mamluk was carried out by the hand of Ja‘d al-Sahriyy on the third of Muharram, together with the sultan’s response to each of the two delegations that arrived.c Then his

Letter to al-Malik al-MuÕaffar Taqi al-Din May God prolong the life of the sultan, our master, al-Malik al-Muzaffar the Great, the Learned, the Just, the Murabit, the Mujāhid, the Victorious, the Conqueror, the Chosen, Taqi al-Din, the Pillar of Islam and the Muslims, the Conquering Imam, the Pride of Humankind, the Support of the Nation, the Honour of the Faith, the Grace of the Umma, the Support of Kings and Sultans, the Victor of the Mujahidin, Repressor of the Unbelievers and Associators, Chosen Commander of the Faithful; may He fortify his nation and strengthen his sortie. May He carry his authority to the enemy.

14 15 16 17 18 19 20 21 22 23 24 25 26 27

‫ وامر بتسيره من المستقر بحماه وتجدد وصول‬،‫بجواب االخري حين ورده‬

‫رسل انطاكية والقصة بطولها تضمنها كتاب عاد لي سيره ووصول‬

‫العدل الي المخيم عا[ﺋ]ﺪا من بالد الفرنج برسالة القومص في السؤال في‬

‫ وكان لقدومه‬،‫ وقد سير ايضا كتابه‬،‫ ومعه اساري‬،‫هدنة انطاكية‬

،‫ محفل معقود‬،‫باالساري الماخودين قطيعة من العدو ومقام مشهور‬

‫ووافق يوم وصولهم يوم وصول رسل المواصلة وقد كانوا ينظنون ان‬

‫ وا الون‬،‫ وان اية تلك العزمة عارضها النسخ‬،‫الشواغل عنهم بالفسخ‬

‫الي االن يراسل سا[ﺋ]ﻼ في الهد]ﻧ[ﺔ وتوقف االمر فيها على اساري قليج الذي‬

‫ تطأ رقبته‬،‫عنده وان لم يجب واال فان العساكر تقطع عقبته‬

‫ وتلف ما كانت البالد قد زرعت‬،‫واالسعار ببالد الفرنج زغت‬

‫والزرع في جانب دمشق اكثره قد عطش فيسره مرجو ان ادركته رحمة‬

‫واكثره قد جال عنه اهله واخلوا قرايا هم ياسا منه وسلوه وكانت‬

‫االسعار بدمشق قد وصلت الي دينار من نصف الغراره ثم تدارك‬

‫جلب الشما[ل] وانحل السعر والباب الناصري لم يتاخر عنه مذ سمعت‬

service was joined with the answer to the second issue when he reached him. And he ordered him to set out from the base at Hama. And the envoys from Antakiya arrived anew. And the account in its entirety was contained in the letter, which returned to me by his journey. And al-‘Adl arrived in the camp, returning from the country of the Franks with an envoy of the count (al-qumis)d about the question of the truce (hudna) of Antakya. And accompanying him were some prisoners. And he also carried a letter. And there were among those who arrived with him a captured party of the enemy, and it was a famous incident and a big cause for celebration. And the day of their arrival coincided with the day of the arrival of the rusul al-muwasila. And they believed that the distractions about them were about the annulment (faskh). And indeed the text opposes any of those decisions. And Ibn Lawun still writes, asking for a gift. And the matter concerning it depends (tawaqqafa) on the matter of the prisoners with Qilij. And if he is not answered, the armies would do nothing but cut off his withdrawal and trample on his neck. And the prices in the country of the Franks rose up. And whatever the country had already grown withered (talafa). And most of the crop around Damascus dried up and relief for this condition was hoped for, if mercy reached it. And its people abandoned the greater part of it, and they vacated their villages, despairing of it, and moved away (sallu). And the prices in Damascus then reached a dinar from [having been] half [that amount] for a sack of grain. Then produce (jalb) reached [us] from the north and prices lowered.e And a messenger did not delay from reaching the court of al-Nasr

28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

‫هذه الحركة رسول من جهة مخالفة في الدين وال موالفة فيه واحتج لهم‬

‫الي التفسيرات والي الضيافات والي عوض هداياهم وهدايا اصحابهم التي‬

‫توخذ منها القليل الذي ال يحتاج اليه ويعوض عنها بالكثير الذي ال‬

‫يستغني عنه والنفقات واسعه والخروج زايد على معتاده ووجوه‬

‫الرجا من تلك الجهات في القرض الح انها اكدت وما اجدت وكل ذلك‬

‫ال يصرف العرم ال يلقنه وال يفرق الراي وال يشتته وامر السلطان‬

‫نصره هللا شمس الدولة بن ﻣﻨ]ﻘ[ﺪ الرسول كان بالموصل ان يكتب الي‬

‫المملوك بصفة ما جري له وقصة طريقة وكتبها في سطور وقد‬

>‫سيرها المملوك لتعرض على موالنا وال يحتاج الي سردها و‫في الثغور المحروسه منتقل تنقل الرحمة المصرفه في البالد و‫ثغورها ويعجل قبل السور سُورها واذا انظرت الي البالد ‫تثري كما تثري الرجال وتعدم وبالد الفرنج عدمت ‫كان من عربانها ونزلوا بالنقرة الدمشقية وارسل القومص ‫عند موالنا فان فيها ذكري للذاكرين ويخاف ان يكون تاخرﺸر