The Production of Meaning in Islamic Architecture and Ornament 9781474482202

Citation preview

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE AND ORNAMENT

Collected Papers in Islamic Art Series Editor: Professor Robert Hillenbrand Series titles include: Studies in Islamic Painting, Epigraphy and Decorative Arts Bernard O’Kane Studies in Persian Architecture Bernard O’Kane Studies in Arab Architecture Bernard O’Kane The Production of Meaning in Islamic Architecture and Ornament Yasser Tabbaa edinburghuniversitypress.com/series/cpia

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE AND ORNAMENT Yasser Tabbaa

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 © Yasser Tabbaa, 2021 Cover image: Yasser Tabbaa Archive, courtesy of Aga Khan Documentation Center, MIT Libraries (AKDC@MIT) Cover design: www.hayesdesign.co.uk Edinburgh University Press Ltd The Tun – Holyrood Road 12 (2f) Jackson’s Entry Edinburgh EH8 8PJ Typeset in Trump Medieval by Servis Filmsetting Ltd, Stockport, Cheshire, and printed and bound in Malta by Melita Press A CIP record for this book is available from the British Library ISBN 978 1 4744 8218 9 (hardback) ISBN 978 1 4744 8220 2 (webready PDF) ISBN 978 1 4744 8221 9 (epub) The right of Yasser Tabbaa to be identified as author 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 Figuresvii Preface xxii PART I  SYRIA CHAPTER 1 Monuments with a Message: Propagation of Jihåd under NËr al-DÈn (1146–1174) 3 CHAPTER 2 Survivals and Archaisms in the Architecture of Northern Syria, c. 1080–1150 26 CHAPTER 3 Circles of Power: Palace, Citadel and City in Ayyubid Aleppo 46 CHAPTER 4 Îayfa KhåtËn, Regent Queen and Architectural Patron75 CHAPTER 5 Defending Ayyubid Aleppo: The Fortifications of Al-Ûåhir GhåzÈ (1186–1216) 94 CHAPTER 6 Originality and Innovation in Syrian Woodwork of the Twelfth and Thirteenth Centuries 105 PART II  IRAQ CHAPTER 7 The Mosque of NËr al-DÈn in Mosul, 1170–1172 129 CHAPTER 8 The Resurgence of the Baghdad Caliphate 158 CHAPTER 9 The Politics of Patronage in Medieval Mosul: NËr al-DÈn, Badr al-DÈn and the Question of the SunnÈ Revival 179 PART III  DOMES CHAPTER 10 The Muqarnas Dome: Its Origin and Meaning CHAPTER 11 The Muqarnas Domes and Vaults of Cairo CHAPTER 12 Andalusian Roots and ʿAbbasid Homage in the Qubbat al-Barudiyyin in Marrakech

201 224 235

vi

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

PART IV  WRITING CHAPTER 13 The Transformation of Arabic Writing: Part 1, Qurʾånic Calligraphy CHAPTER 14 The Transformation of Arabic Writing: Part 2, the Public Text CHAPTER 15 Canonicity and Control: The Sociopolitical Underpinnings of Ibn Muqla’s Reform

259 310 354

PART V  GARDENS CHAPTER 16 The Medieval Islamic Garden: Typology and Hydraulics369 CHAPTER 17 Control and Abandon: Images of Water in Arabic Gardens and Garden Poetry 404 CHAPTER 18 Eternal Hunting Fields: The Frescos at Qusayr ʿAmra as a Pastoralist Interpretation of the Paradise Garden 424 PART VI  SHRINES CHAPTER 19 Invented Pieties: The Rediscovery and Rebuilding of the Shrine of Sayyida Ruqayya in Damascus, 1975–2006 439 CHAPTER 20 Glorifying the Imamate: Architecture and Ritual in the Shiʿi Shrines of Syria 462 CHAPTER 21 The Functional Aspects of Medieval Islamic Hospitals477 Index500

Figures

1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 1.10 1.11 1.12 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10

Aleppo. Qastal al-Shuʾaybiyya, 545/1150, façade 8 Aleppo. Qastal al-Shuʾaybiyya, 545/1150, detail of entablature9 Hama. Mosque of NËr al-DÈn. Façade inscription, 558/1163, detail 11 Hama. Mosque of NËr al-DÈn, Minbar. Inscription on back of chair, 559/1164 14 Mosul. Great Mosque of NËr al-DÈn, 1170–2. Inscriptions on columns 17 Aleppo. Madrasa al-Hallåwiyya. Western apse 18 Aleppo. Qastal al-Shuʾaybiyya. Plan and elevation (Herzfeld)18 Hama. Mosque of NËr al-DÈn. Fragment of Qurʾånic frieze19 Hama. Mosque of NËr al-DÈn. Minbar, 559/1164 20 Jerusalem. Aqsa Mosque. Minbar of NËr al-DÈn, 564/11621 Jerusalem. Aqsa Mosque. Minbar of NËr al-DÈn, detail 21 Damascus. Minaret on Båb Sharqi 22 Split. Palace of Diocletian. Reconstruction of Porta Aurea, c. ad 30028 Cordoba. Great Mosque. Gate on the eastern façade, 98729 Samarra. Jawsaq al-Khaqani. Bab al-ʿAmma, c. 836–7 29 Aleppo. Great Mosque. Domed fountain in courtyard, 96530 Aleppo. Great Mosque. Minaret. 109031 Aleppo. Great Mosque. Minaret. Detail of lower shaft32 Syria. Church at Qalb Lozeh, mid-sixth century. Façade and western flank 33 Cairo. Bab al-Nasr, 1087–8. Detail of tower 33 Aleppo. Minaret of the Great Mosque, 1090, from south34 Cairo. Bab Zuwayla, 1092 35

viii

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

2.11 2.12

Aleppo. Minaret of the Great Mosque, 1090, detail 36 Turkey, Dunaysir (Kiziltepe). Great Mosque, 1204. Detail of arched opening in façade 37 2.13 Turkey, Hisn Keyfa (Hasankeif). Mosque of Süleyman Pasha. Minaret, 1406 37 2.14 Turkey, Mardin. Traditional residence, late nineteenth(?) century. Upper storey windows 38 2.15 Aleppo. Qastal al-Shuʿaybiyya. 1150–1 39 2.16 Aleppo. Qastal al-Shuaʿybiyya. Profile of the entablature40 2.17 Turkey, Harran. Great Mosque. Fragment of entablature, 1174. Urfa Museum 42 3.1 Aleppo, Citadel entrance, from Madrasa al-Sultaniyya 49 3.2 Aleppo, Citadel. Portal of Ayyubid Palace, c. 1200 52 3.3 Aleppo, Citadel, Ayyubid Palace, main courtyard, early thirteenth century 53 3.4 Aleppo, Matbakh al-ʿAjami, iwån, early thirteenth century57 3.5 Aleppo, Citadel. Distant view from south-west 59 3.6 Aleppo, Citadel. Mosque, 1214, from south 60 3.7 Aleppo, Citadel. Maqam Ibrahim, 1168, from north 60 3.8 Aleppo, Madrasa al-Sultaniyya, completed 1221. Plan 61 3.9 Aleppo, Madrasa al-Sultaniyya. Mausoleum of al-Zahir Ghazi with entrance block in background 61 3.10 Aleppo, Citadel and Southern Quarter 62 3.11 Aleppo, Citadel. Opening to tunnel between palace and dår al-ʿadl63 3.12 Aleppo, Bab al-Maqam. Ayyubid with late Mamluk repairs63 3.13 Aleppo, Citadel. Entrance block, thirteenth–fifteenth centuries64 3.14 Aleppo, Citadel. Gate of Serpents, c. 1195 and later 64 3.15 Aleppo, Citadel. The Ayyubid palace, c. 1190–1230. Portal65 3.16 Qalʿat Sahyun. Ayyubid palace, c. 1200–30 65 3.17 Aleppo, Citadel. The Ayyubid palace, 1190–1230. Plan 66 3.18 Hiraqlah (Raqqa), ‘Victory Monument’ of Harun al-Rashid (786–809). Plan 67 3.19 Samarra, Balkuwara Palace, 854–9. Plan of central unit 67 3.20 Raqqa, Qasr al-Banat, c. 1168. Plan 68 3.21 Aleppo, Matbakh al-ʿAjami, first half of thirteenth century. Plan and section, present condition 68 3.22 Aleppo, Citadel. Ayyubid palace, central courtyard 69 4.1 Aleppo, Madrasa al-Firdaws, 1235–6, aerial view from east81 4.2 Aleppo, reconstructed plan of central and southern city82

figures

4.3 4.4

Aleppo, Khånqåh al-Faråfra: foundation inscription 83 Aleppo, Madrasa al-Firdaws: exterior inscription: titles of Îayfa KhåtËn83 4.5 Aleppo, Khånqåh al-Faråfra: plan 84 4.6 Aleppo, Madrasa al-Firdaws, 1235, courtyard, from north 85 4.7 Aleppo, Madrasa al-Firdaws: plan with applied grid 86 4.8 Aleppo, Madrasa al-Firdaws: exterior iwån87 4.9 Aleppo, Madrasa al-Firdaws, 1235, inscription in courtyard87 4.10 Aleppo, Madrasa al-Firdaws, 1235, mi˙råb88 4.11 Aleppo, Madrasa al-Firdaws, courtyard inscription 88 4.12 Aleppo, Khanqah al-Farafra, 1237, courtyard from above  89 5.1 Citadel of Aleppo: ramp and entry block, early thirteenth century 97 5.2 Qalʾat Najm, early thirteenth century, view from the Euphrates99 5.3 Map: al-Zahir GhåzÈ’s Domain 100 5.4 Map: al-Zahir GhåzÈ’s Domain, maximum extent 101 5.5 Qalʾat al-Mu∂Èq (near Afamiya/Apamaea) 101 5.6 Citadel of Aleppo: model 101 5.7 Citadel of Aleppo: entrance block and glacis 101 5.8 Citadel of Aleppo: machicolation 101 5.9 Qalʾat Najm 101 5.10 Citadel of Aleppo: plan of entrance block; Qalʾat Najm: plan of entrance block 102 5.11 Citadel of Aleppo: plan of Ayyubid palace 102 5.12 Qalʾat Najm: plan of Ayyubid palace 102 5.13 Qalʾat Ía˙yËn: portal of Ayyubid palace 102 5.14 Qalʾat Najm: fragment from large Ayyubid inscription 103 5.15 Qalʾat Najm: mosque 103 6.1 Fatimid banister, twelfth century. Damascus Museum 108 6.2 Damascus: cenotaph of Sayyida Få†ima, eleventh century, in floriated KËfic108 6.3 Wooden mi˙råb from shrine of Sayyida NafÈsa, 540/1145109 6.4 Portal of BÈmåristån al-NËrÈ, Damascus, 549/1154 110 6.5 Drawing for a door in Cizre 110 6.6 Aleppo, mi˙råb previously at Maqåm IbråhÈm, 563/1167111 6.7 Analysis of geometric pattern on windows previously at Maqåm IbråhÈm, c. 1200 112 6.8 Minbar at NËrÈ Mosque in Hama, dated 563/1168. Detail and analysis of pattern on the backrest 112 6.9 Minbar of NËr al-DÈn (1168–76). Side panel with signature of one of the artisans 112 6.10 Minbar of NËr al-DÈn. Detail and analysis of pattern on side panel 113

ix

x

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

6.11

Minbar of NËr al-DÈn. Detail and analysis of inside of door113 6.12 Casket of Imåm al-Óusayn, Cairo, 1170–80 114 6.13 Casket of Imåm al-ShåfiʿÈ, Cairo, 574/1178 114 6.14 Minbar at Jåmiʿ al-Óanåbila in Íåli˙iyya, Damascus. Foundation inscription 114 6.15 Minbar at Jåmiʿ al-Óanåbila in Damascus. Side panel and analysis of pattern 115 6.16 Aleppo. Madrasa al-Hallåwiyya, mi˙råb116 6.17 Mi˙råb in the Madrasa al-Óallåwiyya. Detail and analysis of geometric pattern at side panel 117 6.18 Mi˙råb in the Madrasa al-Óallåwiyya. Detail and analysis of geometric pattern in the hood 117 6.19 Detail of mi˙råb in Madrasa al-Hallawiyya, 634/1237  117 6.20 Casket of Fakhr al-KhawåtÈn, wife of Sultan Qillij Arslån IV (r. 1249–67) 118 6.21 Casket for Fakhr al-KhawåtÈn. Analysis of geometric pattern118 6.22 Syrian woodworkers in the twelfth–thirteenth centuries (table) 119 7.1 Mosul, Mosque Al-NËrÈ, minaret 134 7.2 Mosul, Mosque Al-NËrÈ, mi˙råb, dated 543/1148 139 7.3 Mosul, Mosque Al-NËrÈ, 1172, inscribed capital 142 7.4 Mosul, Mosque Al-NËrÈ: dome and portico from courtyard, present condition 144 7.5 Mosul, Mosque Al-NËrÈ: prayer hall interior from north-east, present condition 144 7.6 Mosul, Mosque Al-NËrÈ: exterior from north-west, condition c. 1930 145 7.7 Mosul, Mosque Al-NËrÈ: exterior from south-east, condition c. 1930 145 7.8 Mosul, Mosque Al-NËrÈ: prayer hall interior from east, condition c. 1930 146 7.9 Mosul, Mosque Al-NËrÈ: type 2 column placed against type 1 column, condition c. 1930 146 7.10 Mosul, Mosque Al-NËrÈ: mi˙råb, dated Jumada I, 543/September–October 1148, condition c. 1930 146 7.11 Mosul, Mosque Al-NËrÈ: mi˙råb, after restoration 147 7.12 Mosul: capital from Mår JurjÈs; capital from Mår A˙udemmeh147 7.13 Mosul, Mosque Al-NËrÈ: plan in c. 1915 148 7.14 Mosul, Mosque Al-NËrÈ: reconstruction of plan of first mosque148 7.15 Mosul, Mosque Al-NËrÈ: maqßËra dome from below 149 7.16 Mår Behnåm near Mosul: ribbed vault over the Chapel of the Virgin 149

figures

7.17

Mosul, Mosque Al-NËrÈ: reconstruction of plan of first mosque150 7.18 Mosul, Mosque Al-NËrÈ: a. Capital inscribed with part of Qurʾån 2:255. b. Capital inscribed with part of Qurʾån 9:18 150 7.19 Mosul, Mosque Al-NËrÈ: a. Capital inscribed with part of Qurʾån 24:36. b. Capital inscribed with part of Qurʾån 24:37 151 7.20 Iraqi Museum, Baghdad: two inscription bands with Qurʾån 2: 148–9, previously at the Mosque Al-NËrÈ in Mosul151 7.21 Mosul, Mosque Al-NËrÈ: large stucco panel above the mi˙råb, condition c. 1930, currently at the Iraqi Museum, Baghdad 152 7.22 Mosul, Mosque Al-NËrÈ: minaret, 1170–2 152 7.23 Mosul, Mosque Al-NËrÈ: uppermost portion of minaret152 8.1 Baghdad. Dhafariya (Wastani) Gate, 1221, from north 162 8.2 Baghdad. conical dome of ʿUmar al-Suhrawardi, early thirteenth century 164 8.3 Baghdad. Madrasa al-Mustansiriyya, completed 631/1233, courtyard from east 166 8.4 Baghdad. Madrasa Mustansiriyya, 631/1233, courtyard from east 168 8.5 Baghdad. ʿAbbasid Palace c. 1200, muqarnas vaulted portico 171 9.1 Sanctuary façade, as rebuilt in 1945, at the Mosque al-Nuri. 1172 and later, Mosul, Iraq 182 9.2 Interior, as rebuilt in 1945, in the Mosque al-Nuri. 568/1172 and later, Mosul, Iraq 182 9.3 Exterior from the south-east of the Mosque al-Nuri. 568/1172 and later, Mosul, Iraq; exterior from the north-west of the Mosque al-Nuri. 1172 and later, Mosul, Iraq 183 9.4 Reconstructed plan of the sanctuary of the Mosque al-Nuri. 568/1172 and later, Mosul, Iraq 184 9.5 Interior, c. 1920, of the Mosque al-Nuri. 1172 and later, Mosul, Iraq 185 9.6 Inscribed marble panel from the Mosque al-Nuri. Undated, Mosul, now in the Iraqi Museum 186 9.7 Stucco panel from the Mosque al-Nuri. Mid-thirteenth century, Mosul, now in the Iraqi Museum 186 9.8 Minaret of the Mosque al-Nuri. 568/1172, Mosul, Iraq 187 9.9 View of Mosul from citadel (north), with shrine of Imam Yahya b. al-Qasim to left 187 9.10 Mosul. Shrine of Imam Yahya b. al-Qasim – Mosul, 1239188

xi

xii

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

9.11

Mosul. Shrine of Imam Yahya b. al-Qasim – Mosul, 1239, dome 189 9.12 Mosul. Exterior from the north-west of Mashhad Yahya b. al-Qasim. 637/1239 190 9.13 Mosul. Exterior with inscription frieze of Mashhad Yahya b. al-Qasim. 637/1239 190 9.14 Mosul. Interior, marble ornament and inscriptions of Mashhad Yahya b. al-Qasim. 637/1239 190 9.15 Mosul. Shrine of Imam Yahya b. al-Qasim – Mosul, 1239, interior inscription 191 9.16 Mosul. Interior and corner mi˙råb of the Mashhad Yahya b. al-Qasim. 637/1239 191 9.17 Mosul. Muqarnas vault of Mashhad Yahya b. al-Qasim. 637/1239192 9.18 Mosul. Exterior of Mashhad Imam ʿAwn al-DÈn. 646/1248193 9.19 Mosul. Portal of Mashhad Imam ʿAwn al-DÈn. 646/1248193 9.20 Mosul. Portal detail of Mashhad Imam ʿAwn al-DÈn. 646/1248193 9.21 Mosul. Muqarnas vault of Mashhad Imam ʿAwn al-DÈn. 646/1248 193 10.1 Al-Dawr. Shrine of Imam al-Dawr (1075–90). Exterior203 10.2 Al-Dawr. Shrine of Imam al-Dawr. Detail of squinch 203 10.3 Al-Dawr. Shrine of Imam al-Dawr. Interior of dome 204 10.4 ‘Flood of Baghdad’, from a dispersed Turcoman manuscript (1468) 204 10.5 View of Baghdad. Matrakçi, Beyan-i Menåzil-i Sefer-i ʿIråkeyn (1537)205 10.6 Fez. Mosque of al-Qarawiyyin. Muqarnas vault, 1132–42206 10.7 Damascus. Bimåristån al-Nuri (1154). Dome over the vestibule207 10.8 Damascus. Bimåristån al-Nuri. Vault over niche 207 10.9 Muqarnas domes in Iraq and Syria 208 10.10 Damascus. Madrasa al-Nuriyya al-Kubra (1172). Mausoleum of NËr al-DÈn. Exterior 209 10.11 Damascus. Madrasa al-Nuriyya al-Kubra. Dome over mausoleum. Interior 212 10.12 Damascus. Madrasa al-Nuriyya al-Kubra. Dome over mausoleum. Detail of corner 213 10.13 Baghdad. Shrine of Zumurrud Khatun (1180–1220). Exterior214 10.14 Baghdad. Shrine of Zumurrud Khatun. Springing of dome215 10.15 Baghdad. Shrine of Zumurrud Khatun. Interior 215

figures

10.16 10.17 10.18 10.19 10.20

Shatt al-Nil. Shrine at al-Najmi. Exterior 216 Shatt al-Nil. Shrine at al-Najmi. Detail of muqarnas216 Mosul. Shrine of Imam ʿAwn al-DÈn (1245). Exterior 217 Mosul. Shrine of Imam ʿAwn al-DÈn. Interior of dome 218 Mosul. Shrine of Imam ʿAwn al-DÈn. Upper zone of dome218 11.1 The dome in the shrine of al-Faraj ibn Barquq, Cairo (1405)225 11.2 Isfahan, Great Mosque, North Dome, 1086, transition zone226 11.3 Cairo, Shrine of Sitt Ruqayya, 1135, transition zone 227 11.4 Samarra. Shrine of Imam Dur (1088), interior 228 11.5 Marrakesh. Qubbat al Barudiyyin (1117), dome interior229 11.6 Fez. Great Mosque of al-Qarwiyyin. Almoravid rebuilding of the axial nave (1134–43): muqarnas vault230 11.7 Aleppo. Madrasa al Zahiriyya (c. 1200), portal vault 230 11.8 Damascus, Bimåristån al-Nuri (1154), vault over vestibule231 11.9 Cairo. Shrine of Imam Shafi (1210) 231 11.10 Cairo. Madrasa of Sultan Hassan (1356), portal vault 232 11.11 Cairo. Mosque of al-Muʿayyad (1420) 232 11.12 Madrasa of Sultan Hassan (1356), vault on vestibule 233 12.1 Qubbat al-Barudiyyin in Marrakech, 1117, exterior view (photo Tabbaa) 236 12.2 Map of Marrakech under the Almoravids 238 12.3 Plan of the Qubba, showing location of the cistern and the mosque of ʿAli b. Yusuf 239 12.4 Marrakesh. Qubbat al-Barudiyyin, 1117, exterior view from south 241 12.5 Qubbat al-Barudiyyin, plan 242 12.6 Qubbat al-Barudiyyin, interior view of the dome 242 12.7 Geometric proportions in the plan of Qubbat al-Barudiyyin243 12.8 Qubbat al-Barudiyyin, section 245 12.9 Qubbat al-Barudiyyin, detail of an interior corner 246 12.10 Mausoleum of Imam Dur, near Samarra, c. 1090, interior view 246 12.11 Marrakesh. Qubbat al-Barudiyyin, 1117, detail of transition zone 247 12.12 Qubbat al-Barudiyyin, view of interior ornament 247 12.13 Qubbat al-Barudiyyin, detail of inscription 248 12.14 Qubbat al-Barudiyyin, detail of upper exterior 249 13.1 Papyrus fragment, Egypt, third/ninth century 263 13.2 Qurʾån page in KËfic script, ninth century. Ink on vellum264

xiii

xiv

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

13.3 13.4

Page of Qurʾån, Iraq/Iran, third/ninth century 264 Reconstruction of the method of Ibn Muqla: letters alif, låm, sÈn, dål, ßåd 265 13.5 Tentative reconstruction of Arabic letter forms according to Ibn Muqla 268 13.6 Qurʾån fragment on paper, Iran, twelfth century 268 13.7 Juzʾ of Qurʾån on vellum, Iran, Shaʿbån 292/June 905, signed A˙mad ibn Abiʾl-Qåsim al-KhåyqånÈ269 13.8 Large fragment (first quarter) of Qurʾån, Iran, other parts of same ms. dated 361/972 271 13.9 Verse count. Dublin, The Chester Beatty Library, 1434, ff. 1b and 2a 272 13.10 Semi-KËfic Qurʾån on paper, Iran, 388/998 273 13.11 Colophon, dated 388/998, signed Mu˙ammad ibn ʿAlÈ ibn al-Óusayn al-Íaffår274 13.12 Semi-KËfic Qurʾån fragment on vellum, late tenth century, Iraq or Iran 274 13.13 Verse count. London, British Library, 11,735, f. 4b 275 13.14 Semi-KËfic Qurʾån on vellum, Iraq or Iran, late tenth century275 13.15 Verse count. Istanbul, Topkapı Saray Library, R-38, f. 317b 276 13.16 Semi-KËfic Qurʾån on paper, Iraq/Iran, 394/1004–5 276 13.17 Colophon, dated 394/1004–5, signed Abu Bakr ʿAbd al-­Malik ibn Zarʿah ibn Mu˙ammad al-RËzbåri277 13.18 Fragment from a semi-KËfic Qurʾån on paper, Iran or Afghanistan, mid-eleventh century 278 13.19 Fragment of a semi-KËfic Qurʾån on paper, Iraq or Iran, eleventh century 278 13.20 Qurʾån of Ibn al-Bawwåb, Baghdad, 391/1000–1, signed ʿAlÈ ibn Hilål Ibn al-Bawwåb279 13.21 Qurʾån of Ibn al-Bawwåb280 13.22 Qurʾån of Ibn al-Bawwåb280 13.23 Qurʾån of Ibn al-Bawwåb281 13.24 Small cursive Qurʾån on paper, Baghdad?, dated 402/1011, signed Saʿd ibn Mu˙ammad ibn Saʿd al-KarkhÈ282 13.25 Qurʾån in large gold naskh on paper, possibly Iran; possibly made for a Íulay˙id prince in Yemen, dated 419/1026, signed Al-Óasan ibn ʿAbdallah284 13.26 Heading of sËra 21 (Al-Anbiyåʾ) in thuluth, 419/1026284 13.27 Maghribi KËfic Qurʾån on vellum, North Africa, eleventh century, sËras 99–104 285 13.28 Qurʾån in small naskh script on brownish paper, Iraq or Iran, dated Jumada I 427/March 1036, calligrapher: Abuʾl-Qåsim SaʿÈd . . . ibn TilmÈdh al-JawharÈ, illuminator: Abu ManßËr ibn Nåfiʿ ibn ʿAbdallah286

figures

13.29 Qurʾån in small naskh script on brownish paper, Iraq or Iran, dated 21 Rajab 428/10 May 1037 287 13.30 Qurʾån in small naskh script on brownish paper, Iraq or Iran, 1037 287 13.31 Qurʾån in small naskh script on thick buff paper, Baghdad?, early eleventh century, colophon falsely signed in the name of Ibn al-Bawwåb288 13.32 Qurʾån in small naskh, Baghdad?, early eleventh century289 13.33 Fifth volume of Qurʾån in tawqÈʿ, pen on paper, Bust (Iran), dated 505/1111–12, signed ʿUthmån ibn Mu˙ammad290 13.34 Colophon of Bust Qurʾån, 1111–12 291 13.35 Qurʾån in naskh on paper, Hamadhan (Iran), dated 559/1164, signed Ma˙mËd ibn al-Óusayn al-Katib al-KirmånÈ292 13.36 Qurʾån in naskh on paper, Hamadhan, 1164 293 13.37 Qurʾån in naskh and mu˙aqqaq with thuluth and Eastern KËfic headings, Iraq or Iran, dated 15 Jumada I, 582/3 August 1186, signed ʿAbd al-Ra˙mån . . . al-Kåtib al-MalikÈ ʿZarrin Qalamʾ294 13.38 Verse count. Dublin, The Chester Beatty Library, 1438, ff.2a and 1b 294 13.39 Large section of a Qurʾån in naksh, Iraq or Iran, dated Mu˙arram 592/December 1195, signed AbË NaʿÈm ibn Óamza al-BaihaqÈ295 13.40 Verse count. Dublin, The Chester Beatty Library, 1435, f. 3a 295 13.41 Qurʾån in naskh on paper, Iraq or Iran, datable by waqf to c. 597/1200, signed by calligrapher Mu˙ammad ibn A˙mad al-Jabali and illuminator ʿAbd al-Ra˙mån ibn Mu˙ammad al­-ÍËfÈ296 13.42 Verse count in floriated KËfic contained within circular medallions 296 13.43 Page from the ‘Blue Qurʾån’, gold on blue parchment, North Africa, tenth century 299 13.44 Qurʾån of Ibn al-Bawwab, dated 392/999–1000. Naskh text and thuluth chapter headings 301 14.1 Cairo. Mosque al-Hakim, 990 and later, floriated KËfic inscription on south-eastern tower 314 14.2 Óijåz (Arabia). Gravestone, 250/864. Museum of Islamic Art, Cairo 316 14.3 Cairo. Nilometer: ÊËlËnid inscription (Qurʾån, 2:256), 247/861316 14.4 Cairo. Mosque al-Azhar: alphabet of inscription in the maqsËra, 361/972 318

xv

xvi

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

14.5

Cairo. Mosque al-Óåkim: inscriptions in the maqsËra, late tenth century 319 14.6 Cairo. Mosque al-Óåkim: inscription on casing of north-west minaret, 403/1013 320 14.7 Damascus. Cenotaph of Få†ima: inscription on northern face, 439/1037 320 14.8 Aleppo. Minaret of the Great Mosque: uppermost inscription, 483/1090 320 14.9 Khuråsån (Iran). Left: Samanid dirham minted at Balkh, 292/905. Right: Samanid dinar minted at Nisabur, 340/951 322 14.10 Ghazna (Afghanistan). Cursive inscription on cenotaph of Ma˙mËd ibn SebËktekin, 420/1030322 14.11 Ghazna (Afganistan). Fragment of inscription with name of IbråhÈm, 1059–99; Fragment of inscription with name of YamÈn al-Dawla, late eleventh–early twelfth century 324 14.12 Isfahan. Masjid-i JåmÈ: fragment of inscription on north face of south dome, 478/1086–8 325 14.13 Western Iran. Marble gravestone, 549/1154 326 14.14 Aleppo. Minaret of the Great Mosque: inscription on the third zone, 483/1090 329 14.15 Aleppo. Madrasa al-Óallåwiyya, 543/1149: inscriptions on the portal 330 14.16 Damascus. Bimåristån al-Nuri, 549/1154, foundation inscription331 14.17 Óama (Syria). Mosque of NËr al-DÈn: inscription on back of minbar332 14.18 Aleppo. Maqåm IbråhÈm in the citadel: inscription of NËr al-DÈn, 563/1168 332 14.19 Aleppo. Maqåm IbråhÈm in the citadel: inscription of IsmåʾÈl, 575/1180 332 14.20 Aleppo. Citadel, Lion’s Gate: inscription of al- Ûåhir GhåzÈ, 606/1210 333 14.21 Aleppo. Mosque in the citadel: inscription of al-Ûåhir GhåzÈ, 610/1213 333 14.22 Aleppo. Madrasa al-Firdaws, 633/1235–6: inscription in courtyard334 14.23 Mosul. Mosque al-NËrÈ: mi˙råb, 543/1148. Originally in the now destroyed Umayyad mosque in Mosul335 14.24 Mosul. Mosque al-NËrÈ: inscriptions on capitals, 1170–2336 14.25 Mosul. Mashhad of Imåm ʿAwn al-DÈn, 646/1248: portal to the masjid336

figures

14.26 Mosul. Mashhad of Imåm ʿAwn al DÈn, 646/1248: portal to the masjid, detail of inscription 337 14.27 Mosul. Shrine of Imam ʿAwn al-Din, 1242, detail of thuluth inscription on portal 337 14.28 Tlemcen (Algeria). Great Mosque: inscription on the mi˙råb dome, 530/1136 337 14.29 Fez. Mosque of al-QarawiyyÈn: inscriptions in the cells of the mi˙råb dome, 531/1137 338 14.30 Fez. Mosque of al-QarawiyyÈn: foundation inscriptions above the mi˙råb, 531/1137 339 14.31 Cairo. Citadel: inscription of Íalå˙ al-DÈn on the Mudarraj Gate, 579/1183 341 14.32 Cairo. Fragmentary inscription of Íalå˙ al-DÈn, 583/1187342 14.33 Cairo. Madrasa of al-Íali˙ Najm al-DÈn AyyËb, 641/1243: foundation inscription above the portal 344 15.1 Al-Aßmaʿi, TaʿrÈkh mulËk al-ʿArab, 243/957 354 15.2 New Testament, Timothy 4:1f. Jerusalem 902 358 15.3 Al-Siråfi, Kitåb akhbår al-na˙wiyyÈn al-baßriyyÈn, calligrapher ʿAlÈ b. Shådhån al-RåzÈ, Iraq/Iran, dated 986359 15.4 Qurʾån, Iran (other part of same ms. at the University Library in Istanbul [A6758] is dated 361/972), calligrapher ʿAlÈ b. Shådhin al-RåzÈ361 16.1 Khirbat al-Mafjar, reconstruction drawing of fountain, c. 735 370 16.2 Samarra, Balkuwåra Palace, 849–59, plan 372 16.3 Samarra, stone fountain, originally in the Great Mosque, 849–52, now in Khån Murjån, Baghdad 373 16.4 MadÈnat al-Zahråʾ, 936–81, central palace complex, plan374 16.5 MadÈnat al-Zahråʾ, view of Salon Rico (953–7) from north-west374 16.6 Spain. Madinat al-Zahraʾ, 936 and later, ‘Dar al-Jund’, from south 375 16.7 MadÈnat al-Zahråʾ, Salon Rico, from garden 376 16.8 MadÈnat al-Zahråʾ, pools seen through arches of Salon Rico377 16.9 LashkarÈ Båzår, south palace, eleventh–twelfth century, plan 377 16.10 LashkarÈ Båzår, south palace; I: external iwån, present condition; II: eleventh-century plan of fountain 378 16.11 Qalʾat BanÈ Óammåd, 1015–1152, plan of main palace379 16.12 Qalʾat BanÈ Óammåd, marble slab of a shådirwån, eleventh century 380 16.13 Palermo, La Ziza, 1166–85, aerial view 380

xvii

xviii

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

16.15 Palermo, La Ziza, view of salsabÈl381 16.15 Palermo, Capella Palatina, 1154–66, painting in the ceiling381 16.16 Mardin, Qasr al-Firdaws, 1239–60, plan and section of central iwån382 16.17 Mardin, Qasr al-Firdaws, 1239–60, main iwån from south383 16.18 Mardin, Qasr al-Firdaws, view of cistern 384 16.19 Fus†å† (old Cairo), House VI, eleventh–twelfth century, plan384 16.20 Diyarbakir, palace in the citadel, early thirteenth century, plan 385 16.21 Diyarbakir, palace in the citadel, detail of pool 385 16.22 Qalʾat ÍahyËn, Ayyubid palace, late twelfth century 386 16.23 Qalʾat ÍahyËn, Ayyubid palace, courtyard 386 16.24 Seville, garden in the patio of Qasr al-Mubårak, twelfth century387 16.25 Granada, Alhambra, fourteenth century, exterior towers388 16.26 Story of Bayå∂ and Riyå∂, Bayå∂ lying unconscious by a noria, Spain, thirteenth century, ms. ar. 368, fol. 19 r, Biblioteca Apostolica Vaticana 390 16.27 Drawing of a såqiya in Spain 391 16.28 Water-elevating mechanism, al-JazarÈ, Maʾrifat al-Hiyal, 1354, Suleymaniye Library, Istanbul, 3606 392 16.29 Granada, Alhambra, lion fountain at the pool of the Partal, thirteenth century 393 16.30 Granada, Alhambra, plan of main units 395 16.31 Granada. The Alhambra Palace: Courtyard of the Myrtle, 1333–54, from south 396 16.32 Granada, Alhambra, Court of the Lions, 1354–9 396 16.33 Granada, Alhambra, Fountain of the Lions, eleventh and fourteenth centuries 397 17.1 Damascus. Great Mosque, 705–15, the ‘Barada Panel’ 408 17.2 ‘Barada Panel’ mosaics in the Great Mosque of Damascus, Syria; 705–15 408 17.3 Ground plan of Balkuwara Palace, Samarra, Iraq; c. 850410 17.4 View of Madinat al-Zahra, Spain; 926 412 17.5 A saqiya or water-raising device 414 17.6 (facing page) SalsabÈl fountain in the Zisa Palace, Palermo, Sicily; 1189 415 17.7 Plan and section of the Qasr al-Firdaws east of Mardin, Turkey; 1239–60 416 17.8 View of main iwån in the Qasr al-Firdaws east of Mardin, Turkey; 1239–60 416 17.9 SalsabÈl in the Ayyubid Palace, Aleppo, Syria; c. 1200417

figures

17.10 Fountain in the Court of the Lions, Alhambra Palace, Granada, Spain; fourteenth century; detail of the fountain in the Court of the Lions, Alhambra Palace, Granada, Spain; fourteenth century 420 18.1 Jordan. Qusayr ʿAmra, 743–5, main hall from north-east 425 18.2 Great Mosque of Damascus (705–15) ‘Barada’ mosaic 426 18.3 Jordan. Qusayr ʿAmra, 743–5, the Hunt fresco 428 18.4 Qusayr ʿAmra – iwån with enthroned prince 428 18.5 Qusayr ʿAmra – frescos on west wall 429 18.6 Qusayr ʿAmra – detail of onagers 430 18.7 Qusayr ʿAmra – reconstruction of frescos on west 430 18.8 Taq-e Bostan, fifth century ad431 18.9 Taq-e Bostan Royal Boar Hunt, fifth century ad432 19.1 Aleppo. Mashhad al-Husayn, 1183–96, courtyard from south-west442 19.2 Damascus. Shrine of Sayyida Ruqayya, rebuilt 1988–95, main courtyard 444 19.3 Damascus. The Madrasa al-Muqaddamiyya, courtyard (now destroyed) 445 19.4 Damascus. Shrine of Sayyida Ruqayya and unknown dome, 1978 446 19.5 Coloured cadastral map of region around Sayyida Ruqayya446 19.6 Darayya. Shrine of Sayyida Sukaina bint ʿAlÈ (condition in early 2006) 447 19.7 Damascus. Shrine of Sayyida Zaynab, exterior 447 19.8 Damascus. Bab Saghir cemetery, double shrine of Få†ima and Sukaina bint al-Óusayn448 19.9 Raqqa. Model of the shrine complex of ʿAmmår b. Yåsir, Uways al-QaranÈ and Ubayy b. Kaʿb448 19.10 Aleppo. Mashhad al-Óusayn, c. 1910 449 19.11 Damascus. Map of walled city 449 19.12 Damascus. Shrine of Sayyida Ruqayya, aerial view 450 19.13 Damascus. Shrine of Sayyida Ruqayya, street view 450 19.14 Damascus. Shrine of Sayyida Ruqayya, plan of first expansion451 19.15 Damascus: Shrine of Sayyida Ruqayya, plan of second expansion451 19.16 Damascus. Shrine of Sayyida Ruqayya, courtyard 452 19.17 Damascus. Shrine of Sayyida Ruqayya, central hall 452 19.18 Damascus. Shrine of Sayyida Ruqayya, cenotaph 453 19.19 Damascus. Shrine of Sayyida Ruqayya, dome 453 19.20 Damascus. Shrine of Sayyida Ruqayy, inscriptions on door to cenotaph 454 19.21 Damascus. Shrine of Sayyida Ruqayya, inscription on domeʾs drum 454

xix

xx

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

19.22 Damascus. Shrine of Sayyida Ruqayya, hadith inscriptions455 19.23 Damascus. Shrine of Sayyida Ruqayya, hadith inscriptions455 19.24 Letter signed by President Hafez Assad 456 20.1 Damascus. Shrine of Sayyida Zaynab, rebuilt 1980s, façade from north 463 20.2 Shrine of Sayyida Zaynab, south of Damascus 463 20.3 Women touching the grille of the cenotaph at Sayyida Zaynab’s shrine 464 20.4 Damascus. Shrine of Sayyida Zaynab, worshippers at the cenotaph 464 20.5 Chart of the main tenets of Ithna ʿAshari Shiʿism465 20.6 Ground plan of shrine of Imam ʿAli, Najaf 467 20.7 Men at the cenotaph of the shrine of Sayyida Zaynab 468 20.8 Inscriptions on the shrine of Sayyida Ruqayya, Damascus469 20.9 Qurʾånic verse 28:5 from sËra al-Qaßaß inscribed in a frieze on the shrine of Sayyida Zaynab 470 20.10 Detail of a hadith referring to Imam ʿAli, inscribed on courtyard tiles at the shrine of Sayyida Ruqayya 471 20.11 Detail of a hadith referring to the succession of the Twelve Imams, inscribed on tiles near the cenotaph inside the shrine of Sayyida Ruqayya 471 20.12 Najaf. Shrine of Imam ʿAli, main iwån and dome, from north-east472 20.13 Mirror-glass mosaic work (åʾÈna-kårÈ) in the dome of the shrine of Sayyida Ruqayya 473 20.14 Dome with mirror-glass mosaic at the shrine of Sukaina bint al-Husayn, Darayya, Syria 473 21.1 Damascus. Bimåristån al-Nuri, 1154, façade 483 21.2 Damascus. Bimåristån al-Nuri, 1154, courtyard, from south-east484 21.3 Ground plan of Bimåristån al-Nuri (1154), Damascus 485 21.4 Eastern iwån of Bimåristån al-Nuri, Damascus 486 21.5 Façade of Bimåristån of Arghun al-Kamili (c. 1285), Aleppo486 21.6 First courtyard of Bimåristån of Arghun al-Kamili, Aleppo486 21.7 Exterior view of Darushifaʾ (or çifte Medrese) al-Ghiyathiyyya (1205), Kayseri 487 21.8 Plan of the complex of Qalawun (1285) 489 21.9 View from one of the chambers of the insane at Bimåristån of Arghun al-Kamili (c. 1285), Aleppo 490

Preface

A series of chance encounters with prominent scholars, all dating to 1973–4, prompted my turn to the study of Islamic art and architecture shortly after I had graduated with a BA in Anthropology. The first was reading the pivotal article by Oleg Grabar on the Dome of the Rock, which I came upon in the small library of the Department of Antiquities in Saudi Arabia, where I was working at the time. The second was meeting Michael Meinecke in Riyadh, who subsequently invited me to visit Cairo and stay at the Deutsches Archaologische Institut. My two-week exploration of the highlights of Cairo’s Islamic monuments and my introduction to Meinecke’s research methodology had a lasting impact on me. An extended stay in Baghdad on a UNESCO fellowship introduced me to Iraq’s little-known Islamic monuments and brought me into contact with its renowned and generous scholars, in particular Fuad Safar. Finally, an accidental meeting in Riyadh with R. Bailey Winder, then Dean of Arts and Science at New York University, confirmed my new academic direction and helped me gain admission to the Institute of Fine Arts of NYU. There I studied Islamic art with Richard Ettinghausen, who would die in 1980, two years before I got my PhD. So, quite fortuitously, four notable scholars of Islamic and Middle Eastern art and history would play a decisive role in shaping my future in Islamic art. My entire career is a small gift to their precious memory. I was drawn from the start, even in graduate seminars, to topics that arrived at an interpretation through investigative fieldwork, in continuity with my earlier anthropological training and also a necessary adaptation to the thinly researched conditions of medieval Islamic art at the time. Extensive travel to historical and archeological sites from Iran to Spain, which I undertook regularly from 1973 until fairly recently, grounded my textual knowledge of Islamic architecture and brought me into contact with scholars in these countries. The photographs, slides and negatives that I took; the drawings I made or improved upon; and the field notes I wrote – all these provided me with materials for my dissertation and several other projects.

xxii

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

For better or worse, I was a solitary traveller, a one-person expedition, and that too has coloured my scholarly output. I was never drawn to popular topics and crowded areas of research in which one’s contribution tends to be more critical and theoretical than originally interpretive. Rather than adding to the numerous studies on such popular topics as the Dome of the Rock or the Alhambra, I chose to work on issues and sites that had been relatively untouched since the first decades of the twentieth century. The architectural patronage of Nur al-Din (1146–74) proved ideal for a dissertation topic: it was little studied; it encompassed all Syria plus Mosul; and it proved an excellent vehicle for linking architecture with social and religious history. Although influenced by the social historical approach employed by Grabar in his seminal The Formation of Islamic Art, the dissertation and much of my work did not shy away from engaging religious and sectarian factors into the discussion of architectural patronage. Although it remains unpublished, my dissertation on Nur al-Din opened up several venues for research, including the counter-­ Crusade, transformations in calligraphy and ornament, the Ayyubid architecture of Aleppo, and even the rise of medieval hospitals. My first published article dealt with the resonance of Nur al-Din’s jihåd propaganda in his architecture, mosque furnishings and inscriptions, culminating in the pulpit that he endowed for the Aqsa Mosque in Jerusalem. Regrettably, the sources available to me at the time did not allow for expanding on this theme along the same lines that it had been developed for Crusader studies, although much remains to be done in this area of research. Examining architectural, ornamental and epigraphic changes through the lens of sectarian transformation, the so-called Sunni revival of the eleventh and twelfth centuries, quickly encompassed much of my scholarly output. In these studies I tried to strike a balance between historical and archeological investigation and symbolic interpretation, between studying monuments in their historical moment and also against the slower moving forces of sectarian transformation and technological innovation. Reading broadly in Islamic theology and philosophy and in modern semiotics and deconstruction provided me with interpretive tools that proved effective in understanding the various dimensions of an especially contentious period of Islamic history. Rejecting the ahistorical essentialism of perennialist thought, I sought instead to investigate the production of meaning and symbolic charge in architectural, calligraphic and ornamental forms within the context of parallel dislocations in the political and sectarian realms. Central to this process was reasserting the role of the Baghdad caliphate as the source of legitimation for the many arriviste dynasties ruling the Islamic world in this period and, perhaps more speculatively, as the innovator of ornamental forms, including the

preface

muqarnas; calligraphic styles, such as the proportioned script (alkhatt al-mansub) of Ibn Muqla and Ibn al-Bawwab; and institutions, including the madrasa and the bimaristan. None of these architectural and ornamental forms represented immanent manifestation of a timeless Islamic spirit, as proponents of perennialism had proposed. They were rather created or monumentalised within a symbiotic relationship of a centre, the ʿAbbasid caliphate, possessing legitimacy but lacking power and a periphery of new dynasties possessing power but needing ʿAbbasid legitimation. These incipient symbolic forms, I argued, quickly spread throughout much of the Islamic world as signs of symbolic allegiance to the safeguard of the Sunni Islamic world. My work on the muqarnas dome began in a formalist and typological study of about a dozen such domes in Iraq and Damascus and progressed to other similar domes in North Africa and Spain. This led me to question the significance of the most salient features of these domes – geometric division and unsupported projection – features that distinguish them from earlier hemispherical domes and that, I argued, reflected an entirely different conception of the dome and its referent, heaven. Reading these formal and expressive features in conjunction with the contemporary theosophy of occasionalism, I argued that the muqarnas dome was developed as an architectural manifestation of this pervasive theosophy, whose main proponent was none other than the Baghdad caliphate. These findings would later contribute to a better understanding of the unique early twelfth-century Qubbat al-Barudiyyin in Marrakech, whose blended forms and unusual inscriptions reflected its rootedness in Cordoba and its assimilation of ʿAbbasid symbolic forms. My work on the transformation of Arabic writing, in Qurʾåns and in public inscriptions, was guided by a similar progression from an analytical study of numerous examples, to abstracting their innovative features in relation to what had preceded them, to an interpretation along contemporaneous sectarian and political lines. The conversion of Qurʾånic writing from ambiguously majestic Kufic scripts to the elegantly legible scripts developed by Ibn Muqla (d. 940) and Ibn al-Bawwab (d. 1022) was one of the most pivotal and least studied phenomena in Islamic art. This two-phase adoption of proportioned cursive scripts in place of the earlier Kufic scripts produced an entirely new Qurʾånic codex whose uncompromising clarity and legibility, I argued, reflected a specifically exoteric/zahir view of the scriptures, in conformity with contemporary ʿAbbasid theology and as a challenge to its divergent Fatimid counterpart. The architectural implications of these calligraphic reforms resonated widely in epigraphy, particularly public inscriptions, which, starting in the late eleventh century underwent a fitful and then quite decisive change from Kufic to cursive. Thus, in Syria Nur al-Din mandated this conversion in the middle of the twelfth century, most

xxiii

xxiv

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

likely in continuity with Qurʾånic practice and in opposition to the Fatimids, who had championed the floriated Kufic script, one of the most ambiguous calligraphic scripts. Subsuming the esoteric/batin within the exoteric/zahir, the new public inscriptions embodied and projected the exoteric and inclusive doctrines of the Sunni revival. My turn to the Ayyubid architecture of Aleppo in the 1990s was prompted by its natural continuity with the earlier Zangid period, its excellent state of preservation and also by my deep appreciation of its subtlety and sobriety. The military and palatial architecture of Aleppo, best exemplified by its mighty citadel, demonstrates filiations with Baghdad and Samarra but presents solutions unique to medieval Aleppo. Thus, the foundation of a tribunal (dar al-‘adl) at the foot of the citadel continues earlier ʿAbbasid practice, but the Ayyubid institution, though still an appendage of royal power, is separate from the palace and better integrated with the city and its population. Likewise, the Ayyubid citadel palace, with its cruciform plan and playful water fountain, recalls the ninth-century palaces of Samarra but at a greatly reduced scale, necessitated by its restricted location and the diminished resources of its founders. I have dealt more thoroughly in my book, Constructions of Power and Piety in Medieval Aleppo (1997), with the religious architecture of Aleppo, highlighting in particular issues of patronage and the sectarian dialogue between its Sunni madrasas and Shiʿi shrines. Regarding patronage, the Ayyubid period is especially notable for the patronage of its women of the court, in particular that of the notable regent queen Dayfa Khatun (r. 1237–44). Her two most important buildings, Khanqah (Sufi convent) al-Farafra and Madrasa al-Firdaws, signal her support of Sufis, in particular women Sufis, and her deployment of Sufi transcendental piety as an alternative to Shiʿi rituals, which were quite prevalent in Aleppo at the time. My main interest in Mosul was at first restricted to the Mosque al-Nuri (1170–2), which I had examined in my dissertation as the only monument founded by Nur al-Din outside his sovereign domain. Utilising contemporary sources and hitherto unpublished photographs from the early twentieth century, this article attempted a reconstruction of the original mosque and a new chronology for its main phases. It remains the only scholarly study in English of this mosque, which was completely destroyed in recent years. The symposium The Art of the Seljuqs, held in 2016 at the Metropolitan Museum of Art, presented me with an opportunity to return to the study of medieval Mosul, which, sadly, no longer exists in any recognisable form. Comparing the architectural patronage of of Nur al-Din and Badr al-Din Luʾluʾ (1234–59) highlights their contrasting religious and sectarian policies, with Nur al-Din strictly adhering to Sunnism, while undermining Shiʿism and Christianity, and Badr al-Din promoting Christianity and embracing Shiʿism as an instrument against a particularly militant Sufi movement that had

preface

threatened his sovereignty. This comparison expands the parameters of the Sunni Revival and highlights the mediation of regional political forces in sectarian issues. In some respects, even my work on Islamic gardens was informed by the relationship between the Baghdad caliphate and the outlying dynasties, although much mediated by poetry in this case. Garden forms, hydraulic machines and decorative fountains, I argued, found their way from Samarra and Baghdad to North Africa and the Iberian Peninsula, a translation that may have been facilitated and given meaning through contemporaneous poetic descriptions. The diminution of gardens and the increasing elaboration of pools and fountains, seen in both the eastern and western Islamic world, resonated through both architectural forms and poetic metaphors. A critical revision of the essentialism and ahistoricism of the Islamic paradise garden paradigm led me recently to emphasise the hunt as the central attribute of what might be called the pastoralist garden. Re-examining Umayyad hunt imagery, particularly the extensive hunt cycles at Qusayr ʿAmra (c. 740), directed me to conclude that these hunt frescoes embodied an entirely variant vision of paradise, one that rejected the tedium of the enclosed garden and embraced a more active paradise, with hunt at its centre. This is also an area worthy of further investigation. Finally, my recent and still unfinished work on Shiʿi shrines began with a single monument in Damascus, Sayyida Ruqayya, whose encroachment upon the site of medieval monuments I had observed over several years. The massive expansion of this shrine in recent years and its enormous popularity among Shiʿi pilgrims, a process repeated in many other Shiʿi shrines in Syria, led me to study it both historically and as a contemporary monument made alive through ritual. This in turn led to me to study Shiʿi shrines in general from a phenomenological perspective, linking Shiʿi core beliefs and ritual practices with the main architectural and design features of these shrines. My dear colleague Robert Hillenbrand, a keen observer and promoter of the historiography of Islamic art, has kindly and persistently urged me to compile this selection of articles in the hope that this book would make them more accessible to a new readership. I would therefore wish that these new readers would engage with my often quite controversial articles, critique them, and perhaps better link them with contemporary scholarship. I would also hope that the republication of these articles would propel me to expand on my views on the role of difference, disjunction and controversy, both sectarian and political, in the production of meaning of Islamic art. I believed then, as I do now, that these dynamic interactions and dissonant voices, far more than an essentialist adherence to tawhid or a postmodernist appeal to the commonalities of piety and sacred space, contributed to the unending transformations in Islamic art.

xxv

PART I  SYRIA

CHAPTER ONE

Monuments with a Message: Propagation of Jiha¯d under Nu¯r al-Dı¯n (1146–1174) The jiha¯d (holy war) of NËr al-DÈn against the Crusaders was conducted militarily on the battlefield and ideologically on the home front. Militarily, his period marked the actual beginning of a systematic Muslim counter-crusade which began with his takeover of Edessa and culminated in the conquest of Jerusalem under Saladin.1 The ideological battle, on the other hand, manifested itself most explicitly in contemporary literature, primarily poetry, but also in official letters and juridical treatises on jihåd. This literary evidence has been thoroughly examined by Emmanuel Sivan in a book which discusses ideology and propaganda in the Muslim reaction to the Crusades.2 An equally important, if perhaps less explicit, aspect of the propagation of jihåd is also evident in the inscriptions and some monuments of NËr al-DÈn and later heroes of the Muslim countercrusade. A study by Nikita Elisséeff on the titulature of NËr al-DÈn from his inscriptions dealt, in part, with the jihåd titles of these inscriptions and the monuments which bore them.3 Here I am concerned with the inscriptions and certain monuments of NËr al-DÈn whose function, it will be shown, was to spread the spirit of jihåd against the Crusaders and declare the triumph of Islam in Syria and JazÈra. When surveying the corpus of the inscriptions of NËr al-DÈn, I was struck by the fact that those inscriptions of the greatest significance for jihåd generally occurred on monuments which themselves convey a similar message, although in a more subtle manner. In other words, the various jihåd titles and epithets in these inscriptions served to reinforce the jihåd message of the monument on which they were inscribed. For this reason, it was thought preferable to study these inscriptions not separately, as has generally been done, but in conjunction with the monuments on which they occur. Furthermore, in the hope of connecting these monuments with the actual jihåd of NËr al-DÈn – while obviating Yasser Tabbaa (1986), ‘Monuments with a Message: Propagation of Jihad under Nur al-Din’, in V. Goss and C. Vézar-Bornstein (eds), The Meeting of Two Worlds: Cultural Exchange Between East and West during the Period of the Crusades, Kalamazoo, MI: Medieval Institute Publications, 223–41.

4

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

the need to refer to historical sources – it was thought advisable to begin the discussion of each monument with a brief summary of the historical circumstances which occurred immediately before it was built. The special importance of the inscriptions of NËr al-DÈn can best be understood when contrasted with those of his father Zangi. The contrast between Zangi’s rather indifferent attitude to jihåd against the Franks and NËr al-DÈn’s total dedication to it is clearly evidenced by the great difference in their respective protocols. Zangi possessed a set of titles which were among the longest in medieval protocols. They included four Persian titles – isfahsilår (commander of the army), shahriyår (protector of a province), pahlavån jåhån (guardian of the world) and khusro, Iran; two Persian-derived titles – al-sayyid al-kabÈr (the great lord) and shams al-ma’åli (the sun of heaven); and five Turkish titles – alp ghåzi (invader hero), agh arslån (white lion), inanj qutlugh (loyal fortunate), tughrultekÈn (falcon prince) and atåbek (tutor of the Saljuq crown-prince).4 Zangi’s protocol also laid special emphasis on attributes of legitimacy and sovereignty as bestowed by the Caliphate and the Sultanate. Out of the thirtyone epithets and attributes in his most complete inscription in the mashhad (shrine) Imam Muhassin in Aleppo 1143, eighteen concern these two matters. The ubiquity of foreign titles and the emphasis on legitimacy and sovereignty are clear indications of Zangi’s TurkishPersian orientation and his rank as atåbek of the Saljuq sultan. Finally, one notices the absence of the title al-mujåhid (fighter in the Holy War) from all Zangi’s inscriptions. It seems that even after his victory at Edessa in 1144, Zangi did not merit the most important title of the Holy War.5 Only the very earliest inscription of NËr al-DÈn, in the mashhad al-Muhassin (dated 1146, a few months after his accession to rule) may have contained any Turkish titles.6 It is fragmentary and largely derived from Zangi’s inscription which stands a few feet away from it on the same wall. The little that is preserved from it follows Zangi’s titulature very closely, and it is quite possible that it once contained the title atåbek as well as other Turkish titles. Its epigraphic style (dry, somewhat crude KËfic) is also derived from Zangi’s inscription. Executed only a few months after NËr al-DÈn’s accession, this inscription (in style and content) reveals NËr al-DÈn’s dependence on the dominating figure of Zangi. Beginning with his second inscription at the portal of the madrasa (theological college) al-Hallåwiyya, dated February/March 1149, NËr al-DÈn’s protocol begins a process of total change. Turkish titles disappear first, to be followed by all Persian and Persian-derived titles as well as most titles of sovereignty and legitimacy. These are replaced by purely Arabic titles which emphasise NËr al-DÈn’s dedication to jihåd as well as to the affirmation of SunnÈ orthodoxy and the establishment of justice.

monuments with a message

This remarkable change in the titulature was preceded by a number of decisive battles which NËr al-DÈn fought in the first few years of his reign (1146–50) against the Frankish colonies of north-western Syria: Edessa, Antioch and Tripoli. In 1146, he reconquered Edessa (which had rebelled after Zangi’s death) and massacred much of its Christian population. He followed that, in 1147 and 1148, with the capture of most of Antioch’s possessions east of the Orontes river. Finally, in 1148, he succeeded, with the help of his brother Sayf al-DÈn Ghåzi of Mosul, in repulsing the remnants of the Second Crusade which had besieged Damascus.7 According to Marshal Baldwin, ‘the one thing that the Muslims feared the most, a powerful expedition from Europe, had arrived and been repulsed’.8 Ibn al-ʾAdÈm, the thirteenth-century historian of Aleppo, said: ‘From this point on NËr al-DÈn dedicated his will to jihåd.’9 The madrasa al-Hallåwiyya with its innovative inscription was, therefore, built at the first peak of the jihåd of NËr al-DÈn. This madrasa (originally the Byzantine cathedral of St Helena) was one of four churches which had been converted to Muslim usage by the judge Ibn al-Khashshåb in 1124, in retaliation for a particularly pernicious Frankish siege of Aleppo.10 It does not seem, however, that Ibn-Khashshåb’s conversion entailed any structural modification or addition to these churches: there is no mention in the sources of such activity, and the two remaining converted churches (the madrasas of al-Hallåwiyya and al-Muqaddamiyya) bear no inscriptions which pre-date the period of NËr al-DÈn. It would seem, then, that the conversion entailed little more than the destruction or removal of obvious elements or insignia of the Christian faith (altars and crosses, for example) and the addition of a mi˙råb. Of the four converted churches, the cathedral of St Helena was, by far, the largest and most important. At the time of the Muslim conquest of Aleppo, it was the main centre of worship for the city, and its property extended to the present courtyard of the Great Mosque. Following its confiscation by Ibn al-Khashshåb, it became known as masjid al-Sarråjin (mosque of the saddle makers), possibly after the branch of the bazaar in which it was located.11 NËr al-DÈn completed Ibn al-Khashshåb’s act by giving the converted church a new physical appearance as an important madrasa of the city and the first to be built in almost forty years. The building of a madrasa at such a close proximity to the Great Mosque was designed, in part, to undermine the prominence of the Shiʿi majority of the City which had previously resisted the building of the first madrasa of the city, al-Zajjåjiya.12 An equally important purpose of the al-Hallåwiyya, however, was to show the subjugation of Christianity and the triumph of Islam in the city which had had to bear the brunt of Frankish attacks for half a century. This militant intent of the building will be made clearer when we examine its interior and the inscription of NËr al-DÈn at its entrance.

5

6

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

The cathedral, as correctly reconstructed by M. Ecochard, was of the centrally planned tetraconche type common in north and south Syria in the early Christian period (Figure 1.6).13 It is not certain how much of the Byzantine structure existed at the time of NËr al-DÈn, but judging from the good state of preservation of the western apse, one may postulate that the main apse and two side apses were at least partly preserved. These parts may have been demolished either by the mob which was led by Ibn al-Khashshåb or later by NËr al-DÈn. However, the fact that the colonnade and entablature of the western apse were left intact suggests that the destruction of the other parts, particularly the eastern apse, was not the act of a mob but a deliberate act which was most likely carried out under NËr al-DÈn. While the total destruction of a particular church in this period could reasonably be interpreted as a misguided act of vengeance, the selective destruction of the cathedral of St Helena and the incorporation of important parts of it in an otherwise new Islamic building demands a different interpretation. Under other historical circumstances, one might postulate that the purpose behind this unusual endeavour was NËr al-DÈn’s desire to connect himself with the ancient heritage of Aleppo, but this is highly unlikely in the wake of his major victories against the Franks. I would suggest, then, that the primary intent of preserving these Christian remains was to symbolically assert NËr al-DÈn’s subjugation of and victory over the Christians. Not surprisingly, the inscription of NËr al-DÈn at the portal of the madrasa al-Hallåwiyya displays important developments in the titulature of the sovereign, especially in the realm of jihåd.14 All Turkish titles, including atåbek, are dropped.15 Atåbek was Zangi’s title par excellence and one which was retained by his descendants in Mosul to the time of Badr al-Din Luʾluʾ.16 Its deletion by NËr al-DÈn at this early stage is a clear indication that he no longer considered himself an atåbek of the Saljuq sultan but an independent sovereign of an Arab state. By far the most important development exhibited in this inscription, however, is the use of the title al-mujåhid, for the first time in the protocol of NËr al-DÈn. The title may have been bestowed on him by the ʿAbbasid caliph in appreciation of his valiant effort in saving Damascus from the Second Crusade. This epithet appears in sixteen out of thirty-eight preserved inscriptions of NËr al-DÈn and may, therefore, be considered his most important title after al-’ådil (the just). Emboldened by his success against the Second Crusade, NËr al-DÈn continued his aggressive raids in middle and northern Syria. On 29 June 1149, his troops – aided by Anar of Damascus – fought against Raymond of Antioch in Innab, just east of the lower valley of the Orontes river. The Antiochene forces suffered a disastrous defeat, and Raymond himself perished in the battle. NËr al-DÈn then

monuments with a message

took Apamea and, according to Baldwin, ‘advanced toward Antioch ravaging the country-side as far as the coast where he exultantly bathed in the Mediterranean’.17 The moral implications of this battle were even greater. Gibb said: This, the most spectacular of NËr al-DÈn’s victories over the Franks, and coming at this early stage in his career, seems to have been the turning point in his own conception of his own mission and in the history of Muslim Syria. In the eyes of all Islam, he had become the champion of the faith and he now consciously set himself to fulfill the duties of this role.18 In 1150, NËr al-DÈn built the qastal and madrasa al-Shuʾaybiyya in Aleppo, a small monument of great significance for the propagation of jihåd. This building, of which only the qastal (fountain) remains (now attached to a mosque of later construction), is located inside the Antioch gate, at the bifurcation of the main east–west branch of the bazaar. This is the spot where once stood the mosque of ʿUmar, the first mosque built by the Muslims upon taking Aleppo in the year 637. The tradition, on which all sources agree, states that the Muslims, led by ʿUmar, entered Aleppo from the Antioch gate. The spot where they put down their arms became the location of the first mosque of the city.19 The little mosque served the congregational needs of the early Muslim community until the building of the Great Mosque by al-WalÈd (c. 715), when it must have become a simple masjid with a great tradition. Around the beginning of the tenth century, the masjid was renovated by a certain Aleppan Shiʿi named Abu-l-Hasan al-Ghadåiri. It was venerated throughout the intervening period and became a favourite spot for meditation by various ascetics.20 Under NËr al-DÈn, a totally new building was erected which exists today in an advanced state of ruin. Only the façade is original; the interior is modern and without interest. Herzfeld’s reconstructed drawing shows the original elevation without all the later additions (Figure 1.7). It consists of a projecting portal, open on three sides through pointed arches and with a fountain to its right. A heavy entablature of classical appearance crowns the façade. It consists, in the antique tradition, of architrave, frieze, and a tripartite cornice with dentil, fascia on brackets and sima. Instead of Roman ornament, a profusion of KËfic inscriptions and arabesque scrolls cover all parts of the entablature. A fragmentary historical inscription, partly concealed by a modern roof, mentions the name of Caliph ʿUmar ibn al-Kha††åb and gives the date of the building. NËr al-DÈn’s name and titles (which would be of special interest here) are, unfortunately, concealed by the new additions. Opinions have been divided on whether the entablature was indeed antique (re-used and re-decorated in a contemporary fashion)

7

8

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 1.1  Aleppo. Qastal al-Shuʾaybiyya, 545/1150, façade (photo: Yasser Tabbaa)

or whether it was totally made in the middle of the twelfth century. Sauvaget considered the entablature antique and even proposed that it was part of a Roman monumental arch which existed on the same spot as the later mosque: ‘Cette mosquée n’était rien que l’arch monumental erigée en tête de l’avenue à colonnades que les Arabes,

monuments with a message

Figure 1.2  Aleppo. Qastal al-Shuʾaybiyya, 545/1150, detail of entablature (photo: Yasser Tabbaa)

se l’étant approprié, transformèrent en lieu de culte en murant ses baies.’21 He presented three pieces of evidence for this assertion: the existence of military emblems – ‘emblemes guerriers’ – on the frieze, a certain text by Ibn Shaddåd about ‘un arc portant une inscription grècque’, and the classical appearance of the entablature.22 The socalled military emblems simply do not exist, and Sauvaget may have mistaken the large foliated KËfic script for them. As for the statement of Ibn Shaddåd, it refers not to the present edifice but simply to a Greek inscription which was discovered on the Antioch gate itself.23 The classical appearance of the entablature presents a slightly more difficult problem. When examined closely, however, one notices that although the entablature has the same components as a classical entablature, it displays many discrepancies: the architrave projects at an angle; the frieze is a cavetto instead of a flat surface; and the corners are left undecorated, possibly because the medieval architect was not certain as to how to continue the ornament around the corner. Furthermore, the blocks which make up the entablature are in perfect accord with one another and with the arches below, whose pointed profile can only be medieval.24 Herzfeld, who made a detailed study of this entablature, also concluded that it was medieval. He added, however, that it was not the result of spontaneous imitation of an antique architecture but, rather, the product of ‘an uninterrupted antique tradition’.25 To prove his point he compared the entablature of al-Shuʾaybiyya with those of the gates of Cairo (built in 1087) and the minaret of the Great Mosque of Aleppo (built in 1089).26 There are, indeed,

9

10

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

general ­similarities among these monuments, as in the use of KËfic inscriptions on friezes and cornices supported by brackets. But that is where the similarity stops, for, whereas the gates of Cairo and the minaret of the Great Mosque of Aleppo are vaguely related to antique architecture, the qastal al Shuʾaybiyya is a close and deliberate imitation of a classical entablature. In this respect, it is not part of ‘an uninterrupted antique tradition’ but a rare example of purposeful imitation.27 The use of a projecting portal with a classical-looking entablature may have been intended to express military power in much the same way as a triumphal arch. Projecting portals, with or without classical entablatures, have been used in several Islamic monuments, ranging in date from the Umayyad to the Fatimid period and even later. The military intent of such portals is unmistakable in the Umayyad desert palaces, in the Ribåt (garrison) of Susa, and even in the mosque of al-Håkim in Cairo.28 In terms of intent, then, the projecting portal of the qastal al-Shuʾaybiyya is related to such Islamic examples. In terms of its form, however, this monument harks back to Roman architecture. It is possible that NËr al-DÈn saw and admired the magnificent Roman ruins in Apamea, which he had only very recently conquered, and wanted to express his military victories against the Crusaders with the same vocabulary. The mention of the Caliph ʿUmar ibn al-Kha††åb in the inscription can be interpreted as an allusion either to his founding of the first mosque in Aleppo or to his capture of the city.29 The first interpretation would emphasise the pietistic intent of the monument, to show NËr al-DÈn’s care and reverence of the mosque of ʿUmar. This was, indeed, one of the intents of the building, but not the main one. An allusion to ʿUmar’s capture of Aleppo would, on the other hand, be more consistent with historical circumstances (NËr al-DÈn’s major victory) and the expressive intent of the building from (military triumph). I would suggest, then, that the qastal al-Shuʾaybiyya was primarily intended as a victory monument whose message was to connect the jihåd of Nur al-DÈn against the Crusaders with the jihåd of ʿUmar against the Byzantines. Its location at a heavily travelled spot and the inclusion of a fountain in its façade ensured that many people would see the monument and understand its message. It may be added that this interpretation is in complete harmony with the character of NËr al-DÈn, for he is said to have taken great interest in reading about the Prophet and the Companion Caliphs and in emulating their deeds. In fact, he is often compared to the two ʿUmars, ʿUmar ibn al-Kha††åb and ʿUmar bin ʿAbd al-ʾAzÈz.30 The ten-year period which followed NËr al-DÈn’s capture of Damascus in 1154 was not eventful in terms of jihåd. He had succeeded in fulfilling his primary ambitions and now dedicated his energy to the stabilisation of the internal situation. Towards the end of this period, in 1163, he made a last try at opening a sea outlet for

monuments with a message

Figure 1.3  Hama. Mosque of NËr al-DÈn. Façade inscription, 558/1163, detail (photo: Yasser Tabbaa)

his kingdom, but suffered a disastrous defeat in al-Buqayʾ, near Krak des Chevaliers. This defeat, we are told, greatly disturbed him and led him to abandon all luxuries of life for the sake of worship and fighting for the faith.31 Between 1163 and 1164, NËr al-DÈn built a Friday mosque in Hama, the second such mosque in the city. The historical and Qurʾanic inscriptions of this building and the minbar (pulpit) inside it can be connected to NËr al-DÈn’s new asceticism and dedication to the Holy War. The dedicatory inscription is over seven metres long and executed in large and clear thuluth script.32 It begins with the normal Basmala (pronouncement of the name of God) but proceeds into the Shahåda (declaration of faith), which is rare in historical inscriptions and found in only two other inscriptions of NËr al-DÈn, both in Hama.33 The title al-mujåhid is used, as well as a number of composite epithets of jihåd: keeper of the outposts, saviour of the public, vanquisher of the rebels, killer of the infidels and polytheists. The length and size of the inscription, the Shahåda and the emphasis on jihåd leave no doubt as to its propagandistic intent. It would also seem that the use of the Muslim declaration of faith here and in two other inscriptions in Hama was directed at the Christian population, which may have been quite large at that time.34 The Qurʾanic inscription in the sanctuary corroborates this interpretation. A small fragment of this inscription, done in a technique peculiar to the period of NËr al-DÈn,35 is still preserved on the easternmost pier of the qibla wall (Figure 1.8). By a happy coincidence, it contains the end of one Qurʾånic verse and the beginning of another: (IX. 23–4), so that we can be sure that the original inscription extended some length to both sides. In fact, for the sake of symmetry, we may assume that the inscription started on the other side of the mi˙råb with verses 20–2. These verses urge the believers to sacrifice their homes and possessions for God, exhort them to abandon the infidels even if they be their fathers and friends, and threaten that if family and friends are more important to them than jihåd, then they must await God’s punishment. In short, the messages are austerity, intoleration of non-Muslims, and jihåd.

11

12

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

The mosque contains a minbar with an inscription of NËr al-DÈn dated 1169 (Figure 1.9). According to the research of Becker and Sauvaget,36 the minbar was less a pulpit and more of a throne in the early Islamic period, which symbolised the secular authority of the caliph or his local governor. Neither author took his study beyond the Umayyad period, so the function of the minbar in later periods remains uncertain. There is little doubt, however, that it underwent a gradual loss of its religious function over the next few centuries. The actual process of this change remains to be studied, in order to determine the actual and symbolic function of the minbar in various places and times. In the absence of such a study, there are certain indications in the sources that, at least up to the twelfth century, the minbars of the main cities of Islam were still associated with caliphal and princely sovereignty. In the year 1110, for example, a delegation of Aleppan notables and learned men went to Baghdad to protest against the Frankish attacks on Aleppo. Ibn al-Qalånißi described the measures they undertook to publicise their grievances: the delegation went to the Jåmiʾ al-Sultan in Baghdad where they cried for help, brought down the khatib [speaker of the sermon] from the minbar, and broke it [the minbar]. They cried and wailed for what Islam had suffered at the hands of the Franks . . . And they prevented the congregation from prayer, while the servants and military commanders [muqaddams] tried to calm them down by promises that the Sultan would send soldiers to help them defeat the Franks and the infidels. They returned the following Friday to the Jåmi’ al-Khalifa where they did the same thing . . .37 [emphasis added] The act of breaking the minbars of the Sultan’s and Caliph’s mosque during the Friday sermon must have had the symbolic connotation that the Sultan and the Caliph, by failing to defend Islamic lands against the infidels, no longer deserved to be honoured every Friday on the minbars of Baghdad. Breaking their minbars was tantamount to challenging their political authority over the lands of Islam. There are other indications of the political significance of the minbar during the times of NËr al-DÈn and Saladin. Under NËr al-DÈn, the minbar served, among other functions, as the podium from which his sovereignty was affirmed and his ideology of jihåd was expounded. NËr al-DÈn took great interest in what was being said on his behalf on the minbars of his cities, as a letter to his vizier Ibn al-Qaysaråni clearly states. The letter and Ibn al-Qaysaråni’s reply describe the general topics that the khatÈbs discussed in the Friday sermon. In his reply, NËr al-DÈn urges his vizier not to allow the khatÈbs to indulge in their usual hyperboles about his greatness but, rather, to emphasise his sovereignty, humility before God, justice and, especially, jihåd.38 These topics are, in essence, expanded

monuments with a message

­ ersions of the titulary of NËr al-DÈn as it appears, for example, on v his minbar in Hama. Under Saladin, the secular function of the minbar seems to have been further emphasised, as indicated by Ibn Jubayr’s detailed description of the khutba (Friday sermon) ceremony in that period.39 We may conclude, then, that the minbar built by NËr al-DÈn for his mosque in Hama and another which he commissioned for the Aqsa mosque in Jerusalem bore the message of his sovereignty and dedication to jihåd. This assertion is perfectly consistent with the inscriptions on the Hama minbar. The minbar contains two historical inscriptions on long friezes which border the two balustrades; the left inscription has, however, been removed and is now preserved above the entrance. This is essentially an abbreviated version of the main inscription of the building, with emphasis on jihåd and justice. The right inscription is somewhat different, with emphasis on piety and humility. The cornice bears a Qurʾånic inscription which speaks about the wonder of God’s creation of the Heavens, and the back of the chair has the Shahåda in two lines of large thuluth within cartouches. Thus, the minbar, in its inherent political significance and in the meaning of its inscriptions, repeats and re-affirms the messages of the historical and Qurʾanic inscriptions of the mosque. The minbar which NËr al-DÈn commissioned for the Aqsa mosque of Jerusalem puts forth his most eloquent statement of jihåd (Figure 1.10). Finished in 1169, it was an outstanding monument in every respect until it was completely burnt on its eighth centennial by a Christian fanatic. Aesthetically, it was one of the most splendid minbars ever built.40 Ibn Jubayr, who saw the minbar in 1182 when it was temporarily kept in the Great Mosque of Aleppo, was quite impressed by it: The qarbasa craft has exhausted its resources in this minbar, for I have not seen in any other country a minbar which resembles its shape and the uniqueness of its manufacture . . . It rises like an enormous crown above the mihrab until it reaches the ceiling. Its top part is arched and open with balconies. It is all inlaid with ivory and ebony, and this inlay work continues to the mihrab and beyond to the qibla wall without any apparent division, so that the eyes enjoy one of the most beautiful sights in the world.41 The beauty of the minbar was matched by its great significance and forceful message, which are evident in its historical and Qurʾånic inscriptions and the unusual circumstances of its commission.42 From the year 1164, NËr al-DÈn focused his diplomatic and military efforts on the conquest of Egypt as a source of revenue to finance his war against the Latin Kingdom of Jerusalem and as an ideal place to launch unconstricted attacks against the Holy City.43 Indeed, by 1168 he had Egypt under his control, and his dream to restore

13

14

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 1.4  Hama. Mosque of NËr al-DÈn, Minbar. Inscription on back of chair, 559/1164 (photo: Yasser Tabbaa)

Jerusalem to Islam was a distinct possibility. It was around this time that NËr al-DÈn commissioned the minbar, with the apparent intention of placing it in the Aqsa mosque in Jerusalem. Needless to say, he was not able to fulfil his dream, and the minbar was kept in the Great Mosque of Aleppo in the meantime. The main historical inscription extended on a band bordering the four sides of the left balustrade (Figure 1.11). The superb thuluth script was made more legible by having it painted in white:44 (1) Basmala . . . Has ordered its making the slave, (2) the needy to His mercy, the thankful for His Grace, the defender against the enemies of His faith, al-Malik al ʿAdil NËr al-DÈn, the pillar of Islam and the Muslims, the helper of the oppressed against the oppressors, Abu-l-Qåsim MahmËd ibn Zanki ibn Åq Sunqur (3) the helper (to victory) of the Commander of the Faithful. May God make his victories glorious and preserve his power; may He raise his signs and spread in the two sides of the earth his standards and emblems; may He strengthen the supporters of his reign and humiliate those ungrateful of his favor; may He grant him conquest at his own hands and delight his eyes with victory and closeness to Him. At Your mercy O God of the Worlds. In the months of the year 564. This is the richest of all NËr al-DÈn’s inscriptions in proclamations of the victory of Islam and defeat of the infidels. While it contains a number of standard epithets (or formules banales) such as ‘the needy for His mercy’ and ‘the pillar of Islam and the Muslims’, the inscription also contains a number of unusual formulas and eulogies which seem to refer to the actual situation, namely NËr alDÈn’s wish to conquer Jerusalem. The rare epithet ‘the thankful for His Grace’ seems to have a subtle optative meaning: NËr al-DÈn is

monuments with a message

thanking God for the benefits which He will grant him in the future. Much more suggestive, however, is the eulogy fataha lahu wa-’alå yadaihi (that God may grant him conquest and make it happen at his own hands). NËr al-DÈn not only wishes for a Muslim conquest, he also asks God to bestow the honour of conquest upon him. The last eulogy, wa-aqarra bin naßri waz-zulfa ‘ayna˙u (that God may delight his eyes with victory and closeness to Him), adds a moral and religious tone to NËr al-DÈn’s plea for conquest, for he hopes that this conquest will bring him closer to God.45 Such a long series of emotional invocations is extremely uncommon in historical inscriptions, which rarely transcend the banal. How is one to explain such immediate sentiments; indeed, how is one to interpret the significance of a minbar built for Jerusalem when that city was still firmly in the hands of the Franks and was to remain so for twenty more years? One may glean answers to these inquiries from the statements of two contemporary historians, ʿImåd al-DÈn al-Ißfahåni and Ibn al-AthÈr. The first author wrote: After Jerusalem had been seized, Saladin gave the order . . . to place in the Aqsa mosque an inaugural (rasmi) minbar for the first day, as prescribed by religious law. But, in time a more splendid minbar was needed . . . Saladin then recalled that al-Malik al-ʾÅdil NËr al-DÈn had had one made for Jerusalem more than twenty years before the capture of the city . . . He wrote to Aleppo to claim it and have it transported to Jerusalem, thereby using it for the purpose for which it had been made . . . It is said that God had revealed in advance (alhama) to NËr al-DÈn . . . that Jerusalem would be captured after him . . . for he is one of His intimate confidants and servants honoured to receive His revelations. There was in Aleppo a carpenter (najjår) called al-AkhtarÈni who had no equal in the excellence of his craft. NËr al-DÈn ordered him to build a minbar for the Sacred House of God (Jerusalem), for which he was advised to employ all his skills . . . He, therefore, brought together many artisans, made an excellent design for the minbar, and dedicated several years to complete his work . . . In the meantime, it was said everywhere: ‘This is an impossible thing; this is an illogical opinion . . . Sooner to the sky that than Jerusalem should return to Islam . . . And the Franks, who will control them . . . becoming day after day more numerous . . . Were we not forced to share with them the majority of the provinces of the Hawran? Have they not opposed faith by heresy and have they not day after day defeated the Muslim princes? See to what degree of degradation we have fallen.’ But he [NËr al-DÈn] who possessed the force of certitude and who knew that God will confirm the victory of the true religion: ‘Patience’, said he . . . for he had foreseen with the light of divination that the conquest will be soon and that God will grant his request [concerning the minbar] only after this conquest . . . As it

15

16

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

happened, the blessing which he received from God was extended to Islam after him and sealed by the conquest of Saladin . . . The minbar had stayed in its place in the Great Mosque of Aleppo . . . until the present day, when Saladin ordered the fulfillment of the vow [nadhr] of NËr al-DÈn and had the minbar transported to the place for which it was destined in Jerusalem.46 Van Berchem described ʿImåd al-DÈn’s statement as ‘a living commentary on the inscriptions of the minbar’.47 Indeed, everything in it is in perfect accord, not only with the inscriptions of the minbar but also with the signature of the master carpenter al-AkhtarÈnÈ, which appears on the minbar along with four other signatures. There is, however, one important discrepancy, the author’s claim that NËr al-DÈn knew with certainty that God would not grant him the conquest of Jerusalem but was, rather, saving this favour for Saladin. This is highly unlikely for a number of reasons. First, relations between NËr al-DÈn and Saladin were already quite strained by the year 1168 (when the minbar was commissioned), primarily because of their differences with regard to jihåd policy and the possession of Egypt.48 Far from considering Saladin his successor and, therefore, worthy of the honour of having the minbar taken to Jerusalem, NËr al-DÈn rather thought of him as a disobedient officer, even a rival. Second, if the minbar was, indeed, not destined to be placed in the Aqsa mosque until the death of NËr al-DÈn, then why the pleas and eulogies and why ask God to grant victory to a dead person? Fortunately, ʿImåd al-DÈn himself gave the answer when he, perhaps unwittingly, described the minbar as a nadhr, or ex-voto. The minbar was, then, intended by NËr al-DÈn and understood in his time as an ex-voto for obtaining the capture of Jerusalem.49 Ibn al-AthÈr, the official historian of the Zangid dynasty of Mosul, repeated ʿImåd al-DÈn’s account of the minbar, but with one major difference: rather than re-stating the author’s convoluted story about NËr al-DÈn’s intentions, he quoted NËr al-DÈn as saying ‘we have made this [the minbar] to set up in Jerusalem’.50 While Ibn al-AthÈr has sometimes been rightfully accused of partiality to the Zangids against Saladin, this is one instance, I believe, when he was in the right. On the other hand, it is ʿImåd al-DÈn, the official historian of Saladin, who in this case may have twisted the facts to suit the orders of his patron. Ismåʾil, NËr al-DÈn’s young son and successor, added his own inscription to the minbar,51 stating that the minbar was finished under him. There are reasons, however, to doubt that Ismåʾil had any work performed on the minbar, with the exception of adding his own inscription. The minbar is dated 1169, and there is every reason to believe that it was finished by that date or, at the very latest, by the death of NËr al-DÈn in 1174. It is known, for example, that the minbar was already placed in the Great Mosque of Aleppo during

monuments with a message

the lifetime of NËr al-DÈn, a strong indication that it was, indeed, finished. This suggests that Ismåʾil’s inscription served a symbolic function only, to connect him with this ex-voto, in the hope that God would grant him the favour He had denied his father.52 As it happened, it was neither father nor son but Saladin who conquered Jerusalem in 1187. In the same year, while inspecting the works to restore the Muslim shrines of Jerusalem to their original use, he ordered that a minbar be built for the Aqsa mosque. It would seem certain that he knew about NËr al-DÈn’s minbar, but for reasons of sovereignty, he wanted to have his own. When he was reminded (according to Ibn al-AthÈr’s version of the story) that NËr al-DÈn had already had one specifically made for the Aqsa mosque, he, perhaps unwillingly, had the minbar brought to Jerusalem and placed in the Aqsa mosque. By this act, Saladin fulfilled NËr al-DÈn’s vision and symbolically asserted his sovereignty over Jerusalem, thirteen years after his death. Saladin had to settle for building the mi˙råb. He made sure, however, to have it inscribed with the statement that Jerusalem was re-conquered at his own hands (ʿalå yadayhi).53 If the minbar symbolised the sovereignty of a Muslim ruler over a particular city, the minaret was, perhaps, the most visible symbol of Islam. As such, the minaret served to assert the presence of Islam in a city or a particular locale, in addition to its function as a tower from which to call for the prayer.54 Between the years 1165 and 1170, NË al-DÈn had a number of minarets built: on the gates of the Damascus enclosure (Figure 1.12), in Raqqa, Qalʾat Jaʾbar, Mosul, and perhaps also in Maʾrrat al-Nuʾman.55 The symbolic, propagandistic intent is particularly evident in the otherwise useless minarets of the Damascus enclosure. These minarets were originally associated with tiny mosques – all disappeared now – which would not normally have any minarets at all. It would seem, then, that these minarets, of which only the one at Bab Sharqi (Figure 1.12) could date from the time of NËr al-DÈn, served to re-affirm the Islamic character of Damascus and its function as the centre of the jihåd of NËr al-DÈn.

Figure 1.5  Mosul. Great Mosque of NËr al-DÈn, 1170–2. Inscriptions on columns (photo: Yasser Tabbaa)

17

18

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 1.6  Aleppo. Madrasa al-Hallåwiyya. Western apse (photo: Yasser Tabbaa)

Figure 1.7  Aleppo. Qastal al-Shuʾaybiyya. Plan and elevation (Herzfeld)

monuments with a message

It is not certain who the builder of the minaret of the Great Mosque at Maʾrrat al-Nuʾman was, but the primary intent of such a tall minaret in a relatively small town with a Christian majority must have been to affirm the triumph of Islam. The same must be said about the minaret, indeed the whole mosque, of NËr al-DÈn at Mosul, dated between 1171 and 1173. During his last visit to Mosul, following the death of his younger brother, Qutb al-DÈn MawdËd, in 1170, NËr al-DÈn introduced a number of measures designed to undermine the power and position of the Christians.56 The building of the mosque and the tall minaret are consistent with these acts. Indeed, the interior of the mosque is filled with Qurʾånic inscriptions, on the capitals of the columns and on a long marble band. The inscriptions on the columns once formed parts of continuous Qurʾånic excerpts, but recent rebuilding of the mosque completely disturbed the original sequence. Presently, twenty-two columns bear parts of Qurʾån 2:255, 9:18–19 and 24:36–8, while the frieze has parts of 2:148–50. None of these verses deals with jihåd in any way; they are, rather, of a more general religious nature. In fact, they read like something of a primary lesson about the five pillars of Islam: unity and omniscience of God, prayer, pilgrimage and alms-giving. Only the fast of Ramadhan is missing, although that could have been included on some of the missing columns. I know of no other group of columns in a mosque which bear continuous verses, and their use here may have, indeed, been intended to remind new converts of their duties as Muslims.

Figure 1.8  Hama. Mosque of NËr al-DÈn. Fragment of Qurʾånic frieze (photo: Yasser Tabbaa)

19

20

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 1.9  Hama. Mosque of NËr al-DÈn. Minbar, 559/1164 (photo: Yasser Tabbaa)

By way of conclusion, I would like to briefly review the development of the jihåd message of NËr al-DÈn and the architectural form it assumed. One cannot help but see an element of pomp and stridency in the early monuments (the Madrasa al-Hallåwiyya and the qastal al-Shuʾaybiyya), buildings which use the architectural vocabulary of the Antique or Christian past to express the triumph of Islam over Christianity. This concept was repeatedly expressed by the Zangids of Mosul and Artuqids of Mardin and Hisn Kifa in their architectural reliefs and figural coins.57 NËr al-DÈn, however,

21

monuments with a message

Figure 1.10  Jerusalem. Aqsa Mosque. Minbar of NËr al-DÈn, 564/116 (CIA-Jerusalem, II, pl. XXX)

Figure 1.11  Jerusalem. Aqsa Mosque. Minbar of NËr al-DÈn, detail (CIA-Jerusalem, II)

quickly abandoned this obsession in favour of more Islamic forms of architectural expression: inscriptions, Qurʾånic verses, minbars and minarets. The message expressed by these monuments was no longer pompous and self-laudatory, but humble and directed at the affirmation of Islam and Mußlim unity.

22

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 1.12  Damascus. Minaret on Båb Sharqi (photo: Yasser Tabbaa)

Notes  1. For a summary of NËr al-DÈn’s warfare against the Crusaders see Hamilton A. R. Gibb, ‘The Career of NËr al-DÈn’, in The First Hundred Years, ed. Marshal W. Baldwin (Madison: University of Wisconsin Press, 1969), pp. 513–27, vol. 1 of A History of the Crusades, ed. Kenneth Setton. For a more detailed account see Nikita Elisséeff, NËr al-DÈn, un grand prince musulman de Syrie au temps des Croisades (511– 569/1118–1174), 3 vols (Damascus: Institut Français de Damas, 1967).   2. Emmanuel Sivan, L’Islam et la croisade. Idéologie et propagande dans les réactions musulmans aux croisades (Paris: Libraire d’Amerique et d’Orient, 1969).   3. Elisséeff, ‘La titulature de NËr al-DÈn d’après ses inscriptions’, Bulletin des Études Orientales (hereafter, BEO), 14 (1952–4), 155–96.  4. For a full discussion of Zangi’s titulature and the translation of his titles see Ernst Herzfeld, Materiaux pour un Corpus Inscriptionum Arabicarum. Troisième partie: Syrie du Nord. Inscriptions et Monuments

monuments with a message

d’Alep, vols 76–8 of Memoire. Institut Français d’Archéolgie Orientale du Caire (hereafter MIFAO) (1954–6), pp. 188–92.   5. Elisséeff, ‘Titulature de NËr al-DÈn’, p. 172. The author adds that Zangi bore the title za’Èm al-mujåhidin (leader of the fighters in the Holy War), a grand-sounding title which was, apparently, considered of less importance than the simple title al-Mujåhid.   6. Elisséeff, ‘Titulature de NËr al-DÈn’, p. 157. It is unlikely, however, that his inscription contained all the titles in Zangi’s inscription; consisting of only four lines (as opposed to twelve in Zangi’s inscription), it would have had to be three times as long. There would not have been enough room on the wall for such a long inscription.  7. Elisséeff, NËr al-DÈn, 2: 394–423.   8. Marshal W. Baldwin, ‘The Latin States under Baldwin III and Almaric I, 1143–1174’, in The First Hundred Years, p. 532.  9. Kamål al-DÈn ibn al-ʾAdÈm, Zubdat al-Halab fÈ TårÈkh Halab, ed. S. Dahhan, 3 vols (Damascus: Institut Français di Damas, 1951–68), 2: 291. 10. `Izz al-DÈn ibn Shaddåd, Al-A’låq al-KhatÈra fi Dhikr Umarå al-Shåm wa-l-JazÈra: La Description d’Alep, ed. D. Sourdel (Damascus: IFD, 1953), p. 45. See also Claude Cahen, La Syrie du Nord à l’époque des Croisades et la principauté franque d’Antioche (Paris: Paul Geuthner, 1940), pp. 275–6. 11. `Izz al-DÈn ibn Shaddåd, pp. 110–11. 12. The Madrasa al-Zajjajiyya, the first madrasa of Aleppo, was first started in 1116 by the Artuqid prince Salmån ibn Balak. It was not finished, however, until 1122, since it is said that the Shiʿis of the city kept tearing it down at night. See ʿIzz al-DÈn ibn Shaddåd, pp. 96–7 and Kamål al-DÈn ibn al-ʾAdÈm, 2: 210. 13. Michel Ecochard, ‘Note sur un edifice chrétien d’Alep’, Syria, 17 (1948), 270–83. 14. Elisséeff, ‘Titulature de NËr al-DÈn’, pp. 157, 177–80. 15. Herzfeld, 2: 212–13, makes a mistake by including the title atåbek in the inscription of the al-Hallåwiyya, when it clearly is not there. It is true, as he said, that ‘NËr al-DÈn n’est plus atåbek d’une souverain ture mais lui même souverain arabe’, but we must push back this change in titulature to before 1149. 16. On the titulature of the Zangids of Mosul, including Badr al-DÈn luʾluʾ, see Max van Berchem, ‘Monuments et incriptions de l’Atabek Luʾluʾ de Mossoul’, Opera Minora, 2 vols (Geneva: Éditions Slatkine, 1978), 2: 659–72. 17. Baldwin, ‘Latin States’, p. 533. 18. Gibb, ‘Career of NËr al-DÈn’, p. 515. 19. `Izz al-DÈn ibn Shaddåd, p. 44. 20. `Izz al-DÈn ibn Shaddåd, p. 44. 21. Jean Sauvaget, Alep. Essai sur le development d’une grande ville syriénne des origines au milieu du xix’ siècle (Paris: Paul Geuthner, 1941), pp. 74–5 and 122, n. 1. An antique attribution of this entablature has been accepted by Elisséeff in ‘Les monuments de NËr al-DÈn: inventaire, notes archéologiques et bibliographiques’, BEO, 13 (1951), 9; and Michael Rogers, ‘A Renaissance of Classical Antiquity in North Syria’, Annales Archeologiques de Syria, 13 (1971), 350. 22. Sauvaget, p. 75. 23. `Izz al-DÈn ibn Shaddåd, p. 12.

23

24

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

24. This is illustrated in Herzfeld, vol. 3, pl. LXXXIXa. 25. Herzfeld, ‘Damascus: Studies in Architecture II’, Ars Islamica, 10 (1943), 85, 105. 26. These two monuments are illustrated in Herzfeld, ‘Damascus’, figs 47 and 51. 27. In this respect, it is closely related to two twelfth-century monuments in the same region: the Great Mosque of Diyarbakr – particularly the west and east court façades (1124 and 1164, respectively) – and the Great Mosque of Harran, c. 1180. 28. These monuments are illustrated in John Hoag, Islamic Architecture (New York: Abrams, 1978), figs 25, 67 and 173. 29. Compare Sauvaget’s interpretation in ‘Notes sur quelques monuments musulmans de Syrie à propos d’une étude récente’, Syria, 24 (1945), 223 with Herzfeld’s in ‘Damascus’, p. 31. 30. See the statements of Shihåb al-DÈn abË Shåma, Kitåb al-Rawdatayn fi Akhbår al-Dawlatayn al-NËriyya wa-l-Salåhiyya, ed. Muhammad Hilmi Ahmad, 2 vols (Cairo Ministry of Culture, 1956), 1: 11; and ʿIzz al-DÈn ibn al-AthÈr, al-TarÈkh al-Båhir fi al-Dawla al-al-Atåbikiyya bi-l-Mawsil, ed. Ahmad Tulemat (Cairo: Ministry of Culture, 1963), p. 163. 31. Kamål al-DÈn ibn al-ʾAdÈm, 2: 313–15. 32. Elisséeff, ‘Titulature de NËr al-DÈn’, p. 160; and Répertoire Chronologique d’Épigraphie Arabes, 14 vols (1939–), vol. 9, no. 3248 (hereafter cited as RCEA). 33. Elisséeff, ‘Titulature de NËr al-DÈn’, pp. 158–9; and RCEA, vol. 9, no. 3220. 34. No study has yet been made on the Christian population of medieval Syria. There are, however, certain indications – for example, the large number of Christian commemorative shrines in north Syria – which suggest that the Christians were a sizeable minority in that region. See the list in ʿIzz al-DÈn ibn Shaddåd, pp. 40–2 and 45–59. 35. This technique entails carving out the inscription on a long slab of marble and filling it in with a black paste which hardens. The technique is seen first in the måristån of NËr al-DÈn in Damascus (1154) and subsequently in Hama, Aleppo and finally Mosul, where it becomes a standard feature of Zangid monuments. 36. Carl H. Becker, ‘Die Knazel im Kultus des alten Islam’, Islamstudien. Vom Werden und Wesen der islamischen Welt (Hildesheim: G. Olms, 1967), pp. 450–71; J. Sauvaget, La Mosquée omeyyade de Médine (Paris: Paul Geuthner, 1947), pp. 139–44. According to Sauvaget, the minbar in the Umayyad period was nothing other than le trône du chèf politique de la communauté (emphasis in the original). 37. Abu Yaʾla ibn al-QalånisÈ, Dhayl TarÈkh Dimashq (History of Damascus), ed. Hernandez F. Amedroz (Beirut: Beirut Catholic University, 1908), p. 173; and Ibn al-ʾAdim, 2: 521. 38. Abu Shama, 1: 30. 39. Muhammad ibn Jubayr, Rihlat Ibn Jubair (Beirut: Dar Sader, 1980), pp. 24–5 and 73. 40. Indeed, it is possible that this very minbar, by virtue of its especially sacred location and great political significance, may have provided the inspiration for many Mamluk minbars of Cairo and for the art of geometric ornament in general. 41. Muhammad ibn Jubayr, p. 227.

monuments with a message

42. My interpretation of this monument owes a great deal to a profound and innovative study by M. van Berchem in Matériaux pour un Corpus Inscriptionum Arabicarum. Deuxième partie: Syrie du Sud. Jerusalem, vols 43–4 of MIFAO (1922, 1927), 1: 393–402 (hereafter CIA-Jerusalem). 43. For a detailed study of NËr al-DÈn’s overtures toward the conquest of Egypt see Elisséeff, NËr al-DÈn, 2, ch. 4 and passim. 44. Elisséeff, ‘Titulature de NËr al-DÈn’, p. 163; and RCEA, vol. 9, no. 3281. 45. CIA-Jerusalem, 2: 401. 46. CIA-Jerusalem, 2: 398–400. 47. CIA-Jerudalem, 2: 400. 48. Elisséeff, NËr al-DÈn, 2: 671–84. 49. CIA-Jerusalem, 2: 401. 50. Francesco Gabrieli, Arab Historians of the Crusades (Los Angeles: University of California Press, 1978), p. 145, after Ibn al-Athir. 51. The text of this inscription is in CIA-Jerusalem, 2: 395; also RCEA, vol. 9, no. 3345. 52. IsmåʾÈl added another symbolic inscription to a completed monument of NËr al-DÈn, the maqåm IbrahÈm in the Aleppo citadel. See RCEA, vol. 9, no. 3345. 53. RCEA, vol. 9, no. 3423. 54. See Archibald G. Walls, ‘Two Minarets Flanking the Church of the Holy Sepulchre’, Levant, 8 (1976), 159–61, where the author documents two identical minarets of the fourteenth century situated on the city wall at either side of the church of the Holy Sepulchre. 55. Herzfeld, in ‘Damascus’, p. 39, argued that the minaret of the Great Mosque of Maʾrrat al-Nuʾmån should be dated to 1155 at the latest; i.e. well within the reign of NËr al-DÈn. But his argument is far from being conclusive. 56. Jean M. Fiey, Mossoul Chrétienne. Essai sur l’histoire, l’archéologie, et l’état actuel des monuments chrètiens de la ville de Mossoul (Beirut: Institut Français de Beirut, 1959), p. 38. 57. Estelle Whelan, ‘The Public Figure: Political Iconography in Medieval Mesopotamia’, Diss. New York University, 1979.

25

CHAPTER TWO

Survivals and Archaisms in the Architecture of Northern Syria, c. 1080–1150 For Creswell, chronology was ‘the spinal column of history’,1 a dogma that permeated his voluminous work and influenced the scholarship of many of his successors. Ultimately based on archaeological documentation, this positivist methodology was designed to create order where disorder had been perceived and to equip a new field with a scientific foundation. Typically, his scheme attempted to fit the monuments – or particular phases of them – into a neat chronological sequence and to provide each with a historical background, an objective description, and a large number of comparisons intended to ascertain its date or determine the various influences impinging upon it. The field of Islamic architecture as we know it today grew out of this artificial scheme, rarely departing from its rigid progression and attempting instead to fit all new findings within its pre-established categories.2 But the apparent neutrality and objectivity of Creswell’s chronological evolution are deceiving, for they implicitly endorse a false sense of continuity across the various epochs and geographic regions of early and medieval Islamic architecture. Recalling Riegl’s outmoded concept of Kunstwollen with its central belief in continuous and autonomous evolution of forms and motifs, this method has led to at least two unfortunate tendencies in the field. The first is to assume a rather monolithic picture of Middle Eastern Islamic architecture, in which forms and styles are subject to gradual change and continuous development. The second is to focus scholarly interest on a ‘central style’ of Islamic architecture to the neglect or misunderstanding of those periods which do not directly stem from or contribute to it. On the whole, these ‘intrusive traditions’ were either ignored or artificially brought into accord with the dominant tradition. Also contributing to these centralising tendencies were non-historical essentialist views reflecting the field’s earlier association with Orientalism – which, though totally opposed to positivist Yasser Tabbaa (1993), ‘Survivals and Archaisms in the Architecture of Northern Syria, ca. 1080–ca. 1150’, in Essays in Honor of Oleg Grabar, Muqarnas 10, 29–41.

survivals and archaisms

methodology, were easily and unconsciously incorporated into the discourse.3 One could instead view the development and increasing variation of Islamic architecture between c. 1000 and c. 1350 at least in part as a series of discontinuous events brought about by geographic dislocation, dynastic change, racial intrusions, sectarian differences and other factors. With the possible exception of Cairo, which had a uniquely continuous architectural history in the Middle East, Islamic cities – the centres of architectural patronage – prospered and ebbed in fairly rapid succession: one day at the vanguard of Islamic architecture, the next a forgotten backwater. By accepting, instead of ignoring, these basic discontinuities, we may arrive at a different perhaps less dogmatic and pre-determined – view of Islamic architecture. In particular, the flexibility and non-essentialist nature of this approach might more fairly and accurately address and assess those periods which seem to stand apart from the main tradition. The buildings of northern Syria between the late eleventh and the middle of the twelfth century represent perhaps one of the most intrusive phases of medieval Islamic architecture. Their singularity, in their reliance on ashlar masonry and archaising motifs, has intrigued several writers, including Herzfeld in 1921 and later,4 Creswell in 1952,5 Grabar in 1963 and later,6 and Rogers in 1971,7 all of whom characterised it as a survival or revival of classical antiquity. Emphasising its most non­-Islamic features and comparing it with contemporary Persian architecture, Herzfeld saw it as part of an ‘uninterrupted classical tradition’.8 Most recently, the problem has been discussed by myself, in 1982 and later,9 and by Allen, who wrote a book-length study on it in which he proposed to explain it as a conscious and purposeful revival of classical antiquity.10 In this piece the archaising architecture of northern Syria is regarded first and foremost as a consequence of the region’s relative isolation and insignificant architectural activity between the eighth and the eleventh centuries. In addition to dealing with the stylistic and iconographic aspects of this architecture, the piece will address the much more basic problem of the use of stone, a local material which had been supplanted elsewhere in the Islamic world by baked brick. Third, every attempt is made to provide a regionally based interpretation for this entirely local production, without resorting to ambiguous art historical explanations such as ‘influences’ or grandiose ideas such as the revival of classical antiquity.11 Finally, the piece will outline some of the significant changes that were applied to this archaising architecture, changes that easily and assuredly situate it in its context. Throughout, the distinction between the creation or perpetuation of an archaistic style and the re-use of salvaged ancient materials is made as clear as possible.12 I am interested in the former but not the latter, which is so pervasive a phenomenon in the Islamic world that

27

28

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 2.1  Split. Palace of Diocletian. Reconstruction of Porta Aurea, c. ad 300. (From Ward-Perkins, Roman Imperial Architecture, fig. 311)

it could not have had a single meaning, and whatever meanings it had derived from the local context.13 In any case, the two processes are entirely separate and cannot be given the same explanation. North Syria’s isolation The violent end of the Umayyads in 750, the subsequent eastern shift of the Islamic empire, and the vehemence displayed by the early ʿAbbasids in rejecting Umayyad architecture and creating their own brought a glorious phase of Islamic architecture to an end. For over three centuries that followed Syria languished as a backwater between the dominant forces in Iraq and Egypt; invaded and occupied by the Byzantines in the ninth century; subjected to Fatimid rule and tribal intrusions in the tenth and eleventh centuries; and finally devastated by two great invasions, the Crusades and the Seljuq conquest. It is not surprising, therefore, that very little was built there between 750 and 1100, years when elsewhere some of the greatest buildings of early Islamic architecture were constructed. In the same period Islamic architecture became increasingly independent from its classical sources and shifted progressively toward Perso-Mesopotamian forms and techniques. The overall trend might be adequately illustrated by looking at one significant

survivals and archaisms

Figure 2.2  Cordoba. Great Mosque. Gate on the eastern façade, 987 (photo: Yasser Tabbaa)

Figure 2.3  Samarra. Jawsaq al-Khaqani. Bab al-ʿAmma, c. 836–7 (photo: Yasser Tabbaa)

29

30

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 2.4  Aleppo. Great Mosque. Domed fountain in courtyard, 965 (photo: Yasser Tabbaa)

form, that of the gate. The gates of several Umayyad palaces, such as Khirbat al-Mafjar and Qasr al-Hayr al-Gharbi, are clearly derived from Roman ceremonial gates, for example, the early fourth-century Porta Aurea at the palace of Diocletian at Spalato (Figure 2.1). The same gate form can also be seen in a number of Umayyad desert palaces and at the Great Mosque of Cordoba, primarily in its ninth- and tenth-century phases, although much flattened and made more ornamental (Figure 2.2). It is, however, entirely absent from ʿAbbasid architecture. The Bab al-ʿAmma at Samarra, for example, retains only the tripartite composition of the gate form, but is otherwise eastern in its use of brick, the pointed arch and the iwån form (Figure 2.3). Although few monuments remain in northern Syria from the ʿAbbasid period, there are no indications that Mesopotamian brick architecture had much of an impact on the region either. Certainly no brick architecture can be found in the Great Mosque of Aleppo – which must be regarded as an exemplar of what might have taken place in the intervening period – nor anywhere else in the city. On the contrary, it seems from the few monuments preserved from the short-lived period of architectural activity during the reign of Sayf al-Dawla, the Hamdanid ruler of Aleppo between 945 and 967, that building in stone continued virtually unaffected by ʿAbbasid architecture.14 This is amply illustrated by the first phase of the Mashhad al-Muhassin (or al-Dakka), which consists of two small squinch domes resting on solid masonry walls, and the domed fountain in the courtyard of the Great Mosque, which is not only built of stone but even has obvious classicising features (Figure 2.4).15 In other words,

survivals and archaisms

Figure 2.5  Aleppo. Great Mosque. Minaret. 1090

even in its period of utmost contraction and impoverishment, the stone architecture in Aleppo remained close to its antique sources and largely impervious to eastern influences. These features of local continuity and conservatism are therefore at the crux of understanding what took place in northern Syria between the late eleventh and early twelfth centuries. Previous writers, Allen in particular, have insisted that this period was one in which classical architecture was revived, and this has created a false impression of the extent of the problem. We should emphasise, therefore, that this was an entirely localised phenomenon, centred mainly in the region between Aleppo and Edessa, which are only about 200 kilometres apart, and involving only a few monuments. Stone buildings with similarly archaising features

31

32

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 2.6  Aleppo. Great Mosque. Minaret. Detail of lower shaft (photo: Yasser Tabbaa)

are known neither in Damascus nor Mosul and, with one notable exception, they are not built after the middle of the twelfth century. The rise of stone architecture Seen against the overall development of Islamic architecture between the eight and the eleventh centuries, the late eleventhcentury monuments of northern Syria are quite remarkable. But seen within the context of the stone architecture of northern Syria, they are remarkable only for their profusion and high quality. The new period of architectural ascendance in northern Syria begins with a bang. The minaret of the Great Mosque of Aleppo, erected in 1090, is unmatched in Syria in its rigour of proportion and elegance

survivals and archaisms

33

Figure 2.7  Syria. Church at Qalb Lozeh, mid-sixth century. Façade and western flank (photo: Yasser Tabbaa)

of ornamentation (Figure 2.5). With a square shaft divided into five zones by classicising cornices and inscriptional friezes, this minaret owes very little to its contemporaries in Iran and Central Asia. Indeed, the excellence of its masonry, its relatively uncluttered surfaces and the use of restrained but very well-executed mouldings all recall the architecture of northern Syria in the fifth and sixth centuries ad.16 Perhaps the most striking similarity between this late eleventh-century minaret and the early Christian architecture of northern Syria is the use of a decorative device which is commonly called the continuous moulding (Figure 2.6). In the sixth-century churches of Qalb Lozeh and Meez for example, a single moulding runs along all Figure 2.8  Cairo. Bab the openings of the building, simultaneously framing al-Nasr, 1087–8. Detail of each one and joining them all together (Figure 2.7). tower (photo: Yasser Quite often the moulding even turns around the edge Tabbaa) of the building, connecting its two façades together, a phenomenon also seen in the eleventh-century minaret. Contrary to classical architectural sensibility, which relies on the superimposition of orders to establish hierarchy and achieve monumentality, early Christian Syrian churches almost dispense with the orders and rely instead on large masses that are enlivened and interconnected by the continuous moulding. Another characteristic moulding common to both the minaret and the church is a cuspated moulding that runs along the outer edge of the continuous moulding.

34

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 2.9  Aleppo. Minaret of the Great Mosque, 1090, from south (photo: Yasser Tabbaa)

Of course, despite all these similarities, the minaret of the Great Mosque of Aleppo contains unmistakably Islamic features: muqarnas in its topmost cornice, polylobed arches in the fourth and fifth zones and Arabic inscriptions throughout. It is signed by Hasan ibn Mufarrij (or Mufrih) al-Sirmani, an artisan from the small town of Sarmin near

survivals and archaisms

Figure 2.10  Cairo. Bab Zuwayla, 1092 (photo: Yasser Tabbaa)

Aleppo. Artisans with the same nisba are also responsible for a few other related monuments in the same region, most importantly a stone mi˙råb dated 505 (1111) in Aleppo and the minaret of the Great Mosque of Maʿarrat al-Nuʿman, datable to c. 1190.17 At almost exactly the same time that the minaret of the Great Mosque of Aleppo was being erected, the Fatimid city of Cairo was building its first truly monumental enclosure with three magnificent gates. These gates – Bab al-Nasr (Figure 2.8) and Bab al-Futuh in the north and Bab Zuwayla (Figure 2.10) in the south – were built between 1087 and 1092 by the energetic Armenian vizier Badr al-­Jamali. Universally admired by travellers and scholars alike since the early eighteenth century, these gates were systematically studied by Creswell, who pronounced them among ‘the great masterpieces of military architecture of lslam’.18 Even a cursory look at these gates should suffice to illustrate their similarity to the Aleppo minaret and ultimately to the same school

35

36

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 2.11  Aleppo. Minaret of the Great Mosque, 1090, detail (photo: Yasser Tabbaa)

of late antique architecture in northern Syria. In the most general sense, monumentality is achieved here as in early Christian Syrian architecture, not by the use of orders, but by the sheer mass of lightly ornamented forms conceived in superb ashlar masonry. In terms of

37

survivals and archaisms

Figure 2.12  Turkey, Dunaysir (Kiziltepe). Great Mosque, 1204. Detail of arched opening in façade (photo: Yasser Tabbaa)

ornament, the continuous moulding is also present here, used in conjunction with other specifically north Syrian features such as the cuspated moulding, joggled voussoirs, cushion voussoirs, and arches with stepped extrados. Finally, even the building material is alien to medieval Egypt, all of whose earlier Islamic buildings were built in brick or brick-shaped stones. With all these similarities, it comes as no surprise that according to Maqrizi, these gates were built by three Christian architects from the north Syrian town of Urfa, or Edessa, situated about 200 kilometres north-east of Aleppo.19 A case of artisanal conservatism

Figure 2.13  Turkey, Hisn Keyfa

We therefore have in the minaret of the (Hasankeif). Mosque of Süleyman Pasha. Great Mosque of Aleppo and its cog- Minaret, 1406 (photo: Yasser Tabbaa) nates and the eleventh-century gates of Cairo two groups of monuments built by two teams of artisans from northern Syria, both still working within the specifically north Syrian late-antique vocabulary. How was this architectural style transmitted over the centuries, and what factors led up to its ­transmission?

38

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 2.14  Turkey, Mardin. Traditional residence, late nineteenth(?) century. Upper storey windows (photo: Yasser Tabbaa)

Specifically, how was it possible for artisans from Sarmin and Edessa to perpetuate a distinctive and distinguished architectural style through a long period of minimal patronage? We have two sets of evidence to help us in addressing these questions: the first consists of the signatures of the artisans and tales about their origins, and the second is made up of the monuments themselves in relation to what had preceded them and what came after them in the same region or in adjacent ones. Unfortunately, looking for the answer in the towns of origin of the stone masons, namely Sarmin and Edessa, is rather fruitless. Sarmin, today a small village with hardly any monumental architecture, could not have been any more than the place of birth of these artisans – possibly the location of their first workshop – but certainly not the centre of their artisanal practice. Edessa, a much larger town and an important pilgrimage centre, was totally sacked in 1148 by NËr al-DÈn, thereby wiping out the record of its earlier churches. Nevertheless, one can at least maintain from what was subsequently built in Edessa that the city continued its excellent tradition of stone architecture, both monumental and residential. Expanding our area of inquiry to adjacent territories, we should perhaps take into consideration the parallel development of Armenian architecture, since a fairly continuous series of churches remains from the early medieval period. One can suggest, for example, that in a time of declining patronage, Syrian masons could have sought employment in Armenia and stayed there until the market improved

survivals and archaisms

Figure 2.15  Aleppo. Qastal al-Shuʿaybiyya. 1150–1 (photo: Yasser Tabbaa)

in northern Syria. This intriguing possibility has not been studied exhaustively although, thanks to a recent publication, several stylistic connections between northern Syria and Armenia, including the continuous moulding, can now be identified.20 But there are a number of problems with these purely formal links. The first is that without knowing the relative chronology of Syrian and Armenian monuments, it is impossible to tell who in fact is appropriating the stylistic features of the other. The second is that, although the two styles are generally related, the medieval architecture of northern Syria, unlike that of central Anatolia, does not contain any specific Armenian features. And the third is that, as far as we know, there is no direct evidence for the existence of Syrian masons in Armenia. At this stage of our knowledge, perhaps the best explanation for the presence of similarly archaising features in medieval Syrian and Armenian architecture is that they both drew on a common pool of late-antique vocabulary. With the possibility of a direct Armenian influence diminished, the most likely remaining explanation is that the monuments under consideration represent an architectural survival in a marginalised region of the Islamic world. We may envision a situation where the sharp decline in architectural patronage between the eighth and eleventh centuries would have forced the majority of north Syrian masons out of their ancient craft, leaving only a few to perpetuate the tradition. These masons would have remained in close contact with their past tradition while absorbing little from the changing forms of Islamic architecture. The old forms would have undergone various processes of development while preserving the overall

39

40

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 2.16  Aleppo. Qastal al-Shuaʿybiyya. Profile of the entablature (photo: Yasser Tabbaa)

aesthetic of the original style.21 Such regional conservatism is not as exceptional as it may seem, and one can easily point to similar pockets of traditionalism in North Africa between the twelfth and fifteenth centuries and eastern Anatolia up to the eighteenth century. In fact, one need not search outside northern Syria and southern Anatolia for a situation very similar to the one that seems to have occurred between the sixth and eleventh centuries. Travelling in the hills of southern Anatolia between Hisn Keyfa (present Hasankeyf), Mardin and Urfa (Edessa), one comes across numerous monuments built in excellent ashlar masonry whose predominant decorative feature is precisely the continuous moulding. Acanthus ornament turns to arabesque and even to late Ottoman floral designs, and the script changes from KËfic to cursive, but the continuous moulding meandering over the uncluttered stone surfaces remains unchanged. Examples ranging from the façade of the Great Mosque at Dunaysir

survivals and archaisms

(1204) (Figure 2.12), to the minaret of the Süleyman Pasha Cami at Hisn Keyfa (1406) (Figure 2.13), to the sixteenth-century minaret at the Great Mosque of Mardin, to the nineteenth-century restoration of the Great Mosque of Mayyafariqin demonstrate an astonishing degree of conservatism. Even nineteenth- and early twentieth-century domestic architecture in Mardin employs this decorative device in much the same way as it was used in the fifth-century churches of north Syria. In one nineteenth-century residence (Figure 2.14), the architect has used the continuous moulding to outline the Italianate windows – perhaps something he saw in Istanbul – but was compelled to continue the moulding around the corner as had been done one and a half millennia earlier in the Church of Qalb Lozeh (cf. Figure 2.7). This extraordinary degree of continuity between the eleventh and the nineteenth century in a restricted part of south-eastern Anatolia raises a number of questions. The first is why the architecture of Aleppo and northern Syria took a rather different course of development. There, after the third quarter of the twelfth century, the continuous moulding and other archaising features are generally absent, most likely because after 1170, Aleppo was the capital for an important branch of the Ayyubid dynasty and as such closely related to Damascus, Cairo and the rest of the Islamic world. Its new architecture was no longer restricted to the local tradition, but could draw on all the forms and images that had begun to define medieval Islamic architecture.22 The second and, for our purpose, more important question is the following: since a measure of continuity between the eleventh and the nineteenth century has been demonstrated in this piece, is it not also possible between the sixth and the eleventh centuries? Indeed, as has been discussed above, the absence of an intrusive ʿAbbasid phase during the period in question, the presence instead of certain elements of continuity, and, most importantly, the endurance in the more isolated parts of the region of some basic archaistic elements up to the nineteenth century make it probable that we are dealing with a long-term survival of a regional style, not just from the sixth to the eleventh centuries, but in some respects all the way up to the modern period. This, in turn, greatly diminishes the likelihood that these northern Syrian monuments were part of a sweeping classical revival. These conclusions have only limited implications and are not intended to undermine the case for revivals in Islamic architecture generally. Revivals did indeed take place in some periods of Islamic architecture, although – before the sixteenth century, in any case – they do not seem to have involved the emulation or adaptation of a particular style or elements of it, but rather the imitation of entire monuments which for a variety of aesthetic, political or pious reasons had entered the collective consciousness of medieval Islam.

41

42

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 2.17  Turkey, Harran. Great Mosque. Fragment of entablature, 1174. Urfa Museum (photo: Yasser Tabbaa)

These include the Dome of the Rock in Jerusalem, the Great Mosque of Damascus, the Great Mosque of Isfahan, and perhaps a few others. The legacy of these monuments for the petty dynasties of the Islamic Middle Ages cannot be overestimated, and their inspiration for numerous buildings across the Islamic world is a phenomenon in need of further study.23 Two odd monuments The conclusions presented above still leave unexplained two other monuments – the Qastal (fountain) al-Shu­ʿaybiyya of 1150 in Aleppo and the Great Mosque at Harran, datable to the last quarter of the twelfth century – which are often included in the discussion of this phenomenon. The two monuments do betray certain affinities with classical architecture, but their respective styles are so entirely different from that of the Aleppo minaret and the Cairo gates as to make it inconceivable to include them under our concept of regional continuity.24 In fact their very individual style is perhaps best explained by the specific circumstances of their creation. I have elsewhere investigated these circumstances for the Qastal al-Shuʿaybiyya (Figure 2.15), suggesting that it was built by NËr al-DÈn in 1150 as a victory monument to commemorate his successes against the Crusaders of Antioch.25 The specificity of the interpretation, which is totally in keeping with NËr al-DÈn’s jihåd propaganda, best explains the uniqueness of the monument, which

survivals and archaisms

in fact has no parallel in Aleppo or anywhere else. At the same time, however, the Qastal al-Shuʿaybiyya displays a number of features which counteract its classical appearance and bring it closer to contemporary monuments. These include its highly elaborate KËfic inscriptions and arabesque decorations, the pointed arches supporting the entablature, and especially the peculiar profile of the entablature itself, whose exaggerated horizontal projection gives it a somewhat unsupported appearance in keeping with medieval Islamic practice (Figure 2.16). The Great Mosque of Harran remains an enigma, and only a complete re-evaluation of its previous excavations coupled with renewed efforts for excavation and restoration will clear up its chronology and archaeology.26 Meanwhile, its existing remains, in situ and at the Urfa Museum, seen to display a much greater degree of synthesis between classical and Islamic architecture. Although its heavy and overly elaborate entablatures (Figure 2.17) are comparable with such early Christian monuments as the fifth-century gateway of Rusafa and the fifth-century entablature over the altar of Dayr al-Zaʿfaran near Mardin, they incorporate numerous Islamic features, including arabesques, cursive inscriptions and foliate arches. But like the Qastal al-Shuʿaybiyya, the Great Mosque at Harran (Figure 2.17) is an entirely singular building, with no real precedents and no existing antecedents. This is all the more remarkable in view of the excellence of its stone carving and the happy mix it displays of ancient and medieval features.

Notes 1. K. A. C. Creswell, Early Muslim Architecture, vol. 1, pt. 1 (Oxford University Press, 1969), p. xxxv (hereafter, Creswell, EMA). 2. Exceptions, such as Burckhardt and Papadopoulo, of course exist, especially in the non-historical school of Islamic art. But within the accepted boundaries of the field, far fewer studies have attempted a theme- or problem-oriented approach. Foremost among these is Oleg Grabar, The Formation of Islamic Art (Yale University Press, 1973; rev. edn, 1987). 3. A similar situation is described by James S. Ackerman, ‘Toward a New Social Theory of Art’, New Literary Theory 4, 2 (Winter, 1973): esp. 318–20. 4. The phenomenon is first discussed by him in Ernst Herzfeld, ‘Mshatta, Hira und Bådiya: die Mittellander des Islam und ihre Baukunst’, Jahrbuch der preussischen Kunstsammlungen, 42 (1921): 104–46. But his most complete statement on the question can be seen in ‘Damascus – Studies in Architecture, II’, Ars Islamica 10 (1943): 30–40. 5. K. A. C. Creswell, The Muslim Architecture of Egypt, 2 vols (Oxford: Clarendon Press, 1952), 1: 165–216, passim, esp. pp. 206–16 (hereafter, Creswell, MAE). 6. Although primarily concerned with the survival of late-antique forms in Umayyad art and contacts with the Byzantine Empire and only

43

44

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

­ arginally with medieval Syria, Grabar’s several articles on the quesm tion provide a very useful conceptual framework. See especially, ‘Islamic Art and Byzantium’, Dumbarton Oaks Papers, 18 (1964): 69–88; and ‘Survivances classiques dans l’art de l’Islam’, Annales archeéologiques arabes syriennes, 21 (1971): 371–80. 7. J. Michael Rogers, ‘A Renaissance of Classical Antiquity in North Syria’, Annales archéologiques arabes syriennes, 21 (1971): 347–56. 8. Herzfeld, ‘Damascus-II’, pp. 32 and 39. 9. Yasser Tabbaa, ‘The Architectural Patronage of NËr al-DÈn, 1146–1174’, Ph.D. diss. New York University, 1982, and ‘Classical Survivals (or Revivals?) in Medieval Islamic Architecture’, paper delivered at Society of Architectural Historians, Pittsburg, 1984. 10. Terry Allen, A Classical Revival in Islamic Architecture (Wiesbaden: Ludwig Reichert Verlag, 1986). For Allen (p. ix), the existence of a ‘classical revival in Islamic architecture that developed in Syria and the neighboring lands during the . . . eleventh and twelfth centuries a.d.’ is a given: it is used in the title of his book and stated in the introduction. The book is not intended to demonstrate that such a revival did in fact take place, but rather to describe the monuments affected by it and to provide an interpretation for the phenomenon. 11. The question of scale comes up again and again in a field in which priorities are still very poorly defined, and the natural tendency to inflate the problem that one is addressing can misguide even the most conscientious scholar. 12. Although he simultaneously dealt with both matters, Herzfeld did in fact make this crucial distinction: ‘Damascus-II’, p. 39. Rogers largely ignored it, and Allen (Classical Revival, pp. 57–67 passim) went so far as to base his interpretation on both the re-use of ancient materials and the creation – as he sees it – of a classicising style. 13. For example, it is hardly possible that the extensive re-use of pre-Islamic material at the Great Mosque of Diyarbekir and the Great Mosque of Quwwat al-Islam in Delhi should have the same explanation. In neither case is one entitled to speak of revival, whether of classical antiquity or Hindu architecture. 14. No attempt has been made to study the architectural patronage of the Hamdanids in Aleppo and Mayyafariqin. Meanwhile, see Rmazi J. Bikhazi, ‘The Hamdanid Dynasty of Mesopotamia and North Syria (254–404/868–1014)’, 3 vols, Ph.D. diss., University of Michigan, 1981, pp. 628–30. 15. Ernst Herzfeld, Materiaux pour un Corpus inscriptionum Arabicarum, deuxième partie: Syrie du Nord. Inscriptions et Monuments d’Alep, vol. 1, pt. 1 (Cairo, 1955): 143–4. An inscription that has since disappeared gave the date of this fountain as 354 (965). 16. For a fuller description, see ibid., pp. 151ff. 17. These are discussed in greater detail in Tabbaa, ‘NËr al-DÈn’, pp. 38–47. 18. Creswell, MAE, 1: 165–6. 19. 19. Ibid., pp. 161–2. 20. O. Kh. Khalpakhchian, Architectural Ensembles of Armenia (Moscow: Iskusstvo, 1980); for example, pp. 274–5, figs 7–13, all of St Astvatsatsin (1201) at the Harich monastery. 21. Tabbaa, ‘NËr al-DÈn’, pp. 44–5. 22. Links with past architectural tradition do continue in the later medieval architecture of Aleppo, but on a much deeper level. Of these, we

survivals and archaisms

may mention the use of the peristyle courtyard in some later Ayyubid monuments and the reliance on geometric planning. While both of these architectural devices echo classical practice, they are so deeply embedded in the architecture of northern Syria as to be simply part of the local tradition. 23. See, meanwhile, Ernst Herzfeld, ‘Damascus: Studies in Architecture, IV’, Ars Islamica, 13–14 (1948): 118ff. and K. A. C. Creswell’s discussion of the monumental antecedents of the funerary complex of Sultan Qalaʾun in Cairo in The Muslim Architecture of Egypt, vol. 2: AyyËbids and Early Bahrite Mamluks a.d. 1171–1326 (New York: Hacker Art Books, 1978), pp. 198–204. See also Oleg Grabar’s brief but perceptive remarks on this matter in The Great Mosque of Isfahan (New York and London: New York University Press, 1990), pp. 69–72. 24. Here, one must take issue with Herzfeld’s statement that the qastal represents an uninterrupted classical tradition: see above, p. 4. 25. First in ‘NËr al-DÈn’, pp. 220–4, and later in ‘Monuments with a Message: Propagation of Jihåd under NËr al-DÈn’, in V. Goss (ed.), The Meeting of Two Worlds: Cultural Exchange between East and West during the Period of the Crusades (Kalamazoo, MI: Medieval Institute Publications, 1986): 223–41. 26. For bibliography and history of the excavation of this monument, see Allen, Classical Revival, which also provides hitherto unpublished plans and drawings of the excavation by D. S. Rice.

45

CHAPTER THREE

Circles of Power: Palace, Citadel and City in Ayyubid Aleppo Citadels were intended to dominate cities and in this respect the citadel of Aleppo perhaps succeeds a little too well. Built over a partly natural outcrop with numerous building levels, it soars above the city like a great cone with a chamfered top. Its massive elliptical base, measuring 450 metres on its east–west axis and 325 metres on its north–south axis, tapers steeply to a smaller but still quite large ramparted ellipse, 285 by 160 metres (Figure 3.5). Nearly fifty metres high with its ramparts intact, its already dominant image would have been immeasurably more powerful when its glacis was still paved with polished limestone and its moat filled with water. Its height, impregnability and apparent timelessness have long fascinated poets: A citadel whose base was embraced by the Pleiades while its peaks surpassed the orbit of Gemini Its watchtower would be counted among the celestial bodies were it only to move in their courses Arrogant, it laughs in the face of Time, who has long ridiculed such lofty buildings.1 But this hyperbolic imagery of power seems flawed or, one might say, it is inhabited by a number of paradoxes. The citadel of Aleppo, like other urban citadels, was not just a military garrison but also the royal palace and the centre of government and administration. Although constructed materially and metaphorically as a closed system, it had by necessity to interact with the city and its population. Furthermore, it was a hard shell with a soft centre, an image of power and dominance that envelops and protects an atmosphere of royalty and elegance, the palace of the Ayyubid dynasty of Aleppo.2 These visually arresting contrasts between impregnability and linkage, militarism and luxury may be taken at face value as true reflections of the distant and strained relationship between ruler Yasser Tabbaa (1993) ‘Circles of Power: Palace, Citadel and City in Ayyubid Aleppo’, in Pre-Modern Islamic Palaces: Ars Orientalis 23, 181–200.

circles of power

and ruled in the Islamic Middle Ages. Foreign, in this case Kurdish, dynasties rule an Arab population; their illegitimate status and distance from the population are manifested in the semiotics of the very architecture of power that they create. While this is generally true, this piece attempts in the first place to provide nuance for this seemingly stark separation between the enclosed and the displayed, between citadel and city, and consequently between ruler and ruled.3 This will be done by examining the religious, pietistic, ceremonial and administrative links the citadel had to the city. In the second place, the piece examines the Ayyubid palace in the citadel, proposing that, despite its modest proportions, it was formally and iconographically associated with a well-established iconography of power, an iconography that was first conceived in classical Islamic palaces and later perpetuated in the palaces of the Ayyubid and Artuqid dynasties. Before proceeding with this investigation, it seems necessary to stress the fact that the institution of the palatial citadel is not contemporaneous with the beginning of Islam but represents the third stage in a series of discrete urban transformations. The first stage is represented by a centrally located palace (dår al-imåra) adjacent to the great mosque, as in Kufa or the Round City of Baghdad.4 In the second stage the palace shifted to a suburban location, a development initiated by the ʿAbbasids in Samarra and imitated by many tenth- and eleventh-century dynasties, including the Hamdanids of Aleppo.5 The third and equally pivotal stage occurred when the Samarran model for royal architecture was gradually supplanted by that of the citadel palace. Aleppo was the first city to undergo this transformation, which occurred in the middle of the eleventh century under the Mirdasids.6 Contexts The fortification and building of various parts of the citadel continued unabated during the successive periods of the Seljuqs, the Zangids and the early Ayyubids. But the citadel – in fact the entire fortification of Aleppo – did not come into its own until the time of al-Zahir Ghazi (1186–1216), the most important architectural patron in the medieval history of the city. Contemporary sources and epigraphic evidence (in the form of five preserved inscriptions inside the citadel)7 attest to the fact that he was responsible for the total rebuilding of the citadel, including its impressive ramp and entrance block, its mosque and its palace, and for the restoration of the Maqam Ibrahim. Indeed, one might even suggest that nearly all subsequent work on the citadel and the fortifications of Aleppo by his successors proceeded in accordance with his plan. At the highest point in the citadel, al-Zahir Ghazi in 1214 built a mosque whose lofty minaret even today dominates the city

47

48

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

(Figure 3.6). It was one of the first mosques in an urban citadel, and although more symbolic than functional, it served as a model for most other citadels, including those in Damascus and Cairo. As the primary Islamic institution, the mosque at this site declared Islam triumphant in a city that a century earlier had almost fallen to the Crusaders, while it also affirms the role of the Ayyubids as guardians of the faith. The great height of the minaret (21 metres), on top of 40 metres for the citadel, created a towering structure visible from every comer of the city. Conversely, many parts of the city were at least potentially visible from the minaret, which, from the vantage point of the inhabitants, could have meant a tower of observation and surveillance. That function has been attributed to the minaret of the Juyushi mosque of 1085 in Cairo, and it was perhaps more common than we might imagine.8 Nearly midway between the mosque and the palace stood the small but highly significant Maqam Ibrahim (Figure 3.7). First built by NËr al-DÈn and his son Ismaʿil, this shrine underwent some restorations and additions during the reigns of Ghazi and his son Muhammad.9 As the patron saint of Aleppo 10 and the traditionalist figure par excellence, Ibrahim was venerated by the Zangids and the Ayyubids in order to enhance their authority and legitimacy by deepening their roots in Aleppo and to reinforce their image as the safeguards of SunnÈ Islam. Another even more important shrine for Ibrahim stood about one kilometre south of the city, and it also was the object of some attention by the Ayyubids.11 Another component of the citadel’s context actually lay outside its confines in the region commonly known as Taht al-Qalʿa. Under al­ -Zahir Ghazi, a systematic attempt was begun to convert this region into an extension of the citadel, or better into a zone of official exchange between the Ayyubids and the city. Two major institutions contributed to the official image of this region, which is located just east of the citadel’s entrance block or directly opposite the palace. These are the still extant intramural funerary madrasa of al-Zahir (al-Sultaniyya) and the dår al-ʿadl (court of grievances or tribunal), which has disappeared almost without a trace. The Sultaniyya Madrasa was founded for both Hanafis and Shafiʿis by al-Zahir Ghazi, who died in 1216 before completing it (Figure 3.8).12 It remained in its unfinished state for five years, at which time the atåbek Tughril completed it in 1221. He was most likely responsible for the two rows of student cells that extend between the portal and the tripartite prayer hall. Ghazi’s mausoleum, which was ready to receive him at the time of his death, stands at the south-eastern comer of the building. It projects slightly into the urban space and has windows with inscriptions on three sides (Figure 3.9). With its minaret – topped portal facing the citadel and Ghazi’s mausoleum originally facing the dår al-ʿadl – this two-madhhab madrasa reaffirmed the traditionalist orientation of the Ayyubids and reinforced

circles of power

Figure 3.1  Aleppo, Citadel entrance, from Madrasa al-Sultaniyya (photo: Yasser Tabbaa)

their links with the ʿulama. The location of Ghazi’s mausoleum confirms the existence of a street next to it, since urban mausoleums in Syria and Egypt nearly always had street frontage.13 This street would have run parallel to the old wall of the city, passing through the southern gate of the Bab al­Maqam, and ending at the Maqam Ibrahim and the nearby Ayyubid madrasas. Where the dår al-ʿadl in Aleppo was has been the subject of some debate. Some writers place it to the west of the citadel’s entrance block14 and others to the east of it.15 It suffices here to say that both archaeological evidence and a number of statements by Ibn Shaddad leave no doubt that it was located to the east of the entrance block, underneath the present municipality building, between the old and the new eastern wall of the city. The wall seems to have been shifted slightly to the west for no other reason than to create a defensible space for this important institution.16 It was also provided with a wall on its southern side; its northern side faced the moat of the citadel. These walls were pierced by four gates: the Bab al-Jabal, the secret gate connecting the dår al-ʿadl to the palace; the Bab Dar al-ʿAdl on the south, for the procession of al-Zahir Ghazi; and the two gates called Bab al-Saghir near the edge of the moat, one in the old wall and one in the new (Figure 3.10).17 The dår al-ʿadl communicated directly with the palace through a subterranean passage. The passage itself has largely disappeared, but its location may be determined by following in a southerly direction the vaulted corridor that currently separates the bath from the arsenal (Figure 3.17). After making a 90-degree turn to the right, this corridor ascends to the level of the lower ramparts, where it would

49

50

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

originally have connected with a vaulted passage that led directly to an arched opening that still remains in the curtain wall (Figure 3.11). Peering through this opening, we glimpse the modem municipality building directly ahead of us and leading in its direction, remnants of stairs, at the end of which are ruins of a vaulted tunnel. With a little reconstruction, we can postulate that the stairs leaving the palace area would have led directly to a vaulted passage which would have gone through an opening in the wall of the citadel and descended to the dår al-ʿadl.18 The palace, the symbol of absolute authority, therefore literally presided over the dår al-ʿadl and was directly linked with it through a secret passage. Underground passages, known to have linked various parts of ʿAbbasid palaces in Baghdad,19 are rather commonplace in Islamic citadels. Most often they are used for escape in time of danger or to reach provisions, but the link between palace and tribunal is rather unusual and seems to have more serious implications. On the surface it simply facilitated the sultan’s movements between palace and tribunal during his routine audience on Monday and Thursday, visits that were intended to foster the essential link between authority (mulk) and justice (ʿadl). In his important treatise, the Siyåsatnåma, Nizam al-­Mulk (d. 1092) expatiates upon the concept of justice, which he saw as the foundation of stately power and the primary attribute of correct government: It is absolutely necessary that on two days in the week the king should sit for the redress of wrongs, to extract recompense from the oppressor, to give justice and to listen to the words of his subjects with his own ears, without any intermediary. It is fitting that some written petitions should be submitted if they are comparatively important, and he should give a ruling on each one. For when the report spreads throughout the kingdom that on two days in the week The Master of the World summons complainants and petitioners before him and listens to their words, all oppressors will be afraid and curb their activities, and no one will dare to practice injustice or extortion for fear of punishment.20 A decade later Ibn al-Balkhi expressed a similar view: ‘Those possessed of learning have said: “When a king is adorned by religion and his rule is stable because of justice, kingship will not disappear from his house unless, God forbid, some disorder appears in religion or he commits tyranny.”’21 The dår al-ʿadl was therefore a direct outgrowth of formulations that were themselves based on earlier conceptions of Islamic rule. Despite its physical separation from the citadel and its relative importance under the Ayyubids and the early Mamluks, the dår al-ʿadl was by no means an autonomous institution subject to its own charter and body of laws. It was an appendage of royal power,

circles of power

and it received its ultimate authority neither from the city elders, as was the case in the communal courts of medieval Europe,22 nor from the chief justice (qå∂È, al-qu∂åt) but from the sultan, who alone embodied the state and was responsible for directing its affairs in accordance with the Shariʿa.23 The invisibility of the link between the palace and the tribunal has further covert implications, in that the sultan could presumably observe the proceedings at the dår al-ʿadl any time he wished without being seen and could even interrupt them if he heard something not to his liking. The secrecy of the ruler’s movements through this passage would make it more a vehicle for monitoring and surveillance than for observing with detached interest the progress of justice. This surveillance should be understood in terms of control and coercion, directed at curtailing the authority and independence of the judges.24 The street between the Madrasa al-Sultaniyya and the dår al-ʿadl originally proceeded southwards, where it left the city through the Bab al­Maqam and ended at the shrine of Abraham. Unlike all other medieval Syrian gates, which have a single bent entrance between two towers, the Bab al-Maqam is a straight tripartite gate without towers, a form that could not have served any defensive functions (Figure 3.12).25 Most likely it was built as a ceremonial gate that stood in the middle of the street connecting the secular and religious domains of the Ayyubids.26 The Ayyubid palace in the citadel was, then, linked to the city and the southern suburb on a number of levels: physically through the tunnel and the southern road; ceremonially through the dår al-ʿadl and the Bab al-Maqam; pietistically through the Abrahamic connection; and religiously through the double madrasas of al-Zahir Ghazi, one facing the citadel and the other near the Maqam Ibrahim. Precedents for the palace and its contemporaries The palace complex is located in the southern part of the citadel to the right of the ascending ramp that begins at the entrance block and ends at the mosque of al-Zahir above. Before reaching it, one has to pass through the colossal entrance block, one of the most impressive gates in the medieval world (Figure 3.13). Built in its entirety by Ghazi between 1189 and 1214,27 this entrance block has three monumental gates, two of which, the Gate of the Serpents and the Gate of the Two Lions, are interesting for their figural sculpture.28 The first gate has inscribed in its voussoir a relief sculpture in the form of two knotted and intertwined dragons, each with two heads (Figure 3.14). This is one of four roughly contemporary gates with intertwined dragons on them,29 suggesting that this protective emblem was thought in some way to contribute to the power, impregnability, or even good fortune of the structure. The

51

52

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 3.2  Aleppo, Citadel. Portal of Ayyubid Palace, c. 1200 (photo: Yasser Tabbaa)

fantastic and ferocious aspects of the dragons would have enhanced the imagery of power, while their mysterious knotting would have suggested a magical apotropaic symbolism.30 The third gate is flanked by a pair of recumbent lions which jut out like consoles from the masonry of the door jambs. They have schematic features typical of sculptural representations of lions in medieval Islamic art, such as those on the Fountain of the Lions at the Alhambra Palace. Lions, like dragons, were commonly used in medieval Islamic sculptural ornament, including several examples on military and palatial monuments in Anatolia and Aleppo (see Redford, figs 1, 8, 11).31 Indeed, their use in pairs as guardian figures can be traced back to the ancient Near East, one of the earliest instances being the two terracotta lions flanking the entrance to the early-secondmillennium palace at Tell Harmal.32 All this suggests that lions were associated with kingship and royal glory, which is quite appropriate in the case of the Aleppo citadel, given their proximity to the palace. The palace is entered from the west through a monumental portal fronted by a fairly spacious area that may have served a ceremonial function. Though traditional in form, the portal is striking in its richness and complexity, combining as it does striped masonry, joggled voussoirs and geometric ornament within the standard format of the muqarnas entry (Figures 3.2 and 3.15). It consists of a recessed door whose jambs and lintel are made of joggled voussoirs, which are all overlaid by a network of geometric ornament. A superb muqarnas vault of four tiers and a scalloped half­dome rises above the lintel and is crowned by a panel of black and white stone marquetry. Two

circles of power

r­ectangular windows topped by a braided motif are symmetrically placed on either side of the portal. Although related to the portals of contemporary religious structures, it differs from them in its higher degree of elaboration and complexity. In fact, the only other comparable portal known to me is that of the Ayyubid palace at Qalʿat Sahyun in the ʿAlawite mountains33 (Figure 3.16). In addition to their overall formal similarities, both portals contain an element of structural mystification, especially apparent in their joggled voussoirs. Although too small a sample for any definite conclusions, the occurrence of such suspended forms in other palatial contexts suggests that they may have had royal or authoritative associations. The palace comprises a rectangle, approximately 45 metres north– south and 40 metres east– west, which is bordered on its western and northern sides by access streets (Figure 3.17). Three fairly distinct functions are contained within this rectangle: a palace in the northern two-thirds; a bath in the south-eastern corner; and what appears to be guard rooms and arsenals in the southern third. The palace proper consists of two cruciform units of unequal size centred around squarish courtyards, the smaller being about 4.5 metres per side and the larger 9.5 metres. Much has been written about the problem of cruciform plans, particularly in reference to the madrasa, although the form has not been discussed within the context of palatial architecture. At the centre of the controversy regarding cruciform four-iwån plans is Creswell’s theory about the strict correspondence between the four

Figure 3.3  Aleppo, Citadel, Ayyubid Palace, main courtyard, early thirteenth century (photo: Yasser Tabbaa)

53

54

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

SunnÈ sects and the number of iwåns in Cairene madrasas.34 Critics of this hypothesis have stressed the Persian ancestry of this plan, seeing it originally in the vernacular architecture of eastern lran.35 But this hypothesis is equally problematic in at least two other respects. The first is that it rests on the faulty assumption that vernacular architecture is changeless, so that the existence of four-iwån village houses in the late nineteenth century suffices as a proof for their existence in the tenth century. The second is that the hypothesis is based on a transmission from vernacular to monumental, the opposite of the far more likely process of vernacular architecture appropriating the basic design idea of contemporary monumental architecture, perhaps that of an important palace.36 What is missing in all these four-iwån hypotheses is any mention of palatial architecture as a possible prototype. This omission is especially striking when we consider the relatively large number of extant palaces from the pre-Islamic and early Islamic periods and the fairly continuous history of four iwåns and other quadripartite forms in their plans. Such a quadripartite division is first seen in the Parthian palace at Assur (first to third centuries ad)37 The type becomes especially common in early Islamic palaces, including the seventh-century dår al-imåra at Kufa, the Umayyad palace at the Amman citadel,38 the palace of Abu Muslim at Merv from 750–5,39 and the so-called Victory Monument of Harun al-­Rashid (786–809) at Hiraqlah 40 (Figure 3.18). The last two examples represent a variant of cruciform planning, in which four iwåns converge on a square chamber that is usually covered by a dome. Both variants of cruciform design are found at Samarra (833–92), whose palaces display a bewildering array of innovations and original combinations. In all five primary palaces – Jawsaq al­Khaqani, Qasr al-Jiss, Qasr al-Haruni, Qasr Balkuwara (Figure 3.19), and Qasr al-ʿAshiq – a cruciform design occupies the centre formed by a domed throne hall with four axial iwåns.41 It should also be added that, whereas none of these palaces has the traditional four-iwån plan with an open court, the plan nevertheless does exist in the official architecture of the period, as is demonstrated by a recently excavated example, the so-called resthouse of Caliph alMutawakkil. That building was located adjacent to the mi˙råb of the Abu Dulaf mosque and consisted of two courtyards, each with four iwåns.42 It is clear from this brief look at early Islamic palatial architecture that cruciform plans played a dominant role in their design, occupying the most privileged part of the palace and imparting a sense of order and monumentality to the courtyard form. What functional or symbolic values lie behind this formal order? And how did these ordered forms in turn contribute to the iconography of power? The quadripartite division as a symbol of universal power is deeply rooted in the ancient Near East, where it can be traced back to the time of

circles of power

the Akkadian sovereign Naramsin (c. 2250 BC), who called himself the ‘King of the Four Quarters of the World’.43 Although no medieval Muslim caliph or sovereign is known to have borne this exact title,44 the underlying concept of universal rule is quite explicit in the form of the Round City of Baghdad,45 and somewhat less so in the ceremonial practices of the ʿAbbasid caliphs and their successors and emulators, the Ghaznavid kings. Cursory references in Hilal al-Sabi’46 and somewhat more detailed ones in Baihaqi47 paint a picture of a formalised court ceremonial in which the caliph is described as surrounded on the four directions by contingents of honour guards (ghulåms). Within the fourfold composition of the throne hall, contingents of ghulåms, whose distinctive clothing and headdress made specific references to different parts of the Islamic world, defended the right, left, front and rear of the sovereign.48 It seems fairly certain, therefore, that the cruciform divisions of the courtyards and throne halls at Lashkar-i Bazar and somewhat earlier at Samarra played an important role in organising the form of the ceremonial and in codifying the roles of its participants. Conversely, the complex ceremony and its highly conventional character would seem to recharge the meaning of an ancient symbol, converting it from a static symbol of royalty to a dynamic attribute of authority. Given the royal and authoritarian associations of the four-iwån plan, its predominant use in medieval Islamic palaces comes as no surprise. In fact, it is used in eight other palaces that I know of, ranging in date from c. 1170 to 1260.49 These are: the so-called Qasr al-Banat in Raqqa (Figure 3.20), possibly built by NËr al-DÈn;50 the early thirteenth-­century ʿAjami palace in Aleppo (Figure 3.21); the Ayyubid palaces at Qalʿat Sahyun51 and Qalʿat Najm near Raqqa; the so-called saråy in the Basra citadel; the Ayyubid palace at the Kerak citadel; the late Ayyubid palace on Roda Island;52 and the Artuqid palace in the Diyarbakir citadel.53 Of course, the essential difference between these palaces and their early Islamic predecessors is their reduced scale, a question to which I will return. The northern façade of the Aleppo palace consists of an iwån flanked by narrow doors surmounted by arched windows, a composition repeated by the eastern and western façades (Figure 3.22). This tripartite composition is nearly standard in this period; it appears in all the eight medieval palaces. The same composition is also used in pious buildings, including madrasas and bÈmåriståns beginning with their earliest examples in Syria: the madrasa of Gümüshtekin in Bosra (1136) and the Bimaristan al-Nuri in Aleppo, datable to c. 1150. Tripartite courtyard façades have a long history in Middle Eastern architecture; the composition is first seen in Parthian temples and palaces, such as the Parthian palace at Assur.54 Possibly originating in Roman gates and triumphal arches, this composition is

55

56

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

c­ ombined in Parthian and Sasanian architecture with the iwån form. Subsequently, tripartite iwån compositions are documented at a number of early Islamic palaces, including those at Kufa, Ukhaidir, Samarra and Lashkar-i Bazar, which is known to have been modelled after Samarran prototypes. In view of the pivotal importance of Samarran and east Persian palaces for the overall development of the medieval Islamic palace, one may propose that the tripartite façade of the latter owes its origin to classical Islamic prototypes. But a closer and somewhat more concrete prototype is provided by the early Christian churches of northern Syria, which contain several examples of this composition in their façades and, even more closely, in some of their interior divisions.55 The medieval specimens closely resemble their late-antique predecessors in the quality of their stonework and the use of lintels, but differ from them in the use of the iwån for the central space. It would therefore seem that the tripartite-façade composition as known in medieval Syria and the JazÈra was ultimately based on the fusion of the eastern tripartite-iwån composition with the local architectural tradition of northern Syria. To what extent can we attribute any specific meaning to this highly ubiquitous form? Although one might reasonably argue that the Parthian palace at Assur still retained some of the specific meanings associated with the triumphal arch, the medieval examples are too chronologically and culturally removed to allow for any such associations. At this point the most we can say is that by the middle of the twelfth century this form had become one of the necessary elements of a royal palace or aristocratic residence, perhaps even an important component of palatial iconography. Also part of this aristocratic iconography may have been the elaboration of the voussoirs of the arch framing the main iwån, as we see in a single outstanding example in Aleppo, about 200 metres west of the citadel. This site is commonly called the ma†bakh (kitchen) of al-ʿAjami, but there is no doubt that it was originally a palace, perhaps belonging to the Banu al-ʿAjami, a noble family of Aleppo.56 The voussoir of the arch has here been transformed into two staggered rows of arrow-shaped pendants, creating a powerful and luxurious effect. Although there is no supporting archaeological evidence, it seems reasonable to assume that the arch of the northern iwån of the palace was elaborated in like manner. The northern iwån is also distinguished from all the others by having in its rear wall a muqarnas vault beneath which once flowed a shådirwån fountain (Figure 3.22). Only traces of this fountain exist, but a passage in a later medieval text describes how the water flowed down a marble­inlaid shådirwån, into a channel beneath the iwån, and finally emerged as a water jet in the middle of a pool.57 This type of fountain was quite common in medieval Islamic palaces from Iran to North Africa.58 In addition to its association with images of para-

circles of power

Figure 3.4  Aleppo, Matbakh al-ʿAjami, iwån, early thirteenth century (photo: Yasser Tabbaa)

dise and fertility, the elaborate use of water in palatial courtyards must also have reflected the power and sophistication of the patron. In the specific case of Ghazi, this fountain also celebrates his outstanding efforts at expanding the underground water network in the city and within the citadel, where he was responsible for enlarging a vast underground cistern.59 It becomes clear from this discussion of five components of the medieval palace – sculptural ornament, muqarnas portal, four-iwån plan, tripartite courtyard façade and shådirwån fountain – that they were all part of the repertory of significant forms in medieval Islamic palaces and that, with the exception of the muqarnas portal, they all echoed well-established formal types in early Islamic palaces.60 Indeed, the correspondences between the two kinds of palaces are at times so striking that they have almost led us to forget about the crucial difference of scale. This essential difference can best be understood by comparing in table form the courtyard dimensions of some early and medieval Islamic palaces (Table 1).

57

58

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Table 1  Courtyard dimensions of some early and medieval Islamic palaces name

Medieval Palaces court dimension

Banat Sahyun Aleppo Diyarbalir ʿAjami Najm Bosra Roda

9.95 x 9.15 m 6.65 x 6.73 9.72 x 9.49 3.50 x 3.80 9.10 x 9.90 7.05 x 7.05 8.05 x 6.38 6.50 x 5.50

Average linear dimension average area

7.57 x 7.25 54.88 sq. m

name

Early Palaces court dimension

Kufa Ukhaidir Jawsaq Balkuwara al-ʿAshiq Lashkari So. Palace Ghazna Palace Musʿud III Average linear dimension Average area

37 x 37 m 32.70 x 29 110 x 62 105 x 67 40 x 23 60.5 x 47.5 50 x 23 62.17 x 42.50 2642.23 sq. m

The average courtyard dimensions of a medieval palace come to approximately 7.57 x 7.25 metres; those of an early Islamic palace come to a gigantic 62.17 x 42.50 metres. Dividing the corresponding average linear dimensions of the two groups, we get an average linear ratio of approximately 1 to 8 (8.21) for the longer (usually north– south) dimension and 1 to 6 (5.85) for the shorter one. What this in effect means is that the surface area of the courtyards in early Islamic palaces is on the average 48 times that of medieval ones, a striking disparity that clearly underlines the basic differences between the two groups of palaces and hence between the two periods in which they were built. This difference in scale is so startling that it should lead us to re-evaluate all the implied similarities and the overall relationship of the two classes of palaces to each other. Clearly, what we have is not a relationship of equals, but rather one of petty dynasties looking at a distant golden age – ninth-century Samarra or tenth-century Baghdad – which they are unable to supersede and yet unwilling to ignore.61 The palaces they built were not just miniature versions of classical Islamic palaces, but essentially constituted a new palace type, some of whose most important forms and images were derived from the past. The originality of these palaces stems not so much from their smaller size as from the fact that this down-scaling in itself seems to have led to the sharpening of the appropriated images and the enhancing of their iconography. This is easily observable in the precision of the plans, the complexity of portals, the dramatic use of water, and the employment of a variety of architectural devices, such as muqarnas and pendant arches, whose desired effect was to induce wonder and amazement in the viewer.

circles of power

I have attempted in this piece to place the Ayyubid palace in Aleppo within a complex matrix of relationships that included its urban context, its typological parentage, the symbolic associations of some of its forms, and its links with contemporary and classical Islamic palaces. What distinguished this palace from all its cognates was not so much its formal qualities as its contextual relations, both inside and outside the citadel. Its overall significance derived equally from its formal and typological associations and from its multiple links with Aleppo, its history, its patricians and its general population. Viewed from this multifaceted perspective, the citadel appears less impregnable and isolated and more firmly linked with the city. Through overt and subtle means, it broadcasts political and religious messages that were heard throughout the town. Conversely, the soft and luxurious image of the palace and even its smallish size are partly modified by the apotropaic and authoritative symbolism of its gates and the royal associations of its plan and forms. To some extent, however, the contradictions between the image and intent of citadel and palace remain unresolved. And this may be seen as a reflection of the special situation of the Ayyubid dynasty of Aleppo, an alien dynasty that, perhaps more than any other, attempted to integrate its authority with that of the city.62 Painfully aware of their arriviste status and of the limitations of brute force, the Ayyubids of Aleppo seem to have left no stone unturned in their attempts to establish the foundations of their rule while effectively and justly dealing with an urban population. The measure of power and dominion that they attained was not simply handed to them through their noble lineage and their links with the ʿAbbasid court,

Figure 3.5  Aleppo, Citadel. Distant view from south-west (photo: Yasser Tabbaa)

59

60

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

but was constructed through historical, social, pietistic and symbolic means. The resiliency of this construction ensured a century of prosperity and relative stability for Aleppo, but its inherent weakness could not withstand the onslaught of the Mongols and the more centralised rule of the Mamluks.

Figure 3.6  Aleppo, Citadel. Mosque, 1214, from south (photo: Yasser Tabbaa)

Figure 3.7  Aleppo, Citadel. Maqam Ibrahim, 1168, from north (photo: Yasser Tabbaa)

circles of power

Figure 3.8  Aleppo, Madrasa al-Sultaniyya, completed 1221. Plan. After Lauffray, fig. 1

Figure 3.9  Aleppo, Madrasa al-Sultaniyya. Mausoleum of al-Zahir Ghazi with entrance block in background (photo: Yasser Tabbaa)

61

62

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 3.10  Aleppo, Citadel and Southern Quarter

circles of power

Figure 3.11  Aleppo, Citadel. Opening to tunnel between palace and dår al-ʿadl (photo: Yasser Tabbaa)

Figure 3.12  Aleppo, Bab al-Maqam. Ayyubid with late Mamluk repairs (photo: Yasser Tabbaa)

63

64

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 3.13  Aleppo, Citadel. Entrance block, thirteenth–fifteenth centuries (photo: Yasser Tabbaa)

Figure 3.14  Aleppo, Citadel. Gate of Serpents, c. 1195 and later (photo: Yasser Tabbaa)

65

circles of power

Figure 3.15  Aleppo, Citadel. The Ayyubid palace, c. 1190–1230. Portal (photo: Yasser Tabbaa)

Figure 3.16  Qalʿat Sahyun. Ayyubid palace, c. 1200–30 (photo: Yasser Tabbaa)

66

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 3.17  Aleppo, Citadel. The Ayyubid palace, 1190–1230. Plan. Adapted from a plan of the Directorate General of Antiquities, Damascus

circles of power

Figure 3.18  Hiraqlah (Raqqa), ‘Victory Monument’ of Harun al-Rashid (786–809). Plan. From Toueir, ‘Hiraqlah’, fig. on p. 111

Figure 3.19  Samarra, Balkuwara Palace, 854–9. Plan of central unit. From Creswell, Early Muslim Architecture, vol. 2, fig. 214

67

68

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 3.20  Raqqa, Qasr al-Banat, c. 1168. Plan. After Toueir, ‘Qaßr al Banåt’, fig. 2

Figure 3.21  Aleppo, Matbakh al-ʿAjami, first half of thirteenth century. Plan and section, present condition. Drawing: courtesy of the Directorate General of Antiquities, Aleppo

circles of power

Figure 3.22  Aleppo, Citadel. Ayyubid palace, central courtyard (photo: Yasser Tabbaa)

Notes 1. Cited in ʿIzz al-DÈn ibn Shaddåd, al-aʿlåq al-kha†Èrah fi dhikr umarå al-shåm wa’l-jazÈra, vol. 1, pt 1, Ta’rÈkh Óalab, ed. Dominique Sourdel (Beirut: Institut français de Damas, 1953), 170 (hereafter Ibn Shaddåd, Óalab); my translation. 2. No serious study exists on the Ayyubid dynasty of Aleppo. Meanwhile, see Claude Cahen, ‘AyyËbids’, EI2, 1: 796–807; and R. Stephen Humphreys, From Saladin to the Mongols: The Ayyubids of Damascus, 1193–1260 (Albany, NY: State University of New York Press, 1977), esp. 98–123, passim. 3. The need to provide nuances for this too rigid formulation has already been suggested by urban historians. According to Ira M. Lapidus, Muslim Cities in the Later Middle Ages (Cambridge, MA: Harvard University Press, 1967), 78: ‘Regime and society did not confront each other, reacting only on the interface between them; rather they permeated each other, the stronger pressing its way through the structure of the latter.’ 4. See, for example, S. A. El-Ali, ‘The Foundation of Baghdad’, in The Islamic City, ed. S. M. Stern and Albert Hourani (Oxford: Bruno Cassirer, 1970), 87–102. 5. J. M. Rogers, ‘Samarra: a Study in Medieval Town Planning’, in The Islamic City, 119–56, has brilliantly discussed the consequences of the separation of caliphal palaces from the city. For a recent discussion of the three stages in Islamic urbanism, see Jere L. Bacharach, ‘Administrative Complexes, Palaces, Citadels: Changes in the Loci of Muslim Rule’, in The Ottoman City and Its Parts, ed. I. A. Bierman, R. Abou-El-Haj and D. Preziosi (New Rochelle, NY, 1991), 111–28. 6. This is quite explicitly stated by Ibn Shaddåd (Óalab, 29) who adds that the Mirdasids (1023–79) established a practice (sunna) that was followed by later princes of the city.

69

70

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

7. For the inscriptions of this sovereign, see Gaston Wiet, ‘Une inscription de Malik: Ûåhir GhåzÈ à Latakieh’, Bulletin de l’Institut français d’archéologie orientale de Caire 30 (1931): 273–92. 8. Farid ShåfiʿÈ, ‘The Mashhad al-JuyËshÈ (Archaeological Notes and Studies)’, in Studies Islamic Art and Archituture in Honour of Professor K. A. C. Creswell (Cairo: American University in Cairo Press, 1965), 251 and 246–52, passim. ShåfiʿÈ’s view has been partly refuted by Jonathan Bloom in Minaret, Symbol of Islam, Oxford Studies in Islamic Art 7 (Oxford: Oxford University Press, 1989), 140–2, who offers a pious interpretation for the mosque itself and a symbolic interpretation for the minaret. Bloom’s refutation of Shafi‘i rests largely on the reference to the building as ‘mashhad’ and on the choice of Qurʾånic inscriptions in it, but generally ignores its highly unusual form and placement. For these reasons, I find ShåfʿiÈ’s interpretation more convincing. 9. Ernst Herzfeld, Materiaux pour·un Corpus Inscriptionum Arabicarum, deuxième partie: Syrie du Nord. Inscriptions et Monuments d’Alep, vol. 1, pt. 1 (Cairo, 1955), 125–31, has suggested that the marble revetment and all the carved wood panelling except the mi˙råb belong to the periods of al-Zahir Ghazi and his son al-ʿAziz Muhammad. 10. For Ibrahim’s links with Aleppo, see Herzfeld, Alep, 1: 3–5. According to a widely accepted legend, the very name of the city, Halab, was related to Ibrahim’s milking of his goats (˙alaba IbråhÈm) in the vicinity of the citadel; see, for example, Ibn Shaddåd, Óalab, 15. 11. Herzfeld, Alep, 1: 180. 12. This would make it the only extant two-madhhab madrasa in Aleppo, a type that was in any case quite rare in Syria. Hanafis and Shafi‘is were by far the two most important SunnÈ sects in Aleppo, a fact that is clearly reflected in the large number of madrasas (22 for Hanafis and 21 for Shafiʿis) that were built for them during the Ayyubid period and the negligible number dedicated to the other two sects. See the statistical tables in K. A. C. Creswell, ‘The Origin of the Cruciform Plan of Cairene Madrasas’, Bulletin de l’Institut français d’archéologie orientale 21 (1922): 1–54; updated in Creswell, Muslim Architecture of Egypt, 2: 122–3. See also J. Lauffray, ‘Une madrasa ayyoubide de la Syrie du Nord. La Soultaniya d’Alep’, Les Annales archéowgiques Arabes syriennes 3 (1953): 49–66. 13. The earliest known occurrence of such a mausoleum is in the funerary madrasa of NËr al-DÈn in Damascus (1172), which may have set a precedent for later Ayyubid and Mamluk funerary madrasas. 14. M. Rågheb al-Tabbåkh, Iʿlåm al-nubalåʿ, 2nd edn (Aleppo: Dår alQalam, 1988), 3: 13–14. In Alep, pl. LVIII, Sauvaget clearly placed the ‘palais de justice’ west of the entrance block. This is curious because in an earlier and little-known publication entitled ‘L’enceinte primitif de la Ville d’Alep’, Mélanges de l’Institut français de Damas 1 (1929): 148–9, he quite accurately places the dår al-ʿadl east of the entrance block, within the gates built for it by al-Zahir Ghazi. Sauvaget’s error, however, was copied by Dominique Sourdel, ‘Esquisse topographique d’Alep intramuros à l’époque ayyoubide’, Annales archéologiques Arabes syriennes 2 (1952): fig. 2. 15. Herzfeld, Alep, 1: 15–16 and fig. 5. 16. This may sound strange, but it seems to be the only likely explanation for moving a short section of the eastern wall of the city a distance of no more than 40 metres to the east. See Herzfeld, Alep, 1: 15–16. 17. These details are culled from Ibn Shaddåd, Óalab, 17, 21 and 25.

circles of power

18. Contrary to what one would have expected, this opening is located above instead of below the glacis of the citadel, suggesting that the vault of the connecting tunnel would have been at least partly visible above the glacis. This is indeed a problematic and unsatisfactory solution, but it seems to be dictated by the archaeology of the site. 19. A. A. Duri, ‘Baghdad’, EI2, 1: 898. For the topography of medieval Baghdad, see also A. A. al-Duri, ‘Society and Economy of Iraq under the Seljuqs (1055–1160 ad) with Special Reference to Baghdad’, Ph.D. diss., University of Pennsylvania, 1970, 44–94. 20. The Book of Government or Rules for Kings, The Siyåsatnåma or Siyar al-Mulk of NiΩåm at-Mulk, trans. Hubert Drake (New Haven: Yale University Press, 1960), 14. 21. A. K. S. Lambton, ‘The Internal Structure of the Saljuq Empire’, in Cambridge History of Iran, vol. 5, The Saljuq and Mongol Empires, ed. J. A. Boyle (Cambridge, Eng.: Cambridge University Press, 1968), 205, quoting the Fars-Nåma, ed. Guy Le Strange and R. A. Nicholson, Gibb Memorial Series, n.s., vol. 1 (London, 1921), 34. 22. Georges Duby (ed.), A History of Private Life, vol. 2, Revelations of the Medieval World (Cambridge, MA: Harvard University Press, 1988), 300–2. 23. Although courts of grievances (maΩålim) are known to have existed since the ʿAbbasid period, if not before, the Arabic sources are unanimous in attributing the institution of the dår al-ʿadl to Nur al­DÈn, who built the first one in Damascus soon after 1154. See Nikita Elisséeff, NËr al-Din, 3 vols (Damascus: lnstitut français de Damas, 1967), 3: 843–7, where the author also discusses the specific allusions to justice in the titulature of NËr al-DÈn, al­ malik al-ʿådil. NËr al-DÈn’s motives for the creation of the dår al-ʿadl are discussed in AbË Shåma, Kitåb al-Raw∂atayn, vol. 1, pt 1, ed. M. Hilmi Ahmad (Cairo, 1956), 17–18. See also Yasser al-Tabbaa, ‘The Architectural Patronage of NËr al-DÈn, 1146–1174’, Ph.D. diss., New York University, 1982, 219–25. The institution seems also to have existed in Mosul during the Zengid dynasty; see Rashed al-­ Jumayly, Dawlat al-atåbikah fi’l-Mawßil baʿda Imåd al-DÈn ZankÈ (Beirut, 1970), 247–8. 24. A related, though not identical, clandestine means of surveillance is reported for the Madrasa al-­Mustansiriyya in Baghdad. There, it took the form of an elevated window (shubbåk) that directly overlooked the lecture hall below, allowing the caliph to eavesdrop without being seen. See Hussein Amin, al-madrasa al-Mustanßiriyya (Baghdad, 1960), 142. 25. Straight tripartite gates were not built during the Islamic Middle Ages, and even ones remaining from the Roman period (e.g. the Bab Sharqi in Damascus) were rebuilt with a single bent entrance (see Herzfeld, Alep, 1: 68–9). 26. This gate seems to have enjoyed a ceremonial role until almost the very end of the Mamluk period, as we can deduce from the two round shields added to it by Barsbay between 1428 and 1437. I was not able to find any textual descriptions for this ceremonial. 27. These dates can be surmised from the information provided by Ibn Shaddåd, Óalab, 17 and 24–5. 28. For the chronology of this gate complex, see Herzfeld, Alep, 1: 87–91. Meanwhile, it suffices to note that the entire superstructure above the two Ayyubid towers belongs to the late Mamluk period, when it served

71

72

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

as the palace of the Mamluk governors of Aleppo. This palace, which still awaits study, contained a very prominent and lavishly decorated shubbåk that may have served a ceremonial function. It seems fairly certain that the original Ayyubid entrance block did not contain such a window for appearances. 29. These are (1) the now-destroyed Talisman Gate in Baghdad, which depicts a seated figure (presumably the caliph al-Nasir) grabbing in each hand the tongue of a rather ferocious dragon whose serpentine body fills the rest of the archivolt (F. Sarre and E. Herzfeld, Archäologische Reise im Euphrat- und Tigris Gebiet, 4 vols [Berlin: D. Reimer, 1920], 2: 153–6); (2) the gate of a khan in Sinjar from the reign of Badr al-Din Luʾluʾ (1219–59) depicting on each half of the archivolt a nimbed and bearded man attacking a knotted dragon with a spear (Sarre and Herzfeld, Archäologische Reise, 1: 13–15); (3) a fragmentary example that once decorated a portal in the Konya citadel, showing a knotted double-headed dragon with a ferociously open mouth at each end of its scaly body: discussed by Gönül Öney, Anadolu Selçuklu Mimarisinde Süsleme ve el Sanatlari (Ankara: Türk Matbaacilik Sanayii, 1978), 46, fig. 32; (4) another fragmentary example in a reused state in the western wall of the citadel of Damascus. 30. Cf. Herzfeld, Alep, 1: 259–60, where he suggests a similar symbolism for any of the many knotted forms common in this period. Willy Hartner in ‘The Pseudoplanetary Nodes of the Moon’s Orbit in Hindu and Islamic Iconographies’, Ars Islamica 5 (1938): 113–54, has argued for an astrological interpretation of the dragon in Islamic art as the eclipse monster, but his interpretation seems to work best in a very restricted context. 31. See, for example, Herzfeld, Alep, pls VIII and IX. 32. Pierre Amiet, Art of the Ancient Near East (New York: Harry Abrams, 1980), fig. 446. 33. The only study so far published on this curious palace (wrongly called a hammam locally) is Maurice Ecochard, ‘Notes d’archéologie musulmane, 1: Stéréotomie de deux portails du Xlle siècle’, Bulletin des études orientales 7–8 (1937–8): 98–112. Although undated, this palace is undoubtedly Ayyubid and possibly contemporary with that of the Aleppo citadel. Sahyun (the Crusader name is Soane) was taken in 1186 by Saladin who gave it as iq†åʿ to the amir Nasir al-Din Mankuras ibn Khumartekin and his descendants, who ruled it until 1272. 34. Creswell, ‘The Origin of the Cruciform Plan’, 104–33. 35. These views are summarised in André Godard, ‘L’origine de la madrasah, de la mosquée et du caravansérail à quatre iwans’, Ars lslamica 15–16 (1951): 1–9, and, more recently, in Robert Hillenbrand, ‘Madrasa’, EI2, 5: esp. 1136–40. 36. The only author to have questioned the validity of the vernacular connection has been Janine Sourdel-Thomine in her two excellent articles: ‘La mosquée et la madrasa’, Cahiers du monde mediévale, 1970: 114–15; and ‘Renouvellement et tradition dans l’architecture saljËqide’, in Islamic Civilisation, 950–1150, ed. D. S. Richards (Oxford: Bruno Cassirer, 1973), 256–62. 37. Malcolm A. R. Colledge, Parthian Art (Ithaca, NY: Cornell University Press, 1977), fig. 21. 38. The new plan by Alastair Northedge is illustrated in: W. Allan, ‘New Additions to the New Edition’, Muqarnas 8 (1991), fig. 4.

circles of power

39. See Creswell’s reconstruction in Early Muslim Architecture, vol. 2, fig. 1. 40. Kassem Toueir, ‘Natå’ij al-tanqÈb fi Hiraqlah’, in Les Annales archéologiques Arabes syriennes 33, 1 (1983): 99–112. 41. These plans are illustrated in Alastair Northedge, ‘Creswell, Herzfeld and Samarra’, Muqarnas 8 (1991): 74–93. 42. Northedge, ‘Creswell, Herzfeld and Samarra’, fig. 10. 43. Henri Frankfort, The Art and Architecture of the Ancient Orient (Harmondsworth, Eng.: Penguin Books, 1970), 84f. Although no work of art from the time of Naramsin represents this conception of kingship in an emblematic fashion, Frankfort suggests that it was nevertheless ‘expressed’ in some of his works. Indeed, the closest that we seem to come to a symbolic representation of this concept might be in the depiction of Assyrian camps as a circle or an oval divided into four quadrants. See Amiet, Art of the Ancient Near East, fig. 599. 44. Somewhat related Muslim titles are sul†ån al­ar∂, thåt al-†Ël wa’lʿar∂ (‘sultan of the long and wide earth’) and malik umarå’ al-mashriq wa’l-maghreb (‘king of the princes of the east and the west’); cited in Herzfeld, Alep, 1: 112–13. Furthermore, the concept of the ‘master of the world’ occurs rather frequently in the Siyåsatnåma, esp. 9–13. 45. Oleg Grabar, The Formation of Islamic Art (New Haven: Yale University Press, 1973), 69–70. H. P. L’Orange, Studies on the Iconography of Cosmic Kingship in the Ancient World (Oslo: H. Aschehoug & Co., 1953), 12–15. See also, Charles Wendell, ‘Baghdåd: Imago Mundi and Other Foundation­Lore’, International Journal of Middle Eastern Studies 2 (1971): 99–128. 46. RusËm Dår al-Khilåfa, ed. Mikhåʾil ʿAwwåd (Baghdad, 1964), 91. 47. Abuʾl-Fazl BayhaqÈ, Ta’rÈkh-’È MasʿËdÈ, ed. Q. GhanÈ and A. A. Fayyå∂ (Tehran, 1945), 539–41. These passages are quoted at length by C. E. Bosworth, The Ghaznavids, 2nd edn (Beirut: Librarie du Liban, 1973), 135–8. Bosworth also concludes, p. 104, that ‘the literary descriptions . . . of these levées have in recent years received striking confirmation from the researches of the French Archaeological delegation in Afghanistan in 1949–51 at the Ghaznavid palace of Lashkar-i Båzår at Bust’. 48. BaihaqÈ, as translated by Bosworth, Ghaznavids, 136, states: ‘All around the hall, standing against the panels, were the household ghulåms (ghulåmån­i khåßßagÈ) with robes of Saqllatun, BaghdådÈ and IßfåhånÈ cloth, two-pointed caps, gold-mounted waist sashes, pendents and golden maces in their hands. On the dais itself, to both left and right of the throne, were ten ghulåms, with four-sectioned caps on their heads, heavy, bejewelled waist sashes and bejewelled sword belts. In the middle of the hall [serai] were two lines of ghulåms; one line was standing against the wall, wearing four-sectioned caps. In their hands they held arrows and swords, and they had quivers and bow-cases. There was another line, possibly down the centre of the hall, with twopointed caps, heavy, silver-mounted waist sashes, pendents and silver maces in their hands. The ghulåms of both these lines all wore cloaks of Shushtari brocade.’ This ceremony was immediately followed by a feast for all soldiers and gentry and terminated by a smaller majlis that lasted until the evening prayer. 49. The one discrepancy in this consistent typological sequence appears most unexpectedly in the Damascus citadel, whose palace – if it is indeed that – has been identified by Sauvaget as a series of ­interconnected

73

74

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

two-storeyed chambers in the south­western comer of the enclosure. See Jean Sauvaget, ‘La citadelle de Damas’, Syria, 1930: 217–18 and figs 21 and 22. 50. Kassem Toueir, ‘Der Qaßr al-Banåt in ar-Raqqa. Ausgrabungen, Rekonstruktion und Wiederau bau (1977–1982)’, Damaszener Mitteilungen 2 (1985): 297–319; and Robert Hillenbrand, ‘Eastern Islamic Influences in Syria: Raqqa and Qalʿat Jaʿbar in the Later 12th Century’, in The Art of Syria and the JazÈra, 1100–1250, ed. Julian Raby (Oxford: Oxford University Press, 1985), 21–48. 51. Ecochard, ‘Stéréotomie de deux portails du Xlle siècle’, 98–112. 52. Creswell, Muslim Architecture of Egypt, vol. 2, fig. 38. 53. Oktay Aslanapa, ‘Erster Bericht über die Ausgrabungen des Palastes von Diyarbakir’, Istanbuler Mitteilungen 12 (1962): 115–28. 54. Roman Ghirshman, Persian Art, The Parthian and Sassanian Dynasties, 249 b.c.–a.d. 651 (New York: Golden Press, 1962), 32. At Hatra, another important Parthian site, the tripartite form is also used as a gate and in the exterior façades of temples; Ghirshman, Persian Art, 35–6. 55. Some of these are illustrated in Georges Tchalenko, Villages antiques de la Syrie du nord, 3 vols (Paris: Librarie Orientaliste Paul Geuthner, 1953). 56. No less than thirteen members of Banu al-ʿAjami are mentioned in Ibn Shaddåd, Óalab. They were builders of mosques, madrasas, khånqåhs, baths and palaces. They seem to have inhabited the privileged quarter between the Great Mosque and the citadel, where their two extant monuments – the Madrasa al-Sharafiyya and the ma†bakh of al­ʿAjami – are located. 57. Jean Sauvaget, Les tresors d’ors de Sib† ibn al-ʿAjamÈ (Beirut: Institut français de Damas, 1950), 165. 58. See my ‘Toward an Interpretation of the Use of Water in Islamic Courtyards and Courtyard Gardens’, Journal of Garden History 7, 3 (1987): 197–220. 59. See Ibn Shaddåd, Óalab, 144–50, for a detailed description of Ghazi’s expansion of Aleppo’s water network. 60. Although no trace remains of any painted decoration in the Ayyubid palace, it may once have existed. A short passage in Ibn Shaddåd, Óalab, 26, describes an earlier palace on the same site that was called dår alshukhËß (house of images) because of the many images decorating it. 61. The Ayyubids were tradition-bound in many other respects, including the adoption of classical poetic genres (see, for example, A. F. Hayb, al-Óarakah al-shiʿriyyah zaman al-ayyËbiyyÈn fi Óalab al-shahbå’ [Kuwait: Muʿalla, 1987], 167 ff.) and the use of proper names with historical associations (e.g. Muhammad, Abu Bakr, Yusuf and ʿIsa). 62. The acculturation of the Ayyubids to local Arab culture stands in some contrast to the exclusionary policies of other alien dynasties, including the Rum Seljuqs and especially the Mamluks. Alone among these dynasties the Ayyubids adopted Arabic names, conversed and wrote in Arabic, and a number of princes even became proficient in poetry and theology. On the contrary, there is very little likelihood that the Rum Seljuqs ever learned Kurdish or Armenian, the main languages in Anatolia until recently. While a few Mamluks were indeed given Arabic names, the majority were not. In any case, their well-known exclusionary policies stood firmly in the way of their integration into Egyptian society.

CHAPTER FOUR

D . ayfa Kha¯tu¯n, Regent Queen and Architectural Patron The study of women of high status in the Islamic world has generally focused on the modern and contemporary periods, where ample documentation and stark contrasts have generated complete portraits of these women as individuals and in terms of gender and class. Much less has been done to define the personalities of influential women in medieval Islam and to highlight the areas in which they chose to exercise their authority and patronage. The otherwise excellent recent studies by Mernissi, Ahmed and others have attempted to deal comprehensively with the social history of women in different periods but have generally failed to say anything new about medieval women.1 As is often the case with such diachronically ambitious projects, the middle Islamic period seems to get lost between the theoretical underpinnings of early Islam and the archival richness of later periods. At the very best, women of the middle period are assumed to have behaved and acted in a manner similar to women of the Umayyad and ʿAbbasid courts. More focused studies by Williams, Humphreys, Atil and others have contributed a much-needed perspective about the architectural and artistic patronage of women in the middle Islamic period.2 Despite these innovative works, however, the study of the patronage of women in medieval Islam remains very underdeveloped, particularly when compared to similar studies for the Ottoman Empire or medieval Europe. This state of research and, more specifically, the absence of any published work on Îayfa KhåtËn will help formulate the questions and shape the discourse for this piece. The questions I ask are both general and specific, initially addressing the status and patronage of some women of the Ayyubid court (1174–1260) before focusing on the regency and architectural patronage of Îayfa KhåtËn in Aleppo in the 1230s. I will look first at the general patronage of monuments by court women and then more carefully at Îayfa KhåtËn’s commissions built within the context of her powerful Yasser Tabbaa (2000), ‘Dayfa Khatun: Architectural Patron and Regent Queen’, in D. Fairchild Ruggles (ed.), Women, Patronage and SelfRepresentation in Islamic Societies, Albany: SUNY Press, 17–34.

76

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

but gendered position in a world of men. Although Îayfa achieved an unprecedented measure of autonomous political influence, becoming the first regent queen of her time, the piece will argue that her elevated status and prodigious architectural patronage were at least equally linked to the fact that she was wife, mother and grandmother of successive Ayyubid sultans. Furthermore, the paper will demonstrate that Îayfa KhåtËn, while operating within the Ayyubid system of architectural patronage, exhibited strong preference for foundations that catered to the needs of the Sufi mystics, some of whom were women. Two anecdotes from contemporary observers underline the power enjoyed by women of the court and the high regard in which they were held. The first is from the late twelfth-century Andalusian traveller Ibn Jubayr, whose vivid and awestruck observations are seen through the lens of a more conservative Muslim culture. Recounting his encounter in 1184 with three royal women, or khåtËns as he called them, during the pilgrimage to Mecca, he said: Among the strange affairs that are discussed and listened to by men we witnessed the following. One of the aforementioned khåtËns, the daughter of prince Mas’Ëd . . . came to the Mosque of the Apostle of God – may God bless and preserve him – on the evening of Thursday, the sixth of Mu˙arram, the fourth day of our arrival at Medina, riding in her litter (qubbah), surrounded by the litters of her ladies and handmaidens and led by Koran-readers, while pages and eunuch-slaves, bearing iron rods, moved around her driving the people from her path until she arrived at the venerated Mosque. Wrapped in an ample cloak, she descended and advanced to salute the Prophet – may God bless and preserve him – her servants going before her and the officials of the Mosque raising their voices in prayer for her and extolling her fame. She came to the small raw∂ah between the venerated tomb and the pulpit, and prayed there wrapped in her cloak while the people who thronged around her were kept back by the rods. She then prayed in the ˙aw∂ beside the minbar, and moving thence to the west wall of the venerated raw∂ah, sat in the place where it is said that the Angel Gabriel – peace be upon him – came down. The curtain was then lowered on her, and her pages, slaves, and chamberlains remained behind the curtain receiving her commands. She had brought with her to the mosque two loads of provisions as alms for the poor, and stayed in her place until night had fallen.3 A second account by Ibn Wåßil, contemporary with Îayfa KhåtËn, describes a journey from Cairo to Aleppo taken by al-Malik al-Kåmil with two of his daughters: In the year 629/1232 the sultan al-Malik al-Kåmil left Egypt in order to claim Amida [Diyarbakr] from its owner, accompanied

regent queen and architectural patron

by the ‘Elevated Veil’ Få†ima KhåtËn, wife of al-Malik al-AʾzÈz of Aleppo; Ghåziya KhåtËn, wife of al-Malik al-MuΩaffar of Hamåh; and also the judge Bahåʾ al-DÈn Ibn Shaddåd. The companion [alÍå˙iba] Ghåziya KhåtËn, mother of our lord the sultan al-Malik al-ManßËr – may God bless his soul – arrived in Hama in great spectacle and ceremony. She embellished Hama by her arrival. Meanwhile, Få†ima KhåtËn, accompanied by the judge Bahåʾ al-DÈn Ibn Shaddåd and other dignitaries, continued toward Aleppo. On her way there, her caravan was ceremoniously received at several pre-arranged spots: the first in Hama; the second in Tell al-Sul†ån, where she was received by her aunt Îayfa KhåtËn; and the third just outside of Aleppo, where al-Malik al- ʾAzÈz Mu˙ammad, sultan of Aleppo, came out to meet her. They rode back to the citadel of Aleppo together, ‘with row upon row of soldiers flanking them’.4 These anecdotes attest to the wealth of medieval Islamic court women, their independence, penchant for ceremony, and proclivity towards charitable acts. Whether in public processions or in private moments of worship, these princesses of the Zengid and Ayyubid dynasties of Upper Mesopotamia, Syria and Egypt commanded a great deal of respect and a royal status rarely accorded to court women before them. While this does not indicate anything of their political power or legal status as independent agents, it does indicate that they enjoyed a ‘public’ profile and notable presence. Indeed, they form the first group of women who are named on the inscriptions of their own monuments and in the contemporary texts. Although their person was hidden from view by veils, cloaks and curtains, their actions were highly visible. There are perhaps three factors that might have contributed to this situation and might explain its exceptional nature within the medieval Islamic world. The first is that these were princesses not concubines; that is, they were free women of noble origin rather than merely slaves in a harem. Concubines and harems undoubtedly existed in this period, although they do not seem to have been as common as they had been under the ʿAbbasids5 or as they would be later under the Ottomans.6 The small size and relatively limited means of the ruling dynasties of this period – most of whom ruled over quasi-autonomous city states – may have stood in the way of such accumulation. But it is equally possible that the austerity and militarism that characterised these dynasties, including the Ayyubids, would have also militated against such open display of luxury and consumption. Finally, one has to consider the severely limited space within which these dynasts lived: huge harems were simply inconceivable within the cramped quarters of citadels. Second, although free status was no guarantee for social promotion, Ayyubid princesses also enhanced their prestige by means of political marriages.7 This age-old institution was especially valued

77

78

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

under the Ayyubids, who ruled over Egypt and Syria as a large family confederacy consisting of quasi-autonomous provinces.8 Marriages within this confederacy were often arranged in order to initiate, foster or seal alliances among different members of the royal household. It follows then that court women, through marriage to their male counterparts, elevated their own status and fostered a sense of unity in the larger Ayyubid family tree. Simultaneously linked to two royal branches, these women enjoyed the protection of their natal families and the prestige associated with their spousal families. Leaving their father’s house and living with their husbands, often in a faraway land, they acted as diplomatic representatives and as agents of good will. The third factor, the ultimate seal of status, was giving birth to a male child, a future prince or even ruler. Procreation was obviously central to the survival of the dynasty, but it also guaranteed the continued well-being and prestige of the princess, whose status was now that of umm or wålidah, mother of a sultan or even queen mother. This was such an important status that it was adopted by Îayfa and other princesses as an epithet designating elevated status as well as protection.9 This is precisely the path that was laid out for Îayfa KhåtËn, the favourite daughter of al-Malik al-ʾÅdil, who had succeeded his brother Saladin as sultan of the Ayyubid dynasty. Al-ʾÅdil could claim control to most of Saladin’s possessions with the exception of Aleppo, the only province that Saladin granted to one of his sons, al- Ûåhir GhåzÈ. As a result of this divide, there was considerable tension between GhåzÈ and al-ʾÅdil, and it threatened to dismantle the Ayyubid confederacy. To resolve the tension GhåzÈ asked his uncle al-ʾÅdil for his daughter’s hand in marriage. With the marriage approved, Îayfa KhåtËn was dispatched to GhåzÈ in Aleppo in 1212 at the head of a great retinue that was received amidst great festivities. According to Ibn Wåßil: The khåtËn arrived in Aleppo in great ceremony. She was received by al-Malik al-Ûåhir along with the emirs of Aleppo, its turbaned [scholars], and its notables. Her entry into the citadel was a famous day. She had with her textiles, furnishings, and diverse jewelry that required fifty mules, 100 Bactrian camels, and 300 dromedaries to carry . . . It is also mentioned that she had in her retinue 100 slave girls, all singers who played different instruments; and another 100 slave girls who could make various wonderful crafts . . . When she entered the court of al-Malik al-Ûåhir, he arose and took several steps towards her and showed her great respect.10 This was probably the most important marriage in the history of the Ayyubids, for it helped remedy the seriously threatening rift between GhåzÈ and al-ʾÅdil. Îayfa was therefore instrumental in the

regent queen and architectural patron

maintenance of the unity of the Ayyubid house, a destiny shared by several other women of her standing. Îayfa’s status changed considerably in the following year when she gave birth to the male heir that GhåzÈ had desired, the son who later became sultan al-ʾAzÈz Mu˙ammad. As wife and then mother, Îayfa was little more than a passive vehicle, and any other daughter of her father’s house would have served equally well. However, destiny ultimately called her to step out of the traditional female role and to distinguish herself in the pages of history as a regent queen. Little is reported about Îayfa during the remainder of GhåzÈ’s reign, who died in 1216, or that of their son, al-ʾAzÈz Mu˙ammad, who died in 1236 at the age of twenty-four.11 Mu˙ammad’s son and successor (the later Salå˙ al-DÈn II) was, however, only seven years old at the time of his father’s death, a situation that necessitated the creation of a regency council. This council consisted of the two emirs Shams al-DÈn Luʾuʾ al-AmÈnÈ and ʿIzz al-DÈn ʿUmar b. MajallÈ, the vizier Jamål al-DÈn al-Qif†È, and Îayfa KhåtËn’s own slave Jamål alDawla Iqbål al-ÛåhirÈ al-KhåtËnÈ. The latter acted as Îayfa KhåtËn’s secretary and deputy at the regency council, whose decisions she herself ratified as regent queen in the name of her grandson. Thus, she in effect ruled Aleppo from 1236 until her death in 640/1243.12 She was the first woman to govern an Islamic dynasty, pre-dating by about fifteen years the far more notorious queen Shajar al-Durr, who ruled Egypt from 1250 to 1257 before being ignominiously killed.13 Though the actual sovereign of an Islamic state, Îayfa KhåtËn seems to have exercised considerable restraint and discretion in keeping her voice, her person and her regnal symbols hidden from view. The secretary Iqbål was her mouthpiece to the regency council and to the public alike,14 and she was herself an embodiment of her titulature which described her as the ‘elevated curtain and impregnable veil’, metaphors for virtue and chastity. Furthermore Îayfa abstained from claiming for herself the classical emblems of Islamic rulership, namely demanding that coins be struck in her name and the Friday khutbah (sermon) be pronounced in her name. Keenly aware of the limits of female power and the boundaries of acceptable behaviour, she walked a fine line between her de facto sovereignty and her de juris position as simply the grandmother and advisor to the future sultan.15 Symbolically, the line of male succession in the Ayyubid dynasty of Aleppo continued uninterrupted despite her regency.16 Rather than threatening this lineage, her regency protected it, which may explain the public’s tolerance of her rule. Îayfa KhåtËn’s seven-year reign was militarily uneventful, making it possible for her to dedicate her efforts to civic affairs and the patronage of scholars and mystics, for whom she created several important institutions. Ibn Wåßil wrote: ‘She was just to her subjects, very charitable and loving towards them. She removed various taxes in all the regions of Aleppo. She favored jurists, ascetics, scholars

79

80

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

and people of religion, and extended to them many charities.’17 This testimony is confirmed by Îayfa’s architectural patronage, which focused on khånqåhs (Sufi convents) and mausoleums. Îayfa’s involvement in architectural patronage was the rule rather than the exception among Ayyubid court women. In a recent study, Humphreys has estimated that about one fifth of all religious structures and one fourth of madrasas (colleges) built by the Ayyubids of Damascus were founded by women.18 Although court women under the earlier Fatimids and later Mamluks also commissioned pious foundations, their relative number falls below those founded under the Ayyubids.19 First in Damascus and then in Aleppo, court women of the Zengid and Ayyubid dynasties founded mausoleums, madrasas and khånqåhs in order to promote religious learning and as a means for enhancing their prestige and commemorating their names. Sitt al-Shåm (Lady of Damascus), wife of Saladin, for example, founded the Madrasa al-Shåmiyya in 1185, initiating a type of family mausoleum not previously known in Damascus.20 Akhshu KhåtËn and KhadÈja KhåtËn, respectively wife and daughter of the Ayyubid sultan al-MuʾaΩΩam ʿIsa, built two madrasas of some note in Damascus: al-Måridåniyya in 1227 and al-Íå˙ibiyya before 1245.21 All three madrasas are fairly small, consisting of little more than a mosque with an attached mausoleum, and lacking the residential cells present in larger educational foundations.22 In Ayyubid Aleppo, women’s patronage seems to have found an outlet primarily in the building of khånqåhs, of which six are mentioned by Ibn Shaddåd, the thirteenth-century topographer of Syria and Upper Mesopotamia.23 Interestingly, each of these khånqåhs was also intended for women ascetics, a fact that underlines the centrality of Sufism in female piety. Indeed Sufism, which did not require the same kind of knowledge and erudition as jurisprudence, was somewhat more open for women’s participation and even leadership. Schimmel, for example, proposes that ‘sufism, more than stern orthodoxy, offered women a certain amount of possibilities to participate actively in the religious and social life’.24 In a study specifically dealing with education in Mamluk Cairo, Berkey concludes that ‘formal Sufism and the broader forms of Muslim mysticism may have been an important route by which learned women could acquire standing in the religio-academic world’.25 The emphasis on a direct and personal path to the Divine, the prominence of emotional and ecstatic practices, and perhaps even the possibility of celibacy may have conjoined to make Sufism especially attractive to medieval Muslim women. It is also quite likely that some of these convents served as hospices for unmarried or widowed women without families to shelter them, who would have found in them protection as well as spiritual fulfilment. Although this important feature is not noted for any of the Ayyubid khånqåhs in Aleppo, Berkey discusses several

regent queen and architectural patron

instances in which ribåts (nearly synonymous with khånqåhs) were specifically intended as ‘places of residence for elderly, divorced, or widowed women who had no other place of abode, until their death or remarriage’.26 This was most likely the case with the khånqåh al-Farafra. The two structures founded by Îayfa in Aleppo are the khånqåh al-Farafra located beneath the citadel (ta˙t al-qal’ah) and the Madrasa al-Firdaws in a southern suburb called the Maqamat quarter south of the city wall (Figure 4.2). Both were privileged locations, and thus especially desirable for pious foundations: the former was the residential quarter of the city’s notables; the latter already contained several other Ayyubid madrasas and mausoleums. Indeed, the location of these two monuments closely corresponds to the two monuments built by Îayfa’s husband GhåzÈ two decades earlier, suggesting a degree of parallelism between hers and her husband’s patronage. The main difference is that GhåzÈ’s monuments were madrasas whereas those built by Îayfa were khånqåhs. While this identification is clear in the case of al-Farafra, it goes counter to the accepted function of al-Firdaws as madrasa. Elsewhere I have shown that this building was not intended for jurists but for Sufis, making it a khånqåh, and this identification sits more comfortably with the pattern of women’s patronage in Aleppo.27 Al-Firdaws, the larger of the two, was founded just before Îayfa’s ascendance to regency in 1236; and al-Farafra was founded just a year after. This is a clear indication of her already powerful and privileged standing in the dynasty as the daughter, widow, mother and grandmother of Ayyubid sovereigns. Indeed, these attributes seem to have outweighed Îayfa’s

Figure 4.1  Aleppo, Madrasa al-Firdaws, 1235–6, aerial view from east

81

82

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 4.2  Aleppo, reconstructed plan of central and southern city (Maqåm Ibrahim is about 300 metres south of Madrasa al-Firdaws)

regent status, as an examination of her remaining inscriptions will demonstrate. Three historical inscriptions for Îayfa KhåtËn have survived, one in the khånqåh al-Farafra and two in the madrasa al-Firdaws. The

regent queen and architectural patron

Figure 4.3  Aleppo, Khånqåh al-Faråfra: foundation inscription 635/1237 (Photo: Yasser Tabbaa)

Figure 4.4  Aleppo, Madrasa al-Firdaws: exterior inscription: titles of Îayfa KhåtËn (photo: Yasser Tabbaa)

first (Figure 4.3), dated to the first year of her regency in 635/1237, does not name her as the founder but simply states that the khånqåh was built during the reign of Íalå˙ al-DÈn II, her young grandson, mentioning also his father and grandfather.28 In fact, her own name does not appear at all. Instead, her identity as a patron was subsumed within the identity of the grandson in whose name and for whose dynasty she governed. At the Madrasa al-Firdaws one inscriptional band is located on the exterior along the portal façade and the other on the interior, along the eastern wall of the courtyard. Although dated to 1235–6, two years earlier than the previous inscription, they are somewhat more assertive in their titulature, possibly owing to the extramural location of the madrasa. Both mention Îayfa KhåtËn by name and give her titles, epithets and royal lineage. The outer one (Figure 4.4) reads as follows: This is what has ordered its construction the elevated curtain and impregnable veil, the Merciful Queen . . . Îayfa KhåtËn, daughter of the sultan al-Malik al-’Ådil . . . during the reign of our lord the sultan al-Malik al-Nåßir, the learned, the just, the warrior for the

83

84

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 4.5  Aleppo, Khånqåh al-Faråfra: plan (adapted from a plan of Syrian Dir. Gen. of Antiquities and Museums)

faith, the defender of the outposts, the assisted by God, the victorious, the vanquisher, Íalå˙ al-DÈn son of al-Malik al-’AzÈz son of al-Malik al-Ûåhir GhåzÈ . . . in the year 633/1235–36.29 These are some of the earliest feminine honorific and royal epithets known. The titles ‘al-sitr al-rafi’ w’al-˙ijåb al-manÈʾ’, literally meaning ‘the elevated curtain and impregnable veil’, were sometimes used to describe fortifications. But figuratively they refer to virtue and chastity and should perhaps be translated as ‘the virtuous veil and chaste lady’. These are followed by ‘the Merciful Queen’ and ‘the refuge of world and religion’, titles that were later adopted by Queen Shajar al-Durr of Cairo.30 The courtyard’s historical inscription abbreviates the exterior inscription, omitting the date and title ‘the Merciful Queen’ but adding the unusual though pertinent detail that Îayfa KhåtËn was the mother (wålidat) of the sultan al-Malik al-ʾAzÈz.31 This information augments the exterior inscription and leaves no doubt that the building was built by a royal figure, the mother of the ruling sultan. The two monuments founded by Îayfa share several features of plan and design (Figures 4.5 and 4.7). Both are entered through bent entrances, which were quite uncommon in contemporary madrasas, a straight axial or cross-axial entrance being far more common.32 The bent entrance is more commonly associated with residential archi-

regent queen and architectural patron

Figure 4.6  Aleppo, Madrasa al-Firdaws, 1235, courtyard, from north (photo: Yasser Tabbaa)

tecture, where it ensures privacy and protection. This might explain its use in the khånqåh al-Farafra, which was intended for women Sufis, whose personal security would have been well guarded, but it does not explain its use in the Madrasa al-Firdaws, which was most likely for men. Both buildings employ a fairly typical courtyard plan with an elaborate central pool. The southern quadrant in each is occupied by a mosque that is faced on the north by a large iwån, while the eastern and western flanks are filled with various rooms. The tripartite prayer halls have elaborate marble mi˙råbs, which is the only concession to ornamentation in these austere buildings. Furthermore, both buildings contain what might be read as residential annexes (for their respective shaykhs [scholar-in-residence]) centred around their own little courtyards: one in al-Farafra and two in al-Firdaws. Overall, however, al-Firdaws is larger, better designed, and somewhat less severe in its decoration.33 But there are several significant differences between the two buildings. Perhaps the most apparent is that al-Firdaws does not contain individual residential cells as was customary in most madrasas and khånqåhs of any size. Al-Farafra, on the other hand, contains a total of thirty-four such cells, arranged on two levels. Indeed, a second floor is entirely absent at al-Firdaws, whose flanks continue the undulating rhythm of the squat domes that cover the sanctuary for a total of eleven domes. It seems clear, therefore, that these residential units were sacrificed in order to maximise the number of domes, a point to which I will return.

85

86

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 4.7  Aleppo, Madrasa al-Firdaws: plan with applied grid (adapted from a plan of Syrian Dir. Gen. of Antiquities and Museums)

The second most noteworthy difference is the external iwån at al-Firdaws, which adorses the interior one and which originally faced a garden with a pool (Figure 4.8). Such iwåns are very uncommon in religious architecture, but have a long history in palatial architecture where they were used as annexed or freestanding pavilions facing a body of water and a garden. In its present adaptation, the iwån may have served as an audience chamber for the patron where she would have received prominent Sufis. Metaphorically, its use in this building might be the reason behind the name of the madrasa, since Firdaws (Paradise) and garden are intimately linked.34 This paradisiac imagery is further reinforced by the portal inscription, which begins with Qurʾånic verses (63:68–72) that offer a vision of the faithful entering Paradise where they are offered exquisite food on plates of gold. Indeed, a similarly evocative Qurʾånic passage was also selected for the portal of al-Farafra: ‘Praise belongs to God who has put away all sorrow from us. Surely our Lord is all-forgiving, all-thankful, who of His bounty has made us to dwell in the abode of everlasting life wherein no weariness assails us neither fatigue’ (Qurʾån, 35:34–5). Both excerpts are very rare in public inscriptions

regent queen and architectural patron

Figure 4.8  Aleppo, Madrasa al-Firdaws: exterior iwån (photo: Yasser Tabbaa)

Figure 4.9  Aleppo, Madrasa al-Firdaws, 1235, inscription in courtyard (photo: Yasser Tabbaa)

and seem to refer to Paradise as the reward for the Sufis who worship in these foundations.35 Far more striking are the inscriptions that surround three sides of the courtyard at al-Firdaws for a length of about sixty metres

87

88

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 4.10  Aleppo, Madrasa al-Firdaws, 1235, mi˙råb (photo: Yasser Tabbaa)

Figure 4.11  Aleppo, Madrasa al-Firdaws, courtyard inscription (photo: Yasser Tabbaa)

(Figure 4.11). These texts, which are unpublished, are significant because they describe vividly and passionately the actions and aspirations of the Sufis who worshipped at al-Firdaws.36 Their worship, which took the form of nightly vigils of repetitive chanting and frenzied dancing, was to culminate in a direct vision of God, whose apparition was seen as the ultimate hope and reward of the Sufis. This vision, which the non-Sufis only experienced in Paradise, was a foretaste of the eternal bliss that awaited the Sufis and a reward that further impelled them in their worship. It is, of course, impossible for us to determine the extent to which Îayfa was involved in the epigraphic programme of her madrasa/

regent queen and architectural patron

Figure 4.12  Aleppo, Khanqah al-Farafra, 1237, courtyard from above (photo: Yasser Tabbaa)

khånqåh. But in view of her known inclination to Sufism and particularly acknowledging the astounding originality of these inscriptions, we are entitled, I believe, to grant her some influence in their selection and formulation. If so, what was their intent and that of the madrasa generally? The first motive was to create a Sufi haven in the city where one of the greatest Sufis of all time, SuhrawardÈ MaqtËl, had been executed forty years earlier. Instigated by conservative theologians and ordered by al-Ûåhir GhåzÈ himself, the execution was widely deplored in the Islamic world.37 Al-Firdaws may therefore be seen as a long overdue atonement for a grave error and a recognition of the Sufis as an important social force in Islam and particularly in Ayyubid Aleppo. The second motive was to create a SunnÈ alternative to the highly impassioned Shiʿi ceremonies that were regularly held in some shrines in Aleppo and that exerted a particular attraction in the minds of the populace. Indeed, it is quite likely that the eleven domes of al-Firdaws – unique among all Aleppine madrasas – were intended to compete with the equally large number of domes at the main Shiʿi shrine of al-Husayn just west of the city.38 Through her generous patronage of the Sufis, therefore, Îayfa KhåtËn hoped to ensure the allegiance of this sector of society, who attracted the populace in a manner far more direct than the intellectual appeal of orthodox religious scholars, the ‘ʿulamå’. It also

89

90

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

seems clear that she (as with other Ayyubid dynasts before her) was still engaged in a theological rivalry with Shiʿism, the so-called SunnÈ Revival, whose origins in Syria date back to the first half of the twelfth century.39 Indeed, her Madrasa al-Firdaws might be seen as an attempt to use Sufi ascetic and ecstatic practices as a SunnÈ paradigm for human salvation. Îayfa KhåtËn, Syria’s sole female Ayyubid ruler, impresses us today by her influential, though discreet, almost self-effacing form of sovereignty. Before and after her accession to the regency, she worked within the Ayyubid system, avoiding controversial acts and veiling herself from the public eye by using the regency council as an intermediary and by living a life of profound piety. Deeply interested in Sufism, she dedicated considerable energies and funds to promote its well-being and to lead it into new horizons. Compared to the better-known Queen Shajar al-Durr of Egypt, she appears distinctly demure. Although today we might be more attracted to Shajar alDurr’s more assertive personality and disruptive acts, she was in the end far less effective as a queen. The Egyptian queen suffered a horrific death for her alleged transgressions and left the country with a shredded dynasty that was soon taken over by the Mamluks. Îayfa KhåtËn, on the other hand, left two important monuments and a legacy of piety and tolerance.

Notes 1. Fatima Mernissi, The Forgotten Queens of Islam (Cambridge: Polity Press, 1992); Leila Ahmed, Women and Gender in Islam: Historical Roots of a Modern Debate (New Haven and London: Yale University Press, 1992). By far the weakest chapter in Ahmed’s pioneering work is the one entitled ‘Medieval Islam’, pp. 102–24. 2. Caroline Williams, ‘The Cult of ʿAlid Saints in the Fatimid Monuments of Cairo. Part II: The Mausolea’, Muqarnas 3 (1985): 39–60; Caroline Williams, ‘The Mosque of Sitt Hadaq’, Muqarnas 11 (1994): 55–64; R. Stephen Humphreys, ‘Women as Patrons of Religious Architecture in Ayyubid Damascus’, Muqarnas 11 (1994): 35–54; and Patronage by Women in Islamic Art, ed. Esin Atil, in Asian Art, 5.2 (Spring 1993). 3. Ibn Jubayr, The Travels of Ibn Jubayr, ed. and tr. by R. J. C. Broadhurst (London, 1952): 207–8. The terms raw∂ah (garden) and ˙aw∂ (basin) are used figuratively to highlight the paradisiac associations of the shrine of the Prophet. For other accounts of the same women, see ibid., 189–90, 239 and 246. These three women were: SaljËqah, the daughter of ʿIzz al-DÈn MasʾËd, lord of Konya; the wife of Qu†b al-DÈn MawdËd, the previous ruler of Mosul; and the daughter of Tukush Shah, the prince of Isfahan. 4. Ibn Wåßil, Mufarrij al-kurËb fi akhbår dawlat banÈ AyyËb, ed. J. Shayyal (Cairo, 1953), 5: 29–30. See also ibid., 3: 124. 5. For a discussion of pervasive concubinage in early ʿAbbasid society, see Ahmed, Women and Gender in Islam, 81–7. See also Judith Tucker, ‘Gender and Islamic History’, in Islamic and European Expansion, The

regent queen and architectural patron

Forging of a Global Order, ed. Michael Adas (Philadelphia: Temple University Press, 1993), 37–74. 6. Leslie Peirce, The Imperial Harem: Women and Sovereignty in the Ottoman Empire (New York: Oxford University Press, 1993), 121–2. 7. In Islam, political marriages were given special sanction since they followed in the example of the Prophet himself, most of whose later marriages were political in nature. On Mu˙ammad’s marriages, see Ahmed, Women and Gender in Islam, 49–59. 8. The term ‘Ayyubid confederation’ was first coined by R. Stephen Humphreys in his From Saladin to the Mongols: The Ayyubids of Damascus, 1193–1260 (Albany: State University of New York Press, 1977). See pp. 66–73 for a discussion of the legal and political dimensions of this concept. 9. Shajar al-Durr was in fact referred to in contemporary documents and chronicles as Umm KhalÈl, although that prince (her son from Sultan Najm al-DÈn AyyËb) had died in infancy. See K. A. C. Creswell, The Muslim Architecture of Egypt (MAE), 2 vols (New York: Hacker Art Books, 1978), 2: 135. 10. Mufarrij, 3: 213–14. 11. The volume(s) pertaining to women in the famous biographical dictionary of Kamål al-DÈn Ibn al-ʾAdÈm, Bughyat al-†alab fÈ tårÈkh ˙alab, have been lost. This is especially unfortunate since Ibn al-ʾAdÈm (1192–1262) was one of the greatest chroniclers and biographers in medieval Islam and since his Bughya contains the most detailed information on the important personalities of his time. See note 14. 12. Mufarrij, 3: 229–30. 13. A discussion of Shajar al-Durr’s monuments and a summary of her rise to power and terrible death are provided by Creswell, The Muslim Architecture of Egypt, 2: 135–40. Her monuments are also briefly discussed in Doris Behrens-Abouseif, Islamic Architecture in Cairo: An Introduction (Leiden: E. J. Brill, 1992), 90–3. 14. On Jamål al-Dawla Iqbål al-ÛåhirÈ al-KhåtËnÈ, see David Morray, An Ayyubid Notable and His World, Ibn al-’AdÈm and Aleppo as Portrayed in His Biographical Dictionary of People Associated with the City (Leiden: E. J. Brill, 1994), 69–70. Iqbål is repeatedly criticised by Ibn al-ʾAdÈm for his extremely aggressive and arrogant demeanour. It is possible that these very traits would have made him especially suited for his position as chamberlain and mouthpiece of the regent queen. 15. This is entirely different from Shajar al-Durr, who, flaunting tradition, had coinage struck and the khutba pronounced in her name. This act of transgression was even criticised by the ʿAbbasid caliph himself, who reputedly wrote to the Egyptian leaders that ‘if they had no man, he would send them one’. Cited by Creswell, The Muslim Architecture of Egypt, 2: 135, after MaqrÈzÈ, SulËk, 1: 358. 16. One indication of Îayfa’s uncanonical status is that she is not mentioned under the list of the Ayyubid dynasts of Aleppo in C. E. Bosworth, Islamic Dynasties (Edinburgh University Press, 1963). This omission most likely has to do with the absence of her name from the numismatic evidence, which is generally the foundation of Muslim ­dynastic sequences. Shajar al-Durr, on the other hand, is listed in the same source (p. 63) as the first Mamluk sovereign of Egypt. Since she was at first married to an Ayyubid sultan, she could also be seen as the last of the Ayyubids.

91

92

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

17. Ibn Wåßil, Mufarrij, 5: 313. 18. Humphreys, ‘Women as Patrons’, 35–54. 19. For the Mamluks, see Jonathan Berkey, The Transmission of Knowledge in Medieval Cairo: A Social History of Islamic Education (Princeton, NJ: Princeton University Press, 1992), 162–81. 20. Ernst Herzfeld, ‘Damascus: Studies in Architecture – III’, Ars Islamica 11–12 (1941): 38–42. 21. Ibid., 19–20 and 64–5; and Humphreys, ‘Women as Patrons’, 43–4. 22. The smallness of some Ayyubid madrasas and dearth of students cells within them has been noted by Robert Hillenbrand in ‘Madrasa’, Encyclopaedia of Islam, new edition (Leiden, 1960–), 6: 1139, and nearly verbatim in Islamic Architecture: Form, Function and Meaning (New York: Columbia University Press, 1994), 186–90. Although one might, as Hillenbrand did, question the educational possibilities of a madrasa without residential cells, it is equally possible that students would have resided elsewhere and simply used the madrasa as a day school. 23. `Izz al-DÈn Ibn Shaddåd, Al-A’låq al-kha†Èra fi dhikr umarå’ al-shåm wa’l-JazÈra. Ta’rÈkh ˙alab, ed. D. Sourdel (Beirut: Institut Français de Damas, 1953), 95–6. 24. Annemarie Schimmel, Mystical Dimensions of Islam (Chapel Hill: The University of North Carolina Press, 1975), 432. 25. Berkey, The Transmission of Knowledge in Medieval Cairo, 175. 26. Ibid., 174. 27. Yasser Tabbaa, Constructions of Power and Piety in Medieval Aleppo (University Park: Pennsylvania State Press, 1997), 163–82. Humphreys in ‘Women as Patrons of Religious Architecture’, p. 35, states that ‘this was in fact the single madrasa in Aleppo during the Zangid-Ayyubid era to be founded by a woman (out of more than fifty madrasas in the city dating from those 133 years)’. This anomaly no longer needs to be sustained since it is clear that al-Firdaws functioned primarily as a khånqåh. 28. Ernst Herzfeld, Matériaux pour un Corpus Inscriptionum Arabicarum: Syrie du Nord. Deuxième partie, Monuments et inscriptions d’Alep, 3 vols (Cairo: Institut français d’archéologie orientale, 1954–6), 2: 305. 29. Répertoire chronologique d’épigraphie arabe, 16 vols in progress (Cairo, 1931–), no. 4084. 30. Herzfeld, Monuments et inscriptions d’Alep, 2: 298. 31. Répertoire, no. 4081. 32. This distinction applies mainly to the madrasas of Aleppo, where the two built by al-Ûåhir GhåzÈ, al-Sul†åniyya and al-Ûåhiriyya, have straight axial entrances while two others built by notables, al-Sharafiyya and al-ʾAdÈmiyya, have straight cross-axial entrances. See Herzfeld, Monuments et inscriptions dʾAlep, passim and my Constructions of Power and Piety for plans and discussion of these buildings. Elsewhere in the Islamic world, matters are different, and madrasas with bent entrances are by no means uncommon. 33. For a study of the geometry and design features of the Madrasa alFirdaws, see my ‘Geometry and Memory in the Design of the Madrasat al-Firdaws in Aleppo’, in Theories and Principles of Design in the Architecture of Islamic Societies, ed. Margaret Šev¥enko (Cambridge, MA: The Aga Khan Program for Islamic Architecture, 1988), 23–34. 34. Ibid., pp. 29–34.

regent queen and architectural patron

35. Erica Dodd and Shereen Khairallah, The Image of the Word, 2 vols (Beirut: American University of Beirut Press, 1981): 2: 101–2, list only six other monuments that use this Qurʾånic verse. 36. Tabbaa, Constructions of Power and Piety, 173–81, contains a full transcription, translation, and a detailed discussion of the inscriptions of this monument. 37. Schimmel, Mystical Dimensions of Islam, 260. 38. Tabbaa, Constructions of Power and Piety, ch. 6. 39. For a discussion of the roots of this movement, see George Makdisi, ‘The SunnÈ Revival’, in D. S. Richards, Islamic Civilization, 950–1150 (Oxford: B. Cassirer, 1973), 155–68. For a discussion of its efflorescence during the reign of NËr al-DÈn (1146–74) and the early Ayyubids, see Nikita Elisséeff, NËr al-DÈn un grand prince musulman de Syrie au temps des Croisades (511–569/1118–1174), 3 vols (Damascus: Institut français de Damas, 1967), esp. 750–80; and my ‘The Architectural Patronage of NËr al-DÈn Ma˙mËd ibn ZangÈ, 1146–1174’, Ph.D. diss. (New York University, 1982), 169–85.

93

CHAPTER FIVE

Defending Ayyubid Aleppo: The Fortifications of Al-Z.a¯hir Gha¯zı¯ (1186–1216) Introduction: castles and medieval Islamic culture Two observations by medieval Arab authors help us view citadels and fortifications from a broad contemporary perspective. The first, by Ibn KhaldËn, contains three significant propositions about the centrality of fortifications for urban culture. The fourteenthcentury North African historian states first, that the building of fortifications is the sole privilege and responsibility of rulers, and conversely, that fortifications embody dynastic power. Second, Ibn KhaldËn suggests that fortifications are a necessary and nearly sufficient condition of urban life, for no city can survive without walls and a citadel, and the presence of these two components ensure the possibility of a city. Finally, these fortifications protect urban dwellers (˙a∂arÈs) against nomads (badawÈs), substituting an artificial protective structure for the natural cohesion (ʿaßabiyya) of the desert dwellers.1 The second observation, embedded in an anecdote told first by the twelfth-century historian Ibn al-AthÈr and retold by Aleppo’s thirteenth-century historian Ibn al-ʾAdÈm, specifically refers to Syrian forts and citadels. Shortly after a serious illness that he had suffered in 1186, Íalå˙ al-DÈn was taking a walk with one of his trusted companions, Sulaymån b. Jandar, who told him the following anecdote. ‘If a bird wanted to build a nest for its fledglings, it would reach for the tops of trees in order to protect them. But you have surrendered the forts to your family (i.e. household) while leaving your children on the ground. Aleppo . . . is in the hands of your brother, Hama is with TaqÈ al-DÈn; Homs is in the hands of Asad al-DÈn; and your son al-Af∂al is with TaqÈ al-DÈn in Egypt from where he can expel him at will . . . Íalå˙ al-DÈn said: “You have spoken the truth, but keep it a secret.”’2 This anecdote is intended to explain Íalå˙ al-DÈn’s about-face in Yasser Tabbaa (2006), ‘Defending Ayyubid Aleppo: The Fortifications of alZahir Ghazi (1186–1216)’, in Hugh Kennedy (ed.), Muslim Fortifications in Bilad al-Sham, Leiden: E. J. Brill, 175–83.

defending ayyubid aleppo

taking Aleppo from his brother al-ʾÅdil and giving it back to his son al-Ûåhir GhåzÈ, who had briefly ruled it as a child.3 But the tale is even more telling about the importance of fortresses and citadels in medieval Syria and the Islamic world generally. And it is undoubtedly this citadel culture, whether in Greater Syria or in North Africa, that forms the historical foundation upon which Ibn KhaldËn in the fourteenth century created his theoretical edifice. Interestingly, both Ibn al-ʾAdÈm and Ibn KhaldËn discuss fortifications in terms of dynastic preservation and protection from neighbours (whether competing dynasts or Bedouins) rather than defence against non-Muslim enemies, Crusaders or otherwise. These internal pressures seem to have been the most compelling factors for the creation of citadels and fortresses in the Islamic world, and they help us explain their geographic location and even their internal composition, so different from their Crusader counterparts. One can argue that the fortified citadel city constituted one basic difference, perhaps the most basic, between early and middle Islamic cultures. Although fortresses and defended outposts (thughËr) were known in early Islam, there is no evidence that they were used as centres of royalty or administration. In fact, Ibn Shaddåd specifically writes that the Mirdasids, in the second half of the eleventh century, were the first to establish their palace within the citadel of Aleppo, thereby ‘creating a precedent (sunna) for the kings who followed them’.4 This contemporary observation finds archaeological support in the sudden emergence of citadel cities in the late eleventh century and their proliferation throughout the Islamic world by the twelfth century. Indeed, as Bacharach and I have proposed, the citadel city represents and defines the third phase in Islamic urbanism, whereby the amßår (garrison towns) with their centrally located palaces and the Samarra-style cities with their vast extra-urban palaces represent the first and second phases.5 The creation of these palatial urban citadels in the twelfth century had huge ramifications for medieval Islamic polity, military, and urbanism whose extent remains to be fully explored. But three of these consequences should be briefly mentioned in view of their import for the role of fortifications in Islamic cities. First, the palatial citadel replaced horizontal distance between the ruler and the ruled, characteristic of the second phase of Islamic urbanism, by vertical separation, placing citadels above their corresponding cities and populations.6 Second, though impregnable and separate from the city, these citadels had to be linked with it through various physical and ceremonial ways. This linkage was quite often manifested in public squares beneath the citadel, which in Aleppo also contained the dår al-’adl.7 Third, the necessarily circumscribed domain of these citadels led to quite small palaces, which were on the average only about 1/50th the size of ʿAbbasid or Ghaznavid palaces.8

95

96

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Ayyubid castles: patronage This paper focuses on the city and province of Aleppo between about 1180 and 1260, a period during which Aleppo was ruled by a branch of the Ayyubids, the dynasty first founded by Íalå˙ al-DÈn under a largely symbolic capital and sultan in Cairo. In most other respects, the provincial Ayyubid dynasties of the main cities of Syria were largely autonomous principalities loosely confederated within the Ayyubid household, all of whom owed their ultimate allegiance to the ʿAbbasid Caliphate. Even within this loose framework, the Ayyubids of Aleppo were exceptional in terms of their independence and autonomy. First, they, alone among the other branches, were able to follow an uninterrupted hereditary succession, directly from the legendary Íalå˙ al-DÈn. Second, from about 1190, the Ayyubids of Aleppo adopted the title Sultan, the only Ayyubid province to carry the title normally restricted to the supreme ruler of the dynasty. Third, their coinage, with its Seal of Solomon and an epigraphic formula that invariably included the name of the Aleppo sultan and the ʿAbbasid caliph, sometimes without any mention of the Egyptian sultan, differed from all the other Ayyubid provinces.9 The autonomy of Ayyubid Aleppo was partly due to its relative isolation from Egypt and even Damascus, but it also has to do with the extraordinary efforts at consolidation and centralisation practised in particular by al-Ûåhir GhåzÈ (1186–1216), efforts that are quite amply illustrated by his building campaign of citadels and fortresses. Studying the history of al-Ûåhir GhåzÈ (1186–1216) through chronicles and tracing his architectural patronage through his inscriptions could be somewhat confusing were it not for Gaston Wiet’s brilliant discovery of a pattern to GhåzÈ’s conquests.10 Writing in 1931 Wiet noticed that GhåzÈ’s numerous inscriptions outline a pattern of methodical conquest and refortification of towns and forts surrounding Aleppo. Humphreys, writing in the 1990s reiterates Wiet’s formulation, adding that GhåzÈ sought ‘to bring the great castles of his realm under his direct control . . . thereby reducing or eliminating the autonomous role of the muq†åʾs of north Syria’.11 These fortresses, which generally surround Aleppo at a radius of about 100–150 kilometres, seem to have formed its outermost ring of defence. Although the domain of the Ayyubids of Aleppo remained relatively constant over their eighty-year rule period, it witnessed some expansion during GhåzÈ’s reign. His domain around 1192 included, in addition to Aleppo, the fortresses and towns of Óårim, Aʾzåz, Tall Båshir and Darabsak/al-Shughr, although even some of these forts had to be re-conquered or at least renegotiated from their feudal lords (Figure 5.3). Lådhiqiyya, for example, which had previously been taken by Íalå˙ al-DÈn, had to be taken again in 592/1196 after its muq†aʾ had rebelled against Aleppo.12 We note that these early forays were made

defending ayyubid aleppo

Figure 5.1  Citadel of Aleppo: ramp and entry block, early thirteenth century (photo: Yasser Tabbaa)

in the direction of the Franks to the west and the Seljuqs to the north and that their intention seems to have been not so much conquest and domination but the creation of defensible borders. Furthermore, the motivation was rarely religious, for we know that both GhåzÈ and his uncle al-ʾÅdil made alliance with various Crusader princedoms against each other and against other local dynasts. When we shift to about 1210 we note that GhåzÈ has expanded his domain to the south-west, where he has added the fortresses of Shayzar, al-Mu∂Èq, Kafar†åb, Maʾarrat al-Nuʾmån, and even BåʾrÈn, nearly surrounding Hama (Figure 5.4). Interestingly, he does not take Hama, which remains an independent Ayyubid princedom long after the end of the Ayyubid sultanate at the hands of the Mongols. More surprising perhaps is his expansion to the north-east of Aleppo in the direction of the Euphrates. In addition to Qalʾat Jaʾbar, he takes Manbij and Qalʾat Najm in 1202 and a little later SarËj and Ras al-ʾAyn, thereby extending his domain well into the JazÈra. GhåzÈ was perhaps motivated in this easterly expansion by his ongoing competition with al-ʾÅdil, who had ambitions in the JazÈra.13 But controlling the trade between Aleppo and upper Mesopotamia, especially Mosul and Irbil, was probably an even more important factor. While protecting important crossing points along the Euphrates, these fortress would have also kept the unruly nomadic tribes appeased and at bay. Overall, GhåzÈ’s far-reaching plan of conquest and refortification seems to have been largely motivated by political, military and even commercial factors. Politically, GhåzÈ sought to replace the quasiindependent feudalists by a somewhat more centralised mechanism based on his own appointees. These appointees, whose allegiance

97

98

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

was at least in part ensured by being housed in huge castles with elaborate palaces, were largely drawn from the Ayyubid household and from the Kurdish military aristocracy. Militarily, these rebuilt fortresses created defensible borders, a kind of cordon militaire around his capital, Aleppo. Commercially, GhåzÈ’s expansion to the Euphrates was largely intended to control the trade between Aleppo and upper Mesopotamia, especially Mosul and Irbil. On the other hand, we note that very few of these fortresses were built to ward off the Crusaders, who rarely posed any threat to the Aleppo province during the first half of the thirteenth century. Ayyubid castles: form and meaning How do all these suggestions and proposals correspond to or inform the actual form of GhåzÈ’s fortresses? I would like to look briefly at some of the most salient features of fortresses that were substantially rebuilt by GhåzÈ, including Aleppo, Najm, Óårim, Mu∂Èq and ÍahyËn. We note first that the first four belong to the truncated cone type, which is also known in the citadels of Hama, Homs, Mu∂Èq, ʿAyntåb and other castles extending between central Syria and southern Anatolia (Figure 5.5). Originally, these cones would have been covered with a smooth ashlar glacis and surrounded by a moat, remains of which survive in Aleppo, Najm and Óårim (Figure 5.6). In his recent study of the Hama citadel, Peter Pentz writes that ‘with its moats, glacis, and its curtain wall the resemblance to the famous citadel of Aleppo must have been striking, and the . . . shape of the mound links it to a group of Ayyubid fortifications called the “tell troncoconico”’.14 Second, all these fortresses have a well-defended entrance block which was invariably the most monumental and perhaps costliest part of any Ayyubid citadel (Figure 5.7). Although none of these entrance blocks is as complex or monumental as Aleppo’s magnificent gateway, they share with it the ramp, the multiply bent axis, the long durgahs, and perhaps portcullises; all devices intended to strengthen the gate and impede the progress of the enemy. In particular the entrance block to Qalʾat Najm, with its ramp, four bends and two durgahs, is a simplified version of the Aleppo’s entry block (Figure 5.8). More generally, castles at ʿAyntåb, Shughr-Bakås, Maʾarrat al-Nuʾmån and perhaps elsewhere show considerable affinities, possibly suggesting a well-established building tradition and a modular approach to design. What mainly distinguish Najm and other provincial fortresses from Aleppo is their smaller towers and fewer embrasures, including loopholes and especially box machicolations (Figure 5.9). Indeed, it seems that box machicolation, which simultaneously emerges in Crusader and Islamic castles in the beginning of the thirteenth century, was in the Islamic world mainly used in urban citadels,

defending ayyubid aleppo

Figure 5.2  Qalʾat Najm, early thirteenth century, view from the Euphrates (photo: Yasser Tabbaa)

including Aleppo, Damascus, Jerusalem and Cairo15 (Figure 5.10). Its use in Najm would have been quite anachronistic. Third, these medieval citadels were rather heavily inscribed, both on their exterior and interior. Although many of these inscriptions, including the long frieze on the entrance block of the Aleppo citadel, are Mamluk, the practice is well attested from the early Ayyubid period, in Aleppo, Damascus, Najm, elsewhere. Qalʾat Najm, for example, has three inscriptions from GhåzÈ’s reign and a fourth gigantic inscription was found in the recent excavations and restorations (Figure 5.11). Assuming that this inscription also dates to GhåzÈ’s reign, it would have pre-dated the well-known inscription of al-Ashraf KhalÈl in Aleppo by about eighty years. This might suggest the existence of an earlier Ayyubid inscription in the Aleppo citadel. The propagandistic intent of these gigantic inscriptions is, I think, quite self-evident.16 Internally, most of these castles are focused around their palacebath complex, our fourth common feature. In addition to the wellknown palace in Aleppo (Figure 5.12), there are comparable palaces in Najm (Figure 5.13), ÍahyËn, al-Raqqa, Diyarbakr, Bosra, Karak, Shawbak and perhaps Óårim.17 All are four-iwån structures with a decorative pool in the middle, and three – Aleppo, Diyarbakir, and ÍahyËn – have Shadirwan-type fountains.18 Although much smaller, the plan and sense of luxury in these palaces harken back to their ʿAbbasid prototypes of the ninth century. Thus, instead of a keep bristling with embrasures, the last resort of a besieged Frankish castle, Islamic castles contain a luxurious palace and bath. The difference is quite striking, and it once again underlines the dynastic rather than militaristic nature of many Islamic castles.

99

100

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

The luxuriousness of these palaces is quite literally announced in their elaborate portals, of which two – Aleppo and ÍahyËn –19 have survived from the Ayyubid period (Figure 5.14). Although undated, the portal – indeed the entire palace of ÍahyËn – cannot be any earlier than the first half of the thirteenth century, for it closely resembles the portal of the palace in the Aleppo citadel, which is datable to c. 1200. Similarly, its interior tripartite facades are closely related to Aleppo and to several other Ayyubid palaces. Fifth and last, these castles contained mosques, some of substantial size and at least two with a minaret. The largest of these is that built by GhåzÈ at the highest point of the Aleppo citadel with an especially tall and somewhat archaic minaret. More interesting by virtue of its rarity is the mosque at Qalʾat Najm, whose twin arcades provide a magnificent view of the Euphrates (Figure 5.15). I know of only one other citadel – the Alhambra – that has a mosque with an exterior facing arcades.20 Although the two are quite unrelated, these mosques speak of a life of some luxury and sophistication. I have attempted in this piece to give a brief discussion of the significance of the geographic distribution and internal composition of the castles built by al-Ûåhir GhåzÈ.21 The garland of castles and fortifications built or rebuilt by GhåzÈ around and within Aleppo defended the borders, acknowledged Ayyubid suzerainty, facilitated trade, kept the nomadic tribes at bay and created a suitable abode for the sultan and his various governors. The effectiveness of this defensive system promoted the prosperity of Aleppo and ensured the survival of its Ayyubid rulers for another half century. But whereas this system was flexible enough to absorb various skirmishes and encroachments, it could not withstand the severe and unexpected blow of the Mongol invasion, which not only destroyed Aleppo and its surrounding fortresses but also ended the ʿAbbasid Caliphate. Lacking this ultimate safeguard and sanction, the succeeding Mamluks chose to look elsewhere for their palatial architecture, which departs from established ʿAbbasid types and introduces types whose models and origins require Figure 5.3  Map: al-Zahir GhåzÈ’s further discussion. Domain

101

defending ayyubid aleppo

Figure 5.5  Qalʾat al-Mu∂Èq (near Afamiya/Apamaea)

Figure 5.4  Map: al-Zahir GhåzÈ’s Domain, maximum extent (both after Humphreys)

Figure 5.6  Citadel of Aleppo: model

Figure 5.7  Citadel of Aleppo: entrance block and glacis (photo: Yasser Tabbaa)

Figure 5.9  Qalʾat Najm (photo: Yasser Tabbaa)

Figure 5.8  Citadel of Aleppo: machicolation (photo: Yasser Tabbaa)

102

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 5.10  (left) Citadel of Aleppo: plan of entrance block (after Herzfeld); (right) Qalʾat Najm: plan of entrance block

Figure 5.11  Citadel of Aleppo: plan of Ayyubid palace

Figure 5.12  Qalʾat Najm: plan of Ayyubid palace

Figure 5.13  Qalʾat Ía˙yËn: portal of Ayyubid palace (photo: Yasser Tabbaa)

defending ayyubid aleppo

Figure 5.14  Qalʾat Najm: fragment from large Ayyubid inscription (photo: Yasser Tabbaa)

Figure 5.15  Qalʾat Najm: mosque (photo: Yasser Tabbaa)

Notes 1. Ibn KhaldËn, Al-Muqaddima. See also the excellent analysis of this idea in Philippe Gourdin, ‘Les fortifications du Maghreb d’après les sources écrites: la vision d’Ibn KhaldËn’, in Rika Gyselen (ed.), Sites et monuments disparus d’après les témoignages des voyageurs, Res Orientalis, vol. VIII, Groupe pour l’Étude de la Civilisation du Moyen-Orient: Bure-sur-Yvette (1996), 27ff. 2. Kamål al-DÈn Ibn al-’AdÈm, Zubdat al-halab fi ta’rÈkh Óalab, 3 vols, ed. Sami Dahhan, Damascus (1951–68). 3. For a history of Aleppo under the Ayyubids see R. Stephen Humphreys, From Saladin to the Mongols: The Ayyubids of Damascus, 1193–1260

103

104

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Albany (1977); or more recently Anne-Marie Eddé, La Principauté Ayyoubide d’Alep (579/1183–658/1260) (1999). 4. Ibn Shaddåd, Al-A’låq al-kha†Èra fi dhikr umarå’ al-Shåm wa’l-JazÈra. Ta’rÈkh Óalab, ed. Dominique Sourdel, Damascus (1963), 1: 24 and 29. 5. For the overall transformation of Islamic cities into fortified citadel cities, see in particular Jere Bacharach, ‘Administrative Complexes, Palaces, and Citadels: Changes in the Loci of Muslim Rule’, in Irene Bierman, Rifaat Abou-El-Haj and Donald Preziosi, eds, The Ottoman City and Its Parts, New Rochelle, NY (1991): 111–28; and Yasser Tabbaa, Constructions of Power and Piety in Medieval Aleppo, Philadelphia (1997), ch. 3. 6. The concept of height and dominance was frequently celebrated by contemporary poets and even by later illustrators of the citadel of Aleppo and citadels generally. See the brief remarks in Tabbaa, Aleppo, pp. 53–5. 7. Ibid., 62–6. See also Nasser Rabbat, ‘The Ideological Significance of the Dar al-’Adl in the Medieval Islamic Orient’, International Journal of Middle East Studies 27 (1995), 3–28. 8. Yasser Tabbaa, ‘Circles of Power: Palace, Citadel, and City in Ayyubid Aleppo’, Pre-Modern Islamic Palaces, ed. Gulru Necipoglu, in Ars Orientalis 23 (1993), 181–200. 9. See, in particular, Eddé, La Principauté Ayyoubide D’Alep. 10. Gaston Wiet, ‘Une inscription de al-Malik Zahir GhåzÈ a Latakieh’, Bulletin de l’Institut Français d’Archéologie Orientale 30 (1931): 273–93. 11. Humphreys, Ayyubids of Damascus, 118–19. 12. Wiet, ‘Inscription’. 13. Humphreys, Ayyubids of Damascus, 98–103. 14. Peter Pentz (ed.), Hama IV: The Medieval Citadel and its Architecture, 2 vols, Aarhus, Monographs of the National Museum (1998). 15. Hugh Kennedy, Crusader Castles, Cambridge (1994). 16. For a discussion of the propagandistic aspects of these inscriptions, see Tabbaa, Aleppo, 73–6. 17. See Tabbaa, Aleppo, ch. 4 for a thorough discussion of these palaces. 18. For a discussion of this type of fountain, see Yasser Tabbaa, ‘Typology and Hydraulics in the Medieval Islamic Garden’, in John Dixon Hunt (ed.), Garden and Landscape History: Issues, Approaches, Methods, Dumbarton Oaks (1992), 303–29. 19. Qalʾat ÍahyËn has been renamed for political reasons after Íalå˙ al-DÈn. This should not, however, lead us to conclude that Íalå˙ al-DÈn actually rebuilt it, for it was quite likely rebuilt by GhåzÈ, who was certainly behind its very interesting palace, now called a ˙ammåm. 20. This is the mosque adjacent to the Cuarto Dorado, frequently illustrated on all books on the Alhambra palace. 21. This brief discussion of some medieval Syrian fortresses should not stand for an intensive archaeological investigation and detailed monographic analysis of each fortress, something that is already taking place. Specifically, Qalʾat Najm alone, with its unique and unexplained third storey and well-preserved kitchen facilities, is likely to reveal important secrets about Syrian fortifications.

CHAPTER SIX

Originality and Innovation in Syrian Woodwork of the Twelfth and Thirteenth Centuries Introduction When viewing the corpus of Islamic woodwork produced in the twelfth and thirteenth centuries, one is immediately struck by four main trends. First is quantity: more has survived from this period than from previous periods, a trend that continues unabated into the fourteenth century. Second is the prevalence of geometry: whereas woodwork made in previous centuries is generally vegetal or calligraphic in style, the woodwork objects produced in the twelfth and thirteenth centuries are decidedly geometric. Third is the dominance of Syria: many of the great wood masterpieces of this period were made in Syria, including objects made by Syrian craftsmen outside Syria. Fourth, there are indications – both in view of the increased frequency of craftsmen’s signatures and the inclusion of some of their names in biographical dictionaries – that at least some woodworkers in this period began to enjoy an increased status.1 In this piece I would like to examine the aesthetic and historical foundations of these trends by looking at a coherent group of wood objects made in Syria or by Syrian craftsmen in the twelfth and thirteenth centuries. Following a brief introduction in which I assess the status of the woodworker in this period and a methodological statement on the study of Islamic ornament, I will discuss these objects, highlighting and analysing in particular their innovative geometric patterns. I will conclude by exploring the factors and motives that may have contributed to the efflorescence of the geometric mode in Syria in the twelfth century and to its ultimate spread to other regions, in part due to the influence of Syrian craftsmen. As with all artisans in the Islamic world, the status of Islamic woodworkers (sing. najjår) was perhaps on par with that of shop­ keepers,2 and well below that of jurists or merchants, although it Yasser Tabbaa (2014), ‘Originality and Innovation in Syrian Woodwork of the 12th and 13th Centuries’, in Katia Cytryn-Silverman and Daniela Talmon (eds), Material Evidence and Narrative Sources: Interdisciplinary Studies of the History of Islamic Societies, Leiden: E. J. Brill, 2014, x.

106

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

seems to have improved in the period under discussion. One indication of the improved status of woodworkers in this period is the increased tendency they showed to sign their works: nearly all large objects made in this period bear a signature, and one has as many as four signatures. These artisanal signatures should not be understood as an indication of an elevation of status from artisan to artist – which rarely happened in Islamic art – but rather as a sign of pride in workmanship and, from the patron’s point of view, something like a certificate of quality. Other crafts in the twelfth and thirteenth centuries, including metalwork and stone masonry, are also marked by an increased occurrence of artisan’s signatures, but woodworkers stand apart from these others by dint of their occasional mention in historical sources.3 Miriam Kühn argues in this volume that this trend in fact became quite common under the MamlËks, but it may have started in the twelfth century, when at least one famous woodworker was mentioned in a biographical dictionary. Nasser Rabbat has argued that the inclusion of short biographies of some woodworkers in biographical dictionaries may have been due to their association with important MamlËk princes for whom they had done works.4 While this is entirely likely, there are two additional reasons that may have led to woodworkers being seen as more worthy of being commemorated in this manner. Some woodworkers, and especially those involved in design, were trained in practical geometry, which suggests that they were literate and therefore able to follow abstract designs on paper.5 Perhaps more important is that woodworkers, unlike potters or metalworkers, for example, were responsible for a variety of religious or ritual objects, including doors, minbars, mi˙råbs, ra˙las (folding stands for reading the Qurʾån), cenotaphs and banisters around tombs. This association with important religious objects and the monuments that housed them may have been just as important to the status of woodworkers as their association with a dignitary or a patrician. In providing an interpretation of the flourishing of the geometric mode in the twelfth and thirteenth centuries, I do not seek to assign specific symbolic meanings to geometric patterns but only to suggest some possibilities of purpose and intentionality. Meaning, and even intentionality, in Islamic ornament has been the subject of a rather vigorous debate that was, until recently, generally divided into two camps: archaeologists, following a positivist method, and the so-called ‘perennialists’, who adopted an essentialist approach to the study of Islamic culture. The archaeological approach mainly focused on documentation, taxonomy and chronological development and largely dismissed the communicative and semiotic dimensions of Islamic ornament.6 The resulting lacunae of interpretation were filled in the 1970s and 1980s by a group of artists and religious scholars, who generally viewed Islamic ornament as a reflection of

syrian woodwork

perennial spiritual values and concepts in Islam, in particular the concept of taw˙Èd.7 More recently, scholars have attempted to chart a middle course between the extremes of positivism and essentialism by proposing various historical, psychological and semiotic strategies that attempt to locate (or reject) meaning in Islamic ornament within accepted art historical parameters.8 To summarise, first, these scholars broadly agree that Islamic ornament underwent significant changes during its thirteen centuries of development, changes that reflected regional variation, historical development, and even the limitations of varied media. Second, some agree that these changes were sometimes quite abrupt, as in the eleventh–twelfth centuries in the central Islamic world and the sixteenth century in the Ottoman Empire, leading some scholars to postulate a degree of intentionality. Third, and perhaps most controversial, these writers, including the present one, propose that these rather abrupt shifts were not simply natural and internal developments but rather were self-conscious transformations motivated by combinations of political, religious and technical factors.9 Bloom, who in his earlier work on Fatimid architecture had proposed a measure of intentionality and meaning in the architecture and ornament of the period, has recently retracted to a more positivist position – as, for example, in his chapter in the recently published AyyËbid Jerusalem, in which he grapples with much of the same material as is presented below.10 By focusing on Syrian geometric woodwork of the twelfth and thirteenth centuries, this piece hopes to fill an important gap in our understanding of the development of geometric ornament, generally. In fact, I believe that only through close examination of these Syrian developments can we come to better understand the quite significant differences that separate, for instance, eleventh-century Iranian brick ornament and the magnificent stone geometric ornament of Anatolian monuments of the thirteenth century. Equally, Syrian woodwork of this period, I would argue, was central to the development of Egyptian geometric designs, which began under the Fatimids and gathered momentum after the rise to prominence of the Ayyubids. Finally, in view of the sudden nature of the development of an original style in geometric woodwork, the piece concludes by proposing the means and motives behind this innovation and its transmission to other regions. Fatimid woodwork (c. 1050–c. 1150) Syrian woodwork of the twelfth and thirteenth centuries can be divided historically – and, to some extent, also stylistically – into two periods: Fatimid (c. 1050–c. 1150) and Zangid–Ayyubid (c.1150– c. 1260). The first period is represented in Syria by very few examples, which makes it necessary to include in the present ­discussion

107

108

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

some Egyptian Fatimid works. The two main Syrian pieces are a ‘banister’ in the Damascus Museum, previously part of a funerary enclosure, and the cenotaph of Sayyida Få†ima, in the Båb ÍaghÈr cemetery. The banister, which may originally have been part of an enclosure surrounding a cenotaph, once belonged to a sanctuary called Mußallå al-ʿÏdayn. Dated by inscription to 497/1103–4, it has no geometFigure 6.1  Fatimid banister, twelfth ric patterns but is instead covered in quite century. Damascus Museum (photo: elaborate vegetal patterns that seem to stand Yasser Tabbaa) midway between the bevelled style (an ornamental style first seen in Såmarråʾ that consisted of convex and dense floral patterns with minimal background) and fully developed arabesque (Figure 6.1). Its floriated KËfic inscriptions also place it stylistically within the Fatimid period. The cenotaph of Få†ima, which now is barely visible beneath layers of tapestries and dirty glass, lacks both geometric and vegetal patterns and bears an elaborate floriated KËfic inscription of the Verse of the Throne (Q 2: 255) (Figure 6.2). These two examples are inadequate to the task of giving us a complete picture of Fatimid woodwork, which is much better represented in Cairo. Most commonly, Egyptian Fatimid woodwork dated to before the twelfth century consisted of mortised panels largely decorated with variations on the bevelled style, either completely floral or mixed with human and animal figures in a kind of a rinceau animé.11 A group of three late Fatimid objects, however – comprising a wooden mi˙råb for each of the shrines of Sayyida NafÈsa (533–40/1138–45) and Sayyida Ruqayya (549–55/1154–60) and the minbar of Ashkelon – are adorned with geometric patterns.12 Dating to 484/1092, the little-studied but quite astonishing minbar of Ashkelon, now in Hebron, has flanks that bear hexagonal ornaments with arabesque fillets. If these flanks are part of the original minbar, they would pre-date any existing geometric ornament in

Figure 6.2  Damascus: cenotaph of Sayyida Få†ima, eleventh century, in floriated KËfic (photo: Yasser Tabbaa)

syrian woodwork

109

Syria by about a half century, and raise questions about their source and origin.13 In view of the Syrian-Armenian origin of the patron, Badr al-JamålÈ, one may wonder whether he relied on Syrian craftsmen for its execution, as he had done a decade earlier when he commissioned work to be done on the gates of Cairo.14 What remains unexplained, however, is how this early geometric ornament appeared so suddenly and in a prominent pulpit, and especially why it did not lead directly to further developments in Egyptian woodwork. Curiously, the geometric patterns adorning two movable wooden mi˙råbs, of Sayyida NafÈsa and Sayyida Ruqayya, are somewhat less developed than those on the Ashkelon minbar (Figure 6.3). More in keeping with late Figure 6.3  Wooden Fatimid design, the ornament on these mi˙råbs is more mi˙råb from shrine of striking for its crisp vegetal patterns than for its geometric Sayyida NafÈsa, frames. The fact that they postdate the Ashkelon minbar 540/1145. Museum of by about half a century either reinforces the uniqueness of Islamic Art, Cairo the Ashkelon minbar or perhaps casts doubt on the date (photo: Yasser of its flanks. What seems more likely, however, is that Tabbaa) the two styles of woodwork coexisted in the late Fatimid period: mortised panels with vegetal ornament alongside a style of increasingly more developed geometric patterns.15 In other words, in Egypt, the geometric style gradually and unevenly supplanted the earlier ‘mortised panel’ style, whereas in Syria, as we shall see below, the change was somewhat more abrupt. Zangid woodwork, c. 544–69/1150–74 The development of geometric woodwork in Syria in the second half of the twelfth century is much more straightforward, with none of the overlapping of styles as seem to have been the case in Fatimid Egypt. Some of the finest pieces were produced during the reign of NËr al-DÈn (540–69/1146–74), an important patron of religious architecture and of mosque furnishings, including minbars and mi˙råbs. Being the first ruler in several centuries to preside over Damascus and Aleppo and to extend his political influence to Mosul and Egypt as well, his patronage resonated widely in the Islamic world.16 The many monuments he commissioned in Aleppo and Damascus display a number of innovations – including muqarnas vaulting, cursive monumental inscriptions and ­geometric ­ornament – that some have linked with the ʿAbbåsid capital.17 Since little architecture or woodwork has survived in Baghdad that dates to between the tenth and the thirteenth centuries, it is nearly impossible to trace the development of geometric ornament in what may have been its generative centre. That makes the monuments erected elsewhere by NËr al-DÈn, which exhibit significant developments in

110

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

arabesque and geometric ornament, all the more important. These monuments include the doors of the BÈmåristån al-NËrÈ in Damascus (549/1154), the wooden mi˙råb and windows previously at the Maqåm IbråhÈm at the Aleppo citadel (1167 and c. 1200), the wooden minbar at Hama (1163) and the minbar designed for the Aqßå mosque in Jerusalem (1168, 1176).18 Furthermore, this tradition continues uninterrupted under the Ayyubids, and is continuous with Ayyubid woodwork in Egypt, as well.19 The large doors at the BÈmåristån al-NËrÈ in Damascus are especially important since they fronted the leading medical institution in Damascus in the medieval period and constituted one of NËr al-DÈn’s main acts of patronage. They are also significant as the earliest preserved doors with comFigure 6.4  Portal of BÈmåristån plete geometric ornament and for the fact al-NËrÈ, Damascus, 549/1154 (photo: Yasser Tabbaa) that the identity of their maker is known. The double door is made of wood, sheathed in bronze affixed with brass nails to form an overall geometric pattern.20 The design, which covers the entire door except for a narrow inscriptional frieze at the top, is a fully developed star pattern that is based on a triangular grid. Its primary unit is a sixpointed star inscribed within a hexagon, which is surrounded by six five-pointed stars whose external sides form a larger hexagon. Five such units are used – one in the middle of each leaf and one in the exact centre of the door, with half on each Figure 6.5  Drawing for a door in leaf (Figure 6.4). When the door is closed, Cizre, from JazarÈ, al-Jamiʿ fÈ maʿrifat the design on both leaves can be read as al-hiyal al-Handasiyya a single composition focused on the large star in the middle of the door. The logic and originality of this geometric design testify to the training of its maker, al-muhandis (the geometer) Muʾayyad al-DÈn AbËʾl-Fa∂l b.ʿAbd al-KarÈm Mu˙ammad al-ÓårithÈ (d. 599/1202–3). According to the biographer of physicians Ibn AbÈ Ußaybiʿa, al-ÓårithÈ made this door and the others that once existed in the bÈmåristån, possibly including the window grilles. He was known as a carpenter, stone-mason, and geometer or engineer (muhandis) who had studied Euclid and the Almagest in order to excel in his crafts. He was also

syrian woodwork

said to be versed in mathematics, geometry, astronomy, medicine and even hadith.21 In other words, he was both an artisan and a man of letters and learning, a combination that, although questioned by most writers on Islamic art, may have existed in medieval Islam.22 The now-lost mi˙råb at Maqåm IbråhÈm in the Aleppo citadel was one of the great masterpieces of Islamic geometric woodwork. Rising to a height of about three metres, it consisted of a deep niche covered by a hemispherical conch and flanked by a wide frame of wooden marquetry (Figure 6.6). The complex geometric strapwork encloses delicate vegetal arabesque fillets and a few passages of floriated KËfic inscriptions that give the name of the carpenter, MaʿålÈ b. Sålim, and the date of its completion, 563/1167–8.23 The design makes ingenious use of three different grids: a triangular grid in the niche; a square grid for the flanking frames; and an interesting radial grid with a pentagonal star for the conch, constructed so as to fit perfectly within the hemispherical surface. In both the niche and the flanking frames, the ornament comprises several interlocking geometric forms that create complex geometric patterns, known locally as ‘tafßÈl makhbˆ’, or knotted design, or perhaps as girih, a Persian artisanal term indicating geometric interlaces.24 A quite ingenious pattern is used in the niche, whose triangular grid can simultaneously be read diagonally – as diamonds enclosing six-pointed stars – and vertically, as a series of staggered squares that enclose smaller lozenges. The maqåm also once contained wooden windows, datable to the reign of al-Ûåhir GhåzÈ (581–612/1186–1216); their astonishing patterns will be discussed later (Figure 6.7). The minbar at the NËrÈ Mosque in Hama, dated 563/1168, has long since lost its stairs and side panels, which would have certainly carried geometric designs. The upper structure, which is entirely original, consists of the usual chair with three arched openings and a backrest, all crowned by an elaborate entablature that surrounds a small dome. Only the backrest contains a geometric pattern, a simple design of sixpointed stars that encloses the shahåda. The highlight of this minbar is the vegetal arabesque patterns that decorate the three open arches, and the frieze and cornice above them (Figure 6.8). The minbar commissioned by NËr al-DÈn for the Aqßå mosque in Jerusalem marks the peak of creativity of the Aleppo school of woodcarvers in the twelfth century. The minbar is dated twice, to 564/1168 and 1176, suggesting that it was begun by NËr al-DÈn Figure 6.6  Aleppo, mi˙råb but perhaps not completed until the brief previously at Maqåm IbråhÈm, reign of his son, al-Íali˙ IsmåʿÈl.25 There is 563/1167

111

112

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 6.7  Analysis of geometric pattern on windows previously at Maqåm IbråhÈm, c. 1200

Figure 6.8  Minbar at NËrÈ Mosque in Hama, dated 563/1168. Detail and analysis of pattern on the backrest (photo: Yasser Tabbaa)

no doubt, however, that NËr al-DÈn was the motivating force behind its construction and that it was designed to fulfil the dream of liberating Jerusalem with which he had been preoccupied since the beginning of his reign in 1146.26 As is well known, NËr al-DÈn died with his dream unfulfilled, and the completed minbar was moved to Jerusalem by Íalå˙ al-DÈn in 1187. It was during its brief stay at the Great Mosque of Aleppo that the minbar was seen and greatly admired by the Andalusian traveller Ibn Jubayr, who commented: The art of qarnaßa (ornamental carving) had exhausted itself in its endeavors on the pulpit, for never have I seen a pulpit like it or of such wondrous workmanship. The woodwork stretches from it to the mi˙råb, beautifully adorning all its sides in the same marvelous fashion.27 In 1969, the minbar was burnt by an arsonist and a replica was recently completed in Amman and placed once again in the Aqßå Mosque, further indicating its great historical significance and contemporary resonance. The minbar was signed by four different artisans – Óåmid b. Ûåfir al-ÓalabÈ,28 AbË ʾl-Óasan b. ʿAlÈ, Salmån b. MaʿålÈ and Fa∂åʾil b. Ya˙yå – all from Aleppo. The third of these woodcarvers was most likely the son of MaʿålÈ b. Sålim, who had made the mi˙råb at the Aleppo citadel in 1168. Others from the same family were also responsible for important works in Cairo and perhaps elsewhere, as we shall see below (Figure 6.9). The four artisans who worked on Figure 6.9  Minbar of NËr al-DÈn this minbar and their six recorded signatures (1168–76). Side panel with further attest to its significance and the pride signature of one of the artisans: taken in its creation. Óåmid b. Ûåfir al-ÓalabÈ. Radial No less than twenty-five different geometric design with ten-pointed star

113

syrian woodwork

Figure 6.10  Minbar of NËr al-DÈn. Detail and analysis of pattern on side panel

Figure 6.11  Minbar of NËr al-DÈn. Detail and analysis of inside of door

patterns appear in the minbar, in addition to vegetal arabesques, openwork, muqarnas and inscriptions. Its patterns were based on stars with five, six, eight and twelve points, in an encyclopedic effort to represent nearly every pattern known at the time (Figures 6.10–6.11). Though lacking the overall unity and harmony that characterise minbars of the MamlËk period, this example expresses an unprecedented richness of patterns that had not been previously combined in one object.29 Ayyubid woodwork, c. 1180–1260: transmission to Egypt and Anatolia The excellence of the woodcarvers of Aleppo in the twelfth century is further confirmed by the fact that at least one of them practised his craft in Cairo. Two commemorative caskets commissioned by Íalå˙ al-DÈn, those of Imåm al-Óusayn and Imåm al-ShåfiʿÈ (dated 574/1178),30 are signed by ʿUbayd b. MaʿålÈ, who must be the son of MaʿålÈ b. Sålim, the maker of the Aleppo mi˙råb and therefore a brother of Salmån b. MaʿålÈ, the chief artisan of the minbar of Jerusalem. It seems likely that ʿUbayd was invited by Íalå˙ al-DÈn to move to Cairo in the 1170s, where he continued to practise the family craft under the patronage of Íalå˙ al-DÈn and that of the later Ayyubids. Both caskets are carved on all four vertical faces in bold geometric patterns framed by inscriptional bands and enclosing rich vegetal arabesque (Figures 6.12–6.13). Although the casket of al-Óusayn is certainly a Shiʿi monument, I believe, contrary to Williams and Bloom, Íalå˙ al-DÈn commissioned it and thus that it represents one of several known acts of SunnÈ patronage of a Shiʿi shrine.31 On the basis of these two early Ayyubid examples and the subsequent proliferation of geometric woodwork in Egypt, we are led to conclude that, although some geometric woodwork was known in Fatimid Egypt, it only became widespread under the Ayyubids, who also introduced muqarnas vaulting and cursive monumental inscriptions to the visual vocabulary in Egypt.32

114

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 6.13  Casket of Imåm al-ShåfiʿÈ, Cairo, 574/1178 (photo: Yasser Tabbaa)

Figure 6.12  Casket of Imåm al-Óusayn, Cairo, 1170–80 (photo: Yasser Tabbaa)

Figure 6.14  Minbar at Jåmiʿ al-Óanåbila in Íåli˙iyya, Damascus. Foundation inscription, dated 604/1207, ‘bi-tawallÈ Ma˙åsin b. Nuʿmån b. AbÈ Mu˙ammad al-QalånisÈ’ (photo: Yasser Tabbaa)

Further proof of the vigour and potential influence of Syrian woodworkers can be seen in one important object, previously at the Maqåm IbråhÈm in the Aleppo citadel and datable to c. 1200. This was a double door to a wall cabinet, completely geometric and without any arabesque fillings (Figure 6.7). Herzfeld pronounced it to be ‘the most complicated design ever produced by that branch of art’, adding that the nearly impossible design consists of eleven-pointed stars that are staggered between rows of twelve-pointed and ten-pointed stars.33 Standing at the peak of Syrian wooden geometric ornament, such experiments may have contributed to stone geometric ornament in Anatolia, which postdates these by about one generation, and which would eventually exceed them in their complexity and execution.34 There are three more specimens of Ayyubid geometric woodwork that warrant mention: a minbar, a mi˙råb and a cenotaph. The minbar, which remains in situ at the Jåmiʿ al-Óanåbila in Íåli˙iyya, Damascus, was commissioned by the ruler of Irbil, MuΩaffar al-DÈn GökburÈ, in 604/1207 (Figure 6.14). The side panels, defaced by oil paint, are made of a very unusual geometric pattern consisting of large octagons that enclose four-pointed stars, with equally large

syrian woodwork

115

cross-shaped designs in between the octagons. The background is filled with arabesque, recalling the styles of earlier decades (Figure 6.15). The second piece is a magnificent mi˙råb in the Madrasa al-Óallåwiyya in Aleppo, made in 643/1245–6 during the reign of Sultan Íalå˙ al-DÈn YËsuf II. As is typical of medieval Syrian woodwork, it combines geometric patterns, delicate arabesque fillets and astonishing Figure 6.15  Minbar at Jåmiʿ calligraphy. A splendid inscription surrounds al-Óanåbila in Damascus. Side the mi˙råb on three sides and gives the name panel and analysis of pattern of the commissioner (bi-tawallÈ), ʿUmar b. (photo: Yasser Tabbaa) A˙mad b. Hibatallåh b. Mu˙ammad b. AbÈ Jaråda, who is identified as none other than the great Aleppan historian Ibn al-ʿAdÈm. The mi˙råb bears two other signatures – that of the artisan, AbË ʾl-Óusayn Mu˙ammad al-ÓarrånÈ, and the carpenter, ʿAbdallåh b. A˙mad al-Najjår. As was the case with the al-Aqßå minbar, this mi˙råb makes a distinction between craftsman (ßåniʿ) and carpenter (najjår), wherein the former may have been respons­ible for the design and the latter the actual making. Larger and bolder but less subtle than its lost predecessor at the Maqåm IbråhÈm, this mi˙råb consists of a broad frame that encloses a deep niche surmounted by a semi-circular conch (Figure 6.16).35 The side panels, which are based on a clearly visible square grid, nevertheless contain alternating six-pointed stars with twelve-pointed rays and four-pointed stars within squares surrounded by large octagons (Figure 6.17). The niche, also based on a square grid, is adorned with eight-pointed stars surrounded by eight staggered five-pointed stars. For the hemispherical hood, the artisan ingenuously adapted two interlocked nine-pointed stars, reminiscent of the treatment of the mi˙råb at Maqåm IbråhÈm in the Aleppo citadel. The last piece to consider is a casket, currently on display in the Damascus Museum, which is undated but was commissioned by Fakhr al-KhawåtÈn (Pride of Ladies), wife of the RËm Seljuq Sultan Qillij Arslån IV (r. 1249–67). In pristine condition, it is completely covered by a vigorous, if somewhat repetitive, radial pattern based on a ten-pointed star. Originally made for the Madrasa al-Måridåniyya in Damascus, this casket further attests to the widespread fame of Syrian woodworkers (Figures 6.20–6.21). Conclusions Three main conclusions can be derived from this survey and analysis of the surviving woodwork objects made in Syria or by Syrian craftsmen in the twelfth and thirteenth centuries (Figure 6.24). First, most such pieces bear the signatures of their artisans, from which we have been able to identify seven artisans responsible for at least

116

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 6.16  Aleppo. Madrasa al-Hallåwiyya, mi˙råb dated 634/1237 (photo: Yasser Tabbaa)

syrian woodwork

Figure 6.17  Mi˙råb in the Madrasa al-Óallåwiyya. Detail and analysis of geometric pattern at side panel (photo: Yasser Tabbaa)

Figure 6.18  Mi˙råb in the Madrasa al-Óallåwiyya. Detail and analysis of geometric pattern in the hood (photo: Yasser Tabbaa)

Figure 6.19  Detail of mi˙råb in Madrasa al-Hallawiyya, 634/1237 (photo: Yasser Tabbaa)

117

118

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 6.20  Casket of Fakhr al-KhawåtÈn, wife of Sultan Qillij Arslån IV (r. 1249–67). Damascus Museum (photo: Yasser Tabbaa)

Figure 6.21  Casket for Fakhr al-KhawåtÈn. Analysis of geometric pattern (photo: Yasser Tabbaa)

ten objects. This is quite a departure from the earlier anonymity of Islamic woodworkers and may point to the increasing importance of woodworkers applying geometry to their craft. Second, many outstanding pieces of geometric woodwork discussed above were commissioned by NËr al-DÈn, whose reign also coincides with the introduction of muqarnas vaulting and of cursive monumental inscriptions, most likely from Baghdåd, and their subsequent rationalisation in Syria.36 Though long noted by Herzfeld, Syria’s role in the development of the geometric and vegetal arabesque has been virtually ignored by most writers, who have adopted a somewhat mechanistic mode for the transmission of artistic forms, without sufficiently appreciating the unprecedented importance and centrifugal effect of the patronage of NËr al-DÈn. Indeed, the spread

syrian woodwork

Figure 6.22  Syrian woodworkers in the twelfth–thirteenth centuries (table)

of the girih mode to Egypt after the AyyËbid takeover owes a great deal to the pivotal period of the reigns of NËr al-DÈn and his immediate successor, Íalå˙ al-DÈn. Third, the relative suddenness of the spread of geometric ornamentation in Syria and elsewhere in the twelfth century leads us to question whether these geometric interlaces had any cultural or symbolic associations. Did they convey any meanings beyond the normative associations generally supported by ornament, including emphasis or dissolution of forms, luxury, or just plain decoration? Overall, geometric interlaces seem to lack the intentionality of use and iconographic associations that could be proposed for changes in calligraphic styles or even the rise of muqarnas vaulting. The very ubiquity of geometric patterns and their use in myriad contexts also seem to undermine the associating of them with any specific symbolic associations. What argues in favour of some signification or intentionality, however, is the appearance and significant development of geometric interlaces in monuments whose patrons are linked with the SunnÈ revival, the sectarian movement that began in Baghdåd in the

119

120

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

eleventh century and was wholeheartedly embraced by NËr al-DÈn and others in the twelfth century.37 Progressing from Seljuq Iran, to Zangid and Ayyubid Syria, and to Almoravid Morocco, we witness stridently SunnÈ dynasties whose religious structures and furnishings display prodigious developments in geometric ornament.38 Fatimid Egypt, where we have seen limited but significant developments in geometric patterns, nevertheless continues to demonstrate a preference for contained decorative friezes and mortised floral ornaments well into the twelfth century. The full development and spread of the girih mode really only takes place under the Ayyubids, not reaching its peak until the MamlËks. Also supporting the possibility of meaning in the geometric mode is that, particularly in its earlier examples, it is generally used with a sense of decorum, a studied sense of applying the appropriate ornamentation to objects or architectural forms.39 We note, for example, that geometric ornament was first applied to objects of cultic or symbolic value – including Qurʾån frontispieces, portals, domes, minbars, mi˙råbs and cenotaphs – before it eventually spread to nearly every type of object and monument. The specimens discussed above demonstrate that, whereas vegetal, geometric or even muqarnas ornament often coexisted in one and the same object or monument, they were used with a sense of order and decorum that accounted for place, context and function. Thus, geometric ornament, with or without vegetal arabesque fillets, was most effectively used for doors and door frames, minbars, minarets and, more rarely, domes. Combining the purity and austerity of geometric principles with the celestial allusions of star-patterns, geometric ornament may have reflected the ordered universe, whose atomistic structure was created and sustained by divine intervention. Likewise, the strength and vigour of geometric ornament would have enhanced the image of power and authority that minbars were intended to project, while also calling attention to their founders. Thus, although there is in fact very little contemporary evidence that specifies the meaning of geometric ornament, there are various indications in patronage, production and function that geometric ornament was used in this period with some sense of purpose and intentionality. These specific domains constituted the creative loci of the geometric interlace well into the fourteenth century. The intimate association of the geometric interlace with religiously significant objects clearly worked in both directions: it reaffirmed the symbolic potency of these objects, while also enhancing the religious dimension of arabesque ornament. The simultaneity of association stands at the heart of an ornamental system that entirely consumes the object it covers.

syrian woodwork

Notes 1. As Miriam Kühn argues in her article in this volume, this was a far more common occurrence in the MamlËk period, although the shift towards the acceptance of such may have started somewhat earlier. 2. The social status of Islamic artisans has been the subject of much debate, with most scholars accepting their comparatively low status and few arguing that they would have enjoyed a somewhat more elevated status. Overall, I agree with L. A. Mayer, in Islamic Metalworkers and their Works (Geneva, 1959), 14–15, in which he suggests that most Islamic artisans had a low status but that there were exceptional cases, mainly among woodworkers and architects, of artisans rising above their prescribed status, ‘by sheer force of intelligence and manual skill’. 3. See also S. S. Blair, ‘A Brief Biography of AbË Zayd’, Muqarnas 25 (2008), esp. 162–3, in which she offers perceptive remarks on the high status of the potter AbË Zayd. 4. N. Rabbat, ‘Architects and Artists in MamlËk Society: The Perspective of the Sources’, Journal of Architectural Education 52 (1998): 30–7. 5. See note 8, below. To these references should be added Ibn KhaldËn, The Muqaddima: An Introduction to History, tr. F. Rosenthal (Princeton, 2005), 321, who wrote: ‘In view of its origin, carpentry needs a good deal of geometry of all kinds. It requires either a general or specialised knowledge of measurement and proportion in order to bring forms from potentiality into actuality in the proper manner.’ 6. Scholars in this group include M. Dimand, ‘Studies in Islamic Ornament, I: Some Aspects of Omaiyad and early ʿAbbasid Ornament’, Ars Islamica 4 (1937): 293–337; E. Kühnel, The Arabesque: Meaning and Transformation of an Ornament, tr. R. Ettinghausen (Graz, 1977); K. A. C. Creswell, in various publications; and, to some extent, E. Herzfeld, MCIA 2. Syrie du Nord: Inscriptions et Monuments d’Alep, 3 vols (Paris, 1954–6), who nevertheless consistently viewed ornament within an historical context. 7. Scholars and artists in this group include K. Critchlow, Islamic Patterns: An Analytical and Cosmological Approach (New York, 1976); T. Burckhardt, Art of Islam: Language and Meaning (London, 1976); and, S. H. Nasr, Islamic Art and Spirituality (Albany, 1987). 8. T. Allen, Five Essays on Islamic Art (Sebastopol, 1988); O. Grabar, The Alhambra (Cambridge, 1978); idem, The Mediation of Ornament (Princeton, 1992); G. Necipo©lu, The Topkapi Scroll: Geometry and Ornament in Islamic Architecture (Santa Monica, 1995); and Y. Tabbaa, The Transformation of Islamic Art during the Sunni Revival (Seattle, 2001). 9. Whereas nearly all scholars to have studied Islamic ornament have noted its transformative nature, and even the sometimes sudden nature of these transformations, fewer is the number to have acknowledged or discussed questions of meaning in Islamic ornament. Those to have done so include Grabar, NecipoÌlu and Tabbaa, as noted above. 10. This disavowal of interpretation is quite clear when comparing Bloom’s early works, including his 1980 dissertation, J. M. Bloom, ‘Meaning in Fatimid Architecture’ (Ph.D. diss., Harvard University, 1980), and idem, ‘The Mosque of al-Óåkim in Cairo’, Muqarnas 1 (1980): 15–36, to his more recent works, including ‘Woodwork in Syria, Palestine and Egypt during the 12th and 13th Centuries’, in Ayyubid Jerusalem: The Holy

121

122

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

City in Context, 1187– 1250, ed. R. Hillenbrand and S. Auld (London, 2009), 129–46; S. S. Blair and J. M. Bloom, Cosmophilia: Islamic Art from the David Collection, Copenhagen (Boston, 2006); and J. M. Bloom, Arts of the City Victorious: Islamic Art and Architecture in Fatimid North Africa and Egypt (New Haven, 2007). See, in particular, R. Hillenbrand’s quite vitriolic statement in his preface to Bloom, Arts, against all ‘excessive’ modes of interpretation (p. x). 11. The standard work on Fatimid woodwork remains C. J. Lamm, ‘Fatimid Woodwork, its Style and Chronology’, Bulletin de l’Institut d’Égypte 18 (1935): 59–91. For a more recent treatment, see A. Contadini, Fatimid Art at the Victoria and Albert Museum (London, 1978), esp. 111–13 and pls 51–2. 12. For a study of these monuments in context, see C. Williams, ‘The Cult of ʿAlid Saints in the Fatimid Monuments of Cairo, Part II: The Mausolea’, Muqarnas 3 (1985): 39–60. 13. I have not personally directly examined this minbar, which seems to have undergone some restoration or rebuilding in later periods. Certainly its doors, with a radial star pattern, can only be MamlËk in their date of construction, although that leaves the date of its flanks open to question. 14. For Badr al-JamålÈ and his patronage in support of the eleventh-century gates of Cairo, the best work remains K. A. C. Creswell, The Muslim Architecture of Egypt, 2 vols (Oxford, 1952–9) 1: 164–5. See also Bloom, Arts, 122–8. 15. This is precisely the conclusion that Bloom reached, in ‘Woodwork in Syria, Palestine and Egypt’, 135. 16. Other than my unpublished dissertation, The Architectural Patronage of NËr al-DÈn (New York University, 1982), there is no work that has considered the patronage of NËr al-DÈn. For a short biography, see N. Elisséeff, ‘NËr al-DÈn Ma˙mËd b. ZankÈ’, EI2 8: 130–5. 17. See, for example, Allen, Five Essays, 56–7. 18. I had previously considered the doors of the bÈmårÈstån of NËr al-DÈn in Aleppo to be original to the foundation of the monument (c. 1150), but they are most likely modern additions. See Tabbaa, Transformation, 87–8. 19. Regrettably, all of the woodwork at the Maqåm IbråhÈm in the Aleppo citadel vanished without a trace in the early twentieth century; the minbar of NËr al-DÈn was destroyed by fire in 1968. 20. This rather simple technique is known from another example dated to the early Ayyubid period (the doors of the Madrasa al-Shådbakhtiyya in Aleppo, dated 1189), and seems to have become the standard way of embossing geometric ornaments on doors in Syria. At least two other more demanding techniques are known: the fixing to the door with nails strips of metal; and the use of large sheets of bronze, which are hammered from behind in the repoussé technique to form outstanding and intricate geometric designs. 21. Ibn AbÈ Ußaybiʿa, ʿUyËn al-anbåʾ fi †abaqåt al-a†ibbåʾ (Beirut, 1966): 669–70; and L. A. Mayer, Islamic Woodcarvers and Their Works (Geneva, 1958), 53. See also Bloom, ‘Woodwork in Syria, Palestine and Egypt’, 132, for a biography of Muʾayyad al-DÈn al-ÓarithÈ, and his eventual dismissal of this training by concluding that even the most complex patterns ‘were probably made by combining set patterns rather than by complex geometric analysis’ (p. 134).

syrian woodwork

22. Cf., for example, Allen’s assertion (Five Essays, 36) that ‘geometry was not necessarily part of a cultured man’s education’. In fact, further evidence from the same period of the fact that Islamic designers and carpenters possessed some basic knowledge of geometry comes from one of the most famous treatises on automata – Al-JazarÈ, Kitåb fÈ maʿrifat al-˙iyal al-handasiyya [The Book of Knowledge of Ingenious Mechanical Devices], tr. D. R. Hill (Dordrecht, 1974) – in which the author shows a drawing of a door that he had designed for the Great Mosque of Cizre (Fig. 8.5). 23. Mayer, Islamic Woodcarvers, 48. 24. Herzfeld, Inscriptions et Monuments d’Alep, 123–34. For the term girih, see Necipo©lu, The Topkapi Scroll, 9. 25. M. van Berchem has noted that the minbar may have been finished by the earlier date since it was immediately moved to the Great Mosque of Aleppo (M. van Berchem, MCIA 2. Syrie du Sud. Tome Premièr, Jérusalem ‘Ville’ (Cairo, 1922), 398–402). See also Mayer, Islamic Woodcarvers, 63. 26. Much has been written on this minbar, including my own ‘Monuments with a Message: Propagation of Jihåd under NËr al-DÈn’, in The Meeting of Two Worlds: Cultural Exchange between East and West during the Period of the Crusades, ed. V. P. Goss (Kalamazoo, 1986), 223–40; and Y. Lev, ‘The Jihåd of sultan NËr al-DÈn of Syria (1146–1174): History and Discourse’, JSAI 35 (2008): 271–2. 27. The Travels of Ibn Jubayr, tr. R. J. C. Broadhurst (London: Jonathan Cape, 1952), 260. 28. This name also has been read as Êåfir, since the dot over the letter is missing. But Êåfir is not attested as an Arab name, whereas Ûåfir (victor) is quite common. 29. A thorough description and analysis of this minbar is provided by S. Auld, ‘The Minbar of NËr al-DÈn in Context’, in AyyËbid Jerusalem, ed. R. Hillenbrand and S. Auld (London, 2009), 72–93. 30. rcea ix, no. 3331: ‘Made by ʿUbayd, the carpenter, known as Ibn MaʿålÈ’. 31. C. Williams, ‘Qurʾånic Inscriptions in the TåbËt al-Óusayn in Cairo’, Islamic Art 2 (1989): 3–14, has argued, on the basis of the Shiʿi implications of its Qurʾånic inscriptions, for a late Fatimid date for the casket of al-Óusayn. However, its cursive inscriptions, and especially the signature of an Aleppan artisan on a closely related piece of woodwork, strongly argue for an Ayyubid date, most likely one during the reign of Íalå˙ al-DÈn. Furthermore, this would not be the only instance of an Ayyubid prince contributing to a Shiʿi shrine; at nearly the same time, al-Ûåhir GhåzÈ in Aleppo aided in the building of the Mashhad alÓusayn. See Y. Tabbaa, Constructions of Power and Piety in Medieval Aleppo (University Park, 1997), 109–14. 32. See Y. Tabbaa, Transformation of Islamic Art (Seattle, 2001), chs 3 and 4. See also L. Korn, Ayyubidische Architektur in Ägypten und Syrien: Bautätigkeit im Kontext von Politik und Gessellschaft 564–658/1169– 1260 (Heidelberg, 2004). 33. E. Herzfeld, ‘Damascus: Studies in Architecture – II’, Ars Islamica 10 (1943), 65. 34. The astonishing development evident in geometric ornament in Anatolia – whether in woodwork, stone carving or tile work – is beyond the scope of this piece. It should be noted, however, that Anatolia possesses a number of extremely important wooden minbars and one

123

124

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

mi˙råb from the twelfth and thirteenth centuries that are quite comparable to Syrian examples. 35. This mi˙råb has been discussed in: Herzfeld, Inscriptions et Monuments d’Alep; D. W. Morray, An AyyËbid Notable and his World: Ibn al-ʿAdÈm and Aleppo as portrayed in his Biographical Dictionary of People associated with the City (Leiden, 1994), 41–4; and Tabbaa, Transformation of Islamic Art, 96–100. See also Tabbaa, Constructions of Power, 134–41, for a discussion of the Madrasa al-ʿAdÈmiyya, located extra muros, just east of Aleppo. 36. Herzfeld has long noted the pivotal importance of NËr al-DÈn. The historical, cultural and artistic significance of NËr al-DÈn also have been discussed by several other writers, including myself – mainly in The Architectural Patronage of NËr al-DÈn and in Transformation of Islamic Art, esp. 68–70, 88–94 and 166–8 – and T. Allen, A Classical Revival in Islamic Art (Wiesbaden, 1989). 37. See G. Makdisi, ‘The SunnÈ Revival’, in Islamic Civilization 950–1150: A Colloquium, ed. D. S. Richards (Oxford, 1973), 155–68. For more recent assessments of the political and sectarian dimensions of the SunnÈ Revival, see J. P. Berkey, The Formation of Islam: Religion and Society in the Near East, 600–1800 (Cambridge, 2003), 189–202; and, D. Talmon-Heller, Islamic Piety in Medieval Syria: Mosques, Cemeteries and Sermons under the Zangids and AyyËbids (1146–1260) (Leiden, 2007), 8–9, 147–8. 38. On the spread of geometric ornament to Morocco under the Almoravids, see J. M. Bloom (ed.), The Minbar from the Kutibiyya Mosque (New York, 1998), and Y. Tabbaa, ‘Andalusian Roots and ʿAbbåsid Homage in the Qubbat al-Barudiyyin in Marrakech’, in Muqarnas 25 (2008): 133–46. 39. See E. H. Gombrich, The Sense of Order: A Study in the Psychology of Decorative Art (Ithaca, 1979), 17–19, for a definition of ‘decorum’, which is one of the key concepts in the book.

Bibliography Allen, T. (1989), A Classical Revival in Islamic Architecture. Wiesbaden. Allen, T. (1988), Five Essays on Islamic Art. Sebastopol. Auld, S. (2009), ‘The Minbar of NËr al-DÈn in Context’, in R. Hillenbrand and S. Auld (eds), AyyËbid Jerusalem: The Holy City in Context, 1187–1250, London, pp. 72–93. Berkey, J. P. (2003), The Formation of Islam: Religion and Society in the Near East, 600–1800. Cambridge. Blair, S. S. (2008), ‘A Brief Biography of AbË Zayd’. Muqarnas, 25: 155–76. Blair, S. S. and J. M. Bloom (2006), Cosmophilia: Islamic Art from the David Collection, Copenhagen. Boston. Bloom, J. M. (2007), Arts of the City Victorious: Islamic Art and Architecture in Fatimid North Africa and Egypt. New Haven. Bloom, J. M. (1980), ‘Meaning in Early Fatimid Architecture: Islamic Art in North Africa and Egypt in the Fourth Century A.H. (Tenth Century AD)’. Ph.D. diss., Harvard University. Bloom, J. M. (1983), ‘The Mosque of al-Óåkim in Cairo’. Muqarnas, 1: 15–36. Bloom, J. M. (2009), ‘Woodwork in Syria, Palestine and Egypt during the 12th and 13th Centuries’, in R. Hillenbrand and S. Auld (eds), AyyËbid Jerusalem: The Holy City in Context, 1187–1250, London, pp. 129–46.

syrian woodwork

Bloom, J. M. (ed.) (1988), The Minbar from the Kutibiyya Mosque. New York. Broadhurst, R. J. C. (tr.) (1952), The Travels of Ibn Jubayr. London: Cape. Burckhardt, T. (1976), Art of Islam: Language and Meaning. London. Contadini, A. (1978), Fatimid Art at the Victoria and Albert Museum. London. Creswell, K. A. C. (1952–9), The Muslim Architecture of Egypt. Oxford. Critchlow, K. (1976), Islamic Patterns: An Analytical and Cosmological Approach. New York. Dimand, M. (1937), ‘Studies in Islamic Ornament, I: Some aspects of Omaiyad and Early ʿAbbasid Ornament’. Ars Islamica, 4: 293–337. Elisséeff, N., ‘NËr al-DÈn Ma˙mËd b. ZankÈ’, EI2 8: 130–5. Gombrich, E. H. (1979), The Sense of Order: A Study in the Psychology of Decorative Art. Ithaca. Grabar, O. (1978), The Alhambra. Cambridge. Grabar, O. (1992), The Mediation of Ornament. Princeton. Herzfeld, E. (1943), ‘Damascus: Studies in Architecture – II’, Ars Islamica, 10: 13–70. Herzfeld, E. (1954–6), mcia 2. Syrie du Nord: Inscriptions et Monuments d’Alep. 3 vols. Paris. Ibn AbÈ Ußaybiʿa (1966), ʿUyËn al-anbåʾ fi †abaqåt al-a††ibåʾ. Beirut. Ibn KhaldËn (2005), The Muqaddima: An Introduction to History, tr. F. Rosenthal. 3 vols. Princeton. al-JazarÈ, I. (1974), Ibn al-Razzåz. Kitåb fÈ maʿrifat al-˙iyal al-handasiyya [The Book of Knowledge of Ingenious Mechanical Devices], tr. D. R. Hill. Dordrecht. Korn, L. (2004), Ayyubidische Architektur in Ägypten und Syrien: Bautätigkeit im Kontext von Politik und Gesellschaft 564–658/1169–1260. Heidelberg. Kühnel, E. (1977), The Arabesque: Meaning and Transformation of an Ornament, tr. R. Ettinghausen. Graz. Lamm, C. J. (1935), ‘Fatimid Woodwork, its Style and Chronology’, Bulletin de l’Institut d’Égypte, 18: 59–91. Lev, Y. (2008), ‘The Jihåd of Sultan NËr al-DÈn of Syria (1146–1174): History and Discourse’, JSAI 35: 227–84. Makdisi, G. (1973), ‘The SunnÈ Revival’, in D. S. Richards (ed.), Islamic Civilization, 950–1150: A Colloquium, Oxford, pp. 155–68. Mayer, L. A. (1958), Islamic Woodcarvers and Their Works. Geneva. Mayer, L. A. (1959), Islamic Metalworkers and their Works. Geneva. Morray, D. W. (1994), An AyyËbid Notable and his World: Ibn al-ʿAdÈm and Aleppo as Portrayed in His Biographical Dictionary of People Associated with the City. Leiden. Nasr, S. H., Islamic Art and Spirituality. Albany. Necipo©lu, G. (1995), The Topkapi Scroll: Geometry and Ornament in Islamic Architecture. Santa Monica. Rabbat, N. (1998), ‘Architects and Artists in Mamlûk Society: The Perspective of the Sources’, Journal of Architectural Education, 52: 30–7. Tabbaa, Y. (2008), ‘Andalusian Roots and ʿAbbåsid Homage in the Qubbat al-Barudiyyin in Marrakech’, Muqarnas, 25: 133–46. Tabbaa, Y. (1982), ‘The Architectural Patronage of NËr al-DÈn, 1146–1174’. Ph.D. diss., New York University.

125

126

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Tabbaa, Y. (1997), Constructions of Power and Piety in Medieval Aleppo. University Park. Tabbaa, Y. (1986), ‘Monuments with a Message: Propagation of Jihåd under NËr al-DÈn’, in V. P. Goss (ed.), The Meeting of Two Worlds: Cultural Exchange between East and West during the Period of the Crusades, Kalamazoo, pp. 223–40. Tabbaa, Y. (1994), ‘The Transformation of Arabic Writing: Part 2, The Public Text’, Ars Orientalis, 24: 119–47. Tabbaa, Y. (2007), The Transformation of Islamic Art during the Sunni Revival. Seattle. Talmon-Heller, D. (2007) Islamic Piety in Medieval Syria: Mosques, Cemeteries and Sermons under the Zangids and AyyËbids (1146–1260). Leiden. Talmon-Heller, D. (1922), mcia 2. Syrie du Sud. Tome Premièr, Jérusalem ‘Ville’. Cairo. Williams, C. (1985), ‘The Cult of ʿAlid Saints in the Fatimid Monuments of Cairo, Part II: The Mausolea’. Muqarnas, 3: 39–60. Williams, C. (1989), ‘Qurʾånic Inscriptions in the TåbËt al-Óusayn in Cairo’. Islamic Art, 2: 3–14.

PART II  IRAQ

CHAPTER SEVEN

The Mosque of Nu¯r al-Dı¯n in Mosul, 1170–1172 In his detailed exposition of the meritorious acts of NËr al-DÈn, the thirteenth-century Damascene historian AbË Shåma wrote: ‘[NËr al-DÈn] built congregational mosques in all regions, of which his mosque in Mosul is the ultimate in beauty and excellence. It is especially praiseworthy that he entrusted its construction and the supervision of its expenses to the shaykh ʿUmar al-Mallå,1 may God have mercy on him, who was a pious man. He was told: “such person is not suited for the task”. He replied: “If I were to assign this job to one of my associates, whether soldier or scribe, I know that he would oppress some of the time; and a mosque cannot be founded on oppression. I suspect that this shaykh will not do wrong; but if he does, then it is his sin not mine.”’2 This short anecdote about the mosque of NËr al-DÈn in Mosul is quite telling about one of the last and most important foundations of the Syrian sovereign. First, since Mosul was not directly ruled by NËr al-DÈn, his building a congregational mosque in it represents an unusual act of patronage that requires greater scrutiny.3 Second, further investigation into the identity of his appointee reveals that, more than a simple ascetic, ʿUmar al-Mallå was a Sufi of some note and the pole of opposition to the Christians of Mosul, in particular the Christian vizier Fakhr al-DÈn ʿAbd al-MasÈ˙.4 Finally, it seems clear from NËr al-DÈn’s hands-off approach to the project and from his entrusting it to a ‘pious person’ that some deemed unsuitable, that he was perhaps concerned less with the details of the construction of his mosque and more with its overall message. This piece is primarily concerned with the message of the mosque Al-NËrÈ in Mosul as it pertains to NËr al-DÈn’s dynastic ambitions in upper Mesopotamia, his constant struggle for SunnÈ Islam and his undiminished opposition to Christianity, whether Frankish or local. A discussion of the political and religious circumstances surrounding the creation of this mosque will hopefully shed some light on NËr alDÈn’s unusual act of patronage and perhaps on the ideological intent Yasser Tabbaa (2002), ‘The Mosque of Nur al-Din in Mosul, 1170–72’, Annales Islamologique 36, 339–60.

130

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

of the mosque. But this must be coupled with a careful analysis and reconstruction of the original design and epigraphic programme of the mosque, a very difficult task since the mosque no longer exists in anything resembling its twelfth-century form (Figures 7.1 and 7.2).5 Nu¯r al-Dı¯n in Mosul Before his death in 1146, ʿImåd al-DÈn ZankÈ, father of NËr al-DÈn and founder of the Zankid dynasties in Syria and upper Mesopotamia, ruled over a vast region, extending from Mosul to Aleppo and from Edessa to the outskirts of Damascus. His domain was split between his two oldest sons: Sayf al-DÈn GhåzÈ, the eldest, took Mosul; NËr al-DÈn Ma˙mËd took Aleppo, to which he added Damascus in 1154. Following a joint campaign the two brothers made in 1146 to wrest Edessa from the forces of the second Crusade, NËr al-DÈn made a special trip to Mosul, where he witnessed Sayf al-DÈn’s investiture and acknowledged his suzerainty over the city. When Sayf al-DÈn unexpectedly died in 1149, NËr al-DÈn once again went to Mosul, but this time as the elder of the Zankid household and the nominal suzerain of all its territorial possessions. He arranged for the succession of his younger half brother, Qu†b al-DÈn MawdËd, who, in return, granted him important provinces along the Euphrates and agreed to pronounce his name during the khu†ba. NËr al-DÈn managed to outlive Qu†b al-DÈn, who died in 1170, at which time NËr al-DÈn went to Mosul for the third time, also to supervise the succession and reaffirm his own suzerainty over the entire Zankid domain.6 This third succession was far from straightforward, however, and its various complications are quite telling about NËr al-DÈn’s huge influence in Mosul and, indirectly, about his motivation for building a congregational mosque there. While on his deathbed, Qu†b al-DÈn decided to assign as his successor his oldest son ʿImåd al-DÈn II, a decision approved by NËr al-DÈn since ʿImåd al-DÈn had in fact grown up in his court in Aleppo and was married to one of his daughters. But Qu†b al-DÈn’s vizier, Fakhr al-DÈn ʿAbd al-MasÈ˙, a Christian captive, had another son in mind as successor, the younger Sayf al-DÈn GhåzÈ II, and he managed to sway the dying prince toward this choice.7 Fakhr al-DÈn’s intervention, undoubtedly motivated by his desire to curb NËr al-DÈn’s influence in Mosul, greatly angered NËr al-DÈn and his designated successor. Matters quickly worsened, and in September 1170 NËr al-DÈn began preparation to take Mosul. In order to dress his conquest with proper legal garb, NËr al-DÈn sent an emissary, the famous chronicler ʿImåd al-DÈn al-Kåtib al-IßfahånÈ, to the ʿAbbasid caliph Al-Musta∂Èʾ, asking for the caliph’s permission and blessings. The permission, along with a khil’a (robe of honour), arrived while NËr al-DÈn had already begun his campaign, which

ˉ R AL-DIˉ N THE MOSQUE OF NU

took him first to Sinjår, an important town about 75 miles due west of Mosul.8 Having taken Sinjår by siege, NËr al-DÈn proceeded to Mosul, which he took in the same year after lengthy negotiations with Fakhr al-DÈn. Pleading his case for an orderly succession within the Zankid household, Fakhr al-DÈn finally convinced NËr al-DÈn to accept the succession of his candidate, Sayf al-DÈn GhåzÈ II, while agreeing that the older son ʿImåd al-DÈn would be made governor of the lesser province of Sinjår.9 But this concession came with several conditions, which interestingly had little to do with the succession and everything to do with containing the influence of the Christians in Mosul. NËr al-DÈn’s attitude toward the Christian vizier and Christians generally can in fact already be predicted in his letter to Sayf al-DÈn II during the siege of Mosul: ‘My intention is not the city itself, but to preserve the city for you. For I have received letters telling a thousand tales about ‘Abd al-MasÈ˙’s ill treatment of the Muslims. My aim is to remove this Christian from governing Muslims.’10 This anti-Christian attitude is consistent with other acts of NËr al-DÈn, including his ruthless repression of the rebelling Christians of Edessa (Urfa) in 1146.11 The first of NËr al-DÈn’s conditions concerned Fakhr al-DÈn: he was to leave Mosul for Aleppo, to convert to Islam and to change his name from ʿAbd al-MasÈ˙ (slave of the Messiah) to Abdullah (slave of Allah). With Fakhr al-DÈn thus neutralised, NËr al-DÈn seems to have felt free to impose various repressive economic and legislative measures that were undoubtedly intended to lower the status and limit the authority of the Christians, who were at that time an especially large minority in Mosul.12 Specifically, he expanded the collection of tributes from various Christian villages and communities, increased the jizya tax, and reinstated the rule that Christians should cut their hair short and wear a distinctive belt, zinnår.13 Finally, NËr al-DÈn applied with renewed strictness the ‘Pact of ʿUmar’ which upheld the safety of existing churches, but prohibited any new construction or restoration and subjected such violations to confiscation.14 Although NËr al-DÈn’s earlier atrocities against Edessa in 1146 may be partly attributed to his sense of outrage at the alliance struck by the Armenians of the city with the Crusaders,15 these later acts were more deliberate and systematic in their selective destruction and pillaging of churches and monasteries. NËr al-DÈn even appointed the noted jurist Sharaf al-DÈn b. AbÈ ʿAßrËn as an inspector of the Christian towns of the JazÈra, giving him a free hand to demolish all new structures and confiscate their endowments.16 These repressive acts covered a wide swathe of the Christian JazÈra, including Mardin, Nisibis (NußaybÈn), Mosul and other places.17 In 1171, for example, NËr al-DÈn mandated the conversion of the Monastery of the Virgins (Dayr al-Abkår) near Mardin into a

131

132

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

mosque for Kurds. Later in the same year, ‘he ordered the destruction of all new additions in the churches and monasteries of NußaybÈn and several other places’.18 In June 1172 ‘the Muslims took over the church of St. Thomas in Mardin’ and converted it into a mosque, on the pretext that a certain patron of the church named BarßËm had raped a Muslim woman.19 Even beyond their immediate negative impact, these anti-Christian measures created an atmosphere of fear among the Christians of Mosul and the JazÈra and contributed to later act of pillage and confiscation.20 It is undoubtedly because of these repressive measures that Syriac Christian writers, including Michael the Syrian and Bar Hebraeus, were harshly critical of NËr al-DÈn.21 NËr al-DÈn stayed twenty-four days in Mosul, during which time he ordered the foundation of a congregational mosque in a thinly populated part of the city.22 Surveying this location from a nearby minaret, he ordered the annexation of adjoining houses and shops, but only after their owners had been adequately compensated.23 He then appointed Shaykh ʿUmar al-Mallå as supervisor of the project, entrusting him with the huge sum of 60,000 dinars for the purchase of these properties and the completion of the mosque. The mosque was apparently completed in just under two years, for in 568/1172 NËr al-DÈn visited Mosul for the fourth and last time and performed the Friday prayer in it. He also formalised the waqf allotments of his mosque, which seem to have comprised agricultural lands around Mosul and commercial properties near the mosque, including a large covered market, qaysåriyya, with numerous shops.24 Furthermore, NËr al-DÈn ‘gave considerable alms, appointed a kha†Èb and muezzins for the mosque, and supplied it with rugs and straw mats’.25 Finally, after the mosque had been completed, NËr al-DÈn ordered a madrasa built next to it and even appointed its first teacher.26 Nothing has survived of this madrasa, but its foundation is perfectly consistent with NËr al-DÈn’s patronage of SunnÈ institutions all over Syria. Archaeology: the original mosque Although the present rebuilt mosque has been stripped of its long history (Figures 7.1 and 7.2), Herzfeld’s investigation, some early photographs and the existing architectural and epigraphic remains can help us reconstruct the original mosque and propose a possible chronology for its later phases. Herzfeld’s work in the first decade of the twentieth century is the only serious study of this mosque before its demolition and rebuilding in the 1940s. More recent studies by Iraqi scholars, particularly DaywajÈ and Jumʾa, made minor but important contributions to Herzfeld’s work, particularly in terms of the early chronology of the mosque and its ornament and calligraphy. Finally, in 1979 I came upon a wonderful collection of large

ˉ R AL-DIˉ N THE MOSQUE OF NU

format glass negatives at the Iraqi Institute for Antiquities, from which I was allowed to make prints of the exterior and interior of the mosque prior to its destruction. The old photographs show a ruinous and poorly built mosque located at the southern end of a vast enclosure, about 90 × 65 m, approximately corresponding to the present enclosure (Figure 7.6). The peeling plaster of the southern exterior wall exposed the building material, a conglomerate of rubble and broken bricks, bound with mortar and covered with thick plaster. The sanctuary was choked by parasitical buildings on its south and west, and disfigured by several pierced windows and an unsightly buttress built against the mi˙råb. The dome looked misshapen, with a hemispherical lower half incongruously surmounted by a faceted cone (Figure 7.7). The 60-metre tall minaret (Al-Hadbåʾ) stood, as it still does, at the north-west corner of the enclosure, separated from the sanctuary by an empty court. An exterior mi˙råb could be seen about ten metres north of the wall of the sanctuary.27 Even Herzfeld was confounded by the chaotic interior, protesting that the walls were so thickly covered with plaster as to ‘make the separation of the building phases difficult’.28 Unable to subject the building to a thorough archaeological investigation, Herzfeld instead limited his examinations to the surface (Figure 7.8). Two types of columns were used in the mosque then, both made of the same soft, dark blue marble common in the area of Mosul since Assyrian times (Figure 7.9). Type 1 had a thick octagonal shaft with a square frieze and console but no real capital. Type 2 had a thinner composite shaft with four engaged columns and a lyre-shaped capital. The two types were completely different in their height, thickness, shaft and capitals, and clearly belonged to two different periods. This assertion is confirmed by the fact that several Type 2 columns were placed against the octagonal shaft of Type 1 columns, and a base and an abacus were added to them in order to compensate for the height difference. It is further corroborated by the decoration on the capitals of both column types, which continues behind their point of contact, suggesting originally freestanding columns. From this Herzfeld concluded that the octagonal columns belonged to the first phase while the composite columns belonged to a later phase. This much is perfectly consistent with archaeological evidence, but what caused Herzfeld to err was that he assumed that the main mi˙råb of the mosque was part of the first phase of the building (Figure 7.10). This mi˙råb is dated to Jumada I, 543/September-October, 1148, which led Herzfeld to conclude that the mosque was first built during the short reign of Sayf al-DÈn GhåzÈ I (541/1146-544/1149).29 Noting further that the capitals of the octagonal columns demonstrated ornamental and palaeographic affinities with the mi˙råb, Herzfeld, therefore, attributed them all to the first period. Since these columns constituted the main support system of the mosque,

133

134

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 7.1  Mosul, Mosque Al-NËrÈ, minaret (photo: Yasser Tabbaa)

ˉ R AL-DIˉ N THE MOSQUE OF NU

Herzfeld went on to conclude that the mosque was not only begun under GhåzÈ I, but was actually completed under him. It followed then that, according to Herzfeld, the building phase of NËr al-DÈn was a mere restoration, to which he attributed the bundled columns on the basis of their rather superficial similarity to the composite piers at the Great Mosque of Raqqa, which was in fact restored by NËr al-DÈn in 1168.30 As for the fact that Ibn al-AthÈr clearly stated that this mosque was built by NËr al-DÈn, Herzfeld argued that the Mosulite historian was only eleven to thirteen years old when NËr al-DÈn began his ‘restoration’ and that he must have repeated faulty information.31 This peculiar objection to one of the most important medieval Islamic historians can be rejected out of hand, as it was in fact by DaywajÈ as early as 1949. Dismissing Herzfeld’s objection to the reliability of Ibn al-AthÈr’s account, DaywajÈ proposed that, while always reliable, the historian was especially so for the events that took place during his own childhood, for which he often relied on direct accounts from his father, who was a high official at the Zankid court of Mosul. Furthermore, DaywajÈ added that the details of NËr al-DÈn’s acquisition of the land for the mosque, the specific sum of 60,000 dinars which he endowed for the mosque, and the waqfs he allotted to it were all too specific to be mere fabrication, especially since only twenty-five years separated GhåzÈ’s death and the beginning of the mosque.32 Finally, not one historian either disputed Ibn al-AthÈr’s account or even proposed an alternative one that mentions GhåzÈ I.33 Archaeologically, DaywajÈ pointed out that the mi˙råb of 1148 – the cornerstone of Herzfeld’s periodisation – was not indigenous to the mosque, but had been brought to it from the Umayyad mosque of Mosul by a Shaykh Mu˙ammad al-NËrÈ in 1864, as part of a restoration project.34 Even a cursory examination of the mi˙råb before its most recent restoration is enough to suggest that it was not intended for this mosque but rather brought into it in fragments and reassembled, using no less than a dozen other fragments originating from three or four sources (Figures 7.7 and 7.8).35 Thus, the 1148 mi˙råb did not belong to the original mosque of NËr al-DÈn, although its stunning arabesque ornament is clearly related to the later arabesque decorations on the capitals of the mosque of NËr al-DÈn and to the even later ornament on the mi˙råb that Badr al-DÈn Luʾluʾ added to the mosque.36 If the octagonal columns, therefore, constitute the support system of the NËrid mosque, where did the other columns, with their composite shafts and lyre-shaped capitals, come from? Although Herzfeld is right in tracing the form of their shafts and capitals back to Raqqa and then Samarra, these basic similarities need not argue for an early date, for the Mosul columns show considerable development over their brick and stucco prototype. In addition to being carved out

135

136

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

of stone rather than moulded in stucco, the Mosul capitals have a greater three-dimensional feel, most clearly visible in the treatment of the corners. Similarly, while the shafts of the Mosul columns can be compared to the brick piers with four engaged columns in Raqqa and Samarra, in Mosul the four engaged columns have been collapsed into the square shaft, turning the composite pier into a fully rounded bundled column. Such bundled columns are practically unknown in Islamic architecture.37 But similar bundled columns with lyre-shaped capitals are fairly common in the Christian buildings of Mosul, including the churches of Mår A˙Ëdemmeh, Mår Ishaʾya and Mår JurjÈs, where they generally date to the first half of the thirteenth century (Figure 7.11).38 It is therefore likely that these columns were salvaged from a ruined or destroyed Christian church and brought into the mosque. Unlike the inscribed capitals from the NËrid phase, none of the lyre-shaped capitals contains any Arabic, let alone Qurʾånic inscriptions, further setting them apart from the original mosque. Judging from the crude way by which they were juxtaposed against the octagonal columns, I would suggest that they were added to the mosque at a very late date, possibly in the 1860s, when the mi˙råb was also brought in. The mosque was then built from start to finish in one endeavour by NËr al-DÈn, using heavy octagonal columns with inscribed capitals. How these columns were arrayed and what they supported remain problematic since the mosque has been entirely rebuilt according to a new plan and since early photographs show very little of the vaulting, except for the mi˙råb dome. Fortunately, Herzfeld made a plan and perspective drawing of the mosque as it appeared before its reconstruction and also proposed a restoration plan that purports to show the mosque in its original NËrid design. The pre-reconstruction mosque was a broad and narrow structure (approximately 75 × 20 metres), seven bays wide and only one and a half bays deep, with a large dome over the mi˙råb (Figure 7.13). The seven bays at the qibla wall were alternatingly large squares and smaller rectangles, while those north of them were small rectangular bays, each with a door to the courtyard. In his restoration plan, Herzfeld seems to have doubled the existing mosque along its north–south axis, creating a very large mosque (75 × 38 m), four bays deep and seven bays wide (Figure 7.14). These bays alternate in both depth and breadth between wide and narrow, such that there are two wide and two narrow horizontal rows and three wide and four narrow vertical aisles. The bays are covered with a system of alternating domes and barrel vaults, for a total of six large and eight smaller domes. Therefore, rather than a narrow hypostyle mosque that is focused on a maqßËra dome, Herzfeld proposes a deeper mosque with a rhythmic alternation of bays, but without a dominant maqßËra dome. Even by his own admission, however, Herzfeld’s restoration plan

ˉ R AL-DIˉ N THE MOSQUE OF NU

is based more on aesthetic prerogatives than on adequate archaeological evidence. In fact, at least two archaeological features – the number of available columns and the mi˙råb dome – argue against his plan and its peculiar vaulting system. The plan requires about 140 columns, more than four times the number of octagonal columns existing today.39 In the absence of so many columns, it appears more likely that the NËrid mosque more or less resembled the mosque that Herzfeld saw in 1910 and that is documented in early phot­ ographs. Although it is not unlikely that this mosque had barrel vaults, there is little support for the vaulting system proposed by Herzfeld, particularly the large number of domes. Certainly, none of these proposed domes appears in the old photographs of the mosque’s roof, which is flat with the exception of the large mi˙råb dome (Figure 7.6). Herzfeld omitted the dome over the mi˙råb from the original plan of the mosque Al-NËrÈ, noting that its sixteen-sided pyramidal exterior looked like a later restoration, but failing to take account of its interior appearance because ‘it lay in such darkness’.40 But there are fairly clear clues, both on the exterior and on the interior, that the pre-restoration dome represented two building phases. Looking at early photographs of this misshapen dome, we note that it first springs from its octagonal drum as a regular hemispherical dome before turning, about a third of the way up, into a faceted cone (Figures 7.3 and 7.4). This agglutination is clearer on the interior of the dome, where a zone of pendentives with large muqarnas cells provide the transition to the octagon (Figure 7.19). Twenty-four ribs spring from this octagon but end abruptly at about one-third the distance to the peak, where they seem to vanish under thick layers of plaster. This ribbed dome, which corresponds to the exterior hemispherical dome, represents the first building phase, whereas the superimposed sixteen-sided cone represents a later restoration, possibly from the reign of Badr al-DÈn Luʾluʾ, who built at least two other double-shell pyramidal domes in Mosul.41 Further distinguishing the cone from the dome is that its sixteen sides do not correspond to the twenty-four ribs of the first dome. Ribbed, or gored, domes are indigenous to Mosul and its surroundings, where they are generally datable to the first half of the thirteenth century. One such dome exists at the Mår Behnåm monastery just outside Mosul, where it covers the chapel of the Virgin, a chamber datable to the first half of the thirteenth century (Figure 7.16). This finely constructed dome rests on a sophisticated transition dome and has sixteen ribs, but is otherwise closely linked to the dome of the mosque Al-NËrÈ. Another dome with twenty-four ribs has survived at the shrine of Sittna Zainab in Sinjår, where it dates to the period of Badr al-DÈn Luʾluʾ (1233–59).42 Interestingly, both of these domes have a hemispherical exterior, which may also have been the original shape of the mosque’s dome.

137

138

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

It seems likely, therefore, that the mosque of NËr al-DÈn had a large dome over the mi˙råb, gored on the inside but perhaps hemispherical on the outside. The shallow plan of the mosque and its maqßËra dome (Figure 7.17) would seem to link it with a fairly large group of twelfth- and thirteenth-century Syrian and JazÈran mosques that were modelled after the venerable Umayyad mosque of Damascus, including those at Diyarbakir, Mardin and Dunaysir (Kiziltepe), and possibly also the one at MayyåfåriqÈn (Silvan), which also has similarly alternating bays of varying depths.43 But the large maqßËra dome was clearly based on the model of Iranian Saljuq domes, particularly the one at the Masjid-È JåmÈ in Isfahan, or even the Saljuq dome added to the Great Mosque of Damascus in 1082. Thus, despite its use of local materials and vaulting techniques, the mosque Al-NËrÈ presents a synthesis, albeit awkward and incomplete, of the Umayyad mosque of Damascus and of nearly contemporary Saljuq architecture. The inscriptions Although the mosque is curiously devoid of any historical inscriptions, it is very rich in Qurʾånic inscriptions and pious supplications. These exist in three distinct groups: short friezes on the capitals of the octagonal columns; marble bands with black inlaid inscriptions and a fragment of a stucco frieze; and a large panel of stucco decoration above the mi˙råb. The inscriptions on the columns once formed parts of continuous Qurʾånic verses, but the recent rebuilding of the mosque seems to have disturbed the original sequence (Figures 7.15 and 7.16). Presently twenty-four capitals bear Qurʾån 2:255; 9:18–19 and 24:36–8; the marble friezes have most of 2:148–50; and the short stucco frieze contains part of 3:18. Nearly all the capitals of the octagonal columns have vegetal arabesque decoration on three sides, while the north-facing side bears a Qurʾånic inscription written on a bed of arabesque. The script used in these inscriptions is a rather squat and fleshy thuluth, of a type previously seen in Iranian Saljuq monuments and in some of the buildings of NËr al-DÈn in Syria.44 It is also clearly related to, though somewhat more developed than the cursive inscriptions encircling the 1148 mi˙råb in the mosque. Qurʾån 2:255 is the well-known Throne Verse (åyat al-kursÈ), a verse that describes God’s omniscience, omnipotence and dominion over heaven and earth. This is one of the most, if not the most, frequently used verse in monumental inscriptions, where because of its eschatological significance, it is often written at the springing of the dome or within the mi˙råb niche.45 But it is also often inscribed on portals, minbars and tombstones. For different reasons, Qurʾån 9:18 is also extremely common in mosque inscriptions, for, according to Blair, ‘it is one of only three Koranic references to God’s mosques (masåjid Allåh), a special term

ˉ R AL-DIˉ N THE MOSQUE OF NU

Figure 7.2  Mosul, Mosque Al-NËrÈ, mi˙råb, dated 543/1148 (photo: Yasser Tabbaa)

distinct from any masjid or place of prayer’.46 Its continuation, 9:19, however, is much less common, possibly because it seems to distinguish between passive pious practices –‘giving water to pilgrims and the inhabiting of the Holy Mosque’– and active Islamic practices

139

140

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

–‘struggle in the way of God’– and clearly favours the latter.47 It was most likely included because it refers to jihåd, which would link the verse with the jihåd of NËr al-DÈn against the Crusaders and possibly even freeing Mosul from the rule of a Christian governor.48 Qurʾån 24:36–7 also makes reference to places of worship: ‘In temples God has allowed to be raised up, and His Name be commemorated therein; therein glorifying Him, in the morning and the evenings, are men whom neither commerce nor trafficking diverts from the remembrance of God and to perform the prayer, and to pay the alms, fearing a day when hearts and eyes shall be turned about, that God may recompense them for their fairest works and give them increase of His bounty; and God provides whomsoever He will, without reckoning.’ It was undoubtedly for its evocative linking of the building of ‘temples’ and the ‘recompense’ and ‘bounty’ accruing from this pious act that these verses were commonly placed on mosque portals and mi˙råbs. Fragments of the same verses, written in floriated KËfic, flank the 1148 mi˙råb, originally completely surrounding it, as has been done in the restored mi˙råb (Figure 7.11). In addition to the inscribed capitals, the mosque once contained long friezes with inscriptions, which are now exhibited above the entrance to the Islamic galleries of the Iraqi Museum in Baghdad (Figure 7.20). These consist of four pieces of white marble (total length 4.35 m) on which the inscription is carved out and filled in with a bituminous paste, an ancient Mesopotamian technique where bitumen is widely available. The use of the same technique in two earlier NËrid monuments – the bÈmåristån Al-NËrÈ in Damascus (1154) and the mosque Al-NËrÈ in Hama (1168) – suggests that the Mosul fragments also belong to the NËrid phase of the mosque. We can also propose that, as with the inscribed friezes at the Hama mosque, these friezes were once embedded in the qibla wall, on both sides of the mi˙råb.49 The fragmentary friezes contain parts of 2:148–50, verses rarely used in inscriptions but that seem quite appropriate in a qibla wall. ‘From whatsoever place thou issuest, turn thy face towards the Holy Mosque; it is the truth from thy Lord. God is not heedless of the things you do. From whatsoever place thou issuest, turn thy face towards it, that the people may not have any argument against you.’ Originally intended for non-Muslim or early converts to Islam, who were presumably at a loss as to what direction they should turn in their prayer, these verses assert the proper Islamic orientation ‘towards the Holy Mosque’ at Mecca.50 Another fragmentary inscription, now lost, has been documented by Herzfeld, who, quite correctly I believe, attributes it to the period of NËr al-DÈn.51 The fragment is part of Qurʾån 3:18: ‘God bears witness that there is no God but He – and the angels, and men possessed of knowledge – upholding justice; there is no God but He, the All-mighty, the All-wise.’ This verse is nearly as common as the

ˉ R AL-DIˉ N THE MOSQUE OF NU

Throne Verse, with which it shares the same concept of God’s unity and dominion. Indeed, by using the word shahada (to bear witness), which is the root verb for shahåda (the Muslim declaration of faith), this verse is an even more assertive statement about the absolute unity of God.52 Lastly, the space between the mi˙råb and the springing of the dome was previously completely covered with a large panel of moulded and carved stucco, which is now partly preserved in the Iraqi Museum in Baghdad (Figure 7.21). I hope to discuss this complex and interesting panel in a separate piece on the patronage of Badr al-DÈn Luʾluʾ, from whose period the panel most likely dates.53 But I would like here simply to refer to the central square KËfic inscription, whose text seems to echo similar texts used by NËr al-DÈn in other mosques. The square KËfic text reads as follows: ‘Mu˙ammad, AbË Bakr, ʿUthmån, ʿAlÈ, ʿUmar, Óasan, Óusayn, may God be pleased with them all.’ This formula – which includes the names of the Prophet, the Companion Caliphs and the first two Shi‘i imams – is a kind of ecumenical SunnÈ prayer that was especially common during the period of NËr al-DÈn. As far as I know, it is first seen in epigraphic form in three mosaic inscriptions at the Umayyad mosque of Damascus, one of which mentions the name of NËr al-DÈn and the other two are attributable to him. Datable to the year 554/1159, when NËr al-DÈn carried out important restorations in the Great Mosque, two of these inscriptions are identical in content to the Mosul inscription and one continues to include the name of ʿÅʾisha, and Få†ima.54 Although the stucco panel at the mosque most likely dates to the time of Badr al-DÈn, it is quite possible that the square KËfic inscription was modelled after an earlier NËrid inscription. The main purpose of these inscriptions was not to gloat over the victory of SunnÈs, but rather to present a formula which unites the SunnÈ sects and which may be found acceptable among the moderate Shi‘is. Badr al-DÈn, who was known for his Shi‘i inclinations, may have found in this formula an acceptable compromise between the two dominant sects in Mosul.55 Other than the square KËfic inscriptions, all the inscriptions in the mosque Al-NËrÈ were Qurʾånic, and most of these used rather commonly quoted verses. Two verses, 2:255 and 3:18, present the essence of Islamic theology: God’s unity, His omniscience and omnipotence, and His rather anthropomorphic dominion in heaven and on earth. The other three dwell on the mosques (masåjid) and temples (buyËt) of God and seem to have been specifically chosen for that reason. But despite their rather commonplace content, the inscriptions are still quite striking by their quantity and accessibility. The large number of inscriptions in the qibla wall, in the mi˙råb, and especially on capitals ensured that the worshipper would be immersed in the word of God, which he would be able to experience spatially

141

142

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

by moving from one capital to the next. The same desire to make inscriptions accessible and legible to the congregation may have been behind the use of the black-inlaid inscriptions, a technique primarily intended to enhance the legibility of religious and historical texts. The minaret The minaret of the mosque of NËr al-Din is not only the most distinctive feature of the mosque but of Mosul as well. Its tapered cylindrical shaft (45 metres high) springs from a battered cubical base and ends in a little cupola that rises a few metres above a bracketed balcony (Figures 7.19 and 7.20). The base and the entire shaft are decorated in typically Iranian brick decoration, both basket weave (hazår baf) and strapwork, but without any inscriptions. Three sides of the base are decorated with a simple stepped pattern whereas the western side, which faces an important street, contains an elaborate star pattern. Towering sixty metres over the utterly flat landscape, this is the tallest minaret in Iraq. The freestanding location of the minaret at the elevated northeastern corner of the courtyard, its extreme height and its excessive decoration all link it with contemporary Iranian Saljuq minarets, which have been interpreted by Hillenbrand as ‘expressions alike of conspicuous consumption and conspicuous piety’.56 In other words, these minarets were intended less as a functional appendage of the mosque and more as symbolic features on the urban scale. But whereas Saljuq Iranian minarets – in their excessive number, size and decoration – may have been intended to address tribal or interIslamic differences, the minarets at Mosul and the JazÈra were most

Figure 7.3  Mosul, Mosque Al-NËrÈ, 1172, inscribed capital

ˉ R AL-DIˉ N THE MOSQUE OF NU

likely intended to highlight Islam’s dominance over Christianity. That in itself might explain why some of the tallest minarets in the JazÈra – at Mardin, Hasankeyf, DåqËq, IrbÈl, Mosul and Sinjår – were built in cities with important Christian populations.57 Furthermore, most were built in the aftermath of the period of tolerance that preceded NËr al-DÈn and that resulted in the creation of numerous Christian buildings.58 Conclusion NËr al-DÈn founded his mosque in Mosul at the peak of his powers, when his domain extended from Mosul to Cairo and from Diyarbakir to Damascus and when his spiritual aura nearly outshone that of the ʿAbbasid caliph himself. Far more than an act of restoration to a preexisting mosque, as Herzfeld had proposed, I have argued above that the mosque was entirely built by NËr al-DÈn, according to a typical JazÈran mosque plan, with a shallow sanctuary and a maqßËra dome, but utilising unusual octagonal columns with inscribed capitals. I have also discussed the mosque within the context of NËr al-DÈn’s hegemonic ambitions in Mosul and his confrontational relationship with the Christians of Mosul and the JazÈra. Built soon after NËr al-DÈn had removed the Christian governor of the city and while he was actively engaged in undermining local Christianity and dismantling its structures, the mosque, with its towering minaret, may be seen as a late but definitive statement of Islamic dominion over a city with a substantial Christian population. But the mosque can also be viewed from a strictly Islamic perspective, though one coloured by the traditionalist SunnÈ beliefs of its founder. I have elsewhere discussed NËr al-DÈn’s pivotal role in enforcing the use of cursive writing for all his public inscriptions, proposing that cursive scripts embodied the exoteric, literalist tenets of the SunnÈ revival.59 The Qurʾånic inscriptions at the Mosque Al-NËrÈ statement take his exoteric trend a step further, by placing some inscriptions within reading distance of the congregation and by colouristically enhancing the legibility of others. Surrounding the worshipper by evocative Qurʾånic verses, whose impact was audibly reinforced during the Friday sermon, may well have been one of the most successful innovations of this period of heightened piety and increased orthodoxy.60 Finally, the mosque Al-NËrÈ in Mosul may be seen within the context of NËr al-DÈn’s architectural patronage, which, through its extent, sizes and innovation, exerted considerable influence on the architecture of the central Islamic world and created a precedent for later patrons with similar ambitions, including Baybars, Al-Nåßir Mu˙ammad and Tinkiz. Blending established building designs – the hypostyle mosque plan or the madrasa courtyard plan with innovative, even showy features (muqarnas domes and portals, tall

143

144

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

minarets, or inscriptions on capitals) – the architecture of NËr al-DÈn continues and reinforces an earlier Saljuq trend toward an extroverted Islamic architecture that addresses the spiritual and visual needs of the lay worshipper and user.61

Figure 7.4  Mosul, Mosque Al-NËrÈ: dome and portico from courtyard, present condition (photo: Yasser Tabbaa)

Figure 7.5  Mosul, Mosque Al-NËrÈ: prayer hall interior from north-east, present condition (photo: Yasser Tabbaa)

ˉ R AL-DIˉ N THE MOSQUE OF NU

Figure 7.6  Mosul, Mosque Al-NËrÈ: exterior from north-west, condition c. 1930. (Courtesy of the General Institute for Antiquities, Baghdad)

Figure 7.7  Mosul, Mosque Al-NËrÈ: exterior from south-east, condition c. 1930. (Courtesy of the General Institute for Antiquities, Baghdad)

145

146

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 7.8  Mosul, Mosque Al-NËrÈ: prayer hall interior from east, condition c. 1930. (Courtesy of the General Institute for Antiquities, Baghdad)

Figure 7.9  Mosul, Mosque Al-NËrÈ: type 2 column placed against type 1 column, condition c. 1930. (Courtesy of the General Institute for Antiquities, Baghdad)

Figure 7.10  Mosul, Mosque Al-NËrÈ: mi˙råb, dated Jumada I, 543/ September–October 1148, condition c. 1930. (Courtesy of the General Institute for Antiquities, Baghdad)

ˉ R AL-DIˉ N THE MOSQUE OF NU

Figure 7.11  Mosul, Mosque Al-NËrÈ: mi˙råb, after restoration (photo: Yasser Tabbaa)

Figure 7.12  Mosul: a. Capital from Mår JurjÈs (Reise, 2, fig. 280). b. Capital from Mår A˙udemmeh (Reise, 2, fig. 282)

147

148

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 7.13  Mosul, Mosque Al-NËrÈ: plan in c. 1915 (Reise, 4, pl. 88)

Figure 7.14  Mosul, Mosque Al-NËrÈ: reconstruction of plan of first mosque. (Redrawn after Herzfeld in Reise, 4, fig. 237) (photo: Yasser Tabbaa)

ˉ R AL-DIˉ N THE MOSQUE OF NU

Figure 7.15  Mosul, Mosque Al-NËrÈ: maqßËra dome from below. (Courtesy of the General Institute for Antiquities, Baghdad)

Figure 7.16  Mår Behnåm near Mosul: ribbed vault over the Chapel of the Virgin, first half of the thirteenth century (photo: Yasser Tabbaa)

149

150

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 7.17  Mosul, Mosque Al-NËrÈ: reconstruction of plan of first mosque (Adapted with changes from fig. 11)

a

b Figure 7.18  Mosul, Mosque Al-NËrÈ: a. Capital inscribed with part of Qurʾån 2:255. b. Capital inscribed with part of Qurʾån 9:18

ˉ R AL-DIˉ N THE MOSQUE OF NU

a

b Figure 7.19  Mosul, Mosque Al-NËrÈ: a. Capital inscribed with part of Qurʾån 24:36. b. Capital inscribed with part of Qurʾån 24:37

Figure 7.20  Iraqi Museum, Baghdad: two inscription bands with Qurʾån 2: 148–9, previously at the Mosque Al-NËrÈ in Mosul (photo: Yasser Tabbaa)

151

152

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 7.21  Mosul, Mosque Al-NËrÈ: large stucco panel above the mi˙råb, condition c. 1930, currently at the Iraqi Museum, Baghdad. (Courtesy of the General Institute for Antiquities, Baghdad)

Figure 7.22  Mosul, Mosque Al-NËrÈ: minaret, 1170–2 (photo: Yasser Tabbaa)

Figure 7.23  Mosul, Mosque Al-NËrÈ: uppermost portion of minaret (photo: Yasser Tabbaa)

ˉ R AL-DIˉ N THE MOSQUE OF NU

Notes 1. N. Elisséeff, NËr al-DÈn un grand prince musulman de Syrie au temps des croisades 511–569/1118–1174 (Damascus, 1967), 1, 109 and II, 661 read ʿUmar’s nisba as Al-Mullå, which is a later designation for religious scholars, particularly among Shi‘is. But AbË Shåma, Kitåb al-raw∂atayn fÈ akhbår al-dawlatayn al-nËriyya wa’l-ßalå˙iyya, ed. M. H. A˙mad (Cairo, 1956–62), 11, 480 clearly calls him Al-Mallå (the filler), because ‘he used to fill buckets with lime in return for a wage’. 2. AbË Shåma, Kitåb al-raw∂atayn, II, 20–1. 3. Rulers and other patrons did sometimes build monuments outside their domain of authority, but in most case these acts of patronage were made in and around sacred shrines and other sites of pilgrimage, including Mecca, Medina and Jerusalem for the SunnÈs and Najaf, Karbala, Qumm and Mashhad for the Shi‘is. Even there, most foreign patrons of the medieval period built relatively small structures or additions to larger complexes, although that changed in later centuries. 4. See AbË Shåma, Kitåb al-raw∂atayn, II, 480, where the author describes ‘scholars and jurists, kings and princes visiting [ʿUmar] in his retreat’. And (483–4) where he details Shaykh ʿUmar’s machinations in affecting the banishment of Fakhr al-DÈn to Syria. 5. Between 1945 and 1950 the old mosque Al-NËrÈ was completely destroyed and rebuilt, using both old and new materials, according to a new plan and design. The dome and all the vaulting were completely torn down and replaced by a flat roof and a hemispherical dome, all made of reinforced concrete. And the interior was completely painted in white, turquoise blue and silver, giving it a totally modern and sterile appearance. See especially S. al-DaywajÈ, ‘Al-Jåmi’ al-NËrÈ fÈ al-Mawßil’, Sumer 5 (1949): 276–96 for the entire history of this mosque. 6. N. Elisséeff, NËr al-Din, III, 650–7. See also Y. al-Tabbaa, The Architectural Patronage of NËr al-DÈn (1146–1174), upublished Ph. D. dissertation, New York University, 1982; esp. 23–6 and 143–57. 7. N. Elisséeff, NËr al-Din, III, 657–61. 8. AbË Shåma, Kitåb al-raw∂atayn, II, 478; ʿIzz al-DÈn b. al-AthÈr, Al-Kåmil fÈ al-tårÈkh, ed. C. J. Tornberg (Beirut, 1965) (reprint, with new pagination of the Leiden edition, 1851–76), XI, 362–3. 9. Ibn al-AthÈr, Kåmil, XI, 363–4. 10. Ibn Shaddad, SÈra Íalå˙ al-DÈn, 35, complete bibliographical data; AbË Shåma, Kitåb al-raw∂atayn, II, 482. 11. Edessa had fallen to the Crusaders in 1117 and had since become a Crusader county, the easternmost possession of the Crusaders. In 1144 it was conquered and sacked by ʿImåd al-DÈn ZankÈ, who, however, ordered a stop to the massacre of the local Christian population before its complete annihilation. See J. B. Segal, Edessa the Blessed City (Oxford, 1970), 300–13; and N. Elisséeff, NËr al-Din, II, 377–82; 396–401. 12. J. M. Fiey, Mossoul chrétienne. Essai sur l’histoire, l’archéologie et l’état actuel des monuments chrétiens de la ville de Mossoul (Beirut, 1959), 35–6. 13. Jews were also obligated to wear a distinctive mark, in the form of a red piece of cloth attached to the shoulder. 14. For the full translated text of the pact or covenant of ʿUmar, see for example, N. A. Stillman, The Jews of Arab Lands: A History and Source Book (Philadelphia, 1979), 157–8. 15. J. B. Segal, Edessa, 300–13.

153

154

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

16. Michael the Syrian, 3, 299 further comments on the corruptibility of Ibn AbÈ ʿAßrËn, who extorted bribes from monks and priests in return for not destroying newly-built structures. 17. N. Elisseeff, NËr al-DÈn, II, 661, proposes that NËr al-DÈn did not demolish nor confiscate the properties of any of the eight churches in Mosul. But this might just be an inference based primarily on NËr alDÈn’s strict adherence to Islamic law, which prohibits the destruction or confiscation of previously existing structures. 18. Michael the Syrian, 3, 298–9. 19. Ibid., 308. 20. Ibid., 299–300, where the author describes attacks by Kurds against the monastery Mår Mattå near Mosul. In fact, it seems that anti-Christian acts continued after the death of NËr al-DÈn, for several Christian writers report ‘a wave of pillaging that lasted three months in 1195, during which the “great sultan of the Turks” occupied Nisibis and Mosul’ (J. M. Fiey, Mossoul chrétienne, 37–8). 21. In fact, Michael the Syrian (3, 300) attributes a letter by NËr al-DÈn to the caliph of Baghdad that calls for nothing less than the forced conversion or outright slaughter of all Christians in Muslim lands. In this alleged letter NËr al-DÈn proposes that ‘the dictum of Mu˙ammad the Prophet, which is in the Qur’ån, that Muslims should not harass Christians for a period of five hundred years, has expired with the termination of these years. It follows then that all Christians in regions subject to Muslim control should be killed, unless they should convert to Islam.’ Interestingly, a similar proposal is attributed to the Fatimid caliph Al-Óåkim, who, when petitioned by Christians in 1013 to refrain from their oppression and destruction of their temples, allegedly declared that the four-century-old policy of tolerance had not produced the right results. He then gave the Christians the choice between conversion to Islam and ‘prompt punishment’ for those who refuse. See Ya˙yå b. SaʾÈd al-An†åkÈ, TårÈkh al-An†åkÈ al-ma’rËf bi-ßila tårÈkh U†Èkha, ed. A. U. TadmurÈ (Tripoli, 1990), 289–90, 295–9. 22. The story of NËr al-DÈn’s foundation of this mosque is told in some detail in AbË Shåma, Kitåb al-raw∂atayn, I, 20–1; II, 480. See also Sibt Ibn al-JawzÈ, Mir’åt al-zamån fÈ tarÈ¢ al-A’yån (Heyderabad 1370 AH) VIII, 282. Interestingly, it appears that some coins bearing the name of NËr al-DÈn were struck in Mosul during or around the time of his presence there. One coin, to my knowledge, has survived: Cairo Museum of Islamic Art, 2/17099, whose verso has the legend Al-Malik al-ʾÅdil Ma˙müd b. ZankÈ. It is published in M. B. al-ÓusaynÈ, Al-’umla al-islåmiyya fi al-’ahd al-atåbikÈ (Baghdad, 1966), 47. 23. Some writers, citing ‘a local Christian tradition’, have suggested the existence of a church on the spot where the mosque was built. See F. Sarre and E. Herzfeld, Archäologische Reise im Euphratund TigrisGebiet (Berlin, 1911–20), II, 216, where the alleged church is called St. Paul. See also N. Elisséeff, ‘Les monuments de NËr al-Din’, 34; and F. B. Flood, ‘The Medieval Trophy as an Art Historical Trope: Coptic and Byzantine “Altars” in Islamic Contexts’, Muqarnas 18 (2001), 31 n. 78, who cites Elisséeff. But, as DaywajÈ has previously noted in ‘Al-Jåmi’ al-NËrÈ’, 286, this legend is not mentioned by any reliable historian, all of whom unanimously speak of an empty lot in the middle of Mosul. 24. S. al-DaywajÈ, ‘Al-Jåmi’ al-NËrÈ’, 279–80 makes an attempt to determine the extent of these waqfs.

ˉ R AL-DIˉ N THE MOSQUE OF NU

25. AbË Shåma, Kitåb al-raw∂atayn, I, 20–1; and Sibt b. al-JawzÈ, Mir’åt al-zamån, VIII, 282. These chroniclers also report that during this visit, ʿUmar al-Mallå offered NËr al-DÈn, who was sitting by the Tigris river, the account books for the mosque for his verification. NËr al-DÈn allegedly responded: ‘daʾ al-˙isåb li-yawm al-˙isåb’ (Leave accounting for the Day of Judgement). He then tossed the account books into the river. 26. AbË Shåma, Kitåb al-raw∂atayn, I, 189. 27. This mi˙råb, which dates to the reign of Badr al-DÈn Luʾluʾ, is now kept in the Mosul Museum, along with several other magnificent stone mi˙råbs. 28. F. Sarre and E. Herzfeld. Archäologische Reise, II, 216. Herzfeld also dates the application of this thick plaster to the repairs of the 1860s, which is quite probable since it is unlikely that such a crude restoration would have occurred earlier. 29. Archäologische Reise, II, 216–31. This chronology was accepted by N. Elisséeff, ‘Les monuments de NËr al-Din’, BEO 13 (1949–50), 34; and idem, NËr al-DÈn, I, 109, where the author even attributes the minaret to 1148. The chronology is also repeated, with some caution, in R. Ettinghausen and O. Grabar, The Art and Architecture of Islam 650–1250 (New York, Viking Penguin, 1987), 298. 30. Archäologische Reise, II, 216–19. 31. Archäologische Reise, II, 224. 32. S. al-DaywajÈ, ‘Al-Jåmi’ al-NËrÈ’, 280–3. 33. For example, in the lists of pious foundations, which are normally included in the obituaries of important persons, not one medieval chronicler referred to the foundation of a mosque by Sayf al-DÈn. Ibn alAthÈr, Al-Kåmil, XI, 138–9 mentions the large madrasa Al-Atåbikiyya al-’AtÈqa and a khånqåh, but no mosque. The same information is repeated in AbË Shåma, Kitåb al-raw∂atayn, I, 168. Cf. N. Elisséeff, ‘Les monuments de NËr al-Din’, 34, where he repeats Herzfeld’s faulty chronology without citing any Arabic sources in its support. 34. See S. al-DaywajÈ, ‘Al-Jåmi’ al-NËrÈ’, 282–3; and idem, ‘Al-Jåmi’ alUmawi fÈ al-Mawßil’, Sumer 6 (1950), 216. The removal and reuse of mi˙råbs is not without precedent in Islamic architecture. After all, the so-called Khasseki mi˙råb, possibly dating to the second half of the eighth century, was discovered in a mosque founded in 1658. See F. Sarre and E. Herzfeld, Archäologische Reise, II, 139–44. In some respects, the mi˙råb of Qayrawan, which was first made in 836, but which is popularly attributed the original mosque of Sidi ʿUqba of 670, belongs to a similar phenomenon of the reuse of ancient mi˙råbs as cult objects. See, for example, K. A. C. Creswell, A Short Account of Early Muslim Architecture, revised and supplemented by J. W. Allan (Cairo, 1989), 325. 35. A longer examination of this mi˙råb is beyond the scope of this piece, but it should be noted that its outstanding vegetal ornament and multiple epigraphic styles are potentially quite informative about medieval stone carving in Mosul. All the inscription fragments are Qurʾånic, with the exception of a small fragment in the upper left hand corner, which states: ‘Al-Jåmi’ al-NËrÈ’, most likely referring to a later work of restoration. The epigraphic style looks late, possibly from the fifteenth or sixteenth century. See F. Sarre and E. Herzfeld, Archäologische Reise, I, 17–18. 36. It is even possible that ʿUthmån al-BaghdådÈ, the artisan of the 1148 mi˙råb, may have himself supervised the work on the columns. As

155

156

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

for the Badr al-DÈn mi˙råb, which is today preserved at the Mosul Museum, see Archäologische Reise, II, fig. 235 and III, pl. XCII. 37. Bundled brick piers are known from the Buyid rebuilding of the Masjid-È JåmÈ in Isfahan; see R. Hillenbrand, Islamic Architecture: Form, Function, Meaning (New York, 1994), fig. 2.270. 38. Archäologische Reise, II, 293–5 and figs 280 and 282. Most of these churches seem to have been restored during the relatively tolerant period of Badr al-DÈn. 39. Twenty-four original octagonal columns have been used in the modern mosque, while about ten more were made after their model. Restoration reports I briefly read in 1979 indicate that some of the more damaged columns were simply discarded. 40. Archäologische Reise, II, 226, and fig. 237. 41. These are: Imim Ya˙yå b. al-Qåsim (637/1239–40) and Imåm ʿAwn al-DÈn (646/1248). 42. Archäologische Reise, II, fig. 285. 43. For plans of these mosques, see R. Hillenbrand, Islamic Architecture: Harran, 2.40; Raqqa, 2.42; Dunaysir, 2.192; MayyåfåriqÈn, 2.204; Diyarbakir, 2.206. See also T. Sinclair, ‘Early Artuqid Mosque Architecture’, in J. Raby (ed.), The Art of Syria and the JazÈra, 1100–1250 (Oxford, 1985), 49–68 for more detailed plans. 44. A complete catalogue of such inscriptions is hardly necessary. But see Y. Tabbaa, ‘The Transformation of Arabic Writing, Part 2, The Public Text’, Ars Orientalis 24 (1994), figs 10–15 for comparable examples. 45. See E. Cruishank Dodd and Sh. Khairallah, The Image of the Word: A Study of Qur’ånic Verses in Islamic Architecture (Beirut, 1981), and II, 9–17 for a partial list of the occurrence of this verse. Sh. S. Blair, Islamic Inscriptions (New York, 1998), 73–4, 156–7, 195–6. 46. Sh. S. Blair, Islamic Inscriptions, 69. 47. Translations are from A. J. Arberry, The Koran Interpreted (New York, 1970). 48. On the jihåd of NËr al-DÈn against the Crusaders, see Y. Tabbaa, ‘Monuments with a Message: Propagation of Jihåd under NËr al-DÈn’, in Vl. Goss and Chr. Vezar-Bornstein (eds), The Meeting of Two Worlds: Cultural Exchange between East and West during the Period of the Crusades (Kalamazoo Ml, 1986), 223–40; and C. Hillenbrand, ‘Jihad Propaganda in Syria from the Time of the First Crusade until the Death of Zengi: The Evidence of Monumental Inscriptions’, in Kh. Athamina and R. Heacock (ed.), The Frankish Wars and their Influence on Palestine (Birzeit, 1994), 60–9. 49. See Y. Tabbaa, ‘Monuments with a Message’, 229–31 and fig. 24. 50. I have generally abstained from over-interpreting Qurʾånic passages since their meaning changes historically and in different contexts. On this issue, see Sh. S. Blair, Islamic Inscriptions, 215–17. For this specific interpretation, I have relied on The Holy Qur’ån: English Translation of the Meanings and Commentary, rev. and ed. by The Presidency of Islamic Researchers (Medina, 1992), 60–1. 51. Archäologische Reise, I, 18. The palaeographic style is closely linked to other inscriptions from the second half of the twelfth century. 52. See Sh. S. Blair, Islamic Inscriptions, 198: ‘With its emphasis on attestation, [3:18] is particularly apt on tombstones, whose inscriptions used the same verb, saying that the deceased attested.’ 53. E. Herzfeld (Archäologische Reise, II, 226–7) dated this stucco panel to

ˉ R AL-DIˉ N THE MOSQUE OF NU

a restoration allegedly carried out by Uzun Hassan in the middle of the fifteenth century. On the basis of comparable stucco work in Iran and in Mosul, that attribution now seems much too late, and a date from the period of Badr al-DÈn Luʾluʾ (1233–59) or even from the original foundation of the mosque in 1170 seems much more likely. The earlier date is quite problematic, however, because it would have to rest completely on comparison with stucco work in eleventh- and twelfth-century Iranian monuments: see, for example, A. Hutt and L. Harrow, Iran I (London, 1977) pls 55–6 (Great Mosque at Ardistan) and pl. 72 (madrasa Kuh-È Banan). The later date seems more likely because other monuments built or restored by Badr al-DÈn contain similarly lavish stucco or even stone work, including the relief panels at the Mår Behnam monastery outside Mosul and the astonishing marble revetments at the shrines of Imåm Ya˙yå b. al-Qåsim and Imåm ʿAwn al-DÈn. 54. K. A. C. Creswell, Early Muslim Architecture, I, 348–9 and figs 411–12. 55. Badr al-DÈn Luʾluʾ has not received sufficient scholarly attention. See meanwhile D. Patton, Badr al-DÈn Lu’lu’, Atabeg of Mosul, 1211–1259 (Seattle and London, 1991), esp. 67–9 for a discussion of the ecumenical policy practised by Badr al-DÈn. 56. R. Hillenbrand, Islamic Architecture, 153. See also J. Bloom, Minaret, Symbol of Islam (Oxford, 1989), 19, 66 and 73 for the significance of height in minarets. 57. Although the question of height in minarets has been discussed by Bloom (see n. 55), no one to my knowledge has attempted systemically to link it to the sectarian situation obtaining at the time of the creation of these minarets. But a cursory look at minarets in different parts of the Islamic world seems to suggest such a linkage, though not a congruence. Therefore, minarets in regions that have historically been devoid of large Christian minorities, such as Yemen, Arabia and Libya, tend to be short; and minaret built on the ‘borders’ of Islam, such as India, Afghanistan and Spain, tend to be tall. Elsewhere, the situation varies. 58. See G. Bell, The Churches and Monasteries of the ÊËr ‘AbdÈn (London, 1982), VII, where M. Mundell Mango refers to: ‘the fifty-odd churches built and rebuilt in the twelfth century by the bishop of Mardin’. 59. ‘The public Text’, 142–8. 60. I have previously alluded to NËr al-DÈn’s direct involvement in the nature of sermons given in the various mosques in his domain and how these sermons may have been reinforced by the inscriptions in these mosques. See Y. Tabbaa, ‘Monuments with a Message’, 229–35; and idem, The Transformation of Islamic Art during the Sunni Revival (Seattle and London, 2001), 22–3. 61. In these respects and others, this trend represents a reversal of Fatimid architectural practice, whose propagandistic message remained encrypted within enigmatic symbols and a nearly indecipherable script.

157

CHAPTER EIGHT

The Resurgence of the Baghdad Caliphate Baghdad of the later ʿAbbasids (1050–1258) presents the art historian with multiple paradoxes that do not offer easy solutions. Its legendary image as the Round City surrounded by fabulous palaces is difficult to reconcile with a completely altered medieval urban form. Its ruler, the ʿAbbasid caliph himself, possessed less military or political power than nearly any of the non-Arab sultans or governors who ruled most of the central Islamic lands in the period under consideration. And yet, the ʿAbbasid caliphate remained to its end the source of legitimation for these arriviste dynasties and the symbolic safeguard of the Islamic world. Finally, its once great monuments have largely vanished, and the ones that have survived do not match those in contemporary Cairo or Isfahan in number, size or luxury. There are several explanations for the relative dearth of Islamic monuments in Baghdad and the absence of a unifying congregational mosque in it, including destructive invasions, fragile building materials, the limited building activity of later caliphs and the sectarian nucleation of later medieval Baghdad. There is little question, in the first place, that the devastating effects of various invasions, particularly the Mongol invasion of 1258, contributed to the destruction of medieval monuments. Second, the relatively fragile brick architecture in Baghdad has not withstood the exceedingly harsh climate, with brutally hot summers, rainy winters, and especially, high water table. Third, unlike their architecturally prodigious ancestors of the eighth and ninth centuries, later ʿAbbasid caliphs generally retreated behind their palace gates, leaving architectural patronage to Seljuq sultans and governors, with only minimal participation by patricians and court officials (Le Strange 1901: 331). Finally, Baghdad appears in the Middle Ages as a nucleated city divided along sectarian lines, with each sect centred around its own places of cult and worship:

Yasser Tabbaa (2017),‘The Resurgence of the Baghdad Caliphate (1050–1258), in F. Barry Flood and G. Necipoglu (eds), A Companion to Islamic Art and Architecture, 2 vols, Oxford: Wiley Blackwell, 206–26.

the resurgence of the baghdad caliphate

SunnÈs in Rusafa and al-Aʾzamiyya; Shiʿis in Karkh and al-Kazimiyya; with the shrines of Sufi saints, including ʿAbd al-Qadir al-Gailani and ʿUmar al-Suhrawardi providing other centres of settlement. (Ibn Jubayr 2001: 226–8) Brief political and urban history When the Great Seljuqs under Tughril Beg took over Baghdad in 1055, they sought to dress their takeover of the caliphate under the guise of the restoration of the SunnÈ caliphate. The intricate relationship between the ʿAbbasid caliphate, the Seljuq sultanate and their common adversary, the Fatimid caliphs of Cairo, would largely define the course of the ʿAbbasid caliphate and its capital until the middle of the twelfth century, after which the ʿAbbasids witnessed a period of revival and autonomy that would end with the Mongol invasion of 1258. As such, the history of the late Baghdad caliphate falls into two nearly equal periods: the period of Seljuq domination and SunnÈ revival (1055–c. 1150), and the period of the resurgent caliphate and SunnÈ ecumenism (c. 1150–1258). Although it is sometimes asserted that the Great Seljuqs were the primary force behind the SunnÈ revival, traditionalist policies had in fact already been quite actively pursued by the caliph al-Qadir (r. 991–1031), whose Qadiri creed would become the cornerstone of the new ʿAbbasid orthodoxy and the official dogma of the caliphate.1 Qadirism stood for a traditionalist SunnÈ theology under the spiritual leadership of the ʿAbbasids, while opposing all rationalist interpretations of the dogma, in particular those put forth by the Ismaʾili Fatimids. The Qadiri creed remained the cornerstone of the new ʿAbbasid orthodoxy to the end, but it was modified in the middle of the twelfth century into an ecumenical SunnÈ creed that aimed to unite the four schools of jurisprudence (madhhabs) against Ismaʾilism. Interestingly, Baghdad’s new status as a city state in the twelfth century finds further confirmation in its fortified urban form, whereby its most populous eastern part, largely Rusafa, became enclosed within a wall. Erected by the caliph al-Mustarshid (r. 1118–35) and frequently restored in subsequent centuries, this wall continued to enclose and define Baghdad proper until the nineteenth century (Ibn al-Athir 1987: vol. 9, 28). Within it, Rusafa/East Baghdad assumed a rectangular form: bordered on the south-west by the Tigris River; anchored at its western corner by a citadel; and surrounded on its north-western, north-eastern and south-eastern sides by a curtain wall with round towers and four gates. Commanding the middle of this rectangle and facing the Tigris on the south-west was the Dar al-Khilafa (Palace of the Caliphate), which seems to have been a semi-circular enclosure. Facing Rusafa on the opposite bank of the Tigris stretched the smaller walled city of Karkh, while about 1 km upstream stood the even smaller settlements of al-Aʾzamiyya

159

160

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

and al-Kazimiyya around the shrines of the imam Abu Hanifa, founder of one of the four schools of SunnÈ jurisprudence, and of the two Shiʿi imams, Musa al-Kazim and Muhammad al-Jawad. The second phase of the late ʿAbbasid caliphate effectively began with the death of the Seljuq sultan Masʾud during the caliphate of al-Muqtafi (r. 1136–60), at which point the sultanate had already lost much of its authority. Further enforcing this trend, the two succeeding caliphs al-Mustanjid (r. 1160–70) and al-Mustadiʾ (r. 1170–80) ruled Iraq independently while enjoying cordial and mutually supportive relations with the Zangid ruler of Syria NËr al-DÈn (r. 1146–74) and his Ayyubid general (and later ruler in his own right) Saladin, who in 1172 had brought down the Fatimid caliphate and restituted the ʿAbbasid khutba (Friday sermon) in Egypt. These significant developments prepared the way for the caliph al-Nasir, who acceded to the throne in 1180 and ruled for forty-five years until 1225, making him the longest-reigning ʿAbbasid caliph. Politically, his reign signals the end of the rule of the Seljuq Turks in 1194, when they were ousted from Iran by the Khwarazmshahs, who unsuccessfully attempted to replace them as sultans. This gave al-Nasir greater freedom to assert his uncontested rule over Iraq, from Tikrit to Khuzistan. In addition to his political gains, al-Nasir reversed his father’s repressive measures against Shiʿis and practised a more inclusive and purposeful policy that aimed to foster greater cohesion between SunnÈs and Imami Shiʿis.2 More than anything al-Nasir is known for his patronage of the futuwwa, an urban social group that had long existed in Islamic cities but that was reorganised by him along Sufi-ʾAlid lines and made into a chivalric order and an instrument of government. The futuwwa also promoted a sense of Arab identity and anti-Turkish bias, based on language, common lore and allegiance to the ʿAbbasid caliphate. This, in turn, may account for the astonishing popularity of the typically Arab prose works of the Maqamat (Assemblies) genre in the twelfth and thirteenth centuries (see below and Contadini 2017, ch. 17). Al-Mustansir (r. 1226–42), the penultimate ʿAbbasid caliph, is best remembered for his architectural patronage, in particular his foundation of the Madrasa al-Mustansiriyya on the banks of the Tigris in 1233. He was succeeded by his son al-Mustaʾsim, whose reign would end in 1258 with the Mongol invasion and the final collapse of the ʿAbbasid caliphate in Baghdad. Although the destructive nature of the Mongol invasion is sometimes exaggerated, there is little question that they razed the caliphal city and tore down the Mosque of the Caliphs, which was founded in the tenth century, forever ­depriving Baghdad of a unifying congregational mosque.

the resurgence of the baghdad caliphate

Fortifications Until the beginning of the twentieth century, significant sections of the Baghdad enclosure still existed, consisting of a very thick brick curtain wall, several round towers, a deep moat connected with the Tigris, and four main gates. Of these gates – Muʾazzam (North) Gate, Dhafariya (Wastani) Gate, Halaba (Talisman) Gate and Basaliya Gate – only the second and third have survived long enough for documentation and study. The most impressive of these gates, the Talisman Gate, consisted, before its destruction in 1917, of a massive cylindrical tower with a flat façade dominated by a single pointed arch, which had already been walled up in the seventeenth century (Le Strange 1901: 291). First built sometime around the middle of the twelfth century, it was rebuilt in 1221 by al-Nasir and renamed the ‘Talisman’ Gate after the unusual figural relief sculpture carved above its central arch. Springing from a pair of recumbent lions, this arch is surmounted by a fine relief sculpture that shows a crowned cross-legged figure wearing the baggy trousers associated with the futuwwa, with arms stretched to either side, grasping with his hands the tongues of two dragons whose plaited serpentine bodies fill the entire panel. Composed over a background of very fine arabesque and flanked on both sides by a magnificent inscription in thuluth script that gives the foundation date and full titulature of al-Nasir, this gate was one of the finest works of medieval Islamic military architecture. The iconography of its quite unique relief sculpture has been interpreted as a victory monument that commemorates al-Nasir’s victories in the late twelfth century over the Seljuq sultan of Iran, Tughril III, and the Khwarazmshah, ʿAlaʾ al-Din Tekish.3 Dhafariya (Wastani) Gate, located in the middle of the northeastern wall, across a modern street from the mausoleum of ʿUmar al-Suhrawardi, is the only extant gate of the entire Baghdad enclosure. Like the Talisman Gate, it consists of a circular tower (14.5 m high with a circumference of 56 m) that is effectively surrounded by a moat on all sides. Two bridges, at 90 degrees to each other, each resting on two arches, link the tower simultaneously to the city and to the surrounding countryside, turning the single round tower into a bent-axis gate. Internally, the tower comprises a domed central octagon with eight radiating vaulted chambers that support an upper level of defence with loopholes and brattices along the entire circumference of the tower. As with the Talisman Gate, the tower was crowned by a large inscription, dated 618 (1221), which is still partially preserved. These two inscriptions are noted for their size, length and calligraphic quality, features that attest to the superiority of calligraphy in Baghdad, as we shall see below. The two entrances of this tower, intra-mural and extra-mural, have been squared off in order to create monumental portals, of

161

162

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 8.1  Baghdad. Dhafariya (Wastani) Gate, 1221, from north. Source: Yasser Tabbaa. Reproduced with permission (photo: Yasser Tabbaa)

which the inner one retains much of its original composition and decoration (Figure 8.1). Slightly recessed within the surrounding brickwork, the portal rises from a keel-shaped arch whose voussoirs spring from a pair of recumbent lions, a very common feature in this period for towers and gates and generally interpreted in terms of power and protection (Rabbat 2006: 94–9). The arch is topped by a rectangular panel with a complex geometric pattern of large twelvepointed stars surrounded by alternating squares and rhombuses, over delicate vegetal arabesque patterns. The Dar al-Khilafa, the royal palace-city of the ʿAbbasid caliphs, occupied about one quarter the area of East Baghdad, making it a city within a city, an urban situation that recalls Fatimid Cairo (see Anderson and Pruitt, ch. 9). More of a palatial enclosure than an actual palace, the semi-circular grounds were surrounded by a wall pierced with nine gates, of which one was a secret subterranean gate and another led directly to the Mosque of the Caliphs (Masjid al-Khulafaʾ). Completely vanished and rather poorly described in the sources, it seems that the Dar al-Khilafa comprised pavilions and kiosks facing the Tigris, including al-Hasani, al-Rayhan, al-Firdaws and al-Taj.4 The Harem quarters were at the opposite end, with gardens in between (Le Strange 1901: 263–8). At its northern edge, the Palace of the Caliphate was linked with the Mosque of the Caliphs, first built by al-Muqtafi in the middle of the twelfth century but restored by al-Mustansir, from whose time the minaret still remains.

the resurgence of the baghdad caliphate

Religious architecture Baghdad presents a very incomplete picture of religious architecture in comparison to other Islamic cities, where a measure of architectural continuity is often provided by the congregational mosque of the city. Lacking a composite multiphase mosque – such as the Masjid-i Jamiʾ in Isfahan, the Great Mosque of Damascus or the alAzhar Mosque in Cairo – and the numerous pious institutions typical of medieval Islamic cities, later ʿAbbasid architecture in Baghdad is largely defined by two important thirteenth-century monuments, two tombs with conical muqarnas domes, and a handful of minarets. Furthermore, all these structures date to the first half of the thirteenth century, after the decline of Seljuq power and before the coming of the Mongols in 1258. In terms of minarets, of all the mosques built by the ʿAbbasid caliphs and Seljuq sultans, what remains is a handful of minarets, two of which date from the pre-Mongol period, and a third, once belonging to the Mosque of the Caliphs, was rebuilt in the early Mongol period. The two late ʿAbbasid minarets – belonging to the mosques al-Khaffafin and Qumriyya and dating to the first three decades of the thirteenth century – are nearly identical in their baked brick building material and overall form. Both begin from a polygonal base, largely hidden today within later mosques, proceed to thick cylindrical shafts that develop into a corbelled muqarnas zone supporting a balcony, and end in a tapered cylindrical shaft capped by a little dome. Also both are entered from the roof of the mosque rather than at the ground level. The al-Khaffafin mosque, located just north-west of the Madrasa al-Mustansiriyya, was first founded by Zumurrud Khatun, wife of al-Mustadiʾ and mother of al-Nasir, before her death in 1202 (Janabi 1982: 61). The original mosque has entirely vanished, having been replaced by an undated mosque from the Ottoman period, to which must also date the geometric tile-work at the top of the minaret. Similarly, the minaret of the Qumriyya mosque is undated and has been subsumed within a later mosque, but according to textual sources the original mosque was completed in 626 (1228), which must also be the date of the minaret. Bulky and rather poorly proportioned in comparison to its predecessor, the Qumriyya minaret nevertheless shows some delicate hazar-baf (thousand weaves) brick ornament, which is arrayed as rhomboid squares that enclose small crosses. In addition to mosques, Baghdad was provided with a number of significant funerary monuments, particularly tomb towers. In a well-known Persian painting dated 1475 and entitled ‘The Flood of Baghdad’, the artist shows a bird’s-eye view of both banks of the city with numerous conical domes dotting its landscape.5 These conical domes, sometimes also called muqarnas domes, most likely originated in Baghdad sometime in the eleventh century, although

163

164

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 8.2  Baghdad. conical dome of ʿUmar al-Suhrawardi, early thirteenth century (photo: Yasser Tabbaa)

the earliest existing example – the Imam Dur of 1088 – is located in Samarra.6 Two of these Baghdadi conical domes have survived – the mausoleum of Zumurrud Khatun (Figure 8.2), just beyond the edge of Karkh; and the shrine of ʿUmar al-Suhrawardi, very near the Wastani Gate – extra-urban locations that seem to fit their funerary status. Most likely founded by al-Nasir for his pious and charitable mother, the mausoleum of Zumurrud Khatun consists of an octagonal base from which springs a nine-layered conical muqarnas dome capped by a small cupola. Each face of the octagonal base is divided grid-like into four panels, the lower two rectangular and the upper square, topped by a narrow frieze. All panels are decorated with geometric brickwork in the hazar-baf mode with some geometric strapwork.

the resurgence of the baghdad caliphate

Today, the tomb is entered from a square domed chamber whose ribbed vaulting suggests a Timurid rebuilding of an earlier structure. The walls of the interior octagon, which correspond to the exterior octagon, are perfectly vertical to the height of 9.5 m, above which begin nine zones of corbelled and staggered muqarnas zones that end in a tiny eight-pointed star. Each muqarnas cell has been cut and covered in thick glass whose inner glow produces a dazzling effect unknown in other muqarnas domes. The shrine of the Sufi shaykh Shihab al-Din ʿUmar al-Suhrawardi (d. 1234) occupies the focal point of a shrine complex that was first founded soon after Suhrawardi’s death. It was subsequently augmented with a beautiful Ilkhanid façade in 735 (1334), an Ottoman mosque, and later restorations in the nineteenth and twentieth centuries. These significant additions to the thirteenth-century shrine demonstrate the affection and veneration that were accorded to this pivotal Sufi figure, who became the official Sufi master of Baghdad during the caliphate of al-Nasir (Schimmel 1975: 245–6). The actual mausoleum is an even more attenuated version of Zumurrud Khatun’s tomb, resembling a conical tower more than a dome. Internally, the original conical vault is today largely invisible owing to the addition of a hemispherical dome, possibly under the Ottomans, just above the first muqarnas zone (Hadithi and Khåliq 1974: 39–43). Among the madrasas of Baghdad one in particular, the Madrasa alMustansiriyya, achieved renown throughout the Islamic world. Six years in construction, from 625 (1227) to 631 (1233), this madrasa was seen by contemporary writers as nothing less than a wondrous creation (Figure 8.3). Sibt Ibn al-Jawzi (d. 1256) wrote, ‘There is no equal to this madrasa in the world; and nothing like it has been built in previous years’, a sentiment repeated by Ibn al-Fuwati (d. 1323), who actually taught at this madrasa. Ibn Battuta in 1327 wrote, ‘Its appearance is unique and its elegant organization miraculous, lofty to the reaches of heaven . . . It is the Ka’ba of mankind and the dome of Islam.’ Other writers, including al-Dhahabi (d. 1348), Ibn Kathir (d. 1373), and Ibn ʿAnbah (d. 1426) spoke glowingly of its vast and unique library, its well-equipped stationery store and the personal attention it was accorded by the caliph himself. In more recent times, the Madrasa al-Mustansiriyya has been studied by Ernst Herzfeld, K. A. C. Creswell, Hansjorg Schmid, and by several Iraqi scholars.7 The importance and uniqueness of the Madrasa al-Mustansiriyya can be further attested in the long and detailed list of jurists and luminaries who taught in it over several centuries8 and especially in the great pomp and ceremony that accompanied its dedication. Ibn al-Athir wrote: On Thursday the fifth of Rajab (631/1233) Nasir al-Din, the Vice Minister, and all the governors (walis), chamberlains (hajibs), judges, professors, and jurists; and the shaykhs of ribats and Sufis,

165

166

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 8.3  Baghdad. Madrasa al-Mustansiriyya, completed 631/1233, courtyard from east (photo: Yasser Tabbaa)

preachers, Qurʾan readers, poets, and a group of the select among the merchants and the foreigners were present at the madrasa. The Vice Minister selected for each madhhab [law school] in the madrasa 62 students, and appointed for it two professors and two assistant professors . . . Each professor was presented with a black cloak, a dark blue robe, and a fully caparisoned mule. As for the assistants . . . each was offered a white coat and a silver-threaded turban . . . A dinner was then spread in the entire courtyard of the madrasa. (1987: vol. 11, 284) The Madrasa al-Mustansiriyya is located right on the Tigris River, about 100 m upstream from the outer wall of Dar al-Khilafa and about half that distance from the Seljuq Madrasa al-Nizamiyya. An approximate rectangle – 104.8 m long with a width that expands from 44.2 m to 48.8m at the southern end – the Madrasa al-Mustansiriyya was larger and more lavishly built than any other madrasa of the first half of the thirteenth century (Schmid 1980: pl. 15). Unlike Syrian and Egyptian madrasas and more like Iranian ones, it was intended to be freestanding, as is clear from its lightly ornamented exterior brickwork; the inscription bands on three sides, of which only the one facing the river has survived; and the monumental portal in the middle of its long north-eastern side. For centuries hidden behind a bazaar that extended along its entire entrance façade, the lofty portal now rises about 16.5 m above a narrow street and projects

the resurgence of the baghdad caliphate

from the flanking walls by 3.50 m. A rabbeted frame, dominated by a large twisted cable moulding, articulates this portal, which takes on the form of a shallow pointed iwån that encloses a tympanum with a foundation inscription. This ten-line inscription gives the name and titles of the founder and the date of foundation; includes an appropriate Qurʾånic inscription (Qurʾån: 18:29) about the reward that awaits those who perform good deeds; and specifies that the madrasa is intended for the four SunnÈ madhhabs. A much longer and larger inscription band on the river side remains but has been largely redone during the reign of the Ottoman sultan ʿAbdülaziz in 1282 (1865). Extending for nearly 100 m, the inscription frieze stands above a series of ornamented blind niches with a shouldered arch, all reinforcing the idea that the madrasa was intended to be accessed from the Tigris by boat as well. The interior of the Madrasa al-Mustansiriyya is disposed around a perfectly regular courtyard, balanced on its short ends by tall and spacious iwåns and in the middle of its long sides by the entrance iwån and the tripartite façade of the prayer hall, a straightforward but elegant adaptation of the four-iwån plan that became standardised under the Seljuqs in Iran. All these four focal points are flanked by two levels of superimposed pointed arches that lead to rectangular rooms, totalling thirty-six on each floor, although at least six of these spaces also accommodated staircases linking the two floors. Vaulted corridors on the upper level extend on both sides of the iwåns and the prayer hall, providing access to the upper chambers while also reducing their size by about one-fifth. The lower rooms, which are larger and more easily accessible, were most likely intended for teachers, while the upper ones were probably for students. Beyond its courtyard, the Madrasa al-Mustansiriyya contained on its northwestern side a warren of rooms that may have been the hospital (bÈmåristån) (Mashhadani and al-Naqshbandi 1986: 77–9). Also on this side there are the remains of an iwån that adorses the internal iwån and that may have once faced a garden that would have extended to the masnat (breakwater) known to have existed there. The chambers on the opposite, north-eastern side of the madrasa are larger, better organised, and more accessible through two openings in corners of the courtyard that lead to a tall, narrow vaulted corridor that extends the full width of the structure. Covered by sail vaults with a central skylight, these seven spacious chambers are generally identified in the secondary literature with the library, scriptorium and study halls (Mashhadani and al-Naqshbandi 1986: 97–109). An architectural and photogrammetric survey of the Madrasa al-Mustansiriyya by Hansjorg Schmid fully documents the existing form of the building and demonstrates that the structure was built according to sophisticated rules of geometric planning and harmonic proportions. In terms of the ground plan, Schmid uncovered the foot measure that may have been used in designing the structure (30 cm),

167

168

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 8.4  Baghdad. Madrasa Mustansiriyya, 631/1233, courtyard from east (photo: Yasser Tabbaa)

demonstrating that, with fairly minor adjustments, the plan was generated as whole number multiples of this foot unit, such that, for example, the width of the mosque was 80 feet and that of the two main iwåns 30 feet. Applying this unit measure to the elevations of the courtyard units, the profiles of their arches and the vaulting of the ancillary units, Schmid proposed that these measurements were equally based on whole or fractional multiples of the same foot measure. The precision of the Madrasa al-Mustansiriyya’s geometric planning is well attested in Iranian brick-built monuments, ranging from the Seljuq to the Timurid periods and beyond (Golombek and Wilber 1988: 37–59). Secular architecture: the ʿAbbasid palace When Herzfeld examined the so-called ʿAbbasid Palace in the early twentieth century, he simply referred to it as Iwan al-Qalʾah (Iwan of the Castle), since only an iwån facing an irregular cruciform courtyard seems to have remained from the original structure. The northern side of this courtyard was occupied by a longitudinal structure covered by six small domes, suggesting an Ottoman building. Several episodes of excavation, restoration and rebuilding between 1933 and c. 1970 have uncovered other important wings of this building. Curiously, these archaeological efforts have failed to uncover any epigraphic evidence that world ascertain the name and date of the ‘Abbasid Palace’, an

the resurgence of the baghdad caliphate

appellation that has been disputed by several scholars who identified it instead as a madrasa, specifically al-Madrasa al-Sharabiyya. For the sake of clarity, this piece will accept the original identification of this monument as a palace before addressing in the conclusion those arguments that have identified it as a madrasa. The ʿAbbasid Palace is a two-storey brick structure situated on the Tigris River at the western corner of the city wall, in what would have been the citadel of the city and later the Ottoman governor’s place. A nearly square structure centred around a square courtyard, the palace is entered through a rather inconsequential portal on its eastern side and an impressive but largely rebuilt portal on the southern side, which faces the Tigris. With its rabbeted convex mouldings, a low doorway surmounted by a much taller arch and a panel for inscription in the tympanum, the portal was clearly rebuilt after the model of the Madrasa al-Mustansiriyya’s portal. Its rebuilt status notwithstanding, this was undoubtedly the main original entrance to the palace, a conclusion ascertained by the well-preserved vestibule to which it leads. This vestibule, in turn, takes to two bent-axis passages, the western into the corner of the courtyard and the eastern to a long corridor that connects a series of independent rooms. The square courtyard is dominated on its eastern courtyard by a large iwån whose keel-shaped pointed arch is double the height of the chambers that flank it and that surround the courtyard on three sides. Arranged on two levels, the approximately forty chambers are fronted by a double arcade resting on square piers, behind which runs a narrow corridor. Facing the iwån is a spacious rectangular chamber (13.8 × 4.5 m), which could be read as a mosque, were it not for its incorrect orientation and the absence of a mi˙råb niche in its long wall. Although heavily restored, there is little question that the ʿAbbasid Palace was thoroughly decorated, in the grand style of al-Mustansiriyya, with geometric strap-work and minutely carved brick arabesque ornament, in addition to magnificent muqarnas vaulting. The best preserved passages of original ornament are found in the vestibule, the vaulting of the iwån, and especially the vaulting of the eastern corridor, while all other parts of the structure have been rebuilt and decorated in a similar style. Upon entering the vestibule, one is immediately struck by its fine and opulent ornament, which builds up in a series of carved muqarnas cells that support a flat ceiling, a splendid masterpiece of carved brick decoration. Interlocking eightpointed stars alternating with swastikas form the basic design, which is further embellished by carving minute arabesque and geometric patterns in every cell and even along the ribs of the strapwork, creating a uniquely rich and tactile effect. The tall and spacious iwån (5 × 8.5 × 9.15 m) is framed by a rabbeted pointed arch which is itself enclosed within a rectangular frame. The exterior ornament is all recent restoration, but original ornamental

169

170

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

passages remain on the tympanum and especially the vault of the iwån, whose two halves are symmetrically decorated in geometric patterns whose cells are carved in minute arabesque patters. The design emanates from a circular acanthus whirl surrounded by a ring with sixteen raised bosses and ends in four arrow-shaped medallions, a carpet-like composition that imparts a sense of order and rigour to the intervening geometric ornament. Even more stunning is the vaulting of the eastern corridor, which effectively consists of a series of interlocked small muqarnas vaults that rise as large corbelled cells from the piers surrounding the courtyard and from the inner wall and end each in an eight-pointed star. Each cell is minutely decorated in arabesque ornament, producing an exquisitely tactile design. Although muqarnas vaultings had become commonplace throughout the Islamic world by the first half of the thirteenth century, this muqarnas-vaulted corridor appears without parallel and once again points to the excellence and originality of late ʿAbbasid architecture and ornament (Figure 8.5) (Tabbaa 2001: 123–4). Most scholars have dated this structure on stylistic grounds to the period of either al-Nasir or al-Mustansir, but many have rejected its original function as a palace, proposing that it was originally a madrasa, specifically the Madrasa al-Sharabiyya.9 Arguing for its origin as a madrasa are its stylistic similarities to Madrasa alMustansiriyya and the large number of side rooms on two levels, normally used to house students. But there are more compelling reasons to argue for a palatial origin. First, the name in itself is quite curious for, to my knowledge, there is not a single madrasa that has been renamed a palace. Second, the location of the ʿAbbasid Palace within the Baghdad citadel precinct, and mainly accessible from the riverside, seem to argue for a palace rather than a madrasa. Third, the historical sources are completely silent about this alleged madrasa: they do not specify a madrasa in this location; they say nothing about its caliphal inauguration; and most seriously, they do not offer the usual list of teachers who have taught in this institution. Alternatively, there is a handful of historical sources that mention a so-called Dar al-Masnat (House by the Breakwater) at the citadel, built by al-Nasir c. 580 (1184).10 This could very well be ʿAbbasid Palace, begun by al-Nasir and completed by al-Mustansir. There are also some ambiguities absences in the architecture of this building that seriously undermine its designation as a madrasa, including its incorrect orientation to the qibla by around 30 degees, the absence of a concave mi˙råb in its alleged mosque, and its entry from the riverside rather than the city. Furthermore, the entire building is anepigraphic, lacking a foundation inscription, Qurʾånic passages, and later inscriptions of rebuilding of waqf (endowment) augmentation, which is very unusual for a prestigious madrasa such as this one.

the resurgence of the baghdad caliphate

Figure 8.5  Baghdad. ʿAbbasid Palace c. 1200, muqarnas vaulted portico (photo: Yasser Tabbaa)

Even the close architectural affinities between the ʿAbbasid Palace and the Madrasa al-Mustansiriyya do not conclusively argue for such an identification since medieval madrasas and palaces are often typologically related. Thus, the ʿAbbasid Palace was most likely a palace that was built either by al-Nasir or al-Mustansir, a re-identification that potentially contributes to our incomplete understanding of the topology of medieval Islamic palaces. The arts of the book during the late Baghdad Caliphate No decorative arts in pottery, metalwork, glass, woodwork or ivory can be securely attributed to late ʿAbbasid Baghdad. As such, this discussion will focus solely on the arts of the book: calligraphy, which enjoyed a glorious career under the early and late ʿAbbasids;

171

172

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

and illustrated manuscripts, whose short-lived efflorescence closely parallels the architecture discussed above. Two Iraqi cities, Kufa and Baghdad, have largely shaped the course and trajectory of Islamic calligraphy between the eighth and thirteenth centuries – Kufa in the first three centuries and Baghdad from to the tenth to the thirteenth century and beyond (see George 2012, ch. 4). Much of this development pre-dates the middle of the eleventh century, where this piece begins, including the rise of ʿAbbasid KËfic Qurʾånic calligraphy toward the end of the eighth century; the definitive reforms of Ibn Muqla at the beginning of the tenth century, that may have led to the development of the so-called New Style script (also known as Broken or Eastern KËfic); and the elegant resolution of these reforms at the hand of Ibn al-Bawwab at the eve of the eleventh century (Tabbaa 1991: 119–47). Having said that, there is little question that the legacy of these two calligraphers continued into the eleventh century, when Qurʾåns were simultaneously written in the New Style script and increasingly in the more cursive scripts attributed to Ibn al-Bawwab, including naskh and thuluth, which were produced by the systematisation of the cursive scripts of the ʿAbbasid chancery. Several eleventh-century Qurʾån manuscripts, written in Baghdad and in some Iranian cities, continue to be written in the semi-KËfic script, whose use for writing the Qurʾån should be attributed to Ibn Muqla. Written on paper with a vertical format, these rather modest Qurʾåns generally employ an austere and closely packed script characterised by vertical uprights, angular ligatures and somewhat triangular openings. Fully vocalised and with complete and consistent diacritical marks, the semi-KËfic script was intended to be read by any literate person, a striking difference from the earlier KËfic manuscripts. Some eleventh-century Qurʾån manuscripts, starting with the unique extant Qurʾån dated 391/1000 and calligraphed by the celebrated Ibn al-Bawwab, are written in the new cursive styles that became dominant in the twelfth century. Generally speaking, the cursive Qurʾån manuscripts of the eleventh and early twelfth centuries show the closest affinities with Ibn al-Bawwab’s robust naskh script for the text and his highly sinuous thuluth for the chapter headings. Typically, these manuscripts begin with a decorative single or double frontispiece of plaited or geometric ornament that includes a count of the chapter, verses and sometimes words of the Qurʾån, and end with a colophon naming the calligrapher and the date and place of production. For example, Chester Beatty Library 1430 – a complete Qurʾån manuscript dated 428 (1037), most likely produced in Iraq – and TIEM (Türk ve Islam Eserleri Müzesi) dated 449 (1057) – also a complete Qurʾån attributed to Baghdad and falsely signed in the name of Ibn al-Bawwab – steer very closely to the master calligrapher’s naskh calligraphy (Rice 1955).

the resurgence of the baghdad caliphate

Qurʾån manuscripts produced in Baghdad and further east in the late twelfth and thirteenth centuries continue some earlier calligraphic practices but become considerably larger and more lavishly decorated. The best-known calligrapher of this period is the legendary Yaqut al-Mustaʾsimi (1221–98), whose career survived the disastrous end of the ʿAbbasid caliphate and the murder of his patron, al-Mustaʾsim, for he continued to produce Qurʾåns after the Mongol sack of Baghdad in 1258, and well into the period of the Ilkhans, who governed the city on behalf of the Mongols. According to the Iranian anthologist Qadi Ahmad (d. 1610), Yaqut, whom he calls ‘the cynosure of calligraphers’, ‘followed the tradition of Ibn al-Bawwab, but in the trimming of the qalam and in the clipping of its nib he altered the manner of the earlier masters’ (Qadi Ahmad 1959: 57–8). Cutting the nib at a characteristic angle made for a finer and more elegant cursive script, best seen in two of his Qurʾåns written in the rayhan script with thuluth chapter headings. Yaqut is reputed to have written numerous copies of the Qurʾån, and a handful of these have survived, along with fragments and calligraphic specimens signed by him. He is said to have produced six famous calligraphers who spread his calligraphic method to Iran and Turkey, where it was perpetuated well into the sixteenth century. Indeed, the unparalleled position held by Baghdad in the history of Islamic calligraphy should be augmented by the influence of its calligraphic innovations on epigraphy, specifically monumental inscriptions. In this regard, the thuluth of Ibn al-Bawwab – noted for its squat letter forms, compact lines and the interconnections among normally unconnected letters – seems to have achieved an iconic status, for it is used in numerous buildings throughout the Islamic world from the late eleventh to the middle of the thirteenth century. Inscriptions from eleventh-century Ghazni and Isfahan, twelfth-century Syria and North Africa, and even thirteenth-century Aleppo and Mosul, attest both to the significance of Ibn al-­ Bawwab’s calligraphy and, perhaps more importantly, to the desire of distant dynasties to appropriate it as a symbol of allegiance to the ʿAbbasid caliphate (Tabbaa 1994: 119–48). It also seems likely that the calligraphic excellence achieved by Baghdadi calligraphers in the thirteenth century translated into the high quality of monumental inscriptions that we have seen on Baghdad’s fortifications and on the Madrasa al-Mustansiriyya, although more work needs to be done on this connection. In addition to producing magnificent Qurʾåns, Baghdad of the first half of the thirteenth century also stood at the vanguard of significant developments in manuscript illumination, including scientific manuscripts, such as those of late antique writers and their imitators, including Dioscorides and Pseudo-Galen; the fable book of Kalila wa Dimna; and particularly the Maqamat (Assemblies) of al-Hariri (d. 1122). Since this short-lived school of manuscript i­llumination

173

174

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

achieves its most distinctive mode in the Maqamat, the following discussion will focus exclusively on this proto-narrative literary genre, which was first created in the early eleventh century by Badiʾ al-Zaman al-Hamadhani but achieved its greatest popularity in the early twelfth century in the Maqamat of al-Hariri (see also Contadini 2010, ch. 17). This work tells the peripatetic adventures of a roguish hero, Abu Zayd, as narrated by al-Harith, a slightly gullible merchant, but more than anything dwells on Abu Zayd’s verbal virtuosity. An extremely popular book in its time – hundreds of copies are said to have been autographed by al-Hariri himself – the text does not seem to have been illustrated before the early thirteenth century, mainly in Baghdad but also in Syria and later Cairo as well. One question still baffles: why was this rather impenetrable text with limited narration illustrated, and why did this happen in the first quarter of the thirteenth century, more than a century after the text had been written? Richard Ettinghausen, the first to identify a school of Arab painting, attributed the efflorescence of this painting tradition in the first half of the thirteenth century to the generally able and tranquil reign of caliphs al-Nasir and al-Mustansir (Ettinghausen 1962: 8). Oleg Grabar, examining the social context of the illustrated Maqamat of the thirteenth century, linked them with an ‘urban bourgeoisie’ that, he argued, had developed in Baghdad and other contemporary cities (see Contadini 2010, ch. 17; Grabar 1970: 207–22). Alain George has re-examined the massive appeal of the Maqamat as a text that enjoyed widespread oral recitation, arguing that the style and format of both text and images were specifically intended to be listened to and seen by a seated audience. George also expanded on the picaresque element of the Maqamat, first proposed by Ettinghausen, highlighting its theatrical and comic elements (George 2012). The compositional, stylistic and iconographic features of the illustrated Maqamat are best examined through two famous thirteenthcentury manuscripts of the text, Paris Bibliothèque nationale (arabe 5847), which is signed by its calligrapher and illustrator, Yahya b. Mahmud al-Wasiti, and dated to 634 (1237), and a copy in St Petersburg (ms. S 23). Both were produced in Baghdad in the third and fourth decades of the thirteenth century, and both have generated a tremendous amount of interest among art historians, largely focusing on their stylistic sources, iconography and social history. The two manuscripts share many compositional devices: high rate of illustration; unframed illustrations that are interspersed with the text; large figures placed at the edge of the picture plane; flat interior spaces with a tripartite architectural framework; exterior scenes with minimal indications of landscape; and occasional development of the picture plane as two or three strips on which figures and animals are placed. Stylistically, the two manuscripts are characterised by highly saturated opaque water colours that are sometimes

the resurgence of the baghdad caliphate

variegated to resemble ‘watered silk’; animated figures that display a degree of emotional response; and a comic sensibility, often seen in the manner of interaction of figures with one another. Overall, the illustrated Maqamat, particularly these two manuscripts, depict satirical images of urban dwellers and the official class, giving us lively portraits of all aspects of life in a medieval Islamic city: travel, trials, banquets, parades, picnics, weddings, funerals, markets and architecture. While generally adhering to these compositional and stylistic features, the 1237 Hariri manuscript towers over all other Arab manuscripts by the quality of its paintings and their size (averaging 28 × 26 cm), the subtlety of its gestures and the variety of its compositions. Its softer palette, vividly drawn racial types and daring expansion of the picture plane contribute to more engaging pictures that not only illustrate the narrative but also create a tableau vivant of urban, agrarian and pastoral life. Two paintings in particular – both illustrations of the thirty-ninth maqama – stand out for their strident originality in subject matter and composition. The so-called Hour of Birth shows a cross-section of a two-storey house, of which the lower level is dominated by a corpulent woman giving birth, assisted by her daughter and a midwife, and the upper level shows the Indian master of the house, seated cross-legged and flanked by Abu Zayd and al-Harith, who are busy casting a horoscope for the new-born. Less jarring is a depiction of a magical island at which a boat has moored, resplendent with a griffin, a harpy, parrots, monkeys frolicking in trees and fish swimming in a blue sea (Ettinghausen 1962: 121–2; Hillenbrand 2010: 117–34). It would be nearly an entire century until the Persianate painters of the Ilkhanid court progressed beyond the achievements of late ʿAbbasid painting. The legacy of the architecture and arts of the late Baghdad Caliphate Despite its short duration and few surviving monuments, the architecture and the arts of the book of the late ʿAbbasid caliphate resonated widely within the Islamic world and left a significant legacy for the succeeding Mongol dynasties of the Ilkhanids (1256–1353) and Jalayirids (1336–1432). Forms generated, refined or monumentalised in late ʿAbbasid Baghdad – including muqarnas vaulting, proportioned Qurʾånic scripts and cursive epigraphy – were widely adopted by the middle of the twelfth century in various regions of the Islamic world, including Almoravid North Africa, Zangid and Ayyubid Syria, and even late Seljuq Iran. Does the wide dissemination of these late ʿAbbasid forms entail symbolic allegiance to the ʿAbbasids; does it suggest the appropriation of symbolic forms that were linked with the SunnÈ revival; or can this process simply be explained in aesthetic terms, as the circulation of beautiful and wondrous forms among

175

176

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

sophisticated patrons? Although aesthetic factors cannot be ignored, there are some indications that a few of these late ʿAbbasid ornamental forms possessed a symbolic charge, and that their appropriation by various Islamic dynasties points to a symbolic reciprocal relationship between a centre possessing the means of legitimation but lacking power and a periphery lacking legitimacy but possessing worldly power. As such, these ‘symbolic forms’ would have ameliorated the sharp divide between the myth of Islamic unity and the reality of political fragmentation that prevailed in the last two centuries of the ʿAbbasid caliphate (Tabbaa 2001: 163–4). The vigour and resilience of late ʿAbbasid architecture can also be demonstrated by its impressive continuity even after the Mongol conquest in 1258, as can be seen in the complex of the Madrasa al-Mirjaniyya and Khan Mirjan, built in 1357. The Khan Mirjan is quite unique, even among commercial buildings, in that its courtyard is completely covered by astonishing transverse vaulting, which is seen earlier in Sasanian architecture and later in the Ilkhanid period. Though built more than a century after the Madrasa al-­ Mustansiriyya, the Madrasa al-Mirjaniyya shows remarkable continuities with it, in terms of plan, ornament, and particularly a nearly identical monumental portal. On the other hand, the Madrasa alMirjaniyya boasts four domes, of which the most prominent has an attenuated and fluted drum, a form that recalls later Central Asian architecture. The cross-fertilisation of late ʿAbbasid and Ilkhanid architecture, both in Iraq and Iran, presents interesting possibilities for further investigation.

Notes 1. See, for example, Makdisi 1973: 15–57. See also Tabbaa 2001: 14–21. 2. Hartmann 1975 and most recently Hanne 2007. Among al-Nasir’s patronage of Shiʿi shrines are his construction of the shrine of Fatima al-Zahraʾ in Mecca in 604/1207–8 (see RCEA X, No. 3632) and the even more strident rebuilding of Ghaybat al-Mahdi in the Samarra shrine in 606/1209 (see RCEA X, No. 3658). 3. The use of dragons as apotropaic emblems over military gates is quite well attested in a number of citadels, including Aleppo, Sinjar and Mardin. For dragons in Islamic architecture, see Oney 1970 and Rabbat 2006. 4. Some of these palace names would resonate in later medieval Islamic palaces, including al-Firdaws, which is used for a thirteenth-century palace outside Mardin in southern Turkey, and especially al-Rayhan, which is one of the names of the Palace of the Myrtle at the Alhambra, Hispanicised as Arrayanes. 5. See, for example, Tabbaa 1983, pl. 4. 6. Hadithi and Khåliq 1974. 7. Sibt Ibn al-Jawzi, Mir’åt al-Zamån fi tawårÈkh al-a’yån (Beirut, 2013), vol. 8, p. 739; Kamal al-DÈn Ibn al-Fuwati, al-Hawådith al-jami’a wa’l-

the resurgence of the baghdad caliphate

tajårub al-nåfÈ’a fÈl-mi’a al-såbi’a, ed. M. Jawad (Baghdad, 1932), vol. 3, 57–8. Ibn Ba††Ë†a, Travels in Asia and Africa 1325–1354, tr. and ed. H. A. R. Gibb (London, 1929), 99. These sources are cited by Hisham Nashabe (1989: 73–5). 8. See Nashabe 1989: 140–59 for lists of teachers of various madhhabs and other professionals at the madrasa. See also al-Mashhadani and al-Naqshbandi 1986, esp. 97–138, for more extensive lists of students, librarians and books. 9. Jawad 1945, where the author identifies it as a palace. MaʾrËf 1965 and Janabi 1982: 68–71 argued that it was a madrasa, an argument accepted by Hillenbrand (1994: 185, 223). 10. These include Ibn Jubayr (2001: 237), who describes the caliph al-Nasir arriving by boat to his palace at the upper end of the eastern bank of the Tigris; and Ibn al-JawzÈ 2013: 57 for the events of the year 576/1180, where the author describes a palatial structure in the same location as ‘Dar Tatar’ or ‘Dar al-Masnat’, both earlier names for the ʿAbbasid Palace.

Bibliography Anderson, G. and Pruitt, J. (2017), ‘The Three Caliphates: A Comparative Approach’, in F. B. Flood and G. Necipoglu (eds), A Companion to Islamic Art and Architecture, 2 vols. Oxford: John Wiley, vol. I, pp. 223–49. Contadini, A. (2017), ‘Patronage and the Idea Of An Urban Bourgeoisie’, in F. B. Flood and G. Necipoglu (eds), A Companion to Islamic Art and Architecture, 2 vols. Oxford: John Wiley, vol. I, pp. 431–52. Ettinghausen, R. (1962), Arab Painting. Skira: Geneva. George, A. (2012), ‘Orality, Writing and the Image in the Maqamat: Arabic Illustrated Books in Context’, Art History, 35: 10–37. Golombek, L. and Wilber, D. (1988), The Timurid Architecture of Iran and Turan, 2 vols. Princeton: Princeton University Press. Grabar, O. (1970), ‘The Illustrated Måqåmåt of yhe Thirteenth Century: The Bourgeoisie and the Arts’, in A. Hourani and S. M. Stern (eds), The Islamic City. Oxford: University of Pennsylvania Press. Hadithi, A. and Khåliq, H. (1974), Al-Qibåb al-makhrËtiyya fi’l-’Iråq. Baghdad: Wizårat al-I’låm. Hanne, E. (2007), Putting the Caliph in His Place. Madison, NJ: Fairleigh Dickinson University Press. Hartmann, A. (1975), an-Nåsir li-DÈn Allah (1180–1225): Politik, Religion und Kultur in der späten Abbasidenzeit. Geschichte und Kultur des Islamischen Orients. Berlin: De Gruyter. Hillenbrand, R. (1994), Islamic Architecture: Form, Function and Meaning. New York: Columbia University Press. Hillenbrand, R. (2010), ‘The Schefer ÓarÈrÈ: A Study In Islamic Frontispiece Design’, in A. Contadini (ed.), Arab Painting: Text and Image in Illustrated Arabic Manuscripts. Leiden: Brill, pp. 117–34. Ibn al-Athir (1987), Al-Kåmil fi-l tårÈkh. Beirut: Dar Sader. Ibn al-JawzÈ, S. (2013), Mir’åt al-zamån, vol. 8. Damascus: Al-Risala Press. Ibn Jubayr (2001), The Travels of Ibn Jubayr. Tr. from the Arabic by R. Broadhurst. New Delhi: Goodword Books. Janabi, T. J. (1982), Studies in Mediaeval Iraqi Architecture. Baghdad:

177

178

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Republic of Iraq, Ministry of Culture and Information State Organization of Antiquities and Heritage. Jawad, M. (1945), Al-Qasr al-’abbåsÈ fi’l-qal’ah bi-baghdad. Sumer, 1: 65–86. Le Strange, G. (1901), Baghdad During the Abbasid Caliphate. Oxford: Clarendon Press. Makdisi, G. (1973), ‘The SunnÈ Revival’, in D. S. Richards (ed.), Islamic Civilization, 950–1150. Oxford: Cassirer. Ma’rËf, N. (1965), Al-Madåris al-Sharåbiyya fi Baghdad, wa-Wåsit, waMakka. Baghdad: Matb’at al-Irshåd. al-Mashhadani, M. J. and al-Naqshbandi, U. N. (1986), Al-Mustanßiriyya fi’l-tårÈkh. Baghdad: Wizarat al-Thaqafa. Nashabe, H. (1989), Muslim Educational Institutions: A General Survey Followed by a Monographic Study of al-Madrasa al-Mustansiriyya in Baghdad. Beirut: Librarie du Liban. Öney, G. (1970), ‘Dragon Figures in Anatolian Seljuk art’, Belleten, 32: 171–216. Qadi Ahmad (1959), Calligraphers and Painters, ed. V. Minorsky. Washington, DC: Free Gallery of Art. Rabbat, N. (2006), ‘The Militarization of Taste in Medieval Bilad al-Sham’, in H. Kennedy (ed.), Muslim Military Architecture in Greater Syria. Leiden: Brill, pp. 84–105. Rice, D. S. (1955), The Unique Ibn al-Bawwab Manuscript at the Chester Beatty Library. Dublin: E. Walker. Schimmel, A. (1975), Mystical Dimensions of Islam. Chapel Hill: University of North Carolina Press. Schmid, H. (1980), Die Madrasa des Kalifen al-Mustanßir in Baghdad. Mainz am Rhein: Verlag Philipp von Zabern. Tabbaa, Y. (1983), ‘The Muqarnas Dome: Its Origin And Meaning’, Muqarnas, 3: 61–74. Tabbaa, Y. (1971), ‘The Transformation Of Arabic Writing, Part 1: Qur’anic Calligraphy’, Ars Orientalis, 21: 119–47. Tabbaa, Y. (1994), ‘The Transformation Of Arabic Writing: Part 2: The Public Text’, Ars Orientalis, 24: 119–48. Tabbaa, Y. (1996), Constructions of Power and Piety in Medieval Aleppo. Philadelphia: Penn State University Press. Tabbaa, Y. (2001), The Transformation of Islamic Art during the SunnÈ Revival. Seattle: University of Washington Press.

CHAPTER NINE

The Politics of Patronage in Medieval Mosul: Nu¯r al-Dı¯n, Badr al-Dı¯n and the Question of the Sunnı¯ Revival Introduction Mosul was ruled during the twelfth and first half of the thirteenth century by the Zangids, a Turkish dynasty that drew its legitimacy from its affiliation with the Great Seljuqs, its political alignment with the ʿAbbasid Caliphate and its pursuit of the sectarian policies of the so-called SunnÈ Revival. In many respects, the Zangid dynasty of Mosul was the quintessential Seljuq successor state, having been founded by the Atabeg Aq Sunqur, after whom it was often referred to as al-Dawla al-Atabikiyya (the Atabeg State), and then ruled more or less independently by his son ʿImad al-DÈn Zangi (r. 1127–46) and his successors. Zangi quickly filled the vacuum between the weakened tribal Arab dynasties in the JazÈra and the Crusaders in the Levant, becoming the ruler of a substantial region extending from Mosul to Aleppo. Following Zangi’s assassination in 1146, his domain was divided between his two eldest sons, Sayf al-DÈn Ghazi I and NËr al-DÈn Mahmud. Ghazi was granted the province of Mosul, and his progeny would rule it until the Mongol invasion, although only nominally during the reign of Badr al-DÈn Luʾluʾ (r. 1233–59). NËr al-DÈn was given Aleppo, from which he would expand southwards to Damascus in 1154 and eventually Egypt in 1168. NËr al-DÈn’s southward turn and his increasing preoccupation with Fatimid Egypt did not divert him from the affairs of Mosul, for he was directly involved in confirming his nephew’s role and re-organising the city’s administrative structure. In fact, as I hope to demonstrate, NËr al-DÈn continued to enjoy considerable influence, verging on outright control, over the entire Zangid domain until his death in 1174.1 In spite of their commonalties as post-Seljuq sovereigns, NËr al-DÈn Yasser Tabbaa (2019), ‘The Politics of Patronage in Medieval Mosul: Nur al-DÈn, Badr al-DÈn, and the Question of the Sunni Revival’, in Seljuq Art, Edinburgh: Edinburgh University Press.

180

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

and Badr al-DÈn could not have been any more different in terms of their religious policy and architectural patronage. Comparing their architectural patronage in Mosul reveals considerable differences in the application of these policies, particularly with regard to the Shiʿi and Christian communities in the city. NËr al-DÈn was the ultimate promoter of the SunnÈ Revival, a cause that shaped his purposeful religious policy, his architectural patronage and especially his single-minded determination to counter Shiʿism and end the Ismaʿili Fatimid state. Although better known in medieval Europe as an antiCrusader, it should be emphasised that in the 1160s NËr al-DÈn sent three expeditions to defeat the Fatimids and none to free Jerusalem, which would fall to Salah al-DÈn al-Ayyubi in 1187, sixteen years after the end of the Fatimids. In fact, it seems that NËr al-DÈn’s earlier anti-Crusader zeal turned in his later reign into an anti-Shiʿi and antiChristian policy, implemented through the systematic suppression of Shiʿism in Syria and Christianity in northern Syria and the JazÈra. Badr al-DÈn Luʾluʾ, on the other hand, was a manumitted Armenian slave who began his career as atabeg to young Zangid princes in 1211, before usurping Mosul from them, ruling it from 1233 until his death in 1259.2 His tolerance of sectarian and religious minorities, specifically Shiʿism and Christianity, extended to his patronage of several Shiʿi shrines and the toleration, even support, of the building and restoration of Christian churches and monasteries. Examining the monuments erected by these two patrons in terms of their historical context, architectural features and unusual inscriptions, this essay aims to reassess their quite varied sectarian messages that range from strict adherence to Sunnism to a generous embrace of moderate Shiʿism. The patronage of Nu¯r al-Dı¯n in Mosul (1170–2) NËr al-DÈn’s involvement in the affairs of Mosul can be read through the Mosque al-Nuri, built between 1170 and 1172, which was his most important foundation outside Syria. In his exposition of NËr al-DÈn’s meritorious acts, the Syrian historian Abu Shama wrote: (NËr al-DÈn’s) mosque in Mosul is the ultimate in beauty and excellence. It is especially noteworthy that he entrusted its construction and the supervision of its expenses to Shaykh ‘Umar al-Malla . . . who was a pious man. He was told: ‘Such person is not suited for the task’. He replied: ‘If I were to assign this task to one of my associates . . . I know that he would oppress some of the time and a mosque cannot be founded on oppression’.3 This short anecdote provides interesting insights into the last and most important foundations of the Syrian sovereign, first because the mosque was built in a city on which he exerted considerable

the politics of patronage in medieval mosul

i­nfluence but was not directly ruled by him. Although Muslim sovereigns have often restored or added on to monuments outside their domain of authority, their acts of patronage generally focused on a handful of ancient shrines and sites of pilgrimage, including Mecca, Medina and Jerusalem for the SunnÈs and the shrines of the Imams for Shiʿis. Entire monuments built outside a sovereign’s region of control were far less common and nearly always associated with political control rather than pious offering. Second, short of the hapless man as which ʿUmar al-Malla is often described, he was in fact a Sufi leader of some note and the pole of opposition to the Christians of Mosul, in particular to the Christian vizier Fakhr al-DÈn ʿAbd al-Masih. Third, it seems likely from NËr al-DÈn’s rather hands-off approach to the project that he was more concerned with the intention and message of the mosque than the details of its construction.4 Although NËr al-DÈn began his mosque in 1170, his involvement with Mosul dates back much earlier, to 1146 and 1149, when he twice oversaw the succession of Zangid rule in Mosul. In fact, by the second succession, NËr al-DÈn was already the elder of the Zangid household, and his name was officially pronounced in the Friday sermon. But it was the third succession after the death of his nephew Qutb al-DÈn Mawdud in 1170 that witnessed NËr al-DÈn’s deepest intrusion into the politics of Mosul. At the peak of his powers, NËr al-DÈn would play a decisive role in this succession, going as far as to send a military expedition that threatened to bring Mosul directly under his control if his conditions were not met. Interestingly, these conditions had less to do with the choice of successor and everything with containing the influence of the Christians in Mosul, who were then represented by the vizier Fakhr al-DÈn ʿAbd al-Masih. As NËr al-DÈn is quoted saying: ‘My intention is not the city itself . . . [but] to remove this Christian from governing Muslims.’5 To that end, following his successful campaign in al-JazÈra and Mosul, NËr al-DÈn ordered ʿAbd al-Masih banished to Aleppo and forced him to convert to Islam, changing his name to ʿAbdullah. NËr al-DÈn would subsequently impose various repressive measures against local Christian communities, including increasing the jizya tax, restrictions on the building of Christian structures, the confiscation of monastic endowments and even the outright destruction of churches and monasteries. It is undoubtedly because of these repressive measures that Syriac Christian writers, including Michael the Syrian and Bar Hebraeus, were harshly critical of NËr al-DÈn.6 During his lengthy sojourn in Mosul, NËr al-DÈn ordered the foundation of a Friday mosque near the centre of the old city, a place that is described by Muslim chroniclers as an empty lot and by Christian writers as the location of the Church of St Paul. The mosque was completed in just under two years, for in 1172 NËr al-DÈn visited Mosul for the fourth and last time, during which he drew up the substantial endowment of the mosque, appointed its officials and

181

182

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

also founded a madrasa next to it, which is perfectly consistent with his patronage of SunnÈ institutions all over Syria. Regrettably, the Mosque al-Nuri was entirely rebuilt around 1945 – with completely new arches, vaults and a maqsura dome – making its proper study, even during my first visit in 1979, very difficult (Figures 9.1, 9.2).7 Fortunately, about forty Figure 9.1  Sanctuary façade, as rebuilt in years before this rebuilding Herzfeld 1945, at the Mosque al-Nuri. 1172 and later, attempted a rather hasty documenMosul, Iraq (photo: Yasser Tabbaa) tation of the mosque, producing its current plan and attempting a reconstruction of its original plan and design. Furthermore, in 1979 I came upon an archive of large glass negatives at the Iraqi Institute of Antiquities and Museums that documented the mosque’s appearance around the same time (Figure 9.3). Although Herzfeld suggested that the Mosque al-Nuri had been built several decades earlier and that NËr al-DÈn only restored it, all historical and archeological evidence argues against his chronology and strongly suggests that the entire original structure was built by NËr al-DÈn between 1170 and 1172.8 In short, as I have argued elsewhere, the Mosque al-Nuri in Mosul was founded and completed by NËr al-DÈn with a broad but shallow sanctuary focused on a large dome (Figure 9.4). Its original support system consisted, as it did until recently, of squat octagonal piers made in sections of the soft blue marble common in Mosul since Assyrian times. The octagonal shaft ended in concave triangular cells that supported a square frieze and a square console inscribed on their southern side with continuous and quite legible Qurʾånic inscriptions. Such inscribed capitals are very rarely seen in Islamic architecture, and their extensive use in this otherwise poorly built mosque confirms NËr al-DÈn’s strict adherence to basic Islamic principles, Qurʾånic scripture and prayer. Curiously, some of these columns were buttressed by shorter bundled columns with lyre-shaped capitals, a column type that is found in several churches and monasteries in and around Mosul (Figure 9.5). The presence of these rather archaising capitals and especially an exquisite mi˙råb, dated Figure 9.2  Interior, as rebuilt in to 1148, threw off Herzfeld’s chronology, 1945, in the Mosque al-Nuri. leading him to propose that the mosque 568/1172 and later, Mosul, Iraq (photo: Yasser Tabbaa) had been built around 1148 by Sayf al-DÈn

183

the politics of patronage in medieval mosul

a

b Figure 9.3  a. Exterior from the south-east of the Mosque al-Nuri. 568/1172 and later, Mosul, Iraq. Source: photograph courtesy of the Iraqi Institute of Antiquities and Museums. b. Exterior from the north-west of the Mosque al-Nuri. 1172 and later, Mosul, Iraq (source: photograph courtesy of the Iraqi Institute of Antiquities and Museums)

Ghazi I and that NËr al-DÈn’s works were little more than the restoration of a pre-existing mosque.9 But historical and archeological evidence suggests otherwise, for all contemporary sources attribute the building of the mosque to NËr al-DÈn without mentioning an earlier phase. As for the earlier mi˙råb and the bundled columns, later sources tell us that both were in fact brought to the mosque in the nineteenth century: the mi˙råb most likely from the so-called Umayyad Mosque and the columns from a destroyed church, where

184

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 9.4  Reconstructed plan of the sanctuary of the Mosque al-Nuri. 568/1172 and later, Mosul, Iraq (photo: Yasser Tabbaa)

such columns are generally found.10 The misshapen dome seen in old photographs most likely represents two phases, an original hemispherical dome that was later restored with a conical exterior. In short, the Mosque al-Nuri was entirely built by NËr al-DÈn following a regional JazÈran mosque type that blends the plan of the Great Mosque of Damascus with a Seljuq-style dome.11 The Mosque al-Nuri lacks any historical inscriptions from the time of NËr al-DÈn but contains numerous Qurʾånic and pious inscriptions. These have survived as short friezes on the octagonal capitals, marble bands with inlaid black inscriptions and a large undated stucco panel that is now at the Iraqi Museum. Although the Qurʾånic excerpts on the capitals are quite commonplace – 2:255 (Throne Verse); 9:18 (on the erection of mosques); and 24:36–7 (also on building houses for God) – their use as continuous friezes on capitals is nearly unique in Islamic architecture. The inlaid friezes, which may have originally flanked the mi˙råb, contain most of 2:148–50, rather uncommon verses that address the question of the qibla (Figure 9.6). As for the large and undated stucco panel that once crowned the mi˙råb, it seems to date to a later restoration, according to Herzfeld by Badr al-DÈn, although this is by no means certain. The stucco panel focuses on a square KËfic text that reads: ‘Muhammad, Abu Bakr, ʿUmar, ʿUthman, ʿAli, al-Hasan, al-Husayn, may God be pleased with them all’ (Figure 9.7). This encomium, known in some twelfth-century inscriptions, may underline the ecumenical embrace that characterised the period of Badr al-DÈn.12 Although not especially compelling for their content, the inscriptions of the Mosque al-Nuri command our attention for their visibility, which is enhanced by their northern location on the capitals and by the contrasting colour of the inlaid friezes. Surrounding the worshipper by evocative Qurʾånic verses that would have been audibly reinforced during the Friday sermon, these inscriptions deliver an exoteric message within a saturating atmosphere, directly linked to NËr al-DÈn’s pronouncements and his strident SunnÈ policy.

the politics of patronage in medieval mosul

Figure 9.5  Interior, c. 1920, of the Mosque al-Nuri. 1172 and later, Mosul, Iraq (photo: Yasser Tabbaa)

Equally strident was the minaret of the Mosque al-Nuri, the so called al-Hadba’ (the hunchback), whose lofty bent profile became a symbol of Mosul before its horrific destruction in 2017. Standing at the north-western corner of the enclosure, the circular minaret rose to a height of 56 metres from a slightly battered rectangular base, which is also ornamented on all four sides with brick strapwork (Figure 9.8). Its circular shaft is divided by guilloche mouldings into seven zones which are lavishly ornamented in varieties of hazar baf and strapwork ornament that recall Seljuq Iranian minarets. In sum, the mosque – with its tripartite prayer hall, large maqsura dome and lofty minaret – resembles a host of JazÈran mosques that were

185

186

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

founded in the twelfth and thirteenth centuries in regions that had a large Christian population. The architectural patronage of Badr al-Dı¯n Luʾluʾ Surprisingly little has survived from the Zangid rulers who ruled Mosul before Badr al-DÈn’s takeover in 1233. Sixteen madrasas and ten khånqåhs are known to have been built by these rulers, but with the exception of a handful of stone mi˙råbs and portals, preserved mainly at the Mosul Museum and the Iraq Museum in Baghdad, none of these monuments has survived.13 Saʿid Daywahji has suggested that Badr al-DÈn, in his efforts to promote Shiʿism, had converted some of these madrasas into Shiʿi shrines, such that, for example, the shrine of Yahya b. al-Qasim was built adjacent to an earlier madrasa.14 Several of these shrines, all possessing a pyramidal roof sheltering an interior brick muqarnas vault, once dotted the landscape of old Mosul. Sadly, all these shrines and many more have been levelled by Daesh during its recent occupation of Mosul.

Figure 9.6  Inscribed marble panel from the Mosque al-Nuri. Undated, Mosul, now in the Iraqi Museum (photo: Yasser Tabbaa)

Figure 9.7  Stucco panel from the Mosque al-Nuri. Mid-thirteenth century, Mosul, now in the Iraqi Museum (photo: Yasser Tabbaa)

the politics of patronage in medieval mosul

Two of these shrines – those of Imam Yahya b. al-Qasim and Imam ʿAwn al-DÈn – stand out for their historical importance and degree of preservation and will be the focus of this discussion of Badr al-DÈn’s architectural patronage. The shrine of Imam Yahya stood at the foot of Mosul’s citadel, Bash Tabiya, about midway between the citadel and Qara Serai, a lovely pavilion whose two-storey remains face the Tigris. Such informal pavilions were quite common in medieval Islamic architecture, and the Qara Serai resembles in particular the Alaeddin Keyqubad kiosk at the foot of the Konya citadel, built at approximately the same date.15 The shrine of Imam Yahya b. al-Qasim was founded by Badr al-DÈn Luʾluʾ in 637/1239 adjacent to his Madrasa al-Badriyya, which has entirely vanished. Although no archaeological work has been done in its vicinity, it seems likely that the shrine was part of a complex that included the citadel and the Qara Serai kiosk, a royal/religious complex

187

Figure 9.8  Minaret of the Mosque al-Nuri. 568/1172, Mosul, Iraq (photo: Yasser Tabbaa)

Figure 9.9  View of Mosul from citadel (north), with shrine of Imam Yahya b. al-Qasim to left (photo: Yasser Tabbaa)

188

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 9.10  Mosul. Shrine of Imam Yahya b. al-Qasim – Mosul, 1239 (photo: Yasser Tabbaa)

that was enclosed by a low wall separating it from the city. In fact, contemporary sources describe Badr al-DÈn’s frequent visits to this shrine where he would meet with notables, teachers and jurists in order to discuss the history of Imam ʿAli and the various hadiths that extol his virtues and those of the ahl al-bayt.16 Furthermore,

the politics of patronage in medieval mosul

Badr al-DÈn was temporarily interred in the shrine after his death in 1258 before his remains were allegedly moved to Najaf, to be buried near the shrine of Imam ʿAli.17 The shrine stood until recently about 100 m south of the citadel on the western bank of the Tigris, into which it was gradually sinking long before its ultimate destruction in 2017 (Figure 9.12). The pyramidal dome of the shrine rose from a slightly battered rectangular base whose four sides were once ornamented with brick patterns enhanced with turquoise glazed tiles. Entered from the north, facing the citadel, its main façade was flanked by two large panels of glazed brick geometric patterns and exquisite floriated KËfic inscriptions. These are surmounted by Badr al-DÈn’s foundation inscription, which, in evident humility to the interred saint, gives his name simply as ‘the poor slave Luʾluʾ b. Abdullah’ (Figure 9.13). Traces of glazed bricks on the other three sides and on the pyramidal dome suggest that the monument was originally lavishly decorated with blue-glazed tile ornament, which is rather rare in Mesopotamia. Three steps lead down to a dark interior centred around a new cenotaph; the original magnificent carved wood cenotaph has been moved to the Mosul Museum. A lavishly decorated marble dado surrounded the lower walls and culminated in an elaborate mi˙råb located in the south-eastern corner (Figures 9.14, 9.15). Made of typical soft blue veined marble, the dado was ornamented in two distinct styles

Figure 9.11  Mosul. Shrine of Imam Yahya b. al-Qasim – Mosul, 1239, dome (photo: Yasser Tabbaa)

189

190

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

that seem to belong to two different periods. The first, from Badr al-DÈn’s period, consisted of deeply carved ­arabesque patterns undergirded by two raised inscription bands,  the lower in a majestic thuluth and the upper in an exquisite cursive hand that resembles contemporary manuscript scripts. The second style comprised blue marble panels that were inlaid with white marble inscriptions. These date to the late thirteenth century, when the shrine was restored by a certain Ibrahim b. ʿAli, who is identified as ‘the servant of the sacred shrine’. This inscription included the well-known Shiʿi prayer for the Figure 9.12  Mosul. Exterior from the Twelve Imams, the name of the north-west of Mashhad Yahya b. patron and a Qurʾånic passage al-Qasim. 637/1239 (photo: Yasser (76:7–10), a passage often found Tabbaa) in Shiʿi foundations (Figure 9.15). A third inscription band, made out of carved brick and quite likely part of the original structure, surrounded the shrine just below the springing of the dome. This inscription also gives the name and abbreviated titles of the founder and the name of the supervisor of construction, Sanbak al-Badri, a slave of Badr al-DÈn. The wooden cenotaph, the marble inscriptions and the frieze at the springing of the dome contain inscriptions, Qurʾånic and pietistic, of a distinctly Shiʿi character. In addition to the well-known encomium

Figure 9.13  Mosul. Exterior with inscription frieze of Mashhad Yahya b. al-Qasim. 637/1239 (photo: Yasser Tabbaa)

Figure 9.14  Mosul. Interior, marble ornament and inscriptions of Mashhad Yahya b. al-Qasim. 637/1239 (photo: Yasser Tabbaa)

the politics of patronage in medieval mosul

191

Figure 9.15  Mosul. Shrine of Imam Yahya b. al-Qasim – Mosul, 1239, interior inscription (photo: Yasser Tabbaa)

for the Twelve Imams, the mi˙råb is surrounded by verses 20:12 and 11:72, and the cenotaph with 2:255 – verses whose reference to the sacrifice, blessing and purity of ahl al-bayt have long had a specifically Shiʿi interpretation.18 A majestic muqarnas dome covers the shrine, rising rigorously and logically from four squinches that develop into a sixteen-sided star ending in a small eight-pointed star at the apex (Figure 9.17). The construction of the muqarnas cells out of small glazed brick tiles, also seen at Imam ʿAwn al-DÈn, differs from the earlier plaster domes of southern Iraq and contributes to the pristine geometry still visible in the better-preserved portions. Overall, the location, opulence and inscriptions of the Mashhad of Imam Yahya suggest that it was built as a private shrine/chapel for Badr al-DÈn and his circle, a clear indication of his Shiʿi inclination in the second half of his reign. Although we cannot be certain about Badr al-DÈn’s sincerity, his self-identification as ‘wali Al Muhammad’ and his burial in Najaf strongly argue for his conversion to Shiʿism.19 The shrine of Imam ʿAwn al-DÈn, founded by Badr al-DÈn in 646/1248, occupies a more central but less dramatic location than Imam Yahya’s (Figure 9.18). Taller and more Figure 9.16  Mosul. attenuated than the shrine of Imam Yahya, it is entered Interior and corner mi˙råb of the from the west through an elaborate blue marble portal Mashhad Yahya b. framed by a delicate cursive Qurʾånic inscription (contain- al-Qasim. 637/1239 ing the Throne Verse) and surmounted by a larger thuluth (photo: Yasser inscription that affirms the merits of building mosques Tabbaa)

192

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

(24:36; Figure 9.19). A third inscription, made of white marble inlaid in the stones of the lintel, consists of just six words that prominently give the name and brief titulature of Badr al-DÈn as al-Sultan al-Malik al-Rahim Badr al-Dunya w’al-DÈn. The joggled voussoirs of this lintel end in two suspended keystones, a typically Mosulite form seen in other portals dating to the period of Badr al-DÈn, including the portals Figure 9.17  Mosul. Muqarnas vault at the monasteries of Mar Bahnam and Mar of Mashhad Yahya b. al-Qasim. Ahudemmeh (Figure 9.20).20 637/1239 (photo: Yasser Tabbaa) This marble frame once surrounded a magnificent wooden door whose two jambs were sheathed in brass sheets that were overlaid with brass strips to form a complex geometric pattern. Signed by another mamluk of Badr al-DÈn (ʿUmar b. al-Khidr . . . al-Maliki al- Badri), the portal closely resembles a nearly contemporary one that once belonged to the Great Mosque of Cizre before being moved to the Museum of Islamic and Turkish Art in Istanbul.21 As at the shrine of Imam Yahya, the interior of the Mashhad Imam ʿAwn al-DÈn is a few steps below street level, except that the floor here, during my two visits in 1979 and 1986, was inundated by about two feet of water. The marble revetment around the walls had largely vanished, although its fragments were kept at the Mosul Museum. Taller and more complex than Imam Yahya’s shrine, the dome here springs from fully developed muqarnas squinches and progresses similarly to sixteen- and then eight-pointed stars (Figure 9.21). The fact that the shrine was until recently surrounded by a cemetery with old and recent tombs points to its continued veneration in later centuries. The architectural patronage of Nu¯r al-Dı¯n and Badr al-Dı¯n compared Very little has been written in English on Badr al-DÈn, and none of these works sheds much light on his proclivities towards Christianity and especially Shiʿism. Nevertheless, contemporary sources occasionally and quite critically refer to his benevolent attitude towards the Christians of Mosul, noting, for example, that he had the city decorated for Palm Sunday (Sha’aneen) and that he himself took part in the festivities. Churches and monasteries prospered during his reign, including Mar Behnam, Mar Matta, Mar Ahudemeh and others. Not surprisingly, he was well-regarded by the same Christian historians who had reviled NËr al-DÈn. Badr al-DÈn’s support of Shiʿism, to which he most likely converted, is somewhat more difficult to explain. It could be linked to

the politics of patronage in medieval mosul

193

Figure 9.19  Mosul. Portal of Mashhad Imam ʿAwn al-DÈn. 646/1248 (photo: Yasser Tabbaa) Figure 9.18  Mosul. Exterior of Mashhad Imam ʿAwn al-DÈn. 646/1248 (photo: Yasser Tabbaa)

Figure 9.20  Mosul. Portal detail of Mashhad Imam ʿAwn al-DÈn. 646/1248 (photo: Yasser Tabbaa)

Figure 9.21  Mosul. Muqarnas vault of Mashhad Imam ʿAwn al-DÈn. 646/1248 (photo: Yasser Tabbaa)

the ecumenical policies of the ʿAbbasid Caliph al-Nasir (r. 1180– 1225), for whom a specifically ʿAli-centred mystical Shiʿism was promoted in order to foster an esprit de corps among his futuwwa legions that he had created in order to oppose the last Seljuqs in Baghdad.22 But Badr al-DÈn’s Shiʿism was far more pervasive and not linked to Sufism, for he is known to have converted both madrasas and khånqåhs (Sufi convents) into Shiʿi shrines. Rather,

194

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

it seems that his turn to Shiʿism was due in part to a great Sufi rebellion threatening his reign around the time that he had commenced to usurp rule from the last Zangid sultans.23 This rebellion was led by the charismatic Sufi leader Shaykh Hasan Shams al-DÈn al-ʿAdawi, who, claiming Umayyad descent and appealing to Arab ethnicity, called for opposing both Badr al-DÈn and the Zangids and, quite curiously, the establishment of a new Umayyad dynasty. His tariqa al-ʿAdawiyya, which would later be assimilated into the Yazidi faith, gained numerous adherents who spread throughout the region.24 Badr al-DÈn’s battle against Shams al-DÈn and his ʿAdawiyya disciples proceeded along both ideological and military lines. Ideologically, Badr al-DÈn sought first to discredit Shams al-DÈn by denouncing his alleged lineage and accusing him of the usual charges of religious laxity and sexual perversion. Soon afterwards, he resorted to oppose Shams al-DÈn’s charismatic Sufism by Imami Shiʿism, pronouncing his allegiance to ahl al-bayt and spreading the Shiʿi faith in Mosul, where it was already well represented. Adopting the epithet of wali al Muhammad with its well-known resonance, he proceeded to build elaborate shrines for various Shiʿi saints all over Mosul and in Sinjar and Tikrit while also converting earlier SunnÈ structures for Shiʿi piety. He is not known to have built any mosques or madrasas, nor even a mausoleum for himself, which was common practice among medieval SunnÈ dynasts.25 It was not until 1254, five years before the end of his reign, that he had Shams al-DÈn and many of his followers mercilessly killed. About three-quarters of a century apart and ostensibly belonging to a general trend towards SunnÈ orthodoxy, NËr al-DÈn and Badr al-DÈn are poles apart in terms of their religious and sectarian policy and architectural patronage. Obsessed by SunnÈ orthodoxy, NËr al-DÈn persecuted both Shiʿis and Christians, although he was favourably disposed towards Sufis, and built numerous mosques, madrasas and khånqåhs. Not surprisingly, he is nearly beatified by contemporary and later historians, who often placed him in the ranks of the Companion Caliphs. Numerous books in Arabic and European languages have been written on him, including a three-volume study by Nikita Elisseéff and my own unpublished dissertation.26 Badr al-DÈn, on the other hand, is not nearly as well regarded or studied, although his reign marks the peak of cultural life in medieval Mosul. In addition to his numerous architectural foundations, he also commissioned a famous illustrated manuscript of Kitab alAghani and some astonishing inlaid metalwork.27 Regrettably, all his architectural monuments have nearly vanished without being properly studied. Their remarkable fusion of northern and southern Iraqi forms, borrowings from regional Christian architecture, and the purposeful use of Qurʾånic and pious inscriptions all point to an especially high point in medieval Islamic architecture.

the politics of patronage in medieval mosul

Comparing the architectural patronage of NËr al-DÈn and Badr al-DÈn in Mosul also problematises the notion of the ‘SunnÈ Revival’, which has recently been questioned for its potentially simplistic sectarian polarisation that does not adequately address the commonalities of popular piety.28 There is little question, of course, that shrines often appealed to a wide spectrum of the population regardless of their sectarian or religious affiliation, such that SunnÈs, in particular, sought the blessing and intercession from Shiʿi, Christian, Jewish or even pagan shrines. However, popular piety as a social practice was itself subject to influence and manipulation by dominant political forces, which could encourage, condone, divert or prohibit such pietistic ambivalence and sectarian pluralism. Islamic history is certainly replete with examples of dynasties, both SunnÈ and Shiʿi, not just suppressing pious pluralism but also prohibiting the veneration of saints of the wrong sect and even destroying their shrines. Of these, the Wahhabi destruction of the shrines of ahl al-bayt in the Jannat al-Baqi’ (a cemetery near Medina) in the eighteenth and nineteenth centuries is perhaps the most notorious, although the Safavid destruction of SunnÈ shrines and mausoleums in Baghdad in the sixteenth century provides an opposite and equally powerful example. Although these suppressive political measures never actually stopped cross-sectarian piety, there is little question that politics played a significant role in shaping and even redirecting it towards those shrines that had been sanctioned and supported by the prevailing dynasties. NËr al-DÈn and Badr al-DÈn shaped the history of Mosul in quite different ways, and their legacies corresponded fairly closely to the general sectarian transformation in northern Syria and the JazÈra from the late twelfth century onwards. In a region whose two main cities, Mosul and Aleppo, witnessed a decisive shift in the thirteenth century and later from Shiʿism to Sunnism, it comes as no surprise that contemporary and later chroniclers favoured NËr al-DÈn’s SunnÈ orthodoxy and religious intolerance over Badr al-DÈn’s ecumenical support of Imami Shiʿism and Eastern Christianity. NËr al-DÈn and his successors fostered a dogmatic trend that has continued until today, while Badr al-DÈn represents a short-lived and often forgotten period of tolerance. Notes 1. There exists not a single book in English dedicated to the history of the Zangid dynasty of Mosul. But this history is otherwise available in Said Daywaji’s various Arabic publications, including Tarikh al-Mawsil. See also the short monograph Patton 1992. For NËr al-DÈn and the early Zangids, see Elisséeff 1967. 2. As such, he was the first former slave to become sultan, anticipating the rise of the Mamluks in Egypt by twenty-five years.

195

196

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

3. 4. 5. 6.

Abu Shama 1998, pp. 20–1. Tabbaa 2002. Abu Shama 1998, p. 482. See al-Antaki 1990, pp. 289–90, 295–9. In fact, these Christian sources allege that NËr al-DÈn wrote a letter to the caliph of Baghdad that calls for the forced conversion or outright slaughter of all Christians in Muslim lands. On that, see Fiey 1959, pp. 35–6. 7. Even this rebuilt mosque was blown up, along with the famous brick minaret, in 2017, at the hands of ISIS/Daesh. 8. Sarre and Herzfeld 1911, vol. 2, pp. 216–30. 9. Ibid., pp. 22–5. Herzfeld offered this tentative chronology somewhat apologetically since the Mosque al-Nuri was very poorly lit internally. 10. It is virtually impossible to ascertain the entire chronology of the Mosque al-Nuri or to determine the various sources of the later columns and other architectural fragments. For example, the mi˙råb, dated 543/1148, that was brought to the mosque in the nineteenth century had salvaged pieces from several unknown monuments. 11. Tabbaa 2002, pp. 343–8. 12. Ibid., pp. 348–51. For a complete reading of the inscriptions on this stucco panel see Salman and Totonchi 1975, pp. 38–41. 13. The most complete discussion of the medieval architecture of Mosul, with special emphasis on the period of Badr al-DÈn, remains Sarre and Herzfeld 1911, esp. vol. 2, pp. 238–70. 14. See Daywaji 1982, pp. 316–17, and also Daywaji 1957. Daywaji in fact lists no less than four madrasas that Badr al-DÈn converted into Shiʿi shrines, including Madrasa al-Nizamiyya, which became a shrine for Imam ʿAli al-Asghar; Madrasa al-ʿIzziyya, which became a shrine for Imam ʿAbd al-Rahman; Madrasa al-Nuriyya, which became a shrine for Imam al-Bahir; and Madrasa al-Badriyya to which he added the shrine of Imam Yahya b. al-Qasim. 15. See, for example, Redford 1994, pp. 215–32. 16. Daywaji 1982, p. 360; and Daywaji 1968, pp. 171–82. 17. Daywaji 1968, p. 173. 18. Ibid., pp. 174–5, where Daywaji gives a list of six shrines in and around Mosul, nearly all built by Badr al-DÈn, that contain the same encomium. 19. See Daywaji 1982, p. 316, who adds that Badr al-DÈn held frequent gatherings in which books that praised Imam ʿAli and that mourned the martyrdom of Imam al-Husayn were read. 20. Sarre and Herzfeld 1911, vol. 2, pp. 294–303; and most recently Snelders 2010. 21. For an illustration of the door at Imam ʿAwn al-DÈn, see Daywaji 1982, p. 432; and Sarre and Herzfeld 1911, vol. 2, p. 269. See also Canby, Beyazit, Rugiadi and Peacock 2016, pp. 64 and 221. 22. Hartmann 1975. 23. Badr al-DÈn’s turn to Shiʿism as a reaction to the Sufi ʿAdawi movement is not specifically addressed in historical chronicles, although it has been surmised by modern Iraqi historians, including Daywaji 1982, p. 316; and Tariq 1982. 24. On the Yazidi or Ayzidi faith, see Kreyenbroek and Rashow 2005, pp. 4–5. 25. Shiʿi sovereigns and princes rarely built mausoleums for themselves, but preferred to immortalise their names by associating themselves, through patronage, with the shrines of prominent Shiʿi saints.

the politics of patronage in medieval mosul

26. Tabbaa 1982; Elisséeff 1967. 27. Canby, Beyazit, Rugiadi and Peacock 2016, pp. 61–4. See also Hillenbrand 2017, which for the first time provides colour illustrations of all six frontispieces of this important manuscript. 28. Mulder 2014. See also my review of this book in the Journal of Shi‘i Studies 1 (2016).

Bibliography Abu Shama, Shihab al-Din (1956–62), Kitab al-rawdatayn fi akhbar aldawlatayn al-nuriyya w’al-salahiyya, 2 vols. ed. M. H. Ahmad. Cairo. al-Antaki Yahya b. Sa‘id (1990), Tarikh al-Antaki al-ma‘ruf bi-silat tarikh Utikha, ed. A. U. Tadmuri. Tripoli, Lebanon. Canby, S., Beyazit, D., Rugialdi, M. and Peacock, A. C. S. (2016), Court and Cosmos: The Great Age of the Seljuqs. New York: The Metropolitan Museum of Art. Daywaji, S. (1968), ‘Mashhad al-Imam Yahya bin al-Qasim’. Sumer, 24: 171–82. Daywaji, S. (1957), ‘Madaris al-Mawsil fi’l-‘ahd al-atabiki’, Sumer, 13: 73–95. Daywaji, S. (1982), Tarikh al-Mawsil. Mosul: al-Majma‘ al-‘Ilmi al-‘Iraqi. Elisséeff, N. (1967), Nur al-Din, un grand prince musulman de Syrie au temps des Croisades 511–569 H./1118-1174, 3 vols. Damascus: Institut Français de Damas. Fiey, J. M. (1959), Mossoul chrétienne. Essai sur l’histoire, l’archéologie et l’état actuel des monuments chrétiens de la ville de Mossoul. Beirut: Jesuit Press. Hartmann, A. (1975), An-Nasir li-Din Allah 1189-1225: Politik, Religion, Kültür in der spaten ‘Abbasidenzeit. Berlin: De Gruyter. Hillenbrand, R. (2017), ‘The Frontispiece Problem in the Early 13th-Century Kitab al-Aghani’, in L. Korn and H. Müller-Wiener, Central Periphery? Art, Culture and History of the Medieval Jazira. Wiesbaden, pp. 199–227. Kreyenbroek, P. and Rashow, K. J. (2005), God and Sheikh Adi Are Perfect. Wiesbaden: Harrassowitz Verlag. Mulder, S. (2014), The Shrines of the ʿAlids in Medieval Syria: Sunnis, Shiʿis and the Architecture of Coexistence. Edinburgh: Edinburgh University Press. Patton, D. (1992), Badr al-Din Lu’lu’: Atabeg of Mosul, 1211–1259. Occasional Papers Series, band 3. Taschenbuch. Redford, S. (1994), ‘Thirteenth Century Rum Seljuq Palaces and Palace Imagery’, Ars Orientalis 23: 215–32. Salman, ‘I. and Totonchi, N. (1975), Texts in the Iraq Museum, vol. 8, Arabic Texts. Baghdad: Directorate General of Antiquities. Sarre, F. and Herzfeld, E. (1911), Archaologische Reise im Euphrat- und Tigris-Gebiet, 4 vols. Berlin: Verlag von Dietrich Reimer. Snelders, B. (2010), Identity and Christian–Muslim Interaction: Syrian Orthodox from the Mosul Area. Leiden: Orientalia Lovaniensia Analeteca, vol. 198. Tabbaa,Y. (1982), ‘The Architectural Patronage of Nur al-Din, 1146–1174’. Unpublished Ph.D. diss., New York University. Tabbaa, Y. (2002), ‘The Mosque of Nur al-Din in Mosul, 1170–72’, Annales Islamologique, 36: 339–60.

197

198

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Tabbaa, Y. (2016), ‘Review of Stephennie Mulder, The Shrines of the ʿAlids  in  Medieval Syria: Sunnis, Shiʿis and the Architecture of Coexistence. Edinburgh: Edinburgh University Press, 2014, Journal of Shi’i Studies, 1.

PART III  DOMES

CHAPTER TEN

The Muqarnas Dome: Its Origin and Meaning One of the most original inventions of Islamic architecture is the muqarnas, and one of the most effective and widespread of its applications is without doubt the muqarnas dome or semi-dome. Brick vaults and domes have been known in the Near East since Sassanian times, if not before, but the dome in muqarnas is a truly Islamic creation without precedent in any civilisation. Whether made of wood, stucco, brick or stone, muqarnas vaults were among the most characteristic features of medieval Islamic architecture from Iran to Spain. It is therefore not surprising that a good number of studies have been dedicated to the description and analysis of this architectural form.1 What is surprising, however, is that in spite of all these studies such basic problems as origin, chronology, geographic distribution – not to mention meaning – remain unclear and subject to debate. My intention here is to discuss these problems, though not for all types of muqarnas in all periods. I shall limit myself to muqarnas domes and vaults constructed between c. 1050 and c. 1250, that is, to the earliest known examples. Limiting the scope of the discussion to the earliest known muqarnas domes requires some explanation. My reasons are two. First of all, any attempt to discern meaning in such a common architectural form faces the danger of falling into a morass of over-generalisation unless it is focused in some way.2 The problems of interpreting Islamic architecture have been discussed by Oleg Grabar both very recently3 and in a number of earlier essays.4 Grabar attributes the disjunction, or at least the weak connection, between form and meaning (or symbol) to the ‘low symbolic charge’ of Islamic architectural forms, a characteristic that ultimately led to ‘an ambiguous visual system’.5 While that conclusion is generally correct, it overlooks the point that certain forms in certain specific times and places had a ‘high symbolic charge’ at the moment of their inception. How long these forms continued to be used with full awareness of their highly charged meaning depended on a variety of factors, but Yasser Tabbaa (1985), ‘The Muqarnas Dome: Its Origin and Meaning’, Muqarnas 3, 61–76.

202

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

generally speaking that awareness was lost, and the concomitant debasement of meaningful forms into mere decoration took place much more quickly in Islam than in other cultures and religions.6 It therefore becomes imperative in searching for meaning in any given form of Islamic architecture to begin with the origin of this form and to focus on its earliest development – particularly its first use outside its place of origin. Because equally early ‘muqarnas-like’ elements have been found in both north-eastern Iran and central North Africa, most authorities assume that the muqarnas (and hence the muqarnas dome) either originated in one or the other place or was invented simultaneously in both. Specialists in Iranian architecture postulate a continuous line of development that begins with the tenth-century fragments found near Nishapur and the tripartite squinches of the late tenthcentury Arab Ata mausoleum at Tim, continues with the numerous eleventh-century Seljuq domes, and ends with Ilkhanid and Timurid muqarnas domes and portal vaults.7 The problem with this theory is that no direct link can be established between Seljuq domes and Ilkhanid muqarnas domes and portal vaults. Large smooth Seljuq domes which spring from a multipartite squinch zone continue unchanged until well into the Ilkhanid period, as a comparison between the domes of the Great Mosques of Isfahan (1088) and Veramin (1322–6) shows.8 Therefore, although the differentiation of the squinch into muqarnas cells does indeed occur first in northeastern Iran, the total muqarnas dome of the Ilkhanid period is not a product of this development and must therefore be attributed to some other source. As for the North African development, I doubt if the so-called muqarnas fragments discovered at Qalʿat bani Hammad and dated to the late eleventh century are muqarnas at all.9 They share no properties with true muqarnas cells, and in any case they could never have been assembled to fill the cavity of a dome or even a niche. Not by any stretch of the imagination could they have led to the well-known twelfth-century muqarnas vaults in a number of North African mosques.10 Both north-eastern Iran and North Africa therefore have to be eliminated as likely places for the origin of the muqarnas dome. Another possibility is Iraq, since it has provided us with the earliest example of a fully fledged muqarnas dome in the so-called shrine of Imam al-Dawr, located some twenty kilometres north of Samarra in a village called al-Dawr.11 It is a shrine dedicated to Imam Muhammad ibn Musa ibn Jaʿfar, an alleged son of the fifth Shiʿi Imam. It was begun by the ʿUqaylid prince Muslim ibn Quraysh, who died in 1085, and was completed before 1090 by officials of his court.12 The mausoleum consists of an elongated chamber with tapering walls about twelve metres high and a muqarnas dome almost exactly the same height (Figure 10.1). As an early example of its type, this dome betrays certain affinities with the regular squinch dome. The square of the

THE MUQARNAS DOME

Figure 10.1  Al-Dawr. Shrine of Imam al-Dawr (1075–90). Exterior (photo: Yasser Tabbaa)

chamber is transformed into an octagon through the use of large and heavily profiled squinches (Figure 10.2). Eight smaller squinches (or, in fact, large muqarnas cells) rest on this octagon and form an eight-pointed star with four windows. The rest of the dome is made up of three more eight-celled tiers with ever-diminishing cells, each with a 45-degree rotation and a little cupola on top (Figure 10.3). It would have been quite feasible to build a smooth dome after the squinch zone, but a deliberate choice was made instead to continue the intricate layering of eight-celled muqarnas tiers until the desired height Figure 10.2  Al-Dawr. Shrine of Imam al-Dawr. Detail of squinch (photo: Yasser Tabbaa)

203

204

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 10.3  Al-Dawr. Shrine of Imam al-Dawr. Interior of dome (photo: Yasser Tabbaa)

and complexity were reached. Structurally, the effect is to make the dome appear insubstantial, as the play of light on its intricate surfaces dissolves its mass. This visual display, totally missing in Seljuq Iranian domes, is one of the most important features of muqarnas vaults. This earliest muqarnas dome, appearing as it does in a small village of little historical significance, is unlikely to have been the first of its kind or the model for all later muqarnas domes. First of all, because Islamic architecture prospered mainly in cities, it is in cities that one should look for major innovations. Second, Muslim ibn Quraysh, the patron of the shrine, was in certain respects a vassal of the ʿAbbasid caliph, and despite his Shiʿi inclinations, maintained strong and generally friendly links with the caliphate.13 Third, although no early muqarnas domes are preserved in Baghdad, two miniatures, one dated 873 (1468) and the other 944 (1537), show bird’s-eye views of the city with numerous muqarnas domes (Figures 10.4–10.5). Taken together, these suggest that Baghdad might have been the centre Figure 10.4  ‘Flood of Baghdad’, from a dispersed Turcoman manuscript (1468). British Library Add. 16561, fol. 60a

THE MUQARNAS DOME

Figure 10.5  View of Baghdad. Matrakçi, Beyan-i Menåzil-i Sefer-i ʿIråkeyn (1537)

in which the muqarnas dome originated. It was certainly a very common feature of the cityscape by the late medieval period.14 It is curious therefore that the next dated examples of the muqarnas dome are not from Baghdad at all, nor are they from anywhere in Iraq, but come from such diverse places as Damascus, Palermo, Fez and Tinmal. All date from the twelfth century. Possibly the earliest example is in the mosque of the Qarawiyyin at Fez; it dates from the restoration under the Murabitun (Almoravid) dynasty between 1135 and 1140. The whole axial nave of this mosque is covered by a series of stucco muqarnas vaults of great complexity and excellent execution (Figure 10.6).

205

206

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 10.6  Fez. Mosque of al-Qarawiyyin. Muqarnas vault, 1132–42. From Terrasse, Mosquée al-Qaraouyin, pl. 32 (photo: Yasser Tabbaa)

The sudden appearance of an architectural form at such an advanced level of development has posed a major problem for architectural historians. Henri Terrasse, who devoted a monograph to the mosque of the Qarawiyyin, asserts that, although earlier examples of the muqarnas exist in Iran, the muqarnas vaults at Fez are the earliest of their kind and are therefore the product of local development.15 This argument can be refuted in part because an earlier example does exist, in the form of the shrine of Imam al-Dawr of 1087 already mentioned. A much closer, though slightly later, parallel can be found in the måristån of NËr al-DÈn in Damascus, built in 1154.16 The måristån al-Nuri contains four specimens of muqarnas vaulting used in three different ways: as a portal vault (the earliest of its kind), as a muqarnas dome (Figure 10.7) and as a vault for a niche (Figure 10.8). The similarity between the Fez vaults and those of the måristån (especially the two niches) is quite clear (cf. Figures 10.8 and 10.6). These vaults are also made of stucco and serve no structural purpose – they are merely suspended by a wooden framework from the load-bearing vault above them. They are equally intricate, and they both contain pendants and terminate in eight-pointed stars. The muqarnas vaults of the måristån, like other features in it, such as its co-axial four-iwån plan, reflect Iraqi influence.17 It is reasonable to suggest, therefore, that the vaults of the mosque of the Qarawiyyin are not a local development but rather a direct import from Baghdad. By far the largest number of independent muqarnas domes are found in Iraq and the JazÈra, over an area extending from Basra and Khuzistan in the south to Mosul and Raqqa in the north (Figure 10.9). The majority of these shrines are dated, or datable to, between the middle of the twelfth century and the Mongol invasion,18 a period of great building activity in Iraq and the JazÈra. A preliminary typology, largely based on building material and construction method, of

THE MUQARNAS DOME

Figure 10.7  Damascus. Bimåristån al-Nuri (1154). Dome over the vestibule (photo: Yasser Tabbaa)

Figure 10.8  Damascus. Bimåristån al-Nuri. Vault over niche (photo: Yasser Tabbaa)

the more than twenty shrines known to me can yield the essential features of the muqarnas dome. The first, but least common, type in Iraq is the vault made of stucco and suspended from the exterior vault above it by a wooden framework. This is the method found in the måristån al-Nuri and the mosque of the Qarawiyyin (Figures 10.6, 10.7 and 10.8), and it later became extremely popular in North Africa and Spain. The second type, which became most common in Iraq, is made of brick and consists of a single shell: it is the only type in which the interior articulations of the muqarnas are reflected on the exterior. The shrine of Imam al-Dawr illustrates its earliest manifestation, but it is not really typical of later, more developed examples. Its cells are too large to suggest a true muqarnas, even though their multiple profiling was meant to convey that impression. In addition, the cells spring only from the corners of the octagon, in contrast to later instances where they spring from the walls as well. Three

207

208

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

outstanding examples of this type are the mausoleum of NËr al-DÈn in Damascus, the shrine of Zumurrud Khatun in Baghdad and a little-known shrine called al-Najmi in south Iraq. The mausoleum of NËr al-DÈn, dated 567 (1172), is part of a larger complex which includes a madrasa and a masjid.19 Unlike its prototype at the måristån of NËr al-DÈn, which is made of stucco, the muqarnas dome over this shrine is made of brick, and its interior articulation is reflected on the exterior (Figure 10.10).20 The dome springs from a square base which is divided into twelve parts: four tiny squinches and eight small muqarnas columns (Figure 10.11). The muqarnas columns support the central portion of the four sides which consist of three tiers of corbelled muqarnas cells. In the second zone the middle sections taper upward to resemble a trapezoid with considerable horizontal projection; the one-cell squinches expand into two, three and finally four cells (Figure 10.12). By the third zone, the square base changes into an irregular octagon formed by the shrunken middle sides and the expanded squinches. This zone contains four axial windows, the lowest of them on the exterior. In the fourth zone, the octagon is transformed into an approximately uniform circle of twenty elements. Ten intersecting arches in the fifth zone reduce the number of the cells to ten, and a little scalloped dome of ten elements rests on this drum. The drum also contains ten tiny windows. Thus, by using extremely unobtrusive squinches in combination with corbelled muqarnas cells, it was possible to move gradually and almost imperceptibly from the square to the octagon to the circle.

Figure 10.9  Muqarnas domes in Iraq and Syria

THE MUQARNAS DOME

Figure 10.10  Damascus. Madrasa al-Nuriyya al-Kubra (1172). Mausoleum of NËr al-DÈn. Exterior (photo: Yasser Tabbaa)

The shrine of Zumurrud Khatun in Baghdad has the most graceful profile and one of the most integrated interiors among the monuments of its kind (Figure 10.13).21 One reason for the success of this dome is that its base is octagonal; that is, half the transition has already been made. Above an extremely unobtrusive squinch zone (Figure 10.14), the octagonal base is transformed into a muqarnas dome of sixteen cells. Seven tiers of sixteen cells make up the majority of the dome; their number is cut to ten in the last three tiers (Figure 10.15). Each cell contains a tiny opening covered by thick glass. More interesting than either of these two domes is the ruined shrine of al-Najmi, located on a once major canal called Shatt al-Nil in what is today a desolate salt desert.22 Although poorly preserved, enough has survived to reconstruct it as a tall muqarnas dome with an especially large span, around a hundred metres square (Figure 10.16). This feat was accomplished by the use of large triangular cells and brackets skilfully arranged in such a manner as to form corner squinches which alternate with four muqarnas pendentives, one in the middle of each side (Figure 10.17). In this manner, the square base is gradually transformed into a forty-part base which begins just

209

210

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

above a zone of eight windows. The conical part of the dome would have begun just above this base. The third type of Iraqi muqarnas has an exterior pyramidal roof and is made of fired brick. In the earliest examples, the pyramidal roof was usually covered by green-glazed tiles; traces of revetment are still visible in the shrine of Yahya Abuʾl-Qasim.23 This type is restricted to Mosul, where it seems to have developed around the beginning of the thirteenth century under the influence of Baghdad.24 Five examples are preserved there; the finest of them is the shrine of Imam ʿAwn al-DÈn (Figure 10.18).25 The design of this dome is more precise and rigorous than that of the Baghdad examples or even of the mausoleum of NËr al-DÈn, which it otherwise resembles, perhaps because muqarnas cells are faced with tiny colour-glazed tiles, a feature unique to Mosul. As in the mausoleum of NËr al-DÈn the gradual transition from the square to the octagon is made by the use of muqarnas squinches and small muqarnas colonnettes. Here, however, the squinches are themselves turned into tiny muqarnas domes which end in eight-pointed stars (Figure 10.19). Above the octagonal zone, the dome is further divided into sixteen cells; the number is reduced to eight in the last two tiers (Figure 10.20), producing a large eight-pointed star in the middle surrounded by four small eight-pointed stars. This ingenious and pleasing design is imitated and further developed in the shrine of Shaykh ʿAbd alSamad at Natanz.26 All these muqarnas domes share some basic features: (1) they are made of small but distinct cells; (2) their squinches, colonnettes and other structural features are obscured; (3) layers of stucco, paint or glazed tiles are often used to embellish the cells; and (4) windows are used frequently, though of course in a double-shell design they are only possible at the base of the dome. Having established the origin and the geographic and temporal distribution, typology and basic features of the muqarnas dome, it is reasonable to inquire into its raison d’être. What led the Muslim architect in this period to abandon the smooth hemispherical dome with its age-old symbolic associations and take up this fragmented conical vault? What meanings were intended which differed from those inherent in the hemispherical dome, and how did this new form carry these meanings? We know that the form may have originated in Baghdad sometime in the early eleventh century, that it very rapidly spread to Syria and North Africa, and that it was used first as a funerary monument and later in mosques, hospitals, fountains27 and even palaces. To my knowledge, the only explanation that has been offered is that provided by Oleg Grabar for certain muqarnas domes in his book on the Alhambra. Grabar’s interpretation is not derived from the form of the muqarnas itself, but rather rests on evidence external to it: water symbolism, Qurʾånic and poetic inscriptions, and poems

THE MUQARNAS DOME

written in praise of the muqarnas domes of the Alhambra.28 From this evidence, Grabar comes to the conclusion that the Abencerrajes and the Hall of the Two Sisters represent rotating domes of heaven.29 This plausible but specific interpretation tempts one to work backward using similar, though scantier, evidence to interpret earlier muqarnas domes in the same manner. There are, however, two difficulties in this approach: first, since it is based on external, and often unavailable, evidence it can be used to explain some muqarnas domes but not the phenomenon in general. Second, since any dome can be a dome of heaven it does not really tell us why a muqarnas dome was used. To avoid these objections I chose an approach that was exactly the opposite of Grabar’s and sought the meaning of the form in the form itself. I started with the premise that subdividing matter into tiny interrelated segments implied a certain attitude towards matter, or, more specifically, that the division of a dome into segments implied a certain conception not just of the dome but of its referent, the universe. Muslim philosophers and theologians devoted considerable thought to the nature of matter and the universe and their relationship with God. The Aristotelian concept of an eternal cosmos was rejected by most Muslim theologians from the first because it contradicted the Islamic conception of God as the only absolute and eternal. From very early on and ‘with hardly a single exception, the Muslim theologians accepted the atomic view of matter, space and time and built upon it a theological edifice over which God presided as absolute sovereign’.30 Accordingly, matter was neither eternal and immutable nor infinite in composition, but rather composed of particles which cannot be divided any further: al-juzʾ alladhi lå yatajazzaʾ.31 The Ashʿarites of the tenth and eleventh centuries, in particular al-Baqillani (d. 1013), modified this atomistic theory into one of strict occasionalism – that is, a theory of atoms and accidents (aʿrådh, pl. of ʿardh). Al-Baqillani argued that the world, which to him was everything other than God, was composed of atoms and accidents; accidents could not endure within matter (jawhar) for longer than an instant, but were continuously being changed by God.32 It follows then that the attributes of matter (colour, luminosity, shape, etc.) are transitory accidents that change according to the will of God and that even the preservation of matter – the collocation of its atoms – requires the continuous interference of God. This was a solid argument for the existence of God as the only creator, for since such a world was created and is continually being created, then it must by necessity have a creator.33 I would like to suggest that the muqarnas dome is an architectural manifestation of this thoroughly orthodox Islamic concept. Its likely origin in Baghdad in the early eleventh century coincides well with the triumph of the atomist-occasionalist view of the

211

212

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 10.11  Damascus. Madrasa al-Nuriyya al-Kubra. Dome over mausoleum. Interior (photo: Yasser Tabbaa)

universe as formulated by al-Baqillani and supported by Caliph al-Qadir ­(991–1031).34 lt must have become obvious to al-Qadir, or a mathematician-architect in his court, that the usual smooth dome that rests on squinches could no longer express this truly new Muslim view of the universe: it was too solid and continuous; its particles were imperceptibly small; and it was visibly supported by squinches.35 To reflect an occasionalist view of the universe, the dome would have to be divided into small but distinct units arranged in a complex manner, and (like the universe) supported and kept whole by the will of God – thus the de-emphasis of the squinches, clearly the work of man, a feature common to all muqarnas domes. In fact, the earliest muqarnas in north-eastern Iran was nothing more than the division of the squinch zone into three or more parts, undoubtedly in order to de-emphasise its structural appearance. The dome, like heaven, had to stand unsupported: ‘khalaqa al-samåwåt bi-lå ʿimåd’ (Qurʾån 31:10). The muqarnas cells of many of these domes were painted, and windows were frequently and effectively used. This is nowhere truer than in the muqarnas domes at Imam ʿAwn al-DÈn and of course the

THE MUQARNAS DOME

Figure 10.12  Damascus. Madrasa al-Nuriyya al-Kubra. Dome over mausoleum. Detail of corner (photo: Yasser Tabbaa)

Alhambra. According to Grabar, the changing sun- and moonlight were intended to give the impression of a rotating dome of heaven, a view which is supported by Ibn Zmarak’s poem inscribed in the halls of the Abencerrajes and the Two Sisters.36 This is perhaps a secondary meaning that became attached to the muqarnas dome in its three centuries of evolution, but it was not the original intention. Rather the effect of light on these intricate painted surfaces was meant to reflect certain very important Ashaʿri concepts, namely that shape, colour and luminosity are accidents which by definition are subject to continuous change according to the will of God. The dome is therefore not just a physical manifestation of the occasionalist universe, but also a proof of the existence of a God Who can keep this seemingly unsupported, perishable, and ever-changing dome from collapsing, just as He can keep the universe from destruction.37 The muqarnas dome should therefore be viewed not merely as a decorative device to fill some of the space left vacant by the Muslim injunction against religious images, but rather as a uniquely Islamic

213

214

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 10.13  Baghdad. Shrine of Zumurrud Khatun (1180–1220). Exterior (photo: Yasser Tabbaa)

solution firmly grounded in the theology of its time. Although undeniably decorative – and in later periods admittedly used solely for decoration – at the time of its creation and up to the time of the Alhambra that was not its primary purpose. Neither could it have been the product of mathematical or architectural experimentation alone. Mathematics and architecture were simply the tools used to flesh out a major theological concept about the nature of the universe and its relationship to God. That the occasionalist concept permeated Islamic culture can be seen in the parallel developments

THE MUQARNAS DOME

Figure 10.14  Baghdad. Shrine of Zumurrud Khatun. Springing of dome (photo: Yasser Tabbaa)

Figure 10.15  Baghdad. Shrine of Zumurrud Khatun. Interior (photo: Yasser Tabbaa)

215

216

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 10.16  Shatt al-Nil. Shrine at al-Najmi. Exterior (photo: Yasser Tabbaa)

Figure 10.17  Shatt al-Nil. Shrine at al-Najmi. Detail of muqarnas (photo: Yasser Tabbaa)

THE MUQARNAS DOME

Figure 10.18  Mosul. Shrine of Imam ʿAwn al-DÈn (1245). Exterior (photo: Yasser Tabbaa)

of the eleventh and twelfth centuries in architectural ornament (the arabesque and overall star patterns)38 and even in music (increasing embellishment around a common mode)39 that are also explainable in terms of occasionalist concepts. Conceived in these terms, the manifold functions of the muqarnas dome and its quick westward spread cease to be mysteries. Although it was a form deeply rooted in theology, it had no specific liturgical associations and could therefore be used in both religious and secular contexts to enhance the sanctity of the precinct and induce meditation. Its almost immediate appearance in Syria and North Africa can also be explained in both religious and political terms. On the religious level, the muqarnas dome was adopted by the rising SunnÈ forces of Syria and North Africa in the persons of

217

218

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 10.19  Mosul. Shrine of Imam ʿAwn al-DÈn. Interior of dome (photo: Yasser Tabbaa)

Figure 10.20  Mosul. Shrine of Imam ʿAwn al-DÈn. Upper zone of dome (photo: Yasser Tabbaa)

NËr al-DÈn and a­ l-Murabitun respectively. Doubtless it was used with full a­ wareness of its theological associations, whether as a mausoleum for the martyr (al-shahÈd ) NËr al-DÈn or as vaulting for the axial nave and mi˙råb dome of some North African congregational mosques. On the political level, the muqarnas dome provided a formal link with the ʿAbbasid caliphate, the heartland of orthodoxy and source of legitimation.40 The importance of Baghdad, a city of vanished splendour, to this development cannot be overestimated. As Herzfeld wrote, ‘One must not underrate Baghdad, seat of the caliphate and one of the seats of

THE MUQARNAS DOME

the Seljuk sultanate and a cultural centre down to its conquest by Hulagu in 656 h. (a.d. 1258). To underrate Baghdad is to underrate Rome.’41

Notes 1. I would like to express my thanks to Professors Oleg Grabar and Wolfhart Heinrichs for their helpful comments on this paper. I am also indebted to Professor Heinrichs for his remarks on the influence of atomism on Islamic literature.   The already large literature on the muqarnas and on the problem of geometry in Islamic architecture increases daily, but only a small portion of it deals specifically with the muqarnas dome. One of the earliest historical analytical studies of the muqarnas in general is by J. Rosintal, Pendentijs, trompes et stalactites dans l’architecture oriental (Paris, 1928). Much more important are the short studies by Ernst Herzfeld on a number of muqarnas domes in Iraq, Iran and Syria, published in sections of volume 2 of F. Sarre and E. Herzfeld, Archiäologische Reise im Euphrat und Tigris Gebiet, 4 vols (Berlin, 1911–14); and Ernst Herzfeld, ‘Damascus: Studies in Architecture, I’, Ars Islamica 9 (1942): 10–40. Michel Écochard, Filiation de monuments grecs, byzantins et islamiques: une question de géometrie (Paris, 1977), includes a chapter on brick and stucco muqarnas vaults with numerous analytical drawings. The only book so far published on the muqarnas domes of Iraq is ʿA††a al-ÓadÈthÈ and Hanåʾ ʿAbd al-Khåliq, al-Qibåb al-MakhriËtiyya fiʾl-ʿIraq (Baghdad, 1974). Although useful as an inventory and for some factual information, the book is marred by poorly reproduced photographs. 2. For example, Titus Burckhardt, Art of Islam; Language and Meaning (World of Islam Festival Trust, 1976), and Nader Ardalan and Laleh Bakhtiar, The Sense of Unity: The Sufi Tradition in Persian Architecture (Chicago, 1973). See also a review of these works and others in the same genre by Oleg Grabar, ‘Reflections on the Study of Islamic Art’, Muqarnas 1 (1983): 25–32. 3. Oleg Grabar, ‘Symbols and Signs in Islamic Architecture’, in Architecture and Community: Building in the Islamic World Today, ed. Renata Holod (Millerton, NY, 1983), pp. 25–32. 4. Idem, ‘Islamic Art: Art of a Culture or Art of a Faith’, AARP 13 (1978): 1–6, and ‘Das Ornament in des Islamischen Kunst’, Zeitschrijt der Deutschen Morgenländischen Gesellschajt, Suppl. 3 (1977). 5. Grabar, ‘Symbols and Signs’, p. 27. 6. R. Stephen Humphreys, ‘The Expressive Intent of Mamluk Architecture of Cairo: A Preliminary Essay’, Studia Islamica 35 (1972): 69–119. Although it deals with Mamluk architecture as a case study, the methodological discussion in this paper is very useful for all efforts at architectural interpretation. 7. The Nishapur fragments were first published by Charles K. Wilkinson et al., ‘The Museum’s Excavation at Nishapur’, Bulletin of the Metropolitan Museum of Art 33, pt 2 (1938): 9ff. Most recently, Ulrich Harb in Ilkhanidische Stalaktitengewolbe: Beiträge zu Entwurf und Bautechnik (Berlin, 1978), p. 16, reiterated the same theory about the

219

220

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

origin and development of the muqarnas in Iran without referring to Iraq. Another commonly accepted view is the one expressed by Oleg Grabar in The Alhambra (Cambridge, 1978), p. 175, that ‘the origins of the muqarnas lie in the almost simultaneous, but apparently unconnected, developments in northeastern Iran and central North Africa’. See also idem, ‘The Visual Arts’, in Cambridge History of Iran, vol. 4, From the Arab Invasion to the Saljuqs (Cambridge, 1975), pp. 344–5. 8. Compare figures 247 and 338 in John Hoag, Islamic Architecture (New York, 1977). 9. These fragments have been published by Lucien Galvin in Recherches archéologiques a la Qalʿa des BanË Óammåd (Paris, 1965), pp. 125–7. Galvin reconstructs these fragments as bundles of pendants which decorated the corners of a ceiling (fig. 40). In a later work, Essai sur l’architecture religieuse musulmane, vol. 1, Généralités (Paris, 1970), pp. 157–9, he states that the earliest muqarnas may be found in Iran, although the chain of transmission to the Maghrib is incomplete. This being the case, Galvin concludes that, until we are better informed, the first appearance of muqarnas vaulting must be placed in Qalʿat Bani Hammad. See also Henri Terrasse, La Mosquée al­Qaraouryin à Fez (Paris, 1968), pp. 31–2. Terrasse agrees with Galvin that the Qalʿa fragments came first and led to the developments of the twelfth century. 10. Several North African mosques were built or restored in this period with one or more muqarnas domes over the mi˙råb and the axial nave. Among them are the Great Mosque at Tlemcen, 1136 (Hoag, Islamic Architecture, fig. 124); the Qarawiyyin mosque at Fez, 1132–42 (Terrasse, Mosquée al-Qaraouiyin, pls 28–37); the Great Mosque at Tinmal, 1154; the mosque of al- Kutubiyya at Marrakesh, 1162 (Hoag, Islamic Architecture, fig. 133); and the somewhat earlier so-called Almoravid Qubba (1107–43) at Marrakesh. Although not quite a muqarnas dome, the Qubba nevertheless contains two tiers of highly ornamented squinches which resemble those in the shrine of Imam al-Dawr (ibid., fig. 2). See Richard Parker, A Practical Guide to Islamic Monuments in Morocco (Charlottesville, VA, 1981), pls 10–12.   An astonishing use of a large muqarnas vault is displayed in the Palatine chapel at the Norman palace in Palermo, built in 1142. See U. Monneret de Villard, Le Pitture musulmane al soffitto della Cappella Palatina in Palermo (Rome, 1950), and Annabelle Simon-Cahn, ‘Some Cosmological Imagery in the Decoration of the Ceiling of the Palatine Chapel in Palermo’, Ph.D. diss., Columbia University, 1978. This Christian monument poses some problems for my interpretation of the muqarnas dome as a form with well-established orthodox Muslim associations, especially since it employs a muqarnas vault with figural paintings on its cells. The Cappella Palatina is, in my opinion, a monument that reflects on the most general level the confluence of three architectural strains: Romanesque, Byzantine and Islamic. 11. Herzfeld, ‘Damascus, I’, pp. 18–22 with plan and section; and Hadithi and ʿAbd al-Khaliq, al-Qibab al-Makhrutiyya, pp. 19–26. 12. This information is given in some detail in five inscriptions placed within star-shaped panels located on the inner walls. See Hadithi and ʿAbd al-Khaliq, al-Qibab al-Makhrutryya, pp. 20–1, for their transcription. 13. Khåshiʿ al-MaʿadhÈdÈ, Dawlat banÈ ʿAqÈl fi’l-Mawßil (Baghdad, 1968), pp. 75–9 and 105–7, for the biography of Muslim ibn Quraysh and his

THE MUQARNAS DOME

connection with the ʿAbbasids and the Fatimids. The sources consulted by the author suggest that the Shiʿism of the Uqaylids had strong political and ethnic components. Politically the Uqaylids used their Shiʿism to keep in favour with the Fatimids while at the same time, by virtue of their geographic proximity, remaining close to the ʿAbbasids. Ethnically, the Uqaylids stood for the independence of the ʿAbbasids and other Arab dynasties against the Daylamite Persians and the Seljuq Turks. 14. The dome placed over the shrine of Imam Abu Hanifa in Baghdad by Sharaf al-Dawla, the finance minister of Alp Arslan, in 459 (1064–5) may have been a very early muqarnas dome. It is described in Ibn al-JawzÈ, al-MuntaΩam fÈ TårÈkh al-MulËk waʾl-Umam (Hyderabad, 1940), 8: 245–6, as having been tall and well built. It is illustrated in Ma†ra˚çi, Beyan-i Menåzil-i Sefer-i ʿIråkeyn-i Sultån Süleymån Hån, ed. H. Yurdaydin (Ankara, 1976), folio 53b. Two independent shrines are shown in the miniature, one octagonal with a hemispherical dome and the other square with a conical muqarnas dome. No further information is provided in the text. 15. Terrasse, Mosquée al-Qaraouiyin, pp. 31–2. 16. Herzfeld, ‘Damascus, I’, pp. 2–11. See also Yasser al-Tabbaa, ‘The Architectural Patronage of NËr al-Din, 1146-1174’, Ph.D. diss., New York University 1982, pp. 100–2. 17. Al-Tabbaa, ‘Architectural Patronage’, pp. 103–6. 18. For the dating of these shrines, see Hadithi and ʿAbd al-Khaliq, alQibab al-Makhrutiyya, passim, where the authors attempt to date the undated shrines by comparing them with the firmly dated ones. 19. Herzfeld, ‘Damascus, I’, pp. 11–14, and al-Tabbaa, ‘Architectural Patronage’, pp. 125–9. 20. This is not the case with the måristån’s dome, which originally was smooth on the exterior. See Jean Sauvaget, ‘Notes sur quelques monuments musulmans de Syrie à propos d’une étude récente’, Syria 24 (1945), fig. 1, which shows the exterior of the dome before its faulty restoration. 21. Sarre and Herzfeld, Archäologische Reise, 2: 173–9. 22. Ibid., p. 239. 23. Hadithi and ʿAbd al-Khaliq, al-Qibab al-Makhrutiyya, pp. 51–7. 24. There are no muqarnas domes in Mosul that can be dated to the twelfth century. The shrine of Imam ʿAbd al-Rahman, which was built by the atabek Masʿud ibn Mawdud (1180–93), is conical on the exterior but smooth on the interior. The only dated examples – the shrines of Imam Yahya ibn al-Qasim and Imam ʿAwn al-DÈn – are both from the reign of Badr al-DÈn Luʾluʾ (1222–59). 25. Hadithi and ʿAbd al-Khaliq, al-Qibab al-Makhrutiyya, pp. 63–6; Herzfeld, ‘Damascus, I’, p. 37; and Sarre and Herzfeld Archäologische Reise, 2: 263. 26. The shrine of Shaykh ʿAbd al-Samad at Natanz was built by Uljaytu (1304–17) under Iraqi influence. Uljaytu had already commissioned one muqarnas dome in Iraq for the shrine of Dhuʾl-Kifl (Ezekiel) at al-Kifl before the Natanz shrine. Uljaytu had the shrine rebuilt with a muqarnas dome and with a minaret adjacent to it, undoubtedly to emphasise its change from a Jewish to a Muslim sanctuary. The Natanz shrine, however, resembles, not the Kifl shrine, but the shrine of Imam ʿAwn al-DÈn at Mosul. The points of similarity are the double-shell design, the exterior pyramidal roof with blue glazed tiles, and squinches

221

222

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

that end in eight-pointed stars. The main difference is that the Natanz shrine has a cruciform plan, whereas ʿAwn al-DÈn has the usual square one. There is, however, one muqarnas dome in Iraq with a cruciform plan, the so-called mashhad al-Shams at al-Hilla. See Hadithi and ʿAbd al Khaliq, al-Qibab al-Makhrutiyya, pp. 81–3, and Herzfeld, ‘Damascus, I’, pp. 29–30 and 38. See also Sheila Blair, ‘The Shrine Complex at Natanz, Iran’, Ph.D. diss., Harvard University, 1980. 27. I know of three twelfth-century instances of a water fountain emanating from a muqarnas niche: the first (now almost totally destroyed) in the west iwån of the Madrasa al-Nuriyya al-Kubra (1172); the second in the Ziza palace outside Palermo (1180); and the third in the north iwån of the palace of al-Aziz Muhammad in the citadel of Aleppo (c. 1220). 28. Grabar, Alhambra, pp. 144–8. 29. Ibid., p. 147. 30. Majid Fakhry, A History of Islamic Philosophy, 2nd edn (New York, 1983), p. 33. 31. See Encyclopaedia of Islam, 2nd edn (henceforth El2), s.v. ‘Djuzʾ’. 32. Fakhry, Islamic Philosophy, pp. 209–11; in EP2, s.v. ‘Djawhar’; Francis E. Peters, Allah’s Commonwealth (New York, 1973), pp. 486–8. 33. Fakhry, Islamic Philosophy, p. 211, and Abu Bakr al-Båqillåni, Kitåb al-TamhÈd (Beirut, 1957), p. 18. 34. For a brief history of Caliph al-Qadir, see EI2, s.v. ‘al-Kådir Bi’llah’; Peters, Allah’s Commonwealth, pp. 589–91; and George Makdisi, ‘The Sunni Revival’, in Islamic Civilization, 950–1150, ed. D. S. Richards (Oxford, 1973), pp. 15–57. 35. There are some indications in the sources that in the late tenth century SunnÈs of Khurasan were creating new ceremonies and erecting new monuments (qibåb) designed to be counterparts to those of the Shiʿis; see Makdisi, ‘Sunni Revival’, p. 156, and lbn al-Jawzi, al-Muntazam, 7: 265–6. 36. Grabar, Alhambra, pp. 146–8. 37. Perhaps the first scholar to address the question of atomism in relation to the aesthetics of Islamic art was Louis Massignon in ‘Les Méthodes de réalisation artistique des peuples de l’Islam’, Syria 2 (1924): 47–53, and 149–60, esp. pp. 50–3. His ideas were in part followed up by Bishr Farès in Essai sur l’ esprit de la décoration islamique (Cairo, 1952). In the recent Istanbul conference entitled ‘The Common Principles, Forms and Themes of Islamic Art’ (April 18–22, 1983) a presentation by Ali Louati, entitled ‘Khawå†ir ˙awl al-Wi˙ida al-Jamåliyya li’l-Turåth alFanni al-Islåmi’, suggested that Ashʾarite atomism permeated Islamic artistic, architectural and urban aesthetics in all periods and places. While his paper represents an advance over preceding work, it too overgeneralises. It would be difficult – to give two examples far apart in time and space – to apply atomistic philosophy either to the austere architecture of the Ayyubids of Aleppo or to the naturalistic art and symbolic architecture of the Mughals of India. 38. The best study on arabesque ornament is Ernst Kühnel, The Arabesque, tr. Richard Ettinghausen (Graz, 1976). The earliest examples of true arabesque occur in Fatimid Egypt, on both architecture and especially wooden objects. But its greatest development occurs under the Rum Seljuqs in Anatolia and the Ayyubids in Syria. 39. Massignon, ‘Méthodes’, pp. 154–8. At the Istanbul conference (above, n. 37) the ethnomusicologist Mahmoud Guettat (of the Institute for

THE MUQARNAS DOME

Music at Tunis) dealt with improvisation and embellishment in Islamic Tunisian music. The situation in literature is not so clear, since characteristics that might be called ‘atomistic’ are already present in preIslamic poetry. The problem has been discussed by G. von Grünebaum, ‘The Spirit of Islam as Shown in Its Literature’, in Islam: Essays in the Nature and Growth of a Cultural Tradition, 2nd edn (London, 1961), pp. 95–110. According to von Grünebaum, in reference to Arabic poetry, ‘an exclusive attention seems to be given to the individual verse, phrase, or paragraph, at the expense of the consistent layout of the whole’. He adds a little later (p. 98) that ‘it may be tentatively and somewhat hesitatingly suggested that there exists a certain psychological affinity between the leaping from topic to topic, these momentary shifts of attention and mood, and the occasionalist world view which dominates Muslim theology and scholastic philosophy’. Although von Grünebaum states in his article that Ashʾarism was most successful and effective around the time of al-Baqillani (late tenth and early eleventh centuries), he does not explain what might have influenced Arabic poetry and prose in the earlier centuries or whether Ashʾarism enhanced a trend which had always existed in Arabic literature. 40. There are other instances of architectural borrowing from Baghdad by North African dynasties in this period and before. Early and well-known examples are the polychrome lustre-glazed tiles which were imported from Baghdad under the Aghlabid Ziyadat Allah I (817–38) to decorate the voussoir of the arch of the Great Mosque of Qayrawan. 41. Herzfeld, ‘Damascus, I’, p. 38.

223

CHAPTER ELEVEN

The Muqarnas Domes and Vaults of Cairo Scholars may debate the nature of Islamic architecture in different periods and places, but most would concede a special place for muqarnas vaulting as one of its distinctive forms. During the period of its greatest spread and innovation, between the twelfth and fifteenth centuries, muqarnas vaults and domes helped impart a sense of luxury to medieval Islamic architecture and a measure of unity among its diverse regional styles, from Spain to India. Yet, despite its tremendous visual appeal and unparalleled architectural importance, the muqarnas remains somewhat enigmatic in terms of its origin, earliest development and meaning, whether original or modern. More specifically, its introduction to Egypt and its role in shaping the domes and portal vaults of Mamluk monuments continues to be poorly understood and subject to various interpretations. Even when writing generally on the muqarnas it is important to keep in mind a subtle but significant distinction between the muqarnas as a decorative system and the forms to which it was applied, including domes, vaults, cornices and capitals. In short, it seems clear that the muqarnas is not an architectural form as such but a type of three-dimensional ornament – consisting of a variety of spherical sections, brackets and pendants – which was applied during the medieval period to various architectural forms, resulting in their characteristic transformation into geometrically fragmented forms. This essay will therefore discuss the process of this transformation and will examine its most significant products, namely the portal vault and the dome, which form the points of entry and culmination of many Islamic monuments in Cairo and elsewhere. The origin of the first muqarnas forms has been debated for several decades, and the issue is still incompletely settled.1 The main point of contention is that muqarnas-like forms seem to occur in the eleventh century in different parts of the Islamic world, including North Africa, upper Egypt, Iraq and Iran. This has led some scholars Yasser Tabbaa (2002), ‘The Muqarnas Domes and Portal Vaults of Cairo’, in Bulletin of the American Research Center in Egypt, 182 (Spring–Summer), 1–7.

THE MUQARNAS DOMES AND VAULTS OF CAIRO

Figure 11.1  The dome in the shrine of al-Faraj ibn Barquq, Cairo (1405) (photo: Tim Loveless)

to propose a theory of simultaneous development for the muqarnas, whereby these geographically disparate regions would have arrived at a similar form at approximately the same time. More recently, scholarly opinion has shifted to eastern Iran as the place of origin of

225

226

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

incipient muqarnas forms and to Iraq, specifically Baghdad, as the centre of the most significant developments in muqarnas vaulting.2 Reviewing the archaeological evidence, therefore, we note that in the excavations in Nishapur (eastern Iran) a number of concave triangular stucco pieces were discovered within a generally late tenth-century context and tentatively reconstructed as decorative corner squinches.3 While this reconstruction remains conjectural, a tripartite squinch exists in situ in the late tenth-century shrine of Arab Ata at Tim, near Samarqand.4 Subsequently, the same feature is used with great success in several late eleventh-century Iranian domes, such as the two at the Great Mosque of Isfahan (Figure 11.2). Characterised by their smooth hemispherical profile, these large domes spring from a zone of subdivided squinches whose arched profile is often repeated all along the rim of the dome, creating a continuous zone of transition. But with hardly any exceptions, this process of subdivision is not applied to the dome itself, which in Iran remains virtually unchanged up to the fourteenth century. In other words, the differentiation of the squinch zone into three or five elements is not taken any further to produce complete muqarnas domes.5 A number of small mausoleums in upper Egypt and Cairo, all dating from the late eleventh and twelfth centuries, also contain subdivided transition zones, of which the most developed examples, such as the shrine of Sitt Ruqayya (1135) in southern Cairo, do approximate the appearance of muqarnas domes (Figure 11.3). Although these shrines date from the Fatimid period, they are by and large the products of local piety, not imperial patronage, as the Fatimid caliphs themselves continued to favour smooth domes resting on simple squinches to the end of their dynasty in 1170. Rather than being a local creation, as K. A. C. Creswell had argued, these domes were probably built in imitation of domes that Egyptian pilgrims to Mecca would have seen in and around the Holy Shrine.6 Some of these domes may have been built by various ʿAbbasid caliphs in the tenth and eleventh centuries, and they are the ones vividly described by the Moroccan traveller Ibn Jubayr in the twelfth century. The earliest extant true muqarnas dome is the so-called shrine of Imam Dur, dated 1088, located in the outskirts of Samarra, about fifty miles north of Baghdad. It consists of Figure 11.2  Isfahan, Great an elongated chamber about twelve metres Mosque, North Dome, 1086, high topped by a muqarnas dome of almost transition zone (photo: Yasser Tabbaa) the same height. The square of the chamber

THE MUQARNAS DOMES AND VAULTS OF CAIRO

Figure 11.3  Cairo, Shrine of Sitt Ruqayya, 1135, transition zone (photo: Yasser Tabbaa)

is transformed into an octagon by four squinches and four arches (Figure 11.4). A succession of four more eight-celled tiers with everdiminishing cells, each with a 45-degree rotation, make up the rest of the dome. The layering of increasingly small and multiply profiled cells makes the dome appear insubstantial, as the play of light on its intricate surfaces dissolves its mass. Such visual display, totally missing in Seljuq domes, is one of the most important features of the muqarnas dome (cf. Figures 11.2 and 11.4).7 A form at this level of development, existing as it does in a small village, suggests the existence of earlier models in an urban centre. The most likely place is Baghdad, which in the early eleventh century was witnessing a measure of cultural revival and political independence under the stridently SunnÈ leadership of Caliph alQadir (991–1031). It was during his reign that muqarnas was applied, perhaps for the first time, to the dome, thereby creating a fragmentary and ephemeral structure which seems to reflect the theology of atomism and occasionalism that was propagated by the caliph’s chief apologist, al-Baqillani. Since the dome had for centuries been linked to the celestial sphere of heaven, one might propose that the muqarnas dome may have been intended as an Islamic dome of heaven, or more specifically, as an allegory of the occasionalistic universe of SunnÈ theology.8 By around the middle of the twelfth century a large variety of muqarnas domes and vaults are known in the monuments of the Almoravids in Morocco and the Zangids in Damascus, as well as the Normans in Sicily. The enigmatic Qubbat al-Barudiyyin in

227

228

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 11 4  Samarra. Shrine of Imam Dur (1088), interior (photo: Yasser Tabbaa)

Marrakech (Figure 11.5) datable to 1117, though not a muqarnas dome as such, is difficult to imagine without some knowledge of Iraqi muqarnas domes such as Imam Dur.9 The lobed dome rests on an octagon inside an eight-pointed star made by the intersecting ribs of two rotated squares. The little squares in the corners are covered by tiny muqarnas domes, and the transition to the central dome is so highly elaborated with foliate arches and vegetal ornament that the whole does indeed reproduce the feeling of a muqarnas dome. Fully developed muqarnas domes are seen in Morocco only a few

THE MUQARNAS DOMES AND VAULTS OF CAIRO

Figure 11.5  Marrakesh. Qubbat al Barudiyyin (1117), dome interior (photo: D. Fairchild Ruggles)

decades later when, around 1140, the entire roof of the axial nave of the mosque al- Qarwiyyin in Fez was rebuilt using a succession of superbly crafted muqarnas vaults, all made, in the Iraqi tradition, out of carved stucco (Figure 11.6).10 Some of these stucco vaults are quite comparable to the slightly later ones at the hospital of NËr al-DÈn in Damascus (1154): they are all made of stucco and suspended from the load-bearing roof by a wooden framework; they contain pendants; and they often end in eight-pointed stars (Figure 11.7). The mausoleum of NËr al-DÈn, built in 1171 as part of his madrasa, resembles the dome in his hospital but is larger, taller and more rigorously designed. It is also the last Iraqi-style muqarnas dome to be built in Syria. Muqarnas is also used in the portal vault of the hospital of NËr al-DÈn, marking the earliest extant example of a type that becomes popular in succeeding decades. The muqarnas portal vault and dome is subjected to another important development in Syria, namely its translation into stone, a

229

230

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 11.6  Fez. Great Mosque of al-Qarwiyyin. Almoravid rebuilding of the axial nave (1134–43): muqarnas vault (photo: Yasser Tabbaa)

Figure 11.7  Aleppo. Madrasa al Zahiriyya (c. 1200), portal vault (photo: Yasser Tabbaa)

THE MUQARNAS DOMES AND VAULTS OF CAIRO

Figure 11.8  Damascus, Bimåristån al-Nuri (1154), vault over vestibule (photo: Yasser Tabbaa)

process that required considerable ingenuity in stereotomy and design. This development began in Aleppo, sometime in the last third of the twelfth century, where muqarnas portal vaults gradually replace all other varieties (Figure 11.8). The importance of this development for the course of medieval Islamic architecture is attested to by the fact that several Anatolian buildings bear the signatures of Syrian craftsmen and also by the use of the term muqarnas shami (Damascene) and halabi (Aleppine) in the artisanal language of Mamluk craftsmen.11 Figure 11.9  Cairo. Shrine of Imam Shafi (1210) (photo: Yasser Tabbaa)

231

232

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

With the exception of the sporadic use of muqarnas elements in some Cairene and upper Egyptian domes and its quite innovative application to the façade of the mosque al-Aqmar in Cairo (1125), its complete integration into Cairene architectural practice remains curiously delayed until the beginning of the thirteenth century and even later. Furthermore, unlike Spain, North Africa, Iraq, and, to some extent, Timurid Iran, the complete muqarnas dome is rarely used in Mamluk architecture.12 Rather, muqarnas is contained within broad and complex transition zones – sometimes consisting of three or even four layers of cells – that form complete concentric halos beneath Figure 11.10  Cairo. Madrasa of Sultan Hassan their respective domes, such as the magnificent shrine of Imam Shafiʾi (1211) (Figure 11.9) and the shrine (1356), portal vault (photo: Yasser Tabbaa) of the ʿAbbasid Caliphs (c. 1240). These transition zones, which are ultimately based on differentiated squinches, continue to the end of the Bahri Mamluk period.13 The portals of these buildings, on the other hand, are devoid of muqarnas, continuing instead the ‘radiant’ vaults of late Fatimid architecture. While the Fatimid reluctance to use architectural forms developed in the SunnÈ Islamic world might be understandable in view of their heterodox position, the lack of Ayyubid muqarnas portals is an anomaly without ready explanation. Indeed, it is not until the beginning of the fourteenth century that Cairene architecture develops its own version of the stone muqarnas portal vault and muqarnas transition zones. The first muqarnas vaults in Cairo resemble those in Syria, with their relatively simple geometry, large component cells and a curvilinear profile that ends in a small scalloped dome. But most, including those topping the portals of the madrasa of Sultan Hassan (1356) (Figure 11.10) and the Mosque of al- Muʿayyad Shaykh (begun 1415) (Figure 11.11), are truly grand affairs with many layers of tiny cells that include numerous pendants. Their inspiration seems to be not so much the stone muqarnas of Syria as the stucco muqarnas vaults of Timurid Iran, whose astonishing complexity is perhaps rooted in the pliability of their material. In stone, the effect is equally spectacular as a three-dimensional stellar composition but appears rather inorganically linked with the surrounding architecture (Figure 11.12). By comparison, later Mamluk domes, despite their increasingly elaborate exterior, are quite simple internally, retaining a measure of sobriFigure 11.11  Cairo. Mosque of ety and geometric rigour in their muqarnas al-Muʿayyad (1420) pendentives. They differ from their Ayyubid (photo: Yasser Tabbaa)

THE MUQARNAS DOMES AND VAULTS OF CAIRO

233

and early Mamluk counterparts mainly in that they are muqarnas pendentives rather than muqarnas squinches and that their component cells are smaller and more uniform in shape. Furthermore, these muqarnas pendentives are much taller, since they were intended for domes resting on high drums, such that their longer extension would match the increasing height of the domes and mask some of the Figure 11.12  Madrasa of Sultan awkwardness created by this attenuation. The Hassan (1356), vault on vestibule best examples, such as the dome of the shrines (photo: Yasser Tabbaa) of al-Faraj ibn Barquq ([1405]), al-Muʾayyad (1410) or al-Ghawri (1508), consist of huge muqarnas pendentives that support elaborate domes whose decoration seems to continue in two dimensions the three-dimensional geometry of their transition zones. Cairene architecture lagged behind much of the Islamic world in fully assimilating the muqarnas in its domes and portal vaults, and this delay is equally applicable to other novel features of design, ornament and calligraphy.14 Whether we attribute this significant lag to political and theological factors or to the inherent conservatism of the building craft, there is little question that muqarnas vaulting was brought into Egypt from an external, most likely eastern, source and that it did not become fully assimilated into Cairene architecture until the Mamluk period. Once established, muqarnas vaulting was inventively elaborated and combined with other ornamental features to create astonishing façades and interior spaces that mark the peak of the Islamic architecture of Cairo.

Notes 1. These issues are summarised in Yasser Tabbaa, ‘Muqarnas’, The Dictionary of Art, ed. Jane Turner (London 1996), vol. 22: 321–5. 2. For a review of the literature, see Yasser Tabbaa, The Transformation of Islamic Art during the Sunni Revival (Seattle and London 2001), 103–11. 3. Charles Wilkinson et al., ‘The Museum’s Excavation at Nishapur’, Bulletin of the Metropolitan Museum of Art 33/ii (1938), 9–26. 4. See Tabbaa, The Transformation of Islamic Art, fig. 47. 5. Possibly the first muqarnas dome in Iran is that at the shrine of ʿAbd al-Samad in Natanz, dated 1307. See Sheila Blair, The Ilkhanid Shrine Complex at Natanz, Iran (Cambridge 1986). 6. Jonathan Bloom, ‘The Introduction of Muqarnas into Egypt’, Muqarnas 5 (1988): 21–8. 7. Yasser Tabbaa, ‘The Muqarnas Dome: Its Origin and Meaning’, Muqarnas 3 (1985): 61–75. 8. Tabbaa, ‘The Muqarnas Dome’, 69–70. 9. This dome was discovered in 1947 and subsequently excavated,

234

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

restored and published in Jacques Meunié et al., Nouvelles Recherches Archéologiques à Marrakech (Paris 1957). See also Tabbaa, The Transformation of Islamic Art, 114–18, for a more recent assessment. 10. See in particular, Henri Terrasse, La Mosquée al-Qarawiyyin à Fez (Paris 1968). 11. Muhammad Amin and Laila Ibrahim, Architectural Terms in Mamluk Documents 648–923 h (1250–1517) (Cairo 1990), 113. 12. Nearly complete muqarnas domes are used with spectacular results in a small number of closely related Mamluk monuments, mostly from the second third of the fourteenth century. These include the portals of the palace and mosque of al-Amir Bashtak (both dated 1337), the portal of the palace of Qausun al-Nasiri (1338), the portal of the so-called mosque al-Fijl (1335) and the vestibule of the madrasa of Sultan Hassan (1356).   For illustrations of these vaults, see Michael Meinecke, Die mamlukische Architecktur in Agypten und Syrien, 2 vols (Glückstadt 1992), pls 56c, 53d, 56a and 76c, respectively. 13. See especially Meinecke, Die mamlukische Architecktur, pls 24–6. 14. For example, cursive monumental calligraphy, geometric ornament, and even the four-iwån plan. See Tabbaa, The Transformation of Islamic Art, 68–71, 80–3 and 130–5, for a possible interpretation of this discrepancy.

CHAPTER TWELVE

Andalusian Roots and ʿAbbasid Homage in the Qubbat al-Barudiyyin in Marrakech Without any question, I was attracted to the field of Islamic architecture and archaeology through Oleg Grabar’s famous article on the Dome of the Rock, which I first read in 1972 in Riyadh, while contemplating what to do with the rest of my life.1 More specifically the article made me think about domes as the ultimate aesthetic statements of many architectural traditions and as repositories of iconography and cosmology, concepts that both Grabar and I have explored in different ways in the past few decades. This article, on a small and fragile dome in Marrakech, continues the conversation I began with Oleg long before he knew who I was. The Qubbat al-Barudiyyin in Marrakech is an enigmatic and littlestudied monument that stands at the juncture of historical, cultural and architectural transformations (Figure 12.1). Although often illustrated, and even featured on the dust jacket of an important survey of Islamic architecture,2 this monument is in fact very little known to the English-speaking scholarly world, a situation that reflects its relatively recent discovery and its location in a country long influenced by French culture. The inaccessibility of the relevant literature on the Qubbat al-Barudiyyin makes it imperative to begin this piece with brief reconstructions of its urban, archaeological and historical contexts before addressing its architectural form and ornamental and epigraphic programme and investigating its function, the reasons for its construction and its place among related medieval Islamic domes. Like most North African architecture, the Qubbat al-Barudiyyin has generally been seen as a provincial variant of Andalusian architecture, whether in Cordoba, Seville or Granada.3 Although recent scholarship has occasionally conceded some originality to North African architecture, its local historical significance and links with the central Islamic world remain poorly understood.4 Whereas this Hispano-centric perspective might apply for Moroccan architecture of the fourteenth and fifteenth centuries – when many Yasser Tabbaa (2008), ‘Andalusian Roots and Abbasid Homage in the Qubbat al-Barudiyyin in Marrakesh’, in Gulru Necipolglu (ed.), Essays in Honor of Oleg Grabar, Muqarnas 25, 133–46.

236

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 12.1  Qubbat al-Barudiyyin in Marrakech, 1117, exterior view (photo: Yasser Tabbaa)

Andalusian artisans are known to have resettled in Morocco – it seems anachronistic in dealing with periods when Andalusia itself was ruled by dynasties from Morocco, in particular the Almoravids (1061–1147) and the Almohads (1130–1260). More specifically, to view Almoravid architecture from an exclusively Cordoban perspective goes counter to the political and cultural associations of the Almoravids, who, in addition to their Cordoban connections, actively pursued closer links with the ʿAbbasid caliphate, as I hope to demonstrate below. In dealing with the Qubbat al-Barudiyyin, therefore, I would like to reconsider its various formal and ornamental features from a multiple perspective that reflects the cultural orientation of the Almoravid dynasty, the first SunnÈ Berber state in North Africa, who saw themselves as the heirs of Cordoban glory and the propagators of ʿAbbasid symbolic hegemony. Thus, instead of simply viewing this monument as an instance of Cordoban influence, this paper places the Qubba within a matrix of relationships that addresses

ANDALUSIAN ROOTS AND ʿABBASID HOMAGE

the Almoravids’ dedication to their capital city, Marrakech, and embodies their complex cultural identity and pursuit of legitimacy.5 Furthermore, since the Qubba is in many respects the harbinger of many architectural changes that would soon become commonplace in North African architecture, it is also important to view it not just as a crucible of influences but as an innovative statement that challenged existing forms and even modes of decorum that had previously governed one of the architecturally most conservative regions in the Islamic world. Change in North African architecture, in my view, is always significant. Urban context and archaeology Although the Qubbat al-Barudiyyin appears today nearly submerged beneath the surrounding streets, it once stood proudly at the centre of the newly founded city of Marrakech (Figure 12.2). Marrakech was established by Yusuf b. Tashufin in 1070, as a military garrison and tribal capital that linked the Almoravids’ new empire with their ancestral home in the Atlas mountains. A simple mosque, a small market area (qaßaba) and a mud-brick surrounding wall were built, but the city still retained the appearance of a desert settlement, with none of the normative civic aspects of Islamic cities. All this changed under Yusuf’s son and successor, ʿAli, who acceded to the throne in 1106, and who wished to make Marrakech a great imperial capital along the models of Cordoba and Seville. One of his first building feats was to bring a steady water supply to the area, such that water reached all parts of the city and the surrounding gardens.6 It is not coincidental that the Qubbat al-Barudiyyin was a celebration of this great achievement. In addition, ʿAli b. Yusuf b. Tashufin (r. 1106–42) built several mosques, including his eponymous mosque, dated 524 (1130), which was located near the precise centre of Marrakech, at an informal square where all the main roads from the city gates converged. In 1146, on the feeble pretext that it was improperly oriented, the Almohads vindictively destroyed this mosque, as they did most Almoravid buildings, leaving only a plain, square minaret, completed in 1133, which still survives.7 On the basis of the minaret and traces of the original street pattern, it has been proposed that the mosque measured 120 x 80 metres, making it the largest such building in the entire Almoravid domain and the sole congregational mosque in Marrakech.8 The only other survival from the mosque is its famous minbar, commonly known as the Minbar al-Kutubiyya, the subject of a recent study and conservation project conducted by a joint Moroccan and American team.9 Finally, ʿAli b. Yusuf built a palace, called Dar al-Hajar (Stone House), at the south-western end of the city. This palace, too, was destroyed by the Almohads, who built on its site the al-Kutubiyya

237

238

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 12.2  Map of Marrakech under the Almoravids. (After Gaston Deverdun, Marrakech des origines à 1912, 2 vols. Rabat: Éditions Techniques Nord-Africaines, 1959–66, 1: pl. VIII)

Mosque, with its famous minaret.10 Three courtyards have been uncovered by excavation, but, other than their size, these tell us little of the original glory of this palace.11 It is possible, as Deverdun proposes in his map, that the region between the mosque and the palace of ʿAli b. Yusuf was a ‘Grande Place’ or even a garden, although this suggestion cannot be proven archaeologically. Despite its location in the middle of medieval Marrakech, the Qubbat al-Barudiyyin was actually ‘discovered’ only in 1947. Though twelve metres high from base to apex, the Qubba was literally excavated from seven or eight metres of parasitic structures, debris and ashes resulting from centuries of neglect and

ANDALUSIAN ROOTS AND ʿABBASID HOMAGE

Figure 12.3  Plan of the Qubba, showing location of the cistern and the mosque of ʿAli b. Yusuf. (After Jacques Meunié, Henri Terrasse and Gaston Deverdun, Nouvelles recherches archéologiques à Marrakech. Paris: Arts et Métiers Graphiques, 1957: fig. 1)

misuse as an annexe to a hammam. A preliminary study by Boris Maslow eventually led to a concentrated period of excavation and restoration by a French team of architects and archaeologists led by Henri Terrasse and Jacques Meunié.12 In addition to thoroughly documenting the excavation process, this team also removed some later accretions and conserved the structure but abstained from overly restoring the ornament. As a result, the Qubba stands today as a remarkably well-preserved dome resting on an attenuated rectangular base, a solitary reminder of the glory of the Almoravids in their capital city. What has completely changed is the urban context of the Qubba, as immediately signalled by the fact that the streets surrounding it have risen by about six metres. This change has been so profound that not even the careful excavation of the various layers inside and around the Qubba has succeeded in re-creating its original context or in definitively establishing its intended function.13 What we do know is that the Qubba was originally located about twenty-five metres due south of the mi˙råb wall of the mosque of ʿAli b. Yusuf, a curious location to which I will return (Figure 12.3). Furthermore, the Qubba seems to have been connected to a cistern-and-fountain structure located about ten metres to its north-east, as evidenced by clay pipes connecting the two structures and by the frequent

239

240

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

and quite similar restorations undergone by both structures over several centuries. It seems likely, therefore, that the cistern and the Qubba once belonged to an Almoravid water system, and that the Qubba may have originally sheltered a fountain, a possibility further reinforced by several levels of basins and pipes found beneath it.14 These findings might be taken to suggest that the Qubba served as the ablution fountain for the original Almoravid mosque of ʿAli b. Yusuf.15 But this otherwise plausible identification is contradicted by the location of the Qubba, which is too far away from the mosque and in the wrong direction. An ablution fountain is almost always located in the courtyard of a mosque, a few metres from the sanctuary façade, not twenty-five metres outside the mosque perimeter.16 Such a distant location, requiring the worshipper to go through the market area before reaching the mosque, is unprecedented and highly unlikely. Also complicating the identification of the Qubba as the ablution fountain of the mosque is its foundation date, which, as will be shown below, is 1117 – fifteen years before the foundation of the mosque. No such chronology comes to mind for any other mosque: ablution fountains are either contemporary with mosque foundations or much later.17 All this suggests that whereas the Qubba may have sheltered a fountain it was not necessarily intended as the ablution fountain of the Great Mosque of ʿAli b. Yusuf. Physical form In its present excavated and restored condition, the Qubba stands twelve metres tall and rests on a rectangular base 5.25 x 6.20 metres on its exterior. It is largely built of stone and brick, generally progressing from a stone foundation to stone-and-brick walls to purely brick arches and vaults. All supports and vaults are covered inside and out with a layer of plaster, thicker on the inner dome, where it is carved into various mouldings and vegetal forms. Although brick is not unknown in Spain before the early twelfth century, stone and wood were far more common materials, stone being used for columns, arches and ribs and wood for roofing and pyramidal dome covers. Indeed, the earliest use of brick in a monumental structure in Spain, the mosque at Bab Mardum in Toledo, dated 390 (999/1000), is generally attributed to eastern influence.18 Similarly, the use of piers instead of columns begins sporadically in Spain only in the early eleventh century, becoming much more popular and then predominant in North Africa in the twelfth century. The Qubba is an astonishing structure, inventively and flamboyantly decorated inside and out. The exterior is horizontally divided into three zones separated by mouldings and merlons, featuring

ANDALUSIAN ROOTS AND ʿABBASID HOMAGE

Figure 12.4  Marrakesh. Qubbat al-Barudiyyin, 1117, exterior view from south (photo: Yasser Tabbaa)

open-arched doors on the first level, arcaded galleries on the second and a solid, lightly decorated masonry outer dome on the third, covering a highly ornamented inner shell of plaster-covered masonry. The architect took advantage of the unequal sides of the rectangle to display a varied repertory of arched doors and windows – pointed, horseshoe, trilobed and foliate – set, in the Andalusian manner, within a recessed frame (alfiz). The decoration on the dome is itself divided into two zones, the lower with closely spaced interlacing arches and the upper with chevrons that surround a large sevenpointed star emanating from the centre.19 Viewed in plan, the dome seems to rest on an octagon rotated within a larger octagon that is surrounded by an eight-pointed star made by the intersecting ribs of two rotated squares (Figure 12.5). But this impression vanishes when one looks up inside the Qubba

241

242

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 12.5  Qubbat al-Barudiyyin, plan. (After Meunié, Nouvelles recherches, fig. 15, redrawn with changes by Lacey Grey)

Figure 12.6  Qubbat al-Barudiyyin, interior view of the dome (photo: Yasser Tabbaa)

ANDALUSIAN ROOTS AND ʿABBASID HOMAGE

Figure 12.7  Geometric proportions in the plan of Qubbat al-Barudiyyin (drawing: Lacey Grey)

or views it through its section. What appear in plan as continuous ribs are in fact four trilobed squinches and four trilobed arches in the middle of each side, which are surmounted by another level of eight shallow squinches rotated at 45 degrees (Figure 12.6). These are themselves surmounted by a third level of trilobed arches with muqarnas cells that support a lobed dome. Only when the lines of the first two levels of arches, which are quite distinct in section, are flattened in plan do they appear as intersecting ribs. This is an important point and a striking difference between the Qubba, with its three layers of superimposed squinches, and the domes to which it is often compared – those at Cordoba, which indeed have continuous ribs. The plan of the Qubba contains interesting geometric properties, based on the rotation of squares and the creation of √2 ratios (Figure 12.7). If we take the radius of the dome to be a, the various octagons and squares surrounding it progress sequentially as a√2 for the radius of the main octagon, 2a for half the side of the main square, 2a√2 for the distance from the centre of the dome to the end of the short buttresses, and 4a for the distance to the long side. Plotted on a

243

244

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

checkerboard with squares measuring a per side, all the important dimensions will in fact correspond to a whole number or √2 multiples of it. Although geometric planning was undoubtedly used in the layout of the Great Mosque of Cordoba, in particular the famous additions of al-Hakam in c. 965, all indications suggest that it only became a prominent feature of architectural practice under the Almohads of the second half of the twelfth century – for example at the 1154 mosque at Tinmal.20 The use of harmonic proportions also helps mitigate, but does not entirely resolve, the incongruity of using a rectangular base. Furthermore, it seems likely that the elevation of the dome and the relationship of its various levels to each other were also governed by a similar geometric principle, although this possibility remains to be tested. Examined through its section, the Qubba consists of four zones separated by mouldings: a long plain zone that contains the arched entrances, another long zone with two levels of superimposed arches, a short zone with eight muqarnas squinches, and an eight-lobed ribbed dome on top (Figure 12.8). The attenuated and projecting arches of the corner squinches partially obscure four little lobed cupolas, each resting on two layers of muqarnas cells, producing an unusual three-dimensional effect. These four tiny muqarnas cupolas create something of a starry garland around the central dome, an effect known in some eastern muqarnas domes.21 The complexity of the layered and seemingly interlaced arches, the muqarnas corner domes and the richness of the vegetal ornament create an opulent and mysterious effect that has never been surpassed by other domes in North Africa (Figure 12.9).22 Although the Qubba is not a muqarnas dome as such, its realisation seems impossible without some knowledge of such domes. In effect, it appears as a synthesis of the ribbed domes of Cordoba and the muqarnas domes of Baghdad, a cultural duality that parallels its patron’s links with al-Andalus and the ʿAbbasid caliphate (Figure 12.10). The inner, hemispherical dome rests on an octagonal zone made of eight trilobed muqarnas squinches, solidly framed on the outside – that is, in the space between the two shells – by eight slightly pointed brick arches. The dome itself is constructed of eight ribs that enclose eight lobes between them, a form known in the domes of Cordoba and Toledo. The ribs are made of radiating brick courses resting on a foundation of wooden brackets, which are inserted in the outer masonry of the structure. More wood inserts are used at the middle of the inner dome and near its apex, linking the inner shell with the outer masonry. Double-shell domes are quite common in southern Spain and Morocco, where the inner shell is usually a fragile muqarnas or artesonado dome covered externally by a sturdy pyramidal dome with a tile roof. But a double-shell masonry dome like the Qubba

ANDALUSIAN ROOTS AND ʿABBASID HOMAGE

Figure 12.8  Qubbat al-Barudiyyin, section. (After Meunié, Nouvelles recherches, fig. 14)

is completely unknown in the western Islamic world, and its use here certainly points to an eastern influence. In other words, the design of the dome, as of other parts of the Qubba, partakes of both Andalusian and eastern Islamic architecture.

245

246

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 12.9  Qubbat al-Barudiyyin, detail of an interior corner (photo: Yasser Tabbaa)

Figure 12.10  Mausoleum of Imam Dur, near Samarra, c. 1090, interior view (photo: Yasser Tabbaa)

ANDALUSIAN ROOTS AND ʿABBASID HOMAGE

Figure 12.11  Marrakesh. Qubbat al-Barudiyyin, 1117, detail of transition zone (photo: Yasser Tabbaa)

Ornament and inscriptions The precision of the plan of the Qubba contrasts with the opulence of its decoration, which softens the angles and gives the dome a deeply organic appearance (Figure 12.12). The lavishly carved stucco decorating the second and third zones consists of Andalusian conchshell motifs combined with eastern muqarnas niches and vegetal ornament. Though still displaying fairly organic leaf forms that recall Cordoba or Madinat al-Zahraʾ, the ornament is sufficiently sinuous and interconnected in composition to resemble contemporary arabesque ornament in the eastern Islamic world. It seems likely that this stucco ornament was originally painted, although no traces of paint have survived.23 Geometric ornament is restricted to the intrados of the first tier of arches, of which two panels have been preserved. The one on the south-western intrado is a simple alternation of large and small hexagonal stars. But the geometric pattern on the north-eastern intrados is surprisingly complex and advanced, displaying the inter- Figure 12.12  Qubbat al-Barudiyyin, locking of six-pointed and eight- view of interior ornament pointed stars that we ­commonly (photo: Yasser Tabbaa)

247

248

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 12.13  Qubbat al-Barudiyyin, detail of inscription (drawing: Yasser Tabbaa)

find later in the century. Although geometric ornament was judiciously used in Andalusian religious and secular monuments – in the window grilles of the Great Mosque of Cordoba and Madinat al-Zahraʾ, for example – such freely executed patterns are not known there or in the Maghreb and once again point to an eastern source.24 The Qubba contains a single historical inscription that once ran along all four sides of the interior square, measuring 15 metres (Figure 12.13). Although it was deliberately defaced by the Almohads upon their takeover of Marrakech in 1147, enough of the text remained in 1957 to allow the epigraphist Gaston Deverdun to assign the Qubba to ʿAli b. Yusuf and even to tease out from the inscription an exact date of 511 (1117). Deverdun only tentatively identified the building as the ablution fountain for ʿAli b. Yusuf’s mosque.25 Even in its fragmentary state, the inscription is noteworthy for its declamatory tone and titulature. Most historical inscriptions tend to have an even, factual tone; this one, in a rare, direct plea, calls upon God to aid ʿAli b. Yusuf in victory. A few inscriptions from the period of the Crusades contain similarly declamatory language with pleas to God for victory, which suggests that the Qubba may have served a commemorative purpose, a point to which I will return below.26 Just as interesting as the content, however, is the calligraphic quality of the inscription: a highly sinuous cursive script written on an arabesque background, making it the earliest monumental cursive inscription in western Islam, about twenty years earlier than the cursive inscriptions at the mosques of Tlemcen and al-Qarawiyyin, both of which also date to the reign of ʿAli b. Yusuf.27 In addition to their cursive script, which in itself suggests an eastern source, these inscriptions share several features that link them with early Seljuk and Zangid inscriptions and ultimately with the thuluth script of Ibn al-Bawwab, intimately associated with the city of Baghdad.28 The Qubba in an ʿAbbasid context Returning to the Qubba’s exterior, we need to account for its most unusual decoration, consisting of interlaced circular arches and a

ANDALUSIAN ROOTS AND ʿABBASID HOMAGE

Figure 12.14  Qubbat al-Barudiyyin, detail of upper exterior (photo: Yasser Tabbaa)

seven-pointed radiating star above them (Figure 12.14). Carved masonry domes are known in Cairo, but none is so early, and none contains interlaced arches. A handful of Almoravid domes – i­ ncluding two at the mosque al-Qarawiyyin in Fez and one at the mosque Bayn al-Qahaoui in Sousse – have vertical zigzag decoration, but they do not approach the complexity and elegance of the Qubba. To my knowledge, this feature is not known in any other preserved dome and would be considered a complete oddity were it not for a vivid passage by Ibn Jubayr about similar domes at the Haram of Mecca.29 In this passage, Ibn Jubayr describes four domes (qubab) that must have piqued his interest for their height and excessive decoration, and that he calls qarnasa, which may designate not actual muqarnas but simply minute and extensive decoration. These domes are the Qubba above the sacred spring of Zamzam, the Qubba al-ʿAbbasiyya, the so-called Qubba al-Yahudiyya and the Qubba at Bab Ibrahim.30 The description of the last dome – built by the caliph al-Muqtadir’s governor in Mecca in the early tenth century – is especially noteworthy: Over the portal is a large dome (qubba), remarkable because it is almost as high as the adjacent minaret (ßawmaʿa). Its interior is covered with marvelous plasterwork and qarnasÈ carvings that defy description. The exterior is also made of carved plaster, resembling interlaced drums.31 It seems clear, therefore, that one or more of these Meccan domes closely resemble the Qubba in attenuated form, rich interior stucco

249

250

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

carving and exterior interlaced arches. Furthermore, since these domes were built by ʿAbbasid caliphs, they presumably reflect an architectural type that existed in Baghdad. It follows then that the Qubba, in addition to its obvious Cordoban links, also includes many architectural, ornamental, and epigraphic features that emanate from the central Islamic world, and these might have originated in the ʿAbbasid capital. Once again our attention is directed to that city of vanished glory. What does all this mean? Why would a newly arrived Berber dynasty thousands of kilometres away from the ʿAbbasids have been so determined to use the ʿAbbasids’ artistic language and architectural forms? Why would an early twelfth-century dome in Marrakech so closely resemble domes that probably existed in tenth-century Baghdad and eleventh-century muqarnas domes that most likely originated there? And, finally, why was this monument built, and why did it take on the form of a dome over a fountain? Dynastically, the Almoravids are quite comparable to their near contemporaries the Great Seljuks: both were foreign military dynasties, staunch SunnÈs, and supporters of the spiritual hegemony of the ʿAbbasid caliphate.32 In fact, Almoravid Sunnism had a distinctly ʿAbbasid flavour, for their foremost theologian, Abu ʿImran al-Fasi, was directly influenced by al-Baqillani (d. 1013), the greatest AshʿariShafiʿi theologian of his time and the propagator of the caliph al-Qadir’s SunnÈ politics.33 In the following century, al-Ghazzali himself was initially very influential, exchanging letters with Yusuf b. Tashufin before the local conservative, and apparently ignorant, scholars turned against him and burned his book, I˙yåʾ ʿulËm al-dÈn (Revival of Religious Sciences).34 Finally, both Yusuf and his son ʿAli exchanged letters with respective ʿAbbasid caliphs – as many as seven letters have been recorded – that emphasise the central questions of jihåd against the forces of the Reconquest, of SunnÈ orthodoxy and of allegiance to the ʿAbbasids.35 In return for their homage and gifts, the Almoravids received from the ʿAbbasid caliph robes of honour, decrees of territorial possession and the use of the title Amir al-Muslimin. This title occurs on the Qubba inscription and on the Almoravid coins struck by ʿAli and his father.36 Interestingly, it was preferred by both the Almoravids and the Almohads over the caliphal title Amir al-Muʾminin, out of respect for the office of the caliph and perhaps also to emphasise Islåm over Èmån (faith) in contradistinction to the Ismaʿili Fatimids. Ismaʾilism, with its bipartite division of faith into exoteric and esoteric (Ωåhir and bå†in), normally equated Islåm with exoteric rituals and Èmån with the deeper meanings of the faith.37 This concept was opposed by all SunnÈs, in particular the Malikis of North Africa, who privileged the exoteric dimension of Islam and rejected any esoteric reading.38 Seen within this artistic, courtly and politico-theological discourse, the Qubbat al-Barudiyyin should no longer be viewed as a

ANDALUSIAN ROOTS AND ʿABBASID HOMAGE

poor man’s copy of Cordoban architecture or a pastiche of unresolved influences but must be acknowledged as an original monument that, among other things, reflected Almoravid allegiance to the ʿAbbasids as the symbolic rulers of Islam and the safeguard of Sunnism. More than just a dome above an ablution fountain, the Qubba was intended to pay homage to the ʿAbbasid state and perhaps to evoke the pious acts of the ʿAbbasids at the Haram of Mecca, allusions that perfectly coincide with the Almoravids’ political and religious orientation. None of this is to deny the Qubba’s obvious affiliations with Cordoban architecture. After all, the patron of the Qubba, ʿAli b. Yusuf, was born in Ceuta and raised in Seville, and the architect, according to Terrasse, may have come from Cordoba.39 Indeed, despite its many imported technical and ornamental features, the Qubba still has an overall Andalusian appearance and could not be mistaken for an Iraqi muqarnas dome, for example. This Cordoban filiation could quite likely be attributed to the architect’s training and the immediate sources and models available for his design. Upon this traditional slate, a discriminating patron and an undoubtedly well-informed architect inscribed several imported features and ideas, bringing about a vividly ingenious and stridently original monument. In doing so, they deliberately created an indigenous, local synthesis marking a distinctive Almoravid visual identity. A ceremonial dome? Two final interrelated questions remain: why was a domed fountain chosen to serve as a commemorative monument, and does this idea have precedents in medieval Islamic architecture? In the case of Qubbat al-Barudiyyin, it seems that the choice of a fountain should be linked to ʿAli b. Yusuf’s most important accomplishment in the city of Marrakech, namely, the development and expansion of its waterworks. Using the khettara water-supply system, which resembles that of Iranian qanåts, or subterranean canals, his engineer ʿUbaydullah b. Yunis brought water to most quarters in the city, sending the surplus to the surrounding gardens. The fountain, therefore, may have served as the celebratory focus of this important achievement, a phenomenon known in the Islamic world and elsewhere.40 The second part of the question, why a dome was used for this possibly commemorative or ceremonial purpose, is somewhat more difficult to answer. While most early Islamic freestanding domes were mausoleums, other rare examples of ceremonial domes, including commemorative structures and palace pavilions, did exist. Possibly the earliest of the commemorative type are the above-mentioned ʿAbbasid domes constructed around the perimeter of the Haram at Mecca. Devoid of religious function beyond

251

252

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

defining the borders of the Haram and marking some of its venerated sites, such as the Well of Zamzam, these domes, in particular the one at Bab Ibrahim, declared the ʿAbbasid presence in this sacred precinct.41 It can be argued, however, that the Qubba in Marrakech belongs to the class of palatial domes. It seems likely that the famous dome of the Mouchroutas, built in the twelfth century adjacent to the throne room of the Byzantine imperial palace in Constantinople, as well as a few Norman palatial domes in Sicily dating from the reign of William II (1166–89), corresponds to the typology of the Qubba and ultimately to an earlier ʿAbbasid dome type. According to contemporary descriptions, the Mouchroutas (Arabic makhrˆ, cone) very likely had a conical muqarnas dome.42 Built at a time when the Rum Seljuk prince was living in the Byzantine palace, it may have served a ceremonial function as a reception hall for visiting dignitaries. Although its exotic form and unusual function are perhaps more directly linked to Rum Seljuk ceremonial, the Mouchroutas may ultimately reflect the same ʿAbbasid prototype that produced the Qubbat al-Barudiyyin. The handful of domes and garden pavilions built by the Normans in Sicily also seem to belong to the same type, sharing openness to the exterior, muqarnas vaulting and a water element beneath the dome but not the attenuated profile. Although the three remaining domes – La Zisa (1166), La Cuba (1180) and La Favara – ultimately belong to a palatial ʿAbbasid dome type, their form was used for its broadly imperial and exotic effect rather than as an expression of homage to the ʿAbbasids.43 Was the Qubbat al-Barudiyyin a palatial ceremonial dome? If so, was it intended as a freestanding pavilion within a garden, like those of Norman Sicily, or did it belong to a palace? The architectural context has almost entirely vanished, and the textual references are much too vague to allow us to reconstruct an Almoravid palace. Barrucand has proposed in passing that the Qubba was part of the aforementioned Dar al-Hajar palace complex. But the three excavated courtyards apparently belonging to the Dar al-Hajar are located several hundred metres south-west of the Qubba, which therefore could not have been part of the same complex (Figure 12.2). I would propose instead that the Qubba was built within a garden south of the mosque of ʿAli b. Yusuf, a garden that originally may have extended all the way to the Dar al-Hajar. As such, its primary function would have been to serve as a ceremonial station between the mosque and the other great creation of ʿAli b. Yusuf, his palace complex.44 Within a few decades of the completion of this complex, both mosque and palace were destroyed, and with them must have vanished the garden and its associated structures. Only the solitary Qubba remained, and its survival should most likely be linked to its altered function: from a symbolic victory edifice whose

ANDALUSIAN ROOTS AND ʿABBASID HOMAGE

dome announced the imperial grandeur and sophistication of the Almoravids to an Almohad ablution fountain surrounded by latrines. Could revenge be any sweeter?

Notes 1. Oleg Grabar, ‘The Umayyad Dome of the Rock’, Ars Orientalis 3 (1959): 33–62. 2. John D. Hoag, Islamic Architecture (New York: H. N. Abrams, 1974). 3. Jacques Meunié, Henri Terrasse and Gaston Deverdun, Nouvelles recherches archéologiques à Marrakech (Paris: Arts et Métiers Graphiques, 1957). The Qubba had previously been published in Boris Maslow, ‘La Qoubba BarËdiyyÈn à Marråkuš’, Al-Andalus, Cronica Arqueologica 13 (1948): 180–95. It is also briefly noted in Richard Parker, A Practical Guide to Islamic Monuments in Morocco (Charlottesville, VA: Baraka Press, 1981), 59, and in other guidebooks. Most recently Marianne Barrucand and Achim Bednorz, Moorish Architecture in Andalusia (Cologne: Taschen, 1992), 141, concluded that the Qubba was ‘totally within the tradition of Taifa architecture in Andalusia’. 4. See in particular Jonathan Bloom, ‘The Minbar from the Kutubiyya Mosque’, 3–29, in Jonathan Bloom et al., The Minbar from the Kutubiyya Mosque (New York: Metropolitan Museum of Art, 1998), esp. 23, where the author vehemently asserts the quality and originality of Almoravid art. 5. This connection has also been made by Gülru Necipo©lu, in idem, The Topkapı Scroll, Geometry and Ornament in Islamic Architecture: Topkapi Palace Museum MS H. 1956 (Santa Monica: Getty Center for the History of Art and the Humanities, 1995), 101. 6. Salåmah Mu˙ammad Salmån al-ÓirfÈ, Dawlat al-Muråbi†Èn fÈʿahd ʿAlÈ ibn YËsuf ibn TåshufÈn: Dirasa siyåsiyya wa ˙a∂åriyya (Beirut: Dår al-Nadwa al-JadÈda, 1985), 369–70. 7. For the justification of the destruction of Almoravid mosques by the Almohads see Bloom, Minbar, 35. The problem of mosque orientation has a very long history, both theoretical and practical. In theory, it would certainly have been possible in medieval Islam to build mosques very precisely oriented towards Mecca. In practice, such precision was almost never the case, as most mosques had to be inserted within a pre-existing urban fabric. Although most legal scholars conceded considerable latitude in this matter, it was still possible for dynasts to use ‘incorrect’ orientation as a pretext for destroying mosques. 8. The mosque was as expensive as it was large, reputedly costing 60,000 dinars: see al-ÓirfÈ, Dawlat al-Muråbi†Èn, 374. 9. See Bloom, Minbar. 36. Instead of destroying the minbar, the Almohads were apparently satisfied to have the names of Almoravid patrons removed from its inscriptions. 10. Bloom, Minbar, 35–7 and 49, discusses the circumstances that led to the destruction of this palace and the erection of the Kutubiyya Mosque, referring to the anonymous al-Óulal al-Mawshiyya fÈ dhikr al-akhbår al-Marråkushiyya: Chroniques anonymes des dynasties almoravide et almohade, ed. I. S. Allouche (Rabat, 1936). 11. Barrucand, Moorish Architecture, 141. 12. See n. 3 above.

253

254

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

13. It is of course highly unlikely and completely incongruous that an important and elegant structure like the Qubba would originally have been surrounded by latrines as it is today. 14. Meunié, Nouvelles recherches, 36–41. 15. This is the conclusion reached, for example, by Bloom, Minbar, 27, and by most other scholars. 16. Even mosques without a courtyard, particularly from the Ottoman period, often have freestanding ablution fountains just a few metres outside the façade of the sanctuary. 17. Although many Mamluk and Ottoman mosques were built equipped with ablution fountains, most early mosques had ablution fountains added at a much later date: examples include the Great Mosques of Damascus and Aleppo and that of Ibn Tulun. In no case known to me was the ablution fountain built before the mosque. 18. Barrucand, Moorish Architecture, 88. 19. A seven-pointed star is extremely unusual in Islamic geometric ornament, where even numbers are far more common. See, meanwhile, the interesting but inconclusive comments made by Meunié, Nouvelles recherches, 35–7, where the authors acknowledge the uniqueness of this carved dome and its decoration. 20. Barrucand, Moorish Architecture, 154. See also Christian Ewert, Forschungen zum almohadischen Moschee II: Die Moschee von Tinmal, Madrider Beitrage, Bd. 10 (Mainz: P. v. Zabern, 1984). 21. For example, the muqarnas domes of the shrines of Yahya and ʿAwn al-DÈn in Mosul and the shrine of ʿAbd al-Samad in Natanz. All these domes postdate the Qubba by one or more centuries. Such corner domes also exist in Cordoba and Toledo, but there they are rib vaulted like the main dome. The use of muqarnas in these little domes and elsewhere in the Qubba may in fact be the earliest documented instance of this eastern feature in Moroccan architecture. 22. Gaston Deverdun, Marrakech, des origines à 1912, 2 vols (Rabat: Éditions Techniques Nord-Africaines, 1959–66), states (1: 136) that ‘Islamic art has never surpassed the splendor of this extraordinary cupola’. 23. Most Islamic stucco ornament was in fact painted, usually in light blue and ochre with occasional gold highlights and black outlines. Such painting is known in Samarra, Ghazna, Madinat al-Zahraʾ and the Alhambra. 24. See, for example, Barrucand, Moorish Architecture, 154–5. 25. See n. 3 above. 26. One similar plea for God’s aid in victory comes to mind: the long and impassioned inscriptions on the side of the minbar made by NËr al-DÈn for the Aqsa Mosque in Jerusalem. See, inter alia, Yasser Tabbaa, ‘Monuments with a Message: Propagation of Jihad under Nur al-Din’, in The Meeting of Two Worlds: Cultural Exchange Between East and West during the Period of the Crusades, ed. V. Goss and C. Vézar-Bornstein (Kalamazoo, MI: Medieval Institute Publications, 1986), 223–41. 27. Yasser Tabbaa, The Transformation of Islamic Art during the Sunni Revival (London and Seattle: University of Washington Press, 2001), 66–7. 28. See Yasser Tabbaa, ‘The Transformation of Arabic Writing I: Qurʾanic Calligraphy’, Ars Orientalis 21 (1991): 117–48; ‘The Transformation of Arabic Writing II: The Public Text’, Ars Orientalis 24 (1994): 117–47.

ANDALUSIAN ROOTS AND ʿABBASID HOMAGE

29. This passage was first noted by Jonathan Bloom, who used it in his discussion of the late Fatimid domes and minarets of Upper Egypt. See idem, ‘The Introduction of Muqarnas to Egypt’, Muqarnas 4 (1986): 21–8. 30. Mu˙ammad b. A˙mad Ibn Jubayr, Ri˙lat Ibn Jubayr (Beirut: Dår Íådir, 1964), 66 and 77–8, where Ibn Jubayr further describes the wood and plaster qarnasa in these domes. 31. Ibn Jubayr, Ri˙lat, 83. 32. As noted in Necipo©lu, Topkapı Scroll, 101. 33. Al-ÓirfÈ, Dawlat al-Muråbi†Èn, 168–75. 34. Ibid., 324–8. 35. Ibid., 174–5, and Óusayn Muʾnis, Sabʿ wathåʾiq jadÈda ʿan Dawlat al-Muråbi†Èn (Cairo: Dår al-Maʿårif, 1959), 66–8. 36. According to contemporary chroniclers, the elders of his tribe ­proposed to Yusuf that he adopt the caliphal title Amir al-Muʿminin, but he turned it down because he felt only the ʿAbbasids were its legitimate possessors. Instead, he accepted the title Amir al-Muslimin, which was subsequently used in his coinage and spoken during the Friday khutba. 37. On this question see Irene Bierman, Writing Signs: The Fatimid Public Text (Berkeley: University of California Press, 1998), 67–70, where she presents a particularly interesting interpretation of Fatimid coinage reflecting their concept of imån. 38. N. Cottart, ‘Malikiyya’, in The Encyclopaedia of Islam, 2nd edn (Leiden: Brill, 1960–2004), s.v. 39. Jacques Meunié, Nouvelles recherches: according to Henri Terrasse, 19, ‘It is certain that ʿAli b. Yusuf brought the best Spanish artists for the decoration of his shrines and palaces’. 40. Al-ÓirfÈ, Dawlat al-Muråbi†Èn, 369–70. In fact, the same concept was applied by other patrons in the Islamic world, who often celebrated their hydraulic accomplishments by building fountains and sabÈls. See, in particular, Tabbaa, ‘Monuments with a Message’, 230. 41. The attenuated and decorative character of these domes, according to Bloom, seems to have inspired similar domes in Upper Egypt; the founders of these Egyptian domes wished to emphasise their piety and sophistication by emulating domes they would have seen during the Hajj. See Bloom, ‘Muqarnas’, passim. See also Jonathan Bloom, ‘The Qubbat al-Khadraʾ and the Iconography of Height in Early Islamic Architecture’, in Premodern Islamic Palaces, ed. Gülru Necipoqlu, special issue of Ars Orientalis 23 (1993): 135–42. Of these domes, the one at the mosque of Qus comes the closest to the Qubba in its attenuated form, its exterior and interior ornament, and even its rather indefinite function. 42. On the Mouchroutas see Tabbaa, Transformation, 124, 127; and more specifically Paul Magdolino, ‘Manuel Komnenos and the Great Palace’, Byzantine and Modern Greek Studies 4 (1978): 101–14. 43. Giuseppe Bellafiore, La Zisa di Palermo (Palermo: S. F. Flaccovio, 1978); Yasser Tabbaa, ‘Toward an Interpretation of the Use of Water in Islamic Courtyards and Courtyard Gardens’, Journal of Garden History 7, 3 (Mar. 1990): 202–5. See, most recently, Sibylle Mazot, ‘Fatimid Influences in Sicily and Southern Italy’, in Islam: Art and Architecture, ed. Markus Hattstein and Peter Delius (Cologne: Könemann, 2000), 159–60 and 258.

255

256

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

44. This relationship between the mosque and the Qubba actually recalls the association of the Masjid-i Jami in Isfahan and its north dome, which is generally seen in terms of imperial and ceremonial iconography. See Oleg Grabar, The Great Mosque of Isfahan (New York and London: New York University Press, 1990), 38–41.

PART IV  WRITING

CHAPTER THIRTEEN

The Transformation of Arabic Writing: Part 1, Qurʾa¯nic Calligraphy The conversion of Qurʾånic and monumental calligraphy from angular to cursive is one of the most important but least discussed developments in Islamic art. Occurring first in Qurʾån manuscripts in the tenth century and later in monumental inscriptions, this transformation had a deep and long-lasting impact, shaping the subsequent evolution of Islamic calligraphy for several centuries. It was also a geographically widespread change, and although it began in the central Islamic world – most likely in Baghdad – no Muslim country from India to Spain was left unaffected by it. A development of this magnitude, occurring in the most visible medium of Islamic art, requires an explanation. Furthermore, in view of the fact that in the medieval Islamic world calligraphy fulfilled a central iconographic function (as the transmitter of pietistic messages and political propaganda), this explanation cannot be restricted to the mere formal alteration of the script but must reach to the underlying cultural factors that would have made such a change necessary. The following paper will therefore attempt to address both formal and iconographic aspects of this transformation as they apply to Qurʾån manuscripts. Since the change in monumental epigraphy lagged by about one century behind the Qurʾånic transformation and was contingent upon it, it seems logical to proceed chronologically from Qurʾån manuscripts to monuments. Given the central importance of this problem, why has it attracted so little attention among specialists in Arabic manuscripts and monumental inscriptions? It is certainly not for lack of diligence and creativity: specialists in these areas have, over the last century and a half, examined countless manuscripts and fragments and documented nearly all the important monumental inscriptions in the Islamic world. In the process they have laid the foundations of the two fields of palaeography and epigraphy, which today are among the most developed fields in Islamic studies. Instead it seems that the main reason for this neglect is methodological, emanating from Yasser Tabbaa (1991), ‘The Transformation of Arabic Writing, 1. Qur’anic Calligraphy’, Ars Orientalis 21, 119–47.

260

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

the specialised approaches and rather inflexible agendas prevailing in epigraphy and palaeography. Research in Islamic epigraphy has generally been restricted to the recording and translation of inscriptions on monuments and art objects, and somewhat later to their interpretation.1 Little attention has been given to calligraphic form, whose relevance to the very specialised endeavour of the first epigraphers was not at all perceived. While this is understandable given the enormous scope of epigraphic documentation and the outline format of its early publications, the dismissal of the formal qualities of the script is far more problematic in the recent works of art historians who have used epigraphy as an interpretive tool.2 By simply perpetuating the restrictive methodology of the first epigraphers, they have generally reduced calligraphy to mere information, thus diminishing the artistic meaning and visual impact of inscriptions instead of enriching them.3 The analysis of the formal qualities of scripts has traditionally fallen in the domain of palaeography, although in recent years a number of art historians have also made important contributions towards charting the course of Arabic calligraphy and distinguishing its many varieties. But, with rare exceptions, these writers have been disinclined to consider the reasons behind changes in calligraphic form.4 Instead of searching for underlying cultural causes, most writers on calligraphy have tended to explain the great developments in Arabic and Persian scripts in terms of regional variation, autonomous chronological change, or artisanal improvements determined primarily by the innovations of a few well-known calligraphers and the lesser contributions of minor calligraphers.5 This overly specialised approach is problematic in at least two respects. First, in its emphasis on authenticating the works of the most important calligraphers and its dismissal, or at least negative evaluation, of all ‘questionable’ specimens, it has tended to lose sight of the broad artistic trends of the period and even of the legacy of the calligrapher under consideration. This is especially troublesome in the case of Ibn Muqla, of whose calligraphy no specimens have survived but whose method is known to have influenced one or two generations of calligraphers. Second, traditional palaeography has left unexamined the impact of external factors, such as politics and religion, on the world of the calligrapher – factors that may have directly or indirectly contributed to palaeographic changes.6 Just as the specialised approach of epigraphers limited them to the content of the inscriptions, so the approach of palaeographers has restricted them to problems of dating, provenance and authorship. The question of the transformation of Arabic writing, though closely related to both epigraphy and palaeography, seems to fall in a methodological middle ground between the two disciplines. Being equally concerned with the formal and semiological aspects of this

ˉ NIC CALLIGRAPHY QURʾA

transformation, this piece and its sequel will draw on the findings of both epigraphers and palaeographers while at the same time charting an entirely new course. More than anything, it will attempt to examine the various dimensions of the relationship between the form(s) and meaning(s) of certain new calligraphic styles: how and why these new forms were created; how and why meanings were ascribed to them; and what religious or political requirements these meaningful forms were intended to address.7 We begin, therefore, by examining the formal qualities of the new Qurʾånic calligraphy produced by and under the influence of Ibn Muqla and Ibn al-Bawwåb, basing our investigation on a fairly large number of manuscripts culled from several American, European and Middle Eastern libraries.8 We proceed next to investigate a number of factors that may have contributed to this transformation, including the gradual replacement of vellum by paper; the wide application of geometric principles to Islamic art, including calligraphy; and finally the attempts to create a canonical recension of the Qurʾån in the tenth and eleventh centuries. Methodologically, this piece follows a tradition perhaps first established by Nabia Abbott, whereby evidence derived from the palaeographic analysis of Qurʾånic manuscripts and fragments is juxtaposed against a wide variety of literary sources.9 Although Abbott’s work has enjoyed widespread influence, it should be noted that two recent writers on Qurʾånic palaeography have taken exception to her textual approach on two entirely different grounds. Estelle Whelan has correctly questioned Abbott’s reliance on secretarial manuals for the identification of Qurʾånic scripts, suggesting that this undermines the usefulness of her conclusions for the study of extant manuscripts.10 Yet despite this methodological flaw, Whelan reaffirms in her conclusion the importance of a comprehensive approach to the material, one that takes into account ‘paleographical, codicological, textual, and ornamental’ criteria.11 Déroche, on the other hand, having voiced some objections about the ‘relatively modest’ impact of Abbott’s method on palaeography, nearly dismisses the literary sources and bases his entire investigation on the close examination of large collections of Qurʾån manuscripts.12 While the thoroughness and meticulous care of his approach are indeed admirable, the absence of the cultural backdrop, which in the case of calligraphy is quite elaborate, casts some doubt on his detailed classifications and formalist schemes. At best, such an ahistorical approach may be convenient for purposes of taxonomy and classification, but it falls far too short when one attempts to deal with problems of change and transformation in calligraphic styles – problems that have long been identified by the writers on calligraphic and scribal arts. What seems needed, therefore, is not to silence these sources but to utilise them with a greater sense of purpose and focus than

261

262

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Abbott or others have. Despite their often ambiguous statements, impressionistic ideas and incomplete schemes, these texts can ­ nevertheless provide an adequate framework for posing questions to the available specimens. Most researchers in Islamic art agree that all these sources have their limitations and are often silent about matters that seem to us of the greatest importance.13 The challenge, then, is to establish links among these specialised sources and between them and the works of art in the hope of composing a reasonably coherent picture of a particular cultural or artistic phenomenon. While such a reconstruction may remain incomplete and may even lack the apparent authority of positivist classification, the juxtaposition and interlinking of a variety of texts, including the artistic one, will ultimately enrich the cultural discourse and enhance our experience of its various facets. Often in Islamic art, that is the most one could wish for. Before Ibn Muqla Our attempt to present the transformation of Arabic writing in the fourth and fifth centuries a.h. is made infinitely easier by the substantial palaeographic research on the first three centuries of Islam. The main points of this body of research that are relevant to this paper can be summarised as follows. First, cursive Arabic writing did not originate from an older angular script, but rather the two forms coexisted from the earliest days of Islam (Figure 13.1).14 Second, the early cursive scripts were used exclusively for secular purposes, never for the Qurʾån, which was written in the angular KËfic script (Figure 13.3).15 Third, secular and Qurʾånic scripts were subject to totally different calligraphic rules, those applied to the Qurʾån being far more exacting.16 And finally, most treatises on calligraphy dealt with secular not Qurʾånic scripts since their authors tended to be scribes and officials of the administration.17 With few exceptions, Qurʾånic script from the first two and a half centuries of Islam is extremely uniform, a fact that Arthur Arberry attributed to ‘the tenacious conservatism of many Koranic scribes’.18 There is in fact so little variation in the KËfic script of these Qurʾåns that palaeographers have had to depend on diacritical and orthographic marks and decorations for their dating and classification.19 In contrast, judging from the literary sources and the few preserved specimens, secular scripts exhibited far greater variety.20 By the end of the ninth century Ibn al-Nadim had listed twenty-six styles ranging from large and angular to small and cursive.21 Indeed, such a large number of these ‘secular’ scripts existed by the end of the ninth century that Ibn Wahb al-Kåtib, a contemporary of Ibn al-NadÈm, complained that ‘the scribes were no longer aware of all the different styles of the olden days’.22 Nabia Abbott maintained that many of these scripts represented subtle variations on the major

ˉ NIC CALLIGRAPHY QURʾA

Figure 13.1  Papyrus fragment, Egypt, third/ninth century. Ann Arbor, Kelsey Museum of Archaeology, 67.1.52

scripts,23 but their sheer number and the subsequent need for reform suggest a loss of standard and a general decline in scribal writing. Ibn Muqla and his circle of influence Thus, on the eve of the reforms of Ibn Muqla (886–940), Arabic was being written in a standard Qurʾånic script that only a select few copyists (mu˙arrirËn) had mastered and in an unwieldy variety of secular scripts, many of inferior quality and none following an established standard (Table 1).24 It has been firmly established that, contrary to legend, Ibn Muqla did not create any new scripts25 and certainly was not the inventor of cursive writing, incorrectly referred to today as the naskh script. Known primarily as ßå˙ib al-kha†† almansËb (master of the proportioned script), Ibn Muqla gained fame

263

264

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 13.2  Qurʾån page in KËfic script, ninth century. Ink on vellum. Ann Arbor, University Museum of Art

Figure 13.3  Page of Qurʾån, Iraq/Iran, third/ninth century. Tehran, Iran Bastan Museum, 4251

chiefly for inventing a system of proportional writing based on the principles of geometric design (handa­sat al-˙urËf).26 Some idea of the geometric quality of his script may be derived from the laudatory remarks of the tenth-century writer AbË Óayyån al-Taw˙ÈdÈ: ‘Ibn Muqla is a prophet in the field of writing. It was poured upon his hand, even as it was revealed to the bees to make their honey-cells hexagonal.’27 Ibn Muqlaʾs rules of proportion were not intended for Qurʾånic KËfic but for the large variety of scribal scripts. In other words, Qurʾånic KËfic, which by the tenth century had reached a very high

ˉ NIC CALLIGRAPHY QURʾA

Figure 13.4  Reconstruction of the method of Ibn Muqla: letters alif, låm, sÈn, dål, ßåd. After Soucek, ‘Islamic Calligraphy’, from A˙mad Mu߆afa

standard, was not directly affected by the changes of Ibn Muqla; the reform was at first intended for the more mundane scripts. The result of these reforms, therefore, was not the gradual softening of the angular KËfic script but its supplantation by the redesigned scripts of the chancery. I will return to this important point below. The system of proportion that Ibn Muqla devised was based on measurement by dots. The dot was formed by pressing the nib of the qalam (reed pen) on paper until it opened to its fullest extent, after which it was released evenly and rapidly. This produced a square on edge, or a rhombus. The size of the dot affected only the size of the writing; the relative proportions of letters remained constant for each individual script. Placing dots vertex to vertex, Ibn Muqla then proceeded to straighten the KËfic alif, which had been bent to the right, and adopt it as his standard of measurement (Figure 13.4). His next step was to standardise the individual letters of the various corrupted secular scripts by bringing them into accord with ­geometric

265

266

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

figures. By giving each letter a proportional relation (nisba) to the alif, Ibn Muqla was able to construct a canon of proportions for the entire alphabet.28 This allowed the creation of a number of systematic methods or templates for each of the major scripts, which henceforth could be produced accurately to scale. This standardisation came at a price: a relatively small number of scripts was admitted into the canon of reformed scripts, while others were neglected and gradually slipped into oblivion (Table 1). The canonical scripts, known collectively as al-aqlåm al-sitta or shishqalam, were thuluth, naskh, mu˙aqqaq, riqåʿ, tawqÈʿ and ray˙ån. Of these scripts thuluth was to attain the greatest importance in view of its nearly exclusive use for monumental inscriptions and for sËra headings in the Qurʾån. Naskh, originally a minor and somewhat disdained script, became the preferred style for literary manuscripts and small Qurʾåns, especially during the Ottoman period. Mu˙aqqaq and ray˙ån achieved the peak of their fame in the late thirteenth and fourteenth centuries, when they were used for writing the splendid MamlËk and Mongol Qurʾåns, the former script for large copies and the latter for smaller ones. Riqåʿ was employed for correspondence, while the use of tawqÈʿ was restricted to royal decrees and official letters.29 Although the name of Ibn Muqla is second to none in the history of Islamic calligraphy, no authentic specimen in his hand has been found. Some specimens bearing his name have at different times been suggested as authentic, but all of these have ultimately been dismissed as forgeries.30 Unfortunately, in their zeal to authenticate works by Ibn Muqla, palaeographers may have dismissed certain evidence that, though of no particular use in finding definitive specimens by him, may help us to approximate the appearance of his calligraphy and therefore determine the extent of his influence. This evidence can be summarised under two headings: the method of Ibn Muqla as described in his own treatise Risålah fi’l-kha†† al­-mansËb (Treatise on proportioned writing) and the specimens that have been spuriously attributed to Ibn Muqla. Given Ibn Muqlaʾs far greater fame as the innovator of a method than as a calligrapher, the obsessive search for authentic specimens in his hand may have been misguided. Those efforts could perhaps have been more fruitfully spent in examining Qurʾånic manuscripts that postdate him but may have been influenced by his method. By comparing samples of such manuscripts with the reconstructed alphabet of Ibn Muqla, we may be able to establish the circle of influence of his method.31 Following the instructions given in the Risålah, Nabia Abbott, and more recently A˙mad Mu߆afa,32 have obtained a script characterised by regularity, verticality, semi-angularity, short sublineal curves and the triangular appearance of some of the characters (Figure 13.5). Its alifs are pointed and almost vertical and its knots always open. In all these respects, this reconstructed script

ˉ NIC CALLIGRAPHY QURʾA

Table 1  Development of Arabic calligraphic scripts

1. ‘KËfic’ is used here in its original sense as the mother script of all later Arabic scripts. According to Qalqashandi and others (seen. 15) KËfic has two basic terminal features, flatness (bas†) and concaveness (taqwÈr), such that all calligraphic scripts derived from it contain these properties in specific proportions. As a rule, the larger scripts such as †Ëmår are flat and angular and the smaller scripts such as tawqÈʿ and ghubårare concave and cursive, while the middle scripts such as nißf and thuluth combine the two features in characteristic proportions. 2. The distinction between måʾil and mawzËn has been suggested by Yousuf Thannoun in ‘Kha†† al-thuluth wa maråjiʿ al-fann al­Islåmi’,unpublished conference paper in ‘The Common Themes and Principles of Islamic Art’, Yıldız Palace, Istanbul, 1985. 3. Mu˙aqqaq occurs as a variety of Qurʾånic KËfic in Ibn al-NadÈm. See Abbott, ‘Arabic Paelography’, table 1 and p. 79. I have distinguished it here from two other primary types of Qurʾånic KËfic, mashq and maghribi. The same term is employed again for a fully cursive script first used in the early thirteenth century. Although the scripts have nothing in common, the term mu˙aqqaq (verified) seems to refer to Qurʾånic script of especially high quality. 4. Such scripts as thuluth, naskh, riqåʿ and tawqÈʿ have pre-reform and post-reform existences. Their post-reform appearance is well-known although their original form, with the possible exception of naskh, is likely to remain a mystery. 5. The table conflates the accomplishments of Ibn Muqla and Ibn al-Bawwåb. It should be noted, however, that what I have chosen to call semi-KËfic, or the script influenced by Ibn Muqla, may in fact represent a number of related scripts. 6. To be discussed in Part II.

resembles the so-called semi-KËfic script33 used in many tenth­and eleventh-century Qurʾåns.34 Some specimens of this script are so regular as to be somewhat rigid, which might be seen as the result of strict adherence to the geometric precepts of Ibn Muqla. A second kind of evidence comes from Qurʾånic fragments and album pages that have been spuriously attributed to Ibn Muqla

267

268

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 13.5  Tentative reconstruction of Arabic letter forms according to Ibn Muqla. After Abbott, Rise, fig. 1

Figure 13.6  Qurʾån fragment on paper, Iran, twelfth century. Dublin, The Chester Beatty Library, Ms. Add.

ˉ NIC CALLIGRAPHY QURʾA

Figure 13.7  Juzʾ of Qurʾån on vellum, Iran, Shaʿbån 292/June 905, signed A˙mad ibn Abiʾl-Qåsim al-KhåyqånÈ. Dublin, The Chester Beatty Library, 1417, f. 15b

(Figure 13.6). Although certainly not in his hand and often written two or three centuries after him, these manuscripts nevertheless display striking similarities both to one another and to the group of semi-KËfic Qurʾåns mentioned above.35 Such consistency is significant even in forgeries, for a forger after all has to pay due respect to the original he is copying. In this case, there is no doubt whatsoever that what is being copied is an especially precise form of the semiKËfic script. For these reasons, it seems quite likely that semi-KËfic Qurʾåns – which immediately follow Ibn Muqla – are written in a style that resembles his, at first perhaps by calligraphers who were directly under his guidance. Within the body of Qurʾånic manuscripts, semi­-KËfic Qurʾåns form a fairly distinct group sharing a number of features. In terms of their date, these manuscripts tend to cluster from the late tenth century to about 1100 although the script continued to be used and further evolved in eastern Iran up to the end of the twelfth century and even later.36 The few examples that have been sug-

269

270

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

gested from the first half of the tenth century are quite different from regular semi-KËfic. Of these the earliest is perhaps CBL 1417, which consists of four small juzʾs of a single Qurʾån on vellum, dated to 292/905 and signed by A˙mad ibn Abiʾl-Qåsim al-KhåyqånÈ (Figure 13.7). James suggested that it was ‘one of the earliest Eastern Kufic manuscripts,37 and Déroche38 compared it with BN 382 and TIM 12800, which are dated 325/936–7. But the script in all these manuscripts, though somewhat cursive, is quite irregular, resembling contemporary secular manuscripts such as CBL 3494 (dated 279/892) more than semi-KËfic Qurʾåns.39 In fact the loose form of many of the letters, especially the alif with a pronounced hook to the right, shows no sign of having been influenced by the conventions of Ibn Muqla.40 Eliminating these problematic fragments from our sample, we may then proceed to outline the prominent features of semi-KËfic manuscripts. With consistently clear letter forms and distinct words separated by spaces on either side – instead of the previous practice of dramatically scattering word fragments across the line – the most striking feature of the semi­-KËfic script is its legibility, especially when compared to the preceding ʿAbbasid KËfic script (cf. Figures 13.2 and 13.10). This legibility is further enhanced by a clear and often complete system of orthography and vocalisation. The old system of using large coloured dots for vocalisation and groups of black dots for orthographic marks was abandoned in a two-step fashion:41 first, the orthographic dots were standardised and made smaller (Figure 13.9), and second, a totally new system of vocalisation was introduced, consisting of slashes for fat˙a and kasra, a small waw for dhamma, a small circle for sukËn and other signs for shadda and madda (Figures 13.14 and 13.17).42 This is the system still in use today. The semi-KËfic script is also remarkable for its consistency and regularity, as demonstrated by numerous fragmentary and complete Qurʾåns in European and Middle Eastern libraries.43 It appears generally as a small and rather compact script with very simple letter forms and crisp, angular ligatures. The uprights are nearly vertical although the alif can vary from being slightly bent to the right to being perfectly straight (cf. Figures 13.16 and 13.18). The ‘eyes’ of the letters ßåd, †åʾ, ʿayn, qåf, mÈm and håʾ are always open and have a generally triangular appearance. An exercise in the art of restraint, it has neither the deep sublineal curves of Maghribi KËfic nor the flourishes of later cursive writing. In fact, most examples seem to follow rather closely the method of Ibn Muqla, although divergences begin to occur from the second half of the eleventh century with the widespread use of the so-called Eastern KËfic (Figure 13.8).44 In addition to their distinctive and legible script, semi-KËfic Qurʾåns display at least three other features that distinguish them from their predecessors. The first is that they are almost all written

ˉ NIC CALLIGRAPHY QURʾA

Figure 13.8  Large fragment (first quarter) of Qurʾån, Iran, other parts of same ms. dated 361/972. Dublin, The Chester Beatty Library, 1434, ff, 22b and 23a

on paper instead of vellum (Table 2). Out of our sample of eight manuscripts, two (BL 11, 735 and TKS R-38; Figures 13.12 and 13.15) use vellum and the rest paper. Statistical information on the percentage of the use of vellum in semi-KËfic or Eastern KËfic Qurʾån manuscripts is still unavailable, but the present investigation suggests something under 20 per cent. Although paper began to be used for secular manuscripts sometime in the late ninth century,45 these are the earliest Qurʾåns written on paper and represent the transition from KËfic Qurʾåns on parchment to fully cursive Qurʾåns on paper.46 In fact the earliest known dated paper Qurʾån is written in a very upright and regular semi-KËfic script. This is CBL 1434 (Figures 13.8 and 13.9), a large fragment of a manuscript, portions of which also exist at the Ardabil Shrine and the University Library in Istanbul (A 6758), whose section is dated 361/972 and signed by ʿAlÈ ibn Shådhån al-RåzÈ (of Rayy).47 The second is that these semi-KËfic Qurʾåns abandon the horizontal format of ʿAbbasid KËfic and adopt the vertical format that had been used previously in the so-called ÓijåzÈ Qurʾånic manuscripts of the måʾil script and even more commonly in secular manuscripts.48 The motive for this change has not been determined, but it is unlikely to

271

272

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 13.9  Verse count. Dublin, The Chester Beatty Library, 1434, ff. 1b and 2a Table 2 Semi-KËfic Qurʾåns, tenth–eleventh centuries manuscript

medium

place and date

verse count

fig. nos (13.)

CBL 1434 (juzʾ) TKS HS 22 (1/4 Qurʾån) BL 11, 735 (juzʾ)

paper paper

Iran 972 Iraq/Iran 998

on ff. lb, 2a none in section

8, 9 10, 11

vellum vellum

on f. la; 1st part missing on f. 317b; 1st part missing none

12, 13

TKSR-38 (complete) TKS Y-752 (complete) BL Or. 12884 (complete) MMA 45.10 and 40.164 2a-b (detached folios) TKS R-10 (complete)

Iraq/Iran late 10th century Iraq/Iran late 10th century Iraq/Iran 1004–5 Iraq/Iran mid 11th century Iran/Afghanistan

none detailed count before each åyah

18

Iran 11th century

on f. 235a; 1st part missing

19

paper paper paper paper

14, 15 16, 17

have been due to the switch from vellum to paper since both formats had been used previously with vellum. It is more likely that the use of the vertical format of secular manuscripts went hand in hand with the adoption of scripts that had been primarily used in the chancery and in literary manuscripts. The change in format, therefore, could

ˉ NIC CALLIGRAPHY QURʾA

Figure 13.10  Semi-KËfic Qurʾån on paper, Iran, 388/998. Istanbul, Topkapı Saray Library, HS 22, f. 2b

have been simply an outgrowth of the calligraphic change. But it could also have been intentional, as a way of further differentiating the new Qurʾånic manuscripts from their predecessors. The third feature shared by several semi-KËfic Qurʾåns is that they begin with single- or double­-illuminated folios that refer to the particular recension of the Qurʾån and give a verse count (Figures 13.9, 13.13 and 13.15). As far as we know, this feature did not exist in ʿAbbasid KËfic Qurʾåns49 but begins with the earliest dated semi-KËfic Qurʾån, namely CBL 1434, dated 972. Since most semi-KËfic Qurʾåns

273

274

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 13.11  Colophon, dated 388/998, signed Mu˙ammad ibn ʿAlÈ ibn al-Óusayn al-Íaffår. Istanbul, Topkapı Saray Library, HS 22, f. 257b

Figure 13.12  Semi-KËfic Qurʾån fragment on vellum, late tenth century, Iraq or Iran. London, British Library, 11,735, f. 46b

ˉ NIC CALLIGRAPHY QURʾA

275

exist as fragments with missing frontispieces and colophons, it is impossible to say what percentage of them would have included a verse count. But out of seven complete or nearly complete semi-KËfic Qurʾåns from before the end of the eleventh century, four contain a verse count. Although admittedly a small sample, it does suggest at least that the use of verse count was a prevalent and quite deeply rooted practice in semiKËfic Qurʾåns between c. 950 and c. 1100. The content of the verse count varies slightly from one manuscript to the next, but it generally includes the number of sËras and words in the Qurʾån. CBL 1434 begins with two folios (ff. lb, 2a) that give within circular medallions the number of sËras (114), verses (6,226), words (27,439) and even letters (321,015) Figure 13.13  Verse count. (Figure 13.10). BL 11,735, which is datable to the London, British Library, second half of the tenth century in view of its use of 11,735, f. 4b vellum, contains the second folio (f. la) of a doublefolio verse count (Figure 13.14). It gives exactly the same figures as CBL 1434 for the number of words and letters. TKS R-38, which is also a vellum manuscript, contains in its last folio (f. 317b) the first part of a two-folio verse count that gives the number of sËras in the Qurʾån as 114 and the number of verses as 6,224 (Figure 13.16). In TKS R-10, a paper manuscript, the statistical information also once covered two folios, one of which (f. 235a) is still preserved but in

Figure 13.14  Semi-KËfic Qurʾån on vellum, Iraq or Iran, late tenth century. Istanbul, Topkapı Saray Library, R-38, ff. 12b and 13a

276

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 13.15  Verse count. Istanbul, Topkapı Saray Library, R-38, f. 317b

such a fragmentary condition as to render it practically illegible with any accuracy. Finally, the highly unusual fragments at the Metropolitan Museum of Art, 40.164. 2la-d, contain a rather detailed verse count in very small semi-KËfic script at the beginning of each chapter (Figure 13.19). Semi-KËfic Qurʾåns, therefore, differ from ʿAbbasid Kufic Qurʾåns in terms of their medium, format, script, diacritical marks and verse count. Despite their superficial similarity to the earlier Qurʾåns, they should not be seen simply as a stage in a continuous evolution from angular to cursive but rather as a complete, and as I will argue deliberate, departure from past custom. Once this transformation is accepted on these terms, we must then look for the factors that led up to it. But before doing that we should examine next the second step in the transformation of Qurʾånic calligraphy, the one traditionally associated with Ibn al-Bawwåb.

Figure 13.16  Semi-KËfic Qurʾån on paper, Iraq/Iran, 394/1004–5. Istanbul, Topkapı Saray Library, Y-752

ˉ NIC CALLIGRAPHY QURʾA

Figure 13.17  Colophon, dated 394/1004–5, signed Abu Bakr ʿAbd al­-Malik ibn Zarʿah ibn Mu˙ammad al-RËzbåri. Istanbul, Topkapı Saray Library, Y-752, f. 294a

Ibn al-Bawwa¯b and his circle of influence The second most important stage in the reformation of Qurʾånic calligraphy took place under Ibn al­ -Bawwåb (d. 1022). All the sources agree that Ibn al-­ Bawwåb followed the method of Ibn Muqla but further improved it by making the script clearer, more cursive and more elegant. Ibn Khallikån, the thirteenth­ century historian, said that ‘Ibn al-Bawwåb revised and refined [the method of Ibn Muqla] and vested it with elegance and splendor’.50 Ibn KathÈr, the fourteenth-century Damascene historian, added that ‘[Ibn al-Bawwåbʾs] writing is clearer in form than Ibn Muqla’s and that in the author’s time ‘all people in all climes follow his method except a few’.51

277

278

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 13.18  Fragment from a semi-KËfic Qurʾån on paper, Iran or Afghanistan, mid-eleventh century. New York, Metropolitan Museum of Art, Rogers Fund, 40.164.2b

Figure 13.19  Fragment of a semi-KËfic Qurʾån on paper, Iraq or Iran, eleventh century. Istanbul, Topkapı Saray Library, R-10, f. 31a

ˉ NIC CALLIGRAPHY QURʾA

Figure 13.20  Qurʾån of Ibn al-Bawwåb, Baghdad, 391/1000–1, signed ʿAlÈ ibn Hilål Ibn al-Bawwåb. Dublin, The Chester Beatty Library, 1431, drawing of verse count on ff. 7b and 8a

Only one small Qurʾån has been securely attributed to Ibn alBawwåb, the famous copy at the Chester Beatty Library (1431), dated 391/1000–1 (Figures 13.20–13.22).52 This is the earliest known cursive Qurʾån and undoubtedly one of the earliest made, since Ibn al­Bawwåb was the first to write Qurʾåns in fully cursive scripts (Table 3). Written on brownish paper in a clear and compact naskh, this manuscript is rather easy to belittle: it has neither the majesty and mystery of early KËfic folios nor the grandeur and sumptuous-

279

280

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 13.21  Qurʾån of Ibn al-Bawwåb. Dublin, The Chester Beatty Library, 1431, ff. 241a and 241b

Figure 13.22  Qurʾån of Ibn al-Bawwåb. Dublin, The Chester Beatty Library, 1431, f. 284a. After Rice, Ibn al-Bawwåb

ness of later cursive Qurʾåns. But it is precisely because it looks so familiar and legible to the contemporary reader that this Qurʾånic manuscript is in fact so original. In effect, this copy makes a clear and final break with the majestic but ambiguous script of the first three Islamic centuries, replacing it with a robustly cursive and perfectly legible script that has survived until the present day.53 The two most important cursive scripts are represented in this manuscript, naskh in the text and a variety of thuluth in the

ˉ NIC CALLIGRAPHY QURʾA

Figure 13.23  Qurʾån of Ibn al-Bawwåb. Dublin, The Chester Beatty Library, 1431, f. 284a. After Rice, Ibn al-Bawwåb

opening folios and sËra headings. Naskh was one of the pens in which Ibn al­Bawwåb excelled, and his particular style in writing seems to have been imitated until near the end of the twelfth century (e.g. Figures 13.29, 13.31 and 13.40). James has recently suggested that ‘the naskh of Ibn al-Bawwåb seems to be associated with areas east of Baghdad’,54 an observation that is readily confirmed by several Iraqi and Persian manuscripts of the eleventh and twelfth centuries.55 In fact, the renown of Ibn al­ -Bawwåbʾs Qurʾånic naskh immediately brings to mind the wide appeal of Ibn Muqlaʾs calligraphic method. As with Ibn Muqla, the manuscripts closest in date to Ibn al-Bawwåb (before 1100) adhere the most closely to his hand while those from the succeeding century begin to show some divergence (e.g. Figures 13.35 and 13.36).56 The thuluth used in the statistical pages and the sËra headings of the Qurʾån of Ibn al-Bawwåb is no less remarkable than the naskh used in the text (Figures 13.19 and 13.20). Despite its early date, it shows a number of refinements that remain with Qurʾånic calligraphy for nearly two centuries and that even influence monumental writing. The script is of a type called thuluth-ashʿar, appearing here as a fully

281

282

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

cursive script, thinly outlined in gold. Although somewhat densely written, the script is especially noteworthy for its clarity and legibility, achieved in part by its totally explicit letter forms, with open ‘eyes’ and pointed uprights (tarwÈs of the alif and låm). The overall squatness of the script is relieved by variation in the thickness of its lines and by the very distinctive feature of interconnection: normally unconnected letters and even independent words are connected smoothly to one another through the use of thin, sinuous extensions. This identifying feature occurs a number of times in the introductory folios, as for example on f. 6b, line 5 (Figure 13.21), where the waw of huwa is connected to the sÈn of sabʿËn, and on f. 8a, line 1, where the alif and the båʾ of abÈ are interconnected (Figure 13.22). Some of the manuscripts in the present sample and several later ones slavishly copy this peculiarity, or perhaps embellishment, at first because of the immediate impact Figure 13.24  Small cursive Qurʾån on of Ibn al-Bawwåb but later perhaps as paper, Baghdad?, dated 402/1011, signed Saʿd ibn Mu˙ammad ibn Saʿd al-KarkhÈ. a sign of homage to the great calligraLondon, British Library, Or. 13002, pher (Figure 13.20).57 In fact, it can be f. 120a said that nearly all the manuscripts in Table 3 that employ thuluth (i.e. BL Or. 13002, TIM 431/2, CBL1430, DK 227, DK 144, UMP NEP-27, CBL 1438) copy the style of Ibn al-Bawwåb, down to the feature of interconnection (e.g. Figures 13.23, 13.25 and13.36).58 The influence of Ibn al-Bawwåb extended even farther than that of Ibn Muqla, and calligraphers continued to employ his method for more than two centuries after his death.59 Later calligraphers not only honoured and emulated him, but a few even made forgeries bearing his signature and attempted to sell them as originals. Some of these forgeries were almost contemporary with Ibn al-Bawwåb60 while others postdated him by several centuries.61 Although his impact was mainly felt in the lands east of Baghdad, at least one Qurʾån manuscript from North Africa (TKS A3), datable to the late eleventh century, tries to copy the thuluth of Ibn al-Bawwåb in its sËra titles (Figure 13.28). The calligrapher, who was quite proficient in the Maghribi style in which most of the manuscript is written, betrays a certain naivété when trying to write such a distinctive cursive script.

CBL 1435 (large part of Qurʾån) CBL 1439 (complete Qurʾån)

CBL 1438 (complete Qurʾån)

BN Ar. 6041 (fifth juzʾ of Qurʾån) DK 144 (complete Qurʾån) UMP NEP-27 (complete Qurʾån)

DK 227 (3/4 of Qurʾån)

TIM 449 (incomplete Qurʾån)

CBL 1430 (complete Qurʾån)

TIM 431/2 (complete Qurʾån) BL Add. 7214 (complete Qurʾån)

Baghdad 391/1000–1 Baghdad 402/1011

CBL 1431 (complete Qurʾån) BL Or. 13002 (complete Qurʾån)

script

text: naskh; headings: thuluth text: small naskh; headings: varied Iran 419/1026 text: naskh; headings: thuluth Baghdad 427/1036 text: small naskh; headings: semi-KËfic Iraq or Iran 428/1037 text: small naskh; headings: floriated KËfic Baghdad? Early 11th century text: naskh; headings: Eastern KËfic Iraq or Iran 491/1106 text: large naskh; headings: thuluth Bust (Iran) 505/1111–12 text: naskh verging on thuluth Iraq or Iran 555/1160 text: ray˙ån; headings: thuluth Hamadhan, Iran 559/1164 text: naskh; headings: thuluth and semi-KËfic Iraq or Iran 582/1186 text: mu˙aqqaq and naskh; headings: E. KËfic and thuluth Iran 592/1195 text: naskh; headings: E. KËfic Iraq or Iran 597/1200 text: large naskh; headings: E. KËfic

place and date

manuscript

Table 3  Dated cursive Qurʾåns, 1000–1200

30, 31

28, 29

24, 25 27

20–22 23

fig. nos (13.)

on ff. 2b, 3a on ff. lb, 2a

on ff. lb, 2a

38, 39 40, 41

36, 37

none in this juzʾ 32, 33 only in sËra headings very detailed, on ff. la, lb 34, 35

only in sËra headings

none preserved

only in sËra headings

only in sËra headings in sËra headings and v

on two double folios on ff. 22b, 23a

verse count

284

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 13.25  Qurʾån in large gold naskh on paper, possibly Iran; possibly made for a Íulay˙id prince in Yemen, dated 419/1026, signed Al-Óasan ibn ʿAbdallah. Istanbul, Turkish and Islamic Art Museum, 431/2

Figure 13.26  Heading of sËra 21 (Al-Anbiyåʾ) in thuluth, 419/1026. Istanbul, Turkish and Islamic Art Museum, 431/2

ˉ NIC CALLIGRAPHY QURʾA

Figure 13.27  Maghribi KËfic Qurʾån on vellum, North Africa, eleventh century, sËras 99–104. Note headings in thuluth. Istanbul, Topkapı Saray Library, A3, f. 285a

An even greater illustration of Ibn al-Bawwåb’s prominence among later calligraphers is an early sixteenth­ -century manuscript that attempts to reproduce his various styles.62 Written five hundred years after Ibn al-Bawwåb and dedicated almost in its entirety to his various calligraphic styles, this manuscript must be seen as an homage to the great master and as an attempt to perpetuate his legacy. It is important to note, however, that despite their great renown and immediate influence in the eastern Islamic world, Ibn Muqla and Ibn al-Bawwåb had virtually no impact on Egypt. Out of our two samples of eight semi-KËfic and thirteen early cursive Qurʾån manuscripts, not one was produced in Fatimid Egypt. In fact the vast majority were made in Iraq and Iran, with Baghdad occupying a position of honour. Geography may have played a role: Baghdad, the centre of this calligraphic transformation, was in the period under consideration better connected with Iran than with Egypt. But the absence of any semi-KËfic or cursive Qurʾån manuscripts from Egypt until the beginning of the thirteenth century must have another explanation, to which I will return. In addition to perfecting the cursive proportioned script, the Qurʾån of Ibn al-Bawwåb contains other important developments. The two-folio verse count, which we have seen in several semi-KËfic Qurʾåns, is further expanded here to enumerate the words and letters in each sËra, the total number of words and letters in the Qurʾån, and even the number of dotted and undotted letters (ff. 6b and 7a; Figure 13.21). Such obsessive record keeping seems to stand in sharp contrast to the early KËfic Qurʾåns, none of which included even the most casual verse count.

285

286

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 13.28  Qurʾån in small naskh script on brownish paper, Iraq or Iran, dated Jumada I 427/March 1036, calligrapher: Abuʾl-Qåsim SaʿÈd . . . ibn TilmÈdh al-JawharÈ, illuminator: Abu ManßËr ibn Nåfiʿ ibn ʿAbdallah. London, British Library, Add. 7214, f. 74a

The verse count is followed by another double folio (7b and 8a; Figure 13.22), which mentions the particular recension of this Qurʾån. Placed within hexagonal cartouches, six per page, the inscription reads: [f. 7b:] According to the count of the people of KËfa, which is told after the Commander of the Believers [f. 8a:] ʿAlÈ ibn abÈ Êålib and Muhammad our Prophet, peace be on him Despite being the first documented Qurʾån to refer to any recension, it does not provide quite enough information, since KËfa boasted not one but three canonical Qurʾånic readers, namely ʿÅßim, Óamza

ˉ NIC CALLIGRAPHY QURʾA

Figure 13.29  Qurʾån in small naskh script on brownish paper, Iraq or Iran, dated 21 Rajab 428/10 May 1037. Dublin, The Chester Beatty Library, 1430, f. 1b

Figure 13.30  Qurʾån in small naskh script on brownish paper, Iraq or Iran, 1037. Dublin, The Chester Beatty Library, 1430, ff. 117b and 118a. Note the striking similarity to the script of British Library, Add. 7214 (fig. 28)

287

288

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 13.31  Qurʾån in small naskh script on thick buff paper, Baghdad?, early eleventh century, colophon falsely signed in the name of Ibn al-Bawwåb. Istanbul, Turkish and Islamic Art Museum, 449, f. 286a

and al-KisåʾÈ.63 It is possible that this was a deliberate ambiguity, intended to establish the authority of this Qurʾån by reference to all three of these recensions. Of the thirteen dated eleventh- and twelfth-century Qurʾåns in this sample, the first six (all from the first half of the eleventh century) are quite easily comparable with the Qurʾån of Ibn al-Bawwåb in terms of calligraphy, size and the reliance on accurate verse count (Table 3). The closest to his hand are without doubt BL Add. 7214 (1036), CBL 1430 (1037) and TIM 449, which must be very close in date to these manuscripts but whose colophon has been rewritten to refer to Ibn al-Bawwåb (Figures 13.27–13.31).64 Produced about two decades after the death of the master, possibly by students of his, the naskh hand used in these splendid manuscripts is extremely close to that of Ibn al-Bawwåb. BL Or. 13002, the closest in date to Ibn al-Bawwåb, also uses a similar naskh hand but is on the whole a less accomplished manuscript (Figure 13.25). TIM 431/2 (Figures 13.24 and 13.25), which seems to have been intended for the Íulay˙id ruler of Yemen, clearly belongs to a higher level of patronage than the only extant Ibn al-Bawwåb manuscript. Although the undotted thuluth ashʿar used in some of the sËra headings (Figure 13.27) recalls the

ˉ NIC CALLIGRAPHY QURʾA

Figure 13.32  Qurʾån in small naskh, Baghdad?, early eleventh century. Istanbul, Turkish and Islamic Art Museum, 449

thuluth of Ibn al­Bawwåb, the text itself is written in a totally original mixture midway between naskh and golden thuluth. Indeed, the lavish use of gold, the numerous illuminations and, for the time, the large size of the script seem to describe a royal manuscript that has yet to be properly studied. As would be expected, manuscripts from the twelfth century, which are separated from Ibn al-Bawwåb by five or more generations, display much greater calligraphic freedom. One of the most original is the little-known manuscript BN Ar. 6041, written in Bust (Iran) in 505/1111–12. Déroche has convincingly identified its script as tawqÈʿ, which is a larger and more cursive pen than naskh (Figure 13.32 and 13.33).65 NEP-27 at The University Museum in Philadelphia uses a rather sparse naskh script, which is comparable though not identical to that of Ibn al-Bawwåb (Figure 13.37). CBL 1438, dated 582/1186 and written by ʿAbd al-Ra˙mån al-Kåtib al-MalikÈ ʿZarrin Qalamʾ, has been identified by James as a royal manuscript.66 It displays a bewildering array of scripts, including Eastern KËfic, mu˙aqqaq, naskh, and a thuluth that is remarkably close to the hand of Ibn al-Bawwåb (Figures 13.36 and 13.37). As with the earlier TIM 431/2, its royal status sets it apart from the less

289

290

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 13.33  Fifth volume of Qurʾån in tawqÈʿ, pen on paper, Bust (Iran), dated 505/1111–12, signed ʿUthmån ibn Mu˙ammad. Paris, Bibliothèque Nationale, Ar. 6041, f. 107a

elaborate Qurʾån of Ibn al-Bawwåb. The last two manuscripts in the sample, CBL 1435 and 1439, dated respectively 1195 and 1200, are both written in a clear and sober naskh that resembles the style of Ibn al-Bawwåb in most respects but has taller uprights and far fewer lines per page (Figures 13.38–13.41). Taken as a group, six out of the thirteen manuscripts in the sample contain a verse count (Figures 13.20, 13.23, 13.34, 13.37, 13.39 and 13.41). Restricting the sample to the ten complete Qurʾåns, we still get six manuscripts with full verse count, leaving the possibility of an even higher ratio since folios bearing the verse count are especially vulnerable to destruction and loss.67 In short, then, this group of Ibn al-Bawwåb-related Qurʾåns exhibits some continuity and some development over Ibn Muqlarelated tenth-century Qurʾåns. The new scripts are totally cursive, betraying none of the formulaic regularity of the semi-KËfic group. Orthographic signs, while similar to those of the previous century, are more consistently used; verse counts are expanded and also used in sËra headings; and the particular recension is often clearly stated.

ˉ NIC CALLIGRAPHY QURʾA

Figure 13.34  Colophon of Bust Qurʾån, 1111–12. Paris, Bibliothèque Nationale, Ar. 6041, f 125a

Interpretation Between about 930 and the first decades of the eleventh century, Qurʾånic calligraphy therefore underwent two decisive changes that completely transformed the physical appearance of the Qurʾån, both in sum and in detail. The first change led to the creation of a paper Qurʾån written in a crisp, sometimes rigid script with full diacritical marks, while the second resulted in a variety of fully cursive Qurʾåns, which have remained relatively unchanged until recently. Palaeographic and artistic concerns aside, what really distinguishes these Qurʾåns from the earlier KËfic ones is legibility. Semi-KËfic Qurʾåns are, with the exception of some ornate examples, reasonably legible, while the fully cursive ones can be easily read by any literate person. This raises a number of important questions. What was behind this two-phase but total transformation of Qurʾånic calligraphy? Why was a script as old as Islam itself abandoned in Qurʾånic writing and replaced by the reformed scripts of Ibn Muqla and Ibn al-Bawwåb – scripts that had originated in the secular sphere? In what ways was the old KËfic script considered deficient, and what new associations did the new scripts convey vis-à-vis their predecessor? And lastly,

291

292

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 13.35  Qurʾån in naskh on paper, Hamadhan (Iran), dated 559/1164, signed Ma˙mËd ibn al-Óusayn al-Katib al-KirmånÈ. The University Museum, University of Pennsylvania, NEP–27, f. 1a (neg. #21586). Opening page showing verse count according to different schools

what was the reason for the compulsion to count verses and record them accurately? A technical explanation may stress the change from parchment to paper in Qurʾåns – a change that had begun in other manuscripts as early as the late ninth century. Ernst Kühnel, for example, has suggested in passing that the spread of paper as an inexpensive writing medium in the succeeding centuries led to the development of a less cumbersome and pretentious style.68 Actually, the impact of paper on literacy and book production in the Islamic world, which remains a poorly studied subject, was far more complicated than Kühnel suggests.69 There is little doubt that the availability of Chinese paper from the eighth century and its widespread manufacture in the

ˉ NIC CALLIGRAPHY QURʾA

Figure 13.36  Qurʾån in naskh on paper, Hamadhan, 1164. The University Museum, University of Pennsylvania, NEP–27, f. 2a, al-Fåti˙a (neg. #21590)

Islamic world by the tenth century contributed to the expansion of literacy. As a cheap writing medium became available, the number of scribes (nussåkh) increased and so did the number of scripts, which in turn led to the relaxation of calligraphic standards and the general decline in the quality of writing.70 Some system was urgently needed for secular writing, and it was provided, as explained above, by Ibn Muqla in the form of al-kha†† al-mansËb. But although paper and literacy were significant factors in this transformation, they do not explain why the new script was used for the Qurʾån, thereby ending a four-century­old tradition of angular writing. A technical explanation is simply untenable for a matter of such importance.

293

294

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 13.37  Qurʾån in naskh and mu˙aqqaq with thuluth and Eastern KËfic headings, Iraq or Iran, dated 15 Jumada I, 582/3 August 1186, signed ʿAbd al-Ra˙mån . . . al-Kåtib al-MalikÈ ʿZarrin Qalamʾ. Dublin, The Chester Beatty Library, 1438, f. 125a. Throughout the manuscript one line of Mu˙aqqaq alternates with several lines of naskh; Eastern KËfic is used for Bismillåh and sËra titles; small gold thuluth is employed for emphasis

One is also tempted to see in the proportioned script yet another of the many applications of geometric principles to Islamic art in the late tenth and eleventh centuries. Certainly, the method of Ibn Muqla has more to do with geometry than with calligraphy – a fact that was not lost on contemporary authors. We might also add that there is a clear difference between the visible geometry of the angular KËfic script and the invisible geometry of the proportioned script of Ibn al-Bawwåb, often described as a script without any visible external edges (an lå turå min al-khåriji zawåyåhu).71 It is this assimilated geometry that pervades a variety of artistic forms in the eleventh century, namely geometric strap work and muqarnas. In fact, it can hardly be accidental that these calligraphic and architectural Figure 13.38  Verse count. Dublin, changes occur simultaneously and within The Chester Beatty Library, 1438, ff. 2a and 1b the same geographic regions.72

ˉ NIC CALLIGRAPHY QURʾA

Figure 13.39  Large section of a Qurʾån in naksh, Iraq or Iran, dated Mu˙arram 592/December 1195, signed AbË NaʿÈm ibn Óamza al-BaihaqÈ. Dublin, The Chester Beatty Library, 1435, ff. 117b and 118a

Figure 13.40  Verse count. Dublin, The Chester Beatty Library, 1435, f. 3a

295

296

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 13.41  Qurʾån in naskh on paper, Iraq or Iran, datable by waqf to c. 597/1200, signed by calligrapher Mu˙ammad ibn A˙mad al-Jabali and illuminator ʿAbd al-Ra˙mån ibn Mu˙ammad al­-ÍËfÈ. Dublin, The Chester Beatty Library, 1439, f. 366b

Figure 13.42  Verse count in floriated KËfic contained within circular medallions. Dublin, The Chester Beatty Library, 1439, f. 1 r

ˉ NIC CALLIGRAPHY QURʾA

Yet, despite the key role played by geometry in the formation of the proportioned script, it could not have been the primary cause for this development, and one has to search for the factors that led up to its application in Qurʾånic calligraphy. What the geometry of Ibn Muqla did was to make certain mundane scripts such as naskh sufficiently worthy for writing the Qurʾån. But the question remains as to why this development was demanded in the first place. Why was this new script preferred over the KËfic for writing the Qurʾån? To answer these questions we must look closely at certain contemporary ideas about the content of the Qurʾån. The need to produce a universal recension of the Qurʾån was strongly felt in the early Islamic period; finally, under the third caliph ʿUthmån, the official recension was finished, and all other variants were allegedly destroyed.73 Only one reader, Ibn MasʿËd, refused to destroy his version of the Qurʾån or stop teaching it when the ʿUthmånic recension was made official.74 His codex, which differed from the ʿUthmånic recension in several important respects, was later taken over by the Shiʿi Fatimids.75 As time went on, even the so-called canonical version once more became a source of great confusion because of the ambiguity of the script, ‘to the point that it became impossible to distinguish ʿUthmånic from non-ʿUthmånic ones’.76 By the end of the tenth century, the differences in the texts became more pronounced as a result of the general use of more precise scripts, making it possible for the authorities to enforce a greater measure of uniformity. Under the patronage of Caliph al-Muqtadir, a jurist named A˙mad Ibn Mujåhid produced Qurʾånic codices based on the seven canonical readings belonging to important qurråʾ of the eighth century. His views, set forth in a book called Kitåb al-Sabʿa,77 were adopted by none other than the Ibn Muqla, in his position as vizier of the ʿAbbasid state, and made official in the year 322/934. In fact, Ibn Muqla’s involvement in the creation of a canonical body of Qurʾånic recensions went much further than that. He was certainly involved in the trials of two of the variant readers, Ibn Miksam and Ibn ShanabËdh.78 The persecution of Ibn ShanabËdh, who had persisted in teaching the Qurʾån according to the nonʿUthmånic variant of Ibn MasʿËd, by Ibn Mujåhid and Ibn Muqla is especially noteworthy. He was brought to trial at a court presided over by the vizier Ibn Muqla, where he at first quite confidently and belligerently defended the variant readings that had provoked the charges. But after he had been flogged, he completely disavowed his previous position and signed a document stating that in the future he would adhere to the ʿUthmånic text.79 Ibn al­-NadÈm, who also mentions Ibn ShanabËdh’s flogging at the order of Ibn Muqla, quotes his alleged recantation: ‘I used to read expressions differing from the version of ʿUthmån ibn ʿAffån, which was confirmed by consensus, its recital being agreed upon by the Companions of the Apostle of Allah. Then it became clear to me that this was wrong, so that I am

297

298

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

contrite because of it and from it torn away. Now before Allåh, may his Name be glorified for Him is acquittal, behold the version of ʿUthmån is the correct one, with which it is not proper to differ and other than which there is no way of reading.’80 We have therefore in the person of Ibn Muqla both the calligrapher who created the calligraphic system that led to the conversion of the form of the Qurʾån and the vizier who enforced the caliphal order to establish a body of canonical Qurʾånic readings. There is every indication that the two matters are related: that the creation of al-kha†† al-mansËb and its adoption for copying the Qurʾån were inspired by the canonisation of the text of the Qurʾån. This reforming zeal is further reflected in the emphasis on correct verse count in the Qurʾåns of the tenth century. The new script, with its improved orthography and the correct numeration, would have left no doubt in the mind of Muslims that they were reading one of the new orthodox recensions, certainly not a Qurʾån with aberrant readings. It is very likely that the act of al-Muqtadir and his vizier Ibn Muqla was politically motivated. The caliphate and orthodox Islam were under attack from many different sides by heterodox groups of various Shiʿi persuasions. Closest to Baghdad were the Qaråmi†a, who had, during the reign of al-Muqtadir, occupied Baßra and KËfa and threatened Baghdad several times. Farther away, but nonetheless a threat to the orthodox caliphate, were the Fatimids, who in the first quarter of the tenth century had conquered central North Africa and Sicily and were pushing eastwards. In the face of these overwhelming threats the caliphate could resort to one of the very few weapons it had left, namely its nominal position as the safeguard of the Islamic community and the enforcer of the correct religion. Establishing the canonical recensions of the Qurʾån and creating a new unambiguous script for these standard versions were acts in keeping with that role. The second reform of the Qurʾånic script, the one that led to the Qurʾån of Ibn al-Bawwåb, also may have had its source in contemporary events. In 945 the ʿAbbasid caliphate fell under the control of a foreign dynasty, namely the Buyids, who, to make matters worse, were Shiʿis. By the second half of the tenth century, in fact, most of the Islamic world was controlled by Shiʿi dynasties, with the Fatimids even proclaiming a Shiʿi counter-caliphate centred in Cairo. Only the Ghaznavids in north-eastern Iran actively supported the staunch orthodoxy of the ʿAbbasid caliph. While the office of the caliph was immensely weakened under the Buyids, it was still possible for a strong caliph to reclaim some measure of power and authority, especially in times of disunity among the actual rulers. This is precisely what happened during the long reign of the assertive caliph al-Qådir (991–1031). Taking advantage of the strong popular reaction against the Buyids, al-Qådir gradually introduced measures that would undermine Shiʿi law and

ˉ NIC CALLIGRAPHY QURʾA

keep the Shiʿis out of governmental offices. In 1011, he issued a manifesto condemning the Fatimid doctrine, denigrating their genealogy and declaring the Ismåʿili Fatimids to be among the enemies of Islam.81 In 1017 al-Qådir attempted something not tried since the caliph al-MaʾmËn in the ninth century, namely to promulgate an official theology that condemned all opposing doctrines. The socalled Epistle of al-Qådir (al-Risåla al-Qådiriyya) took aim primarily at the Muʿtazili-Shiʿites but also numbered much more moderate groups among its enemies. It forbade kalåm and all other forms of theological argumentation and negation. It even permitted the imprisonment, exile and execution of all those jurists and rulers who persisted in their unorthodox practice. Finally, it decreed the cursing of heterodox rulers at the pulpits of mosques and encouraged rebellion against them.82 The cornerstone of the arguments in the Epistle of al-Qådir, as explicated by his chief apologist al­BåqillånÈ, concerned the nature of the Qurʾån: It was not created in time, as the Muʿtazilis and others believed, but simply recorded the eternal words of God.83 Moreover, it was uncreated in whatever form it existed: maktËb (written), ma˙fËz (memorised), matluw (recited) or masmËʿ (heard). It had only one meaning, not two – a surface meaning (Ωåhir) and a deeper reading (bå†in) – as the Muʿtazilis and Ismaʿilis maintained. Third, the Qurʾån of Ibn MasʿËd, which was used by the Fatimids,

Figure 13.43  Page from the ‘Blue Qurʾån’, gold on blue parchment, North Africa, tenth century. Private collection, chapter XLII, verses 10–23

299

300

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

constituted an unacceptable alteration of the Qurʾånic text.84 The first two tenets were related, for a Qurʾån that was created in time can be interpreted with greater freedom than one that is, like God, eternal. And a Qurʾån with two levels of meaning must be interpreted by those who know for those who do not. Conversely, an eternal Qurʾån with a clearly manifest truth cannot be further interpreted, and therefore one had to accept the traditional exigesis presented by the jurists in the first two centuries of Islam. Therein lies the political importance of al-Risåla al-Qådiriyya. By closing the door to interpretation after the first two centuries of Islam and by insisting on the incorrectness of the recension of Ibn MasʿËd, it was undermining the religious foundations of the Fatimid and Buyid states and affirming the legitimacy of the ʿAbbasid caliphate. The Qurʾån of Ibn al-Bawwåb represents the creation of a perfectly cursive and easily legible script suitable for expressing the clear and explicit nature of the Word of God. Although ultimately based on the script of Ibn Muqla, the uncompromising clarity of the new script must be seen as a direct reflection of the QådirÈ creedʾs insistence on the single and apparent truth in the Qurʾån. Conversely, the reformed Qurʾån was equally intended to challenge the earlier KËfic Qurʾåns, whose use seems to have continued in Fatimid Egypt until the establishment of the Ayyubid dynasty in the late twelfth century. As noted above, not one semi-KËfic or early cursive Qurʾån seems to have been produced under the Fatimids. In fact very few Fatimid Qurʾåns of any description are known, and to my knowledge, only the so-called ‘Blue Qurʾån’ has been attributed with any degree of authority to the early Fatimid period in North Africa (Figure 13.44).85 Scholars have often commented on the archaising nature of the script, whose unvocalised and undotted letters seem to recall ninth-century Qurʾåns.86 In fact, the ambiguity of the script is further enhanced in this manuscript by the fact that it is written in gold over dark blue. The gold shimmers and seems to flow over the receding blue background, creating an evanescent effect that appears to affirm the Muʿtazili belief in the created and mysterious nature of the Word of God. One can hardly imagine a greater contrast than that between a page from the ‘Blue Qurʾån’ and one from the Qurʾån of Ibn al-Bawwåb. The symbolic implications of the proportioned script have long been lost, but its usefulness remains as a clear and legible script. Yet, at the time of its inception and particularly its adoption throughout the only recently SunnÈ Islamic world, it literally reflected the triumph of a theological view and all its political ramifications. The actual image – not just the content – of the Word became the symbol of the most important principle of the SunnÈ revival, a movement that redefined the course of medieval Islam.

ˉ NIC CALLIGRAPHY QURʾA

Figure 13.44  Qurʾån of Ibn al-Bawwab, dated 392/999–1000. Naskh text and thuluth chapter headings

Acknowledgements For his help in the area of Qurʾånic readings, I would like to express my special gratitude to Professor Wolfhart Heinrichs. I would also like to thank Professor Oleg Grabar for his valuable comments on several drafts of this piece. I am also grateful to Dr Sheila Blair for her many helpful comments and bibliographic notes and to Muhammad Zakariyya for providing me with the ‘insider’s view’ as a practising calligrapher and for his assistance in the question of qiråʾåt. Obviously, I am alone responsible for the conclusions reached in this piece.

Notes 1. The rise of Islamic epigraphy went hand in hand with the creation of two related bodies of historical inscriptions. The first and most important

301

302

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

was the Répertoire chronologique d’épigraphie arabe, ed. E. Combe et al., 16 vols (Cairo: Institut Français dʾArchéologie Orientale, 1932–64). The second, which combined epigraphy with architectural documentation, was Matériaux pour un Corpus Inscriptiorum Arabicarum (hereafter MCIA). It included the following publications: Max van Berchem, MCIA, Première partie: Égypte, Mémoire de la Mission Archéologique Française au Caire 19 (1894, 1903); van Berchem, MCIA, Deuxième partie: Syrie du Sud, Jerusalem, 2 vols, Mémoire. Institut Français d’Archéologie Orientale (MIFAO) 43–4 (1922, 1927); and Ernst Herzfeld, MCIA, Troisième partie: Syrie du Nord, Inscriptions et Monuments d’Alep, 3 vols (MIFAO) 76–8 (1954–6). 2. The first attempts at using epigraphy for the interpretation of objects and monuments were made somewhat tentatively within the format of MCIA by Max van Berchem and later by Ernst Herzfeld. But the method was further developed by Oleg Grabar in his highly important study, ‘The Dome of the Rock in Jerusalem’, Ars Orientalis 3 (1959): 39–62. Even here, however, the austerity and rigidity of Umayyad KËfic and its effective illegibility from the viewer’s standpoint may have further problematised the interpretation. 3. For example, even as recent a work as E. C. Dodd and S. Khairallah, The Image of the Word: A Study of Quranic Verses in Islamic Architecture, 2 vols (Beirut: American University of Beirut, 1981), pays very little attention to calligraphic form. See also my review of this publication in International journal of Middle Eastern Studies 17, no. 2 (May 1985): 263–6. 4. This is true of even the most recent publications, including Anthony Welch, Calligraphy in the Arts of the Muslim World (Austin: University of Texas Press, 1979), 22–35; Priscilla Soucek, ‘Islamic Calligraphy’, in The Arts of the Book in Central Asia, 14th–16th Centuries, ed. Basil Gray et al. (Boulder: Shambhala Publications, 1979), 7–34; and David James, Qurʾåns of the Mamluks (New York: Thames & Hudson, 1988). 5. The trilogy of Ibn Muqla, Ibn al-Bawwåb and YåqËt al­-MustaʿßimÈ is repeatedly invoked by all writers on Islamic calligraphy, including Yasin Safadi, Islamic Calligraphy (London: Thames and Hudson, 1978), 13–19, and all those mentioned in the preceding note. But the cultural and political context in which these calligraphers worked is very rarely explored. Glenn Lowry has raised a similar objection regarding this restrictive view of the development of Islamic calligraphy in his excellent essay ‘Introduction to Islamic Calligraphy’, in Shen Fu, Glenn Lowry and Ann Yonemura, From Concept to Context: Approaches to Asian and Islamic Calligraphy (Washington, DC: Smithsonian Institution, 1986), 104. 6. One of the few exceptions is Nabia Abbott, The Rise of the North Arabic Script and Its Kurʾånic Development, with a Full Description of the Kurʾån Manuscripts in the Oriental Institute (Chicago: University of Chicago Press, 1939), 33–41. Although the book is rich in its references to political and religious factors, these are not considered as possible causes for changes and variations in calligraphy. 7. Although somewhat based on semiotic theory, this exploration of the relationship between form and meaning is ultimately justified by contemporaneous literary theory. It has been shown, for example, that the literary critic ʿAbd al­-Qåhir al-JurjånÈ (d. 1078) had trodden similar

ˉ NIC CALLIGRAPHY QURʾA

grounds as early as the eleventh century. See Al-JurjånÈʾs Theory of Poetic Imagery, ed. Kamal Abu-Deeb (Warminster: Aris & Phillips, 1979), where the author elucidates the complex and rather modern principle of the image or form of meaning (ßËrat al-­maʿna), which he sees as a structural whole made up of inner relations. Calligraphy, which conveys a specific message within a complex artistic form, seems ideal for this kind of investigation. The sequel to this piece, which will deal with public inscriptions, will further explore this problem. 8. These include the Metropolitan Museum of Art in New York (MMA), The University Museum in Philadelphia (UMP), the British Library in London (BL), The Chester Beatty Library in Dublin (CBL), the Bibliothèque Nationale in Paris (BN) and the Türk ve Islam Eserleri Müzesi (TIM). I acknowledge here my gratitude to the librarians and curators of these institutions for giving me access to their important collections. I was able to examine the original manuscripts and obtain the necessary photographs at all these libraries and museums except Dar al-Kutub al-Misriyya in Cairo (DK), where, due to conditions prevailing at the time of my visit in 1986, I had to content myself with the inadequate substitute of microfilms. 9. Abbott, Rise, 34–6. 10. Estelle Whelan, ‘Writing the Word of God: Some Early Qurʾån Manu­ scripts and Their Milieux, Part I’, Ars Orientalis 20 (1990): 135 n. 110. 11. Whelan, ‘Writing the Word’, 125. 12. François Déroche, Les manuscrits du Coran, I: Aux origines de la calligraphie coranique (Paris: Bibliothèque Nationale, 1983), 14. 13. See, for example, the recent essay by Oleg Grabar, ‘Patronage in Islamic Art’, in Islamic Art and Patronage: Treasures from Kuwait, ed. Esin Atil (New York: Rizzoli International Publications, 1990), 27–40. 14. This is easily demonstrated by a large number of early papyrus fragments that have been expertly examined by Abbott, Rise, 34–6, and Adolf Grohmann, From the World of Arabic Papyri (Cairo: Al-Maʿaref Press, 1952), among others. 15. The Arabic sources present two overlapping definitions of KËfic writing. The first, occurring in the earlier sources or describing early developments, refers to the mother script from which all subsequent Arabic scripts, whether angular or cursive, were developed. Qalqashandi, quoting an earlier source, says: ‘The Arabic script, which is now known as KËfic, is the source of all contemporary pens . . . The KËfic script has a number of pens which can be traced to two sources: concaveness and flatness.’ In Íub˙ al-aʿshå fÈ Íinåʿat al-Inshåʾ (Cairo: Turathuna, 1964), 3: 11. The second and much later definition of KËfic refers only to the angular script that dominated early Islamic calligraphy in Qurʾåns and monuments. This is the standard contemporary usage of the term and the one I employ in this piece. 16. Nabia Abbott, ‘Arabic Paleography: The Development of Early Islamic Scripts’, Ars Islamica 8 (1941): 68-69. 17. Abbott, ‘Arabic Paleography’, 76. 18. The Koran Illuminated: A Handlist of the Korans in the Chester Beatty Library (Dublin: Hodges, Figgis & Co., 1967), xvii. 19. Many of these difficulties have been discussed by Adolf Grohmann in ‘The Problem of Dating Early Qurans’, Der Islam 33 (1958): 213–31. This method has been further refined by Déroche in Les manuscrits du Coran, I. Relying on these and other palaeographic features, the author

303

304

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

has attempted further to subdivide the well-known categories of maʾil (slanted script) and early KËfic into smaller and more precise groups or families of manuscripts. But even within the central KËfic groups (i.e. neither maʾil nor Eastern KËfic), there is a remarkable degree of consistency in the letter forms and the overall appearance of the scripts. 20. Abbott, ‘Arabic Paleography’, 76. 21. The Fihrist of al-NadÈm, ed. and tr. Bayard Dodge (New York: Columbia University Press, 1970), 13–15. 22. Cited in Abbott, ‘Arabic Paleography’, 67. 23. Rise, 97. 24. According to Abbott, ‘Arabic Paleography’, 88, ‘confusion and neglect seem to have gained sway until Ibn Mukla came to the rescue of the Arabic scripts’. 25. Nabia Abbott, ‘The Contribution of Ibn Muqla to the North Arabian Script’, American Journal of Semitic Languages 56 (1939): 70–83; and Edward Robertson, ‘Mu˙ammad ibn ʿAbd al-Ra˙mån on Calligraphy’, in Studia Semitica et Orimtalia (Glasgow: MacLehose, Jackson and Co., 1920), 57–83. 26. Robertson, ‘Ibn ʿAbd al-Ra˙mån’, 59–60, who cites Ibn Khallikån as his source. 27. Franz Rosenthal, ‘AbË Óayyån al-Taw˙ÈdÈ on Penmanship’, Ars Islamica 13–14 (1948): 9. Also cited in D. S. Rice, The Unique Ibn al-Bawwåb Manuscript in the Chester Beatty Library (Dublin: Emery Walker, 1955), 6. 28. The method of Ibn Muqla is described in one early treatise and in a number of secondary sources. The anonymous treatise, Risåla fiʾlkitåba al-mansËba in Majallat Maʿhad al­Mak˙tˆåt al-ʿArabiyah 1, ed. M. Bahjat al-Athari (1955), describes Ibn Muqlaʾs method in detail and with diagrams. Of the secondary sources, see: Robertson, ‘Ibn ʿAbd al-Ra˙mån’; Abbott, Rise, 33–8; and Safadi, Islamic Calligraphy, 16–18. 29. See Safadi, Islamic Calligraphy, 52–77 passim, for a discussion of the six scripts. 30. Some of these forgeries were produced within a few decades of Ibn Muqlaʾs death, while others postdate his death by up to two centuries. 31. The only attempt that I know of to do just that is Frances C. Edwards’ unpublished Master’s essay, ‘A Study of Eastern Kufic Calligraphy’ (University of Michigan, 1981), esp. 34–60. This excellent and highly original study deliberately stays away from the legacy of Ibn Muqla and focuses instead on the two known works of ʿAli ibn Shådhån al-Råzi, the earliest dated semi-KËfic (Edwards uses Eastern KËfic instead) Qurʾån and a book on the grammarians of Baßra. See note 47. 32. Abbott, Rise, 35, presents a tentative reconstruction of Ibn Muqlaʾs method and script. A˙mad Mu߆afa has written a thesis on the subject (University of Edinburgh, 1983), but it is unavailable for consultation except for some illustrations that have been published by Soucek, ‘Islamic Calligraphy’. 33. The term ‘semi-KËfic’ is something of an established error since it seems to suggest, incorrectly, a gradual softening of the original KËfic script. Other terms for this script, such as ‘broken KËfic’ or ‘broken cursive’, have been suggested recently by Estelle Whelan, ‘Early Islam: Emerging Patterns: 622–1050’, in Islamic Art and Patronage, 51. I, however, find the adjective ‘broken’ problematic in two respects: the first is that it has a somewhat pejorative tone, which can hardly be an

ˉ NIC CALLIGRAPHY QURʾA

appropriate description of Ibn Muqla’s accomplishment, and the second is that it recalls an entirely different late Persian script, the shikasta, or ‘broken’. The term ‘new Abbasid’, proposed by Déroche, is perhaps the most appropriate since it seems to refer to the reforms of Ibn Muqla, who was almost certainly behind the development of this script or group of scripts. 34. As far as I know, this connection has not before been made, although both Nabia Abbott and especially Eric Schroeder have noted the possibility of the indirect influence of Ibn Muqla on Qurʾånic writing of the tenth and eleventh centuries. See Eric Schroeder, ‘What Was the BadÈʿ Script?’ Ars Islamica 4 (1937): 232–48. Unfortunately, Schroeder’s attempts to determine the legacy of Ibn Muqla are diminished by the small number of Qurʾånic fragments he examined and by his insistence on identifying the so-called ‘BadÈʿ ‘script of Ibn Muqla. In fact, Schroeder’s hypothesis was decisively refuted by M. Minovi, ‘The So-called BadÈʿ Script’, Bulletin of the American Institute of Art and Archaeology 5 (1939): 142–6. Schroeder actually had to retract his views in ‘The So-called BadÈʿ Script: A Mistaken Identity’, Bulletin of the American Institute of Art and Archaeology 5 (1939): 146–7. See also Rice, Ibn al-Bawwåb, 3 n. 1. 35. Muslim scholars in particular have attributed a number of folios to the hand of Ibn Muqla, often on the basis of a marginal notation by a later owner of the manuscript. See, for example, Naji Zayn al-DÈn, Mußawwar al-Kha†† al-ʿArabi (Baghdad: Wizårat al-Iʿlåm, 1968), 45, no. 80; Habib Fazaili, Atlas-i Kha†† (Isfahan, 1971), 176; and Ahmed Mousa, Zur Gachichte der Islamische Buchmalerei in Aeygpten (Cairo: Government Press, 1931), 46, no. 30. More sceptically, Nabia Abbott has reproduced some folios attributed to Ibn Muqla in ‘Arabic Paleography’, 80–1, figs 1–2. Both the British Library and The Chester Beatty Library (Add. Ms.) have Qurʾånic fragments claiming to be in the hand of Ibn Muqla. 36. Perhaps the latest dated manuscript to adhere closely to Ibn Muqlaʾs method is Mashhad Shrine Library, 84, dated 620/1223, a specimen that can easily be mistaken for an eleventh­century Qurʾån. See Martin Lings, The Quranic Art of Calligraphy and Illumination (London: World of Islam Festival Trust, 1976), 19, pl. 21. 37. David James, Qurʾåns and Bindings from the Chester Beatty Library (London: World of Islam Festival Trust, 1980), 26. 38. Les manuscripts du Coran, I, 51. 39. See Whelan, ‘Writing the Word’, 134 n. 96, for a detailed discussion of CBL 1417. Her conclusion that its rather ungainly script ‘differs from “broken KËfic” in significant ways’ is entirely in agreement with mine. 40. In fact, Arberry, Koran Illuminated, 10, commented that the script of CBL 1417 ‘appears to have no near parallel’. 41. There was apparently a great deal of resistance to the use of these marks in the Qurʾån. Målik, one of the early readers of the Qurʾån, was asked: ‘May Qurʾåns be written according to the innovated system of vocalization? He answered: ‘No! Only according to the original script.’ Translated from quotation in Theodore Noldeke, Gahichte des Qorans, part 3, Die Gachichte des Korantexts, ed. G. Bergstrasser and O. Pretzl (Hildesheim: Georg Olms Verlag, 1961), 20. In private correspondence, Muhammad Zakariyya supplied me with four other versions of this anecdote of Ibn Målik, attesting to its great popularity. They all repeat

305

306

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

more or less the same tale, but three of them add the jurist’s more lenient view with regard to muß˙afs written for children. 42. See Safadi, Islamic Calligraphy, 13–14, for a good summary of this complicated development. 43. This consistency notwithstanding, Edwards in ‘Eastern Kufic’ has identified within the works of ʿAli ibn Shådhån al­Råzi two distinct scripts, a ‘monumental’ script used in some titles and chapter headings and a ‘classical’ script used for the text. The ‘monumental’ script, with its bold strokes, high uprights and alifs with a hook to the right, bears a close resemblance to the unelaborated Eastern KËfic script, suggesting perhaps an earlier date for the origin of that script. For examples of the ‘monumental’ script, see Salahuddin al-Munajjid, Al-Kitåb al-ʿArabi al­-Makh†Ë† (Cairo: Maʿhad al-Makh†Ë†åt al-ʿArabiyyah, 1960), pls 19 and 22. 44. Specimens of Eastern KËfic are abundantly illustrated in Safadi, Islamic Calligraphy, 50ff., and elsewhere. Whether or not we accept the existence of Eastern KËfic in the late tenth century (see previous note), it seems quite clear that this script is dependent on semi-KËfic for all its character forms. Their differences have to do primarily with the increased size of the Eastern KËfic script and especially its elongated uprights. 45. Among the earliest must be CBL 3494, GharÈb al-ÓadÈth of Ibn Qutayba, dated 279/892. Illustrated in Munajjid, Al­-Kitåb al-ʿArabi, pl. 15. 46. See Nabia Abbott, ‘A Ninth Century Fragment of the Thousand Nights’, Journal of Near Eastern Studies 8, no. 3 (1949) and n. 67 below. 47. See James, Qurʾåns and Bindings, 27; Fehmi Edhem Karatay, Istanbul Universitesi Kutuphanesi Arapça Yazmalar Katalogu (Istanbul: Istanbul Universitesi, 1953), pl. 5. In 376/986 the same scribe/calligrapher copied a text on Baßran grammarians, Kitåb Akhbår al-Na˙awiyyin al-Baßriyyin, now located at the Library of Shahid Ali, no. 1842, Istanbul. It is illustrated in Munajjid, Al-Kitåb al-ʿArab . i, pl. 22. See also Whelan, ‘Writing the Word’ and Edwards, ‘Eastern Kufic’, 34ff. 48. This intriguing connection between the format of Qurʾåns written in the maʾil script and that of secular manuscripts has not been explored. Although the ÓijåzÈ manuscripts have been generally assumed to be earlier than ʿAbbasid Qurʾåns, their link with lesser manuscripts may suggest a lower level of patronage or provincial origin. This would cast further doubt on the already problematic chronological distinction between these two types of manuscripts. 49. Since very few complete Qurʾån manuscripts, with frontis- and finispieces, exist from before the middle of the tenth century, it remains uncertain whether or not they contained verse counts. Meanwhile, only one KËfic manuscript on vellum (CBL 1404) contains a verse count, but it is a later addition. See James, Qurʾåns and Bindings, 23. In his most recent publication, Qurʾåns of the Mamluks, 24, James suggested that in at least one KËfic Qurʾån on vellum the verse count is given on the opening illuminated folios. This is an undated manuscript in the British Library (Add. 11,735). This manuscript, however, is very clearly semi-KËfic. 50. Wafiyyåt al-Aʿyån, ed. Ihsan Abbas (Beirut: Dar Sader, 1970), 3: 342. 51. Al-bidåya waʾl-nihåya, 5th edn (Beirut: Maktabat al-Maʿarif, 1983), 12: 14. 52. The most important study of this manuscript is Rice, Ibn al­Bawwåb. A facsimile edition of it was also made by Club du Livre Facsimile (Paris, 1972).

ˉ NIC CALLIGRAPHY QURʾA

53. The value of clarity in this new script has been stressed by Lings in Quranic Art of Calligraphy, 53. Lings’ observation that the clarity of the new script corresponds to the clarity of the revelation is generally valid, except that, like much of the book, it tends to stand outside of history. One would like to know why it was specifically in the tenth and eleventh centuries that the old ambiguous scripts were replaced by the new clear ones. 54. James, Qurʾans of the Mamluks, 17. 55. These would include the following manuscripts in the present sample: BL Or. 13002, TIM431/2, BL Add. 7214, CBL 1430, TIM 449, CBL 1435. James, Qurʾans of the Mamluks, 251, n. 3, mentions another related manuscript in the Library of the University of Leiden (inv. no. Cod. 437 Warn.), copied in Ghazna c. 1050. This manuscript, Kitåb Khalq al-NabÈ wa Khulqih by AbË Bakr Mu˙ammad b. ʿAbdallah, was published by S. M. Stern, ‘A Manuscript from the Library of the Ghaznawid Amir ʿAbd al-RashÈd’, in Paintings from Islamic Lands, ed. Ralph Pinder-Wilson (Columbia: University of South Carolina Press, 1969), 7–31. 56. In fact James, Qurʾåns of the Mamluks, 17, has identified two groups of manuscripts from the second half of the twelfth century that begin to show divergence from the esteemed naskh of Ibn al-Bawwåb. The first variant is heralded by a small Syrian manuscript, now at the Keir Collection no. 27, bequeathed by NËr al-DÈn Ma˙mËd ibn Zanki to the Madrasa al-Óanafiyya in Damascus in 1167. Another manuscript from the estate of NËr al-DÈn exists at the Damascus Museum. Although undated and not totally identical with Keir 27, it shares with it many palaeographic similarities, and the two juzʾs may in fact belong to the same muß˙af. The second group has long been identified by Richard Ettinghausen in ‘A Signed and Dated Seljuk Koran’, Bulletin of the American Institute for Persian Art and Archeology 4, no. 2 (December 1935). This manuscript is listed tenth in Table 3 above. 57. Perhaps one of the latest and most spectacular manuscripts to emulate both the naskh and thuluth of Ibn al-Bawwåb and even the overall format of his unique Qurʾån is the YåqËt al­-Mustaʿßimi Qurʾån at the Bibliothèque Nationale (Arabe 6716). Although I have not examined this manuscript at first hand, the two recently illustrated folios in Splendeur et Majesté: Corans de la Bihliotheque Nationale (Paris: Institut du Mon de Arabe and Bibliothèque Nationale, 1987), 63, should suffice to underline the close similarity between the two manuscripts. This intriguing connection between the two greatest Muslim calligraphers has not yet been explored. 58. The thuluth of Ibn al-Bawwåb is challenged around the end of the twelfth century by a variant originating in Afghanistan and eventually spreading to India. The outstanding manuscript TKS EH 42, dated 573/1177 and signed by AbË Bakr A˙mad ibn ʿAbdullah al-GhaznawÈ, which is otherwise written in an excellent Eastern KËfic, contains at least two folios (fig. 7) in this new variant. The distinct features of this script are to be found mainly in its uprights, which tend to be tall, vertical and unpointed – quite possibly influenced by the similarly exaggerated uprights of Eastern KËfic. A very similar style occurs in some Ghaznavid and GhËrid monuments in Afghanistan and northern India. See, in particular, Michael J. Casimir and Bernt Glatzer, ‘Šåh-i Mashad, a Recently Discovered Madrasah of the Ghurid Period in Gargistan (Afghanistan)’, East and West, n.s. 21, nos 1–2 (March–June 1971), figs

307

308

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

14–19. The style, which may have deeper roots in Ghaznavid epigraphy, can also be seen in the earliest GhËrid monuments in India, in particular the Qutb Minår. 59. Rice, Ibn al-Bawwåb, 9–10. 60. E.g. Diwån of Salåma ibn Jandal TKS B-125; Rice, Ibn al­Bawwåb, 17–22. 61. Another copy of the same DÈwån (TIM, 2015), also bearing the signature of Ibn al-Bawwåb, has been dated by Rice to the second half of the fourteenth century; Ibn al-Bawwåb, 22–3. There are other more blatant forgeries of Ibn al-­Bawwåb, one datable to the fourteenth century and the other even later; Rice, Ibn al-Bawwåb, 26–8. 62. Muhammad ibn Hasan al-ÊÈbÈ, A˙san Ma˙åsin Kitåhat al­Kuttåh, ed. Salahuddin al-Munajjid (Beirut: Dar al-Kitåb a­JadÈd, 1967). 63. A. T. Welch, ‘Al-urʾån’, EI2 5: 409. 64. See Rice’s brilliant discussion of this manuscript in Ibn al-Bawwåb, 24–6. 65. François Déroche, Catalogue des manuscrits arabes I, 2: Les manuscrits du Coran (Paris: Bibliothèque Nationale, 1985), 121. 66. According to James, Qurʾåns and Bindings, 35, ‘The words “al-Kåtib al-MalikÈ,” “The Royal Calligrapher”, imply that the calligrapher was the secretary or calligrapher of a Seljuk ruler, possibly Qillij Arslån’. 67. The use of verse counts seems to stop sometime in the thirteenth century, by which time the thorny question of qiråʾat had presumably been settled. According to James, Qurʾåns and Bindings, 25, ‘By the fourteenth century . . . verse counts at the beginning of Qurʾåns in the Eastern Islamic world had almost entirely disappeared, and there are virtually no Bahri Mamluk Qurʾåns with verse counts at the front’. 68. Ernst Kuhnel, Islamische Schriftkunst (Berlin: Heintze & Blanckertz, 1950), 23. 69. See, meanwhile, Johannes Pedersen, The Arabic Book, tr. G. French (Princeton: Princeton University Press, 1984), 59–67; and C. Huart and A. Grohmann, ‘Kåghid’, EI24: 419–20. 70. It is perhaps in this period that the seeds of discord between scribes and Qurʾåns copyists were first sown, the latter perhaps feeling threatened by the unprecedented spread of literacy. On the distinction between scribe (nassåkh or warråq) and calligrapher (kha††åt) see Pedersen, Arabic Book, 43ff. and 83ff. Whelan in ‘Writing the Word’ has further explored this distinction between the two professions, relating it to their widely divergent intellectual backgrounds and religious inclinations. 71. ÊÈbÈ, A˙san, 6. 72. One excellent example of this simultaneity occurs in the Almoravid restoration of the Mosque al-QarawiyyÈn at Fez, 1135–40, where both muqarnas and cursive inscriptions are introduced for the first time. See Henri Terrasse, La Mosquée al-Qaraouiyin à Fes (Paris: Librarie Klincksieck, 1968), pls 51–3. See also my ‘The Muqarnas Dome: Its Origin and Meaning’, Muqarnas 3 (1985): 61–74. 73. See, in particular, Arthur Jeffery (ed.), Materials for the History of the Text of the Qurʾån (Leiden: E. J. Brill, 1937), ix–x and 5–8. Jeffery (p. 8) seems to have been fully aware of the political implications of ʿUthmån’s act, proposing that it was ‘no mere matter of removing dialectal peculiarities in reading, but was a necessary stroke of policy to establish a standard text for the whole empire’.

ˉ NIC CALLIGRAPHY QURʾA

74. Welch, ‘Al-urʾån’, 407. 75. Welch, ‘Al-urʾån’, 407. The so-called ‘Shiʿa readings’ were considered the most objectionable. 76. Welch, ‘Al-urʾån’, 407. 77. Welch, ‘Al-urʾån’, 408–9. 78. Welch, ‘Al-urʾån’, 408–9; Jeffery, Materials, 9–10; and especially Henri Laoust, ‘La pensée et l’action politique d’al­-Måwardi (364/450– 974/1058)’, Revue des Études Islamiques 36 (1968): 64–6. Ibn Muqla’s deep embroilment in the politics and statecraft of the time makes it more likely that he was less a calligrapher and more the innovator of a correct method. 79. Paret, ‘Ibn ShanabËdh’, EI2 3: 935–6. 80. Fihrist 1: 70–1. 81. See Laoust, ‘Måwardi’, 50–4 passim, and George Makdisi, Ibn ʿAqÈl et la résurgence del’islam, traditionaliste au XI’ siècle, V’ siècle del’Hegire (Damascus: lnstitut français de Damas, 1963), 299–305. 82. Makdisi, Ibn ʿAqÈ 301, and Laoust, ‘Måwardi’, 236–7. 83. Fakhr al-DÈn al-BåqillånÈ, Kitåb al-TamhÈd, ed. R. J. McCarthy (Beirut: Institut Français, 1957). 84. Makdisi, Ibn ʿAqÈl, 305, and Welch, ‘Al-urʾån’, 409–10. 85. Specimens of this dispersed Qurʾån exist in numerous collections, and these have been frequently published in a number of exhibitions. See, for example, Welch, Calligraphy. Unfortunately, no one has attempted to reassemble all the available folios and subject them to thorough analysis. See, meanwhile, Jonathan Bloom, ‘Al-Maʾmun’s Blue Koran?’ Revue des Études Islamiques 54 (1986): 59–65, which argues against a Persian origin for this manuscript and affirms its early Fatimid status. 86. Welch, Calligraphy, 48; James, Qurʾåns and Bindings, 27.

309

CHAPTER FOURTEEN

The Transformation of Arabic Writing: Part 2, the Public Text1 In the first part of this work2 I discussed the two-phase phase transformation of Qurʾånic writing from angular to cursive, phases that were associated with the calligraphers Ibn Muqla (d. 940) and Ibn al-Bawwåb (d. 1022). I also presented an evaluation and an interpretation of this highly important change, relating it to contemporary ideas about the nature of the Qurʾån, which was definitively proclaimed as the eternal and uncreated word of God. These ideas were themselves shown to be closely linked with the newly emergent movement of the SunnÈ revival, a movement that sought to reaffirm the legitimacy of the ʿAbbasid caliphate and the traditionalist basis of Islamic thought while opposing and undermining contrary beliefs and political systems, in particular those of the Fatimids. In this paper I will discuss the parallel transformation of monumental inscriptions from angular to cursive – a transformation that postdated the Qurʾånic one by nearly a century but that seems to have been, at least in part, propelled by similar conditions. Extending the discussion from the sphere of Qurʾån manuscripts to that of public inscriptions proved to be far more difficult than I had envisioned. Despite the greater accessibility of the material, in the form of a large number of dated public inscriptions from the eleventh to the thirteenth century, at least two major hurdles pose themselves before this quest. The first is that, whereas substantial textual material exists on scribal and Qurʾånic calligraphy, and of course on the Qurʾån itself, nearly nothing is known about the makers and the making of monumental inscriptions before the Ottoman period. Even when monumental inscriptions do end with the name of an artisan, this signature often refers to the architect or the building supervisor, not to the calligrapher.3 The second is that, although public inscriptions often contain Qurʾånic passages, they are not usually exclusively Qurʾånic. Indeed, a sizeable portion of monumental inscriptions is quite secular in nature, being primarily concerned with titulature, patronage and waqf and somewhat less so with poetry and mystical evocations. Yasser Tabbaa (1994), ‘The Transformation of Arabic Writing, 2. The Public Text’, Ars Orientalis 24, 119–48.

the public text

Is it, then, legitimate to use the findings of the preceding piece, which were exclusively based on Qurʾånic material, to interpret the transformation of Arabic monumental writing generally? This question will concern us in this piece, but we can tentatively say at this point that, whether religious or secular in content, most monumental inscriptions were public and official, thus reflecting some of the ruling dynasty’s concerns, which were always theocratic in nature. In a largely aniconic artistic culture, these public inscriptions were by necessity one of the primary visual means for political expression (often tied up with religious concepts) and one of the few effective ways for a dynasty to distinguish its reign from that of its predecessor. While it is true that most dynasties also resorted to other more symbolic means of political expression, such as gates, minarets or domes (and in fewer cases sculpture), public inscriptions remained throughout much of the medieval Islamic period the chief means for transmitting political and religious messages and for portraying these messages in a dynastically distinctive manner. The dual nature of monumental inscriptions – informative and symbolic, denotive and connotive – has been examined in a number of recent and penetrating studies.4 Despite their differing research objectives, the authors of these studies have attempted to move beyond traditional epigraphic documentation, in which their work is ultimately grounded, into questioning some of the premises associated with the visuality and receptivity of public inscriptions. Richard Ettinghausen, pointing out the great complexity and limited legibility of some inscriptions (written in floriated or plaited KËfic, for example), has suggested that the reading and comprehension of such inscriptions was ‘reserved for a limited number of persons’.5 Even for those select few, Ettinghausen adds, the ‘reading’ of these texts was ‘nearly always based on previous knowledge and not on direct word by word reaction’, while ‘for the vast majority of the congregation and passers-by the inscription remains incomprehensible as a verbal communication in the modern sense’.6 In other respects, the high placement of some of these inscriptions, their complexity and the existence of serious epigraphic mistakes among them lead him to conclude that ‘readability was only a secondary concern’ – a concern that was superseded by ‘the Gestalt of the inscription as a whole and the inclusion of the caliph’s name’.7 While Ettinghausen questions the informative aspect of at least some Arabic inscriptions, Erika Dodd takes this discourse one step further to suggest that the reading of inscriptions was unnecessary or redundant in view of the immanent and transcendent nature of the word of God in Islam: A verse from the Koran does not have to be read for it to have meaning. It exists eternally, of and for and by itself, and it does not exists in the reader, nor does it depend on the reader and it

311

312

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

does not even have to be read to be appreciated. For the ordinary Muslim layman, the simple presence of a Koranic verse was as evocative as an icon for a Christian worshipper, and produced a similar emotional response.8 Although Dodd does not specifically address the duality of meaning in Arabic public inscriptions, it is perhaps implicit in her distinction between ‘reading’, by which she probably means decipherment, and ‘emotional response’, which seems to result from the symbolic or connotive aspects of the text. Thus, in her passionate search for a deep and pervasive meaning in Islamic art, Dodd collapses the duality of public inscriptions into their symbolic, internalist aspects while neglecting their formal specificity and manifold variations. In short, Ettinghausen grapples with the formal complexities of some monumental inscriptions only to suggest that such inscriptions were not intended to be read, whereas Dodd ignores the question of form altogether. More recently, Irene Bierman has dedicated a series of detailed and penetrating studies to Fatimid inscriptions, examining them in terms of their public impact, the ambiguities inherent in the KËfic script used and the content of the inscriptions. Regarding the ‘kufic script as inherently difficult to decipher, for reading any kufic text requires a more supportive context to help distinguish the graph (letter) shapes than reading a text written in one of the other Arabic scripts where all twenty-eight letters are differentiated’, Bierman asks ‘how and why certain scripts were employed in certain contexts to express certain connotations’.9 More specifically, Bierman’s analysis of some of the most distinctive forms of floriated KËfic, such as the låm-alif and the word Allah, leads her to conclude that ‘the unusual knotting of these upright letters seems . . . to resonate with those Ismå’ÈlÈ beliefs that reveal, by means of letter symbolism, an aspect of the esoteric (al-bå†in) meaning of the Qur’an behind the plain (al-Ωåhir) religious message of the written text’.10 In a long and complicated essay on Islamic calligraphy, Oleg Grabar has presented a number of brilliant observations that reflect on his long-term involvement in the field while pointing out new avenues for research. Rejecting the large-scale and indiscriminate application of the term calligraphy, or ‘writing with the intent of being beautiful’, to all manner of Arabic writing, Grabar prefers postponing the use of this term until after the tenth-century reforms of Ibn Muqla and Ibn al-Bawwåb,11 or more generally, until after Arabic writing had developed sophisticated and overtly described systems of replication, aesthetics and criticism. While such a degree of textual self-consciousness is often desirable for validating the emergence of a particular art form, it is by no means necessary, and one might postulate alternate means for the creation of aesthetic standards in calligraphy or other medieval art forms, standards that reflect an

the public text

internalist appreciation of early Qurʾånic scripts, not their externalist criticism. Of perhaps greater interest for this piece are Grabar’s hypotheses concerning the role of public inscriptions in Islamic society, specifically the various dichotomies they establish between literalist reading and formalist appreciation; private understanding and the conditions of public display; and single-gaze or ‘monoptic’ perception and time-consuming decipherment. At the risk of greatly oversimplifying Grabar’s complex discourse, it seems that most of his dichotomies are variations and elaborations of the main one that we have established above, namely, the distinction between the denotive and connotive aspects of official Arabic writing. Indeed, the chief importance of Grabar’s essay resides not so much in his sustained historical analysis of this important problem but in his various attempts to place these distinctions within a sociocultural context, specifically to locate this problem within the matrices of political power and religious knowledge. He therefore draws attention to the distinction between official and populist currents in public inscriptions, contending that the former were ‘fostered by the courts . . . and practiced by highly skilled professionals’ as ‘a means to control and distinguish’, while the latter had much more flexible standards and displayed greater variety.12 In official writing, which is the only kind that concerns us here, Grabar suggests that aesthetic values – such as skill, complexity and ambiguity – were inextricably linked with questions of power and status, such that the ability to build, own or fully appreciate objects or monuments with complex inscriptions becomes one of the criteria for justifying the power of a social or political elite. We shall see below how this important equation of knowledge and power was played out during the transformation of public writing in the eleventh and twelfth centuries. With the possible exception of Bierman, there are two main problems with these otherwise outstanding attempts to reach a symbolic understanding of Arabic public calligraphy: a synchronic, non-historical or, in Grabar’s case, multi-historical perspective that does not address specific factors of historical change; and an overall reluctance to engage the question of complexity in calligraphic styles. For Ettinghausen, complexity is, more than anything, a hindrance to understanding; for Dodd, it is an insignificant feature, since the inscriptions were not intended for reading; for Grabar, it is an important indicator of social and political privilege; while for Bierman, the complexity and ambiguity of the floriated KËfic script stand out as specific embodiments of Fatimid IsmåʾÈlÈ theology. This piece places questions of complexity and historical change at the centre of discussion as it explores the transformation of Arabic public inscriptions in the eleventh and twelfth centuries. In order to provide a context and a point of contrast for this transformation, I will begin by reviewing the problem of the creation of the ­floriated

313

314

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 14.1  Cairo. Mosque al-Hakim, 990 and later, floriated KËfic inscription on south-eastern tower (photo: Yasser Tabbaa)

KËfic script under the Fatimids, suggesting in the process some of the political and theological issues associated with its development. Second, I will trace the subsequent development of cursive scripts from their rather vernacular origins in Iran to their definitive formulation in twelfth-century Syria, pointing out the role of NËr al-DÈn in promoting this process. Third, I will follow the spread of highly standardised cursive scripts in twelfth-century Syria, Iraq, and elsewhere. Fourth, I will examine the differing situation in Fatimid Egypt, where the use of floriated KËfic writing persists for several decades after the perfection of monumental cursive writing elsewhere and where the ultimate introduction of cursive scripts in the last quarter of the twelfth century coincides with the establishment of the Ayyubid dynasty. Floriated Ku¯fic Of all the varieties of monumental KËfic, floriated KËfic is perhaps the most elegant, combining as it does angular characters with curvilinear plant forms. Indeed, in its fully developed form, exemplified by the inscriptions of the Fatimid mosques al-Azhar and alÓåkim or the late eleventh-century minaret at the Great Mosque of Aleppo, floriated KËfic may be considered the peak of achievement in early Arabic epigraphy. The beauty and inherent complexity of the script have attracted the attention of numerous scholars, both European and Arab, who have gone a long way towards analysing its characters and ornamental forms and proposing theories for its origin and development.13 Although Flury argued at one point for an

the public text

Anatolian origin of this script,14 most palaeographers today concur that ʿAbbasid KËfic underwent subtle and inconsistent changes between c. 830 and 960, which led to the creation of the so-called foliated KËfic, before being definitively transformed into floriated KËfic in the second half of the tenth century.15 The main point of difference among these authors centres on whether floriated KËfic developed gradually out of foliated ʿAbbasid KËfic over a period of a century or whether it was suddenly created in the early Fatimid period. Grohmann, arguing for a gradual multi-centred development,16 concluded that three neighboring countries have participated decisively in the evolution of floriated Kufic: Palestine, with the first traceable connection of a floral element with a letter; Egypt, where in the middle of the third century of the Hijra the decoration of letters with palmettes had already reached a high perfection and where the decoration of the aspices had been invented possibly in connection with, or imitation of, Coptic forerunners; and the Hijaz, where the tombstone of 250 h [fig. 1] shows genuine floriated Kufic definitely established.17 There are several problems, however, with this view of gradual development from foliated to floriated KËfic. The first is that Egyptian inscriptions that postdate the early attempts at floriation but pre-date the Fatimid takeover of Cairo in 969 have a more traditional form, suggesting that ‘this initial phase of floriated Kufic was not able to impose itself fully, not even in Egypt’.18 The second is that nearly all the early foliated inscriptions come from funerary stelae, not from official inscriptions, of which the few that remain tend to maintain a sober and austere KËfic style. This is evident from the complete absence of foliation in the ʿAbbasid and ÊËlËnid inscriptions at the miqyås (Nilometer) – dated 247/861 (reign of al-Mutawakkil) and 259/873 (reign of Ibn ÊËlËn) (Figure 14.3).19 Apparently, then, early developments towards the floriated script were restricted to funerary stelae and other nonofficial inscriptions and had no impact on the official inscriptions of the period, whose style in any case seems to have been derived from Iraq. The third and most serious challenge to Grohmann’s scheme of gradual development is that the inscriptions of the early Fatimid period, namely those at al-Azhar (361/972) and al-Óåkim mosques (before 403/1013), differ completely from the ʿAbbasid KËfic inscriptions of the preceding century and even from the earlier foliated inscriptions. Whereas only a small proportion of the characters of pre-Fatimid foliated KËfic sprout ornamental leaves, nearly every character in the inscriptions at al-Azhar and more emphatically at al-Óåkim is embellished with leaves that completely transform the letter form and the overall appearance of the script (Figure 14.4).

315

316

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 14.2  Óijåz (Arabia). Gravestone, 250/864. Museum of Islamic Art, Cairo. From Grohmann, Arabische Palåographie, 2: fig. 32

Figure 14.3  Cairo. Nilometer: ÊËlËnid inscription (Qurʾån, 2:256), 247/861. From Dodd and Khairallah, Image of the Word, 1: fig. 22 (photo: Yasser Tabbaa)

the public text

Jumʾah was the first definitively to reject the likelihood of a continuous development of floriated KËfic style and to embrace the opposing view of a sudden transformation. After having painstakingly analysed the earliest inscriptions at al-Azhar mosque, located around the hood of the mi˙råb and elsewhere in the sanctuary, he concluded that ‘these inscriptions cannot be said, whether in terms of their writing style or decoration, to be a natural development of third century Egyptian writing’. He added further that this ‘style of writing . . . differs in its totality from the developed writing styles of Egypt in the third century, a matter that makes us wonder whether the Fatimids may have brought with them upon their departure from North Africa a special style of writing which had developed greatly and rapidly during half a century’.20 This interesting hypothesis is, however, very difficult to prove since no immediately pre-Fatimid official inscriptions remain in Egypt and since no early Fatimid inscriptions have survived from al-Mahdiyya or other Fatimid cities in North Africa. But this does not change the fact that the Fatimid inscriptions at al-Azhar represent a totally original style in floriated KËfic and that they are the earliest official inscriptions to utilise this ornamented script. The inscriptions at al-Óåkim mosque, executed over a long period extending from 370/972 to 403/1013, demonstrate the prevalence of the floriated KËfic in official Fatimid inscriptions and the adaptability of the script to a variety of media, including stone, stucco and wood. The stucco inscriptions at the springing of the mi˙råb dome and the stone friezes that encircle different levels of the minarets exhibit the basic aesthetic feature of the script: ‘a quite particular connection of writing and floral tendril growing out of the letters and forming with them an organic unit, serving at the same time to fill the space ideally’21 (Figures 14.4 and 14.5). Ambiguities between text and ornament, foreground and background are thereby created, and these ambiguities are enhanced by the fact that the characters of the script are themselves internally transformed by means of ‘curvatures, counter-curvatures, knots, and indentations’.22 A splendid example of this kind of virtuosity can be seen in the cenotaph of Få†ima at the Båb ÍaghÈr cemetery in Damascus, dated 439/1037, in which Sourdel-Thomine noted ten different types of the låm-alif character (Figure 14.7).23 Following its development under the Fatimids, floriated KËfic spread outside of Egypt in the eleventh century, at first to regions directly controlled by the Fatimids, especially Palestine and southern Syria, or subject to their propaganda (da’wah), such as western Iran, and subsequently to other parts of the Islamic world. Outstanding specimens of floriated KËfic, generally dating c. 1050–c. 1150, survive in southern Anatolia (e.g. the Great Mosque of Diyarbakr [1085, 1126 and 1156]);24 Aleppo (e.g. the minaret of the Great Mosque [1090] (Figure 14.8) and Qastal al-Shuʾaybiyya [1150–1]);25 Damascus (e.g.

317

318

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 14.4  Cairo. Mosque al-Azhar: alphabet of inscription in the maqsËra, 361/972. From Grohmann, Arabische Palåographie, 2: fig. 248

cenotaphs of Få†ima [439/1047] and Sukaina [early twelfth century] at the Båb ÍaghÈr cemetery);26 Palestine (e.g. the minbar at Ascalon); Spain and North Africa (e.g. the Great Mosque of al-Qarawiyyin at Fez [1135]);27 and Iran (e.g. Masjid-i Óaydariyya at QazvÈn [twelfth century]).28 There is little possibility that these were autonomous developments, and one would have to agree with Grohmann that ‘it is certainly from here [Egypt] that its development has advanced to Mesopotamia on one side and to North Africa on the other’29 (Table 1).

the public text

Table 1.  The Development of Monumental Scripts.

Figure 14.5  Cairo. Mosque al-Óåkim: inscriptions in the maqsËra, late tenth century. From Flury, Hakim und Ashar, pl. IV, 1–4

It seems clear, therefore, that despite sporadic earlier developments of the KËfic script, floriated KËfic was effectively created under the Fatimids, who were also the first to use it for official inscriptions. What were the motives for the creation of this script, and what did

319

320

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 14.6  Cairo. Mosque al-Óåkim: inscription on casing of north-west minaret, 403/1013 (photo: Yasser Tabbaa)

Figure 14.7  Damascus. Cenotaph of Få†ima: inscription on northern face, 439/1037. Author’s drawing after Moaz and Ory, Båb al-ÍaghÈr, pl. IVb (photo: Yasser Tabbaa)

Figure 14.8  Aleppo. Minaret of the Great Mosque: uppermost inscription, 483/1090. From Herzfeld, Alep, 2: pl. LIII (photo: Yasser Tabbaa)

the new privileged script mean within the context of early Fatimid propaganda? In a theocratic state embroiled from the start in political and sectarian controversy, it would seem likely that the creation of a new public form of expression was intended to reaffirm the dynasty’s claims to legitimacy while distinguishing it from earlier dynasties. A more specific religious meaning has been proposed by Bierman, who suggested that ‘the unusual knotting of the upright letters seems . . . to

the public text

resonate with those Isma’ÈlÈ beliefs that reveal, by means of letter symbolism, an aspect of the esoteric (al-bå†in) meaning of the Qur’ån behind the plain (al-Ωåhir) religious message of the written text’.30 Indeed, this reading is consistent with one of the fundamental tenets of the IsmåʾÈlÈ doctrine, namely the distinction between the exterior or exoteric and the inward or esoteric aspects of religion. ‘The Ωåhir consists in the apparent, generally accepted meaning of the revealed scriptures and in the religious law laid down in them. It changes with each prophet. The bå†in consists in the truths (˙aqå’iq) concealed in the scriptures and laws which are unchangeable and are made apparent from them by the ta’wÈl, interpretation, which is often of cabalistic nature relying on the mystical significance of letters and numbers.’31 It is this duality of meaning and the valorisation of bå†in over Ωåhir that was to be challenged by the transformed scripts of the eleventh and twelfth centuries. Precursors to the transformation (1030–1150) In view of the dominance of the monumental floriated KËfic script during the eleventh and first half of the twelfth century, it is not surprising that the demise of this luxurious script and its ultimate supplantation by cursive writing have attracted some attention. What is surprising is that from the start this transformation has been associated with the SunnÈ challenge to Fatimid IsmåʾÈlÈ authority, or the so-called SunnÈ revival or reaction. The broad outlines of this process were laid out a century ago by van Berchem: I have demonstrated that around the middle of the sixth century a.h. the square script, called Kufic, hitherto universally used in inscriptions, was replaced by the cursive style, commonly called naskhi. This phenomenon seems to be connected with the Sunni reaction which, leaving Iran in the fifth century, gradually invades Baghdad, Mesopotamia, northern Syria, Palestine, and Egypt, being conveyed by the Seljuqs, the Zangids, NËr al-DÈn, and Saladin. The relationship of the two phenomena . . . is readily explained by bearing in mind that the Sunni reaction, which in Syria coincided with the Mongol invasion and the Crusades, was accompanied by religious, political, military, and administrative changes. This revolution naturally extended to architecture and to the arts and crafts that depend on it.32 Van Berchem’s far-reaching proposals were taken up by none other than Ernst Herzfeld, who in a series of epigraphic and architectural studies succeeded in elaborating his mentor’s highly suggestive thesis and in attributing the bulk of the transformation to NËr al-DÈn.33 Curiously, the matter has been nearly totally forgotten since then, as most art historians and epigraphers shifted their focus from the

321

322

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

documentation and formal analysis of inscriptions to their iconography.34 But van Berchem’s central thesis is too important to ignore and yet too problematic and incomplete to accept uncritically. In the following section I shall therefore point out some of these problems and fill in some gaps in the earliest development Figure 14.9  Khuråsån (Iran). of official cursive inscriptions. Left: Samanid dirham minted at Balkh, In a slightly earlier publication van 292/905. Note cursive signature of Berchem had noted that cursive scripts A˙mad ibn Mu˙ammad. American appear sporadically on the coinage of Numismatic Society 1927.179.69. the late Samanids and the Ghaznavids, Right: Samanid dinar minted at Nisabur, 340/951. Note cursive signature making them the earliest instances of of NË˙ ibn Naßr. American Numismatic cursive inscriptions in a public context.35 Society 1963.173.2 Actually, this contention is only partly correct since the cursive inscription in all Samanid and early Ghaznavid coinage is restricted to the name of the reigning prince (e.g. Naßr ibn NË˙), whereas the rest of the inscription is in KËfic script (Figure 14.9). What we have here,

Figure 14.10  Ghazna (Afghanistan). Cursive inscription on cenotaph of Ma˙mËd ibn SebËktekin, 420/1030 (author's drawing after photograph in Flury, ‘Ghazna’, fig. 9)

the public text

t­herefore, is not the earliest instance of the transformation of the official script but simply the use of cursive script for the ‘signature’ of the ruling sovereign. The earliest official cursive inscriptions are, however, from the eastern Islamic world, where they seem to begin sometime near the end of the reign of the great Ghaznavid sultan Ma˙mËd (998–1030).36 The well-known cenotaph of MahmËd of Ghazna, dated 421/1030, consists of a large rectangular platform surmounted by a triangular grave cover, all made of marble of possible Indian origin.37 It contains six inscriptional bands: one in the middle of each of the four rectangular sides and one in each of the sloping pediments. Only one of these inscriptions, located on the northern side of the upper grave cover, is written in a cursive (or at least non-KËfic) script; the other five are written in an Eastern KËfic script, with typically tall uprights and restrained floriation (Figure 14.10). The cursive inscription, which consists of six lines within a trilobed arch, reads as follows: Has died, may God’s mercy be upon him, and may He illuminate his chamber38 and brighten his face, the evening of Thursday, seven [days] remaining of the month of RabÈ’ the Latter in the year one and twenty and four hundred, may he be forgiven.39 Although more cursive than angular, this extremely peculiar script has no known parallels in monumental inscriptions. It seems to stand midway between Ghaznavid Eastern KËfic and the early thuluth script, combining the elongated uprights and some of the character forms of the former with the cursiveness of the latter. The script has other tentative or ‘transitional’ features, including elongated U-shaped fillers (commonly seen in floriated KËfic inscriptions),40 inconsistent use of orthographic marks and variation in the size of characters, such that the words on the third and fourth lines are larger than those of the first and last lines. In view of its transitional character, further underlined by its location within an entirely KËfic context, this was quite likely one of the very earliest official cursive inscriptions.41 Although no other inscriptions seem to imitate the calligraphic style of Ma˙mËd’s cenotaph, a number of later Ghaznavid and GhËrid inscriptions employ perfectly cursive scripts, which is subsequently seen in the central and western Islamic world, and an attenuated, monumental script, which is most commonly seen in the GhËrid monuments of Afghanistan and India.42 The later script, which might be related to the script on Ma˙mËd’s cenotaph, will not concern us here. Among the earliest specimens of the compact style are two fragmentary inscriptions, the first bearing the name

323

324

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

of the Ghaznavid ruler Abuʾl-MuΩaffar IbråhÈm (1059–99) and the second containing the words YamÈn al-Dawla, which was the laqab (attribute) of the late Ghaznavid sovereign Bahråm Shåh (1118–52) (Figure 14.11, a–b). This script, which is almost always written on a bed of arabesque, is characterised by its legibility, squatness, high degree of cursiveness and near absence of diacritical marks. In all these respects, this script closely resembles the thuluth of Ibn al-Bawwåb as seen in the verse counts and chapter headings of his unique m ­ anuscript – a script that, as I have demonstrated previously, spread in the eleventh and twelfth centuries among the calligraphers of the eastern Islamic world.43 Indeed, this early monumental cursive script emulates an even more specific feature of the great master’s style, namely Figure 14.11  Ghazna ­interconnection  – the tendency to connect normally (Afganistan). a. Fragment of independent characters by a thin, sinuous line. inscription with name of IbråhÈm, 1059–99. This particular hallmark of the master’s script was b. Fragment of inscription rather slavishly copied by many of his students with name of YamÈn and followers.44 Its use in the earliest cursive offial-Dawla, late eleventh– cial inscriptions as well as in later inscriptions in early twelfth century Syria and North Africa suggests close affinities (author’s drawings after between Qurʾånic and monumental writing and photographs in Flury, points once again to the pivotal importance of ‘Ghazna’) the reforms of Ibn Muqla and Ibn al-Bawwåb and their patrons, the ʿAbbasid caliphs.45 I will return to this important connection later. It is somewhat surprising that the earliest official cursive inscriptions are not from Baghdad, where one would expect them given the BaghdadÈ origin of Ibn Muqla and Ibn al-Bawwåb, but from one of the easternmost regions of the Islamic world. In the absence of any supporting evidence, it is difficult to say whether cursive official inscriptions were used by the ʿAbbasid caliphs in the first half of the eleventh century. But in view of the large-scale destruction of most early and medieval Islamic monuments in Baghdad, it is possible that such inscriptions once existed and may have provided a model for the Ghaznavid development. It is also possible that the early cursive Ghaznavid inscriptions were directly based on the manuscript hand of Ibn al-Bawwåb, which became instantly popular in the eastern regions of the Islamic world. Whatever the conduit may have been, the borrowing by the Ghaznavids of an official ʿAbbasid form of expression fits well with their cultural affiliation with the caliphate. Like the ʿAbbasids, the Ghaznavids were staunch SunnÈs at a time when it might have been more advantageous to accept some form of Shiʿism. They were also

the public text

Figure 14.12  Isfahan. Masjid-i JåmÈ: fragment of inscription on north face of south dome, 478/1086–8. From Grabar, Isfahan, fig. 24

loyal supporters of the ʿAbbasids and bitter opponents of their archenemy the Fatimids, who under the caliphate of al-Óåkim (996– 1021) were ever more active in their IsmåʾÈlÈ propaganda. Ma˙mËd of Ghazna wasted no opportunity in courting the favour of his exact contemporary, the caliph al-Qådir (991–1031). For his immediate recognition of the caliphate of al-Qådir (whose succession was vexed by another pretender), Ma˙mËd was awarded a manshËr (charter) for Khurasan, a khil’a (robe of honour), and his first caliphal titles YamÈn al-Dawla (the right arm of the state) and AmÈn al-Milla (the defender of the community [i.e. the orthodox]). Other titles, such as NiΩåm al-DÈn and Naßir al-Óaqq, were awarded him in 403/1012–13 when he executed the Fatimid propagandist Taharti in Bust.46 By the early decades of the twelfth century, cursive monumental inscriptions had become fairly commonplace, both as architectural friezes and in epitaphs, but they did not supplant KËfic inscriptions until much later. Indeed, quite commonly cursive and different varieties of the KËfic script were used in the same monument – an exercise of virtuosity common in a large number of Seljuq monuments.47 The Seljuq script seems to develop straight out of the compact thuluth style first seen at Ghazna in the second half of the eleventh century, as a comparison of one of the earliest monumental cursive Seljuq inscriptions (on the exterior of the mi˙råb dome of the Great Mosque of Isfahan, dated 1086–7) with the slightly earlier Ghaznavid fragments will demonstrate (cf. figs 10 and 11). Both scripts are highly cursive, especially for monumental inscriptions, betraying in this respect their likely origin in paper calligraphy. Their character forms are not sufficiently distinct: specifically, the alifs are not pointed, the knots or ‘eyes’ are not always open (e.g. the mÈm in the Seljuq inscription), and some of the characters (e.g. the råʾ and the nËn) seem to flow imperceptibly from the preceding character. These ‘deficiencies’ are corrected in the inscriptions of the twelfth century, which begin to differ from the earlier style in their pointed uprights, open ‘eyes’, uniform and fairly distinct characters, and in the separation between the inscription and the arabesque ornament beneath it. These features are quite amply illustrated by a related group of gravestones from western Iran or the JazÈra that

325

326

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 14.13  Western Iran. Marble gravestone, 549/1154. Freer Gallery of Art, Washington, DC, 48.16

show remarkable virtuosity in their different varieties of cursive and KËfic scripts.48 One of the finest of these is a splendidly carved marble gravestone, dated 549/1154, with three types of KËfic scripts and an equal number of cursive scripts, all of especially high quality (Figure 14.13). It is worth noting, however, that despite the overall development of the cursive scripts, especially the shahåda in the upper rectangular panel, none of them is dotted or vocalised.49 Commenting on these calligraphic changes, van Berchem has noted that in the east the development of cursive writing, and especially its supplantation of the KËfic script, is very gradual indeed: KËfic inscriptions continue in historical epigraphy until the end of the twelfth century, become increasingly rare during the thirteenth century, and practically disappear by the end of that century. This led him to conclude that ‘in the east the change was a purely practical and autonomous process: cursive writing from daily life slowly and without plan or design supplanted a monumental script that no one could read’.50 While it is true that the transformation was gradual and at times sporadic and that the mixed use of KËfic and thuluth in the same building does pose some problems of interpretation, it may

the public text

be precipitous to conclude that the change was ‘purely practical and autonomous’ and ‘without plan or design’. It is perhaps more prudent to suggest that in certain instances, for example under the early Ghaznavids and in the first Seljuq monuments, the use of a cursive script for monumental inscriptions was indeed purposeful and was motivated by external forces whose nature is just beginning to be understood. But it remains problematic that the new cursive writing took so long to establish itself and that for more than a century it was used simultaneously with a totally different calligraphic style, the floriated KËfic. This matter requires further investigation and cannot be settled within the scope of this essay. Nu¯r al-Dı¯n, 1146–74 Several scholars have commented on the decidedly different situation in Syria, specifically under the reign of NËr al-DÈn Ma˙mËd ibn ZangÈ (1146–74). Van Berchem was the first to note that the change from angular to cursive scripts in Syria was as sudden as it was quick, having been put into effect within just a few years by the orders of NËr al-DÈn as an ‘intentional act for the achievement of a vast plan, part of a reform’.51 Herzfeld stated the matter even more emphatically by placing this transformation ‘at a point almost exactly defined by the year 548 [1153]’, when NËr al-DÈn abandoned the form and content of earlier Seljuq protocols and embraced the changes produced by ‘the deep movement of the Sunnite reaction’.52 Most recently, these observations have been reiterated by SourdelThomine, who concluded that ‘NËr al-DÈn ordered the adoption of the cursive script in official inscriptions, to the detriment of the angular script, which without disappearing completely, was reduced to repetitions of ancient types’.53 Despite the plausibility, even overall veracity, of the conclusions drawn by these eminent scholars, the chronological sequence of inscriptions in Syria from the late eleventh to the middle of the twelfth century presents two important problems. The first is that one early cursive inscription does exist in Syria: the third inscriptional frieze on the minaret of the Great Mosque of Aleppo, dated 483/1090 (Figure 14.14). Curiously, while the four other inscriptional bands on the minaret are written in floriated KËfic of the highest possible quality, the cursive inscription is comparatively mediocre, perhaps displaying the mason’s lack of experience in the new style. Like other contemporary Seljuq inscriptions, it is written on a bed of arabesque and contains no dots or vowel marks. Even more curiously, it is a Shiʿi inscription giving the names of the Twelve Imams preceded by a taßliya. Two possible (but perhaps insufficient) explanations can be offered for this apparent discrepancy. The first is that this minaret was erected during the period of Seljuq control of Aleppo and represents a local attempt to imitate the Iranian Seljuq

327

328

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

practice of using both styles of calligraphy in the same monument. The second is that the minaret was begun by Ibn al-Khashshåb, member of a Shiʿi patrician family, who continued as supervisor (mutawalli) of its construction after the Seljuq takeover of the city.54 Could the inscription, then, be seen as an act of rapprochement between the Seljuqs and the Shiʿi majority in Aleppo?55 Or was the inscription Ibn al-Khashshåb’s idea, a way of making a legible Shiʿi statement with a cursive inscription? The last possibility would also seem to explain later instances of the Twelver Shiʿi use of the cursive script, seen above in the Freer gravestone of 1154 (Figure 14.13). The second problem is that NËr al-DÈn did not use the new cursive style from the beginning of his reign. In fact, his earliest known inscription at the mashhad al-Dikka in Aleppo, dated 1146, is written in a rather simple KËfic style that closely resembles his father’s (ZangÈ) inscription of 1128 on the same building.56 The poor quality of the inscription, its derivative style and titulature, and the fact that it commemorated a building act on a Shiʿi monument are all symptomatic of the precarious start of NËr al-DÈn’s career.57 His very next dated inscription (Shawwål 543/February 1149) at the portal of the Madrasa al-Óallåwiyya, however, is written in an excellent thuluth script that closely resembles some of the better specimens of late Seljuq cursive writing on brick or stucco (Figure 14.15). It is a pleasing and legible style characterised by compactness, pointed uprights and generally open knots, and full use of diacritical and orthographic marks. The character forms are uniform in appearance and begin to display characteristic tapering in the thickness of the line (easily visible in the låm-alif), a feature already seen in the earliest Ghaznavid inscriptions and even earlier in the Qurʾånic calligraphy of Ibn al-Bawwåb and his successors.58 The cramped space forced the mason-calligrapher to overlap some of the letters; except for that problem, the inscription is very easy to read. This inscription, in effect, initiates the total transformation of monumental calligraphy for Syria and ultimately also for Egypt. With two exceptions, to which I will return, all the succeeding inscriptions from the period of NËr al-DÈn and his Ayyubid successors are written in the cursive thuluth script. We are led to inquire, therefore, what exactly took place in the early career of NËr al-DÈn that led him to embark on this fundamental transformation. Although later sources, written under the patronage of NËr al-DÈn and Ayyubid sovereigns, are deliberately vague about NËr al-DÈn’s early years, a close reading of one of the very few preserved Shiʿi histories of the period, Ibn abÈ Êayyiʾ, suggests that, like his father, he was initially far more tolerant of Shiʿism and quite ambivalent in the pursuit of SunnÈ orthodoxy.59 His personal and public transformation is a complex process, discussed elsewhere by myself and others. It suffices to say here that two major factors contributed to this momentous change in direction: early and somewhat unexpected successes against the

the public text

Crusaders, and improved links with the ʿAbbasid caliphate. Between 1146 and 1149 NËr al-DÈn was able to recapture the north Syrian city of Edessa, to aid in defeating the Second Crusade, and to deal a major defeat to prince Raymond of Antioch (who perished in battle). According to Gibb, ‘in the eyes of all Islam, he had become the champion of the faith and he now consciously set himself to fulfill the duties of this role’.60 The ʿAbbasid caliph wasted no time in recognising the victories of NËr al-DÈn by bestowing on him various honorific titles, the most important of which was al-mujåhid (the fighter for the faith). This title appears for the first time on the Madrasa al-Óallåwiyya and becomes subsequently one of his most common epithets. But the caliphate had other concerns than the Crusades, namely the restoration of SunnÈ orthodoxy all over the Islamic world, particularly in Egypt, where the IsmåʾÈlÈ Fatimids had long posed a political threat and theological challenge to the ʿAbbasids. The chief apologist for the ʿAbbasid cause at the time was the powerful theologian and vizier Ibn Hubayra, whose call for the unification of SunnÈ Islam and for the destruction of the Fatimids seems to have struck an immediate chord with NËr al-DÈn. The two are known to have corresponded about these matters, and it was at the vizier’s urging that NËr al-DÈn finally attempted in 1163 to wrest Egypt from the hands of the Fatimids and retake it in the name of the caliphate.61 Thus, early triumphs against the Crusades, the machinations of Ibn Hubayra and the ʿAbbasid caliphs, and undoubtedly a personal proclivity toward orthodoxy and asceticism, all motivated NËr al-DÈn’s pursuit of SunnÈsm, making him the primary force behind the SunnÈ revival. Beginning as a subsidiary theme to the more pressing problem of the counter-Crusade, the revival of the Sunnah soon became the central motive of NËr al-DÈn’s policy, and it is therefore legitimate to view all his major acts through this traditionalist reaction. The calligraphic transformation was one of the most visible signs of this broad movement, which had lain dormant in Syria during the turbulent decades of the first half of the twelfth century but was now being propagated by the ʿAbbasid caliphs and NËr al-DÈn. At its most

Figure 14.14  Aleppo. Minaret of the Great Mosque: inscription on the third zone, 483/1090 (photo: Yasser Tabbaa)

329

330

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 14.15  Aleppo. Madrasa al-Óallåwiyya, 543/1149: inscriptions on the portal (photos: Yasser Tabbaa)

basic, the use of cursive writing for public inscriptions declared, by virtue of its total difference from earlier public inscriptions, the end of the Fatimid period and the beginning of a new era. More specifically, the use of a script with demonstrable links to the ʿAbbasid caliphate was intended to reinforce the legitimacy of NËr al-DÈn’s rule in Syria and in all other territory conquered in the name of the caliph. Finally, by virtue of its legibility and unambiguousness, the new public writing shattered the cherished duality of meaning implicit in Fatimid inscriptions. Before further exploring the implications of these issues, I would like to investigate this calligraphic transformation in NËrid Syria and in other parts of the Islamic world. The next dated inscription by NËr al-DÈn, at the Qas†al al-Shuʾaybiyya in Aleppo (545/1150), immediately presents a problem, since it is written in a highly elaborate script that recalls the late eleventh-century floriated KËfic on the minaret of the Great Mosque. I have elsewhere investigated this important monument, proposing that it was rebuilt by NËr al-DÈn as a commemorative structure intended to celebrate his triumphs against the Crusaders while evoking the earlier victories of ʿUmar ibn al-Kha††åb, who had conquered Aleppo in 16/637.62 The use of an archaising script seems to comport well with the deliberately archaising appearance of the architecture and the commemorative nature of the monument.63

the public text

Figure 14.16  Damascus. Bimåristån al-Nuri, 549/1154, foundation inscription (photo: Yasser Tabbaa)

With the exception of this inscription and a decree in Damascus dated 551/1156, whose floriated KËfic may have had to do with the bureaucratic nature of the inscription, all the other inscriptions of NËr al-DÈn are written in monumental thuluth.64 Two other specimens from his period and one from the closely related era of his young son IsmåʾÈl will suffice to establish the overall character of the mature thuluth script before the Ayyubids. The first inscription comes from a wooden minbar commissioned by NËr al-DÈn in 559/1164 for his mosque at Óama (central Syria).65 This inscription, which simply states the shahåda, is written within a cartouche in a large and clear thuluth (perhaps originally highlighted with paint) on a bed of arabesque scrolls (Figure 14.17). Fully cursive and entirely legible, the script is also characterised by pointed and highly tapered uprights (e.g. alif and låm-alif), open knots and interconnection, seen here in the way that the rå’ and sÈn of the word rasËl are joined. The inscriptions of NËr al-DÈn and his son IsmåʾÈl at the maqåm of IbråhÈm in the citadel of Aleppo, dated 563/1168 and 575/1180 respectively, are among the best executed thuluth inscriptional plaques of the twelfth century. The script can best be described as a fleshier version of the first cursive NËrid inscription of 1148, a dense and rather short script whose squatness is relieved by the tapering of the beginning and end of its letters and by the judicious use of interconnection. Comparing the inscription of NËr al-DÈn with that of his son, we note more overlapped letters and greater reliance on interconnection, but without an undue loss of legibility.

331

332

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 14.17  Óama (Syria). Mosque of NËr al-DÈn: inscription on back of minbar (drawing: Yasser Tabbaa)

Figure 14.18  Aleppo. Maqåm IbråhÈm in the citadel: inscription of NËr al-DÈn, 563/1168 (drawing: Yasser Tabbaa)

Figure 14.19  Aleppo. Maqåm IbråhÈm in the citadel: inscription of IsmåʾÈl, 575/1180 (drawing: Yasser Tabbaa)

the public text

Both scripts are of course fully vocalised and equipped with all the required orthographic marks, making them the most easily legible and unambiguous inscriptions of their time. The canonisation of the Thuluth of Ibn al-Bawwa¯b: 1170–1260 By the time of the death of NËr al-DÈn in 1174, the monumental cursive script that he had helped introduce into Syria had become standard for all public inscriptions, not just in Syria but also in Upper Mesopotamia, Anatolia, North Africa and Spain (see Table 1). Although it is unlikely that every dynast in all these regions was following the example of NËr al-DÈn, some of them may have been, while others may have received their cultural clues from the ʿAbbasid caliphate itself. We will, therefore, examine the situation in each of these regions, beginning with Syria under the Ayyubids.

Figure 14.20  Aleppo. Citadel, Lion’s Gate: inscription of al- Ûåhir GhåzÈ, 606/1210. From Herzfeld, Alep, pl. 38

Figure 14.21  Aleppo. Mosque in the citadel: inscription of al-Ûåhir GhåzÈ, 610/1213 (drawing: Yasser Tabbaa)

333

334

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Syria In Aleppo, inscriptions produced under al-ÛåhÈr GhåzÈ (1195–1216), the greatest architectural patron in the medieval history of the city, build on the high calligraphic tradition established by NËr al-DÈn.66 The two best preserved and most accomplished inscriptions of GhåzÈ are located in the citadel: a foundation inscription, dated 606/1210, at the end of the entrance block on the tympanum of the Lion’s Gate and another foundation inscription, dated 610/1213, above the entrance to the mosque. The first and aesthetically superior inscription proved too difficult to photograph and draw, and Herzfeld’s otherwise excellent drawing of the gate deprives the inscription of all its calligraphic flair (Figure 14.20).67 In truth, this is a magnificent inscription, a masterpiece that balances monumentality with fluidity and legibility with embellishment. Thinner and somewhat more attenuated than its Zangid predecessors, it still manages to maintain the tapered appearance of the uprights and the interconnection of some of the letters, as in the way the zåy and the yåʾ of al-GhåzÈ are linked. The second inscription is nearly identical to the first, possibly even made by the same calligrapher, but differs from it primarily in being more cramped – a condition that forced the calligrapher to overlap some of the words and to rely a little too much on interconnection (Figure 14.21). Altogether, the calligraphic script created under al-Ûåhir GhåzÈ may be seen as the first truly monumental cursive style in stone. It is therefore not surprising that it continues with minor changes to the very end of the Ayyubid period. A series of long inscriptional friezes from the Madrasa al-Firdaws, dated 633/1235–6, attests to the continuity and subtle development of the early Ayyubid script, which becomes less tapered, a little more attenuated and minimally interconnected (Figure 14.22). Interestingly, inscriptions in the rest of Syria and in Palestine, which never achieve the superior quality of the Aleppo inscriptions, undergo a similar process of development between the twelfth and the first half of the thirteenth century. In Jerusalem inscriptions dated 575/1180, 587/1191, and 589/1193 closely resemble the heavy thuluth style of the Zangid period. An inscription dated 604/1208, on the other hand, is closer in its thinness and attenuation to the inscriptions of al-ÛåhÈr GhåzÈ and his successors.68

Figure 14.22  Aleppo. Madrasa al-Firdaws, 633/1235–6: inscription in courtyard (photo: Yasser Tabbaa)

the public text

Figure 14.23  Mosul. Mosque al-NËrÈ: mi˙råb, 543/1148. Originally in the now destroyed Umayyad mosque in Mosul (photo: Yasser Tabbaa)

Upper Mesopotamia No early cursive monumental inscriptions have been preserved in Baghdad, which unfortunately precludes an examination of the impact of Ibn Muqla and Ibn al-Bawwåb on their native city. Indeed, the earliest preserved monumental inscriptions do not occur until  the period of the Caliph al-Nåßir (1180–1225).69 The situation is a little more encouraging in Mosul, where, outside of a handful of early twelfth-century tombstones inscribed in a crude cursive style, the earliest monumental cursive inscription is the one surrounding the inner frame of the mi˙råb of the mosque al-NËrÈ, dated 543/1148 (Figure 14.23).70 The inscription is written on a bed of arabesque and seems to stand midway in terms of development between early Iranian inscriptions and the Zangid inscriptions of Aleppo. Indeed, the entire composition of this flat mi˙råb with friezes of floriated KËfic inscriptions framing an inner cursive inscription is clearly modelled after a Seljuq Iranian prototype. Interestingly, the mi˙råb is signed by a certain Mu߆afa al-BaghdådÈ, whose nisba suggests that he originally came from Baghdad. This is one of the very few references to the existence of mason-calligraphers in the ʿAbbasid capital. As in Syria, monumental cursive writing seems also to have been introduced en masse into Mosul under NËr al-DÈn, who, though never its actual ruler, exercised considerable control over it during the latter part of his reign.71 The mosque that he founded there between 1170 and 1172 contains numerous inscriptions on the capitals of its massive piers (Figure 14.24). Although bearing a general resemblance to the NËrid inscriptions in Aleppo, they still retain two features of early Seljuq cursive inscriptions, namely the absence of diacriticals

335

336

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 14.24  Mosul. Mosque al-NËrÈ: inscriptions on capitals, 1170–2 (drawings: Yasser Tabbaa)

and the presence of an arabesque background. Other inscriptions from this mosque, possibly dating from the first NËrid phase, consist of long friezes in white marble inlaid with black marble.72 These are somewhat closer to contemporary Aleppine inscriptions in their character form, their use of diacriticals and their minimal background ornamentation. Other than these twelfth-century inscriptions, the only pre-Mongol monumental inscriptions in Mosul are those decorating the various shrines erected during the reign of Badr al-DÈn Luʾluʾ (1222–59). Two of these shrines, the mashhads of Imåm Ya˙ya Abuʾl-Qåsim (637/1239–40) and Imåm ʿAwn al-DÈn (646/1248), preserve a number of excellent cursive inscriptions on marble, which are comparable in quality to the best inscriptions in Aleppo.73 The portal to the mosque of the later shrine displays to advantage the great variety of cursive scripts used in Mosul in the few decades preceding the Mongol invasion (Figure 14.25). The uppermost frieze, serving the function of a cornice, is in monumental thuluth (or thuluth jaliyy), a large and slow-moving script with minimal overlapping of words and practically no interconnection.74 Another large script, rendered in white marble on bluish alabaster, presents the name and titles of Badr al-DÈn across the lintel. It is a highly attenuated script that brings to mind the late Ayyubid inscriptions of Aleppo. The third, and for us most interesting, calligraphic style in this portal is represented by a long frieze that enframes the portal on three sides. The inscription, which gives the fairly common Ayat al-KursÈ (Verse of the Throne), is written in a splendid compact thuluth style that recalls, even surpasses, the twelfth-century inscripFigure 14.25  Mosul. tions in Aleppo. With no less than twelve instances of Mashhad of Imåm ʿAwn interconnection, this inscription might appear to have al-DÈn, 646/1248: portal sacrificed legibility for the sake of cursiveness and to the masjid (photo: Yasser Tabbaa) artistic nuance (Figure 14.26). Remarkably, however,

the public text

Figure 14.26  Mosul. Mashhad of Imåm ʿAwn al DÈn, 646/1248: portal to the masjid, detail of inscription (photo: Yasser Tabbaa)

Figure 14.27  Mosul. Shrine of Imam ʿAwn al-Din, 1242, detail of thuluth inscription on portal (photo: Yasser Tabbaa)

Figure 14.28  Tlemcen (Algeria). Great Mosque: inscription on the mi˙råb dome, 530/1136. From Marçais, Occident, fig. 150

it remains perfectly legible throughout – a feature that must be attributed to the excellence of its calligraphy and the unobtrusive nature of its interconnections, whose extreme thinness further enhances the tapering of the letter forms. It is astonishing that a

337

338

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

calligraphic nuance first introduced in the late tenth century, and whose ultimate origin may have been quite accidental, should still find considerable resonance in monumental writing two and a half centuries later. North Africa In North Africa, including Sicily, the floriated KËfic script remained dominant until about the middle of the twelfth century, when it was challenged, both in coinage and on monuments, by the cursive script.75 Appearing first in some Tunisian tombstones from the late eleventh and early twelfth centuries, the style is initially seen within a monumental context late in the period of the Almoravids (1056–1147).76 The earliest cursive monumental inscription in North Africa is a long frieze that encircles the base of the famous ribbed filigree dome of the Great Mosque of Tlemcen (Algeria), dated 530/1136 (Figure 14.28).77 The script closely resembles other ‘Seljuq’ thuluth inscriptions that we have so far seen in Ghazna, Isfahan, Aleppo and Mosul. Undotted, unvocalised, and displaying some of the characteristic tapering of letters, this historical inscription is also written on a bed of arabesque. On the basis of the published photographs and drawings, it is impossible to determine whether it contained any interconnected letters. A more extensive cycle of early cursive inscriptions is found farther west, at the mosque of al-QarawiyyÈn at Fez. The inscriptions belong to the major Almoravid building phase, in which the entire axial nave of the mosque was rebuilt in 531/1137 with a series of muqarnas vaults.78 The cursive inscriptions coexist with a plethora

Figure 14.29  Fez. Mosque of al-QarawiyyÈn: inscriptions in the cells of the mi˙råb dome, 531/1137. From Terrasse, al-Qaraouiyin, pl. 53

the public text

Figure 14.30  Fez. Mosque of al-QarawiyyÈn: foundation inscriptions above the mi˙råb, 531/1137. Author’s drawing after Terrasse, al-Qaraouiyin, pl. 51

of highly complex floriated KËfic inscriptions, resembling in this respect a group of Qurʾånic manuscripts written in the MaghribÈ script but utilising the thuluth script of Ibn al-Bawwåb for their chapter headings.79 Seemingly restricted to a medallion above the mi˙råb and to short friezes framing the cells of the two muqarnas vaults nearest to the mi˙råb, these inscriptions are almost identical to the Tlemcen inscription, except that some of them are written on an unadorned background (Figure 14.29). The foundation inscription above the mi˙råb consists of four short lines of slightly more developed thuluth, which attempts, though not very successfully, to utilise to the fullest the feature of interconnection. Nearly every word is connected with the following one, and that, combined with links that are nearly the same thickness as the script itself, results in a dense and hard-to-read inscription (Figure 14.30). The overall naivété of these inscriptions seems perfectly consistent with the newness of cursive writing in North Africa and with the apparent desire to steer close to an imported model with all its idiosyncrasies. This model, as suggested above, was the new calligraphic style in the ʿAbbasid capital – a style that had been formulated by Ibn al-Bawwåb and popularised by his many students. Copying one of the most important cultural symbols of the ʿAbbasid caliphate was perfectly consistent with the Almoravids’ strong links with the ʿAbbasids, whom they recognised from early on as the spiritual heads of Islam, and who in turn recognised them as rulers of al-Maghreb in the name of the caliph and SunnÈ Islam.80 The

339

340

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

numerous letters exchanged between YËsuf ibn TåshufÈn and his son ʿAlÈ and the various ʿAbbasid caliphs attest to the Almoravids’ veneration for the ʿAbbasids, whose name was included on their coinage and pronounced during the Friday khu†ba.81 The appropriation of this cultural symbol and its incorporation within the most important mosques of the Almoravids was therefore intended as a sign of homage to the ʿAbbasids and as a means to enhance the legitimacy of the Almoravid state. Egypt, from Fatimid to Ayyubid In discussing the inscriptions of the mosque of al-Íåli˙ Êalåʾiʾ, dated 555/1160, van Berchem concluded that they demonstrate that the KËfic script was used in historical inscriptions until the end of the Fatimid dynasty.82 Commenting on this transformation, Creswell declared that ‘henceforth the beautiful decorated KËfic script, the glory and pride of Fatimid art, was to be used no more for historical inscriptions but employed solely for decorative bands of quotations from the Qurʾån, and that to an ever decreasing extent’.83 Despite relatively minor recent objections to these conclusions, they remain as sound today as they were a century ago.84 Indeed, the earliest public cursive inscription in Cairo is Ayyubid: it is dated 575/1179 and once belonged to a madrasa built by Íalå˙ al-DÈn next to the shrine of Imåm ShåfiʾÈ.85 Although this inscription was not available for analysis, it seems perfectly appropriate that the earliest cursive inscription in Egypt should belong to the shrine of the most important theologian of SunnÈ Islam, and one that the Ayyubids in particular held in special regard.86 Commemorating the building of the shrine of Imåm ShåfiʾÈ and his wooden cenotaph by cursive inscriptions reinforces the fundamental transformation undergone by Egypt under the early Ayyubids. As was the case in Syria and Upper Mesopotamia, this transformation was to a large extent initiated by NËr al-DÈn, who had in 1164, 1167 and 1169 sent three expeditions to Egypt intended to retake it from the Fatimids and bring it back into the orthodox fold. All three forays were led by ShirkËh and his nephew Íalå˙ al-DÈn, who, after the death of his uncle in 1169, assumed real authority in Egypt. But despite his relative independence and the increasing tension between him and NËr al-DÈn, it should be emphasised that Íalå˙ al-DÈn ruled Egypt in his mentor’s name until the latter’s death in 1174.87 Religiously and ideologically, the legacy of NËr al-DÈn extended much farther than that, for Íalå˙ al-DÈn, true heir of his suzerain, became the champion of orthodox Islam and recaptured Jerusalem in the name of the ʿAbbasid caliph. A much better known inscription from the period of Íalå˙ al-DÈn still remains in situ in the Mudarraj Gate of the Cairo citadel (Figure 14.31). Dated 579/1183–4, it commemorates the completion of Íalå˙

the public text

Figure 14.31  Cairo. Citadel: inscription of Íalå˙ al-DÈn on the Mudarraj Gate, 579/1183 (photo: Yasser Tabbaa)

al-DÈn’s work on his foremost military installation, which in fact had become his official palace after he had abandoned his earlier residence in the Fatimid city.88 Made nearly a full century after the Seljuq inscriptions in Isfahan and Aleppo, fifty years after the Almoravid inscriptions in North Africa, and thirty-five years after the formulation of a monumental cursive style under NËr al-DÈn, this inscription is quite astonishing in its crudeness and carelessness.89 With a spindly line, inconsistent letter forms and neither points nor vowel marks, the script displays none of the refinements that had long been established in cursive monumental calligraphy. A similarly naive writing style is employed in another inscription bearing the name of Íalå˙ al-DÈn, a fragment preserved at the Islamic Museum in Cairo (Figure 14.32). Though considerably thicker, the script is stiff and untapered, and the letter forms are nearly indistinguishable from one another. Both inscriptions reflect the inexperience of local calligraphers in this new calligraphic style, which is quite surprising given the royal nature of the texts. Indeed, not until the latter part of the AyyËbid period did any monumental cursive inscriptions approach in quality those seen in Syria and Iran.90 By the time of the last important Ayyubid sultan, Najm al-DÈn AyyËb (1240–9), cursive monumental calligraphy in Cairo had already reached a level of development at least comparable to that in Syria and Iran (Figure 14.33). Conclusion With their conquest of Egypt in 969, the Fatimids introduced, or perhaps re-invented, the floriated KËfic script and facilitated its propagation in most of the Islamic world. Characterised by many writers

341

342

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 14.32  Cairo. Fragmentary inscription of Íalå˙ al-DÈn, 583/1187 (photo: Yasser Tabbaa)

as the most elegant variation of KËfic calligraphy, if not Islamic calligraphy as a whole, this script was gradually supplanted between c. 1075 and c. 1150 by cursive monumental scripts of the thuluth variety. This process began sporadically and rather indefinitely under the Ghaznavids and their successors the Great Seljuqs, from whose time we have the earliest in situ cursive inscription on a monument. Used at first simultaneously with floriated KËfic on the same monument, cursive calligraphy began to achieve its ultimate dominance during the reign of NËr al-DÈn in Syria. Under his guidance, not only did the cursive script replace the floriated KËfic in nearly all Syrian monumental inscriptions, but a truly monumental thuluth script was developed for the first time in stone. Simultaneous with the reforms of NËr al-DÈn was the tentative introduction of cursive official writing into central and western North Africa under the Almoravids. This independent development notwithstanding, there is ample evidence to suggest that it was NËr al-DÈn who catalysed, if not directly mandated, the switch from KËfic to cursive public inscriptions in Syria, various parts of the JazÈra, and ultimately in Egypt, the last stronghold of the floriated KËfic script. Although it has often been difficult to provide early examples of official cursive writing from Baghdad, I have stressed the central role of the ʿAbbasid caliphate throughout this piece and the preceding one. It was in Baghdad that the earliest semi-KËfic and fully cursive

the public text

Qurʾåns were produced, and it is quite possible that some of the earliest monumental cursive inscriptions were also made there. More likely, however, the earliest monumental cursive inscriptions in Iran were direct imitations of calligraphic specimens produced by the master calligraphers of Baghdad during the eleventh century – Ibn al-Bawwåb and his circle. Indeed, early cursive monumental writing, with its excessive cursiveness and its penchant for interconnection, betrays a direct dependence on paper calligraphy and specifically on the hand of Ibn al-Bawwåb. The appropriation and public display of the Qurʾånic script of Ibn al-Bawwåb by the newly emergent SunnÈ dynasties strongly suggest a degree of awareness of the religious and political implications of the new calligraphic style. Religiously, the supplantation of the highly ambiguous floriated KËfic script by clear and legible cursive scripts implied the acceptance and endorsement of the SunnÈ belief in the single and unambiguous nature of the word of God, whether in Qurʾåns or in public texts. The long-held duality in the meaning of the Qurʾånic message, which had been transformed by the Fatimids into an esoteric cult,91 was visibly challenged by a script whose legibility and accuracy left little room for variant readings and therefore variant interpretations. Without completely doing away with the dual nature of early Arabic official writing, especially as exemplified by the floriated KËfic, the new cursive script shifts the balance decisively in favour of the denotive over the connotive aspects of writing. Subsuming the mystical within the informational and the bå†in within the Ωåhir, the new public inscriptions perfectly embodied and eloquently propagated the exoteric and encompassing tendencies of the SunnÈ revival. Politically, the public display of a calligraphic style with indisputable links to the ʿAbbasids was intended to recognise the spiritual reign of the caliphate as well as symbolically affirm the legitimacy of the dynasty paying homage. This process is paralleled in the diplomatic sphere by the caliph’s bequest of titles and official garments in return for gifts received and the inclusion of his name on the coinage and in the khu†ba. Practised by most dynasties of the late eleventh and twelfth centuries – including the Great Seljuqs, Zangids, AyyËbids, Almoravids and Almohads – this reciprocal process aided the greatly weakened but newly assertive caliphate while providing some basis of legitimacy for these arriviste dynasties. Indeed, two of these dynasties, the Turkish Zangids and the Kurdish AyyËbids, finally brought down the Fatimid state and restored Egypt to SunnÈ orthodoxy. This symbiotic relationship between a centre possessing legitimacy but lacking power and a periphery lacking legitimacy but wielding real power had existed for several centuries, but it seems to have acquired a symbolic level of representation in the eleventh century. I have argued here and elsewhere that the late ʿAbbasid

343

344

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 14.33  Cairo. Madrasa of al-Íali˙ Najm al-DÈn AyyËb, 641/1243: foundation inscription above the portal (photo: Yasser Tabbaa)

caliphate was engaged in the production of symbolic forms (e.g. the muqarnas dome) and that these forms found wide acceptance in much of the SunnÈ Islamic world.92 Often originating in the nonofficial, even vernacular sphere, these forms were systematised in the tenth and eleventh centuries according to geometric processes,93 producing elegant types that were then used in highly significant contexts. Thus iconically charged, these forms became the veritable symbols of the SunnÈ revival and the resurgent caliphate and were as a result adopted and further developed by SunnÈ dynasties in various parts of the Islamic world. Neither cursive monumental writing nor the muqarnas dome94 entered official usage in Egypt before the end of the Fatimid period, but they became quite common once Egypt had joined the ranks of the SunnÈ world and declared its allegiance to the caliphate. Decentred during the tenth and eleventh centuries, the deeply fragmented Islamic world of the twelfth century was groping for

the public text

legitimate political authority and a spiritual centre. Real unity, which was impossible to achieve and for the most part undesirable, was replaced by ceremonial allegiance, and the caliphal symbols were introduced in order to mitigate the gap between reality and myth. The ‘semiotic contract’95 struck between the ʿAbbasid caliph and his distant and largely independent subjects ensured the wide dissemination of these symbols in a remarkably short time and their further development in the succeeding centuries. But the iconographic life of artistic forms was shortened by the absence in medieval Islam of an institution or an ecclesiastical body that would engage in a sustained textual interpretation of them. Thus, increasingly elaborate forms came to convey only a generalised Islamic meaning.

Notes 1. The concept of the public text in Arabic inscriptions was first discussed by Irene A. Bierman, ‘The Art of the Public Text: Medieval Islamic Rule’, in World Art: Themes of Unity in Diversity, ed. Irving Lavin, Acts of the XXVIth International Congress of the History of Art (University Park and London: The Pennsylvania State University Press, 1989), 283–90. 2. Yasser Tabbaa, ‘The Transformation of Arabic Writing: Part 1, Qur’ånic Calligraphy’, Ars Orientalis 21 (1992): 119–48. 3. L. A. Mayer, Islamic Architects and Their Works (Geneva: A. Kundig, 1956), 6. 4. Richard Ettinghausen, ‘Arabic Epigraphy: Communication or Symbolic Affirmation’, in Near Eastern Numismatics, Iconography, Epigraphy and History: Studies in Honor of George C. Miles, ed. Dickran Kouymijian (Beirut: American University of Beirut, 1974), 297–318; Erika Dodd, ‘The Image of the Word: Notes on the Religious Iconography of Islam’, Berytus 18 (1969): 35–62; Erika Cruikshank Dodd and Shereen Khairallah, The Image of the Word: A Study of Quranic Verses in Islamic Architecture, 2 vols (Beirut: American University of Beirut, 1981); Bierman, ‘Art of the Public Text’. Although Oleg Grabar’s earlier studies, in particular ‘The Umayyad Dome of the Rock in Jerusalem’, Ars Orientalis 3 (1959): 33–62, were primarily concerned with the informative and iconographic value of inscriptions, he has in his most recent work paid considerable attention to the formal and symbolic aspects of writing in Islamic art. See Oleg Grabar, The Mediation of Ornament (Princeton: Princeton University Press, 1992). 5. Ettinghausen, ‘Arabic Epigraphy’, 303. 6. Ettinghausen, ‘Arabic Epigraphy’, 306. 7. Ettinghausen, ‘Arabic Epigraphy’, 307. 8. Dodd and Khairallah, Image of the Word, 1: 25–6. 9. Bierman, ‘Art of the Public Text’, 284. 10. Bierman, ‘Art of the Public Text’, 285. 11. Grabar, Mediation, 60 and 66–9. Grabar eliminates early KËfic Qurʾåns from the realm of calligraphy, since in his view they do not demonstrate an adequate level of elaboration, or, perhaps more importantly, since their elaboration does not seem to be based on a set of well-defined

345

346

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

rules and standards of aesthetics. While it is true that early Qurʾåns are rather austere in their frugal use of gold and of fully illuminated folios, their worth and artistic merit resided in their exclusive use of vellum, an expensive medium that would have restricted their patronage to the upper stratum of society, and in their exacting and highly uniform calligraphic style. As for the absence of codes of beauty and manuals of calligraphy for the early scripts, there is little doubt that they once existed, although they probably differed considerably from the later manuals, with their step-by-step instructional approach. It would otherwise be impossible to explain the sudden development of a monumental script under the Umayyads from what must have been a very poorly developed system of writing. Grabar, Mediation, 94–5 and 103–4. 12. Grabar, Mediation, 113–14. 13. Max van Berchem, ‘Notes d’archéologie arabe. Monuments et inscriptions fatimites’, Journal Asiatique, 8’ sér., 17 (1891): 411–95, 18 (1891): 46–86; Sam Flury, Die Ornamente der Hakim – und Ashar – Moschée (Heidelberg: C. Winter, 1912); Sam Flury, Islamische Schriftbånder Amida-Diarbekr XI. Jahrhundert. Anhang: Kairuan, MayyåfåriqÈn, Tirmidh (Basel: Frobenius, 1920); partly translated as ‘Bandeaux ornamentés à inscriptions arabes: Amida-Diarbekr, IX siècle’, Syria 1 (1920–1): I, 235–49, 318–28 and II, 54–62; Sam Flury, ‘La mosquée de Nayin’, Syria 11 (1930): 43–58; Sam Flury, ‘Le décor épigraphique des monuments fatimides du Caire’, Syria 17 (1936): 365–76; Adolf Grohmann, ‘The Origin and Early Development of Floriated Kufic’, Ars Orientalis 2 (1957): 183–213; Janine Sourdel-Thomine, ‘Le coufique alépin de l’époque seljoukide’, Mélanges Louis Massignon (Paris: Maisonneuve et Larose, 1957), 3: 301–17; Aida S. Arif, Arabic Lapidary KËfic in Africa. Egypt – North Africa – Sudan (London: Luzac, 1967); Ibrahim Jumʾah, Diråsa fi ta†awwur al-kitåbåt al-kËfiyya’alå al-a˙jår fi mißr fi al-qurËn al-khamsa al-Ëla lil-hijra (Cairo: Dar al-Fikr al-ʾArabi, 1969); Mu˙ammad Kåmil Fåris, ‘al-Kha†† al-kËfÈ al-muwarraq fÈ ma’ålim Óalab al-athariyya’, Majallat ‘ådiyyat Óalab 1 (1975): 237–80. 14. Flury, Islamische Schriftbånder, 11. 15. Grohmann, ‘Floriated Kufic’, 207–8. William and Georges Marçais proposed that the floriated KËfic script may have come from North Africa to Egypt with the Fatimids (Les monuments arabes de Tlemcen [Paris: A. Fontemoing, 1903], 88). Grohmann, however, has convincingly refuted this position by pointing out that no true floriated KËfic existed in Tunisia during the early Fatimid phase. His fig. 2 depicts a tombstone from Qayrawan, dated 341/952, whose script has some foliated letters but cannot be called floriated. 16. Grohmann, ‘Floriated Kufic’, notes the existence of a number of ninth-century funerary stelae in Egypt and Palestine that contain early specimens of floriated KËfic. One inscription in the Óaram of Jerusalem dating between 301 and 304/913 and 917 does indeed show a KËfic script that had begun to sprout leaves. 17. Grohmann, ‘Floriated Kufic’, 212. 18. Grohmann, ‘Floriated Kufic’, 209. This important point is amply illustrated and discussed in Jum’ah, Diråsa, 209–14, who concludes that the inscriptions of the first three-quarters of the fourth/tenth century tend to be of poorer quality and simpler ornamentation than those of the preceding century. 19. Etienne Combe et al., Répertoire chronologique d’épigraphie arabe

the public text

(Cairo: Institut français d’archéologie orientale, 1932–40), vol. 2, no. 461; hereafter RCEA. 20. Jum’ah, Diråsa, 230–1. 21. Grohamnn, ‘Floriated Kufic’, 209. 22. Janine Sourdel-Thomine, ‘Kitåbåt’, EI2, 5: 217. Lisa (Volov) Golombek was, however, the first to note the process of the internal transformation of KËfic characters in ‘Plaited Kufic on Samanid Epigraphic Pottery’, Ars Orientalis 6 (1966): 107–34. 23. Golombek, ‘Plaited Kufic’, fig. 8. 24. Flury, Islamische Shriftsbånder, pls III–XIII. 25. Ernst Herzfeld, Matériaux pour un Corpus Inscriptionum Arabicarum, Inscriptions et monuments d’Alep (Cairo: Institut français d’archéologie orientale, 1954), vol. 2, pls 56–63 and 86–91. 26. Khaled Moaz and Solange Ory, Inscriptions arabes de Damas: les stèles funéraire (Damascus: Institut français, 1977), pls 4–5 and 43–4. 27. Henri Terrasse, La mosquée al-Qaraouiyin à Fès (Paris: Librarie Klincksieck, 1968), figs 39–50. 28. For excellent specimens of Iranian floriated KËfic see Sourdel-Thomine, ‘Kitåbåt’, figs 22–5. 29. Grohmann, ‘Floriated Kufic’. Indeed, Max van Berchem, ‘Notes d’archéologie arabes’, 18: 72, was the first to note the central role of the Fatimids in creating and disseminating the floriated KËfic: ‘it developed under the Fatimids of Egypt, the ʿAbbasids, the last Umayyads of Spain, and other Muslim dynasties until the introduction of cursive script. One finds it in all the Fatimid texts of Egypt, in a large number of the inscriptions of Syria, the Caucasus, Iran and Mesopotamia, Sicily, North Africa, and Spain’ (my translation). 30. Bierman, ‘Art of the Public Text’, 285–6. 31. W. Madlung, ‘Ismå’Èliyya’, EI2, 4: 203–5. 32. Max van Berchem, ‘Recherches archéologiques en Syrie. Lettre à M. Barbier de Meynard’, Journal Asiatique, 9’ sér., 6 (1895): 499; my translation. 33. See, for example, Ernst Herzfeld, ‘Damascus: Studies in Architecture, III’, Ars Islamica 11–12 (1956): 38. 34. I was not immune to this trend, and in my doctoral dissertation, ‘The Architectural Patronage of NËr al-DÈn, 1146–1174’ (New York University, 1983), I paid very little attention to the transformation of the script of official writing under NËr al-DÈn. 35. Van Berchem, ‘Notes d’archéologie arabe’, 18: 73. 36. In her recent compendium, The Monumental Inscriptions from Early Islamic Iran and Transoxiana (Leiden: E. J. Brill, 1992), Sheila Blair has proposed a number of tenth-century Buyid commemorative inscriptions from Persepolis as specimens of early cursive writing (p. 13 and figs 11–13 and 32–3). In fact, however, there is nothing cursive about these inscriptions, and they do not resemble or even ‘betray the shift toward’ the fully cursive style of Ibn al-Bawwåb. Furthermore, they are far more angular in the photographs than in the transcriptions, which considerably soften the ligatures, shorten the uprights and increase the space between words. The original (e.g. fig. 31, an inscription dated 392/1001–2) is quite angular, resembling an unornamented version of contemporary floriated KËfic. This kind of informality is to be expected in inscriptions that stand midway between graffiti and foundation inscriptions.

347

348

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

37. Excellent colour slides in the University of Michigan collection indicate that both cenotaph and triangular cover are made of the same material – a slightly translucent pinkish-grey marble or alabaster – and that both employ the same size slabs. 38. I have suggested reading the first word in the third line as ˙ujratuhu (his chamber, possibly even alluding to the hujra of the Prophet in Medina) rather than the pejorative and therefore very unlikely ˙ufratuhu (his hole). According to Lane’s Lexicon, there is no possibility of translating ˙ufra as ‘sepulchre’. 39. Sam Flury, ‘Le Décor épigraphique des monuments de Ghazna’, Syrua 6 (1925): 61–90, esp. 87–9; RCEA, vol. 6, no. 2377. 40. See, for example, Sourdel-Thomine, ‘Kitåbåt’, fig. 22: tomb of Mu˙ammad b. Wandarin, 407–11/1016–21. 41. Flury, ‘Ghazna’, 87–9, has argued that the entire pyramidal tomb cover, with its floriated KËfic and early cursive inscriptions, must postdate the tomb, suggesting that it may have been added to the original cenotaph after the fall of the Ghaznavids, as a sign of veneration to a great Islamic hero. His argument for this late date rests on three objections, one archaeological and two palaeographic. Archaeologically, Flury (on the basis of André Godard’s observations) has argued that the pyramidal top is uncommon for Gaznavid tombs and that it poorly connected with the rectangular cenotaph beneath it. Yet Flury himself shows a cenotaph (pl. XX, 1–2) for an unknown Ghaznvid dignitary with exactly the same composition. See also n. 37 above. Palaeographically, Flurry notes first that the KËfic inscription departs significantly from other Gaznavid inscriptions. But, once again, an examination of just the Ghaznavid inscriptions he illustrates in his article suffices to demonstrate the wide variety of KËfic writing under the Gaznavids, which is precisely the conclusion that Flury reaches at the end of the article (p. 90). Perhaps a better explanation for the uniqueness of Ma˙mËd’s KËfic inscription is technical: it is more deeply carved and modelled than other Gaznavid inscriptions. The second palaeographic objection has to do with the cursive inscription, which indeed does not resemble any other cursive inscription, Ghaznavid otherwise. But this perhaps indicates its early rather than late date, suggesting that it may have been produced at a time when calligraphers were not yet able to break away from the long and powerful tradition of the floriated KËfic. To assume, as Flury did, that this inscription was made in the late twelfth century, we would have to demonstrate that a model existed for this script, either in the early eleventh century or during the time of its alleged manufacture in the late twelfth century. Since neither of these postulates is at all likely, we would have to accept that this inscription was made around the time of Ma˙mËd’s death in 1030. Janine Sourdel-Thomine, ‘A propos du cénotaphe de MahmËd à Ghazan (Afghanistan)’, in Essays in Islamic Art and Architecture in Honor of Katherina Otto-Dorn, ed. Abbas Daneshvair (Malibu: Undena Publications, 1981), 127–35, has also questioned the early date of the cenotaph on the basis of stylistic and palaeographic evidence. But the stylistic peculiarities of the carving that she notes are most likely attributable to the fact (or at least great likelihood) that most of these stone cenotaphs were carved by Indian masons since Iranian artisans would have been quite unskilled for the task. Palaeographically, Sourdel-Thomine presents absolutely no comparisons or possible models for the cursive inscriptions and no explana-

the public text

42.

43. 44. 45.

46.

47.



tion as to how, indeed why, they could have been produced in the late twelfth century. For early specimens of this script, see Sourdel-Thomine, ‘A propos du cénotaphe’, fig. 26 (minaret near Balkh, dated 502/1109–10) and Adolph Grohmann, Arabische Paläographie, 2 vols (Vienna: Herman Bohlaus Nachforschungen, 1967 and 1971), 2: fig. 19. Tabba, ‘Qur’ånic Calligraphy’, 135–6. Tabbaa, ‘Qur’ånic Calligraphy’, figs 19–21. While the connection of Ibn Muqla with the ʿAbbasid caliphate cannot be questioned, the same is not true Ibn al-Bawwåb, who never achieved Ibn Muqla’s official status and who briefly served as keeper of the Buyid library in Shiraz. Indeed, his connection with the Buyids and the fact that the eulogy in his one preserved Qurʾån includes a reference to the ‘Pure Family of the Prophet’ has led David Storm Rice (The Unique Ibn al-Bawwåb Manuscript in the Chester Beatty Library) [Dublin: E. Walker, 1955], 13) to conclude that ‘Ibn al-Bawwåb shared the Shi’ite persuasion of his patrons, the Buwayhids’. There are many problems with this unfounded assertion. First, the Buyids themselves were only Shiʿis ‘in some vague sense’. Their ambivalence toward Shiʿism was manifested in their numerous concessions to the ÓanbalÈ caliphs; Roy Mottahedeh, Loyalty and Leadership in an Early Islamic Society (Princeton, NJ: Princeton University Press, 1980), 28. Second, the ʿAbbasids, as direct descendants of the Prophet and as ÓanblÈs, were loath to relinquish their claims to the family of the Prophet. Finally, there is absolutely no possibility that Ibn al-Bawwåb was Shiʿi, since his biography in Iban KhallÈkån (Wafiyyåt al-a’yån, 3: 342) states that ‘he died in Baghdad and was buried next to the imåm A˙mad ibn Óanbal’. He was, therefore, most likely a ÓanbalÈ and, as such, theologically opposed to Shiʿism. Indeed, his employment by the Buyids as librarian and calligrapher, rather than proving his Shiʿism, might be one indication that weakened Buids of the early eleventh century were quite willing to soften, or even abandon, some of their Shiʿi positions. See, in particular, Henri Laoust, ‘La pensée et l’ action politique d’al-MåwardÈ’, Revise des études islamiques 36 (1968): 11–92. Laoust demonstrates (pp. 65–6) that between 1007 and 1008 (almost the exact date of the Qurʾån of Ibn al-Bawwåb) the Buyid Bahåʾ al-Dawla was forced to banish the famous Shiʿi theologian Shaykh MufÈd and replace him with the AshʾarÈ theologian AbË Óåmid al-IsfaråʾÈnÈ, primarily because the former had persisted in using the Qurʾån of Ibn MasʾËd. C. E. Bosworth, ‘The Imperial Policy of the Early Gaznavids’, Islamic Culture 1 (1962): 50. Bosworth cites Jurbadhaqani’s description of this event: ‘When the news of the execution of the envoy from Egypt reached Baghdad, and the firmness of the Sultan’s faith became known, the tongues of calumniators and the reproofs of censorious ones were silenced, and his name was always mentioned with praise and honour at the court of the commander of the Faithful.’ Among the Iranian Seljuq monuments that bear both floriated KËfic and thuluth inscriptions are the following: 1. Isfahan, Masjid-È JåmÈ: mi˙råb dome (interior inscription KËfic, exterior cursive), dated 478/1086–7: fragment illustrated in Oleg Grabar, The Great Mosque of Isfahan (New York: New York University Press, 1990), fig. 34. 2. QazvÈn, Masjid-È JåmÈ, dated 509/1116: A. U. Pope, Survey of Persian Art (Ashiya, Japan: SOPA, 1964), 8. pl. 305; hereafter SPA.

349

350

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

3. QazvÈn, Masjid-È Óaydariya, first half twelfth century; SPA, 8, pl. 315. 4. BuzËn, Imåm-zåde Karrår, mi˙råb inscription, dated 528/1133–4; SPA, 8, pl. 311. 5. Tombstone, dated 533/1138; SPA, 8, pl. 520. 6. Ardistån, Masjid-È JåmÈ, dated 555/1160; SPA, 8, pls 522–4. 48. Some of these are illustrated in SPA, 7, pls 518c, 519e (dated 1141), and 520 (dated 1138). 49. Although this particular tombstone contains the Twelver Shiʿi formula, not all the other tombstones in this group were Shiʿi. Indeed, those illustrated in SPA (n. 47) are quite straightforward in their SunnÈ pious formulas. 50. Max van Berchem, Matériaux pour un corpus inscriptionum arabicarum, Syrie du Sud, Jerusalem, 2 vols (Cairo: Institute français d’archéologie orientale, 1922–7), 1: 88ff. and 254ff.; my translation. 51. Van Berchem, Jerusalem, 1: 87. The same idea is restated in ‘Notes d’ archéologie arabe I’, 18: 73–5; and ‘Inscriptions arabes de Syrie, VI, Les inscriptions de NËr al-DÈn et l’origine du charactere arrondi dans l’épigraphie syrienne’, M≠moire de l’institut égyptien 3 (1897): 34–9. 52. Herzfeld, ‘Damascus-III’, 38, and Alep, 1: 210–11. See also my ‘Monuments with a Message: Propagation of Jihåd under NËr al-DÈn, 1146–1174’, in The Meeting of Two Worlds: Cultural Exchange between East and West during the Period of the Crusades, ed. Vladimir Goss (Kalamazoo: Medieval Institute Publications, 1986), 223–40. 53. `Kitåbåt’, E12, 5: 217. 54. On Ibn al-Khashshåb’s involvement with the minaret, see Herzfeld, Alep, vol. 1, pt. 1, 160–2. 55. This would seem to fit in with the later Seljuqs’ ambivalence toward SunnÈ orthodoxy. Indeed, the Seljuq prince Ridwån went so far as briefly to pronounce the khu†ba in the name of the Fatimid caliph; see Kamål al-DÈn ibn al-ʾAdÈm, zubdat al-˙alab fi tårÈkh ˙alab, ed. Sami Dahhan (Damascus: Institut français de Damas), 2: 127–9. 56. Herzfeld, Alep, pl. LXXIXa. 57. Tabbaa, ‘Propagation of Jihåd’, 224. 58. Tabbaa, ‘Qur’ånic Calligraphy’, figs 20–1. 59. Claude Cahen, ‘Une chronique Chiite du temps des Croisades’, Comptes rendus, Academic des inscriptions et belles lettres (1935), 263–4. 60. H. A. R. Gibb, ‘The Career of Nur al-Din’, in A History of the Crusades, ed. K. M. Setton, vol. 1, The First Hundred Years, ed. M. W. Baldwin (Madison, WI: University of Wisconsin Press, 1969), 515. 61. On NËr al-DÈn’s links with Ibn Hubayra, see Nikita Elisséeff, NËr ad-DÈn, un grand prince musulman de Surie au temps des Croisades (511–569 h./1118–1174) (Damas: Institut Français de Damas, 1967), 3: 350–1; Herbert Mason, Two Statesman of Mediaeval Islam (The Hague: Mouton, 1972), 14 and 23; and Tabbaa, ‘NËr al-DÈn’, 178–81. 62. Tabbaa, ‘Propagation of Jihåd’, 227–9. 63. Such deliberate archaisms, though quite uncommon, are not unknown in Islamic art and architecture. See Terry Allen, A Classical Revival in Islamic Architecture (Wiesbaden: Ludwig Reichert Verlag, 1986), passim, where the author, however, speaks instead of a self-conscious revival of classical architecture. See also my ‘Survivals and Archaisms in the Architecture of Northern Syria, ca. 1080–1150’, in Essays in Honor of Oleg Grabar, ed. Margaret Šev¥enko Muqarnas 10 (1993): 29–41.

the public text

64. For a complete inventory of these inscriptions, see Nikita Elisséeff, ‘La titulature de NËr al-DÈn d’après ses inscriptions’, Bulletin des études orientales 14 (1952–4): 155–96. 65. See Tabbaa, ‘Propagation of Jihåd’, 229–31. The minbar contains three other long friezes in splendid thuluth carved on a deep bed of arabesque: two surrounding each side of the balustrade and a third just below the crenellated cornice. A non-royal inscription from the now destroyed mosque of Shaykh Mu˙ammad, dated 558/1163, displays an astonishing degree of development, particularly in the use of interconnection. This tendency has here been taken to an extreme without straining the legibility and elegance of the script. Each line contains several examples of interconnection, and some of the lines are even connected with one another vertically – a feature not encountered in any other inscription of the period. 66. The numerous inscriptions of al-ÛåhÈr GhåzÈ have been collated in Gaston Wiet, ‘Une inscription de Malik Ûåhir GåzÈ à Latakieh’, Bulletin de l’institut français d’archéologie orientale 30 (1931): 273–92. 67. Herzfeld, Alep, cf. the photograph of the inscription (pl. XXXVIIb) and the elevation drawing of the gate (Pl. XXXVIII). 68. Van Berchem, Jerusalem, nos. 118, 150, 38 and 155. 69. Monumental inscriptions from the period of this energetic caliph are quite plentiful. One of his longest and finest inscriptions once existed in the Båb al-Êilasm (Talisman Gate) in Baghdad, which was destroyed early in this century. Another fine inscription carved in wood frames the door to the so-called Båb al-Ghaibah (Gate of Disappearance) in Samarra, part of the Shiʿi sanctuary dedicated to the two imåms ʿAlÈ al-HådÈ and MËså al-ʾAskarÈ. See Anon., Bab ul Ghaibah at Samarra (Baghdad, 1938). 70. This inscription was first discussed by Max van Berchem in F. Sarre and E. Herzfeld, Archäologische Reise im Euphrat- und Tigris-Gebiet, 4 vols (Berlin: Verlag von Dietrich Reimer, 1911–20), 1: 17. On the basis of the mi˙råb’s early date and the assumption that it was made for the mosque of al-NËrÈ Herzfeld (Reise 2: 224–7) proposed that the mosque was begun more than twenty years before NËr al-DÈn, a contention that is not supported by any textual evidence. The truth is that this mi˙råb was brought to the mosque al-NËrÈ in the first decades of this century. It may have been originally intended for the Umayyad mosque of Mosul, which is now destroyed. I addressed the various chronological problems of this mosque in ‘NËr al-DÈn’, 147–51. 71. Elisséeff, NËr al-DÈn’, 2: 657–62. 72. These and other fragments of the mosque al-NËrÈ in Mosul are now preserved at the Iraqi Museum in Baghdad. I discussed them in ‘NËr al-DÈn’, 153–4, fig. 289. 73. These two unusual and sumptuously decorated mausoleums are discussed in Sarre and Herzfeld, Reise, 2: 249–70. Although built for alleged descendants of various Shiʿi imåms, these shrines do not represent the revival of political Shiʿism. Indeed, the official veneration of Shiʿi shrines in the late twelfth century was even practised by the contemporary ʿAbbasid caliph al-Nåßir, whose policy of rapprochement with the Shiʿis was intended to strengthen his power base which the local population. See Mason, Two Statesmen, 99 and 116. 74. Examples of this large thuluth are quite common in Mosul: 1.  Qara Sarai, the palace of Badr al-DÈn Luʾluʾ, 630/1232: a large ­inscription carved in stucco.

351

352

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

2. Shrine Imåm Ya˙ya Abu ʿl-Qåsim (637/1239–40): exterior inscription above the portal, a short frieze above the corner mi˙råb, and a long frieze at the springing of the dome. 3. Shrine of Imåm ʿAwn al-DÈn (646/1248): inscription above the portal to the shrine. 75. This development has not been studied in depth; meanwhile, see Georges Marçais, Marçais, L’ Architecture musulmn d’ occident (Paris: Art et métiers graphiques, 1954), 250, and Terrasse, al-Qaraouiyin, 51 and 80. For the parallel development in Sicily, see ʿAbd al-Munʾim Raslån, Al-˙a∂åra al- islåmiya fi ßiqilliya wa junËb ‘Ȇåliya (Jeddah: al- Kitåb al-JåmiʾÈ, 1980), 80–2. 76. L. Golvin, ‘Kitåbåt’, EI2, 221. Golvin adds that ‘it is in fact impossible, in the absence of precise documentary evidence, to propose an Andalusian influence, as it seems that cursive writing did not appear in Andalusia until much later’. This is an important observation, for it underlines a switch in the prevailing cultural influences on al-Maghrib, from Andalusia to the central Islamic world. 77. This highly original dome, the first in North Africa to use muqarnas in its squinches, has been studies by Marçais, L’ Architecture musulman, 195–7 and figs 125–6. 78. Terrasse, al-Qaraouiyin, chs 2–7, discusses different facets of the Almoravid reconstruction. In the Capella Palatina – a monument that attempts to combine the latest in Byzantine, Norman and Islamic art – the use of cursive calligraphy may be seen as just another example of an up-to-date borrowing from the Islamic world. 79. See, for example, Tabbaa, ‘Qur’ånic Calligraphy’, fig. 27. 80. Link between the Almoravids and the ʿAbbasid caliphs have been alluded to by several writers, including Mu˙ammad A. ʿAnån, Duwal al-†awå’if mundhu qiyåmihå ˙attå al-fat˙ al-muråbi†È, 3rd edn (Cairo: al-Khanji, 1988), 314ff. and more recently, Salåmah M. S. al-HirfÈ, Dawlat al-muråbi†Èn fi ‘ahd ‘AlÈ b. YËsuf b. Tåshufin (Beirut: Dar alNadwa al- Jadida, 1985), 168–75. 81. Al-HirfÈ, Dawlat al-muråbi†Èn, 170–2. The author mentions no fewer than seven exchanges of letters between the Almoravids and the ʿAbbasids, between 1059 and 1118, covering the reigns of AbË Bakr ibn ʿUmar (1059–73), YËsuf ibn TåshufÈn (1061–1106) and ʿAlÈ ibn YËsuf (1106–42). In the correspondence of 479/1086, YËsuf received the ʿAbbasid caliph’s approval of his newly assumed title amÈ al-muslimÈn along with a lengthy letter from AbË Óåmid al-GhazzålÈ, in which the great theologian praises YËsuf as one of the great heroes of Islam. 82. Van Berchem, ‘Notes d’ archéologie arabe I’, 18.69. 83. K. A. C. Creswell, The Muslim Architecture of Egypt, 2 vols (Oxford: Oxford University Press, 1959), 2: 35. 84. Caroline Williams, ‘The Qur’anic Inscriptions of the Tabut al-Husayn in Cairo’, Islamic Art 3 (1987): 3–14, has made a strong case for a Fatimid dating of this commemorative cenotaph despite the fact that it contains both floriated KËfic and cursive inscriptions. But Williams’ argument, which largely rests on the improbability that Íalå˙ al-DÈn would have ordered a cenotaph with such vividly Shiʿi inscriptions, is not completely foolproof. Indeed, the decline of political Shiʿism in the late twelfth century seems to have brought about a renewed tolerance of its pietistic aspects. One of the prime movers in this ecumenical policy was the ʿAbbasid caliph al-Nåßir (1180–1225), who was known

the public text

for this tolerance of Twelver Shiʿism and for his patronage of some Shiʿi shrines, including Båb al-Ghayba in Samarra. This policy may have sanctioned similar acts of tolerance in Mosul under Badr al-DÈn Luʾluʾ (see n. 73) and under al-Ûåhir GhåzÈ of Aleppo, who restored two important Shiʿi shrines in Aleppo, one of which was dedicated to al-Husayn (see n. 66). In her recent book, Islamic Architecture in Cairo: An Introduction (Leiden: E. J. Brill, 1992), 77, Doris Behrens-Abouseif has noted in passing that one of the window grilles in this mosque contains a cursive inscription. Since she does not illustrate this inscription, it is impossible to determine the veracity of her observation. But even if such an inscription does exist, its inconspicuous location would seem to undermine its importance. Could it have been a subtle attempt by a craftsman to introduce a feature that was not part of the intended decorative programme? 85. Creswell, Muslim Architecture of Egypt, 2: 64. 86. Imåm ShåfiʾÈ lived in Cairo for the last fifteen years of his life, where he died in 820. Although a small shrine had always existed at his burial, Íalå˙ al-DÈn began bulding a new one with a magnificent wooden cenotaph (also bearing cursive inscriptions) and an associated madrasa the very same year that he had declared the end of the Fatimids and his own sovereignty. 87. Elisséeff, NËr al-DÈn, 2: 573ff. 88. Nasser Rabbat, ‘The Citadel of Cairo, 1176–1341: Reconstructing Architecture from Texts’ (Ph. D. diss., Massachusetts Institute of Technology, 1991), 17ff. 89. RCEA, no. 3380. Even the Répertoire, which is not noted for its aesthetic judgement, described this inscription as ‘d’un trait lache et peu soigné’. 90. A quick survey of the photographs in Creswell, Muslim Architecture in Egypt, vol. 2, suggests that this began during the reign of al-Malik al-Kåmil (1218–88). 91. See, for example, Marius Canard, ‘Få†imids’, EI2, 2:859; and W. Madelung, ‘Ismå’Èliyya’, EI2, 4: 203–5, who provides a detailed presentation of Fatimid esoteric theology. 92. Yasser Tabbaa, ‘The Muqarnas Dome: Its Origin and Meaning’, Muqarnas 3 (1985): 61–74. 93. The basic outline of these processes has been preserved for calligraphy by the treatise of Ibn Muqla, and for geometric ornament by AbuʾlWafåʾ al-BuzjånÈ and others. 94. See Jonathan Bloom, ‘The Introduction of the Muqarnas into Egypt’,Muqarnas 5 (1988): 27, where the author concludes that ‘muqarnas squinches belonged to vernacular architecture in the Fatimid period, and would have been inappropriate for building commissioned by the court’. 95. I owe this idea, like many others, to Professor Grabar; see his ‘The Meaning of History in Cairo’, in The Expanding Metropolis (Cambridge, MA: Aga Khan Publications, 1985), 12.

353

CHAPTER FIFTEEN

Canonicity and Control: The Sociopolitical Underpinnings of Ibn Muqla’s Reform

Figure 15.1  Al-Aßmaʿi, TaʿrÈkh mulËk al-ʿArab, 243/957. Paris, Bibliothèque Nationale, Arabe 6726, fol. 2v

Nearly every historical treatise on Arabic calligraphy begins with a similar history of writing, a historical narrative whose naivété and repetitiveness veil its heuristic significance. Told sequentially as the solitary acts of saintly figures and calligraphers, the history of calligraphy highlights the role of individual calligraphers but provides minimal cultural context for their accomplishments. The canonical list of calligraphers includes legendary or near-legendary figures, among whom are the prophets Seth, Enoch and Moses and the caliphs ʿAli and ʿUthmån. Calligraphers of the early ʿAbbasid period tended to be high officials, such as al-Fa∂l b. Sahl and al-A˙wal, culminating in the vizier Ibn Muqla.1 Under the Buyids and later ʿAbbasids, calligraphers were professional scribes, such as Ibn Yasser Tabbaa (1999), ‘Canonicity and Control: The Sociopolitical Underpinnings of Ibn Muqla’s Reform’, Ars Orientalis 29, 91–100.

canonicity and control

­al-Bawwåb and YåqËt, who had demonstrated a special gift in the art of calligraphy. This ‘history’ is obligingly, but uncritically, included in most modern studies on Arabic calligraphy, or it is dismissed as lacking a factual basis. Approached on its own terms, however, this canonical narrative potentially raises a number of important issues that bear directly on the changing role of calligraphy and calligraphers in an evolving Islamic society. The first concerns the descending social status of calligraphers: the earliest calligraphers were men of high rank and religious learning; Ibn Muqla was a patrician who became a vizier; Ibn al-Bawwåb was a man of humble origin who rose to the rank of scribe and librarian; YåqËt was a slave. The second has to do with the decreasing independence of calligraphers concomitant with their increasing reliance on patronage. Even disregarding such legendary calligraphers as the caliphs ʿAli and ʿUthmån, evidence suggests that the first calligraphers – those who wrote the earliest Qurʾåns – were learned scholars who were not in the direct employ of sovereigns or princes.2 Later calligraphers, on the other hand, particularly after Ibn al-Bawwåb, relied greatly or even exclusively on princely patronage, culminating in those calligraphers who were employed by the kitåbkhana.3 The third and most general observation about this hierarchical tale is that it is not so much a history as a mythology of Arabic writing, or, more specifically, a legend of the downward and outward spread of literacy from the elevated source of prophets and caliphs to the diversity and multiplicity of a complex multi-ethnic culture. Ibn Muqla (886–940) stands in the chronological and ideological middle of this process as the first calligrapher to carry out a comprehensive reform of Qurʾånic calligraphy, conducted, as I have previously argued, at the behest of the ʿAbbasid state.4 His calligraphic reform entailed the creation of geometric templates for each letter of the alphabet, resulting in a system of proportional writing (al-kha†† al-mansËb) that was then applied to the six main scripts of his time. The geometric precision of this process, combined with the systematisation of orthographic and vocalisation signs, produced clear and legible scripts that were deemed worthy of the Qurʾån. Often called semi-KËfic, although ‘new ʿAbbasid KËfic’ is perhaps more appropriate, these new script(s) were marked by a clarity and legibility that reflected the Ashʾari (or generally SunnÈ) belief in the exoteric nature of the word of God and the uncreated nature of the Qurʾån. But Ibn Muqla’s calligraphic reform also had social and political dimensions that I did not sufficently emphasise in my earlier articles, for his reform was very likely intended to challenge the authority of the calligraphers of early KËfic Qurʾåns while insisting on the authority of the ʿAbbasids in controlling this process. Engendered in an increasingly literate Islamic world, Ibn Muqla’s reform acknowledges, even endorses, the inevitability of increased writing (especially of the Qurʾån) but places new limits on this

355

356

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

f­undamental change. In the following discussion, I would like to turn to these social and political aspects of the reform by focusing on Ibn Muqla’s links with the ʿAbbasid state and the new role created for calligraphy and calligraphers subsequent to this transformation. I shall therefore attempt to reconcile my theological interpretation of this phenomenon with sociological and political information about Ibn Muqla as calligrapher, reformer of the Qurʾånic script and vizier to three caliphs. Since no works of Ibn Muqla have survived, I shall refer to his closest known successor, ʿAlÈ b. Shådhån al-RåzÈ (active 972–86), a calligrapher who produced both secular and Qurʾånic manuscripts.5 Despite Ibn Muqla’s well-known influence on Qurʾånic writing, it should be made clear from the start that he was neither a Qurʾånic calligrapher nor someone especially noted for his religious knowledge.6 As a scribe and state official, he stands apart from early Qurʾånic calligraphers, who, according to Ibn DurustËyah, were men knowledgeable in the Qurʾån and other religious matters.7 Calligraphically speaking, scribes (kuttåb) and early Qurʾånic calligraphers (kha††a†Ën) were worlds apart. Whereas the former were keenly interested in clarity and legibility, the latter were more concerned with maintaining the integrity and sanctity of the sacred text, concerns that were better served by using a nearly illegible script.8 It follows then that Ibn Muqla’s encroachment on the world of Qurʾånic calligraphers and his decisive impact on the development of Qurʾånic calligraphy were not simply internal developments in the craft but ones necessarily motivated by external factors. Before we attempt to describe these factors, it seems necessary to review the situation of writing around the time of Ibn Muqla’s calligraphic reform. Concerning Qurʾånic writing, its great uniformity in the first three centuries of Islam bespeaks a highly conservative and restrictive attitude towards the transcription of the Qurʾån.9 With ambiguous and often undifferentiated letter forms and a scattered disposition on the page, KËfic Qurʾåns of the ninth and tenth centuries were practically illegible except to those who had already memorised the text (i.e. ˙uffaz).10 In other words, these Qurʾåns were created not so much to be read but to validate the act of recitation and to venerate the word of God. The manuscripts speak of privilege and a restrictive attitude to the act of reading: rare materials, exquisite ornament and a nearly indecipherable script.11 In contrast, secular scripts – which can be subdivided into scribal scripts and book scripts – were quite legible, despite their considerable variation.12 Writing at the end of the ninth century, Ibn alNadÈm listed twenty-six scripts used by the scribes of his time, ranging from large and angular to small and cursive.13 The task of matching these scribal scripts to extant specimens has proved to be very difficult, not the least because so few early medieval documents and letters have survived.14 Book scripts, on the other hand, ranging

canonicity and control

from semi-angular to cursive, were quite commonly used in literary and scientific manuscripts of the ninth and tenth centuries (Figure 15.1). Although some of these were copied by the author of the treatise himself, more commonly authors left the task of making clean copies to professional copyists.15 Interestingly, these ‘transitional’ book scripts were also commonly used in a variety of Arabic Christian texts, including gospels, psalters and monastic anthologies (Figure 15.2). A cursory survey of this little-known phenomenon suggests that Christian manuscripts were written in new ʿAbbasid KËfic scripts as early as the last quarter of the ninth century, whereas those written in cursive scripts generally date to the second half of the tenth century.16 In other words, the use of book scripts in Christian manuscripts long pre-dates the transformation in Qurʾånic writing but is generally contemporary with their use in Arabic secular manuscripts. Indeed, the use of these scripts for Christian texts attests to their popularity and strengthens the case for their ‘secular’ background, from an Islamic perspective. On the eve of the reforms of Ibn Muqla, Arabic was being written in an ambiguously majestic Qurʾånic script and in an unwieldy variety of secular scripts, mostly used by scribes (kuttåb) for writing documents and letters and by booksellers or copyists (warråqËn) for the copying of various manuscripts. Ibn Muqla’s rules for proportional writing (al-kha†† al-mansËb) did not emerge from Qurʾånic KËfic but were rather based on these multifarious book scripts, which were also initially the subject of the reform.17 In other words, Qurʾånic KËfic, which by the tenth century had reached a very high standard, was not directly affected by the changes of Ibn Muqla; the reform was intended for the more mundane scripts used by scribes and copyists rather than calligraphers. The result of these reforms, therefore, was not the gradual softening of the angular KËfic script but its supplantation by the redesigned scripts of the copyists. Ibn Muqla thus created order where disorder had been perceived within scribal writing, a feat that earned him heroic stature among later Muslim biographers. Since success is often equated with quality, the success of Ibn Muqla’s proportional writing made him the father of the new Arabic calligraphy, despite the fact that he may not have been an especially gifted calligrapher himself. Indeed, the emphasis by connoisseurs from medieval times to the present on finding authentic specimens in the hand of Ibn Muqla has diverted attention from properly investigating the recipe and legacy of his success, which was certainly not entirely based on his calligraphic hand. Although we lack any authentic specimens in Ibn Muqla’s hand, there is little question that it would have resembled the earliest examples of new ʿAbbasid KËfic Qurʾåns. I have presented this argument previously18 and would like here simply to demonstrate the possible impact of Ibn Muqla’s method on ʿAli b. Shådhån al-RåzÈ,

357

358

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 15.2  New Testament, Timothy 4:1f. Jerusalem 902. Paris, Bibliothèque Nationale, Arabe 6725, fol. 5v

who is known to us both as a copyist of a literary tract and as a Qurʾånic calligrapher. The secular work is a book entitled Kitåb akhbår al-na˙wiyyÈn al-baßriyyÈn (Tales of the Grammarians of Basra), dated 376/986, a period from which we have several other related manuscripts (Figure 15.3). Written in a reasonably legible, fully vocalised ʿAbbasid KËfic script, this manuscript probably represents the high end of secular manuscripts produced in the late tenth century. ʿAli b. Shådhån’s Qurʾån, dated 361/972, is the earliest dated Qurʾån manuscript in the ʿAbbasid KËfic script and also the first Qurʾån written on paper (Figure 15.4). Closely related to, if perhaps more conservative than the calligrapher’s later secular manuscript, this Qurʾån nevertheless demonstrates the close linkages between Qurʾånic and non-Qurʾånic calligraphy in the aftermath of Ibn Muqla’s reforms. Written about one generation after the death of Ibn Muqla, this Qurʾån manuscript represents the direct influence of Ibn Muqla’s calligraphic method, the transmission of

canonicity and control

Figure 15.3  Al-Siråfi, Kitåb akhbår al-na˙wiyyÈn al-baßriyyÈn, calligrapher ʿAlÈ b. Shådhån al-RåzÈ, Iraq/Iran, dated 986. Istanbul, Suleymaniye Library (Šehid Ali 1642), fol. 191a

this method from secular to Qurʾånic manuscripts and the impact of paper p ­ roduction on both processes. Indeed, the widespread use of paper after the tenth century in chancery documents and secular manuscripts may have contributed to the speed of execution required by scribes and book copiers. Cheaper and more widely available than earlier parchment and papyrus, paper greatly facilitated the work of these scribes and promoted the expansion of literacy.19 The growth in the number of scribes and the literate population seems to have been accompanied by the relaxation of calligraphic standards and a general decline in the quality of

359

360

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

writing.20 Some system was urgently needed for the reform of secular writing, and this was provided by Ibn Muqla in the form of al-kha†† al-mansËb. The switch from vellum to paper also led to the transfer of ‘differentiations in value from the medium itself to what was put on it’,21 a point that is addressed below. Although generally discussed in aesthetic terms, Ibn Muqla’s innovations were primarily concerned with clarity and legibility, concerns that seem consistent with his role as a state official.22 While Ibn Muqla’s reform grew out of earlier trends towards clarity in scribal and manuscript writing, his reform was the most systematic and pervasive. This reform was engendered within an atmosphere of increasing literacy – brought about by paper – and was intended to remedy a situation resulting from this burgeoning of the literate population. It resulted in the creation of a series of templates for the canonical calligraphic scripts, which guaranteed quality and consistency. But this standardisation came at a price: a relatively small number of scripts formed the canon of reformed scripts, while others were neglected and slipped into oblivion.23 The power implications of this standardisation and canonicity are, I think, fairly straightforward. Brinkley Messick in his recent book The Calligraphic State expatiates on the links between the introduction of new writing systems in Yemen at the beginning of the twentieth century and the rise of a new power structure.24 Specifically, he argues that the switch that took place from organically formed spiral texts to texts with a standardised linear format implied enforced changes in the relation between form and content and between the state and the population. Although the change in modern Yemen from manuscript to print culture is more abrupt and the sources on it more ample, both situations describe a process by which new writing systems are deployed for affirming power and asserting control. Indeed, the ʿAbbasid reforms entailed control of the scripts, control of the scribes who had to be retrained in these scripts, and ultimately control of the content, the texts for which these scripts were to be used. Although these reforms are attributed by contemporary writers directly to the creative genius of Ibn Muqla,25 there is no question that their success and quick impact resulted from their adoption by the ʿAbbasid state. As vizier to three successive ʿAbbasid caliphs – al-Muqtadir, al-Qåhir and al-Rå∂i – Ibn Muqla was embroiled in the politics and intrigue of the ʿAbbasid state. I have previously explored his involvement under al-Muqtadir with the creation of a canonical body of Qurʾånic recensions (qiråʾåt) that were intended to put an end to discord while forever abolishing the legitimacy of aberrant recensions, particularly that of Ibn MasʿËd.26 Indeed, Ibn Muqla was certainly involved in the trials of two of the variant readers, Ibn Miqsam and Ibn ShanabËdh, the latter of whom was beaten and tortured into acquiescence. 27

canonicity and control

Figure 15.4  Qurʾån, Iran (other part of same ms. at the University Library in Istanbul [A6758] is dated 361/972), calligrapher ʿAlÈ b. Shådhan al-RåzÈ. Dublin, The Chester Beatty Library, 1434, fol. 164b

In essence, therefore, the ʿAbbasid state used trusted members of its administration to try, judge and punish Qurʾånic scholars who were deemed divergent. Although state functionaries with no particular claim to religious knowledge, Ibn Muqla and his cohorts were placed in a position to enforce a particular religious dogma and to punish those who persisted in departing from it. This is a curious situation, though not the first time that the ʿAbbasid state had resorted to such repressive measures: the mi˙na of Ibn Óanbal presents a similar, though ideologically opposite, case.28 In effect, the trials ordered by al-Muqtadir and conducted by Ibn Muqla demoted traditional

361

362

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Qurʾånic readers and valorised a state version of the Qurʾån that was promoted and even copied by men of the administration. The fact that calligraphers of the KËfic Qurʾån were probably drawn from ʿʿulamaʾ circles may have contributed to the ultimate supplanting of their style and manner of writing by the newly canonised calligraphic modes. Thus, Ibn Muqla created a new calligraphic system, eventually applied to the Qurʾån, and was the vizier who enforced the caliphal order to establish a body of canonical Qurʾånic readings. The two roles are undoubtedly related: the adoption of al-kha†† al-mansËb for copying the Qurʾån was inspired by the canonisation of the text of the Qurʾån. The new script, with its improved orthography and the correct numeration, would have left no doubt in the mind of Muslims that they were reading one of the new orthodox recensions, certainly not a Qurʾån with an aberrant reading. The canonisation of the text is made clear and visible by the new canonical script, and the two processes conjoin to reaffirm the absolute control of the content and the form of the Sacred Book by the ʿAbbasid state. Control is therefore essential to the creation of proportional writing and its application to the Qurʾån, thereby ending three centuries of KËfic writing. Although exactly how scripts were transferred from the secular to the religious domain remains incompletely known, the highlights are fairly clear. Three main processes were at work: the reform of scribal writing, the canonisation of the Qurʾånic text and the application of proportional writing to the Qurʾån. Linked together by webs of power, these processes led to the transformation of the form of the Qurʾån. Although little discussed by most modern writers, this was perhaps the most significant ‘artistic’ innovation of the middle ʿAbbasid period, instigated by the ʿAbbasid state. Finally, it is curious that the rise of calligraphy as an art form, one that becomes the object of criticism and collecting, only begins after the reforms of Ibn Muqla and the creation of the new ʿAbbasid scripts. Oleg Grabar has proposed two explanations for this curious phenomenon, which in fact goes counter to contemporary preference for the KËfic script over later cursive scripts. The first is that as paper replaced vellum in the tenth century, there was a shift in value from the medium to what was written upon it, hence to calligraphy. The second is that the spread of literacy, also related to the availability of paper, would have created a market in which calligraphic products circulated as objects of cultural value.29 To these, I can add a third factor: name recognition. Ibn Muqla and Ibn al-Bawwåb are not simply the first calligraphers known to us by name, but their names were also associated with the caliphs and princes for whom they worked. They become rubrics of recognition: later calligraphers imitate their style, and even forgers attribute works to their names. The two initiate the genealogy of calligraphers with whom I began this paper, hut they were not the lone actors implied by the sources.

canonicity and control

They were rather part of an intricate social, political and theological construction that shaped their careers and gave meaning to their creative efforts. Acknowledgements This piece was first presented as a conference paper at ‘Inscription as Art in the World of Islam’, Hofstra University, 25–7 April 1996. I take this opportunity to thank Dr Habibeh Rahim for organising the conference and for allowing me to publish this paper outside its projected format. I also thank the two anonymous readers, who have helped me tighten the argument of this piece and expand on some of its historical dimensions. I regret that I was not always able to follow the many excellent suggestions made by the second reader. Readers of Ars Orientalis will note that this piece is related to my two earlier articles in this journal (1991 and 1994), in which I dealt with the transformation of Arabic writing, Qurʾånic and epigraphic, from angular to cursive scripts.

Notes 1. See e.g. The Fihrist of al-NadÈm, ed. and tr. Bayard Dodge (New York: Columbia University Press, 1970), 8–12; Qalqashandi, Íub˙ al-Aʾshå (Cairo, 1962), 3: 10–14; and Mu˙ammad b. Óasan al-ÊÈbÈ, Jåmiʾma˙åsin kitåbat al-kuttåb, ed. Salahuddin al-Munajjid (Beirut: Dar al-Kitab alJadid, 1962), 13. See also Calligraphers and Painters. A Treatise by Qå∂È A˙mad, son of Mir Munshi (ca. a.h. 1015/a.d. 1606), tr. T. Minorsky (Washington, DC: Smithsonian Institution, 1959), 52–5, for a genealogy that also includes ancient Persian kings and Shiʿi imams. For the early history of Arabic writing, see Salahuddin al-Munajjid, Diråsåt fi tårikh al-kha†† al-ʿarabi mundhu bidåyatihi ilå nihåyat al-ʿaßr alumawiyy (Beirut: Dar al-Kitab al-Jadid, 1972), 23; and Yasin Safadi, Islamic Calligraphy (Boulder: Shambhalla, 1979), 7–8. 2. On the distinction between scribe (nassåkh or warråq) and calligrapher (kha††å†) see Johannes Pedersen, The Arabic Book (Princeton: Princeton University Press, 1984), 43ff. and 83ff. See also Estelle Whelan, ‘Early Islam, Emerging Patterns, 622–1050’, in Islamic Art and Patronage: Treasures from Kuwait, ed. Esin Atil (New York: Rizzoli International Publications, 1990), 27–40. 3. YåqËt al-Mustaʿßimi, for example, was an Abyssinian slave in the court of the last ʿAbbasid caliph, al-Mustaʿßim. The patronage of calligraphers by princes is amply demonstrated for later periods. For example, Thomas W. Lentz and Glenn D. Lowry, Timur and the Princely Vision: Persian Art and Culture in the Fifteenth Century (Los Angeles: Los Angeles County Museum of Art and Arthur M. Sackler Gallery, 1989), 159ff. 4. Yasser Tabbaa, ‘The Transformation of Arabic Writing: Part I, Qurʾånic Calligraphy’, Ars Orientalis 21 (1991): 119–48. 5. Holly C. Edwards, ‘A Study of Eastern Kufic Calligraphy’ (MA thesis, University of Michigan, 1981). This excellent thesis meticulously

363

364

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

analyses a group of tenth- and eleventh-century secular and Qurʾånic manuscripts, while highlighting the work of ʿAlÈ b. Shådhån al-RåzÈ. 6. For a summary of Ibn Muqla’s entirely bureaucratic and administrative career, see Dominique Sourdel, ‘Ibn Mukla’, EI2, 3: 886–7. 7. Estelle Whelan, ‘Writing the Word of God: Some Early Qurʾån Manuscripts and Their Milieux, Part I’, Ars Orientalis 20 (1990): 122, where the author suggests that ‘Ibn DurustËyah included copyists of the Qurʾån among the ʿulamåʾ, which is also confirmed by the manuscripts themselves’. See Ibn DurustËyah, Kitåb al-kuttåb, ed. Ibrahim al-Samarraʿi (Beirut: Dar al-Jil, 1992), 20, where the author expressly exempts Qurʾånic calligraphers from the rules of orthography discussed in his book. 8. On the conservatism of Qurʾånic scripts and calligraphers, see Nabia Abbott, ‘Arabic Paleography: The Development of Early Islamic Scripts’, Ars Isfumica 8 (1941): 83, who proposes that Qurʾånic ‘writing, including spelling rules and scripts, became established as a Sunna or sacred practice, as one learns from Ibn Durustuya and others’. 9. François Déroche, Catalogue des manuscrits arabes, Les manuscrits du Coran, I: Aux origines de la calligraphie coranique (Paris: Bibliothèque Nationale, 1983). The author has attempted further to subdivide the well-known categories of Måʿil (slanted script) and early KËfic into smaller and more precise groups or families of manuscripts. But keeping within the central KËfic groups one notices a remarkable degree of consistency in the letter forms and the overall appearance of the scripts. 10. On the problem of legibility, see in particular Richard Ettinghausen, ‘Arabic Epigraphy: Communication or Symbolic Affirmation’, in Near Eastern. Numismatics, Iconography, Epigraphy and History: Studies in Honor of George C. Miles, ed. Dickran Kouymijian (Beirut: American University of Beirut, 1974), 297–318; Erika Dodd, ‘The Image of the Word: Notes on the Religious Iconography of Islam’, Berytus 18 (1969): 35–62; and Oleg Grabar, The Mediation of Ornament (Princeton: Princeton University Press, 1992), 60–8. 11. Oleg Grabar has recently dealt with complex writing systems as emblems of privilege in the medieval world, proposing that the acquisition of objects with nearly indecipherable scripts became one of the criteria for justifying and sustaining the power basis of a social and political elite. See Mediation of Ornament, ch. 2. 12. Whelan, in ‘Early Islam’, expanding on Abbott, was the first to make a cogent case for subdividing non-Qurʾånic writing into scribal or secretarial and book scripts. See also Nabia Abbott, ‘Arabic Paleography’, Ars Islamica 8 (1941): 76f. 13. Fihrist, 13–15. 14. In The Rise of the North Arabic Script and Its Kurʾånic Development (Chicago, 1938), Nabia Abbott did in fact succeed in identifying one script, al-musalsal. But, overall, her efforts in this regard were not successful. 15. Pedersen, Arabic Book, 45: ‘it was not uncommon in the time of the early ʿAbbåsids for an author to have his special warråq’. 16. For other specimens, see Georges Vajda, La palaeographie arabe (Paris, 1953), pl. 4; and especially Evgenivs Tisserant, Specimina Codicvm Orientalivm (Rome, 1914): pl. 54: Vat. ar. 7: Florilegium Monasticum, dated 885; new ʿAbbasid KËfic script with elaborate letter forms; pl. 55: Borg. ar. 71: Evangelia, ninth century; new ʿAbbasid KËfic script;

canonicity and control

pl. 45a: Vat. ar. 18: Evangelium, sec Lucam, dated 993; cursive (naskh) script. 17. The canon of proportions in the treatise of Ibn Muqla has been graphically reproduced by Ahmad Mustafa in his unpublished MA thesis, ‘The Scientific Construction of Arabic Alphabets’ (University of London, 1979). Though the thesis is unavailable for consultation, a splendid full-colour plate from it has been reproduced in Priscilla P. Soucek, ‘The Arts of Calligraphy’, in The Arts of the Book in Central Asia, ed. Basil Gray (UNESCO: Serindia Publications, 1979), 21. 18. Tabbaa, ‘Qurʾånic Writing’, 122–5. 19. The relation between paper production and the expansion of literacy has not been sufficiently explored in the Islamic world. For medieval Europe, see Brian Stock, The Implications of Literacy: Written Language and Models of Interpretations in the Eleventh and Twelfth Centuries (Princeton: Princeton University Press, 1983). 20. It is perhaps in this period that the seeds of discord between scribes and Qurʾån copyists were first sown, the latter perhaps feeling threatened by the unprecedented spread of literacy. See Pedersen, Arabic Book, 43ff. 21. Grabar, Mediation of Ornament, 77. 22. According to Anthony Welch, Calligraphy in the Arts of the Muslim World (Austin: University of Texas Press, 1979), 29, ‘good writing was the indispensable tool for anyone aspiring to high governmental rank’. 23. Tabbaa, ‘Qurʾånic Writing’, 122. 24. Brinkley Messick, The Calligraphic State: Textual Domination and History in a Muslim Society (Berkeley, Los Angeles, and Oxford: University of California Press, 1993), esp. p. 3 and chs 6 and 12. 25. Ibn Muqla’s geometricisation of Arabic writing is often compared to the way God inspired the honeybees to make their cells hexagonal. Cited in Franz Rosenthal, ‘AbË Óayyån al-Taw˙ÈdÈ on Penmanship’, Ars Islamica 13/14 (1948): 9. 26. Tabbaa, ‘Qurʾånic Writing’, 141–2. 27. A. T. Welch, ‘Al-Kurʾån’, EI2, 5: 409; Arthur Jeffery (ed.), Materials for the History of the Text of the Qurʾån (Leiden: E. J. Brill, 1937), 9–10; and esp. Henri Laoust, ‘La pensée etl’action politique d’al-Måwardi (364/450–974/1058)’, Revue d’Études Islamiques 36 (1968): 64–6. 28. On the mi˙na see Henri Laoust, Les schismes dans l’Islam (Paris: Payot, 1983), 107–11; or Henri Laoust, ‘Ahmad b. Hanbal’, EI2, 1: 272–7. 29. Grabar, Mediation of Ornament, 76f.

365

PART V  GARDENS

CHAPTER SIXTEEN

The Medieval Islamic Garden: Typology and Hydraulics The study of the Islamic garden has focused on two primary areas of research: ancient palaces and gardens known primarily through excavation, and the later magnificent gardens of Persia and Mughal India.1 The whole medieval period is virtually unknown with the exception of the Alhambra palace, whose unique survival, relative accessibility and high architectural merit have made it the subject of several books and articles.2 But despite all the attention its gardens and courtyards have received, they remain severed from their immediate medieval heritage,3 and no attempt to compare them with ancient and early Islamic gardens or with later Persian ones has adequately explained their outstanding architectural qualities and iconographic associations.4 One of the main concerns of this piece, therefore, is to bridge the gap that separates the early Islamic gardens in Mesopotamia from their later medieval counterparts in Spain. For this purpose it will first look at the palaces and gardens of ninth-century Samarra and then proceed to outline their legacy in the gardens and courtyards of the tenth and eleventh centuries. Next, it will attempt to trace the continuity and evolution of these garden types in the fragmentary courtyard gardens of the petty dynasties of the twelfth and thirteenth centuries before finally discussing, if only briefly, the courtyards and gardens of the Alhambra Palace.5 Since many of these intermediary courtyard gardens are known only from their plans and waterworks, their formal investigation will be based on their typological and hydraulic aspects which, in any case, constitute the backbone of the Islamic garden. By providing for the first time a fairly complete picture of the typology and hydraulics of the medieval Islamic garden, it may be possible to raise some questions about its meaning and symbolism. Specifically, we may wish to query the concept of the paradise garden, a concept that, I feel, has been applied indiscriminately to Yasser Tabbaa (1992), ‘Typology and Hydraulics in the Medieval Islamic Garden’, in J. D. Hunt (ed.), Landscape and Garden History: Issues, Approaches, Methods, Washington, DC: Dumbarton Oaks, 303–29.

370

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

all gardens in the Islamic world.6 It is perhaps time to look critically at this widespread notion and to test it against evidence emanating from the technological, political, literary and aesthetic dimensions of the garden. Was it concern with depicting the luxuries of the Islamic Paradise or with reflecting the glory of the state and its level of technology and sophistication that stood behind the formation of the medieval Islamic garden? Since direct references to the symbolism of the garden or its component parts are wanting except on the most general level, we shall have to draw on a wide variety of sources, which may at least give us some idea as to how these gardens were perceived and experienced. On the basis of these sources, it will be argued that a certain garden aesthetic developed – based equally on ancient and Islamic religious imagery, royal prerogatives and the sensibility of the literati.7 At least two regions in the ancient Near East – Persia and Mesopotamia – seem to have provided prototypes for the Islamic garden. The most famous of these is the Persian pairidaiza (paradeisos in Greek), a walled garden with a water axis, which was known at least from the sixth century bc, as we gather from Xenophon’s famous descriptions of the gardens of the Achaemenids.8 The earliest extant example of this kind is at Pasargadae, the administrative capital of the Achaemenids, where excavations have revealed a large palace overlooking a vast garden with two pavilions, watered by stone-lined channels.9 This type is known in a handful of examples from the Sassanian period, including the ʿImårat-i Khusrau at Qasr-i Shirin built by Parviz II in the early seventh century ad. As at Pasargadae, the palace stands on a high terrace with deeply vaulted chambers in the centre of a great paradeisos encompassed by a wall.10 Whether some of these paradises were divided by orthogonally intersecting water channels in the later form of the jahår-bågh or whether these quadripartite gardens represent another garden type is not altogether clear.11 The second garden type was the ˙ayr, an extra-urban park or zoological garden, which is known to have existed in Mesopotamia and Syria from at least the fifth and sixth centuries ad.12 Its tradition continued in the Umayyad desert palaces of Syria and Jordan. Figure 16.1  Khirbat al-Mafjar, Although none of these eighth-century reconstruction drawing of fountain, palaces can be said to have had a garden, c. 735 (after R. W. Hamilton, Khirbat al-Mafjar, Oxford, 1959, fig. 56) many of them contained walled enclosures

the medieval islamic garden

with vast areas of arable land, irrigated by aqueducts and drained by means of sluice-gates.13 One of the most elaborate of these palaces, the Khirbat al-Mafjar near Jericho, contained, in addition, an ornamental fountain inside a domed pavilion, which may have been originally surrounded by a small garden (Figure 16.1).14 This situation changes abruptly at Samarra, the capital of the ʿAbbasid dynasty from 836 to 892, which must be seen as a watershed in the history of Islamic palaces and gardens. Located some sixty kilometres north of Baghdad on the eastern bank of the Tigris river, Samarra possessed numerous palaces and gardens of astounding size and luxury.15 The largest two palaces, the Jawsaq al-KhåqånÈ (c. 836) and the Balkuwåra (849–59), enclosed spacious courtyards intersected by channels and dotted with fountains and basins. Two of these fountain types are noteworthy for they occur frequently in subsequent gardens. The first, found in the private quarters of the Jawsaq al-KhåqånÈ, is of the domed or canopied fountain type, which we have come across at Khirbat al-Mafjar. The second, perhaps derived from Sassanian palaces, consists simply of two basins at slightly varying elevations connected by a thin channel. At the Balkuwåra palace, a succession of axial courtyards separated by monumental gates led up to the cruciform throne hall from the city and river side (Figure 16.2). These spacious courtyards, with a continuous water channel around their perimeters and cruciform channels meeting at a pool with an island in their midst, provide the earliest Islamic example of the jahår-bågh, a garden type that may have existed in ancient Persia. The parterres were most likely sunken below the level of the channels and their flanking paths, as was the custom in Persian gardens and as it continued to be in later Islamic gardens, especially in North Africa and Spain. The Jawsaq al-KhåqånÈ overlooked the Tigris and the distant landscape from a terrace placed at the top of a spacious staircase, while another terrace provided views of the pools and gardens of the palace itself. This interest in the commanding view was further refined at the Balkuwåra palace, where each court rose slightly from the preceding one until one reached the central throne room at the highest level. According to E. Herzfeld, ‘one standing in the central room sees towards the north-west the mighty lines of halls, the three Courts of Honor with their gateways, and the streets of the outer quadrangle; to the south-west the halls, the garden, the river, and the limitless undulating plain of the JazÈraʾ.16 In an alluvial plain with little topographical differentiation, height and the ability to see palaces, gardens and other accoutrements of civilisation must have been synonymous with power and munificence. In addition to these courtyard gardens, Samarra had at least one large garden, the so-called Óayr al-Wu˙Ësh, a zoological park built south of the city by the caliph al Mutawakkil. Only a sketchy plan, largely based on literary sources, exists of the park and shows it as

371

372

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 16.2  Samarra, Balkuwåra Palace, 849–59, plan (after K. A. C. Creswell, Early Muslim Architecture, II, Oxford, 1969, fig. 214)

a vast enclosure with a pavilion facing a pool at its southern end. A large canal traversed the park from north to south and a branch of it fed the pool, the legendary Birkat al-Mutawakkil, which once boasted animal-shaped fountains made of precious materials. A palace with an arcaded terrace overlooked the pool and the animal park from a southern, shaded vantage point. South of the palace itself extended a smaller garden that contained an artificial mound in its middle.17 Only a single fountain has survived from Samarra, the so-called Pharaoh’s Tray, originally in the courtyard of the Great Mosque of Samarra of c. 850 (Figure 16.3).18 When in operation, this large fountain (more than three metres in diameter) would have sent a single jet of water straight up in the air to fall back in the basin. The over-

the medieval islamic garden

Figure 16.3  Samarra, stone fountain, originally in the Great Mosque, 849–52, now in Khån Murjån, Baghdad (after Bargebuhr, The Alhambra, pl. 20)

flow would have fallen into a thin meandering channel around the base and escaped through a long spout, quite likely into a lower pool. The short-lived magnificence of Samarra had a significant civilisational aspect that extended far beyond its confines and reverberated in the Islamic world for several centuries. In terms of courtyard and garden, its legacy may be seen in the dispersion of the two garden types that existed in it: the quadripartite or jahår-bågh garden and the spacious ˙ayr with a freestanding pavilion. These were copied by dynasties that may have wanted to benefit from the reflected glory of Samarra, including the Spanish Umayyads at MadÈnat al-Zahråʾ in Spain and the Ghaznavids at LashkarÈ Båzår in Afghanistan.19 MadÈnat al-Zahråʾ, the tenth-century palatial city of the Cordoban Caliphate, was consciously modelled after the palaces of Samarra (Figure 16.4). ʿAbd al-Ra˙mån III, who founded it in 936, styled his kingship after the Samarran ʿAbbasid model, and MadÈnat al-Zahråʾ was meant to be an expression of his new caliphate.20 Still, his royal pleasance was far from being a mere copy of Samarra, but was rather an interesting mix of forms and ideas derived equally from local geographic conditions, local building tradition and yearning for the culture of the East.21 Compared to the palaces of Samarra, those at MadÈnat al-Zahråʾ appear less formal, more intimate, and better integrated with the landscape (Figure 16.5). Situated on a fairly steep slope ­overlooking

373

374

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 16.4  MadÈnat al-Zahråʾ, 936–81, central palace complex, plan (after G. Michell (ed.), Architecture of the Islamic World, London, 1978, p. 213; with changes by author)

Figure 16.5  MadÈnat al-Zahråʾ, view of Salon Rico (953–7) from northwest (photo: Yasser Tabbaa)

the medieval islamic garden

Figure 16.6  Spain. Madinat al-Zahraʾ, 936 and later, ‘Dar al-Jund’, from south (photo: Yasser Tabbaa)

a verdant valley, the main pavilions of MadÈnat al-Zahråʾ were arranged in three large terraces cut into the side of the mountain such that each pavilion commanded views of its own garden but also caught glimpses of the gardens below and the distant landscape. In addition, it has been suggested by Hernandez Jimenez, one of the site archaeologists, that the ramparts of the city possessed a number of mirador towers, just as at the Alhambra, a suggestion that tallies with Nuwayri’s description of MadÈnat al-Zahråʾ having such miradors overlooking gardens – ‘basåtÈn ta˙t manåΩirihi’.22 The pavilions and gardens at MadÈnat al-Zahråʾ are smaller and at first do not seem to have much to do with those at Samarra. But they conform to the two basic types seen there: the arcaded pavilion facing a body of water and the courtyard garden. The best preserved of these pavilions is the Salon Rico, a basilical hall that commands a central position on the middle terrace (Figure 16.7). Its arcaded porch looks onto an extensive walled garden divided into four quarters by paved walkways, the north arm of which is taken up by a large rectangular basin and a smaller pavilion surrounded by smaller basins (Figure 16.8). This pavilion may have been open on all four sides to the large tank and three smaller basins. This would have given it the appearance of a small island, an effect which may have existed at Samarra and which becomes part of the Islamic garden repertory.23 At the other end of the Islamic world, in modern-day Afghanistan, another palace and garden complex was built after the model of Samarra, namely, the eleventh-century LashkarÈ Båzår built by the Ghaznavids (Figure 16.9).24 Its dramatic setting above the Helmand

375

376

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 16.7  MadÈnat al-Zahråʾ, Salon Rico, from garden (photo: Yasser Tabbaa)

river and the vistas provided by its terraces and iwåns are features that are clearly derived from Samarra. These parallels and others have prompted its excavator to say that ‘the architecture of LashkarÈ Båzår seems, in more than one respect, a late and provincial representative of Abbasid architecture’.25 The external iwån in the south palace seems to have been modelled after the one at the Balkuwåra palace and, like it, may have been originally reached by a monumental staircase from the river. Typologically, the palaces of LashkarÈ Båzår are based on Samarran models but are more geometrically rigorous while at the same time displaying greater concern with fountains and gardens. Fountains and water channels abound throughout the south palace, of which the one crossing the width of the external iwån is especially interesting (Figure 16.10). It consists of a central lobate octagonal tank flanked by paved water channels. Water would have flowed from both channels into the end pool, creating an element of movement and intimacy in the iwån. Such channels become common in many Islamic courtyards, especially in North Africa. Vast enclosed gardens stretched among the three main palaces of LashkarÈ Båzår, of which the one adjacent to the central palace contained a freestanding pavilion in its centre. This was a square, coaxially symmetrical building with iwån openings on all four sides, a type that follows an independent course of development in later Persian gardens.26 A third palatial city should be briefly mentioned as it provides a link between the great gardens of the early Islamic empires and the smaller ones of the medieval petty dynasties. This is the

the medieval islamic garden

Figure 16.8  MadÈnat al-Zahråʾ, pools seen through arches of Salon Rico (photo: Yasser Tabbaa)

Figure 16.9  LashkarÈ Båzår, south palace, eleventh–twelfth century, plan (after Grabar, The Alhambra, fig. 71)

377

378

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 16.10  LashkarÈ Båzår, south palace; I: external iwån, present condition; II: eleventh-century plan of fountain (after Schlumberger, Lashkari Bazar, fig. 13)

eleventh-century Qalʾat (citadel) of banÈ Óammåd in today’s Algeria (Figure 16.11).27 The main palace in this complex consists of a large rectangular hall with a deep narthex and an arcaded porch overlooking a vast water tank. Its closest parallel is the Salon Rico at MadÈnat al-Zahråʾ, but here the palace has become firmly integrated with the water tank, which has been surrounded by a peristyle. An important discovery was made among the ruins of Qalʾat BanÈ Óammåd: a slab of gray marble (1.30 m long and 37 m wide), carved with a chevron design between two borders and containing an image of three fish in low relief at its beginning point (Figure 16.12).28 This was to serve as the inclined plane of a water chute, a so-called shådirwån, the earliest known to the author. It seems to have been discovered in association with certain carved stucco pieces that have been reconstructed by L. Golvin as a muqarnas or stalactite vault.

the medieval islamic garden

Figure 16.11  Qalʾat BanÈ Óammåd, 1015–1152, plan of main palace (after Grabar, The Alhambra, fig. 77)

On the basis of these two finds, one may reconstruct a fountain system with water emanating from a spout beneath a muqarnas vault, pouring down a shådirwån into a channel that empties into a central pool. We have, therefore, from the middle of the eleventh century the earliest datable occurrence of a salsabÈl, a fountain type in great vogue in succeeding centuries. The courtyard gardens of the twelfth and thirteenth centuries are small, scattered and poorly preserved. Typologically, they tend to continue on a miniature scale the pavilion-on-pool type and the

379

380

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 16.12  Qalʾat BanÈ Óammåd, marble slab of a shådirwån, eleventh century (drawn after Govin, Qalʾat, fig. 39)

courtyard-with-a-central-pool type. An element common to both types, however, is a water axis connecting the pavilion or the main iwån with the central pool or tank. Most commonly, this water axis takes on the form of a shådirwån and salsabÈl, a fountain type that undergoes much refinement in this period.29 Possibly the earliest preserved example of the pavilion-on-pool type with a connecting salsabÈl is, surprisingly, the Norman La Ziza palace at Palermo, one of a number of pavilions – La Cuba and La Favara were two others – built by the Norman kings of Sicily during the course of the twelfth century.30 Like the Salon Rico, La Ziza consists of a pavilion (here multi-storey), which faces a water tank containing an island with a smaller pavilion (Figure 16.13). The central hall contains a beautifully preserved salsabÈl. Water flows from a spout beneath a stone muqarnas vault, down a marble shådirwån, in and out of two small pools, and finally through a long, partly submerged channel that empties into an external tank (Figure 16.14). From Leandro Alberti, who visited La Ziza in 1526, we can add further that the island supported a square pavilion containFigure 16.13  Palermo, La Ziza, 1166– ing a vaulted room and was approached by a 85, aerial view (after Caronia, La Zisa di Palermo, fig. 144) little stone bridge.31

the medieval islamic garden

That La Ziza is largely based on Islamic – perhaps North African – models is, I believe, beyond dispute. As such, it is perfectly congruent with the general attitude of the Norman kings of Sicily who adopted many Muslim ways, including the wide use of Arabic.32 It was certainly not an isolated example; others existed in Norman Sicily, and a salsabÈl is depicted in one of the paintings of the Palatine Chapel in Palermo (Figure 16.15). The painting shows two male musicians seated on either side of a quadrilobed square pool with a jet, the water of which spouts from a lion head and flows down a shådirwån. These fountains Figure 16.14  Palermo, La Ziza, view of salsabÈl (after Caronia, La Zisa di Palermo, fig. 73)

Figure 16.15  Palermo, Capella Palatina, 1154–66, painting in the ceiling (after R. Ettinghausen, Arab Painting, Geneva, 1962, p. 48)

381

382

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 16.16  Mardin, Qasr al-Firdaws, 1239–60, plan and section of central iwån (drawing: Yasser Tabbaa)

are frequently mentioned in Sicilian Arabic poetry, of which two examples are: ‘And the lions of its shådirwån spout ­kawthar-like waters’, and ‘Its waters are like ingots of silver which melt on the steps of a shådirwån’.33 Another well-preserved example of the pavilion-on-pool type exists on the eastern outskirts of Mardin in south-eastern Turkey. This is a structure known today simply as al-Firdaws (paradise), quite probably the one remaining pavilion (jawsaq) from the cluster built by the Artuqid dynasty during the first half of the thirteenth century (Figure 16.16).34 According to Ibn Shaddåd, the choice of its location was determined by the existence of a water source (ʿAyn al-TËta) on which the present pavilion seems to be built.35 The water from this source flows right through the building and emerges from three different spouts located in the rear walls of three iwåns (Figure 16.17). The larger, central iwån contains a very elaborate salsabÈl, which directs water through no fewer than four small pools and a number of drops and chutes before leading it to a large cistern facing

the medieval islamic garden

Figure 16.17  Mardin, Qasr al-Firdaws, 1239–60, main iwån from south (photo: Yasser Tabbaa)

the pavilion. The side iwåns also contain salsabÈls – miniature versions of the central one – which also empty into the cistern at either end (Figure 16.18). The succession of pools, the use of partly open and partly covered channels, indeed the whole composition of a pavilion fronted by a large cistern with a fountain connecting the two brings to mind the palace of La Ziza. In both, the water element gives a sense of motion and helps to connect interior and exterior spaces. The courtyard-with-salsabÈl type is represented by a few bourgeois residences in Fus†å† and a handful of palaces in Syria and southern Turkey. In Fus†å† at least two of the excavated houses had complete salsabÈls fed through underground channels that began at the well of the house (Figure 16.20). Each contained the succession of spout, shådirwån, short canal, small pool, long canal and central pool. These houses, whose plan is derived from the Samarran house type, are datable no earlier than the late eleventh century.36 An early thirteenth-century Artuqid palace in the citadel of Diyarbakir contains the remains of a salsabÈl with a high shådirwån.37 The channel flows down the main (southern) iwån of a cruciform palace and empties into a square pool with chamfered edges, which contains a small island with a fountain in its centre (Figure 16.21).

383

384

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 16.18  Mardin, Qasr al-Firdaws, view of cistern (photo:Yasser Tabbaa)

Figure 16.19  Fus†å† (old Cairo), House VI, eleventh–twelfth century, plan (after Creswell, Muslim Architecture of Egypt, I, fig. 61)

Both salsabÈl and pool were lavishly decorated with glass and stone mosaics and tiles, late survivals of an ancient technique (Figure 16.22). Remains of a highly original fountain exist in the late twelfth-century Ayyubid palace at Qalʾat ÍahyËn in the ʿAlawite mountains of north-western Syria (Figure 16.23).38 A central pool with a ring encircling it and four channels, each leading to one of the corner chambers flanking the iwån, was carved out of the actual bedrock (Figure 16.24). The natural slope of the site was utilised to run water down the northern chambers, into the pool (where a fountain may once have existed), and out of the southern chambers. Although somewhat crude and poorly preserved, this courtyard bears a general similarity to that in the Court of the Lions at the Alhambra palace. Of the many gardens known to have existed in Spain from the twelfth and thirteenth centuries only a few survive, in a ruinous state. Two of these have been uncovered beneath the present gardens of Alcazar in Seville, and a

the medieval islamic garden

Figure 16.20  Diyarbakir, palace in the citadel, early thirteenth century, plan (after Aslanapa, ‘Ausgrabungen’, fig. 1)

Figure 16.21  Diyarbakir, palace in the citadel, detail of pool (after Aslanapa, ‘Ausgrabungen’, fig. 4)

third, the Castillejo Monteagudo near Murcia, was excavated and later destroyed.39 The best preserved, identified by its excavator as the patio of al-Muʾtamid’s famous palace, Qaßr al-Mubårak, is a rectangular courtyard garden with ­cruciform water channels converging on a circular central pool (Figure 16.25). The parterres are five feet below the channels, deep enough to have their surrounding walls shaped into a continuous arcade. Someone walking on the path or

385

386

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 16.22  Qalʾat ÍahyËn, Ayyubid palace, late twelfth century (drawing: R. Brotherton, redrawn with additions by Yasser Tabbaa)

Figure 16.23  Qalʾat ÍahyËn, Ayyubid palace, courtyard (photo: Yasser Tabbaa)

the medieval islamic garden

Figure 16.24  Seville, garden in the patio of Qasr al-Mubårak, twelfth century. Drawing: author after Ettinghausen and MacDougall (eds), The Islamic Garden, fig. xvii/4)

seated on an elevated pavilion at one end would have been above the level of the bushes and possibly even the fruit trees.40 The garden would have appeared as an exquisite object, a carpet for example, which could be appreciated from above in one look. Although this garden is a bit exaggerated, it does not totally depart from a dominant aesthetic trend in medieval Islamic gardens, namely, a preference for the small and exquisite over the large and ponderous. Ibn LËyËn, the fourteenth-century Grenadine agronomist and man of letters, stressed that a garden should not be so large as to become a colossal eyesore, but must be proportioned to produce maximum delight in one view.41 Of course, the ability to appreciate a garden in one view was also predicated on the existence of an elevated vantage point. This was provided in the earliest gardens by the use of terracing, but special viewing towers, manΩaras, were effectively used in medieval and post-medieval gardens. References to lofty pavilions and elevated sites with commanding views are common both in twelfth-century Persia and in Spain, where we can turn again to Ibn LËyËn: ‘With regard to houses set among gardens, an elevated site is to be recommended, both for vigilance and layout . . . In the center of the garden let there be a pavilion in which to sit, with vistas on all sides.’42 Towers were also frequently used to provide vistas of the surrounding landscape. They seem to have been equally at home in the gardens of MadÈnat al-Zahråʾ as in those built in eleventh-century Isfahan by the Great Seljuqs. According to R. Pinder-Wilson, all these gardens possessed lofty pavilions (qaßr) with a high tower (ßar˙) commanding a fine view over the garden, which led him to conclude that in Saljuq gardens ‘the pavilion with a commanding prospect is a prominent feature’.43

387

388

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 16.25  Granada, Alhambra, fourteenth century, exterior towers (photo: Yasser Tabbaa)

Of course, the best-preserved garden towers from the Islamic Middle Ages are those that surround the enclosure of the Alhambra palace (Figure 16.26). The majority of these towers were not intended for defensive purposes, as one can tell from their location, external form and excessive interior ornamentation. At least two of these towers, those of the Captive and the Infantas, were spacious enough to accommodate a small group of diners perched above the hills of Granada.44 A third feature of medieval Islamic gardens was the interpenetration of house, water and garden. Every attempt was made to enhance the sense of being part of the garden without actually strolling in it. One is reminded of the often-quoted remarks of John Chardin, the seventeenth-century French traveller: ‘The Persians don’t walk so much in the garden as we do, but content themselves with a bare prospect . . . For this end they set themselves down in some part of the garden at their first coming into it, and never move from their seats till they are going out of it.’45 Of course, the openness of the pavilions facilitated this detached interaction; these were always totally open on one side through a colonnaded porch, a talår, or an iwån. In Persia they were often open on all four sides. The fourth and perhaps most important feature of medieval and later Islamic gardens is the celebration of water. A great deal has been written about the symbolism of water in the ancient Near East and Islam, particularly as an image of fertility and of Paradise. Indeed, the image of Paradise as a garden beneath which rivers flow

the medieval islamic garden

occurs no fewer than twenty-five times in the Qurʾån. The confluence of these ancient and Islamic ideas has led to the widespread and uncritical acceptance of the paradisiacal symbolism of water, especially in the form of single or intersecting channels: the rivers of Paradise. Although one need not totally reject the notion of Paradise in Islamic gardens, one can easily question its validity as a factor in the actual design of gardens and their water systems and the reaction that these forms elicited. Gardens existed in those areas of the Islamic world that already had an advanced system of water storage and irrigation.46 The collection and transportation of this precious resource demanded great effort and expenditure, and our immediate reaction upon seeing such projects as the Aghlabid cisterns in Tunisia or any of the thousands of qanåts (underground water conduits) in Iran is one of awe at the power of the state that created them and the knowhow of the craftsmen who designed and built them. Similarly, the manipulation of water into regularly intersecting channels, elaborate fountains, gushing water chutes and still pools was partly a statement of the power of the state and a reflection of its technological achievements.47 It follows, then, that the water mechanisms of the Islamic garden were on a certain level quite similar to the armature of the Graeco-Roman city: each civilisation chose to celebrate what it considered to be its greatest accomplishment.48 Water in the garden also reflected the level of technology that made it possible in the first place. This technology could be as simple as laying out the garden or pavilion near an elevated source of water and simply controlling the natural flow of the water down the slope. But at least two water-raising mechanisms were used throughout the Islamic and Iberian worlds. The first was the nå’Ëra or noria, a water wheel with hinged compartments that dip into the stream and carry the water to the top of the wheel where it is discharged into an aqueduct. Because norias relied for their motion on the push of water, they could only be used in fairly rapid streams, which, in the Islamic world, restricted their usage to Syria, parts of North Africa and Spain. Norias were used in the Roman world and described by Vitruvius, but only became commonplace during the Islamic Middle Ages.49 Although they were primarily used for agriculture – as in Óamåt (Syria), which has some exceptionally large ones – they seem to have been used in gardens and palaces in Spain. Some were illustrated in the thirteenth-century Spanish manuscript Bayå∂ wa Riyå∂ (Figure 16.27), and even poems were written about them, often comparing their melancholy groan to the wailing of parted lovers.50 Far more practical for the arid climate of most of the Islamic world was the såqiya, a water-elevating machine that uses animal power. The central mechanism of the såqiya consists of two gears meshing at right angles (Figure 16.28). The vertical cogwheel is mounted on

389

390

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 16.26  Story of Bayå∂ and Riyå∂, Bayå∂ lying unconscious by a noria, Spain, thirteenth century, ms. ar. 368, fol. 19 r, Biblioteca Apostolica Vaticana

an axle over the source of water, usually a well. On the other side of the cogwheel from the teeth is a drum carrying a chain-of-pots or ‘potgarland’ – the pegs for carrying the potgarland are carried on a separate wheel on the same horizontal axis as the cogwheel. A drawbar is pulled by an animal tethered to it. Water is raised to the surface and discharged into a head tank.51 A. Watson suggests that ‘the machine was probably transmitted to Spain from Syria, when the Muslims introduced their irrigation methods to Spain’, and indeed Ibn al-Baßßål, the eleventh-century agronomist, described it as a standard machine for irrigation.52 Although these practical machines were in common use for agricultural purposes, other more exotic and visionary mechanisms were used specifically in palatial gardens. These were the automata, known today primarily from the eleventh-century treatise of BanË MËsa and the 1206 treatise of BanË al-Jazari.53 Al-JazarÈʾs treatise is especially noteworthy for our purposes because it includes a whole chapter entitled ‘On the construction in pools of fountains which change their shape, and of machines for the perpetual flute’. Six types of water-raising mechanisms are illustrated and discussed, some practical, others for amusement. A description of the functioning of these

the medieval islamic garden

391

mechanisms is outside the scope of this paper, but something about their aesthetics and effect may be derived by a brief examination of just one of them (Figure 16.29). In al-JazarÈʾs third device, water from a nearby lake filled a tank whose bottom contained a drain that discharged a steady stream of water over a mechanism concealed in a lower basin. The water fell over a wheel with scoops, setting it in motion. Through a system of two sets of meshing gears, the motion was transferred first horizontally then vertically to a potgarland that dipped into the top tank and carried the water to a channel, which emptied into the lake as a cascade. Since the lower mechanism was totally hidden from sight, all that the viewer could see was a såqiya, which appeared to be propelled by the wooden model of a cow. R. D. Hill concludes that Figure 16.27  Drawing of a såqiya in Spain ‘the machine was intended as a (after D. R. Hill, A History of Engineering in decorative lakeside attraction with Classical and Medieval Times, London, 1984, fig. 8.4) an element of mystification about it’.54 Thus, in addition to their other practical applications, these devices were primarily intended to produce amazement in the viewer. In Arabic these automata fell under the heading of ʿilm al-˙iyal, literally the science of ruses or tricks but probably implying something about the ingenious or miraculous nature of the device. S. H. Nasr suggested that this branch of science ‘has always been related in the Muslim mind with the occult sciences and magic’,55 but the mechanisms in al-JazarÈ are probably closer to the idea of ʿajå’ib or wondrous creations. In this respect, they are in keeping with other wondrous elements in the Islamic garden, including manmade trees,56 column fountains, animal fountains and other artificial devices. In all these fountains wonder is produced by artfully concealing the source of water and its passage through the fountain. The fifteenth-century Egyptian historian al-MaqrÈzÈ recounted the description of a ninth-century garden with ‘palm trees whose trunks were covered with gilt copper; behind this covering were lead pipes which brought water up the side of the trees and sprayed it out from various openings into pools’.57 Quite possibly, these ‘tree fountains’

392

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 16.28  Water-elevating mechanism, al-JazarÈ, Maʾrifat al-Hiyal, 1354, Suleymaniye Library, Istanbul, 3606 (photo: courtesy of Suleymaniye Library)

provided the original idea for the stylised column fountains that became popular in the twelfth century and later.58 One of the most curious garden devices was a glass pavilion erected by Ya˙ya ibn IsmåʾÈl al-MaʾmËn, king of Toledo (1043–75). According to the contemporary historian al-MaqqarÈ: He constructed in the middle [of his palace area] a lake, in the centre of which lake he built a pavilion (qubba) of stained glass, and encrusted with gold. The water was caused to rise to the top of the pavilion by an artful device invented by his engineers, so that the water would descend from the summit of the pavilion, encompassing it, the various streams uniting themselves with one another. In this fashion, the glass pavilion was within a sheet of water which was shed across the glass and which was flowing incessantly while al-MaʾmËn sat within the pavilion without being touched by the water; and even torches could be lighted in it, producing, thereby, an astonishing and marvellous spectacle.59

the medieval islamic garden

Figure 16.29  Granada, Alhambra, lion fountain at the pool of the Partal, thirteenth century (photo: Yasser Tabbaa)

Zoomorphic fountains representing a variety of animals are mentioned at Samarra and elsewhere, but the lion was especially favoured in the later Middle Ages (Figure 16.30).60 The exotic effects of these animals were often praised by the poets of Sicily and Andalusia. Ibn ÓamdÈs, the eleventh-century Sicilian poet, described a pool of water with lion-shaped spouts: As to the lions that inhabited a jungle lair of rulership, they abandoned their roaring for the murmuring of water. It is as if gold enveloped their bodies And crystal melted in their mouths. You might imagine, as the sun brings out their color, that they are fire and their licking tongues light.61 It will seem quite evident that our attempt to underline the aesthetic qualities of the medieval Islamic garden was to no small degree based on poetry. This was not accidental. Garden poetry had already developed in the ninth-century courts of Samarra into a major poetic form, raw∂iyyåt, and spread from there to Sicily and Spain where it reached its peak of accomplishment. In the process, the raw∂iyyåt gradually shed the pretensions of court panegyric and embraced the less formal and more personal tradition of the literati. These generally short poems delighted in describing the sensuous qualities of the garden: the lushness of its foliage, the coolness of its shade and

393

394

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

the heaviness of its perfume.62 More than anything, however, they dwelt on water: running freely in a meadow, generously overflowing a pond, rising in a jet like a rod of rock crystal, or flowing in a channel like melted silver.63 The poetical inscription around the basin of the Fountain of the Lions at the Alhambra palace is quite typical in this regard: Silver melting which flows between jewels, one   like the other in beauty, white in purity A running stream evokes the illusion of a solid substance   for the eyes, so that we wonder which one is fluid Don’t you see that it is the water which is running over the rim   of the fountain, whereas it is the monument which offers    long channels for the water.64 Some of these poems were written for the court, but many more seem to have been personal reflections, usually occasioned by an amorous picnic or a gathering of poets and other literati. These gatherings often assumed the character of poetic games in which the participants would compete against each other in poeticising about a favourite garden subject. Al-GhuzËlÈ described several such assemblies: ‘I met with a group of Alexandrian literati in a garden . . . with a pool. . . . One of those present scattered jasmine petals in the pool . . . and we all dropped our heads to stimulate and emote our minds, then we produced what we had written. I said: “They scattered jasmine on the surface of water/so that we imagined stars in the middle of the water.”’65 It is not quite certain to what extent garden poetry and the involvement of the literati in the garden experience influenced garden design. It seems clear, however, that the raw∂iyyåt served as a literary parallel to the garden, highlighting its images and mechanisms and connecting them with current literary motifs and ancient myths. A dialectical process thus formed between gardens and garden poetry, such that as the perception of gardens came to be seen more and more in poetic terms, the garden itself may have aspired to these literary ideals. More work on this intriguing connection needs to be done. The courtyards and gardens of the Alhambra palace stand at the end of a continuum that, as suggested above, began in Samarra. Typologically speaking, they still conform to the main types established above: the Court of the Myrtles and the Partal belong to the pavilion-on-pool type, while the Court of the Lions represents a variation on the courtyard-with-a-central-pool type (Figure 16.31). Indeed, all the different features of the medieval Islamic garden – miniaturisation, the interpenetration of architecture and garden, the dominant view, the use of exotic effects and the manifold qualities of water – lead to the Alhambra, especially the Court of the Lions,

the medieval islamic garden

Figure 16.30  Granada, Alhambra, plan of main units (after G. Michell, Architecture of the Islamic World, p. 213)

which seems to summarise several centuries of artistic development (Figure 16.33). There is no confusing beauty with size here: the Court measures 28.50 by 15.70 metres, but is designed in plan and elevation according to a sophisticated system of proportion.66 Second, the use of the peristyle, the projecting pavilions and the water channels help to unify interior and exterior spaces, thereby perfecting an ideal that had occupied Muslim architects for several centuries.67 Third, with its well-defined axes and originally sunken parterres (about 80 cm below the level of the water channels) the whole design of the courtyard would have been comprehensible in one look, particularly from the second-floor rooms. In fact, it has been graphically demonstrated by Pietro-Moreno that the room located above the antechamber of the Hall of the Two Sisters provides uninterrupted views, not

395

396

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 16.31  Granada. The Alhambra Palace: Courtyard of the Myrtle, 1333–54, from south (photo: Yasser Tabbaa)

Figure 16.32  Granada, Alhambra, Court of the Lions, 1354–9 (photo: Yasser Tabbaa)

just of the Fountain of the Lions, but also of the Garden of the Daraxa and the tamed landscape beyond. A lower manΩara, the Mirador de la Daraxa, directly overlooks the Garden of the Daraxa and provides a much-needed release from the complexity and intensity of the Court of the Lions.68 The Fountain of the Lions stands in the middle of the courtyard as the architectural and iconographic focus of the entire courtyard (Figure 16.34). Leaving aside its Solomonic associations (which have been discussed thoroughly by F. P. Bargebuhr),69 it is first and foremost an image of great wonder (ʿajÈba), following in a tradition that goes back to ninth-century Samarra, if not earlier. It is also a representation of the power of the patron and the culmination of a great hydraulic feat that began on the hills east of the Alhambra.70 Finally, its cruciform channels, whose waters flow beneath exquisite pavilions, reflect an image of paradise, which is itself derived

the medieval islamic garden

Figure 16.33  Granada, Alhambra, Fountain of the Lions, eleventh and fourteenth centuries (photo: Yasser Tabbaa)

from the confluence of Qurʾånic evocations and ancient Near Eastern archetypes.71 These multi-layered associations, combined with the magnificent appearance of the lions and the sound of water gushing from their mouths, totally dominate the Court, which, conversely, may be seen as a vehicle for the celebration of the Fountain. The greatness of the Court of the Lions does not reside in its absolute originality but in the complexity of its formal and iconographic associations and in the manner that these diverse ideas were brought together in an exquisite ensemble. In these respects the art of garden design was wholly consonant with the art of later medieval Arabic poetry, especially its Andalusian variety, whose excellence was not measured by its startling originality but in ‘the ingenious way in which the poet has rephrased traditional themes and motifs’.72 Fully mature and deeply conscious of their long and rich traditions, neither genre was particularly concerned with stepping outside its boundaries but perfectly content to further refine and interpret the forms and ideas of the past. Subtlety replaces stridency; embellishment supersedes novelty; and meditation rises above aggrandisement as the artist searches for the most elegant expression of noble ideas. A comparison of early and late medieval Islamic courtyards and gardens reveals the evolution of Islamic garden design and iconography and underlines the importance of the intermediary gardens of the twelfth and thirteenth centuries. Moving from the oversize gardens of Samarra and LashkarÈ Båzår to the intimate ones at the Alhambra palace, the preceding discussion has noted a simultaneous evolution of some features and the abandonment of others. Bizarre

397

398

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

forms and spectacles such as ponds made of tin, trees with goldsheathed trunks and artificial singing birds – what R. Ettinghausen aptly described as ‘one-upmanship on nature’73 – are either totally dismissed or transformed in a way that would integrate them within the overall unity of the garden or courtyard. Other features, having to do primarily with hydraulics and the perfection of the courtyard garden, are greatly developed as to become the essential qualities of the western Islamic garden.74 Notes 1. A number of general but somewhat problematic works exist: J. Lehrman, Earthly Paradise, Garden and Courtyard in Islam, Berkeley–Los Angeles, 1980; and more recently J. Brooks, Gardens of Paradise: The History and Design of the Great Islamic Gardens, London, 1987. Focused and more scholarly are D. N. Wilber, Persian Gardens and Garden Pavilions, 2nd edn, Dumbarton Oaks, Washington, DC, 1979; and R. Ettinghausen and E. B. MacDougall (eds), The Islamic Garden, Dumbarton Oaks Colloquium on the History of Landscape Architecture 4, Washington, DC, 1976. The best works to date on later Islamic gardens are E. B. Moynihan, Paradise as a Garden in Persia and Mughal India, New York, 1979; and S. Crowe et al., The Gardens of Mughal India, London, 1972. J. Wescoat has re-evaluated the Mughal contribution to landscape design from the perspective of the geographer. See his ‘The Islamic Garden: Issues for Landscape Research’, Environmental Design 1 (1986), 10–19. 2. For bibliography, see O. Grabar, The Alhambra, Cambridge, MA, 1978. 3. To my knowledge the only article that has attempted to deal with these fragmentary gardens, and then only for Spain, is J. Dickie, ‘The Islamic Garden in Spain’, in Ettinghausen and MacDougall, The Islamic Garden, 87–106. 4. This would include Grabar in The Alhambra, 103 ff, where the formal parallels for the courtyards of the palace reach at one end into the Roman villa tradition and at the other into the pavilions of Safavid Isfahan but largely miss the medieval palace tradition. A similar objection has already been voiced by J. Dickie in his excellent review of Grabar’s book, ‘The Alhambra: Some Reflections Prompted by a Recent Study by Oleg Grabar’, Art and Archeology Research Papers 16 (1979), 53–66. 5. Since Iran has been inaccessible for quite some time now and since its gardens have been discussed by Wilber and others, I have chosen to leave out Timurid gardens, focusing instead on the gardens and courtyards of the central and western Islamic world. Timurid gardens have been discussed also by R. Pinder-Wilson, ‘The Persian Garden: Bagh and Chahar Bagh’, in The Islamic Garden, 69–86. 6. See e.g. Lehrman, Earthly Paradise, 31–4. T. Burckhardt may have been the first to insist on the close association between Islamic garden and paradise. See, for example, his Moorish Culture in Spain, tr. A. Jaffa, London, 1972, 208f., where he says: ‘every such garden [walled garden] is in Islam an image of paradise.’ It should be noted that D. Fairchild Ruggles, ‘A Mythology of an Agrarian Ideal’, Environmental Design 1

the medieval islamic garden

(1986), 24–7, has presented a summary of her criticism of this concept, the full text of which will hopefully appear soon in her doctoral dissertation on MadÈnat al-Zahråʾ for the University of Pennsylvania. 7. Although the literati did not exist as a class as they did in China or have the same intellectual clout, they were nevertheless quite influential in questions of taste, proper conduct and literature. 8. See especially Wilber, Persian Gardens, 3–21; and Pinder-Wilson, ‘The Persian Garden’, 71–2. 9. Ibid., and D. Stronach, ‘Excavations at Pasargadae: Third Preliminary Report’, Iran, Journal of the British Institute of Persian Studies 3 (1965), 31 f. 10. Pinder-Wilson, 72–4. 11. Ibid., 79–81. Pinder-Wilson states that the term chahar bagh is first used in the twelfth century but does not become common for denoting gardens until the Timurid period (fifteenth century). He further suggests that ‘the word chahar bagh may ultimately derive from the Soghdian s’r’b’gh “tower”’ and adds that ‘if this derivation is correct, then chahar bagh was the popular etymology of a word of which the original meaning had been lost’. 12. Since none of these ˙ayrs is preserved (only known in poetry and chronicles), it is not possible to see how they may have differed from the Persian game reserves. According to J. Sourdel-Thomine, ‘Óayr’, Encyclopaedia of Islam, 2nd edn, III (1971), 71, some of the ˙ayrs of Syria and Mesopotamia were used not just for hunting and sport but also for agriculture. It is not clear, however, whether her observation is based on actual findings in these pre-Islamic ˙ayrs or on the backward projection of evidence found in Umayyad ˙ayrs. 13. See. O. Grabar, The Formation of Islamic Art, New Haven–London, 1973, 42–3, 141–5, for observations on the agricultural aspect of Umayyad palaces and bibliography on the subject. See also, V. Strika, ‘The Umayyad Garden: Its Origin and Development’, Environmental Design 1 (1986), 72–5. 14. R. W. Hamilton, Khirbat al-Mafjar, Oxford, 1959, 5–7, 110–20. 15. The palaces and gardens of Samarra are known today through their often hyperbolic descriptions in contemporary court poetry and also thanks to the archaeological efforts of H. Viollet, E. Herzfeld and A. Sousa in the first half of this century. The works of these authors are very inaccessible, but J. M. Rogers, ‘Såmarra, A Study in Medieval Town Planning’, in A. H. Hourani and S. M. Stern, The Islamic City, Oxford, 1970, 119–56, contains the most pertinent bibliography and an excellent summation of the views expressed in these works. The literary evidence concerning Samarra has been compiled by Y. al-SåmarråÈ, in Såmarrå fÈ adab al-qarn al-thålith, Baghdad, 1963, and in TårÈkh Såmarrå, 2 vols, Baghdad, 1968. 16. Quoted in K. A. C. Creswell, Early Muslim Architecture II, Oxford, 1940, 269. 17. Ahmad Sousa (in Arabic), Baghdad, 1948, 298 and 306–13. 18. This fountain is mentioned in al-MuqaddasÈ (d. 895), A˙san al-TaqåsÈm fÈ Ma’rifat al-AqålÈm, ed. H. de Goeje, Leiden, 1906, 122. 19. Samarra had a much more direct influence on the palace and garden begun by Ahmad ibn TËlËn (868–84) and enlarged by his son Khumåruwayh outside Fus†å† (old Cairo). Before being appointed as governor of Egypt, Ibn TËlËn was a soldier at Samarra, and his architecture in Egypt bears

399

400

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

the imprint of the Samarran style. Nothing survives of the palace or Khumåruwayh’s magnificent garden, but the fifteenth-century Egyptian historian al-MaqrÈzÈ has left us a lengthy description of it: Khi†a† alMaqrÈzÈ, Bulaq, 1854, II, 95–7. The description focuses on the exotic elements of the garden. 20. See, in particular, A. Chejne, Muslim Spain, Its History and Culture, Minneapolis, 1974, 34–5 and 366–8. The author remarks (p. 34) that ‘the Umayyads up to 929 had abstained from assuming the title of the caliphate’, despite the decline of the Baghdad caliphate. He attributes ʿAbd al-Rahmån’s decision in 929 to adopt the title of caliph as a reaction to the threat of the heterodox Fatimids. 21. The literature on MadÈnat al-Zahrå’ is still scanty and scattered despite more than three decades of archaeological activity. The posthumous publication by F. Hernández Giménez, MadÈnat al-Zahrå’: Arquitectura y Decoración, Granada, 1985, is seriously impaired by the virtual absence of photographs and original drawings. 22. This observation is based on D. F. Ruggles’ unpublished paper ‘MadÈnat al-Zahrå’: Abbasid Vistas of Umayyad Landscape’, delivered at the Middle East Studies Association meeting in Los Angeles, 1988. I am grateful to the author for allowing me to refer to her unpublished work. Al-NuwayrÈ, Nihåyat al-arab, Cairo, 1980, XXIII, 398. 23. J. Dickie, ‘The Islamic Garden in Spain’, in Ettinghausen and MacDougall, The Islamic Garden, 96. 24. The primary work on the history and architecture of the palaces at Lashkari Båzår is D. Schlumberger et al., Laskhari Bazar: Une résidence royale ghaznévide et ghoride, 2 vols, Paris, 1963–78. 25. D. Schlumberger, ‘Le palais ghaznévide de Lashkari Bazar’, Syria, 29(1952), 268. For the connection between Ghaznavids and ʿAbbasids, see in particular C. E. Bosworth, The Ghaznavids: Their Empire in Afghanistan and Eastern Iran, 994–1040, Beirut, 1973, 51–4 and 137–8. 26. Schlumberger, Laskhari Bazar, fig. xliii. Pavilions with cross-axial symmetry reach a high degree of refinement in the Timurid and Safavid period, culminating in such exquisite examples as the seventeenth-century Hasht Behisht pavilion in Isfahan. For their typology in the Timurid period, see D. Wilber and L. Golombek, The Timurid Architecture of Iran and Turan, Princeton, 1988. 27. The main source on this important site is L. Golvin, Recherches archéologiques à la Qal ‘at des BanÈ Hammåd, Paris, 1965. 28. Ibid., 122–7. See also G. Marçais, ‘SalsabÈl et Sådirwån’, in Études d’Orientalisme dédiées à la mémoire de Lévi-Provençal, II, Paris, 1962, 639–48. 29. See my ‘The Salsabil and Shadirvån in Medieval Islamic Courtyards’, Environmental Design 1 (1986), 34–7, and ‘Toward an Interpretation of the Use of Water in Islamic Courtyards and Courtyard Gardens’, Journal of Garden History 7 (1988), 197–220. 30. See G. Caronia, La Zisa di Palermo: Storia e restauro, Rome, 1982, 24 and fig. 24. Caronia presents the most detailed account of the history, architecture and restoration of this palace. See also G. Bellafiore, La Zisa di Palermo, Palermo, 1977. 31. Alberti’s account is quoted in part in N. Miller, ‘Paradise Regained: Medieval Garden Fountains’, in E. B. MacDougall (ed.), Medieval Gardens, Dumbarton Oaks Colloquium on the History of Landscape Architecture 9, Washington, DC, 1986, 142–3. Caronia, La Zisa, ­contains

the medieval islamic garden

a watercolour reconstruction (fig. 143) of the Ziza complete with island pavilion and bridge. 32. This is evidenced by the number of Arabic inscriptions found from Norman Sicily and by Ibn Jubayr’s account (Ri˙lat Ibn Jubayr, Beirut, 1980, 297–300) that the rulers of Sicily adopted many Muslim ways. 33. To my knowledge, the Arabic poetry of Sicily has not been collected in an anthology. For these two lines, see ʿImåd al-DÈn al-IßfahånÈ, KharÈdat al-Qaßr wa-JarÈdat al-’Aßr, I, M. al-MarzËqÈ, ed., Tunis, 1966, 23; and DÈwån Ibn HamdÈs, I˙sån ‘Abbås, ed., Beirut, 1960, 494–7. 34. I have identified this building in ‘The Use of Water’, 209 and 219–20. 35. Ibn Shaddåd, Al-A’låq al-Kha†Èra fÈ Dhikr Umarå’ al-Shåm wa’l-JazÈra, III (2 parts), ed. Y. ʿAbbåra, Damascus, 1978, 543–4 says: ‘The banË Artuq built a wall around a plain area in the eastern vicinity of the city. Next, al-Malik al-Sa’Èd, the owner, made in it jawåsiq (pavilions) and gardens near a spring called ‘Ayn al-TËta. This place was called al-Firdows.’ 36. K. A. C. Creswell, Muslim Architecture of Egypt, I, Oxford, 1952, 121–30. 37. This palace has been excavated by O. Aslanapa, ‘Erster Bericht uber die Ausgrabungen des Palastes von Diyarbakir’, Istanbuler Mitteilungen 12 (1962), 115–28. The site is currently an off-limits military prison. 38. M. van Berchem and E. Fatio, Voyage en Syrie, 2 vols, Mémoire, Institut Français d’Archéologie Orientale 37–8 (1914–15), 267–83, pls 59–62; Paul Deschamps, ‘Le chateau de Saone et ses premiers seigneurs’, Syria 16 (1935), 73–88. 39. Dickie, ‘The Islamic Garden in Spain’, 96–8, figs 4–8. 40. Ibid., 97. According to Dickie, nothing has been published on the Seville excavations. 41. Quoted and commented on in G. Marçais, ‘Les jardins de l’Islam’, in Mélanges d’Histoire et d’Archéologie de l’Occident Musulman, Tome 1, Articles et Conférences de Georges Marçais, Algiers, 1957, 238–9. There is, of course, little doubt that the preference for smaller gardens was also related to climatic factors. 42. Ibid., and for an English translation, see Dickie, ‘The Islamic Garden in Spain’, 94. 43. Pinder-Wilson, ‘The Persian Garden’, 76. 44. Grabar, The Alhambra, 91–2. Towers become an even more important feature in the later gardens of Turkey and Mughal India. They are often illustrated in Ma†råkçÈ as octagonal towers with a viewing balcony: see NaßË˙u’s-silå˙È Ma†råkçÈ, Beyån-i Menåzil-i Sefer-i ‘Iråken-i Sultån Suleymån Hån, ed. H. G. Yurdayin, Ankara, 1976, pl. 90a. The same tower form still exists in Nishat Bagh in Srinagar, Kashmir, similarly situated at the edge separating the terraced garden from the valley below. 45. E. Lloyd (tr.), Sir John Chardin’s Travels in Persia, London, 1927; quoted in Pinder-Wilson, ‘The Persian Garden’, 72. 46. A. Watson, Agricultural Innovation in the Early Islamic World: The Diffusion of Crops and Farming Techniques, 700–1100, Cambridge, MA, 1973, 117–19; and Ruggles, ‘Agrarian Ideal’, 26–7. 47. These points were not lost on state chroniclers and court painters who wrote about and illustrated many watered gardens in Mughal India. 48. For the significance of the armature in Roman city planning, see W. L. MacDonald, The Architecture of the Roman Empire, II. An Urban

401

402

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Appraisal, New Haven–London, 1986, 29–32. While our analogy may sound forced, it clearly describes the situation in the seventeenth-­ century Jahår Bågh avenues of Isfahan and Delhi, straight avenues that run along both sides of a wide water channel. See Wilber, Persian Gardens, 48–50. 49. A. Y. Hassan and D. R. Hill, Islamic Technology, an Illustrated History, Cambridge, MA, 1986, 38–40. 50. See the anthology of al-GhuzËlÈ, Ma†åli’ al-budËr fÈ manåzil al-surËr, Cairo, 1905, which contains a chapter (1, 35–45) on fountains and water wheels. More accessible is J. A. Bellamy and P. O. Steiner (tr.), Ibn Sa’Èd al-MaghribÈ, The Banners of Champions: An Anthology of Medieval Arabic Poetry from Andalusia and Beyond, Madison, WI, 1989, esp. 115–17. 51. Hassan and Hill, Islamic Technology, 40–2. 52. Ibid., 40. 53. Both treatises have been translated and edited by D. R. Hill: The Book of Ingenious Devices (Kitåb al-Óiyal), Dordrecht–Boston–London, 1974; and The Book of the Knowledge of Ingenious Geometry (Mechanical Devices), Hingham, MA, 1974. See also N. Miller’s perceptive remarks about the possible impact of these treatises on fountain design in late medieval Europe, ‘Medieval Garden Fountains’, 142–4. 54. Hassan and Hill, Islamic Technology, 44. 55. S. H. Nasr, Islamic Science, an Illustrated Study, London, 1976, 145. 56. Wilber, in Persian Gardens, 8–9, has collected a number of references to these artificial trees. 57. Al-MaqrÈzÈ, 96. See also R. Ettinghausen’s introduction to The Islamic Garden, 4, where he mentions another early Islamic garden characterised by its curious devices. 58. Perhaps the earliest of these is the twelfth-century fountain located in the south-western corner of the cloister at the Norman cathedral at Monreale, a fountain whose chevron-carved shaft may be a stylisation of a palm tree trunk. In this fountain the water rises through a pipe embedded in the column and shoots out of holes in the anthropomorphic head, which is probably a later addition. See W. Krönig, The Cathedral of Monreale and Norman Architecture in Sicily, Palermo, 1965, pl. 86. Related fountains exist at the Alhambra palace and the Topkapı Saray in Istanbul. 59. This passage is quoted in F. P. Bargebuhr, The Alhambra, a Cycle of Studies on the Eleventh Century in Moorish Spain, Berlin, 1968, 144. Such artifice, according to the author, seems to have been quite common in the medieval gardens of Spain. 60. See B. P. Maldonaldo, Estudios sobra la Alhambra, II, Granada, 1977, 47–9, for a short list of such lion fountains. 61. GhuzËlÈ, Matåli’, 37. 62. Watson, Agricultural Innovation, 83. 63. See Dickie, ‘Islamic Garden in Spain’, 89: and Tabbaa, ‘Use of Water’, for further references to this poetic genre. Cf. Ruggles, ‘Agrarian Ideal’, 24, where the writer generally minimises the role of poetry in the interpretation of gardens. 64. Grabar, The Alhambra, 124. 65. Al-GhuzËlÈ, Matåli’, 38–42. 66. Marçais suggested a rhythmic proportional system for the spacing of the columns of the peristyle. Much more of this kind of geometric analysis

the medieval islamic garden

could be done to the courtyard, as has been suggested by Grabar in The Alhambra, 182–5. 67. See the many examples preceding and Grabar, The Alhambra, 165–6. 68. Reproduced in Dickie, ‘Some Reflections’, 56, 64. See also J. Dickie, ‘Toward an Aesthetic of Granadine Art’, Oriental Art 26, 3 (Autumn 1980), 327, where the author discusses the function of the mirador. 69. Bargebuhr, Cycle, 120–39, passim; and Grabar, The Alhambra, 127–9. 70. Grabar, The Alhambra, 28. 71. Or as Dickie puts it in ‘Some Reflections’, 56: ‘the Court of the Lions is a translation into Islamic terms of the ancient peristyle under the influence of the (Perso-Central Asian) quadripartite division of space as seen in the Islamic garden’. 72. Bellamy and Steiner, Ibn Sa’Èd, vii. 73. Ettinghausen and MacDougall, The Islamic Garden, 4. 74. The Alhambra may seem like a dead-end for garden development. This is probably true for Spain, but some of its forms and ideas may have influenced the later gardens of Morocco, the famous riyå∂. These once existed in very large numbers, as we can gather from travellers’ accounts, and some still survive in a good state of preservation. In addition to Marçais’s short but perceptive remarks in ‘Les jardins de l’Islam’, 240–3, the only study of these gardens is L. Gallotti, Le jardin et la maison arabes au Maroc, 2 vols, Paris, 1926. They await a proper study, which may entail an endeavour similar to the one currently being carried out by J. Wescoat for the gardens of Lahore.

403

CHAPTER SEVENTEEN

Control and Abandon: Images of Water in Arabic Gardens and Garden Poetry

Philologists and art historians studying the Islamic world have long discussed the correspondence between poetry and gardens. Louis Massignon (1926) spoke of the manner by which Arabic poetry and song permeated the serene atmosphere of Islamic gardens, combining with their pavilions, flowers and water to produce a state of ecstasy or tarab in the visitor. Richard Ettinghausen (1976) understood the relationship between garden and poetry in terms of excessive stylisation and artifice whose aim was not so much to reproduce as to outdo nature. Donald Wilber (1979, 3–21) emphasised the imperial nature of Persian gardens and garden poetry, whose primary intent, he proposed, rested in exalting the ruler by linking him with the legendary gardens of ancient times. Scholars working on western Islamic gardens – including James Dickie (1968, 240–2), D. Fairchild Ruggles (1993; 2000, esp. 128–43 and 203–15), Cynthia Robinson (2002) and the present author (Tabbaa 1987, 1992) – have noted the presence of a literati spirit that seems to mediate the distance separating religious and imperial iconography and the changing reality of smaller gardens and intimate poetry. Finally, scholars interested in water technology and hydraulics – including Donald Hill (1993), Andrew Watson (1983) and Ruggles (2000) – have underscored the pivotal importance of the Islamic world in creating and developing hydraulic techniques that stood Yasser Tabbaa (2009), ‘Control and Abandon: Images of Water in Arabic Gardens and Garden Poetry’, in Jonathan Bloom and Sheila Blair (eds), Rivers of Paradise: Water in Islamic Culture, Yale University Press, 29–57.

control and abandon

at the foundation of garden design both in and outside the Islamic world. Certainly one reason for the richness of the modern discourse on gardens and poetry is that Muslims – Arabs, Persians and Turks – wrote rather copiously about gardens, both in poetry and prose. Poems, descriptions, travellers’ accounts and treatises on the hydraulics, design and agronomy of gardens in the Islamic world are relatively common and help inform the contemporary discourse on these subjects (the most recent bibliography on the technological side of gardens is to be found in Ruggles 2000). The linkage between these two creative realms – the textual and the architectural – was especially intimate, as I hope to demonstrate, in Islamic Spain, where something like a reciprocal process seems to have formed between gardens and garden poetry, such that as gardens and fountains became increasingly viewed in poetic terms, the gardens themselves may have aspired to these literary ideals. Such congruence between the textual and the visual is quite rare in Islamic culture, which – with the exception of calligraphy and later Persian painting – remains fairly silent about its artistic creations. The mere existence of such congruence deserves closer investigation, and its prevalence over diverse regions and historical epochs invites a comparative and diachronic approach that takes into account the changing perceptions of water and gardens across space and time. Such an approach is both necessary and possible. It is necessary in order to critique the rather static but dominant view of Islamic gardens as earthly manifestations of the paradisiacal garden described in the Qurʾån (see, in particular, Burckhardt 1976 and Lehrman 1980). And it is now possible, thanks to the regional studies that have so far been conducted in different parts and periods of the Islamic world. This essay will attempt to examine the exchange and interplay between poetry and water and garden design and to highlight the physical, technological and aesthetic aspects of this relationship. In order to accomplish this objective, while also addressing historical change, the essay will adopt a comparative and diachronic approach that examines the continuities and differences in the perception and implementation of gardens in three different epochs in the central and western Islamic world: pre-Islamic-Umayyad, ʿAbbasid-Cordoban, and medieval. More specifically, the essay will highlight the important but still understudied medieval period (twelfth–­fourteenth century), whose citadel culture permeated the entire Islamic world and contributed to important transformations in hydraulics, fountains and garden design. Not surprisingly, this period also witnessed significant developments in garden poetry and a more intimate linkage between poetry and gardens.

405

406

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Gardens in the pre-Islamic and Umayyad imagination Long before their outpouring into the relatively watered regions of the Near East, the Arabs inhabited one of the most arid regions on earth. Despite the existence of the mercantile city of Mecca and such oasis towns as Medina and Tayma, settled communities were limited and transient phenomena whose seasonal existence rested on irregular wadis and sparse watering holes. This was the memory that the invading Arabs would carry with them north, east and west, a memory that was evoked and immortalised by the Jahili poets of pre-Islamic times. Although the ultimate objectives of a Jahili qasida (ode) were pride (fakhr) and panegyric (madih), all qasidas begin with a passage of desert travel, in which the poet returns to the site of a vanished settlement where he had once experienced a pleasant sojourn or a love affair. The poet evokes these sad and sweet memories by means of the meagre remains, dried wadis and tumbledown wells of the settlement. Often called buka ʿala al-atlal (‘weeping by the ruins’), this genre of poetry encapsulates the essence of the relationship of the poet as desert traveller to the transient settlements and dried-up oases that characterised the Arabian landscape. Perhaps the most famous of these odes is the beginning of the one by ʿImruʾl-Qays (d. circa 550; Irwin 2002, 7): Halt, friends both! Let us weep, recalling a love and a lodging   by the rim of the twisted sands between al-Dakhoul and Haumal, Toodih and al-Miqrat, whose trace is not yet effaced   for all the spinning of the south winds and the northern blasts; there, all about its yards, and away in dry hollows   you may see the dung of antelopes spattered like peppercorns. Nearly as famous are these lines from the Mu ʿallaqa of Zuhair b. Abi Sulma (early seventh century): I stood again near the encampment of Awfa after an absence of twenty years, and with some efforts, I know her abode again after thinking awhile. I recognised the three stones blackened by fire at the place where the kettle used to be placed at night, and the trench round the encampment, which had not burst, like the source of a pool. Look, oh my friend! Do you see any women traveling on camels, going over the high ground above the stream of Jurthum? When they arrived at the water, the mass of which was blue from intense purity, they laid down their walking sticks, like the dweller who has pitched his tents.1

control and abandon

Water, greenery and grazing animals partook of this poetic vision only as memories. Here are the scattered stones of a dried well; there lie the weathered stakes of a tent; further on spread the spattered dung where animals once grazed. Permanent settlements are hardly mentioned; flowing water and irrigated fields are excluded from this vision, as if their mere mention would taint the poet as soft and effeminate, not a true desert wanderer. It is a romantic, even romanticised vision, for whereas these poets most probably did wander the desert on camelback, they must have also spent a good part of their time at the courts of various Arab kingdoms, whether Ghassanid, Lakhmid or Kinda. Islam transformed this memory into a promise. The believer was no longer a desert wanderer bemoaning the loss of things past but was now a believer, a follower of the true path, a path that culminated in Paradise. Mentioned about thirty-five times in the Qurʾån (including 2:25, 6:141, 15:47), Paradise is depicted as the blissful abode of the righteous believers: gardens beneath which rivers flow; elevated chambers attended to by houris; four rivers of sacred water, milk, wine and honey; a shady grove drenched in water and populated with all things beautiful. These paradisiacal images are serene, sensuous, fecund, overwrought, and unrealistic. Clearly, this is not a normal garden but the eternal garden of Paradise, and its refined imagery was intended to transport the true believer into a world that he or she would inhabit after death. More than anything, the garden of Paradise emerges as the exact opposite of the world that surrounded the Arabs of Arabia: watered, shady, plentiful, and rather divorced from reality. It is not a place in which one lives and through which one walks, but an encompassing image of immersion, contemplation and pleasure. Finally, the paradisiacal images presented by the Qurʾån could not, except in the most impressionistic way, have served as a blueprint for the creation of real gardens. For pre-Islamic Arabs, therefore, water was a memory; for Muslims, it was an eternal promise. In other words, water inhabited a distant past and an even more distant future, but was not a present phenomenon that captured the imagination of the Jahili poet or the visionary images of the Qurʾån. The transformation of memory into an existential presence and of promise into reality took several centuries, and it required the confluence of visual, poetic and iconographic ideas, as well as significant developments in hydraulic technology. Surprisingly, this transformation did not take place under the Umayyads, who, despite their prodigious architectural activity, are not known to have built gardens.2 Moreover, Umayyad poetry, which is quite prodigious, remains silent about these architectural accomplishments, largely dwelling on the traditional themes of Jahili poetry, preserving a romantic image that was increasingly at odds with the changing architectural reality. One possible exception

407

408

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 17.1  Damascus. Great Mosque, 705–15, the ‘Barada Panel’ (photo: Yasser Tabbaa)

Figure 17.2  ‘Barada Panel’ mosaics in the Great Mosque of Damascus, Syria; 705–15 (photo: Yasser Tabbaa)

is the poetry intended for the caliph al-Walid II (r. 743–4), which seems to engage visual themes from Khirbat al-Mafjar near Jericho, but even these poems were not so much about the palace and its pools as they were about the extravagant feasting and drinking that took place in them (Hamilton 1988). Even the famous mosaics of the Great Mosque in Damascus (705–15), possibly the most outstanding depiction of the gardens of Paradise in Islamic art, went virtually unnoticed by contempo-

control and abandon

rary poets and prose writers alike (Figure 17.2). With their riparian landscape inhabited with tall trees and fanciful structures, these mosaics were less inspired by formal gardens than by Roman and Byzantine mosaics, the natural by landscape surrounding Damascus, or Qurʾånic imagery. Interestingly, it was not until well into the ʿAbbasid period that such authors as the tenth-century geographer alMuqaddasi (1966) mentioned these mosaics, not in terms of Paradise but as a literal representation of the palaces of the Umayyads. This description seems to be an ‘ʿAbbasid’ interpretation of the Umayyad mosaics, an interpretation that is undoubtedly inspired, as we shall see next, by the increasingly palatial nature of ʿAbbasid rule. Gardens of the ʿAbbasids and the Córdoban Caliphate Both the absence of gardens in the Umayyad period and the disjunction between the architectural and the poetic realms changed quite significantly in the ʿAbbasid period, a transformation that some writers have linked to the enormous palaces and gardens that were erected in Samarra in Iraq during the city’s brief florescence (836–92). No fewer than six enormous palaces were built, all of which enclosed spacious courtyard gardens overlooking even larger gardens that often cascaded all the way down to the banks of the Tigris (the site reports and findings of the excavations were published in a number of volumes, of which the most useful for architecture is Herzfeld 1948; see also Sousa 1948–9 and al-Samarraʾi 1968, esp. i: 87ff.). Although variations exist, the most typical of these gardens are those of the Balkuwara palace (854–9), which contains within its vast enclosure several quadripartite gardens whose four parterres are divided by cruciform water channels, with a little island at the point of intersection (Figure 17.3). This garden type represents a reinvention of the chahar-bagh type that had existed since the time of the Achaemenids (c. 550–331 bce), for whom earthly gardens reflected the gardens of Paradise, itself a word derived from the Persian pairidaiza and its Greek derivative paradeisos. Regardless of its archaeological origin or etymology, there is little doubt that this garden type had the widest and most profound impact on the development of Islamic gardens, from Spain to India. Although often understood as an image of Paradise, the links of this garden type with power are also quite unmistakable: most clearly visible in the regularisation and control of water and gardens under the direct gaze of the palace (on this concept, see Necipo©lu 1993; Ruggles 2000, 94–109). In addition to these quadripartite gardens, Samarra also possessed a less formal garden called hayr al-wuhush (‘game preserve’), built by the caliph al-Mutawakkil (r. 847–61) south of the city. This was a gargantuan enclosure (about 350 square kilometres) traversed from north to south by a large canal, a branch of which fed the legendary pool al-Birka al-Jaʿfariyya. On its sides the pool, about 200 metres

409

410

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 17.3  Ground plan of Balkuwara Palace, Samarra, Iraq; c. 850

square, once had animal- and bird-shaped fountains made of semiprecious materials. A large pavilion with a colonnaded portico stood at the northern end of this pool (Sousa 1948–9, ii: 298). In his famous panegyric written for the caliph al-Mutawakkil in 861, the court poet al-Buhturi included twenty lines describing the al-Birka al-Jaʿfariyya: Torrents of water pour into it rushing Like horses let loose from their race gates. It is as if ingots of white molten silver are flowing in its streams The Sun’s eyebrow sometimes makes it laugh While the spattering of rain sometimes makes it cry. When the stars gaze at themselves in its sides at night You would think a firmament had been built into it. In a recent study of this poem, Stefan Sperl (2006) discusses the contrasts al-Buhturi makes between the uncontrolled ravages of the natural landscape, swept by wind, sand and rain, and the protected enclosure surrounding the pond. The violent winds sweeping the encampments of ʿImruʾl-Qays and Zuhair only gently caress the surfaces of this pond, whose lasting presence beautifies and nourishes the surrounding landscape. In effect, the caliph himself transforms nature, creating landscapes that supersede and transcend

control and abandon

nature’s creations, in turn reinforcing the image of the caliph as the divinely sanctioned ruler and defender of the faith. Beneath his feet extend waters that have been shaped and controlled into channels unequalled in the natural world. Although in his earlier poems al-Buhturi spoke of his desert wanderings and his longing for desert life, from c. 860 until his death in 897 he wrote several qasidas that transformed the earlier formats of nasib and fakhr by including extensive descriptions (wasf ) of the palaces, gardens, and pools of Samarra. Al-Buhturi’s poetic transformation, undoubtedly motivated by the wondrous effect of Samarra’s palaces and gardens, had a profound influence on ʿAbbasid poets, and especially on poets in the Maghrib and al-Andalus. His vivid and breathless descriptions of the palaces, gardens, pools and fountains of Samarra foretell later developments in the gardens and fountains of Sicily and Spain, thereby suggesting a possible mode of oral transmission for these architectural features (Sousa 1948–9, ii: 306–13, where the author reconstructs the pool of al-Mutawakkil with zoomorphic fountains, narrow channels with running water, and little island pavilions in the middle of large pools with several paved platforms in its middle). Furthermore, the entire genre of garden poetry – with its emphasis on luxurious surroundings and the similitude of natural elements, especially water, to precious materials – would be exported in succeeding centuries westward to North Africa, Sicily and Spain, where it enjoyed a long life. The short-lived magnificence of Samarra resonated in the Islamic world for several centuries. Madinat al-Zahra, the tenth-century palatial city of the Córdoban Caliphate, was consciously modelled after the palaces of Samarra. ʿAbd al-Rahman iii, who founded it in 936, styled his kingship after the Samarran ʿAbbasid model, and Madinat al-Zahra was meant to be an expression of his new caliphate. Although Madinat al-Zahra has often been compared to Samarra, it is in fact more comparable to the two main royal palaces of Samarra – the Jawsaq al-Khaqani and the Balkuwara – which it resembles by its location near a river, the terraced layout of its palaces, quadripartite division of its gardens, and pools, channels and fountains. Despite their many similarities, Madinat al-Zahra is not simply a copy of the palaces of Samarra, but represents an original mix of forms and ideas derived variably from the topography of the site, local building traditions, and a yearning for the culture of the East. Compared with the palaces of Samarra, Madinat al-Zahra appears less formal, more intimate, and better integrated with the landscape. Situated on a steep slope, the main pavilions are arranged on three terraces cut into the side of the mountain, such that each pavilion caught glimpses of its own gardens and its own landscapes. The Salon Rico, the best-preserved of the pavilions at Madinat al-Zahra, consists of a basilical hall that looked onto a large pool through five arches (Figure 17.4). The pool ends in a smaller pavilion

411

412

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 17.4  View of Madinat al-Zahra, Spain; 926 (photo: Yasser Tabbaa)

that was itself surrounded on its three other sides by smaller tanks and appears like a small island surrounded on four sides by pools. The composition ultimately recalls the palace and pool of the hayr al-wuhush in Samarra and is repeated in several Islamic gardens, both western and eastern. By the eleventh century most of the basic forms of Islamic gardens seem to have assumed their definitive form. These include the use of terracing, the fourfold garden with intersecting water channels, the basilical pavilion facing a water tank, and the freestanding pavilion that is open on all four sides to water and gardens. Towers with commanding views (manzara) were also parts of this garden vocabulary. Finally, these gardens included a variety of fountains, at first quite varied and exotic but eventually streamlined into a standard variety of architectural fountains. Gardens and fountains in medieval citadels By the twelfth century, however, a different garden and water aesthetic seems to emerge, and it is eventually accompanied by a different poetic sensibility. Beyond the far more limited resources of the petty dynasties of the Middle Ages, it seems that the most decisive factor was the rise of the citadel as the primary centre of defence and royalty. As palace culture around the beginning of the twelfth century came to be centred within fortresses and citadels, a transformation that affected the entire Islamic world, there was a necessary diminution in the size and luxury of palaces, which, as I have shown elsewhere (Tabbaa 1993), were on average one-fiftieth the size of their early Islamic predecessors. In addition to these generally cramped palaces, some medieval dynasts continued to build modest

control and abandon

suburban gardens and pavilions (jawsaqs), examples of which are still preserved in Sicily and in a unique survival in southern Turkey (Redford 2000). Furthermore, the very height of citadels led to quite significant changes in the nature of fountains and the hydraulic devices required to raise water to high places. Two water-raising mechanisms were used throughout the Islamic and Iberian worlds. The first is the naʿura or noria, a waterwheel with hinged compartments that dip into the stream and carry the water to the top of the wheel, where it is discharged into an aqueduct. Since norias rely on the push of water for their motion, they can be used only in fairly rapid streams, which in the Islamic world restricts their usage to Syria, parts of North Africa and Spain. Although they are primarily used for agriculture – as in Hama (Syria), where some exceptionally large ones survive – they also seem to have been used to irrigate gardens and palaces in Spain (see Ruggles’s essay in this book). Some were illustrated in the thirteenth-century Spanish manuscript Bayad wa Riyad, and poems were even written about them, often comparing their melancholy groan to the wailing of parted lovers (Tabbaa 1992, 322). Far more practical for the arid climate of most of the Islamic world was the saqiya, a water-elevating machine that generally consists of two interlocking wheels, one horizontal and one vertical, that are placed above a well (Figure 17.5). The horizontal wheel is driven by one or two animals, and the vertical wheel is attached to a potgarland that dips in the well and raises the water to an upper channel. Watson (1983) suggests that ‘the machine was probably transmitted to Spain from Syria, when the Muslims introduced their irrigation methods to Spain’, and indeed the eleventh-century Andalusian agronomist Ibn al-Bassal (1959) described it as a standard machine for irrigation. As with the noria, the saqiya was used mainly in agricultural fields, but its use in gardens and even in citadels is well attested in the citadels of Aleppo and Cairo. In Aleppo a large saqiya was constructed over the well of the Satura, and in Cairo a large saqiya was built near the Mamluk Iwan al-Kabir, driven by two buffalos (Rabbat 1995). The elaborate fountains that stood at the end of these hydraulic accomplishments differed from their early Islamic predecessors by their deeper integration with the architecture. One of the most interesting of these fountains is sometimes called a shådirwån, although salsabÈl is more commonly used, especially in later periods (on the etymology of these two terms, see Tabbaa 1987). A salsabÈl typically consists of a water spout with or without an animal head in the back wall of an iwån; an inclined, carved marble slab called a shådirwån on which water flows; a long and thin channel running through the centre of the iwån; and a pool in the middle of the courtyard into which the salsabÈl water flows. The first known shådirwån slab comes from a palace in Algeria (Golvin 1965, 122–7 and pls 43–4), but the earliest preserved salsabÈl

413

414

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 17.5  A saqiya or water-raising device

(Figure 17.6) is that found in the Zisa palace outside of Palermo, begun by the Norman king William I and completed by his successor William II in 1189 (thorough documentation and analysis of this unique monument in the restoration report of Bellafiore 1978; also discussed in Marçais 1954, 121–3). In this salsabÈl water flowed from a spout beneath a stone muqarnas vault, down a marble shådirwån, in and out of two small pools, and finally through a long, partly submerged channel that emptied into a large pool with a little pavilion in its middle (see also Miller 1983, 143). A similar fountain once existed in the nearby La Cuba palace, all part of a garden complex that the Normans referred to as genoard, a corruption of the original Arabic term jannat al-ard or ‘earthly paradise’ (Finley et al., 643). That salsabÈl fountains were fairly common in Norman Sicily is evidenced by the depiction of one such fountain among the paintings of the Palatine Chapel in Palermo and references to them by contemporary Sicilian poets. One of these paintings shows two male musicians seated on either side of a quadrilobed square pool with a jet, whose water spouts from a lion head and flows down a shådirwån. In a poem describing a palace of King Roger of Sicily, the poet al-Butheiri al-Siqilli wrote: ‘And the lions of its shadirwan spout kawthar-like

control and abandon

Figure 17.6  SalsabÈl fountain in the Zisa Palace, Palermo, Sicily; 1189

waters’ (cited in Imad al-Din al-Katib Isfahani 1966, i: 23). The poetry of perhaps the most famous Sicilian poet, Ibn Hamdis (d. 1132), contains several references to shådirwåns and other fountains found in the palaces of the Hammadids of Algeria and the Normans of Sicily. His descriptions of the palace of al-Mansur4 in Bijaya mention pools into which water pours from the mouths of birds, giraffes and lions: ‘Its waters are like ingots of silver that melt on the steps of a shadirwan’ (al-Nuwayri 1998, 286; Ibn Hamdis 1960, 494–7). Such extramural garden pavilions or water palaces were once quite common in the central Islamic world, and some have been documented archaeologically, but, to my knowledge, only one has survived nearly intact (Redford 2000): a little-known palace called al-Firdaws, located east of Mardin in south-eastern Turkey (Figure 17.7). In all likelihood, it is the one remaining pavilion (Arabic jawsaq) from the

415

416

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 17.7  Plan and section of the Qasr al-Firdaws east of Mardin, Turkey; 1239–60 (photo: Yasser Tabbaa)

Figure 17.8  View of main iwån in the Qasr al-Firdaws east of Mardin, Turkey; 1239–60 (photo: Yasser Tabbaa)

control and abandon

Figure 17.9  SalsabÈl in the Ayyubid Palace, Aleppo, Syria; c. 1200 (photo: Yasser Tabbaa)

group of palaces and pavilions built by the Artuqid prince Najm al-DÈn Il-Ghazi I (r. 1239–60) and named al-Firdaws by him (Ibn Shaddad 1978, 543–4, identifies this palace). It consists of a large central iwån flanked by a pair of small superimposed iwåns and another group of two small iwåns to the west of the large iwån. All the iwåns face a large cistern and overlook the southern valley below. Five salsabÈls, one in each of the lower iwåns, originate from stone spouts in the iwån’s back wall and descend through a number of drops and chutes, passing through several small pools before emptying into the cistern from three different points (Figure 17.8). The succession of pools, the use of partly open and partly covered channels, and even the whole composition of a pavilion fronted by a large cistern with a fountain connecting the two, bring to mind La Ziza palace in Palermo. Palaces built in castles were generally of the four-iwån plan, sometimes with a salsabÈl flowing out of the main iwån, an arrangement found in the castles of Aleppo, Sahyun and Najm in Syria and Diyarbakir in Turkey. In the palace of the citadel of Aleppo (Figure 17.9), which dates from c. 1200, the shådirwån is located in the northern iwån, which is the largest and most elaborate iwån in the palace (Herzfeld 1955–7, i: 135–7). Sheltered by a muqarnas vault, the water cascades down the shådirwån, collects in a little pool, flows underneath the floor to the end of the iwån, falls in a little cascade into an open channel, and empties into the central pool (Aslanapa 1962). Whether in Palermo, Aleppo, Mardin, or other medieval palaces and residences, varieties of the salsabÈl fountain became quite common. Overall, these fountains contribute a sense of movement to the architecture and help interconnect interior and exterior spaces, a characteristic feature of many Islamic palaces.5 Perhaps

417

418

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

more, these fountains – with their cascades, narrow channels and spurting fountains – celebrate water: its stillness and movement, its apparent solidity as it runs in narrow channels as opposed to its actual malleability, and its use as a thin decorative veil over stone. In some respects, the pavilions themselves are little more than the setting and the perfect vantage point for the celebration and contemplation of water by poets and patrons alike. Medieval Islamic poets were especially fond of the manifold qualities of water, which they enhanced by linking them variously with precious or semi-precious substances, such as molten silver, rock crystal, lapis lazuli, pearls and tin-plated mirrors; with the features of beautiful women or men, such as blue eyes and white teeth; or even with astral images, including stars and galaxies. A fountain shooting water upwards was often compared to a rod of rock crystal; water flowing in a stream resembled molten silver; a brook meandering through a meadow recalled ‘a hollow disk of silver on a green garment’; a clear pool evoked deep blue eyes (for the most complete anthology of Andalusian garden poetry, see al-Shakʿa 1972, 247–370). Although some of this imagery had already been stock-in-trade since the ʿAbbasid period, its increased quantity and focus show greater engagement with nature, garden and water. For Ibn Khafaja (d. 1138), possibly the most important poet of nature in western Islam, this engagement with nature became a near obsession. Spared by a small personal fortune from having to write panegyric for profit, Ibn Khafaja focused his talents on the description of nature and gardens, making it the primary theme of his poetry and even the point of departure for all his other poetic genres. His solitary walks in nature – rare among Arab poets – and his intense observation of both the grand and the minute in the natural and cultivated worlds demonstrate a significant departure from traditional norms and a nearly complete disavowal of the classical qasida. The sights, sounds and scents of gardens engulf him and send him into a dream-like trance, a rapture that he tries to convey through delicate and at times quite moving images (see, for example,ʿAbbas 2001, 63–72). In his poem ‘The Dancing Garden’, Ibn Khafaja writes (my translation): With shiny flowers, its sides are blown by a wind that twists its branches and spreads its perfume. A black-eyed youth drank wine in it traversing it from darkness of night till the break of dawn. Its flowers are a necklace, the branches sidelocks the trunk a wrist and the bay a bracelet In a garden whose shady groves resemble darkened lips from which peak flowers as white as teeth A tree danced in it, having drunk the earth. Pigeons cooed and the breeze clapped.

control and abandon

And in a poem entitled ‘Twisted like a Bracelet’, he says: By God, a river flowing on level ground   better tasting than the lips of a beauty Meandering like a twisted bracelet, resembling,   with flowers flanking it, a galaxy in the sky So placid you would think it liquid silver   flowing through a green dress With tree branches drooping into its banks   like eyelashes surrounding deep blue eyes. Oftentimes did I drink wine in it   a golden sunset on silvery waters. Poetic bouts and amiable competitions were often held in gardens, where the poets would take turns extemporising about such things as a flower arrangement, a fountain, a pool, and even a natural stream. The anthologist al-Ghuzuli (d. 1412) describes several such gatherings, including the following: ‘we were gathered in an intimate séance with the litérateur Abu Ishaq . . . in Fayyum. There was a garden with a pool and a water jet, and we took turns describing it.’ The poems generally dealt with either the shape of the jet, which was typically compared to a silver or crystal column, and its effect on the standing water, comparing it to a starry sky or the crying eyes of a longing lover or a homesick person. In another episode, the author writes: I met with a group of Alexandrian literati in a garden with a pool . . . One of us scattered jasmine petals in the pool . . . We all dropped our heads to stimulate and emote our minds and then showed what we had written . . . I wrote a couplet: ‘They scattered jasmine petals on water, looking like the stars in the sky.’6 While lacking the poignant longing of ʿImruʾl-Qays or the imperial grandeur of al-Buhturi, the short descriptive poems of Andalusian poets reflect a different feeling, more closely linked to the sensibilities of the literati (zurafa) than the rigidity of royal patronage. In this respect, these poems seem to echo the changing appearance and experience of medieval gardens, whose intimate scale and focused view invited closer inspection and deeper contemplation. Small gardens designed to be perceived with one look, fountains that manipulate water, streams that connect spaces, pools that provoke reflection, an intoxicating intensity of sights, sounds and perfume: these are some of the themes and motifs shared by Andalusian poetry and gardens. Ibn Zamrak’s poem, inscribed around the rim of the basin of the Fountain of the Lions at the Alhambra Palace in Granada (pls 2.9a, 2.2b), fits in quite well with this sensibility (Grabar 1978: 124–5):

419

420

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Silver melting which flows between jewels, one like the other in beauty, white in purity A running stream evokes the illusion of a solid substance For the eyes, so that we wonder which one is fluid. Don’t you see that it is the water that is running over the rim of the fountain, Whereas it is the structure that offers channels for the water flow. The poem, written especially for the fountain and inscribed on it, forms a persistent commentary of the fountain, an indelible literary dimension of the architectural monument. The Figure 17.10  (upper) Fountain in the Court distance between this fusion of the of the Lions, Alhambra Palace, Granada, visual and the poetic, and ʿImruʾ Spain, fourteenth century; (lower) Detail of the fountain in the Court of the Lions, al-Qays’s evocation of memory Alhambra Palace, Granada, Spain, through meagre remains, mark a fourteenth century (photo: Yasser Tabbaa) seventh-century journey travelled by Arabic poetic consciousness as it negotiated a space within the increasing sophistication and variation of medieval Islamic society. Paradise, revisited This piece has quite self-consciously put off discussing perhaps the most prevalent image of Islamic gardens, namely Paradise, an image that has been used rather ahistorically by several writers (e.g. Burckhardt 1976; Lehrman 1980). I did this not to minimise the importance of paradisiacal imagery in gardens and poetry, but rather to allow other visual and poetic images to develop within fairly specific historical and cultural contexts before, once again, linking them with the image of Paradise. We have therefore seen gardens as images of power and magnificence, as vehicles for agronomic and technological sophistication, as the abode for the literati, as wondrous creations redolent with exotic images, and as sequestered spaces enclosing a refined nature. How did these images and ideas interact with the concept of the Qurʾånic Paradise, a concept still used by some as the primary model and measure of Islamic gardens? Are they simply extensions of that rather static idea, commentaries on a grand discourse? Or do these

control and abandon

images and ideas represent expansions and transformations of the Qurʾånic Paradise, attempts to reconstruct the concept in ways that seem more appropriate to changing times and culture (Ruggles 2000: 209–20)? If anything, our attempt to examine the interaction between poetry and gardens would seem to suggest that this lengthy process had the effect not only of enriching this capacious concept but also of actually transforming it from within. Paradise therefore emerges not so much as a timeless essence reflecting a Qurʾånic Paradise but as a human construction that varied according to historical and cultural conditions. Despite their significant transformations between the early and medieval periods and their interaction with poetic ideas, Islamic gardens maintain a coherent image and a sense of decorum that admits certain images while keeping others outside their walls. This was not a place to look for the violent and the shocking, for the torrential and the sublime, and certainly not for the bizarre or the grotesque. Islamic gardens, and to a large extent garden poetry as well, kept these images and ideas at bay: in an extramural world and a distant consciousness that had no place in gardens or poetic imagination. Meandering streams were savoured, but neither the Tigris nor the Nile received much attention by poets. A decorative shådirwån with a thin sheet of water was relished, but heavy waterfalls were ignored by poets and garden-makers alike. A sense of meditative detachment – ‘a great silence’, in the words of the Persian poet Saʿdi – pervades the experience of the Islamic garden; it is an experience that saturates the senses and elevates the mind but does not call for the active engagement of the visitor (Wilber 1979: 17). Viewed from a high vantage point or enjoyed from an open pavilion, gardens were rarely the subject of promenades or exploration. Places of earthly delights and sensual pleasures, they gave their noble visitors a foretaste of Paradise, a divine image upon which they could inscribe their feelings and aspirations.

Notes 1. http://www.fordham.edu/halsall/source/64Ohangedpoems. html#The%20Poem%20of%20Zuhair 2. It might be argued that the cultivated estates around some of the larger Umayyad palaces, such as Khirbat al-Mafjar and Qasr al-Hayr al-Gharbi, were gardens. But, with the possible exception of the small garden around the courtyard fountain of Khirbat al-Mafjar, these were primarily agricultural developments that do not seem to have been aesthetically planned or used for the purposes of leisure. 3. See also www.comune.palermo.it/musei/zisa/index_en.html 4. This refers to al-Mansur ibn al-Nasir ibn ʿIlnas or Aʿla al-N as (1090– 1104), the second Hammadid prince to settle in the maritime capital of Bougie. For another poem of Ibn Hamdis describing the same palace, see Gabrieli 1959, 54–8.

421

422

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

5. According to the garden historian Georges Bazin (1990, 1), ‘this sense of unity between interior and exterior is an essential aspect of Islamic culture’. 6. Similar themes occur in al-Ghuzuli’s Mataliʿ al-budur fi manaz il alsurur, which contains a chapter (1905, i: 35–45) on pools, fountains and waterwheels.

Bibliography ʿAbbas, I. (1968), ‘Rihlat Ibn al-ʿArabi ila al-mashriq kama sawwaraha “Qanun al-ta’wil”’, Al-Abhath, 21/r: 59–92. ʿAbbas, I. (2001), Tarikh al-adab al-Andalusi: ʿShi'rsral-tawaʾif waʾlMurabitin. Amman: Dar al-Shuruq. Al-Shakʿa, M. (1972), Al-Adab al-Andalusiyy: MawdËʿåtuh wa FunËnuh. Beirut. Aslanapa, O. (1962), ‘Erster Bericht uber die Ausgrabungen des Palastes von Diyarbakir’, Istanbuler Mitteilungen, 12: 5–28. Bazin, G. (1990), Paradeisos: The Art of the Garden. Boston, Toronto and London: Little, Brown. Bellafiore, G. (1978), La Ziza di Palermo. Palermo: S. I. Floccovio. Burckhardt, T. (1976), Art of Islam: Language and Meaning. London: World of Islam. Castejon, R. (1945), Excavaciones del plan nacional en Medina Azahra (Cordoba), campaia de 1943. Madrid: Ministerio de Educacion Nacional. Dickie, J. (1968), ‘The Hispano-Arab Garden: Its Philosophy and Function’, Bulletin of the School of Oriental and African Studies, 31: 237–48. Ettinghausen, R. (1976), ‘Introduction’, The Islamic Garden, ed. E. B. MacDougall and R. Ettinghausen. Washington, DC: Dumbarton Oaks. Gabrieli, F. (1959), ‘II palazzo hammadita di Bijaya descritto da Ibn Hamdis’, Aus der Welt der islamischen Kunst: Festschrift fur Ernst Kuhnel, ed. R. Ettinghausen. Berlin: De Gruyter. al-Ghuzuli (1905), Matali’al-budur .fi manazil al-surur. 2 vols. Cairo: s.n. Golvin, L. (1965), Recherches archeologiques a la Qalʾades banu Hammad. Paris: Maisonneuve et Larose. Grabar, O. (1978), The Alhambra. Cambridge, MA: Harvard University Press. Hamilton, R. W. (1959), Khirbat al Mafjar. An Arabian Mansion in the Jordan Valley. Oxford: Clarendon Press. Hamilton, R. W. (1988), Walid and His Friends: An Umayyad Tragedy. Oxford Studies in Islamic Art, vol. 1. Oxford: Oxford University Press. Hill, D. R. (1993), Islamic Science and Engineering. Edinburgh: Edinburgh University Press. Ibn al-Razzaz al-Jazari (1974), The Book of Ingenious Mechanical Devices, trans. and ed. D. R. Hill. Boston, MA: Reidel. Ibn Hamdis (1960), Diwan Ibn Hamdis, ed. Ihsan Abbas. Beirut: Dar Sadir. Ibn Shaddad (1978), Al-A claq al-Khatira fi Dhikr Umaraʾ al-Sham waʾlJazira, vol. III, 2 parts, ed.Yahya ʿAbbara. Damascus: Wizarat al-Thaqafah waʾl-Irshad al­Qaumi. Imad al-Din al-Katib al-Isfahani (1966), Kharidat al-Qasr wa-]aridat al- Asr, vol. I, ed. Muhammad al-Marzuqi et al. Tunis: al-Dar al-Tunisiyya liNashr. Irwin, R., ed. (2002), Night, Horses and the Desert: An Anthology of Classical Arabic Literature. New York: Anchor Books.

control and abandon

Lehrman, J. (1980), Earthly Paradise: Garden and Courtyard in Islam. Los Angeles: University of California Press. Marais, G. (1954), L’Architecture musulmane d’Occident. Paris: Arts et Metiers Graphiques. Massignon, L. (1962), ‘Salsabil et Sadirwan’, Études d’Orientalisme dédiées à la mémoire de Lévi-Provençal, vol. II. Paris: CNRS, pp. 639–48. Miller, N. (1983), ‘Paradise Regained: Medieval Garden Fountains’, Medieval Gardens, ed. Elizabeth B. MacDougall. Washington, DC: Dumbarton Oaks Publications, pp. 137–53. al-Muqaddasi (1966), Ahsan al-taqasim Ji maʾrifat al-aqalim. Beirut: Dar Sader. Necipoglu, G. (1993), ‘Framing the Gaze in Ottoman, Safavid, and Mughal Palaces’, Ars Orientalis, 23: 303–42. Nuwayri, Shihab al-Din Ahmad (1998), Nihayat al-arabfi funun al-adab, ed. Fahim M. Shaltut. Cairo: Dar al­Kutub al-Misriyya. Rabbat, N. (1995), The Citadel of Cairo: A New Interpretation of Royal Mamluk Architecture. Leiden: Brill. Redford, S. (2000), Landscape and the State: Seijuq Gardens and Pavilions in Alanya, Turkey. Oxford: Archaeopress. Robinson, C. (2002), In Praise of Song: The Making of Courtly Culture in Al-Andalus and Provence, 1005–1134 AD. Leiden: Brill. Ruggles, D. F. (1993), ‘Arabic Poetry and Architectural Memory in al-­ Andalus’, Ars Orientalis, 23: 171–8. Ruggles, D. F. (2000), Gardens, Landscape, and Vision in the Palaces ef Islamic Spain. University Park: Pennsylvania State University Press. Tabbaa, Y. (1987), ‘Towards an Interpretation of the Use of Water in Islamic Courtyards and Courtyard Gardens’, Journal of Garden History, 71, 3: 197–220. Tabbaa, Y. (1992), ‘Typology and Hydraulics in the Medieval Islamic Garden’, Landscape and Garden History: Issues, Approaches, Methods, ed. J. D. Hunt. Washington, DC: Dumbarton Oaks, pp. 303–29. Tabbaa, Y. (1993), ‘Circles of Power: Palace, Citadel, and City in Ayyubid Aleppo’, Ars Orientalis, 23: 181–200. Watson, A. (1983), Agricultural Innovation in the Early Islamic World: The Diffusion of Crops and Farming Techniques, 700–1100. Cambridge: Cambridge University Press. Wilber, D. (1979), Persian Gardens and Garden Pavilions. Washington, DC: Dumbarton Oaks.

423

CHAPTER EIGHTEEN

Eternal Hunting Fields: The Frescos at Qusayr ʿAmra as a Pastoralist Interpretation of the Paradise Garden The perennialist discourse on the Islamic garden presents it as a timeless manifestation of Janna, the garden of paradise as described in the Qurʾån. Central to this discourse is an uncritical reading of the Qurʾånic descriptions of paradise and their ahistorical juxtaposition against an ancient garden type, namely the chahar-bagh or quadripartite garden.1 For many writers, this curious juxtaposition leads to an inevitable but largely unsupported conclusion that this garden type represented the essence of the Islamic paradise garden and was therefore symbolically charged with paradisiac imagery. In other words, the ahistorical fallacy of conjoining a garden type that dates back at least to the seventh century bce with Qurʾånic descriptions that only became known more than a millennium later is further exacerbated by a circular argument with regard to the symbolic nature of this garden type, such that the chahar-bagh garden becomes an earthly image of the Qurʾånic paradise.2 This narrow conception of the Islamic paradise garden and its static symbolism has until recently impeded the development of alternate conceptions of and perspectives on Islamic gardens, particularly in terms of their typology, hydraulics, agronomy, and, as will be examined below, their use as game and hunting reserves. Indeed, since the relationship, let alone correspondence, between the impressionistic Qurʾånic imagery and actual garden design cannot be demonstrated for the early Islamic period, we may therefore propose that the makers of the first gardens were not bound by one essential type but could have drawn on other garden forms, including the game and hunting reserves that were once quite prevalent in northern Arabia and Syria. The main question that this article addresses, therefore, is whether an alternate garden form with a distinctive iconography had existed in the early Islamic period, one that did not conform to Yasser Tabbaa (2019), ‘Eternal Hunting Fields: The Frescos at Qusayr ʿAmra as a Pastoralist Interpretation of the Paradise Garden’, in Despoina Zavraka (ed.), Islamic Gardens, Kavala, Greece: MOHA Research Centre.

eternal hunting fields

the prescriptions of the Qurʾånic paradise, and whether such gardens or idyllic landscapes were represented in Umayyad art or alluded to in Umayyad poetry. To this end, this piece will deal primarily with the figural frescos at the Umayyad palace of Qusayr ʿAmra, viewing them in relation to Umayyad and pre-Islamic representations of the royal hunt and their paradisiac iconography. Focusing primarily on the royal and hunting fresco cycles at Qusayr ʿAmra, the piece proposes that these paintings simultaneously described existing game reserves and hunting practices, while also exalting the founder, al-Walid II (r. 742–3), through his successful hunt, to the rank of world sovereigns. By viewing these complex frescos in relation to the royal iconography of the Sassanian relief sculptures at Taq-e Bostan, the piece hopes to further substantiate al-Walid II’s aspiration to universal kingship and a place in a paradise of his own choosing. Furthermore, by juxtaposing the iconographic programme at Qusayr ʿAmra against the more established paradisiac imagery at the Great Mosque of Damascus, the piece further proposes that visual constructions of the Islamic paradise responded to setting, patronage, and prevailing social and political conditions.3 The visual dimension Although the first image of a Qurʾånic paradise was quite literally reproduced in the Umayyad Great Mosque of Damascus (705–15), it is important to note that none of the Umayyad desert palaces attempted to create a garden that evoked a Qurʾånic paradise nor one that conformed to the chahar-bagh mode4.

Figure 18.1  Jordan. Qusayr ʿAmra, 743–5, main hall from north-east (photo: Yasser Tabbaa)

425

426

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 18.2  Great Mosque of Damascus (705–15) ‘Barada’ mosaic (photo: Yasser Tabbaa)

Adapted to the arid conditions of the Syrian Desert, some Umayyad palaces included irrigated orchards while most were intended as pleasure palaces, with bathing facilities and close links with hunting and pastoral life. Whereas some of the larger Umayyad palaces, such as Khirbat al-Mafjar, may have been a manor house, Qusayr ʿAmra could only have served as a small pleasure pavilion and a hunting lodge, a designation it shares with other small palaces, such as Hammam al-Sarah. These palaces, in particular Qusayr ʿAmra and Khirbat al-Mafjar, have been studied in the context of Umayyad ceremonial and poetry, highlighting their role in the princely pastimes, which include poetic bouts, drinking, bathing, the company of women and the hunt. Fowden, preceded by Grabar and Ettinghausen, also examined al-Mafjar and ʿAmra in terms of an emerging Umayyad royal imagery that drew on Byzantine and Sassanian royal iconography.5 No one, to my knowledge, has explored the frescos and other figural representations in these palaces as a possible interpretation of the paradise garden. Specifically, the hunting theme, which is quite prevalent in Umayyad figural iconography, has not been adequately explored in relation to the Persian Paradeisos, the eternal hunting fields, as an alternate image of Paradise. Of all Umayyad palaces, Qusayr ʿAmra is perhaps the most enigmatic, particularly because of its diminutive size, absence of a mosque, and of course, its extensive figural frescos. Thematically, the fresco paintings of Qusayr ʿAmra encompass a formal royal scene in the southern apse; the so-called Six Kings panel at the southern end of the western wall; the ‘Royal Family’ panel at the southern

eternal hunting fields

wall of the western aisle; several panels of bathing, nudity and sports on the western wall and throughout; extensive cycles of hunting and related activities on the western and eastern walls; images of labour in the upper left eastern wall; images from classical mythology that remain rather ambiguous; and an astronomical cycle painted in a little cupola above the caldarium of the bath. Some of these images – enthroned sovereign, mythological themes and images of labour – have precedents in western or eastern late antiquity. Others, in particular the Six Kings and the elaborate hunting scenes, have no direct precedent and seem to reflect the intentions of the patron himself. The frescos of Qusayr ʿAmra have been discussed individually in terms of their formal and iconographic sources but, with Fowden’s exception, not as a unified iconographic programme. Although it is unlikely that their quite disparate imagery belongs to a single narrative or iconographic programme, it is nevertheless quite evident that at least the images in the reception hall focus upon the enthroned sovereign in the iwån, or conversely, these images are subject to his gaze. Not yet caliph when he had built this palace, al-Walid II nevertheless presents himself as a Byzantine Emperor or Cosmokrator, seated on an elaborate throne; haloed by an arch of partridges, representing heaven; rising above a Nilotic scene of boats and sea creatures, representing water; and flanked by two attendants holding fans, alluding to his royalty. The vault above him is painted in thick foliage, possibly alluding to paradise, within which are constructed six arches, three on each side, inhabited by men and women, who could be members of his entourage or associates who had preceded him to heaven. The western wall to the right of the royal apse is painted in three registers, of which the lowermost is ornamental and the uppermost has a long hunting scene, while the middle register is vertically divided into three panels: the Six Kings panel at the southern end, the giant woman panel in the middle, and the acrobats panel at the northern end (Figure 18.3). The Six Kings stand in two rows, of which those in the front have been identified by their names, inscribed above their heads in Arabic and Greek as ‘Caesar’, ‘Kisra’, ‘Negus’ of Abyssinia, ‘Roderick’ the Visigothic king of Spain. The other two may have been the emperor of China and the ruler of the Turks. Interpreted by Grabar as the kings of the world paying homage to the Umayyad prince, the names of these kings were also used to date Qusayr ʿAmra to the time of al-Walid I (d. 715), making it the earliest Umayyad palace.6 But a newly discovered inscription located at the top of the south panel of the western aisle, above the so-called ‘Royal Family’ fresco, leaves no doubt that the founder of this palace is al-Walid b. Yazid, or al-Walid II, who lived about thirty years after his illustrious ancestor, a temporal discrepancy that leads us to conclude that the Six Kings represented in the panel had been dead for about two generations.

427

428

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 18.3  Jordan. Qusayr ʿAmra, 743–5, the Hunt fresco (photo: Yasser Tabbaa)

Figure 18.4  Qusayr ʿAmra – iwån with enthroned prince (photo: Yasser Tabbaa)

Why did al-Walid II, therefore, represent these Six Kings, who had preceded him in death, in a gesture of homage? Are they welcoming the Umayyad prince to the world of kings, as is generally accepted, or are they in fact welcoming him to the afterworld, an afterworld already inhabited by royal companions and equipped with their princely pursuits? Interestingly, these kings are represented beardless, contrary to the usual representations of the Sassanian Shah, which were always bearded, and the Byzantine Emperor, which were bearded after the

eternal hunting fields

Figure 18.5  Qusayr ʿAmra – frescos on west wall (photo: Yasser Tabbaa)

seventh century.7 On the other hand, according to numerous hadiths, men in the Qurʾånic paradise will all be beardless youths who will remain perpetually thirty-three years old.8 This is a little-noted but important point to keep in mind as we investigate the bathing and hunting panels to the right of the Six Kings and elsewhere in the palace. Are there other indications in the adjacent bathing and hunting panels of a paradisiac vision to which the Six Kings beckon the seated prince? The centre of this panel is occupied by a tall woman, naked except for a bikini bottom and arm bracelets, standing at the edge of a pool that seems to be located within a colonnaded courtyard in two levels or possibly a domed atrium (Figure 18.7). She is watched from an arched balcony by a large group of featureless men and assisted by a dressed woman, while a third woman peers at her from behind a curtain. Her size, centrality, nudity and physical beauty all suggest that she was more than just a concubine but perhaps a prized possession of the prince, combined, according to Fowden, with iconographic references to standing figures of Aphrodite, goddess of love, seduction and procreation.9 Her position in the middle panel seems to mediate between the formality of the Six Kings panel and the complete abandon of the panel to its right, which shows a group of eight men, all wearing swimsuits, engaged in acrobatics and other sporting activities. They perform on a rather barren landscape, with some indication of greenery on the lower left corner, underneath a leafy grape vine that stretches from just behind the rightmost column of the previous panel all the way to the edge of the acrobats’ panel. The same blue sky behind the Six Kings reappears behind the acrobats, providing the incongruous triptych with some formal continuity. Scenes of hunting and related activities stretch the full length of the upper western wall and are also found on the northern wall of the

429

430

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 18.6  Qusayr ʿAmra – detail of onagers (photo: Yasser Tabbaa)

Figure 18.7  Qusayr ʿAmra – reconstruction of frescos on west (source: C. Vibert-Guigue)

west aisle and the northern and southern walls of the eastern aisle, dominating the iconographic programme of Qusayr ʿAmra (Figure 18.7). The panel on the western wall, seven metres long, depicts two horsemen corralling onagers, wild donkeys, towards a roped

eternal hunting fields

Figure 18.8  Taq-e Bostan, fifth century ad

and fenced enclosure, as a group of men, partly hidden behind black banners, raise torches intended most likely to steer the terrified animals further into the enclosure. The panel on the southern wall of the eastern aisle shows hunters, led by a bearded man, killing onagers with spear or sword, while the panel facing it on the northern side shows men skinning and cutting up the carcasses. All the hunting panels are characterised by a quick sketching technique, visceral realism and complete focus on the action, with minimal suggestion of landscape. Although there are no immediate formal or iconographic parallels to these scenes of hunt and gore, they were quite likely based on ancient Near Eastern hunting representations, of which Sassanian art provides the most likely precedent. With the exception of numerous abbreviated hunting scenes on silver objects, the only preserved Sassanian monumental hunting cycle is the rock relief at Taq-e Bustan, datable to the period of Khosrow II (590–628). A vast iwån carved into a sheer cliff, the rear wall shows the investiture of Khosrow II by Ahura Mazda and most likely the goddess Anahita (Figure 18.8). He is haloed and surrounded above by vegetal scrolls and honoured by the goddess of fortune Nike. Both sides of the iwån are carved in relief sculpture in a more minute style that, with its original paint, seems to hearken to a prototype in fresco painting. The entire composition revolves around a boar hunt taking place in the marshes, with the Sassanian shah, presumably Khosrow II, as the royal hunter (Figure 18.9). The shah is twice represented, first as hunter in a boat shooting arrows at wild boars and, to the right after a successful hunt, as an exalted shah being serenaded by female singers and musicians. He is now possessed

431

432

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 18.9  Taq-e Bostan Royal Boar Hunt, fifth century ad

with the nimbus of farr or khwarna, a state of royal good fortune that is liminally located between this world and the afterworld. Striking formal and iconographic parallels between the central themes of Qusayr ʿAmra and Taqi-e Bostan can be discerned. Both show an exalted sovereign situated at the back of an iwån surrounded by courtiers and naturalistic and abstract foliage. Extensive scenes of hunting and merry-making unfold around these sovereigns, adding a narrative dimension to their otherwise emblematic representation. The hunting scenes take place within a well demarcated enclosure: in Qusayr ʿAmra at a desert corral to which the animals had been driven; and in Taq-e Bostan at a royal hunting enclosure that is surrounded by a natural and man-made fence. The scurrying animals – onagers in Qusayr ʿAmra and wild boars in Taq-e Bostan – are represented in a similarly overlapped manner that suggests action and multitude. These analogies and departures do not account for the entirety of the iconographic programme of either Umayyad or Sassanian Qusayr ʿAmra or Taq-e Bustan, but they do suggest, I would argue, that the main fresco cycle at the Umayyad palace was based on this Sassanian model or one of its vanished cognates. But the famous Sassanian boar hunt in the marsh has been adapted in ʿAmra to desert conditions, and its iconography seems to have been transformed in ways that are more in keeping with an emerging Islamic tradition. Most significant is the absence of the royal hunter,

eternal hunting fields

al-Walid II himself, from all scenes of hunt, a curious departure from the Taq-e Bostan reliefs where the Sassanian Shah is represented, both in the iwån and twice in the hunting reliefs. Unlike the active role played by the Sassanian Shah as the royal hunter and as a divine king possessed with the nimbus of farr, the Umayyad prince, seated in his privileged position, passively observes the hunt and other sensuous and sporting events.10 Rather than being invested by goddesses, the Umayyad prince is being welcome into the world of kings, a world of hunting, bathing and women. Or is it the afterworld? Secular paradise and religious paradise Is al-Walid II being accepted into the ‘Family of Kings’, as Oleg Grabar long suggested, or is he being beckoned to the hunting fields of paradise by royals who had preceded him there? Or does the composition engage both ideas – royal and paradisiac – exalting the Umayyad prince in an earthly paradise that is continuous with the heavenly paradise? Interestingly, a similar iconographic overlap has long been proposed for the Barada mosaics at the Great Mosque of Damascus, which have been read both as a literal image of paradise and as the royal domain of the Umayyads. It is quite likely that the paradisiac mosaics at the Great Mosque of Damascus were known to al-Walid II, and there are in fact intriguing formal and iconographic parallels between the two structures. The most interesting passage of the mosaics of the Great Mosque of Damascus is the so-called Barada panel, named after the river that traverses Damascus from west to east. Extending nearly the full length of the western side of the courtyard, this mosaic panel rises about five metres above the mosque pavement from a zone that is revetted in quartered marble panels and articulated with pilasters that support an inscriptional frieze and extend above it all the way up to the border of the panel (Figure 18.2). The panel itself is surrounded on all four sides by a rich border of continuous circles that contain a radial pattern within them, acting as a kind of visual barrier between the realm of the viewer and the paradisiacal realm above. The composite panel on the western wall of Qusayr ʿAmra is similarly arrayed. Both registers of the panel are enclosed within a painted border of tangential circles that enclose a radial pattern and the entire panel is raised above four pairs of painted pilasters that enclose between them an alternation of rosettes and painted quartered marble. In other words, the artist has translated the costly mosaics and quartered marble at the Great Mosque of Damascus into painted ornament and faux marble. But the representations of paradise above this dado could not be any more different. The riparian, serene, luxurious and uninhabited paradise of the Great Mosque of Damascus – clearly based on Qurʾånic descriptions – has been replaced by an arid, lively, sensuous and quite human-centred

433

434

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

r­epresentation of paradise. Instead of the deferred promises of ‘gardens beneath which rivers flow’, the paradise at Qusayr ʿAmra directly presents us with a sovereign who shares with us his vision of paradise, a vision that juxtaposes al-Walid II’s worldly princely pursuits against his otherworldy aspirations as an exalted ruler in the world of kings. Literary parallels If the literal representation of paradise at the Great Mosque of Damascus is clearly inspired by Qurʾånic descriptions, what would have been the literary subtext of the frescos at Qusayr ʿAmra? Not only are there no Qurʾånic parallels to al-Walid’s vision of a pastoralist paradise, but the Qurʾånic descriptions strive to distance the Islamic paradise from the harsh desert environment of Arabia and especially from its nomadic inhabitants. Lacking deserts and oases, pastoralists and camels, hunters and preys, this paradise was intended for a sedentary population not the Bedouins, the majority of Arabia’s population.11 A harsh, uncontrollable and godless space, the desert is even equated in the Qurʾån with hell: ‘on the day the earth and mountains will rock violently, and the mountains turn to a heap of poured out sand’ (73:14).12 As for the Bedouins (al-Aʾrab), they are invariably impugned for their untethered lifestyle, obstinate attachment to idols and fickle faith.13 Prevented from transhumance, dispossessed of their animals and barred from hunting, for the Bedouin life in the Qurʾånic paradise would have been a life without freedom, courage or dignity. But the Qurʾånic dismissal of the desert ethos stands at odds from its deep appreciation by pre-Islamic and Umayyad poets, whose qasidas (odes) embraced it for its tribal cohesion, unaffected morality and free lifestyle. At the heart of these odes was the hunt wherein the poet praises his mount and the elegance and alacrity of his prey – whether gazelle, onager or oryx – and his own abilities in chase and hunt. Although the qasida achieved its peak in the pre-Islamic period, the form continued well into the Umayyad period, but with greater emphasis on sensuous pleasures and on panegyric. Countless lines were composed on wine, women and the hunt, particularly by the poets in the court of al-Walid II:14 We caught and would have killed an antelope That ran auspicious from the right But then it gently turned its eyes and looked – The very image of your look! We let it go. And know but for our love For you, it surely would have died. Now, little antelope, you’re free and safe. So off you go, happy among other antelopes.

eternal hunting fields

One of Walid’s favourite drinking songs ends with the following lines: So my hereafter’s sure: no fire for me! I’ll teach The folk to ride a donkey’s penis! Tell him who looks for heaven to run along to hell!15 Sung and performed at al-Walid II’s courts in Qusayr ʿAmra and Khirbat al-Mafjar, these irreverent poems resonated with frescos and statuary of royalty, nudity, sportsmanship and the hunt to produce an immersing sense of well-being with subtle allusions to a similar afterlife. A pastoralist paradise The paradisiac iconography of Qusayr ʿAmra or even Khirbat al-Mafjar seem informed less by Qurʾånic descriptions and much more by a curious, even brilliant, adaptation of Sassanian royal iconography as refracted through the lens of Arabic poetry. Although the figural imagery in these palaces does not present a consistent iconographic programme of paradise, it does offer tantalising images – drawn from contemporary poetry, interwoven with Sassanian iconography and sustained by al-Walid’s eccentric vision – of an afterworld seen through the eyes of the first Arab-Islamic royal dynasty. Shunning the tedium of the enclosed garden, Umayyad royal iconography instead embraces a more active vision of paradise, liminally positioned between the here and the hereafter, with hunt at its centre. This vision may seem aberrant or marginal to our eyes, which have become accustomed to the more prevalent and dominant ‘Persian’ tradition of paradisiac representation. But in other respects, the pastoralist vision of paradise I have attempted to illustrate in this piece could in fact be truer to an Arab-Islamic ethos that had persisted despite vigorous attempts to suppress it. Notes 1. See, for example, Jonas Lehrman, Earthly Paradise: Garden and Courtyard in Islam (University of California Press, 1980); and more recently, Emma Clark, The Art of the Islamic Garden (The Crowood Press, 2004). 2. To my knowledge, there is not a single tafsir that presents the Qurʾånic descriptions of paradise as an actual garden or that suggests that these descriptions contributed to the creation of earthly gardens. On the other hand, see D. Fairchild Ruggles, Islamic Gardens and Landscapes (Philadelphia: University of Pennsylvania Press, 2008), x where the author dismisses the existence of paradisiac symbolism in early Islamic gardens and p. 89, where she proposes that the Qurʾånic description of paradise ‘reflects a preexisting vocabulary of garden forms’.

435

436

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

3. Indeed, that was the conclusion I reached nearly two decades ago in discussing the Madrasa al-Firdaws in Aleppo in Constructions of Power and Piety in Medieval Aleppo (Penn State Press, 1996), pp. 181–2. 4. See, for example, Yasser Tabbaa, ‘Typology and Hydraulics in the Medieval Islamic Garden’. In Landscape and Garden History: Issues, Approaches, Methods, ed. J. D. Hunt. (Washington, DC: Dumbarton Oaks, 1992), pp. 303–29. 5. Richard Ettinghausen, From Byzantium to Sassanian Iran and the Islamic world: Three modes of artistic influence (Brill, 1972). Garth Fowden, Qusayr ‘Amra: Art and the Umayyad Elite in Late Antique Syria (University of California Press, 2004). 6. M. Almagro et al., Qusayr ‘Amra: Residencia y banos omeyas en el desierto de Jordania (Madrid, 1975). Oleg Grabar, ‘The Painting of the Six Kings at Qusayr Amra’, Ars Orientalis 1 (1954): 184–7. 7. (last accessed 12 July 2012). See also Richard Ettinghausen, Arab Painting (Skira, 1962), 32. 8. (last accessed 27 October 2002). 9. Fowden, 230–7. 10. For further discussion of the role of the sovereign as a privileged observer, see William Tronzo, The Cultures of His Kingdom: Roger II and the Cappella Palatina in Palermo (Princeton University Press, 1997): 145–8. 11. Kevin Reinhart, ‘The Here and Hereafter in Islamic Religious Thought’, in Blair and Bloom (eds), Images of Paradise in Islamic Art (Dartmouth, Hood Museum, 1991), 18. 12. Christine Jo Dykgraaf, Metaphorical and Literal Depictions of the desert in the Qur’an. ag.arizona.edu/oals/ALN/aln50/dykgraaf.html 13. Qurʾån 9:101: ‘Some of the Arabs of the desert around you are hypocrites . . . we will punish them twice, and they will be sent to a harrowing doom.’ 14. Robert Hamilton, Walid and his Friends: An Umayyad Tragedy (Oxford University Press, 1988): such as 20, 35, 122–5. 15. Hamilton, quoting Isfahani, Kitab al-Aghani vi, 119–20.

PART VI  SHRINES

CHAPTER NINETEEN

Invented Pieties: The Rediscovery and Rebuilding of the Shrine of Sayyida Ruqayya in Damascus, 1975–2006 In the fall of 1979 I was doing a field survey of twelfth-century madrasas in Damascus for my dissertation on the architectural patronage of NËr al-DÈn.1 Winding my way through the alleys between the Great Mosque and the northern gate of Båb al-Amåra (also known as Båb al-FarådÈs), I came upon the still impressive remains of the Madrasa al-Muqaddamiyya, founded in 563/1168, which consisted of a substantial courtyard, centred around a large pool and surrounded by two iwåns and a prayer hall (Figure 19.3). The remains of another twelfth-century madrasa, al-Mujåhidiyya al-Juwwåniyya,2 stood north of it, while in between rose a small, unornamented dome with an inconsequential entrance that bore a very late inscription identifying it as Sayyida Ruqayya.3 Overall, this quarter, called al-ʿAmåra, displayed the rich cultural accumulation and profound neglect characteristic of the walled city of Damascus before its more recent gentrification (Figure 19.4). There were rumours circulating at the time that the quarter of al-ʿAmåra was slated for tanzÈm, an ostensibly benign term that frequently entailed the destruction of historic and traditional buildings in the name of modernisation. A more insistent rumour, with conspiratorial undertones, speculated that ‘Iranians’ were buying up properties around Sayyida Ruqayya as part of a project to turn it into a major Shiʿi shrine (Figure 19.5). I immediately dismissed this rumour, since registered medieval buildings, such as the madrasas al-Muqaddamiyya and al-Mujåhidiyya, were protected by the Laws of Antiquities and, being intramural, were further protected by the Law for the Protection of the Ancient City of Damascus. But in fact, that is precisely what has happened: both of these madrasas and several other traditional residences were swallowed up by the expanded shrine of Sayyida Ruqayya, an ongoing process that Yasser Tabbaa (2007), ‘Invented Pieties: The Revival and Rebuilding of Shi‘ite Shrines in Contemporary Syria’, in Linda Komaroff (ed.), Festschrift for Priscilla Soucek, Artibus Asiae, 142–71.

440

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

has already completely transformed the architectural character of al-ʿAmara quarter and even modified its ethnic and sectarian makeup, from SunnÈ Arab to Shiʿi Iranian. In the intervening twenty-five years, a similar process would be repeated in several sites all over Syria (and at least one site in Lebanon), wherein relatively unknown Shiʿi shrines would be greatly expanded and lavishly refurbished, largely under Iranian patronage. The shrine of Sayyida Ruqayya, the subject of this article, is therefore part of a much broader phenomenon, which is the subject of an ongoing study.4 I present this article to Priscilla Soucek because it evokes for me the time when I came to know her, first as a supervisor and then as a colleague. Furthermore, I hope that this article, which highlights a little known dimension of Shiʿi Iranian culture, will resonate with Professor Soucekʾs lifetime commitment to Iranian art. The newness, extent and unprecedented success of the Shiʿi shrines of Syria, in particular the Sayyida Ruqayya, raise several important factual, analytical and interpretive questions. Since so little is known about these shrines, it might be useful, first, to give a short survey of their location and the saintly figures they commemorate and to examine the legends surrounding their ‘rediscovery’ and revived piety. As for the Sayyida Ruqayya specifically, this piece will examine its quite intrusive architectural form, excessive ornament and numerous inscriptions and investigate the factors and actors behind its ongoing expansion and astounding success. Finally, the piece will examine the prevailing controversy in Syria with regard to these shrines and our own position as art historians in face of the radical disruption to architecture, urbanism and even historical memory presented by these shrines.5 A brief survey The project on which this paper is based will ultimately deal with about ten large and twenty smaller Shiʿi shrines that have been rebuilt in Syria between about 1980 and 2006. Proceeding from southern to northern Syria, we encounter first the maqåm of Sukaina bint ʿAlÈ in Dårayya, a small town about ten kilometres south-west of Damascus. This shrine was rebuilt once in the early 1980s and is currently being entirely rebuilt on a gigantic scale (Figure 19.6). Next is the maqåm of Hijr b. ʿAdÈ al-KindÈ, located in ʿAdhra, about twenty kilometres east of Damascus. About five kilometres east of Damascus is the maqåm of Saʿd b. ʿUbada in the village of MelÈ˙a or ManÈ˙a.6 By far the most famous Shiʿi shrine in Syria is the maqåm of Sayyida Zaynab in the erstwhile village of Rawiya, about five kilometres south of Damascus. Twice rebuilt and refurbished in the last three decades, it is now the focus of Shiʿi pilgrimage to Syria and a thriving cultural centre (Figure 19.7).7 Damascus proper possesses three clusters of Shiʿi shrines: the maqåm of Sayyida Ruqayya, the

invented pieties

maqåm of Raʾs (head of) al-Óusayn in the Great Mosque,8 and at least a dozen smaller shrines in the Bab SaghÈr cemetery, including those of Sukaina bint ʿAlÈ, Få†ima al-Sughra bint al-Óusayn, Umm KulthËm bint ʿAlÈ, and the group shrine for RuʾËs Shuhadåʾ (heads of the martyrs of) Karbala (Figure 19.8). There is a shrine outside Homs dedicated to AbË Dharr al-GhifårÈ, while Baʿalbak has a rebuilt shrine for a daughter of Imam al-Óusayn. Aleppo has two important medieval shrines to Imam Mu˙assin (or Mu˙sin) and Imam al-Óusayn, of which the latter has been greatly rebuilt and expanded in recent years.9 Finally, an enormous triple shrine to ʿAmmår b. Yåsir, Uways al-QaranÈ and Ubayy b. Kaʿb was just completed in Raqqa, near the site of the famous battle of Siffin between the forces of ʿAlÈ and Muʿåwiya (Figure 19.9). Long in planning and construction because of SunnÈ and tribal opposition, this shrine perhaps best represents the extent of Shiʿi triumphalism and state support in Syria.10 Sacralisation and history The sudden appearance or transformation of these shrines raises questions about the process of their sacralisation. How do Shiʿi shrines become sacred? What physical remains, historical evidence, or pious tales were required to infuse sanctity into these particular spots to prompt their future rebuilding? Generally speaking, in Shiʿi Islam the process of discovery and sanctification of a shrine is not based on verifiable material evidence, such as relics, but is often the product of a dream or a vision, which is then subsequently validated by some form of consensus.11 The pretext for building the mashhad al-Óusayn in Aleppo in the late twelfth century, for example, was a dream that a certain shepherd had while napping at the foothill of Jabal Jawshan, about one kilometre west of the city wall (Figure 19.10). In his dream, the shepherd saw that one of his goats had sunk into the solid rock, upon which he was approached by a holy man who directed him to have a shrine for al-Óusayn built at that spot. When the shepherd awoke, he pulled the goat from the rock and a spring of water gushed forth. This event was deemed sufficient proof of the sanctity of the location, and the shrine was begun shortly thereafter.12 Whereas the history of Shiʿi shrines such as the mashhad alÓusayn in Aleppo and the Sayyida Zaynab south of Damascus go back at least to the twelfth century, the shrine of Sayyida Ruqayya is a much more recent creation whose pious justification is quite difficult to pinpoint. Despite its current fame, the shrine was not mentioned by any of the topographers of Damascus, including Ibn ʿAsåkir in the twelfth century, Ibn Shaddåd and YåqËt in the thirteenth century and Ibn ʿAbd al-Hådi in the fifteenth century.13 Furthermore, it is also ignored by all Orientalists of the late nineteenth and early twentieth centuries, including Sauvaire, Wultzinger and Watzinger,

441

442

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 19.1  Aleppo. Mashhad al-Husayn, 1183–96, courtyard from south-west (photo: Yasser Tabbaa)

Sauvaget, and Kurd ʿAlÈ, an omission that suggests that it was not an important structure (Figure 19.11).14 Most recently, Akram al-ʿUlabi, the foremost historian of Damascus, proposed that the shrine is a complete fabrication, as ‘there is no trace of Ruqayya in all Damascus’.15 Even the prominent Shiʿi historian Sayyid Mu˙sin al-AmÈn was dubious about the authenticity of this shrine, following his short entry on it by deferring judgement to God: ‘WʾAllah aʿlam bi-ßi˙˙at hådha al-ʿamr.’16 Fortunately, a short description of the shrine of Sayyida Ruqayya, reflecting its condition around 1970, gives us some idea of its earlier history and form before its recent rebuilding: It is an attractive masjid with a striped stone portal made in 1323/1905 that leads to a short hall with two doors, one leading to the shrine and the other to the servant’s residence. Nothing is noteworthy in the masjid except for three inscribed stone panels near the mihrab. The first includes some ˙adÈths on the merits of Ahl al-Bayt, followed by: ‘The dignified Mirza Båbå al-KaylanÈ has restored the spot known as maqåm Sayyida Ruqayya, the daughter of our Master ʿAlÈ and the site of Raʾs al-Óusayn . . . in the year 1125/1714.’ On the second: ‘This blessed location contains the burial of . . . in the year 880/1475 . . . The masjid has an ordinary mi˙råb and minbar. As for the dome of the shrine, it resembles Mamluk domes, but it has been renewed and painted. The cenotaph

invented pieties

has an ornamented copper frame, beside which is a cupboard that contains a rock, said to be the footprint of the Prophet . . . The third stone contains four lines of poetry, which I was not able to read.’ Although inconclusive, this description suggests, first, that the shrine has a long history, dating back at least to the fifteenth century. Second, although the shrine seems to have been Shiʿi for several centuries, it was not identified with Ruqayya. In fact, it is impossible to determine when exactly this shrine came to be identified with her. Finally, there is little doubt that nothing of the original shrine has been preserved in the new shrine. The new shrine of Sayyida Ruqayya Seen from the Miʾdhanat al-ʿArËs, the northern minaret of the Great Mosque of Damascus, the main dome of Sayyida Ruqayya presents quite an intrusive and ornamented form within the generally dreary exteriors of Damascene architecture. Its high drum, bulbous shape and white tiles with black ornamental and calligraphic details clearly distinguish it from other Damascene domes, including that of the Great Mosque itself, and project a deliberately Iranian image (Figure 19.12). On the other hand, the dome of Sayyida Ruqayya is much less flamboyant than the golden dome of Sayyida Zaynab or the turquoise domes of the Raqqa shrine. This might represent some form of compromise in view of the intramural location of the Sayyida Ruqayya and its proximity to the Great Mosque of Damascus.17 Seen from the narrow street that leads to the covered SËq al-ʿAmåra, the façade of the Sayyida Ruqayya unfolds in a series of setbacks that simultaneously conform to the diagonal street while also projecting a measure of unity and monumentality (Figure 19.13). This sense of unity is achieved through the white stone facing of the walls and the long inscriptional band that crowns them and wraps around the entire structure. Two iwån entrances and a third in the northern wall punctuate this long wall and give it a sense of monumentality and focus. All three entrances are covered in tilework with floral ornament and inscriptions and topped by tiled muqarnas vaults, a luxurious and welcoming display that stands out among the surrounding drabness. The plan of the shrine, somewhat like its façade, attempts a high degree of symmetry within a truncated triangular plot (Figure 19.14). This symmetry is based on an axis that extends from the main portal, through the middle of an open courtyard, to the northern end of the shrine. Two vaulted squarish spaces of nearly equal size flank this axis, which takes on the form of a corridor, while a third vaulted space, off-axis, occupies the southern end of the chamfered triangle. This axiality is further emphasised by two iwåns that define the middle points of the rectangular courtyard.

443

444

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 19.2  Damascus. Shrine of Sayyida Ruqayya, rebuilt 1988–95, main courtyard (photo: Yasser Tabbaa)

An addition to the eastern side of Sayyida Ruqayya, almost complete as of February 2006, will nearly double the size of the shrine (Figure 19.15).18 It consists primarily of two large square spaces, of which the northern one is a large domed mosque that will be linked to the first mosque, while the southern one is a courtyard with arched porticos that addorses the original courtyard. The mosque is entered from the middle of its northern side through a monumental iwån entrance that is perfectly axial with the mi˙råb and the centre of the courtyard, further enhancing the symmetry of the entire shrine. The southernmost entrance to the Sayyida Ruqayya, the only one currently open, leads through a triangular vestibule to a spacious courtyard that, like the rest of the shrine, is oriented to the cardinal directions.19 Paved in white and black marble and surrounded on all four sides by colonnaded porticos, the immaculately clean courtyard appears both serene and luxurious (Figure 19.16). More Syrian than Iranian, the slightly pointed horseshoe-shaped arches are arranged in groups of three, with a tall central arch flanked by two smaller ones. The four corners of the courtyard are defined by small cupolas, and the middles of its longer sides are dominated by tall iwåns, one reflecting the entrance iwån and the other leading to the shrine. An inscriptional frieze surrounds the courtyard and crowns both iwåns, while numerous inscriptional panels are inserted in the walls shaded by the porticos. The northern iwån leads to a vaulted hall, to the right of which is the mosque and to its left stands the dharÈ˙ (cenotaph) of Ruqayya, which is also entered directly through a magnificent golden door at

invented pieties

the north-eastern corner of the courtyard. Every wall, pier, vault or opening within these spaces is extravagantly decorated with polychrome tiles, gilt and enamelled copper, and numerous inscriptions of exceptional quality and legibility (Figure 19.17). But the chamber of the cenotaph, as is typical of all Shiʿi shrines, far exceeds and easily overshadows the surrounding extravagance. At its centre is a massive cenotaph whose golden grill is crowned by an enormous baldachin with gilt and polychrome enamel decoration (Figure 19.18). Directly above the cenotaph is a stupendous dome which is covered with mirror-inlaid muqarnas (Figure 19.19), a technique that had been used in Shiʿi shrines since the eighteenth century and can today be found in most Shiʿi shrines, including Najaf, Karbala, Kazimayn, Mashhad and Qumm. The overall effect is exhilarating and hypnotic, quite likely intended to create a feeling of displacement and transcendence among the pious visitors. In addition to its dazzling Shiʿi forms and ornament, the Sayyida Ruqayya projects a somewhat more subtle Shiʿi message through its numerous inscriptions. These inscriptions surround the shrine both externally and internally, encircle the domeʾs drum and its springing, and ornament the walls at different levels with about three dozen inscribed cartouches and medallions. There are also numerous inscriptions on the massive gate to the cenotaph and on the cenotaph itself. Such heavy reliance on inscriptions can in fact been seen in most of the Syrian Shiʿi shrines, including Sayyida Zaynab and the triple shrine in Raqqa.

Figure 19.3  Damascus. The Madrasa al-Muqaddamiyya, courtyard (now destroyed) (photo: Yasser Tabbaa)

445

446

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 19.4  Damascus. Shrine of Sayyida Ruqayya and unknown dome, 1978 (photo: Yasser Tabbaa)

Figure 19.5  Coloured cadastral map of region around Sayyida Ruqayya (courtesy of the Department for the History of Damascus)

invented pieties

Figure 19.6  Darayya. Shrine of Sayyida Sukaina bint ʿAlÈ (condition in early 2006) (photo: Yasser Tabbaa)

Figure 19.7  Damascus. Shrine of Sayyida Zaynab, exterior (photo: Yasser Tabbaa)

A thorough study of these inscriptions is outside the confines of this article, but a brief account of their content will hopefully shed a light on the overall iconography of the shrine. In terms of content, the inscriptions are of three different types: Qurʾånic verses, hadith and various invocations (duʿåʾ), of which some are in Persian (Figure 19.20). The Qurʾånic verses include some common verses, such as

447

448

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 19.8  Damascus. Bab Saghir cemetery, double shrine of Få†ima and Sukaina bint al-Óusayn (photo: Yasser Tabbaa)

Figure 19.9  Raqqa. Model of the shrine complex of ʿAmmår b. Yåsir, Uways al-QaranÈ and Ubayy b. Kaʿb (photo: Yasser Tabbaa)

the verse of the Throne (Ayat al-KursÈ) around the domeʾs drum; the verse of Light (Surat al-NËr) at the springing of the dome, and other somewhat less common short verses, including Surat al-Kawthar (a spring of Paradise), above the main entrance. But the inscriptions here and in other Syrian Shiʿi shrines also include Qurʾånic verses that have historically been used predominantly in Shiʿi buildings (Figure 19.21). These verses, that include 33:33 and 76:1–2, refer to ahl al-bayt (namely ʿAlÈ and his

invented pieties

Figure 19.10  Aleppo. Mashhad al-Óusayn, c. 1910 (collection E. Poche)

Figure 19.11  Damascus. Map of walled city (from Wulzinger and Watzinger)

­ escendants), emphasising their purity (tahåra) and their God-given d right to lead the Muslim community. One of the most frequent and evocative of these inscriptions is 5:28, which speaks of the reward of legitimate rule that God will bequeath upon those who have been oppressed in the world, an allusion that Shiʿis have long believed refers to their own community.20

449

450

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 19.12  Damascus. Shrine of Sayyida Ruqayya, aerial view (photo: Yasser Tabbaa)

Figure 19.13  Damascus. Shrine of Sayyida Ruqayya, street view (photo: Yasser Tabbaa)

Whereas the Qurʾånic verses have both a pan-Islamic sense and a specifically Shiʿi intent, the hadith selection in the Sayyida Ruqayya and other Syrian Shiʿi shrines is exclusively Shiʿi. In fact, rarely does one come across such clear and emphatic assertions of the basic tenets of Twelver Shiʿism, assertions that are made more forceful by the eye-level location of these inscriptions and their easily legible

invented pieties

Figure 19.14  Damascus. Shrine of Sayyida Ruqayya, plan of first expansion (courtesy of Muhammad al-Suß)

Figure 19.15  Damascus: Shrine of Sayyida Ruqayya, plan of second expansion (courtesy of Muhammad al-Suß)

451

452

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 19.16  Damascus. Shrine of Sayyida Ruqayya, courtyard (photo: Yasser Tabbaa)

Figure 19.17  Damascus. Shrine of Sayyida Ruqayya, central hall (photo: Yasser Tabbaa)

calligraphic style (Figures 19.20–19.21). Surrounding the courtyard and the shrine like icons in an Orthodox church, these calligraphic medallions emphasise the perennial concepts of Imami Shiʿism, including the God-given legitimacy of ahl al-bayt and their special role as the safeguards of Islam. Others specify the exact number and order of the twelve Shiʿi imams, while still others emphasise the concept of walåʾ, the special relationship of association between ʿAlÈ and God.21 Those around the courtyard largely focus on the spiritual merits of Ruqayya. The combined effect of the stupendous ornament and the highly selective and forceful inscriptions create an atmosphere of luxury and directness, of religious texts made accessible through the ­mediation

invented pieties

of ornament. It is a specifically Shiʿi sense of decorum in which the word of God and the traditions of Mu˙ammad and ʿAlÈ are infused with colourful and sparkling patterns, much like the cenotaphs of the Imams, during the ʿAshËra processions are covered in brilliant silks and satins.22 It is an aristocratic rather than an egalitarian heritage, in which the princes and sultans are none other than the Imams themselves.23 Iran in Syria Why has Iran, the largest Shiʿi country in the world, been so intent on building Shiʿi shrines in a country, Syria, with a very small Shiʿi population of its own?24 How and why was this process facilitated? Even more curiously, how do we explain the astonishing success of these shrines in attracting hundreds of thousands of visitors from Iran and the Arab Figure 19.18  Damascus. Shrine of Sayyida Ruqayya, cenotaph (photo: world? To begin with, it should be emphasised Yasser Tabbaa) that unlike their equivocal and highly contested status in SunnÈ Islam, shrines and visitation are central to Shiʿi ritual practice and spiritual needs. As a faith that rests on the remembrance and commemoration of the death and ­martyrdom

Figure 19.19  Damascus. Shrine of Sayyida Ruqayya, dome (photo: Yasser Tabbaa)

453

454

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 19.20  Damascus. Shrine of Sayyida Ruqayya, inscriptions on door to cenotaph (photo: Yasser Tabbaa)

of Shiʿi imams and their various male and female descendants, Shiʿism has long sanctioned and recommended the visitation (ziyåra) of the tombs of these saints. According to Yann Richard, ‘the merit gained from this pilgrimage visit comes from the fact that one has torn oneself from the daily round and traveled a long road to . . . share in the pains [of the imams], weep with them, and together with them wage a symbolic battle against the forces of evil’.25 In other words, travel for ziyåra or pilgrimage is a natural condition for Shiʿis, and the rise of a new shrine is often sufficient incentive in and of itself to visit it. Historically, the Shiʿi shrines of Syria were mainly visited by the Shiʿis of Jabal ʿAmil in south Lebanon, whose visitation of the shrine of Sayyida Zaynab was sometime called Hajj al-fuqaråʾ (the pilgrimage of the poor).26 For the majority of Shiʿis, however, the ultimate pilgrimage destination is none other than the Iraqi shrines of Imam ʿAlÈ in

Figure 19.21  Damascus. Shrine of Sayyida Ruqayya, inscription on domeʾs drum (photo: Yasser Tabbaa)

455

invented pieties

Figure 19.22  Damascus. Shrine of Sayyida Ruqayya, hadith inscriptions (photo: Yasser Tabbaa)

Figure 19.23  Damascus. Shrine of Sayyida Ruqayya, hadith inscriptions (photo: Yasser Tabbaa)

Najaf and the shrine of Imam al-Óusayn in Karbala. Iraq also possesses other important shrines, including al-Kazimayn in Baghdad and al-ʿAskari in Samarra. Iran does have two important shrines – the shrine of Imam Reza (ʿAlÈ al-Ri∂a) in Mashhad and that of his sister Få†ima al-MaʿsËmah in Qumm. But these rank below the shrines of Najaf and Karbala, whose visitation is a central component of Shiʿi piety, and burial in their grounds merits great rewards in the afterlife. The Iran–Iraq war, which began shortly after the Iranian Revolution of 1979 and continued throughout the 1980s, put a stop to the age-old Iranian pilgrimage to the sacred shrines in Iraq. It should also be noted that the Ayatullah Khomeini utterly despised Saddam and his Baathist regime while showing a great deal of affection for Syria, the only Arab country to stand beside him during the Iran–Iraq war and the one to accept the remains of ʿAlÈ ShariatÈ, who is in fact buried on the grounds of Sayyida Zaynab. Equally motivated by the friendly overtures of the Syrian regime and by the persisting animosity between Iran and Iraq, Khomeini and successive Iranian governments have embarked since the late 1970s on a campaign of shrine building in Syria. This was followed in succeeding years by planned tours, initially made for war victims,

456

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 19.24  Letter signed by President Hafez Assad (Archive of Ministry of Awqaf)

but has more recently turned into well-organised junkets with itineraries advertised on various websites. In short, the rebuilding of the Shiʿi monuments of Syria was at least partly motivated by the desire to provide alternatives to the great shrines of Najaf, Karbala and Baghdad, whose visitation had become practically impossible during the decade of the1980s. Although the prevailing political situation during the Iran–Iraq war generally explains the quick rise of Sayyida Ruqayya and other Syrian Shiʿi shrines, there is evidence to suggest that the process of

invented pieties

r­approchement between Syria and Iran actually pre-dates this war and the Iranian Revolution by several years. Beginning in the early 1970s, this process was intended to foster relations between the Syrian Alawites and the larger Shiʿi community, and it seems likely that negotiations for the development of the shrines began then.27 Indeed, a document I have recently examined in Damascus, signed by President Hafez Assad and dated to September 1974 (Figure 19.24), clearly mandates the purchase of twenty-three properties around Sayyida Ruqayya for ‘the expansion of the shrine of Sayyida Ruqayya and the creation of adjacent cultural institutions’. It seems obvious, therefore, that the Law of Antiquities and the protected status of the walled city of Damascus, mentioned above, were superseded by a much greater authority, that of the president of Syria, Hafez Assad himself. Although the original motivation for the creation of these Syrian shrines may well have been political, there is little to suggest that their numerous visitors continue to be motivated by Khomeiniʾs vitriolic politics. Rather, most pilgrims visit these shrines largely for pietistic and financial reasons: to honour the memory of Ahl al-Bayt and to engage in commerce. Gender has also emerged as a significant issue in the three main shrines for female saints in Damascus, which now attract a disproportionate number of female visitors, Shiʿis and SunnÈs alike.28 These women seek the intercession of these female saints for various familial problems, including marriage, conception and cure. In short, the shrines and their visitors have created their own meanings, and this ongoing interaction will continue to transform and enrich these meanings. Views on the shrines These shrines, particularly the Sayyida Ruqayya, have produced quite polarised opinions in Syria, ranging from vocal condemnation to silent acceptance, depending on the sectarian, political and economic background of the author. Not surprisingly, the Directorate General of Antiquities, which adheres to an antiquarian, even Orientalist view of culture, has consistently opposed the expansion of Sayyida Ruqayya. Several letters from Dr Afif Bahnassi, then the Director General of Antiquities and Museums, to the Prime Minister outline the various problems that the expansion of this shrine would create for the traditional fabric of walled Damascus.29 Similarly, articles published in the early 1980s in the newspapers Tishreen and al-Thawra strongly criticise this building and even speak of Iranian encroachment (taghalghul irånÈ).30 These sentiments were reiterated by Akram al-ʿUlabi, who condemned the shrine of Sayyida Ruqayya as ‘a terrifying (mukhÈf) concrete structure’, adding that ‘the region has become today an important site for Shiʿite pilgrimage, particularly the Shiʿites of Iran, who were the ones to execute this project’.31 Somewhat more temperately, Sabrina Mervin proposed

457

458

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

that the renovation and expansion of the shrine of Sayyida Ruqayya was directly and entirely funded by the Iranian government with the co-operation of the Ministry of Awqåf. This, according to her, led to the destruction of an entire quarter, thereby provoking the silent grumbling of the inhabitants of the old city.32 It seems clear, therefore, that despite the quite adamant opposition to the Sayyida Ruqayya by government officials, journalists and citizens alike, its expansion has progressed without an interruption. This is of course not surprising in view of the unchallenged authority of President Hafez Assad and the momentum of Syrian–Iranian relationships during the past thirty years. But the success of these shrines also rests on the approval, or at least acquiescence, of the Ministry of Awqåf of the rebuilding and expansion of the Sayyida Ruqayya. Invariably favouring religious function over historic form, the Awqåf has historically accepted the rebuilding of numerous mosques and shrine with little concern for questions of architectural or urban preservation. The history of opposition and conflict between the institutes of archaeology and the ministries of Awqåf in nearly all Muslim countries need not concern us here. But it is important to note that the two entities espouse quite contrasting visions, not just about historic preservation but about the meaning of history: on the one hand, an archaeological positivist perspective that above all values the forms of monuments and the facts of history; and on the other, a phenomenological traditionalist perspective that values the function of monuments and the morals of history. It is not the least surprising, therefore, that the Iranian forces behind these shrines have chosen to throw in their lot with the Awqåf and totally ignore the complaints of archaeologists and preservationists. What the Antiquities partisans castigate as ‘transgressions’ the Awqåf endorses as ‘improvements’.33 Finally, these shrines, with their intrusive architecture and uncommon ritual practices, seriously test our proclivity toward religious tolerance and cultural pluralism. Although these questions have usually been treated from a post-colonialist perspective that dwells on the polarities of East and West or Islam and Europe, the issues raised by the Shiʿi shrines of Syria call for a different mode of analysis, one founded on the deeply rooted pluralism of Islam, as religion, culture and pietistic practice. As resurgent Shiʿism reclaims its past and forcefully projects it into the present, we must be prepared to understand this process with a profound sense of empathy and an appreciation for cultural difference.

Notes 1. Yassser al-Tabbaa, ‘The Architectural Patronage of NËr al-Din, 1146– 1174’ (Ph.D. diss., New York University, 1982). Only the earliest stages

invented pieties

of my dissertation were supervised by Richard Ettinghausen. After his death in 1980, Priscilla Soucek and Oleg Grabar acted as supervisors. 2. Both of these madrasas are mentioned, though not well-documented, in the works of Ernst Herzfeld, ‘Damascus: Studies in ArchitectureIII’, Ars lslamica 11–12 (1945): 1–71 and idem, ‘Damascus: Studies in Architecture – IV’, Ars Islamica 13–14 (1948): 120–53; and Jean Sauvaget, Les monuments historiques de Damas (Damascus: Institut Français dʾÉtudes Arabes, 1932). 3. Nearly nothing has been written on this shrine before its recent rebuilding. But a report written by Mattar Khashan, Inspector of the Antiquities of Damascus, dated March 1977, describes the shrine as ‘a plain building of stone and mudbrick, with an epigraphic ablaq portal . . . It consists of a masjid and a mashhad . . . The dharÈ˙ (cenotaph) of Ruqayya is made of silver plated steel.’ 4. Research for this project has so far been supported by two consecutive short travel grants from Oberlin College. Earlier versions of this piece have been presented at the Society of Architectural Historians, The Institute of Pine Arts, The Institute of Islamic Studies at McGill University, and at The Arts and Material Culture of Iranian Shiʿism, held at the University of Oxford. 5. Clearly, these shrines also raise important social, economic and legal issues, that cannot be covered in this paper. These issues revolve mostly around the phenomenon of religious tourism, al-siyå˙a al-dÈniyya, which is actively encouraged in Syria. 6. This shrine and others are mentioned in Ibn Jubayr, Ri˙la (Beirut: Dar Sader, 1980), 252–3. 7. This shrine was also mentioned by Ibn Jubayr, when even then (c. 1180 ah), it was known as the beginning and end points of a Shiʿi procession during ʿAshura. For its more recent history see Sabrina Mervin, ‘Sayyida Zaynab: banlieue de Damas ou nouvelle ville sainte chiite?’ in Arabes et lraniens: CEMOTI 22 (1996): 149–62; and Irene Calzoni, ‘Shiite Mausoleums in Syria with Particular Reference to Sayyida Zaynabʾs Mausoleum’, in La Shiʿa nellʾimpero Ottomano (Roma, 15 April 1991) (Rome: Accademia Nazionale dei Lincei, 1993), 193–201. 8. This well-known shrine is located in rooms at the eastern end of the Great Mosque of Damascus. It should also be noted that, according to Shiʿi piety, the existing mi˙råb in the Great Mosque commemorates Imam Zayn al-ʿÅbidÈn, the fourth Shiʿi imam. 9. Yasser Tabbaa, Constructions of Power and Piety in Medieval Aleppo (University Park, PA: Pennsylvania State University Press, 1996), discusses the original Ayyubid shrine. See also Calzoni, ‘Shiite Mausoleums’, who touches on its more recent transformation that entailed, among other things, roofing the courtyard and adjoining a three-storey structure south of it. 10. See the quite perceptive article regarding the social and cultural tensions precipitated by the creation of this shrine: Myriam Ababsa, ‘Les mausolées invisibles: Raqqa, ville de pèlerinage chiite ou pole étatique en Jazira syrienne?’, Annales de Géographie 622 (November–December 2001): 647–63. 11. On the role of dreams and visions as pretexts for the creation of Shiʿi shrines, see Ibn Shaddåd Al-Aʿlåq al-kha†Èra fÈ dhikr umaråʾ al-shåm waʾl-jazåra, TårÈkh MadÈnat Halab, ed. D. Sourdel (Damascus: Institut Français dʾÉtudes Arabe å Damas, 1956), 42–50, who mentions no fewer

459

460

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

than seven instances of this interesting phenomenon. In five of these shrines – mashhad ʿAlÈ, masjid Ghawth, masjid al-NËr, mashhad Qaranbiya, and mashhad ʿAlÈ – the vision was of ʿAlÈ. In the mashhad al-Dikka it was a flash of light, and in mashhad it was a miraculous occurrence. In some cases, the vision is coupled with a rock with a ‘sacred footprint’ or more rarely, an inscription. 12. Jean Sauvaget, ‘Deux sanctuaires chiites dʾAlep’, Syria 9 (1928): 224–37 and 320–7. 13. Mu˙ammad Asʿad Talass, Thimar al-maqasid fÈ dhikr al-masajid of Yusuf ibn ʿAbd al-Hadi of (Beirut: Maktabat Lubnan, 1975). It is interesting that Sayyida Ruqayya is not mentioned once in this relatively late account of Damascus. It is, however, mentioned in the editorʾs addendum, 229–30. 14. Karl Wulzinger and Carl Watzinger, Damaskus, die islamische Stadt (Berlin, W. de Gruyter, 1924); Henri Sauvaire, Description de Damas (Paris: Imprimerie nationale, 1895–6); Sauvaget, Les monuments historiques de Damas (Beirut: Imprimerie catholique, 1932); and Muhammad Kurd ʿAli, Kitab khi†a† al-Sham (Damascus: al-Ma†baʾah al-Óadithah, 1925–8). 15. Akram Hasan al-ʿUlabi, Khi†a† Dimashq (Damascus: Dar al-Tabbaa, 1997), 221 and 335. 16. Muhsin al-Amin, Aʿyån al-ShÈʿa (Beirut: Ma†baʾat al-Inßaf, 1960), vol. 7, 34. 17. This is my own speculation, for I have not yet found any communications between the developers and the Ministry of Awqåf. 18. I would like to thank the supervising architect of this addition, Mu˙ammad al-SËß, for providing me with copies of these plans. The name of another architect, ʿImåd al-Haydari from Iraq, occurs in some documents, although it is difficult to ascertain the nature of his involvement. 19. Interestingly, the use of a triangular vestibule to shift orientation from the street to the shrine is attested in two well-known Shiʿi buildings: the Mosque al-Aqmar of 1125 in Cairo and the Masjid-i Shah in Isfahan, dated 1611–23. 20. See, for example, Jonathan Bloom, ‘The Mosque of al-Hakim in Cairo’, Muqarnas 1 (1982): 15–36. 21. The shrine at Raqqa is inscribed with several ˙adÈths that extol the merits and sacrifices of the two Shiʿi heroes ʿAmmår ibn Yåsir and Uways al-QaranÈ. 22. See, for example, Kamran Aghaie, Shiʾi Symbols and Rituals in Modern Iran: The Dynamics of Culture, Politics, and Society (Seattle and London: University of Washington Press, 2004). 23. On the aristocratic nature of Shiʿism, see Richard, Shiʿism. 24. Out of a population of 18 million, Syria has perhaps fewer than 100,000 Imami Shiʿis. 25. Richard, Shiʿism. 26. Mervin, ‘Sayyida Zaynab’, 155. 27. On the rapprochement between the Alawites of Syria and the Imami Shiʿis of lran, see Asad Patrick Seale, The Struggle for Power in the Middle East (Berkeley: University of California Press, 1988), 352. See also Abbas William Samii, ‘Syria and Iran: An Enduring Axis’, Mideast Monitor 2(2) (April/May 2006). http://www.mideastmonitor.org 28. Mervin, ‘Sayyida Zaynab’, 158.

invented pieties

29. These documents are located in the files of the Directorate for the History of Damascus (Mudiriyyat TårÈkh Dimashq). They include: – letter 2512/s, dated August 1981, from Afif Bahnassi to the Prime Minister – letter 3121/s, dated 18 March 1982, from Afif Bahnassi to the Prime Minister – letter 723/s/5, dated 2 March 1983, from Afif Bahnassi to the Minister of State and Minister of Culture –  letter 3121/s/22, dated 18 September, 1982, from Najah ʿAttar (Minister of Culture) to the Chairman of the Committee of Ministers, asking for stoppage of works –  letter dated 17 August 1981, from Muhammad al-Khatib to the Antiquities, voicing a citizenʾs complaints against the shrine 30. Tishreen no. 1940 (13 August 1981); Al-Thawra, no. 4410 (7/7/1977); and no. 5681 (22/8/1981). 31. ʿUlabi, Dimashq, 221 and 335. 32. Mervin, ‘Sayyida Zaynab’, 157. 33. The opposition between these two institutions, which perhaps began since the foundation of the first departments of antiquities and the drafting of laws of antiquities, is rife for study, on intellectual, historical and legal grounds.

461

CHAPTER TWENTY

Glorifying the Imamate: Architecture and Ritual in the Shiʿi Shrines of Syria The past three decades have witnessed a substantial increase in the number of Shiʿi shrines in Syria coupled with an unprecedented surge in the number of visitors to these shrines (Figure 20.2). In addition to their number, these shrines stand out for their flamboyant architectural style, which sharply contrasts with the overall drabness of their surroundings and even transgresses the rules and statutes that mandate a measure of conformity in buildings created within historical contexts. If anything, the showy presence of these shrines seems to have contributed to their huge popularity among Shiʿi pilgrims, SunnÈ visitors and non-Muslim tourists. This has turned the shrine of Sayyida Ruqayya into the second most visited religious structure in the walled city of Damascus, after the venerable Umayyad Mosque, and the shrine of Sayyida Zaynab, a few kilometres south, into a veritable pilgrimage city visited by hundreds of thousands every year.1 Just in the past fifteen years or so, scholars have begun to recognise the unique opportunities and interesting problems presented by the study of Shiʿi shrines and the rituals of visitation in Syria. At a time when Iraq is all but closed to Western scholarship and Iran nearly so, the Syrian shrines were, until quite recently, completely open to non-Muslim visitors and quite accessible for study.2 Their recent expansion or complete rebuilding, if not outright invention, often with Iranian funding, presents interesting problems in architecture, sectarian relations and even politics.3 Their popularity among women pilgrims, including some SunnÈ women, also raises interesting sociological and gender issues, which have been discussed by historians and anthropologists (Figure 20.3).4 Their unprecedented expansion in Syria and Lebanon has also raised some acclaiming voices but many more voices of disapprobation. Finally, their use as arenas of Shiʿi rituals, from simple prayers and evocations to complex taʿziyas and ‘Ashura’ celebrations, have equally attracted the attention of historians of religion.5 Yasser Tabbaa (2014), ‘Glorifying the Imamate: Architecture and Ritual in the Shi’i Shrines of Syria’, in Fahmida Suleman (ed.), The People of the Prophet’s House, I. B. Tauris and British Museum Press, 54–63.

glorifying the imamate

Figure 20.1  Damascus. Shrine of Sayyida Zaynab, rebuilt 1980s, façade from north (photo: Yasser Tabbaa)

Figure 20.2  Shrine of Sayyida Zaynab, south of Damascus (photo: Yasser Tabbaa)

The problems and paradoxes raised by Shiʿi monuments in Syria require a different approach than has generally been deployed for Islamic architecture. Neither their rejection as garish intrusions upon the traditional urban fabric, nor their normalisation within a pan-Islamic discourse have contributed to a deeper understanding of their place within contemporary Islamic architecture and culture. In response to this pan-Islamist essentialist dogma, some scholars have recently called for questioning the value of interpretation in Islamic architecture unless supported by contemporary and specific textual

463

464

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 20.3  Women touching the grille of the cenotaph at Sayyida Zaynab’s shrine (photo: Yasser Tabbaa)

Figure 20.4  Damascus. Shrine of Sayyida Zaynab, worshippers at the cenotaph (photo: Yasser Tabbaa)

evidence that directly refers to the monument at hand. Disavowing meaning and intentionality in Islamic architecture – Shiʿi architecture in particular – has led some scholars to conclude that there is nothing Shiʿi about Shiʿi architecture.6 But this unwarranted conclusion seems to be an answer to the wrong question, a question that is asked from a strictly formalist perspective rather than one that takes into account the culture of Shiʿism. In other words, what should be queried is not so much

glorifying the imamate

whether there are essential Shiʿi descriptors in Shiʿi architecture but if Shiʿi architecture, especially of shrines, responds or corresponds to the fundamental tenets of Shiʿism that are rooted in past practice but also alive in the present. It follows then that by viewing the uncommon forms, images and practices presented by these shrines as manifestations of their own system of representation we should be able, I believe, to gain a deeper understanding of their place within the cultural history of Shiʿism, generally and in Syria. To that end, this essay attempts a phenomenological approach in which the Shiʿi shrines, particularly those in Syria, are examined against normative and widely accepted Shiʿi core beliefs and ritual practices. Of the core beliefs of Twelver Shiʿism, none is more central and foundational than the Imamate of the Ahl al-bayt, that is, the belief in the secular and religious leadership of the successors of ʿAli and his male descendants through Fatima, the daughter of the Prophet Muhammad. These Imams, who are infallible (maʿßËm) by definition, have been invested by God with wilåya (spiritual authority) and are due the walåʾ (allegiance) and walåya (devotion) of the believers.7 Walåya, in turn, is activated in real time by ziyåra, the ritual of visitation of the shrines of the Imams and their male and females descendants. The ziyåra both honours the Imams and achieves the purpose of shafåʿa and tawassul (intercession of the Ahl al-bayt). This shafåʿa is achieved, both physically and metaphorically, through the nËr or blissful radiance of the Imams, to which physical proximity is especially meritorious (Figure 20.4).8 These central tenets of Shiʿism converge upon Figure 20.5  Chart of the the shrine, the nexus of Shiʿi beliefs and the focus main tenets of Ithna of most Shiʿi ritual practices, whether daily, ʿAshari Shiʿism weekly, or on high holidays. By viewing the architectural features of Shiʿi shrines in the light of these central tenets, I hope to demonstrate the various levels of interaction and integration between the two realms – religious and architectural – and even the mutually enriching process that links them together. Avoiding the potential cyclicality of this methodology, which could result from an overly stringent selection of evidence, I have aimed for a less selective and more comprehensive approach that aims towards a deeper appreciation of the meaning and intentionality of these shrines and a keener understanding of their differences and distinctions.9 Commemoration of the Imamate of the Ahl al-bayt Perhaps the most basic concept that underlies Shiʿi shrines is that they are not mosques but shrines, whose primary function is to commemorate an Imam or one of his descendants. Variably referred to

465

466

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

in Arabic and Persian as maqåm, mashhad, marqad, mazår or more rarely ʿatabåt-i muqaddasa, these shrines should be distinguished from mosques proper, whether a masjid (daily mosque) or masjid jåmiʿ (Friday mosque).10 Although most Shiʿi shrines are used for prayer, and communal prayer is often held in them, their primary function is commemoration not prayer. This rather obvious difference is actually quite significant for the architecture of these shrines, for communal prayer requires direction, whereas visitation calls for circumambulation. Thus, whereas mosques must be directed towards Mecca and are generally arrayed to allow for parallel rows of congregants, a shrine is primarily centred on the cenotaph, dharÈ˙ (Persian, zarÈ˙), or spot of commemoration. This is not to say that the worshippers prostrate in the direction of the cenotaph – prayers must always be performed in the direction of Mecca – but rather that the predominantly commemorative function of these shrines has been accommodated within a centrifugal architectural design, focused on the dharÈ˙. In smaller and simpler shrines, which typically consist of a domed square, this directional tension is partly resolved by placing the cenotaph in the centre and including a mi˙råb (prayer niche) in the qibla wall, in such a manner that the duʿåʾ (supplications) pronounced for the ‘saint’ would not be confused with the prayer (ßalåt) which is only due to God.11 But in the larger and grander Shiʿi shrines that must accommodate thousands of visitors, this simple solution does not easily present itself. As a result, the plan for nearly all larger Shiʿi shrines abandons the directionality of mosques for a plan with multiple enclosed squares or rectangles that focus on the dharÈ˙ itself (Figure 20.6). We find this plan in all the Iraqi shrines – Najaf, Karbala, Samarra, and Kazimayn in Baghdad – and, in a somewhat more additive form, in the great Iranian shrines of Qum and Mashhad.12 The plan is also used in Syria for the shrine of Sayyida Zaynab, daughter of Imam ʿAli and sister of Imam Husayn, where the dharÈ˙ rests within a domed square sanctuary, which is surrounded by a spacious courtyard, itself enclosed by a ring of chambers pierced by axial entrances. At the more modest shrine of Sayyida Ruqayya, daughter of Imam Husayn, the plan is adapted to the restrictions of the triangular plot. But, even here, the architect has made an attempt to establish the precedence of the cenotaph by having it domed, surrounded by an ambulatory, and preceded by a spacious courtyard.13 The centrality of the domed cenotaph in all Shiʿi shrines quite likely reflects the centrality of the Imam in Shiʿi theology and ritual practice. Expanding centripetally in a series of squares from the inner sanctum, the plan reflects and confirms the centrality of the Imam to Shiʿi belief and his blissful presence in the lives of the believers. In fact, the radiating sense of the design of the shrine seems to allude to the luminous essence of the Imams, the blissful light emitted by them till the end of time. We shall return to this point below.

glorifying the imamate

Figure 20.6  Ground plan of shrine of Imam ʿAli, Najaf

Wila¯ya/wala¯ya (authority/veneration) and ziya¯rat (visitation) Veneration and visitation are spatially congruent as they take place at the shrine, often as near to the cenotaph as possible. The two concepts are also mutually consequent, for the ultimate objective of the ziyåra is the veneration or devotion (walåya) of the spiritual authority of the Imam, while the wilåya (authority) of the Imams mediates between humans and the transcendent God of Islam. A hadith attributed to the sixth Imam, Jaʿfar al-Sadiq, affirms, ‘The sigh of the sorrowful for the wrong done to us is an act of praise of God, his sorrow for us is an act of worship, and his keeping of our secret is a struggle in the way of God.’14 By visiting the shrines of the Imams, the faithful in effect ‘share in the pains of the Imams, weep with them, and together with them wage a symbolic battle against the forces of evil’.15 As a faith that rests on the remembrance and commemoration of the death and martyrdom of Shiʿi Imams and their various male and female descendants, Shiʿism has long sanctioned and recommended the visitation of the tombs of these holy figures.

467

468

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 20.7  Men at the cenotaph of the shrine of Sayyida Zaynab (photo: Yasser Tabbaa)

Although this essay does not focus on Shiʿi ritual – which has been discussed in numerous publications – it unreservedly acknowledges the centrality and pivotal significance of ritual practices for understanding Shiʿi shrines. Indeed, a visit to any shrine on any day of the week will vividly illustrate the role of Shiʿi ritual in animating the spaces and surfaces of the shrine and in linking the worshipper to the living memory of the Imam. Alone or led by a guide, the visitors engage in physical and verbal rituals – including touching, rubbing, grabbing or even circumambulating the cenotaph; or reciting prescribed prayers and supplications (duʿåʾ) – that serve to enhance the wilåya of the Imam and to amplify the walåya of the worshipper to his presence (Figure 20.7). All parts of the ritual are in fact fully described in little booklets, written in Arabic or Persian, at the entrance to the shrines, which also give a hagiography of the saint, the significance of the shrine and the merits derived from the ziyåra.16 Shafa¯ʿa and tawassul (intercession) In addition to venerating the Imam and mourning his martyrdom, visitors to Shiʿi shrines also hope to derive spiritual or worldly benefit from the shafåʿa of the Imam or his descendant. Although intercession is often discussed in theological or sociological terms and quite correctly linked with the palliative and integrative function of ritual, this concept can also be examined through the shrines themselves, whose architecture, ornament and inscriptions can in fact highlight and deepen the experience of the visitors. By coming into direct contact with the images and words in the shrines, the visitor is brought closer to the foundations of Shiʿism and to the memory of the venerated personage. One of the main ways that Islamic architecture makes itself legible is of course through inscriptions, which are especially important in

glorifying the imamate

Figure 20.8  Inscriptions on the shrine of Sayyida Ruqayya, Damascus (photo: Yasser Tabbaa)

Shiʿi architecture. In fact, there is an embarrassment of epigraphic riches in all Shiʿi shrines, especially those of relatively recent construction or rebuilding, exceeding any comparable SunnÈ monument in the number, size and variety of their inscriptions. For example, in the relatively small shrine of Sayyida Ruqayya, inscriptional friezes crown all exterior walls and completely surround the courtyard; encircle the dome’s exterior drum and its internal springing zone; highlight the portals and the cenotaph; and ornament the inner walls with numerous inscribed cartouches and medallions (Figure 20.8). Islamic architecture, particularly of earlier periods, has thrived on the study of inscriptions, seeking to find in them an epigraphic programme comparable in significance and intent to the iconographic programmes of Christian architecture.17 This approach has yet to be applied to the study of Shiʿi shrines, including those of Syria, and this brief account is only a preliminary attempt in that direction. In terms of content, the inscriptions at the shrines of Sayyida Ruqayya and Sayyida Zaynab are of four types: (1) Qurʾånic verses, (2) foundation and waqf (endowment) inscriptions, (3) hadith (4) and various invocations (duʿåʾ), of which a few are in Persian. The Qurʾånic verses include some common verses, including the Throne Verse (2:255, Åyat al-kursÈ) and the Light Verse (24:35, Åyat al-nËr), found respectively on the outer drum and inner springing of the dome at Sayyida Ruqayya; and other somewhat less common short verses, including sura al-Kawthar (108, a River in Paradise), above the shrine’s main entrance. Despite their normative presence in many Islamic monuments, Shiʿi and SunnÈ alike, these verses are interpreted quite differently by Shiʿi theologians, most commonly by interceding the Imam between God and the worshipper. Twelver Shiʿism, which was strongly influenced by Ishraqi Sufism, generally interpreted the Light Verse as God’s bestowing light upon Muhammad, Fatima and the Twelve Imams, who would guide humanity thereafter. Thus, the phrase ‘nËrun ʿalå nËr’ or ‘light upon light’ was understood to mean ‘Imam succeeding Imam’, an interpretation that affirms the continuing

469

470

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 20.9  Qurʾånic verse 28:5 from sËra al-Qaßaß inscribed in a frieze on the shrine of Sayyida Zaynab (photo: Yasser Tabbaa)

role of the Imams in spreading God’s wisdom and grace for future generations.18 Furthermore, these inscriptions also include Qurʾånic verses that have historically been used predominantly in Shiʿi buildings. These verses, including 33:33 (from sura al-A˙zåb) and 32:24 (from sura alSajda), refer to the Ahl al-bayt, emphasising their purity (†ahåra) and their Imamate, or God-given right to lead the Muslim community. One of the most frequent and evocative of these inscriptions is 28:5 (from sura al-Qaßaß), which speaks of the reward of legitimate rule that God will bequeath upon those who have been oppressed in the world, an allusion that the Shiʿa have long believed refers to their own community (Figure 20.9).19 Whereas the Qurʾånic verses have both a pan-Islamic sense and a specifically Shiʿi intent, the hadith selection in these shrines is exclusively Shiʿi in content and recension. At the Sayyida Ruqayya, these hadiths are generally presented within roundels or cartouches placed around the inner walls of the shrine and the arches and spandrels below the dome, surrounding the shrine like icons in an Orthodox church. Placed at eye-level and written in a reasonably legible calligraphic style, these inscriptions were meant to be read, or at least to reinforce the similar verbal exhortations repeatedly made by the various kha†Èbs (who recite sermons or khu†bas) in the shrine. Relatively rare in SunnÈ mosques and religious institutions, the abundance of hadith inscriptions in Shiʿi shrines not only identified the Shiʿi nature of the shrine but also contributed pietistic, theological and even polemical dimensions to its significance. In terms of content, these hadith inscriptions emphatically reiterate the basic tenets of Twelver Shiʿism, including the wilåya (spiritual authority) of the Twelve Imams (Figure 20.10); their rightful succession (Figure 20.11); their ʿißma (infallibility) and †ahåra (purity); and their shafåʿa (intercession) for the worshipper with God.

glorifying the imamate

471

Two examples of these hadiths, out of several dozens from the shrine of Sayyida Ruqayya, will hopefully illustrate their overall significance. As with all hadiths, both of these begin with ‘the Prophet said’ or ‘the Messenger of God said’, and some of the hadith inscribed in the shrine end by citing the reference of the hadith, such as Sulayman al-Qunduzi’s (d. 1877) YanåbÈʿal-mawadda or Muhammad Baqir al-Majlisi’s (d. 1699) Bi˙år al-anwår. In the first example (Figure 20.10) the hadith says: ‘The Prophet said: He who has accepted ʿAli’s authority (walåʾ) has accepted my authority and the authority of God.’ The second hadith (Figure 20.11) says: ‘The Messenger of God said: My trustees after me are twelve, the first of whom is ʿAli and the Figure 20.10  Detail last is al-Qaʾim, and he [the Prophet] said al-Mahdi, of the of a hadith referring descendants of Fatima.’ Here, the latter inscription also to Imam ʿAli, inscribed on includes the source of the hadith as YanåbÈʿ al-mawadda. courtyard tiles at In addition to their specific intent and purposeful loca- the shrine of tion within the monument, these inscriptions, by their Sayyida Ruqayya sheer number and length, may also suggest a more gen- (photo: Yasser eralised intention. In other words, this overabundance of Tabbaa) inscriptions – amounting in the shrine of Sayyida Zaynab and all the main Iraqi and Iranian shrines to hundreds of metres in length – may not have been used for its specific significance but rather to beautify and sanctify every portion of the shrine while also immersing the worshipper in the divine presence. Being surrounded by Qurʾån and hadith inscriptions also emphasises the active presence of God in the world through the mediation of His words and the example of the Imams.

Figure 20.11  Detail of a hadith referring to the succession of the Twelve Imams, inscribed on tiles near the cenotaph inside the shrine of Sayyida Ruqayya (photo: Yasser Tabbaa)

472

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Nu¯r (radiance of the Imams) All Shiʿi Imams are said to be possessed with a divine light, a blissful illumination that has been imbued within them since eternity by God and that continues to radiate from their shrines. In a wellknown hadith, Imam Husayn describes the Imams as ‘silhouettes of light revolving around the throne of the All Merciful’, attesting to the firmly held belief in the blissful light (nËr) that emanates from the Fourteen Infallibles: Prophet Muhammad, Fatima and the Twelve Imams.20 Although there are some tantalising examples of a visual iconography of light in some medieval Islamic monuments,21 it is only in the Safavid and Qajar periods that a consistent cosmology and iconography of light is developed, creating comprehensive and saturating luminescent interiors, largely dominated by domes.22 Since complete Safavid domes with radiant patterns are quite rare and non-existent outside Iran, the following will focus exclusively on the most typical and striking ornamental techniques in all Shiʿi shrines, namely mirror-glass mosaic or åʾÈna-kårÈ. Briefly, this showy ornamental technique was first used as early as the late seventeenth century in some Safavid palaces, where mirror-glass fragments were inserted among the geometric and muqarnas (honeycomb) wall decoration.23 But it is only in the nineteenth century that this technique was applied quite densely and comprehensively to the interior of the

Figure 20.12  Najaf. Shrine of Imam ʿAli, main iwån and dome, from north-east (photo: Yasser Tabbaa)

glorifying the imamate

473

domes of Shiʿi shrines, becoming so popular that even earlier tiled domes were apparently torn down and resurfaced with mirror-glass mosaic. The Syrian shrines of Sayyid Ruqayya and Sayyida Zaynab – and even some of the smaller ones in Darayya, Raqqa and Baalbek – are no different. Polychrome tiles, gilt and enamelled doors, and faceted mirror-work cover every surface of these buildings (Figure 20.13). A peak of excess is reached in the massive cenotaph, invariably made as a silver grille, crowned by a lavishly decorated baldachin, and sheltered Figure 20.13  Mirror-glass mosaic by a stupendous muqarnas dome inlaid with work (åʾÈna-kårÈ) in the dome of mirror-glass mosaic. More than reflecting or the shrine of Sayyida Ruqayya even fracturing images or forms, the mirror (photo: Yasser Tabbaa) mosaic seems to radiate with an inner glow, even in the dimly lit interiors, quite likely a metaphor of the eternal radiance of the Imam (Figure 20.14). The combined effect of the constant prayers, forceful inscriptions and brilliant ornament creates an immersive atmosphere of luxury

Figure 20.14  Dome with mirror-glass mosaic at the shrine of Sukaina bint al-Husayn, Darayya, Syria (photo: Yasser Tabbaa)

474

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

and directness, of sacred texts made accessible through the mediation of ornament. It is a specifically Shiʿi sense of decorum, in which the Word of God and the traditions of the Imams are infused with colourful and sparkling patterns, much like the cenotaphs of the Imams during the ʿAshuraʾ processions are covered in brilliant silks and satins.24 The ultimate objective of these multiple visual and aural stimuli is not quite transcendence, as it would be for Sufism, but rather total immersion in the presence of the Imams; proximity to their radiant essence; and the glorification of their memory. Conclusion This essay has attempted to discuss Shiʿi shrines – mainly in Syria but in other regions as well – from a phenomenological perspective, exploring the various ways that the central tenets of the faith were reflected, affirmed and even amplified by the most salient and significant feature of these shrines. Viewing the architecture of Shiʿi shrines through the lens of Shiʿi core beliefs and ritual practices, the essay has highlighted the dynamic interplay between these religious concepts and architectural and ornamental forms. Rather than insisting on essential Shiʿi features in the architecture of Shiʿism, or dismissing them thereof, this paper has sought to envisage Shiʿi shrines – in their plan, design, ornament and inscriptions – within a discursive relationship with Shiʿi concepts and principles. As such, the architecture of Shiʿism emerges, in Syria and elsewhere, not as an aggregate of imported Iranian features, but one whose significant features reflect and amplify the main tenets of Shiʿism. If there is a single overriding theme that would bring together the architectural, ornamental, epigraphic and ritualistic aspects of Shiʿi shrines, it is perhaps immersion and saturation. From the moment worshippers enter the shrine, they are engulfed by images, words, sounds, and even scents and textures, that transform their sense of place and time and deepen their sense of belonging. Their journey is neither personal nor transcendent but communal and immersive, leading to the presence of the Imam.

Notes 1. This is not based on a scientific survey but on direct observation over several years of the huge number of pilgrims visiting this shrine. It is also based on the presence of the Sayyida Ruqayya shrine on all online tourist itineraries to Syria, whether for religious tourism or otherwise. 2. Unfortunately, the current state of insurgency and civic unrest in Syria makes it all but impossible to conduct research in it for some time to come. 3. See Yasser Tabbaa, ‘Invented Pieties: The Revival and Rebuilding of Shiʿite Shrines in Contemporary Syria’, in Linda Komaroff, ed., Artibus

glorifying the imamate

Asiae (special issue, Festschrift for Priscilla Soucek), 66 (2006), pp. 142–71, for an assessment of the overall phenomenon and a more indepth discussion of the shrine of Sayyid Ruqayya. This article also proposes that the earlier political rhetoric of the Iranian Revolution, which contributed to the creation and expansion of this shrine, has now given way to a more generalised Shiʿi piety that has found wide pan-sectarian appeal. 4. These include Sabrina Mervin, ‘Sayyida Zaynab: banlieue de Damas ou nouvelle ville sainte chiite?’, Cahiers d’Études sur la Méditerranée Orientale et le monde Turco-Iranien, 22 (1996), pp. 149–62; and Irène Calzoni, ‘Shiite Mausoleums in Syria with Particular Reference to Sayyida Zaynab Mausoleum’, in Biancamaria Scarcia Amoretti (ed.), La Shiʿa nell’impero Ottomano. Roma 15 April 1991 (Rome, 1993), pp. 193–201. 5. Khalid Sindawi, ‘The Zaynabiyya Hawza in Damascus and its Role in Shiʿi Religious Instruction’, Middle Eastern Studies, 45, 6 (2009), pp. 859–79; and Khalid Sindawi, ‘The Shiite Turn in Syria’, Current Trends in Islamist Ideology, 8 (2009), pp. 82–107. 6. See, for example, Jonathan Bloom, Arts of the City Victorious: Islamic Art and Architecture in Fatimid North Africa and Egypt (New Haven, CT and London, 2007), whose interpretive reticence counters the author’s earlier explorations of meaning in Fatimid architecture, including his article, ‘The Mosque of al-Hakim in Cairo’, Muqarnas, 1 (1983), pp. 15–36. 7. For an explanation of the terms wilåya and walåya in SunnÈ and Shiʿi contexts see, Mawil Y. Izzi Dien and Paul E. Walker, ‘Wilåya’, EI2, vol. 11, pp. 208–9; and Maria Massi Dakake, The Charismatic Community: Shiʿite Identity in Early Islam (Albany, NY, 2007). On the institution of walåʾ see Patricia Crone and Arent Jan Wensinck, ‘Mawlå’, EI2, vol. 6, pp. 874–82. 8. Entire books have been written on any one of these Shiʿi tenets and ritual practices. Nearly all are discussed in Mohammad Ali AmirMoezzi, The Spirituality of Shiʿi Islam: Beliefs and Practices (London, 2011). For light, see Mulla Sadra Shirazi, On the Hermeneutics of the Light Verse in the Qurʾån – TafsÈr åyat al-nËr, ed. Latimah-Parvin Peerwani (London, 2004). 9. The author intends to apply the basic methodology used in this paper to a book-length study of Shiʿi shrines in the Arab world and Iran. 10. Robert Hillenbrand suggests that the multiplicity of terms for the mausoleum might be due to the fact that ‘the role of the mausoleum in Islamic society was not defined with any precision’. Robert Hillenbrand, Islamic Architecture: Form, Function and Meaning (Edinburgh and New York, 1994), pp. 255–7. 11. This commonplace design can be found in numerous small shrines, including many published in Hillenbrand, Islamic Architecture, pp. 256–97. In fact, it seems that the elaboration of the mausoleum into a veritable shrine, with preceding courtyards and ancillary chambers, does not begin before the Mongol or even Timurid periods. 12. For plans of these shrines, see most recently James W. Allan, The Art and Architecture of Twelver Shiʿism: Iraq, Iran and the Indian Subcontinent (London, 2012). Allan has also argued that the courtyard surrounding the dome, which he calls the bahw, is known from perhaps as early as the twelfth century.

475

476

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

13. Prayer halls (mußallås) do exist in all these shrines, but the space dedicated for communal prayer tends to be either the courtyard around the shrine or a sequestered space, as in the Sayyida Ruqayya. This could in fact be related to the contested views on the Friday congregational prayer in Shiʿism, with Usuli and Akhbari legal schools taking different positions on it. For example, Juan Cole proposes that ‘many Shiʿi ulama held Friday congregational prayer invalid in the absence of the Twelfth Imam’. See Juan R. Cole, Roots of North Indian Shiʿism in Iran and Iraq: Religion and State in Awadh, 1722–1859 (Berkeley, CA, 1988), p. 20. 14. Mahmoud M. Ayoub, Redemptive Suffering in Islam: A Study of Devotional Aspects of ʿÅshËrå in Twlever ShÈʿism (Berlin, 1978), p. 142. 15. Yann Richard, Shiʿite Islam: Polity, Ideology, and Creed, tr. Antonia Nevill (Oxford, 1995), p. 11. 16. Much of the information in these booklets appears to be derived from various ritual compendiums, of which the most popular is Shaykh ʿAbbas al-Qummi (d. 1941), MafåtÈ˙ al-jinån: wa yalÈhi kitåb al-båqÈyåt al-ßåli˙åt (Beirut, 2011), long known in Persian, but recently translated into Arabic. 17. The paradigm for this methodology is quite likely Oleg Grabar’s ‘The Umayyad Dome of the Rock in Jerusalem’, Ars Orientalis, 3 (1959), pp. 65–85. 18. See below, note 19. 19. See, for example, Jonathan Bloom, ‘The Mosque of al-Hakim in Cairo’, pp. 15–36. 20. Much has been written on the integration of Ishraqi Neoplatonist philosophy with Imami Shiʿism, particularly in the Safavid period. See, for example, William C. Chittick (ed.), The Essential Seyyed Hossein Nasr (Bloomington, IN, 2007), pp. 111–19; and Amir-Moezzi, Spirituality of Shiʿi Islam, pp. 133–68. 21. These include the façade of al-Aqmar Mosque in Cairo (built in 1125); the portal with a frieze of lamps at the Mashhad al-Husayn in Aleppo (late twelfth century); and even the star-and-cross lustre tiles in various Iranian shrines, which at least one prominent scholar has linked with Shiʿi shrines. See Oliver Watson, ‘The Masjid-i ʿAlÈ, QuhrËd: An Architectural and Epigraphic Survey’, British Institute of Persian Studies, 13 (1975), pp. 59–74. 22. The iconography of light in Shiʿi architecture, the subject of a current study by the present author, has not been adequately studied nor distinguished from more general studies of the iconography of light in Islamic architecture. See, for example, Doris Behrens-Abouseif, Beauty in Arabic Culture (Princeton, NJ, 1999), pp. 197–200, where the author largely dismisses any meaning or symbolism in Islamic art, including light. 23. Little work has been done on Iranian mirror-glass mosaic. Eleanor G. Sims, ‘ÅʾÈna-kårÈ’, EI2, vol. 7, pp. 692–4. 24. See, for example, Kamran S. Aghaie, The Martyrs of Karbala: Shiʿi Symbols and Rituals in Modern Iran (Seattle, WA and London, 2004).

CHAPTER TWENTY-ONE

The Functional Aspects of Medieval Islamic Hospitals The importance of the hospital has been largely neglected by most scholars dealing with Islamic urbanism; its significance is, however, indirectly reflected by the fact that the budget for the Mansuri hospital was the largest of any public institution in late medieval Cairo.1 Introduction Following a sporadic history that began in Baghdad toward the end of the eighth century, the hospital became one of the characteristic institutions of most central Islamic cities. But despite its popularity and relative importance after the twelfth century, the Islamic hospital has attracted very little scholarly attention, lagging in this respect behind all other medieval Islamic institutions, including the madrasa (college), the shrine and the khånqåh (Sufi convent). Other than the outdated Tarikh al-bimaristanat fi l-Islam (History of Hospitals in Islam) and the more recent but problematic books by Gönül Cantay and Françoise Cloarec, no serious book has been written on medieval Islamic hospitals.2 Both books suffer from narrow geographical focus, perfunctory analysis of the architectural remains, and poor utilisation of the primary sources, whether literary or archival.3 The following contribution, part of a larger project on the history of the Islamic hospital from the eighth to the fifteenth century, attempts a comprehensive and interdisciplinary approach to hospitals, integrating their institutional and socio-medical histories with their existing architectural remains. In other words, the charitable or even social function of Islamic hospitals is but one aspect of an institution that fulfilled equally medical and educational purposes, while also having a distinctive architectural form.4 In addition to discussing the role of the medieval Islamic hospital within the contexts of charity and human entitlements, this piece discusses its early Yasser Tabbaa (2003), ‘The Functional Aspects of Medieval Islamic Hospitals’, in Michael Bonner and Amy Singer (eds), Poverty and Charity in Middle Eastern Contexts, Albany: SUNY Press, 95–119.

478

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

history under the patronage of the ʿAbbasid caliphs of Baghdad, its development in the medieval period into a multi-functional institution, and the factors contributing to its gradual decline in the later middle ages. On the basis of findings in literary, architectural and a scattering of archival sources, I would suggest that the combined effect of the rise of post-Seljuq city-states in the twelfth century, the creation of alternate sources of patronage in the form of waqfs (endowments) and the professionalisation of medicine all contributed to the creation of the medieval Islamic hospital as a medical and charitable institution. Thus, the prosperity of these hospitals rested on their continued funding through waqf endowments and on a permissive intellectual atmosphere that tolerated the practice of Galenic or humoural medicine despite its ultimate opposition by most orthodox theologians. Conversely, the hospitals themselves, with their scientific and rationalistic outlook, helped to mitigate the effects of dogmatic theologies and anti-scientific trends, preserving ancient medical practices well into the fifteenth century and beyond. Etymology The various functions that have historically been fulfilled by hospitals are reflected in the wide-ranging terms applied to this institution throughout its long history. The first hospitals, those developed under Constantine in the early fourth century, were called nosokomeion (house of the ill). Those built in Anatolia and the JazÈra during the fifth and sixth centuries were generally called xenodocheion (house for strangers or travellers). The change in terminology did not necessarily result from a change in function, although it does seem likely that the later institutions, in view of their remote locations, may have been as much hospices as hospitals.5 Furthermore, despite the greater specificity of the older term, the latter term, xenodocheion, was the one that took hold and that was subsequently assimilated into the Persian term bÈmåristån (house of ill), as we learn from a letter by Timothy, the patriarch of the Eastern Church in Baghdad from 780 to 823. In this letter dated 790, Timothy writes to his good friend and doctor Sergius, whom he had previously appointed as metropolitan of Elam (Khuzistan): We have built a ksndwykn [= xenodecheion], that is a bymrstn [bÈmåristån] in the Royal Cities [al-Madaʾin or Ctesiphon/ Seleucia], and have spent more or less 20,000 [zuke]. It has been roofed over already and completed, and pray that our Lord may give in it healing to the sick and to those who are bodily or spiritually sick.6 This text is interesting in at least two respects. First, the earliest Eastern word for hospital, bÈmåristån, is Persian rather than Syriac.

medieval islamic hospitals

This seems to suggest that the institution of the hospital may have been well established in Iran before the eighth century, though not necessarily at or within the region of Jundishapur.7 Second, the foundation of this Christian hospital very near to Baghdad closely coincides with the foundation of the earliest Islamic hospital, that built by the ʿAbbasid caliph Harun al-Rashid in 786 in Baghdad. This temporal and geographic coincidence suggests a kind of competition between Muslims and Christians, a point to which I shall return. Interestingly, the term bÈmåristån is mainly used in the Arab world, often shortened to måristån, whereas in Turkey various terms such as dar al-∑ifa or ßifahane (house of cure), or even dar al-’afiya (house of health), are used.8 But the terms bimarhane and hastahane (both variants on bÈmåristån), and timarhane (house of the mentally ill), are also noted. Most recently, the term mustashfa (a derivation of the root shafa [to cure]) has become the standard Arabic name for modern hospitals. Therefore, one is led to ask why so many terms were used for a single institution, even during the medieval period. This is all the more curious in view of the fact that the hospital in Islam was not nearly as important as the madrasa, which has just one name in all Islamic countries. In fact, the variety of terms applied to the Islamic hospital, rather than affirming its importance, seems to point in the other direction: towards an institution whose identity never completely coalesced and never became properly rooted in medieval Islamic cities or in an increasingly conservative culture.9 The hospital in early and medieval Islam The pre-modern history of the Islamic hospital can be divided into three fairly distinct periods: early (785–1000), medieval (1150–1500) and Ottoman (1500–1800). This piece focuses on the second period, from which the earliest hospitals have survived, while using the first period primarily for comparative purposes. In effect, the first period begins with the ʿAbbasids and not the Umayyads (661–750), who did not build any hospitals, with the possible exception of a leprosarium outside Damascus.10 But soon after the Muslim capital was moved from Damascus to Baghdad, the ʿAbbasid caliphs seem to have recognised the importance of medicine and hospitals and built hospitals in their capital city. This took place during the reign of Harun al-Rashid (789–809), apparently under the direct supervision of Syriac-speaking Christian Nestorian physicians, of whom the Bakhtishuʾ family was by far the most prominent. No less than eight generations of doctors from the Bakhtishuʾ family are known to have served as private doctors of ʿAbbasid caliphs between the middle of the eighth and the end of the tenth century, and of Buyid sultans well into the middle of the eleventh century.11 Although some members of this family were occasionally

479

480

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

associated with various hospitals in Baghdad, the Bakhtishuʾ family were primarily court physicians. As such, they were instrumental in the transmission of Syriac-Greek medicine and pharmacology to the ʿAbbasid court and from there to the urban populations of Baghdad and Samarra.12 Six more hospitals were founded in Baghdad in the ninth and early tenth centuries, culminating in al-ʾAdudi hospital, built in 978. This large hospital was founded by the Buyid overlord ʿAdud al-Dawlah (949–83) on the western bank of the Tigris. It had twenty-four physicians (tabib or taba’i’i), a chief physician (sa’ur) and specialists in surgery (jarrah) and ophthalmology (kahhal), making it the most illustrious hospital of the ʿAbbasid Caliphate. It was partly destroyed by flood in 1045 and substantially rebuilt by the Seljuqs in 1068, when the number of physicians was raised to twenty-eight. It continued to operate under increasingly difficult conditions until its final destruction by the Mongols in 1258.13 What led the ʿAbbasids and their Buyid overlords to found so many hospitals and to support the greatest physicians of their time? The early ʿAbbasids saw themselves as the cultural heirs to the great Sassanian Empire, the first patrons of medicine and hospitals in the Orient. Their capital city Baghdad stood at the physical and spiritual centre of the civilised world, and its Neoplatonic form was meant to embody and radiate this centrality and to outshine the glory of the Sassanian kings. Medicine, like poetry, historiography or architecture, was a direct product of ʿAbbasid court culture: it relied on court support and patronage, and its spread and ultimate fate were closely linked to those of the ʿAbbasids. By patronising the great physicians of their time and establishing hospitals in their capital city, the ʿAbbasids were claiming for themselves the heritage of earlier civilisations, while also living up to the Muslim ideals of bounty and charity. It also seems likely that by founding hospitals the ʿAbbasids were providing an Islamic alternative to Christian charitable institutions, including hospitals. According to Franz Rosenthal, ‘Christian intermediaries provided the first and decisive stimulus for Muslims to become aware of medicine’s duties to society’, adding that such a concept was readily accepted because ‘Islam was particularly receptive to all ideas concerning the well-being of society’.14 Finally, one might add that the humoural theory at the foundation of Galenic medicine was not entirely incompatible with Islam, in part because it reaffirmed the unity of body and soul, while leaving a space for God as the Creator and ultimate arbiter of the natural world. Arab medicine, just like Galenic medicine before it, was founded on a view of humans’ constitution dominated by the four humours (amzija) – blood (sanguine), phlegm (phlegmatic), yellow bile (choleric) and black bile (melancholic) – whose varying mixture within the body determined the organism’s health or illness. Physicians

medieval islamic hospitals

hoped to maintain or correct an appropriate balance of humours within their patients through any means at their disposal: medical, pharmaceutical, dietetic, musical, atmospheric, and even spiritual. Overall, physicians were more concerned, and perhaps more successful, in the maintenance of health than in the cure of disease. The period between 850 and 1100 is rightfully considered to be the peak of Islamic medicine. In many respects, this was the period of the court physician: although some of these physicians practised in the hospital, from a shop in the market-place, or from their own home, the main locus of their activities was the court of the ʿAbbasids and Buyids in Baghdad.15 Numerous anecdotes underline the special favour in which these physicians were held by the caliphs and the familial proximity they enjoyed in their company.16 It mattered little that most of these early physicians were Christians; what mattered was that they enhanced the prestige of the caliphate while fulfilling an important social service. Correspondingly, by interacting with the privileged culture of the court, Christian physicians and their Galenic medicine became fully integrated in the adab or belleslettres of Islamic culture. With very few exceptions, hospital building before the twelfth century remained primarily an ʿAbbasid phenomenon restricted to their capital in Baghdad. Thus, their political decline in the first half of the eleventh century adversely affected hospitals and even medicine.17 Such decline is attested to by two main factors: the end of hospital building after the great al-ʾAdudi hospital; and, more important, the departure from Baghdad of several important physicians to Egypt in the eleventh century and to Syria in the twelfth. Therefore, it is permissible to speak of a significant disjunction in the history of Islamic medicine, a disjunction most clearly seen between the peak of medical knowledge and the peak of activity in hospital construction. This disjunction is historical: whereas medicine reached its peak in the ninth and tenth centuries, large-scale hospital building only began in the late twelfth century and reached its height in the thirteenth. It is also geographical: the first centre was indisputably Baghdad; subsequently, it was Damascus followed by Aleppo, Cairo and several Anatolian cities. There was a further disjunction regarding patronage, which was primarily caliphal in the first phase and somewhat more diffuse in the second, involving lesser princes and often their wives or sisters. Yet this interruption of the ʿAbbasid hospitals carried within it the germ of the later revival of this institution – one that began in the middle of the twelfth century in Damascus under the direct impetus of immigrant Baghdadi physicians. According to Max Meyerhof and Joseph Schacht, ‘the revival of medicine in Syria in the twelfth and thirteenth centuries was ultimately related to the demise of the Ududi hospital’.18 Although medicine in Syria certainly benefited from the demise of the ʿAbbasids, it took the personal efforts of the Syrian

481

482

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

sovereign NËr al-DÈn Mahmud ibn Zangi (1146–74) to lay down its institutional foundations in the form of hospitals. Although primarily known for his military foundations against the Crusaders and for his patronage of SunnÈ foundations, especially madrasas, NËr al-DÈn also built several hospitals: two large ones in Aleppo and Damascus, and smaller ones in Hama, Raqqa and Harran.19 Furthermore, his project was continued by Saladin – who built his hospitals in Cairo, Alexandria, Jerusalem, Acre and Ramla20 – and by other dynasts and often their wives or sisters – who built hospitals in Anatolia, upper Mesopotamia and Egypt. By the end of the thirteenth century there were approximately sixty hospitals in the Islamic world, mainly concentrated in the central Islamic world. NËr al-DÈn and Saladin’s interest in hospitals should ultimately be traced to their emulation of Baghdad and to their passionate struggle for the SunnÈ revival. However, one cannot ignore the element of competition with Christian charity, both with local Christians and with the new Crusader kingdom and its principalities. Concerning the former, one notes that hospital building in the medieval period was located in regions (Syria, Palestine, the JazÈra, Anatolia and Egypt) with still sizeable Christian communities who would have received medical care through the many remaining churches and monasteries. Evidence is scarce, but there are some indications that Christians frequented those churches or monasteries that were linked to the healing powers of a particular saint or an ancient cult.21 For example, the church of Mar Tuma in Mosul was located on an alley named after the physician Iliyya, who once practised medicine at the church itself.22 Elsewhere in the JazÈra, including Nusaybin (Nisibis) and especially Edessa (Urfa), many churches and monasteries possessed sacred relics that served as places of magical cure and also contained hospices and even hospitals.23 Many of these hospices, hospitals and places of cultic healing continued their earlier mission well into the medieval period, and they would have served as an example and source of competition for the various petty dynasties that dominated the central Islamic world in the twelfth and thirteenth centuries. A lack of competition might explain equally well the absence of hospitals in Islamic India and their rarity in North Africa, two regions that did not contain large Christian populations in medieval times. Architectural form The renewed activity in hospital construction is heralded by the hospitals of NËr al-DÈn, of which only the one in Damascus (1154) is well-preserved.24 The BÈmåristån al-Nuri in Damascus stands at an important juncture of the institutional and architectural history of the Islamic hospital, from where it informs us about its earlier Baghdadi prototype

medieval islamic hospitals

Figure 21.1  Damascus. Bimåristån al-Nuri, 1154, façade (photo: Yasser Tabbaa)

and provides a model for later Islamic hospitals. As with nearly all extant pre-Ottoman hospitals, it is a four-iwån (a vaulted hall with an arched opening) structure entered through a single monumental gate (Figure 21.1).25 The four iwåns converge on a large rectangular pool with four corner niches. The eastern iwån served as a lecture hall, as is evident from several early descriptions and the presence of niches for books (Figure 21.2).26 The corner rooms were reserved for the patients. The bÈmåristån was also equipped with a large latrine, with six stalls centred around a pool.27 It should suffice to treat the remaining fourteen pre-Ottoman bÈmåriståns as a group, highlighting their most salient architectural features. Note first that they are centrally located structures, easily accessible from any point in the city.28 Second, most of them utilise the cruciform four-iwån plan – one used by some contemporary madrasas but one that is also closely linked to palace architecture.29 Third, they are invariably entered through a single elaborate and

483

484

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 21.2  Damascus. Bimåristån al-Nuri, 1154, courtyard, from south-east (photo: Yasser Tabbaa)

c­ontrolled entrance, unlike mosques but quite like palaces and madrasas. What in a palace would have been guard rooms flanking the entrance serve here most likely as pharmacies or infirmaries with windows facing the street (figure 21.3). Fourth, they almost always have a single floor consisting of tall iwåns and lower enclosed chambers of various sizes (Figure 21.4).30 Fifth, all have at least one central pool with circulating water, and most contain a space for latrines. Sixth, almost none has a mosque, although in some cases a mosque was built at a later point.31 Seventh, several hospitals contain the mausoleum of the founder and many have Qurʾånic inscriptions (Figures 21.5 and 21.2).32 Generally large, well-built and amply provided establishments, these hospitals were described by travellers such as Ibn Jubayr in the twelfth century and Evliya Çelebi in the seventeenth in laudatory terms usually reserved for palaces.33 The waqf and creation of the functional hospital According to Rosenthal, ‘the noblest expression of the deep concern of medieval Muslim society with matters of public health was a highly developed hospital system, a network of urban institutions with large staffs, providing numerous services and frequently having teaching facilities attached to them’.34 This view, which is shared by most writers on the subject, is somewhat exaggerated, particularly in its emphasis on the systematic and comprehensive nature of these hospitals.35 Although preceding western European hospitals by several centuries, Islamic hospitals were not built in sufficiently

medieval islamic hospitals

Figure 21.3  Ground plan of Bimåristån al-Nuri (1154), Damascus (drawing: Yasser Tabbaa)

large numbers to serve more than a sample of the urban population. A few examples should suffice. Damascus and Aleppo, with thirteenthcentury populations of 60,000–70,000, had only two hospitals each. Cairo, with a much larger population, also had only two hospitals throughout the medieval period. Smaller cities and towns in Syria, Iraq, and especially Anatolia had just one hospital each. Many other towns had no hospitals at all. In contrast Florence had about thirty hospitals at the beginning

485

486

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Figure 21.4  Eastern iwån of Bimåristån al-Nuri, Damascus (note inscriptions) (photo: Yasser Tabbaa)

Figure 21.5  Façade of Bimåristån of Arghun al-Kamili (c. 1285), Aleppo (photo: Yasser Tabbaa)

Figure 21.6  First courtyard of Bimåristån of Arghun al-Kamili, Aleppo (photo: Yasser Tabbaa)

medieval islamic hospitals

Figure 21.7  Exterior view of Darushifaʾ (or çifte Medrese) al-Ghiyathiyyya (1205), Kayseri (note mausoleum of the founder Gevher Nesibe on the right) (photo: Yasser Tabbaa)

of the fourteenth century, and a century later there were ten more. Paris had sixty hospitals, and the smaller towns of Narbonne and Arles had fifteen and sixteen, respectively. According to Miri Rubin, ‘in England alone some 220 hospitals were founded in the twelfth century and some 310 in the thirteenth’.36 Such numbers can be produced for nearly all European cities and towns. This striking discrepancy can be minimised by the fact that the medieval European hospital primarily comprised almshouses, hospices and pilgrimage rest stops. These hospitals, according to Carole Rawcliffe, ‘offered little or nothing in the way of professional medical or surgical facilities’.37 In many cases, the truly ill were even turned away because their presence would have exhausted the limited resources of these hospitals and disturbed their ultimately spiritual purpose. In contrast, Islamic hospitals, especially those built between the twelfth and fourteenth centuries and under the Ottomans, were intended for both medical treatment and charitable relief. But even here, it is often difficult to distinguish between the ideal conditions prescribed by their endowment deeds (waqfiyyas) and the day-today functioning of these hospitals. Waqfiyyas prescribe, but in the absence of account books or daily records for any pre-Ottoman hospital, there is no consistent way to find out whether or for how long their prescriptions were followed. Nevertheless, as a group, these endowments unanimously prescribe their related hospitals as places primarily for the treatment and convalescence of the poor – both men and women. For example the waqfiyya of al-Nuri, excerpted

487

488

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

in Abu Shama, specifically limits overnight stay at the hospital to the poor and indigent, while restricting its use by all others to mere dispensation of drugs.38 The waqfiyya for the bÈmåristån al-Nasiri in Cairo also specifies the needy (dhawi al-haja) as its primary users.39 Possibly, the only exception is the waqfiyya of Qalawun’s famous hospital in Cairo, which, despite emphasising the poor and needy as its target beneficiaries, also includes ‘the strong and the weak, the wealthy and the poor, the commander and the commanded, the owner and the slave, and the famous and the unknown’.40 But this is probably just pious hyperbole because it is very unlikely that any persons of means would have sought long-term treatment outside of their own homes.41 At most, such persons would only have used hospitals as pharmacies or infirmaries, which is perhaps the sense in which the all-inclusive statement of the waqfiyya of Qalawun should be understood. In general, the continued functioning of a given medieval Islamic institution, including hospitals, was directly related to the size and prescriptive details of its endowment, and perhaps indirectly to the prevailing intellectual climate. For example, the extremely long waqfiyya of Qalawun’s hospital and several early descriptions of this hospital give us a reasonably good idea of the appearance and functions of its various wards (Figure 21.6). According to these descriptions, the building was divided into sections for various diseases or parts of the human body, each with a specialist in charge. There were sections for fevers, a section for ophthalmic medicine, and one for surgical cases. Bedridden patients (murada’) were segregated from convalescent patients (nuqaha’): the former were in enclosed spaces, whereas the latter benefited from the water and air circulation of the partly open iwåns. Sexes were strictly separated, so that many of these wards were duplicated for women. Mental patients had their own wards, typically composed of an open courtyard surrounded by secure individual cells (Figure 21.7). The hospital also had a large kitchen, storerooms, dispensaries, and several chambers with latrines. The chief physician had his own quarters in the hospital and gave lectures on medicine in a specific location. Finally, the hospital contained a mortuary – a constant reminder of the presence of death within the possibility of cure.42 The later continuity of these prescribed, normative functions must be checked against travellers’ accounts, references in literary sources and, where they exist, documents. Interestingly, the hospital of Qalawun seems to have continued its original function up to the seventeenth century, when it was glowingly described by the Ottoman traveller Evliya Çelebi: It has no equal in Anatolia or among the Arabs and Persians . . . There is a magnificent pool in the middle of a great court, which is paved with polished marble . . . A dome with an ornamented

medieval islamic hospitals

Figure 21.8  Plan of the complex of Qalawun (1285). From Pascale Coste, Monuments du Kaire, de 1818 à 1826 (Paris: Typide Firmin Didot Frères, 1839), pl. xvii

ceiling rests on twelve pillars over this splendid pool [actually, four columns and four piers]. At each of the four sides of this court is a great hall . . . In them are the beds of those afflicted with illness. The sick wear bedclothes and have silk sheets. Some of those who are ill relax next to the flowing ornamental fountains when they are close to recovering their health. The servants look after them with great care. Some of our insane brothers are in

489

490

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

gloomy cells while others are in open rooms . . . [others] are bound like lions with chains around their necks. During our time there were 306 ill and insane people in the Hospital of Qalawun.43 Curiously, European travellers who visited this hospital before and after Evliya Çelebi’s visit generally depicted it in the most dismal terms. Already by the end of the sixteenth century, the Venetian physician Prospero Alpin described a state of utter decline in Egyptian medicine, but without specifically referring to the hospital of Qalawun.44 Two Frenchmen, Benoît de Maillet and Jean de Thévenot, who visited Cairo in the seventeenth century, commented on the dilapidated state of the hospital and the general lack of medical treatment in it.45 Finally, a series of more scientific assessments of this hospital by French physicians and architects in the late eighteenth and early nineteenth centuries describe a building and an institution in utter decline, with only a handful of chained lunatics suffering from total neglect.46 The conflicting descriptions of Muslim versus European travellers are all the more curious because this hospital seems to have been substantially stabilised and rebuilt by the Mamluk governor ʿAbd al-Rahman Katkhuda in 1746.47 Although it is quite likely that the hospital of Qalawun, much like the rest of Egypt, fell on very hard times in the second half of the eighteenth century,48 we cannot altogether discount the difference of perspective that would have coloured European travellers’ perceptions of a hospital that differed so markedly from the ones in Europe. However, even if we assume that the hospital of Qalawun in Cairo had declined beyond repair by the end of the eighteenth century, it had nonetheless fulfilled its intended function for nearly five hundred years – a feat directly attributable to its substantial endowment. Conversely, the hospital founded by al-Muʾayyad Shaykh in Cairo in 821/1418 had to cease all its hospital functions immediately after the death of its founder in 1421 because its endowment did not specify enough funds to support its operation. By the following year, the patients had to be removed and the building was converted into a Friday mosque, having been furnished with a pulpit and provided with the appropriate personnel.49

Figure 21.9  View from one of the chambers of the insane at Bimåristån of Arghun al-Kamili (c. 1285), Aleppo (photo: Yasser Tabbaa)

medieval islamic hospitals

Factors of decline: intellectual With the exception of the poorly endowed and short-lived hospital of al-Muʾayyad in Cairo (1418), very few hospitals were built in the later Mamluk or classical Ottoman period, outside of the three Ottoman capitals.50 Even the earlier hospitals suffered greatly from the fifteenth century onward from embezzlement and general neglect. Although it is easier to attribute this decline to economic factors, I think it reflected an earlier and broader cultural reorientation that valorised the juridical and other Islamic sciences over all branches of the ancient sciences, including medicine. Doris Behrens-Abouseif, for example, notes that ‘specialized physician’s biographies were no longer written after the thirteenth century’, proposing that this was ‘an indication of a decline in the physician’s image’.51 It also seems likely that the complete futility of Galenic medicine in the face of the bubonic plague, particularly the Black Death of the 1340s, would have greatly eroded the credibility of physicians, while enhancing belief in magical cures and Prophetic medicine.52 Although ‘spiritual’ medicine seems to have gained over Galenic medicine in the later middle ages, evidence suggests that the two branches had coexisted since early times. The proliferation of hospitals and the unquestioned growth in scientific medicine does not necessarily mean that the majority of the population held rational views about medicine and sought treatment in hospitals. In fact, the literature is full of references to all kinds of shrines, sacred springs, rocks, trees, ancient columns, or marks left by passing saints to which baraka (blessing) and curative powers were attached. These sacred spots transcended specific religious affiliation, as we can see in al-Shabushti (Kitab al-adyira) and al-Harawi (Kitab al-ziyarat), and seem to have been used by people of various creeds.53 Even more important than these healing shrines was street or dukkan medicine, which seems to have been practised with little or no control in medieval Islam.54 This cultural shift from Galenic to Prophetic medicine and faith healing negatively affected the quality and quantity of doctors in the Islamic world. Physicians dealt with these changes in various ways. They gradually substituted Dawla with Din-composed epithets (e.g. Fakhr al-DÈn, ‘pride of the [Islamic] religion’, instead of Amin alDawla, ‘trustworthy one of the dynasty’) – ones that were intended to soften their scientific image and highlight their pious demeanour. Some incorporated Prophetic medicine into their practices, insisting on the value of prayer and dhikr (the ritual Sufi ‘remembering’ of God’s name) and rarely mentioning wine as an appropriate medication. But most eventually abandoned the profession for more prestigious careers in jurisprudence or administration. This seems to have opened the way for Christian and Jewish physicians, who would presumably settle for jobs with lower status and the increasingly

491

492

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

harsh treatment of their patrons.55 By the sixteenth century some European physicians were even imported. Factors of decline: economic Economically, bÈmåriståns, other than those founded by the Ottomans themselves, suffered in two main ways under Ottoman administration. First, it seems clear from biographical dictionaries and archival documents that the large staff required for their operation was greatly abbreviated and that it rarely included doctors. Thus, while there is plenty of evidence for continuity in the appointments of nazirs (supervisors) and khadims (custodians), there is little mention of physicians.56 In other words, these hospitals (and madrasas as well) were being run at a much lowered capacity – a shrinkage that could be attributable to their dwindling endowments or to acts of usurpation. For example, a royal decree dated 1130/1718 stipulates that one nazir should be appointed for the two hospitals in Aleppo (al-Nuri and al-Kamili), his sole responsibility being to provide soup once a day for the inmates and bury them when they died.57 The hospital had become a mortuary. The decline of proper medical care in bÈmåriståns also explains why most of those in Anatolia were given over in the Ottoman period to various Sufi fraternities who practised some sort of faith healing within them. For example, the Kastamonu hospital, founded in 1272, had already been converted into a tekke (dervish lodge) by the seventeenth century, with the patients’ rooms turned into a mosque and a tomb.58 Likewise, the hospital at Çankiri, outside Ankara, built in 633/1235 by Farrukh al-Lalla during the reign of Alaeddin Kayqubad, became a place of spiritual medicine and snake charming.59 Second, it seems abundantly clear that late Mamluk and Ottoman administrators employed legal and illegal means to usurp the substantial endowments of medieval institutions, including hospitals, and redirect them to other ends. Usurpation of waqf properties through istibdal (replacement, substitution) and khuluww (vacating) has been noted recently by Behrens-Abouseif, who proposes that the Ottomans had no alternative since most of Egypt had become waqf by the end of the Mamluk period.60 Even though this policy, which was also practised in Damascus and Aleppo, may have served the economic and political purposes of the Ottomans, it could only have been detrimental to the continuity of the few remaining hospitals in these cities.61 A case in point is the BÈmåristån al-Nuri in Damascus, founded in 1154, expanded in 1242, and restored in 1285 and 1410. By the end of the fifteenth century, its endowments had been so thoroughly pilfered that, according to the historian Yusuf al-Basrawi, it was ‘ruined beyond repair’.62 In the sixteenth century, we know of two seizures

medieval islamic hospitals

that effectively ended the financial viability of this institution. In the first, water rights coming from the Qanawat (quarter) and the hammam (public bath) al-Ghazzi, both waqfs of the bÈmåristån, were appropriated between 1529 and 1587 by a succession of officials and notables, who diverted the water and annexed parts of the hammam to their own residences.63 But the final blow for the bÈmåristån may have been the usurpation of the village of Qtaife and its two khans (hostels), which were still ‘a waqf for the bimaristan in Damascus’ as late as 1477.64 Within a few decades, however, the khans were declared ruins requiring istibdal, making it legally feasible to rededicate the Qtaife waqf for a new complex, consisting of a mosque, imaret (soup kitchen), ribat (inn or hospice), hammam and shops. But this complex and its redirected waqfs were intended for pilgrims on their way to Mecca, which was far more important for the Ottomans than the hospital of NËr al-DÈn .65 Such occurrences can be multiplied and they stand at the root of the decline of hospitals and other pious institutions in the Arabic-speaking world under the Ottomans. Conclusion The medical and charitable role of Islamic hospitals has been overstated by some polemicists seeking further affirmation for Islamic achievements in science. It has also been generally understated by medical historians who have tended to view these hospitals anachronistically, as a phenomenon that reaches its peak of development just when Islamic medicine is thought to have begun its decline. I have attempted in this piece to avoid these polarities by studying various aspects of Islamic hospitals – medical, charitable, economic and architectural – over a long period to allow for emphasis, comparison and change. From its origins as an appendage of the ʿAbbasid caliphate in Baghdad to its revival in Syria by NËr al-DÈn to its dispersion within the cities and towns of the central Islamic world, this piece has shown that the hospital underwent considerable development in function, financial support and architecture. Although we remain illinformed about the earliest hospitals in Baghdad, with the exception of al-ʾAdudi, the short period of their functioning suggests that they may have especially suffered from inadequate funding. Medieval hospitals, on the other hand, were relatively large, well-built structures that were founded and maintained by means of the waqf, a charitable mechanism that was little used for public institutions before the twelfth century. Specifying funds for the doctors, personnel, patients and the building itself, waqfs ensured the functional continuity and formal integrity of pious and charitable foundations long after the death of their founders. Nevertheless, as I have noted, the conditions of these waqfs were occasionally transgressed because hospitals, which had been

493

494

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

­ esignated for the treatment of the poor, were used by persons of d means or as prisons. More important, the waqfs of some hospitals were pilfered or confiscated in later periods, contributing to the decline of these hospitals and to the curtailment of most of their medical services. Although the abrogation of waqfs is attested in all periods and places, it seems to have become increasingly common in the late Mamluk and Ottoman periods. Just as detrimental to the continued prosperity of Islamic hospitals in the later Middle Ages was the increasingly negative attitude among the ruling and learned classes towards the ancient sciences, including Galenic or humoural medicine. These traditionalist pressures led to the demotion of medicine as a desirable career among the Muslim learned classes, who generally turned their attention to more properly ‘Islamic’ pursuits, such as jurisprudence and governmental service. The gap was gradually filled by Prophetic medicine and by other forms of faith healing that did not require a specialised institution for their practice. But Islamic hospitals were not just passive recipients of external pressures and changing cultural modes. Their architectural presence, relative financial independence and an increasingly professionalised class of physicians helped them negotiate an important place within this significant cultural shift and important epistemological divide. As long as the intellectual climate permitted, and provided their endowments remained intact, the bÈmåriståns continued to treat the sick and the insane, provide drugs and teach medicine. And even after external pressures finally forced them to abandon their intended medical and therapeutic functions, Islamic hospitals continued to provide food, shelter and solace for the poor and the weak. For several centuries during medieval Islam, the Islamic hospital occupied an increasingly restrictive cultural space, perpetuating and refining earlier medical practices but rarely going beyond them. Despite its auspiciously early start and its efflorescence during the twelfth and thirteenth centuries, the medieval Islamic hospital could not withstand cultural and epistemological forces that opposed, or at least challenged, its basic scientific outlook. Compromises with religious orthodoxy contributed to the continued toleration of the hospital and its medical practices long after the rejection of other rational sciences. But in the end, the hospital could only delay, not prevent, the ultimate decline of humoural medicine – a vibrant tradition that had flourished in Islam since the eighth century. Acknowledgements I would like to thank the following institutions for supporting my ongoing work on Islamic hospitals: Center for Advanced Study in the Visual Arts, National Gallery of Art, Washington, DC (1994–5);

medieval islamic hospitals

Social Science Research Council (1996–7); and Fondation Max van Berchem (1997).

Notes 1. Michael W. Dols, ‘The Leper in Medieval Islamic Society’, Speculum 58 (1983): 901. 2. Ahmad ʿIsa, Tarikh al-bimaristanat fi l-Islam (1939; reprint, Beirut: Dar al-Hayat, 1981); Gönül Cantay, Anadolu Selçuklu ve Osmani Darüßßifalari (Ankara: Atatürk Kültür Merkezi Yayini, 1992); Françoise Cloarec, Bimaristans, lieux de folie et de sagesse (Paris: Éditions l’Harmattan, 1998). To these books one must add the numerous articles on hospitals and medicine written by Arslan Terzio©lu over the past thirty years, which have now been collected in Beiträge zur Geschichte der türkisch-islamischen Medizin, Wissenschaft und Technik, 2 vols (Istanbul: Isis Press, 1996). 3. I am referring here to the important work being done on Ottoman hospitals by Miri Shefer. 4. In this respect, the medieval Islamic hospital differed from its Western medieval counterpart, which was essentially an institution of charity. On this distinction see Jole Agrimi and Chiara Crisciani, ‘Charity and Aid in Medieval Christian Civilization’, in Western Medical Thought from Antiquity to the Middle Ages, ed. Mirko D. Grmek (Cambridge: Harvard University Press, 1998), 182–7. 5. For the early history of hospitals in Byzantium, see Timothy S. Miller, The Birth of the Hospital in the Byzantine Empire (Baltimore, MD: Johns Hopkins University Press, 1985). 6. Cited in Michael W. Dols, Majnun: The Madman in Medieval Islamic Society (Oxford: Clarendon Press, 1992), 113–14. See also idem, ‘The Origins of the Islamic Hospital: Myth and Reality’, Bulletin of the History of Medicine 61 (1987): 367–90. 7. Concerning the myth and reality of Jundishapur, see Nabia Abbott, ‘Jundi Shapur: A Preliminary Historical Sketch’, Ars Orientalis 7 (1968): 71–3; Dols, ‘The Origins of the Islamic Hospital’, 369–70; Lawrence I. Conrad (‘The Arab-Islamic Medical Tradition’, in The Western Medical Tradition, 800 BC to AD 1800, ed. Lawrence I. Conrad et al. [Cambridge: Cambridge University Press, 1995], 101) argues against the existence of a hospital at Jundishapur, suggesting that it was a later myth intended to provide a precedent for the hospitals of Baghdad. 8. BÈmåristån continues to be used in waqfiyyas (endowment deeds) and other legal documents throughout history. The shortened måristån, on the other hand, enters literary usage from early on and becomes the norm. 9. Another indication of the marginal status of Islamic hospitals is that they often went unnoticed by urban topographers and, therefore, were omitted from their lists of princely or pious institutions. This is certainly the case in several medieval topographies of Damascus and Aleppo, including Ibn ʿAsakir and Ibn Shaddad. Maqrizi, on the other hand, does list the hospitals of Cairo. See Gary Leiser and Michael Dols, ‘Evliya’ Celebi’s Description of Medicine in 17th c. Egypt’, Sudhoffs Archiv 71 (1987) and 72 (1988). 10. See Lawrence Conrad, ‘Did the Caliph al-Walid Build a Hospital in Damascus?’ Aram 6, no. 1–2 (1994): 2–22 where the author definitively

495

496

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

demonstrates that the Umayyad caliph al-Walid did not build any hospitals. 11. The biographies and genealogies of the Bakhtishuʾ physicians are detailed in Ibn Abi Usaybiʾa, ‘Uyun al-anba’ fi tabaqat al-atibba’ (Beirut: Dar al-Hayat, 1974), 183ff. 12. For a summary of this period, see Emilie Savage-Smith, ‘Tibb’, EI2 10: 452–3. 13. ʿIsa, Tarikh al-bimaristanat, 187–97. 14. The evidence for the existence of Christian hospitals in Baghdad is limited to the text cited above concerning the hospital in Ctesiphon, which is a suburb of Baghdad. Franz Rosenthal, The Classical Heritage of Islam, tr. E. and J. Marmorstein (Berkeley: University of California Press, 1965), 183. 15. The well-known historian of Islamic medicine, Emilie Savage-Smith, has informed me that in her comprehensive reading of al-Razi’s Kitab al-hawi, she did not come across a single case of a patient seeking treatment in a hospital. See also Conrad, ‘The Arabic-Islamic Medical Tradition’, 131–3. 16. See, in particular, Ibn Abi Usaybiʾa, Tabaqat al-atibba’, 188–284. 17. The only hospitals known to have been built outside of Baghdad before the twelfth century are the following: Ibn Tulun (872) and Kafur (957), both in Cairo; Wasit (south of Baghdad) (1022); Mayyafariqin (modern Silvan in southeastern Turkey) (1031); and Antioch (c. 1035). See the lists in ʿIsa, Tarikh al-bimaristanat. The first two of these hospitals must be seen as direct transplants from Baghdad because Cairo, at the time, was subject to great ʿAbbasid influence. 18. Max Meyerhof and Joseph Schacht, The Medico-Philosophical Controversy between Ibn Butlan of Baghdad and Ibn Ridwan of Cairo (Cairo: n.p., 1937), 16. 19. On the architectural patronage of NËr al-DÈn, see Nikita Elisséeff, ‘Les monuments de Nur al-Din: inventaire, notes archéologiques et bibliographiques’, Bulletin d’ études orientales 13 (1951): 5–49; and Yasser Tabbaa, ‘The Architectural Patronage of Nur al-Din, 1146–1174’ (Ph.D. diss., New York University, 1983), esp. 225–32. 20. ʿIsa, Tarikh al-bimaristanat, 71–82, 230–3. 21. The early source on eastern Christian monasteries – Abuʾl-Hasan ʿAli al-Shabushti, Kitab al-Adyira, ed. Kurkis Awwad (Beirut: Dar al-Ra’id al-ʾArabi, 1986) – contains several references to Christian shrines with places of healing, for example, pp. 176, 284, 301, 311. 22. Saʾid al-Daywaji, Tarikh al-Mawsil (Mosul: Al-Majmaʾ al-ʾIlmi, 1982), 218. 23. J. B. Segal, Edessa, the Blessed City, tr. Yusuf I. Jabra (Oxford: Oxford University Press, 1970), esp. 87–100, which contain numerous instances of magical cure. 24. NËr al-DÈn’s hospital in Aleppo is a huge field of ruins south of the Great Mosque. The only remains of his hospital in Hama are two inscriptions, located today in a restaurant. 25. For the foundation and description of the bÈmåristån of NËr al-DÈn, see the still useful study by Salahuddin al-Munajjid, Bimaristan Nur al-DÈn (Damascus: Al-Majmaʾ al-ʾIlmi al-ʾArabi, 1946); and Ernst Herzfeld, ‘Damascus: Studies in Architecture – I’, Ars Islamica 9 (1947): 2–12. 26. An especially vivid description of this hospital is provided by the twelfth-century Andalusian traveller Ibn Jubayr in his Rihlat Ibn Jubayr (Beirut: Dar Sader, 1980), 255–6.

medieval islamic hospitals

27. This part of the building was converted approximately twenty years ago into a display room. 28. This is true for all pre-Ottoman hospitals and for all Ottoman ones except the hospital complex of Beyazid II at Edirne (1421), which is located about two kilometres west of the city. 29. On the uses of the four-iwån plan for palaces and madrasas, see Yasser Tabbaa, Constructions of Power and Piety in Medieval Aleppo (University Park: Pennsylvania State University Press, 1997), 84–92, 129–34. 30. This design feature may have been introduced in consideration of the patients, whose fragile health would not have allowed them to climb stairs. Equally, questions of sunlight and air circulation may have been involved. 31. The lack of mosques is quite astonishing, particularly when we compare these hospitals with their medieval European counterparts. Small mosques were subsequently added at al-Nuri in Damascus and Arghun al-Kamili in Aleppo. In hospitals containing the mausoleum of the founder (see note 32), that space may have doubled as a mosque. 32. Several pre-Ottoman hospitals included mausoleums: al-Ghiyathiyya in Kayseri; Keykaʾus I in Sivas; Turan Melik in Divri©i (626/1228); the so-called Ta∞Mescit in Çankiri (633/1235); al-Qaymari (654/1256) in Salihiyya, Damascus; and Qalawun in Cairo. Interestingly, most of these hospitals are in Anatolia, where the custom of including the founder’s mausoleum continues in the Ottoman period. 33. Leiser and Dols, ‘Evliya’ Celebi’s Description’. 34. Franz Rosenthal, ‘The Physician in Medieval Muslim Society’, Bulletin of the History of Medicine 52 (1978): 490. 35. A somewhat more measured view of the role of the hospital in the Islamic world has been recently offered by Conrad in ‘The ArabicIslamic Medical Tradition’, 135–8. 36. Miri Rubin, Charity and Community in Medieval Cambridge (Cambridge: Cambridge University Press, 1987), 1. 37. Carole Rawcliffe, Medicine for the Soul: The Life, Death, and Resurrection of an English Medieval Hospital, St. Giles’s Norwich, c. 1249–1550 (Stroud, UK: A. Sutton, 1999), xiii. 38. Abu Shama, Kitab al-rawdatayn fi akhbar al-dawlatayn al-nuriyya wa’l-salahiyya, 2 vols, ed. Muhammad H. Ahmad (1956; reprint, Cairo: Matbaʾat Dar al-Kutub al-Misriyah, 1998), 1: 21. 39. Waqfiyya 1/2, dated 29 Ramadan 613/1216. 40. Waqfiyya 2/15, dated 12 Safar 685/1285. Excerpted in ʿIsa, Tarikh albimaristanat, 134–49; see esp. lines 301–4. 41. The main exception to this rule are the students and employees of some of the larger pious complexes, who could receive medical treatment in hospitals adjoining their place of study. I am aware of three such annexed hospitals: the madrasa al-Mustansiriyya in Baghdad (1233); the funerary madrasa of Sultan Hasan in Cairo (1356); and the so-called Rabʾ-i Rashidi near Tabriz, founded by Rashid al-DÈn (d. 1318) as a mosque, tomb, khanqa (dervish lodge), hospital and hospice. 42. This description is directly based on the waqfiyya of this hospital. See ʿIsa, Tarikh al-bimaristanat, 141–6, esp. lines 331–88. 43. Leiser and Dols, ‘Evliya’ Celebi’s Description’, 24–5. 44. Prospero Alpin, La médecine des Égyptiens, 1582–1584, 2 vols, tr. R. de Fenoyl (Cairo: IFAO, 1980): 1: 12–15.

497

498

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

45. Both accounts are cited in Dols, Majnun, 131–2. 46. Also cited in Dols, Majnun, 121. The most important of these accounts are J. J. Marcel, ‘Précis historique et descriptif sur le Moristan ou le grand hôpital des fous au Kaire’, in Contes du Cheykh al-Mohdy (Paris: IFAO, 1835), 2: 151–6, which recounts the description of the French doctor-in-chief Desgenettes in 1798; and E. Jomard, ‘Description de la ville de Kaire’, in Description de l’Égypte: état moderne, ed. M. Jomard and M. Jacotin (Paris: Imprimerie de C. L. F. Pankouche, 1822), 1: 211. 47. Wizarat al-Awqaf, Cairo, no. 940. For the substantial endowments of ʿAbd al-Rahman Katkhuda, see Doris Behrens-Abouseif, ‘The ‘Abd al-Rahman Katkhuda Style in 18th Century Cairo’, Annales Islamologiques 26 (1992): 117–26. 48. For the declining economic and political situation in the second half of the eighteenth century, see Afaf Lutfi al-Sayyid Marsot, Egypt in the Reign of Muhammad Ali (Cambridge: Cambridge University Press, 1984), 3–18. 49. Taqi al-Din Ahmad al-Maqrizi, Al-Mawa’iz wal-i’tibar fi dhikr al-khitat wal-athar, 3 vols (Beirut: Dar al-ʾIrfan, 1959), 3: 324. 50. Several hospitals were built in the different quarters of Istanbul and before that in Bursa and in Edirne. To these should be added an Ottoman hospital in Manisa and two hospitals in Mecca and Medina, brought to my attention by Miri Shefer. 51. Doris Behrens-Abouseif, ‘The Image of the Physician in Arab Biographies of the of the Post-Classical Age’, Der Islam 66 (1989): 342–3. 52. To my knowledge, the effect of the Black Death on the medical profession has not been studied. Meanwhile, see Michael W. Dols, The Black Death in the Middle East (Princeton: Princeton University Press, 1979), 83–142, passim, in which he discusses the total inefficacy of the miasmic theory in dealing with the bubonic plague and the increasing reliance on magic. See also Behrens-Abouseif, ‘Image of the Physician’, 343. 53. Al-Shabushti, Kitab al-Adyira; and Ali al-Harawi, Kitab al-isharat ila ma’rifat al-ziyarat, ed. Janine Sourdel-Thomine as Guide des lieux de pélerinage (Damascus: Institut Français de Damas, 1953). 54. Conrad, ‘The Arabic-Islamic Medical Tradition’, 128–33. 55. Behrens-Abouseif, ‘Image of the Physician’, 345–6; and Anne-Marie Eddé, ‘Les médecins dans la société syrienne du VIIe/XIIIe siècle’, Annales Islamologiques 29 (1995): 91–109. 56. For example, the following documents in Dar al-Wathaʾiq al-Qawmiyya, Damascus: Awamir Sultaniyya, Aleppo, 5/182, 183, 184, 185, 186, 542 only mention appointments of officials, not doctors. 57. Awamir Sultaniyya, Aleppo, 5/186. A similar conclusion regarding the state of the two hospitals of Aleppo (al-Nuri and al-Kamili) around the end of the eighteenth century was reached by M. Kamal Shihade in ‘Al-Tibb wal-madaris al-tibbiyya’ (Ph. D. diss., University of Aleppo, 1995), 209–10. I was able to read this first-rate dissertation at the library of the Institute for the History of Arabic Science at the University of Aleppo. 58. Arslan Terzio©lu, Mittelaterliche islamische Krankenhäuser (Berlin: Technical University, 1968), 131–5. 59. Ibid., 126. 60. Doris Behrens-Abouseif, Egypt’s Adjustment to Ottoman Rule: Institutions, Waqf, and Architecture in Cairo (15th and 17th Centuries) (Leiden: E. J. Brill, 1994), 154–7.

medieval islamic hospitals

61. Abraham Marcus in The Middle East on the Eve of Modernity: Aleppo in the Eighteenth Century (New York: Columbia University Press, 1989), for example, notes that ‘no new hospitals were built, and none of the hundreds of recorded charitable endowments of the period donated property for the support of the existing ones’ (266). 62. ʾAlaʾ al-Din b. Yusuf al-Basrawi, Tarikh al-Basrawi: safahat majhula min tarikh Dimashq fi ‘asr al-mamalik, ed. Akram al-ʾUlabi (Damascus: Dar al-Maʾmun, 1988), 217, 241. 63. Jean-Paul Pascual, Damas à la fin du XVIe siècle d’après trois actes de waqf ottomanes, tome 1 (Damascus: Institut Français de Damas, 1983), 74, n. 1. 64. M. M. al-Arnaʾut, Mu’tayat ‘an Dimashq wa-Bilad al-Sham al-junubiyya fi nihayat al-qarn al-sadis ‘ashr (Damascus: Wizarat al-Thaqafa, 1993), 53–4. 65. On the Ottomans’ overriding interest in the pilgrimage to Mecca, see M. Adnan Bakhit, The Ottoman Province of Damascus in the Sixteenth Century (Beirut: Librairie du Liban, 1982), 107ff.; Karl K. Barbir, Ottoman Rule in Damascus, 1708–1758 (Princeton: Princeton University Press, 1980), 108ff. Although the Takiyya al-Sulaymaniyya, founded by Süleyman I in 962/1554–5, did not include a hospital as such, it did comprise a large soup kitchen and several residential units, all intended for pilgrims.

499

Index

Note: bold indicates illustrations ʿAbbasid Palace, Baghdad, 168–71, 171 ʿAbbasids, 6, 28, 30, 41, 47, 55, 77, 95–6, 99–100, 158–76, 179, 204, 218, 226, 236, 244, 250–2, 297, 299, 310, 315, 324–5, 329–30, 339–40, 342–5, 354–63, 372, 373, 409–12, 478–81, 493 Abbott, Nabia, 261, 262–3, 266 ʿAbd al-MasÈh, Fakhr al-DÈn, 129, 130–1, 181 ʿAbd al-Ra˙mån III, 373, 411 ʿAbdülaziz, Sultan, 167 Abu Dulaf mosque, Samarra, 54 Abu Hanifa, 160 AbË ʿl-Hasan b. ʿAlÈ, 112 Abu Muslim palace, Merv, 54 AbË Shåma, 129, 180 acanthus ornament, 40, 170 Achaemenids, 370, 409 Acre, 482 al-ʿAdawi, Shams al-DÈn, 194 ʿAdud al-Dawlah, 480 al-ʿAdudi hospital, Baghdad, 480, 481, 493 Afghanistan, 323, 373, 375–6 Aghlabid cisterns, Tunisia, 389 Ahl al-bayt, 188, 194, 195, 448–9, 452, 457, 465, 470 Ahmed, Leila, 75 Akhshu KhåtËn, 80 al-AkhtarÈni, 15, 16 ʿAlaʾ al-Din Tekish, 161 Alaeddin Keyqubad kiosk, Konya, 187 Alawites, 456

Alberti, Leandro, 380 Aleppo Antioch gate, 7, 9 Ayyubid palace, 47, 49–60, 52, 53, 65, 66, 69, 99–100, 102, 417 Bab Dar al-ʿAdl, 49 Bab al-Jabal, 49 Bab al-Maqam, 49, 51, 63 Bab al-Saghir, 49 BÈmåristån of Arghun al-Kamili, 486, 490, 492 Bimaristan al-Nuri, 55, 492 citadel, 46–60, 49, 59–66, 69, 95, 97, 98–100, 101, 102, 334, 413, 417 citadel entrance block, 47, 49, 49, 51–2, 61, 64, 97, 98, 99, 101, 102, 334 citadel mosque, 47–8, 60, 100, 333, 334 city walls, 49 dår al-ʿadl, 48–51, 95 fortifications, 47, 94–103 Gate of the Serpents, 51–2, 64 Gate of the Two Lions, 51–2, 333, 334 Great Mosque, 5, 7, 9–10, 13, 14, 16–17, 30, 30–2, 32–5, 34, 36, 112, 314, 317, 320, 327–8, 329 hospitals, 481, 482, 485 khånqåh al-Farafra, 81, 81–7, 83, 84 Madrasa al-Firdaws, 81–90, 83, 85–9, 334, 334 Madrasa al-Hallåwiyya, 4–6, 18, 20, 115, 116–17, 328, 329, 330

501

index

Madrasa al-Sultaniyya, 48–9, 49, 61 Madrasa al-Zahiriyya, 231 Madrasa al-Zajjåjiya, 5 Maqam Ibrahim, Aleppo citadel, 47, 48, 60, 110, 111, 111, 112, 115, 331, 332 Maqam Ibrahim, Southern Aleppo, 48, 51 Mashhad al-Dakka, 4, 30, 328 Mashhad al-Óusayn, 89, 441, 442, 449 Ma†bakh al-ʿAjami, 55, 56, 57, 68 mausoleum of al-Ûåhir GhåzÈ, 48–9, 61 Mosque of ʿUmar, 7 patronage, 47, 57, 79–90, 96–8, 109–10, 334 Qastal al-Shuʾaybiyya, 7–10, 8–9, 18, 20, 39, 40, 42–3, 317, 330–1 Taht al-Qalʿa region, 48, 81 trade, 97, 98, 100 ʿUmar’s capture of, 10 underground passages, 49–51, 63 Alexandria, 482 Algeria, 378–80, 413, 415; see also North Africa; and individual locations Alhambra mosque, Granada, 100 Alhambra Palace, Granada, 52, 210–13, 369, 385, 388, 388, 393, 394–7, 395–7, 419–20, 420 ʿAli, Imam, 188, 189, 193, 354, 355, 452–3, 465 ʿAlÈ b. Shådhån al-RåzÈ, 356, 357–8 ʿAli b. Yusuf b. Tashufin, 237, 248, 250, 251, 340 Allen, Terry, 27, 31 Almohads, 236, 237–8, 244, 248, 250, 253, 343 Almoravids, 120, 175, 205, 217, 227, 236–40, 249–53, 338–40, 342, 343 Alpin, Prospero, 490 al-AmÈn, Sayyid Mu˙sin, 442 Amman, 54 Anar of Damascus, 6 Anatolia architecture, 37, 38, 40–3, 42, 107, 231 calligraphy, 173, 317, 333

gardens, 382–3 hospitals, 481, 482, 485, 492 ornament, 40–3, 107, 114 see also individual locations Antioch, 5, 7, 42 Antioch gate, Aleppo, 7, 9 Andalusia architecture, 235, 244–5, 251 ornament, 247, 248 poetry, 397, 418–20 see also individual locations Apamea, 7, 10 Aq Sunqur, 179 al-Aqsa mosque, Jerusalem, 13–17, 21, 110, 111–13 Arab Ata mausoleum, Tim, 202, 226 arabesque ornament, 7, 40, 43, 108, 110–11, 113, 115, 118, 120, 135, 138, 161–2, 169–70, 247–8, 324–5, 327, 331, 336 Arberry, Arthur, 262 arcades, 100, 169, 372, 375, 378; see also porticos archaism, 26–43 arches ʿAbbasid Palace, Baghdad, 169 foliate arches, 43, 228, 241 Great Mosque of Aleppo minaret, 34 horseshoe arches, 241, 444 interlaced arch decoration, 248–50, 249 Madrasa al-Mustansiriyya, Baghdad, 167 Ma†bakh al-ʿAjami, Aleppo, 56, 57 pointed arches, 7, 9, 43, 161, 167, 169, 241 polylobed arches, 34 Qastal al-Shuʾaybiyya, Aleppo, 7, 9, 43 Qubbat al-Barudiyyin, Marrakech, 227–8, 229, 241, 243, 244 shrine of Sayyida Ruqayya, Damascus, 444 stepped extrados, 37 Talisman Gate, Baghdad, 161 trilobed arches, 241, 243, 323 triumphal arches, 10, 55–6 architectural revivals, 31, 41–2

502

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Aristotle, 211 Armenian architecture, 38–9 artisans conservatism, 37–42 inclusion in biographical dictionaries, 105, 106 signatures, 16, 34–5, 38, 105, 106, 112, 112, 115–18, 231, 310, 335 status, 105–6, 355 Artuqid palace, Diyarbakir, 55, 99, 384, 385 Artuqids, 20, 47, 382 asceticism, 7, 11, 80, 90 Ashʾaris, 211, 213, 250, 355 Ashkelon minbar, 108–9, 318 ashlar masonry, 27, 36, 40, 98 al-Ashraf KhalÈl, 99 ʿAshËra celebrations, 453, 462, 474 Assad, Hafez, 456–7, 458 Assur, 54, 55, 56 Atil, Esin, 75 atomism, 211–17, 227 authority, 12, 48, 50–1, 55, 59, 120, 160, 355, 465, 467–8 ʿAyntåb, 98 Ayyubid palace, Aleppo, 47, 49–60, 52, 53, 65, 66, 69, 99–100, 102, 417 Ayyubid palace, Karak citadel, 55, 99 Ayyubid palace, Qalʿat Najm, 55, 99, 102 Ayyubid palace, Qalʿat ÍahyËn, 53, 55, 65, 99–100, 102, 384–5, 386 Ayyubid palace, Roda Island, 55 Ayyubids, 41, 46–51, 59–60, 75–90, 96, 99, 107, 110, 113–15, 120, 175, 300, 314, 328, 333–4, 340–1, 343 al-Aʾzamiyya, 159–60 Aʾzåz, 96 al-Azhar mosque, Cairo, 314, 315, 317, 318 al-ʿAziz Muhammad, 48, 79 Bab al-ʿAmma, Samarra, 29, 30 Bab Dar al-ʿAdl, Aleppo, 49 Bab al-Futuh, Cairo, 35–7 Bab al-Jabal, Aleppo, 49 Bab al-Maqam, Aleppo, 49, 51, 63

Bab Mardum mosque, Toledo, 240, 244 Bab al-Nasr, Cairo, 33, 35–7 Bab al-Saghir, Aleppo, 49 Båb ÍaghÈr cemetery, Damascus, 108, 317, 318, 448 Bab Sharqi minaret, Damascus, 17, 22 Bab Zuwayla, Cairo, 35, 35–7 Bacharach, Jere, 95 Badr al-DÈn Luʾluʾ, 6, 135, 137, 141, 179–80, 184, 186–95, 336 Badr al-Jamali, 35, 109 Baghdad ʿAbbasid Palace, 168–71, 171 al-ʿAdudi hospital, 480, 481, 493 al-Aʾzamiyya, 159–60 Basaliya Gate, 161 calligraphy, 161, 171–3, 248, 259, 285, 324, 342–3 citadel, 159, 169, 170 city walls, 159, 161 Dar al-Khilafa, 159, 162 Dhafariya (Wastani) Gate, 161–2, 162 fortifications, 159, 161–2, 173 Halaba (Talisman) Gate, 161 hospitals, 477–82, 493 illustrated manuscripts, 172, 173–5 Iraqi Museum, 140, 141, 184, 186 Kazimayn shrine, 445, 454, 466 al-Khaffafin mosque, 163 Khan Mirjan, 176 Karkh, 159 al-Kazimiyya, 159–60 location of palace, 47 Madrasa al-Mirjaniyya, 176 Madrasa al-Mustansiriyya, 160, 165–8, 166, 168, 169, 170, 173 mausoleum of Zumurrud Khatun, 164–5, 208, 209, 214–15 Mongol invasion, 158, 159, 160, 173, 176 Mosque of the Caliphs, 160, 162, 163 Muʾazzam (North) Gate, 161 muqarnas domes and vaults, 204–5, 206, 210, 211, 218, 226, 228, 244 patronage, 158, 160, 173, 478

503

index

Qumriyya mosque, 163 resurgence of caliphate, 158–76 Round City, 47, 55, 158 Rusafa, 43, 159 underground passages, 50 Safavid destruction of SunnÈ shrines, 195 sectarian divisions, 158–9, 195 shrine of ʿAbd al-Qadir al-Gailani, 159 shrine of Abu Hanifa, 160 shrine of Muhammad al-Jawad, 160 shrine of Musa al-Kazim, 160 shrine of ʿUmar al-Suhrawardi, 159, 161, 164, 165 SunnÈ Revival, 119–20, 159 al-BaghdådÈ, Mu߆afa, 335 Bahnassi, Afif, 457 Bahråm Shåh, 324 Baihaqi, 55 Bakhtishuʾ family, 479–80 Baldwin, Marshal, 5, 7 Balkuwara palace, Samarra, 54, 67, 371, 372, 376, 409, 410, 411 banisters, 108, 108 BanË MËsa, 390 al-Baqillani, 211–12, 227, 250, 299 Bar Hebraeus, 132, 181 Barada Panel mosaics, Great Mosque of Damascus, 408, 408–9, 425, 426, 433–4 Bargebuhr, F. P., 396 BåʾrÈn, 97 Barrucand, Marianne, 252 Basaliya Gate, Baghdad, 161 Bash Tabiya, Mosul, 187 basins, 371, 372, 375, 391, 394 Basra, 298 al-Basrawi, Yusuf, 492 baths, 53, 99 Bayå∂ wa Riyå∂, 389, 390, 413 Bayn al-Qahaoui mosque, Sousse, 249 Becker, Carl H., 12 Bedouins, 95, 434 Behrens-Abouseif, Doris, 491, 492 bent entrances, 84–5 Berkey, Jonathan, 80–1 Bibliothèque Nationale (BN), Paris manuscript Ar. 382, 270 manuscript Ar. 5847, 174–5

manuscript Ar. 6041, 283, 289, 290–1 manuscript Ar. 6725, 358 manuscript Ar. 6726, 354 Bierman, Irene, 312, 313, 320–1 Bijaya, 415 BÈmåristån of Arghun al-Kamili, Aleppo, 486, 490, 492 BÈmåristån al-Nasiri, Cairo, 488 BÈmåristån al-NËrÈ, Aleppo, 55, 492 BÈmåristån al-NËrÈ, Damascus, 110, 110, 140, 206–7, 207, 229, 230, 331, 482–3, 483–6, 492–3 bÈmåriståns see hospitals biographical dictionaries, 105, 106 Birkat al-Mutawakkil, Samarra, 372, 409–11 Black Death, 491 Blair, Sheila S., 138–9 Bloom, J. M., 107, 113 ‘Blue Qurʾan’, 300, 300 Bosra, 55, 99 box machicolations, 98–9, 101 brick architecture, 27, 30, 37, 158, 163, 240 brick ornament, 107, 142, 163, 164, 166, 169, 189 bridges, 161 British Library (BL) manuscript 11,735, 270, 272, 274, 275, 275–6 manuscript Add. 7214, 283, 286, 288 manuscript Or. 12884, 272 manuscript Or. 13002, 282, 283, 289 bubonic plague, 491 al-Buhturi, 410–11, 419 building materials brick, 27, 30, 37, 158, 163, 240 fragility of, 158 re-use of salvaged materials, 27–8, 136 stone, 27, 30–7, 229–31, 240 al-Buqayʾ, 11 al-Butheiri al-Siqilli, 414–15 Buyids, 298–9, 354, 479–81 Byzantines, 5–6, 10, 28, 252, 426, 427–9

504

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Cairo al-Azhar mosque, 314, 315, 317, 318 Bab al-Futuh, 35–7 Bab al-Nasr, 33, 35–7 Bab Zuwayla, 35, 35–7 BÈmåristån al-Nasiri, 488 casket of Imåm al-Óusayn, 113, 114 casket of Imåm al-ShåfiʿÈ, 113, 114 citadel, 48, 99, 340–1, 413 Fus†å† residential architecture, 383–4, 384 gates of Cairo, 9–10, 33, 35, 35–7, 109, 340–1, 341 hospitals, 481, 482, 485, 488–90 illustrated manuscripts, 174 Islamic Museum, 341 Juyushi Mosque, 48 Madrasa of Sultan Hassan, 232, 232, 233 Madrasa of al-Íali˙ Najm al-DÈn AyyËb, 344 Mosque al-Aqmar, 232 Mosque al-Håkim, 10, 314, 314, 315, 317, 319, 320 Mosque of al-Muʿayyad Shaykh, 232, 232 Mosque of al-Íåli˙ Êalåʾiʾ, 340 al-Muʾayyadi hospital, 490–1 Mudarraj Gate, 340–1, 341 muqarnas domes and vaults, 224–33, 225, 227, 231–3 Nilometer, 315, 316 Qalawun’s hospital, 488–90, 489 shrine of al-Faraj ibn Barquq, 225, 233 shrine of Imam Shafiʾi, 231, 232, 340 shrine of Sayyida NafÈsa, 108, 109, 109 shrine of Sayyida Ruqayya, 108, 109, 226, 227 shrine of the ʿAbbasid Caliphs, 232 calligraphy see monumental calligraphy; Qurʾanic calligraphy; secular calligraphy canals, 251, 372, 383, 409 Çankiri hospital, 492

canonisation of calligraphic scripts, 266, 333–8, 360, 362 of the Qurʾan, 261, 297–300, 360–2 Cantay, Gönül, 477 caskets casket of Fakhr al-KhawåtÈn, Damascus, 115, 118 casket of Imåm al-Óusayn, Cairo, 113, 114 casket of Imåm al-ShåfiʿÈ, Cairo, 113, 114 castles, 94–103; see also citadels; fortifications; fortresses Çelebi, Evliya, 484, 488–90 cenotaphs cenotaph of Ma˙mËd of Ghazna, 322, 323 cenotaph of Sayyida Få†ima, Damascus, 108, 108, 317, 318, 320, 448 cenotaph of Sayyida Sukaina, Damascus, 318, 448 as central focus of a shrine, 466 shrine of Imam Yahya b. alQasim, 190, 190–1 shrine of Sayyida Ruqayya, Damascus, 444–5, 453 shrine of Sayyida Zaynab, Damascus, 464, 468 ceremonial domes, 251–2 ceremony, 47, 52, 55, 77, 89, 95, 165–6, 426 Ceuta, 251 Chardin, John, 388 charity, 76, 77, 80, 477–8, 480, 482, 487–8, 493 Chester Beatty Library (CBL) Add. manuscript, 268 manuscript 1417, 269, 269–70 manuscript 1430, 172, 283, 287, 288 manuscript 1431, 279–81, 280–2, 283, 286–8 manuscript 1434, 271, 272, 272, 275, 361 manuscript 1435, 283, 290, 295 manuscript 1438, 283, 294 manuscript 1439, 283, 290, 296 manuscript 3494, 270

505

index

Christian architecture, 5–6, 20, 33, 37, 41, 43, 56, 136–7, 192, 194, 469; see also churches; monasteries Christian hospitals, 479, 480, 482 Christian manuscripts, 357, 358 Christian physicians, 481, 491–2 Christianity, 5–6, 11, 19, 20, 129, 131–2, 140, 143, 180, 181, 192, 194, 479, 480, 481, 482 chronology, 26–7, 106, 201 churches Armenian churches, 38 building and restoration of, 180 columns, 136, 183–4 continuous moulding, 33, 41 conversion to Muslim use, 5–6, 131–2 destruction and pillaging of, 131–2, 181 Mår Ishaʾya, Mosul, 136 Mår JurjÈs, Mosul, 136 Mar Tuma, Mosul, 482 northern Syria, 33, 33, 41, 56 as places of cure, 482 Qalb Lozeh, 33, 33, 41 Qalb Meez, 33 St Helena cathedral, Aleppo, 5–6 St Thomas’ church, Mardin, 132 tripartite façades, 56 see also Christian architecture circumambulation, 466, 468 citadels administrative function, 46, 47 Aleppo, 46–60, 49, 59–66, 69, 95, 97, 98–100, 101, 102, 334, 413, 417 Amman, 54 ʿAyntåb, 98 Baghdad, 159, 169, 170 baths within, 53, 99 Bosra, 55, 99 Cairo, 48, 99, 340–1, 413 ceremonial function, 47, 52, 95 Damascus, 48, 99 Diyarbakir, 55, 99, 384, 417 entrance blocks, 47, 49, 49, 51–2, 61, 64, 97, 98, 99, 101, 334 as expressions of power, 46–7, 59–60, 94, 95 gardens and fountains, 412–20

Óårim, 96, 98, 99 Homs, 98 Jerusalem, 99 Karak, 55, 99 Konya, 187 links to cities, 46–7, 50–1, 59, 95 Maʾarrat al-Nuʾman, 98 military function, 46 mosques within, 47–8, 60, 100, 102 Mosul, 187 palaces within, 47, 49–60, 52, 53, 65, 66, 69, 95, 99–100, 102, 412–20 patronage, 96–8 Qalʿat al-Mu∂Èq, 97, 98, 101 Qalʿat Najm, 98–9, 99, 100, 101, 102, 103, 417 Qalʾat ÍahyËn, 98, 99–100, 417 Shawbak, 99 see also castles; fortifications city walls, 49, 94, 159, 161 Cizre, 192 classical architecture, 7, 9–10, 27, 31, 42–3 Cloarec, Françoise, 477 coins, 20, 79, 96, 250, 322, 322, 338, 340, 343 colophons, 172, 274, 275, 277, 288, 291 columns churches, 136 mausoleum of NËr al-DÈn, Damascus, 208 Mosque al-NËri, Mosul, 17, 19, 133–8, 142, 143, 146–7, 150–1, 182, 182–4, 335–6, 336 muqarnas columns, 208 commemoration, 42, 80, 113, 161, 248, 251, 330, 340, 453, 465–6, 467 concubines, 77, 429 Constantine, 478 Constantinople see Istanbul continuous moulding, 33, 36–7, 39, 40–1 Cordoba Great Mosque, 29, 30, 243, 244, 247, 248, 251 MadÈnat al-Zahråʾ, 247, 248, 373–5, 374–7, 387, 411–12, 412

506

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Court of the Lions, Alhambra Palace, 385, 394–6, 396, 419–20, 420 Court of the Myrtles, Alhambra Palace, 394, 396 court physicians, 480, 481 courtyards ʿAbbasid Palace, Baghdad, 168, 169 Alhambra Palace, Granada, 369, 385, 394–7, 396, 419–20, 420 Ayyubid palace, Aleppo, 53, 53, 55–7, 69, 100 Ayyubid palace, Qalʿat ÍahyËn, 100, 384–5, 386 Balkuwara palace, Samarra, 371, 372 BÈmåristån of Arghun al-Kamili, Aleppo, 486 BÈmåristån al-NËrÈ, Damascus, 484 Court of the Lions, Alhambra Palace, 385, 394–6, 396, 419–20, 420 Court of the Myrtles, Alhambra Palace, 394, 396 courtyard gardens, 369, 371, 375, 379–80, 383–7, 394–7, 409 Dar al-Hajar, Marrakech, 238, 252 Jawsaq al-Khaqani, Samarra, 371 Khan Mirjan, Baghdad, 176 khånqåh al-Farafra, Aleppo, 85 Madrasa al-Firdaws, Aleppo, 85, 87–8, 89 Madrasa al-Muqaddamiyya, Damascus, 439, 445 Madrasa al-Mustansiriyya, Baghdad, 167 palace courtyard dimensions, 57–8, 95 Qasr al-Mubårak, Seville, 386–7, 387 residential houses, 383–4, 384 shrine of Sayyida Ruqayya, Damascus, 444, 444–5, 452, 466 tripartite courtyard façades, 55–6, 57, 100, 167 see also gardens craftsmen see artisans Creswell, K. A. C., 26, 27, 35, 53–4, 165, 226, 340

cruciform plans see four-iwån plans Crusades, 3, 5, 6–7, 10, 28, 42, 48, 98, 130, 140, 180, 321, 329, 330 La Cuba pavilion, Sicily, 252, 380, 414 cursive script cenotaph of Ma˙mËd of Ghazna, 322, 323 citadel mosque, Aleppo, 333, 334 on coins, 322, 322, 338 Gate of the Two Lions, Aleppo, 333, 334 Great Mosque of Aleppo minaret, 327–8, 329 Great Mosque of Tlemcen, 248, 337, 338 Madrasa al-Firdaws, Aleppo, 334, 334 Madrasa al-Hallåwiyya, Aleppo, 328, 330 Madrasa of al-Íali˙ Najm al-DÈn AyyËb, Cairo, 344 Maqam Ibrahim, Aleppo citadel, 331, 332 Masjid-È JåmÈ, Isfahan, 325 monumental calligraphy, 40, 43, 138, 143, 190, 248, 248, 314, 321–45, 322, 324–6, 329–39, 341–2, 344 Mosque of NËr al-DÈn, Hama, 11 Mosque al-NËri, Mosul, 138, 335–6, 335–6 Mosque al-Qarawiyyin, Fez, 248, 338–9, 338–9 Mudarraj Gate, Cairo, 340–1, 341 NËr al-DÈn minbar, al-Aqsa, 13, 14–15 NËr al-DÈn minbar, Hama, 331, 332 promoted by NËr al-DÈn, 143, 314, 321, 327–33, 335–6, 340, 342 Qubbat al-Barudiyyin, Marrakech, 248, 248 Qurʾanic calligraphy, 172, 173, 259, 262–3, 280–301 secular calligraphy, 262, 356–7 shrine of Imam ʿAwn al-DÈn, Mosul, 191–2, 336–7, 336–8 shrine of Imam Shafiʾi, Cairo, 340

507

index

shrine of Imam Yahya b. alQasim, Mosul, 190, 336 Syrian woodwork, 109 transformation from angular script to, 259–301, 310–45 see also naskh script; thuluth script cushion voussoirs, 37 cuspated moulding, 33, 37 Daesh, 186 Damascus al-ʿAmåra quarter, 439–40 Båb ÍaghÈr cemetery, 108, 317, 318, 448 Bab Sharqi minaret, 17, 22 BÈmåristån al-NËrÈ, 110, 110, 140, 206–7, 207, 229, 230, 331, 482–3, 483–6, 492–3 casket of Fakhr al-KhawåtÈn, 115, 118 cenotaph of Sayyida Få†ima, 108, 108, 317, 318, 320, 448 cenotaph of Sayyida Sukaina, 318, 448 Damascus Museum banister, 108, 108 citadel, 48, 99 Great Mosque, 41, 138, 141, 408, 408–9, 425, 426, 433–4, 443, 462 hospitals, 481, 482–3, 485 Jåmiʿ al-Óanåbila minbar, 114–15, 114–15 Madrasa al-Måridåniyya, 80, 115 Madrasa al-Mujåhidiyya al Juwwåniyya, 439 Madrasa al-Muqaddamiyya, 5, 439, 445 Madrasa al-Íå˙ibiyya, 80 Madrasa al-Shåmiyya, 80 mausoleum of NËr al-DÈn, 208, 209, 212, 213, 229 muqarnas domes and vaults, 227 NËr al-DÈn’s capture of, 5, 6, 10 patronage, 109–10 shrine of Sayyida Ruqayya, 439–58, 444, 446, 450–5, 462, 466, 469, 469–71, 471, 473, 473

shrine of Sayyida Zaynab, 441, 443, 445, 454, 455, 462, 463–4, 466, 468, 469, 470, 471 dår al-ʿadl, Aleppo, 48–51, 95 Dar al-Hajar, Marrakech, 237–8, 252 dår al-imåra, Kufa, 54 Dar al-Khilafa, Baghdad, 159, 162 Dar al-Kutub al-Misriyya (DK), Cairo manuscript 144, 283 manuscript 227, 283 Darabsak, 96 Darayya, 447, 473 Darushifaʾ al-Ghiyathiyya, Kayseri, 487 Îayfa KhåtËn, 75–90 Dayr al-Abkår, Mardin, 131–2 Dayr al-Zaʿfaran, Mardin, 43 DaywajÈ, Saʾid, 132, 135, 186 decorum, 120, 452–3, 474 Déroche, François, 261, 270, 289 desert palaces, 10, 30, 370, 425–35 Deverdun, Gaston, 238, 248 Dhafariya (Wastani) Gate, Baghdad, 161–2, 162 al-Dhahabi, 165 diacritical marks, 262, 276, 291, 335 Dickie, James, 404 Diyarbakir Artuqid palace, 55, 99, 384, 385 citadel, 55, 99, 384, 417 Great Mosque, 138, 317 Dodd, Erika, 311–12, 313 Dome of the Rock, Jerusalem, 41, 235 domes Alhambra Palace, Granada, 210–11, 212–13 Bab Mardum mosque, Toledo, 244 BÈmåristån al-NËrÈ, Damascus, 206, 207, 207, 229, 230 on central halls of cruciform palaces, 54 ceremonial domes, 251–2 decoration, 248–50, 249 double-shell domes, 137, 210, 244–5 fountain, Great Mosque of Aleppo, 30, 30 Great Mosque of Cordoba, 243, 244

508

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

domes (cont.) Great Mosque of Damascus, 443 Madrasa al-Firdaws, Aleppo, 85, 89 Madrasa al-Mirjaniyya, Baghdad, 176 maqßËra domes, 136, 138, 143, 149, 182 Mår Bahnåm monastery, Mosul, 137, 149 Mashhad al-Dakka, Aleppo, 30 Mashhad al-Husayn, Aleppo, 89 Masjid al-Haram, Mecca, 249–50, 251–2 Masjid-È JåmÈ, Isfahan, 138, 226, 226 mausoleum of NËr al-DÈn, Damascus, 208, 209, 212, 213, 229 mausoleum of Zumurrud Khatun, Baghdad, 164–5, 208, 209, 214–15 mirror-inlaid domes, 445, 453, 472–3, 473 Mosque al-NËri, Mosul, 133, 136–8, 143, 144, 149, 182, 184 Mosque al-Qarawiyyin, Fez, 205–6, 206, 207, 229, 230, 249 Mouchroutas dome, Istanbul, 252 muqarnas domes, 137, 163–5, 164, 191–2, 192–3, 201–18, 203–4, 206, 207, 209, 212–18, 224–33, 225–33, 244, 250, 252, 344, 445, 453 palatial domes, 252 pyramidal domes, 137, 189, 210, 244 Qubbat al-Barudiyyin, Marrakech, 227–8, 229, 235, 240–53, 242, 246–7, 249 ribbed domes, 137, 244 Seljuq domes, 202, 227 shrine of al-Faraj ibn Barquq, Cairo, 225, 233 shrine of Imam ʿAli, Najaf, 472 shrine of Imam ʿAwn al-DÈn, Mosul, 192, 193, 210, 212, 217–18 shrine of Imam Dur, Samarra, 164, 202–4, 203–4, 206, 207, 226–7, 228, 246

shrine of Imam Yahya b. alQasim, Mosul, 189, 191, 192, 210 shrine of al-Najmi, Shatt al-Nil, 208, 209–10, 216 shrine of Sayyida Ruqayya, Damascus, 439, 443, 445, 453, 454, 466, 473, 473 shrine of Sayyida Sukaina bint ʿAlÈ, Darayya, 473 shrine of Sayyida Zaynab, Damascus, 443 shrine of Sayyida Zaynab, Sinjår, 137 shrine of Shaykh ʿAbd al-Samad, Natanz, 210 shrine of ʿUmar al-Suhrawardi, Baghdad, 159, 161, 164, 165 triple shrine, Raqqa, 443 double-shell domes, 137, 210, 244–5 dragons, 51–2, 161 Dunaysir, 37, 40, 138 durgahs, 98 Eastern KËfic script, 172, 269–70, 271, 290, 294, 323 Ecochard, Michel, 6 ecumenism, 141, 159, 193 Edessa, 3, 4, 5, 31, 37–8, 40, 130, 131, 329, 482 Egypt calligraphy, 285, 300, 314, 315–18, 328, 340–1, 342 hospitals, 481, 482, 485, 488–90 NËr al-DÈn’s conquest of, 13–14, 16, 179, 329, 340 ornament, 107, 108–9, 110, 113, 119, 120 woodwork, 107, 108–9, 110, 113 see also individual locations Elisséeff, Nikita, 3, 194 entablatures Dayr al-Zaʿfaran, 43 gates of Cairo, 9–10 Great Mosque of Aleppo minaret, 9–10 Great Mosque of Harran, 43 Madrasa al-Hallåwiyya, Aleppo, 6 NËr al-DÈn minbar, Hama, 111 Qastal al-Shuʾaybiyya, Aleppo, 7–10, 9, 40, 43

509

index

entrance blocks Aleppo citadel, 47, 49, 49, 51–2, 61, 64, 97, 98, 99, 101, 102, 334 Qalʾat Najm, 98, 102 epigraphy see monumental calligraphy Epistle of al-Qadir, 299–300 esoteric (bå†in) interpretations, 250, 299, 312, 321, 343 essentialism, 26–7, 106–7 Ettinghausen, Richard, 174, 311–12, 313, 398, 404, 426 exoteric (Ωåhir) interpretations, 143, 184, 250, 299, 312, 321, 343, 355 exterior arcades, 100 exterior iwåns, 86, 87, 376 Fa∂åʾil b. Ya˙yå, 112 faith healing see spiritual medicine al-Fasi, Abu ʿImran, 250 Fatimids, 28, 35, 80, 107–9, 120, 159–60, 180, 226, 232, 250, 297–300, 310, 312, 314–21, 325, 329, 340–1, 343 La Favara pavilion, Sicily, 252, 380 female patronage, 75–90 female piety, 80–1 female pilgrims, 457, 462, 464 female status, 75–90 Fez Mosque al-Qarawiyyin, 205–6, 206, 207, 229, 230, 248, 249, 318, 338–9, 338–9 al-Firdaws pavilion, Mardin, 382, 382–4, 415–17, 416 ‘Flood of Baghdad, The’, 163, 204 floral ornament, 40, 108, 120, 443 floriated KËfic script, 108, 108, 111, 140, 189, 296, 312–21, 314, 318–20, 327, 330–1, 335, 338–40, 341–3 Flury, Sam, 314–15 foliate arches, 43, 228, 241 foliated KËfic script, 9, 315 fortifications, 47, 94–103, 159, 161–2, 173; see also castles; citadels; city walls; fortresses fortresses, 95, 96–100

foundation inscriptions, 83, 83, 114, 167, 189, 331, 333, 334, 339, 469 fountains Alhambra Palace, Granada, 52, 393, 393, 394, 396–7, 397, 419–20, 420 Artuqid palace, Diyarbakir, 99, 384 Ayyubid palace, Aleppo, 56–7, 99, 417 Ayyubid palace, Qalʿat ÍahyËn, 99, 384–5, 386 Balkuwara palace, Samarra, 371 in citadels, 412–20 Qalʾat BanÈ Hammåd, 378–80 Qastal al-Shuʾaybiyya, Aleppo, 7–10, 8–9, 18, 20, 39, 40, 42–3 al-Firdaws pavilion, Mardin, 382, 415–17 Fountain of the Lions, Alhambra Palace, 52, 394, 396–7, 397, 419–20, 420 in gardens, 370, 371–3, 373, 376, 378–85, 391–4, 393, 396–7, 397, 405, 410–20, 415, 417, 420 Great Mosque of Aleppo, 30, 30 Great Mosque of Samarra, 372–3, 373 Óayr al-Wu˙Ësh, Samarra, 372, 410, 411 Jawsaq al-Khaqani, Samarra, 371 Khirbat al-Mafjar, 370, 371 Lashkar-È Båzår, 376 Qubbat al-Barudiyyin, Marrakech, 239–40, 250, 251, 253 residential houses, 383–4 salsabÈl fountains, 379–84, 381, 413–14, 415, 417, 417 shådirwån fountains, 56–7, 99, 378–84, 380, 413–15, 417, 421 La Ziza, Sicily, 414, 415 zoomorphic fountains, 393, 393, 410, 411, 415 four-iwån plans, 53–5, 57, 99, 167, 206, 417, 483 Fowden, Garth, 426, 427, 429 frescos, 424–35, 425, 428–30 friezes Aleppo citadel entrance block, 99 BÈmåristån al-NËrÈ, Damascus, 110, 110, 140

510

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

friezes (cont.) Great Mosque of Aleppo minaret, 33, 327–8 Great Mosque of Damascus, 433 Great Mosque of Tlemcen, 337, 338 Madrasa al-Firdaws, Aleppo, 334, 334 Madrasa al-Mustansiriyya, Baghdad, 167 Mosque al-Håkim, Cairo, 317 Mosque of NËr al-DÈn, Hama, 140 Mosque al-NËri, Mosul, 19, 138, 140, 151, 182, 184, 186, 336 Mosque al-Qarawiyyin, Fez, 339 NËr al-DÈn minbar, Hama, 111 Qastal al-Shuʾaybiyya, Aleppo, 7, 9 shrine of Imam ʿAwn al-DÈn, Mosul, 336–8, 337 shrine of Imam Yahya b. alQasim, 190, 190 shrine of Sayyida Ruqayya, Damascus, 444, 469 frontispieces, 120, 172, 275 Fus†å†, 383–4, 384 futuwwa, 160, 161, 193 al-Gailani, ʿAbd al-Qadir, 159 Galenic medicine, 478, 480–1, 491, 494 gardens Alhambra Palace, Granada, 369, 388, 388, 393, 394–7, 395–7, 419–20, 420 artistic representations, 381, 381, 408, 408–9, 414, 424–35, 425–6, 428–32 Artuqid Palace, Diyarbakir, 384, 385 Ayyubid Palace, Qalʾat ÍahyËn, 384–5, 386 Balkuwara palace, Samarra, 371, 372, 376, 409 in citadels, 412–20 courtyard gardens, 369, 371, 375, 379–80, 383–7, 394–7, 409 La Cuba, Sicily, 414 Dar al-Khilafa, Baghdad, 162 al-Firdaws, Mardin, 382, 382–4, 415–17, 416

fountains, 370, 371–3, 373, 376, 378–85, 391–4, 393, 396–7, 397, 405, 410–20, 415, 417, 420 ˙ayr gardens, 370–1, 373, 409–11 Óayr al-Wu˙Ësh, Samarra, 371–2, 409–11 hydraulic technologies, 389–93, 390–2, 404–5, 413–14, 414 ʿImårat-i Khusrau, Qasr-i Shirin, 370 jahår-bågh gardens, 370, 371, 373, 409, 424, 425 Jawsaq al-Khaqani, Samarra, 371 Khirbat al-Mafjar, 371, 408, 426 Lashkar-È Båzår, 373, 375–6, 377–8 MadÈnat al-Zahråʾ, Cordoba, 373–5, 374–7, 387, 411–12, 412 Madrasa al-Firdaws, Aleppo, 86 Madrasa al-Mustansiriyya, Baghdad, 167 Mesopotamia, 369, 370 North Africa, 371, 376 palace gardens, 162, 369–98, 408–20 and paradise, 86, 369–70, 388–9, 396–7, 405, 407–9, 420–1, 424–35 Pasargadae, 370 pavilions, 252, 370–3, 375–6, 376, 379–82, 382–4, 387–8, 392, 404, 410–18 Persia, 369, 370, 371, 376, 387, 388, 404, 426 and poetry, 381–2, 393–4, 397, 404–21, 426 pools, 371–3, 376, 377, 379–87, 389, 393, 393–4, 409–15, 417–19 Qalʿat banÈ Hammåd, 378–80, 379–80 Qasr al-Mubårak, Seville, 386–7, 387 terraces, 370, 371, 372, 375–6, 387, 411–12 towers, 375, 387–8, 388, 396, 412 walled gardens, 370, 375 and water, 369–98, 404–21 La Ziza, Sicily, 252, 380–1, 380–2, 414, 415 zoological gardens, 370, 371–2

511

index

gates Antioch gate, Aleppo, 7, 9 Bab al-ʿAmma, Samarra, 29, 30 Bab Dar al-ʿAdl, Aleppo, 49 Bab al-Futuh, Cairo, 35–7 Bab al-Jabal, Aleppo, 49 Bab al-Maqam, Aleppo, 49, 51, 63 Bab al-Nasr, Cairo, 33, 35–7 Bab al-Saghir, Aleppo, 49 Bab Zuwayla, Cairo, 35, 35–7 Basaliya Gate, Baghdad, 161 Dhafariya (Wastani) Gate, Baghdad, 161–2, 162 Gate of the Serpents, Aleppo citadel, 51–2, 64 Gate of the Two Lions, Aleppo citadel, 51–2, 333, 334 gates of Cairo, 9–10, 33, 35, 35–7, 109, 340–1, 341 gateway of Rusafa, 43 Great Mosque of Cordoba, 29, 30 Halaba (Talisman) Gate, Baghdad, 161 Muʾazzam (North) Gate, Baghdad, 161 Mudarraj Gate, Cairo, 340–1, 341 in northern Syrian architecture, 28–37, 29 Porta Aurea, Palace of Diocletian, 28, 30 Roman gates, 28, 30, 55 al-Ghadåiri, Abu-l-Hasan, 7 see also entrance blocks; portals geometric ornament, 52, 105–20, 109–18, 162, 164, 169–70, 172, 189, 192, 247–8, 297 geometric plans, 167–8, 243, 243–4 George, Alain, 174 Ghaznavids, 55, 95, 298, 322–5, 327, 328, 342, 373, 375–6 al-Ghazzali, 250 GhËrids, 323 al-GhuzËlÈ, 394, 419 Gibb, Hamilton, 7, 329 girih ornament, 111, 119 glacis, 46, 98, 101 glazed tiles, 189, 191, 210, 443, 445, 473 GökburÈ, MuΩaffar al-DÈn, 114

Grabar, Oleg, 27, 174, 201, 210–11, 213, 235, 312–13, 362, 426, 427, 433 Granada Alhambra mosque, 100 Alhambra Palace, 52, 210–13, 369, 385, 388, 388, 393, 394–7, 395–7, 419–20, 420 Great Mosque of Aleppo, 5, 7, 9–10, 13, 14, 16–17, 30, 30–2, 32–5, 34, 36, 112, 314, 317, 320, 327–8, 329 Great Mosque of Cizre, 192 Great Mosque of Cordoba, 29, 30, 243, 244, 247, 248, 251 Great Mosque of Diyarbakir, 138, 317 Great Mosque of Damascus, 41, 138, 141, 408, 408–9, 425, 426, 433–4, 443, 462 Great Mosque of Dunaysir (Kiziltepe), 37, 40, 138 Great Mosque of Harran, 42, 42–3 Great Mosque of Isfahan see Masjid-È JåmÈ, Isfahan Great Mosque of Maʾarrat alNuʾman, 17, 19, 35 Great Mosque of Mardin, 40–1, 138 Great Mosque of Mayyafariqin, 41, 138 Great Mosque of Raqqa, 135–6 Great Mosque of Samarra, 372–3 Great Mosque of Tlemcen, 248, 337, 338 Great Mosque of Veramin, 202 Grohmann, Adolf, 315, 318 hadith, 111, 188, 429, 447, 450, 467, 469, 470–1, 472 hadith inscriptions, 447, 450, 455, 469, 470–1, 471 al-Óåkim, Caliph, 325 Halaba (Talisman) Gate, Baghdad, 161 al-ÓalabÈ, Óåmid b. Ûåfir, 112 Hall of the Two Sisters, Alhambra Palace, 211, 213, 395 Hama citadel, 98 hospitals, 482

512

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Hama (cont.) Mosque of NËr al-DÈn, 11, 11–13, 19, 20, 110, 111, 140 NËr al-DÈn minbar, 11–13, 14, 20, 110, 111, 112, 331 al-Hamadhani, Badiʾ al-Zaman, 174 Óamdanids, 30, 47 Óammadids, 415 Hammam al-Sarah, 426 Óanafis, 48 al-Harawi, 491 harems, 77, 162 Óårim, 96, 98, 99 al-Óariri, 173–4 al-ÓårithÈ, Muʾayyad al-DÈn AbËʾl-Fa∂l b. ʿAbd al-KarÈm Mu˙ammad, 110–11 Óarran, 42–3, 482 al-ÓarrånÈ, AbË ʿl-Óusayn Mu˙ammad, 115 Harun al-Rashid, Caliph, 479 Hasankeif see Óisn Keyfa ˙ayr gardens, 370–1, 373, 409–11 Óayr al-Wu˙Ësh, Samarra, 371–2, 409–11 hazar baf ornament, 142, 163, 164, 185 Hernandez Jimenez, Felix, 375 Herzfeld, Ernst, 7, 9, 27, 114, 118, 132–7, 140, 143, 165, 168, 182–4, 218, 321, 327, 334, 371 ÓijåzÈ Qurʿanic manuscripts, 273 Hilal al-Sabiʾ, 55 Hill, Donald, 404 Hillenbrand, Robert, 142 Hiraqlah, 54, 67 Óisn Keyfa, 37, 40, 143 historical inscriptions, 7, 11, 13–16, 82–4, 248, 340 Homs, 98 horizontal format manuscripts, 272–5 horseshoe arches, 241, 444 hospitals al-ʿAdudi hospital, Baghdad, 480, 481, 493 Aleppo, 481, 482, 485, 492 Anatolia, 481, 482, 485, 492 Baghdad, 477–82, 493 BÈmåristån of Arghun al-Kamili, Aleppo, 486, 490, 492

BÈmåristån al-Nasiri, Cairo, 488 BÈmåristån al-NËrÈ, Aleppo, 55, 492 BÈmåristån al-NËrÈ, Damascus, 110, 110, 140, 206–7, 207, 229, 230, 331, 482–3, 483–6, 492–3 Cairo, 481, 482, 485, 488–90 Çankiri hospital, 492 Christian hospitals, 479, 480, 482 Damascus, 481, 482–3, 485 Darushifaʾ al-Ghiyathiyya, Kayseri, 487 decline of, 490–3 Egypt, 481, 482, 485, 488–90 etymology and terminology, 478–9 Europe, 485–7 four-iwån plans, 206, 483 Kastamonu hospital, 492 Madrasa al-Mustansiriyya, Baghdad, 167 mental patients, 479, 488, 489–90 mortuaries, 488 al-Muʾayyadi hospital, Cairo, 490–1 patronage, 478, 480, 481–2 Qalawun’s hospital, Cairo, 488–90, 489 tripartite courtyard façades, 55 waqf endowments, 478, 487–8, 492–4 wards, 488 humoural medicine, 478, 480–1, 494 Humphreys, R. Stephen, 75, 80, 96 hunting, 424–35 al-Óusayn, Imåm, 113, 441, 466, 472 Hussein, Saddam, 455 hydraulic technologies, 389–93, 390–2, 404–5, 413–14, 414 Ibn ʿAbd al-Hådi, 441 Ibn AbÈ ʿAßrËn, Sharaf al-DÈn, 131 Ibn AbÈ Êayyiʾ, 328 Ibn AbÈ Ußaybiʿa, 110 Ibn al-ʾAdÈm, 5, 94, 115 Ibn ʿAnbah, 165 Ibn ʿAsåkir, 441 Ibn al-AthÈr, 15, 16, 17, 94, 135, 165–6 Ibn al-Balkhi, 50

513

index

Ibn al-Baßßal, 413 Ibn Battuta, 165 Ibn al-Bawwab, 172, 173, 248, 261, 277–91, 292, 297–300, 310, 324, 328, 339, 343, 354–5, 362 Ibn DurustËyah, 356 Ibn al-Fuwati, 165 Ibn ÓamdÈs, 393, 415 Ibn Óanbal, 361 Ibn Hubayra, 329 Ibn Jubayr, 13, 76, 112, 226, 249, 484 Ibn Kathir, 165, 279 Ibn Khafaja, 418–19 Ibn KhaldËn, 94, 95 Ibn Khallikån, 279 Ibn al-Khashshåb, 5–6, 328 Ibn LËyËn, 387 Ibn MasʿËd, 297, 299, 360 Ibn Miqsam, 297, 360 Ibn Mujåhid, A˙mad, 297 Ibn Muqla, 172, 260, 261, 263–77, 285, 292, 294–8, 300, 310, 324, 354–63 Ibn al-Nadim, 262, 298, 356 Ibn al-Qalånißi, 12 Ibn al-Qaysaråni, 12 Ibn Shaddåd, 9, 49, 77, 80, 95, 382, 441 Ibn ShanabËdh, 297–8, 360 Ibn Wahb al-Kåtib, 262 Ibn Wåßil, 76–7, 78, 79–80 Ibn Zamrak, 213, 419–20 Ibrahim, 48 IbråhÈm, Abuʾl-MuΩaffar, 324 Ibrahim b. ʿAli, 190 Ilkhanids, 165, 173, 175, 202 illustrated manuscripts, 172, 173–5, 194 ʿImåd al-DÈn II, 130–1 ʿImåd al-DÈn al-Ißfahåni, 15–16, 130 Imami Shiʿism see Twelver Shiʿism ʿImårat-i Khusrau, Qasr-i Shirin, 370 Imperial Palace, Constantinople, 252 ʿImruʾl-Qays, 406, 410, 419, 420 India, 323, 369 inscriptions see foundation inscriptions; hadith

inscriptions; historical inscriptions; monumental calligraphy; Qurʾanic inscriptions intentionality, 106–7, 119–20, 129–30, 181, 464, 465 intercession, 465, 468–71 interconnection (of script), 173, 282, 324, 331, 334, 336–8, 339, 343 Iqbål al-ÛåhirÈ al-KhåtËnÈ, Jamål al-Dawla, 79 Iran architecture, 138, 142, 167, 168, 176, 202, 224–6, 232 building of shrines in Syria, 439–40, 443, 453–8, 462 calligraphy, 172, 173, 289, 317, 318, 325–6, 335, 343 ornament, 107, 120, 142 Revolution, 455, 456 see also Persia; and individual locations Iran–Iraq war, 455–6 Iraqi Institute for Antiquities, 133, 182 Iraqi Museum, Baghdad, 140, 141, 184, 186 Irbil, 97, 98, 114, 143 irrigation, 371, 389, 390, 413, 426 Isfahan Masjid-È JåmÈ, 41, 138, 202, 226, 226, 325 Ishraqi Sufism, 469 Islamic Museum, Cairo, 341 Ismåʾil, al-Malik al-Salih, 16, 48, 111, 331 Ismåʾilis, 159, 180, 250, 299, 312, 320–1, 325, 329 Istanbul, 192, 252 Italianate windows, 41 iwåns ʿAbbasid Palace, Baghdad, 168, 169–70 Ayyubid palace, Aleppo, 55–6, 417 BÈmåristån al-NËrÈ, Damascus, 483 exterior iwåns, 86, 87, 376 al-Firdaws pavilion, Mardin, 382, 383, 416, 417

514

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

iwåns (cont.) four-iwån plans, 53–5, 57, 99, 167, 206, 417, 483 khånqåh al-Farafra, Aleppo, 85 Lashkar-È Båzår, 376 Madrasa al-Firdaws, Aleppo, 85, 86, 87 Madrasa al-Muqaddamiyya, Damascus, 439 Madrasa al-Mustansiriyya, Baghdad, 167 Ma†bakh al-ʿAjami, Aleppo, 56, 57 shrine of Imam ʿAli, Najaf, 472 shrine of Sayyida Ruqayya, Damascus, 443, 444 Taq-e Bostan, 425, 431–2, 431–3 tripartite iwån compositions, 55–6, 57, 100, 167 Jaʿfar al-Sadiq, 467 jahår-bågh gardens, 370, 371, 373, 409, 424, 425 Jahili poetry, 406–8 Jalayirids, 175 James, David, 269–70, 281, 290 Jåmiʿ al-Óanåbila minbar, Damascus, 114–15, 114–15 Jannat al-Baqiʾ cemetery, Medina, 195 al-Jawad, Muhammad, 160 Jawsaq al-Khaqani, Samarra, 29, 30, 54, 371, 411 al-JazarÈ, 390–1 Jerusalem al-Aqsa mosque, 13–17, 21, 110, 111–13 citadel, 99 Dome of the Rock, 41, 235 hospitals, 482 monumental calligraphy, 334 NËr al-DÈn minbar, 13–17, 21, 110, 111–13, 112–13 Saladin’s conquest of, 3, 15–17, 180, 340 as site of pilgrimage, 181 jihåd, 3–21, 42, 140, 250 jizya tax, 131, 181 joggled voussoirs, 37, 52, 53, 192 Jumʾah, Ibrahim, 317

jurisprudence, 80, 159, 160, 167, 491, 494 justice, 4, 12–13, 50–1 Juyushi Mosque, Cairo, 48 Kalila wa Dimna, 173 Karak, 55, 99 Karbala, 445, 454–5, 466 Karkh, 159 Kastamonu hospital, 492 Katkhuda, ʿAbd al-Rahman, 490 Kayseri, 487 al-Kazim, Musa, 160 Kazimayn shrine, Baghdad, 445, 454, 466 al-Kazimiyya, 159–60 KhadÈja KhåtËn, 80 al-Khaffafin mosque, Baghdad, 163 Khan Mirjan, Baghdad, 176 khånqåhs conversion to Shiʿi shrines, 193–4 female piety, 80–1 khånqåh al-Farafra, Aleppo, 81, 81–7, 83, 84 Madrasa al-Firdaws, Aleppo, 81–90, 83, 85–9 patronage, 80–90 al-kha†† al-mansËb (proportional writing), 266, 294, 298, 355, 357, 360, 362 al-KhåyqånÈ, A˙mad ibn AbiʾlQåsim, 269 khettara water-supply system, 251 Khirbat al-Mafjar, 30, 370, 371, 408, 426, 435 Khomeini, Ayatullah, 455, 457 Khosrow II, 431 Khwarazmshahs, 160, 161 kiosks, 162, 187 Kitåb al-Aghani, 194 Kitåb akhbår al-na˙wiyyÈn albaßriyyÈn, 358, 359 Kitåb al-Sabʿa (Ibn Mujåhid), 297 kitåbkhana, 355 Kiziltepe see Dunaysir Konya, 187 Kufa, 47, 54, 56, 172, 298 KËfic script Ashkelon minbar, 318 al-Azhar mosque, Cairo, 314, 315, 317, 318

515

index

cenotaph of Sayyida Få†ima, Damascus, 108, 108, 317, 318, 320 cenotaph of Sayyida Sukaina, Damascus, 318 on coins, 322 Eastern KËfic script, 172, 269–70, 271, 290, 294, 323 floriated KËfic script, 108, 108, 111, 140, 189, 296, 312–21, 314, 318–20, 327, 330–1, 335, 338–40, 341–3 foliated KËfic script, 9, 315 Great Mosque of Aleppo minaret, 314, 317, 320, 327 Great Mosque of Diyarbakr, 317 Mashhad al-Dakka, Aleppo, 328 Masjid-È Óaydariyya, QazvÈn, 318 monumental calligraphy, 7, 9, 43, 140, 141, 184, 189, 312–27, 314, 316, 318–20, 328, 330–1, 338–40, 341–3 Mosque al-Håkim, Cairo, 314, 314, 315, 317, 319, 320 Mosque al-NËri, Mosul, 140, 141, 184, 335 Mosque al-Qarawiyyin, Fez, 318, 338–9 Mosque al-Íåli˙ Êalåʾiʾ, Cairo, 340 new ʿAbbasid KËfic script see semi-KËfic script Qastal al-Shuʾaybiyya, Aleppo, 7, 9, 43, 317, 330–1 Qurʾanic calligraphy, 172, 262–77, 264, 271, 273–8, 285, 291–2, 294, 296, 297, 300, 301, 342–3, 355–8, 362 secular calligraphy, 357, 358 semi-KËfic script, 172, 267–77, 273–8, 285, 291, 342, 355, 357, 362 shrine of Imam Yahya b. alQasim, Mosul, 189 Syrian woodwork, 108, 108, 111 Kühn, Miriam, 106 Kühnel, Ernst, 292–3 Kunstwollen, 26 Kurd ʿAlÈ, Muhammad, 442

Lådhiqiyya, 96 Lashkar-È Båzår, 55, 56, 373, 375–6, 377–8 latrines, 483, 484 Lebanon, 440, 454, 462 legibility monumental calligraphy, 311, 324, 328, 330, 331, 333, 334, 336–7, 343, 445, 450, 470 Qurʾanic calligraphy, 270, 276, 281, 282, 291, 299, 300, 355, 356, 360 legitimacy, 4, 48, 158, 176, 218, 237, 299, 310, 330, 343–5, 452 lintels, 56, 192 lions, 52, 161, 162, 393, 396–7 literacy, 293–4, 355, 360, 362 literature, 3, 160, 173–4, 261, 262, 356–7; see also poetry loopholes, 98, 161 MaʿålÈ b. Sålim, 111, 112, 113 Maʾarrat al-Nuʾman citadel, 98 conquest by al-Ûåhir GhåzÈ, 97 Great Mosque, 17, 19, 35 machicolations, 98–9, 101 MadÈnat al-Zahråʾ, Cordoba, 247, 248, 373–5, 374–7, 387, 411–12, 412 madrasas conversion to Shiʿi shrines, 193–4, 439–40 four-iwån plans, 53–4, 167 Madrasa al-Badriyya, Mosul, 187 Madrasa al-Firdaws, Aleppo, 81–90, 83, 85–9, 334, 334 Madrasa of Gümüshtekin, Bosra, 55 Madrasa al-Hallåwiyya, Aleppo, 4–6, 18, 20, 115, 116–17, 328, 329, 330 Madrasa al-Måridåniyya, Damascus, 80, 115 Madrasa al-Mirjaniyya, Baghdad, 176 Madrasa al-Mujåhidiyya al Juwwåniyya, Damascus, 439 Madrasa al-Muqaddamiyya, Damascus, 5, 439, 445

516

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

madrasas (cont.) Madrasa al-Mustansiriyya, Baghdad, 160, 165–8, 166, 168, 169, 170, 173 Madrasa of al-Íali˙ Najm al-DÈn AyyËb, Cairo, 344 Madrasa al-Íå˙ibiyya, Damascus, 80 Madrasa al-Shåmiyya, Damascus, 80 Madrasa al-Sharabiyya, Baghdad see ʿAbbasid Palace, Baghdad Madrasa of Sultan Hassan, Cairo, 232, 232, 233 Madrasa al-Sultaniyya, Aleppo, 48–9, 49, 61 Madrasa al-Zahiriyya, Aleppo, 231 Madrasa al-Zajjåjiya, Aleppo, 5 patronage, 80, 132, 482 tripartite courtyard façades, 55, 167 Ma˙mËd of Ghazna, 323, 325 Maillet, Benoît de, 490 al-Malik al-ʿÅdil, 78, 95, 97 Malikis, 250 Mamluks, 50, 60, 80, 90, 99, 106, 120, 224, 232–3, 266, 490–1, 492, 494 al-Ma ʾmËn, Ya ˙ya ibn IsmåʾÈl, 392 Manbij, 97 al-Mansur, 415 manΩaras, 387, 396, 412 Maqam Ibrahim, Aleppo citadel, 47, 48, 60, 110, 111, 111, 112, 115, 331, 332 Maqam Ibrahim, Southern Aleppo, 48, 51 Maqamat (al-Hariri), 173–5 Maqamat genre, 160, 173–5 al-Maqrizi, Taqi al-Din Ahmad, 37, 391 maqßËra domes, 136, 138, 143, 149, 182 Mår A˙Ëdemmeh monstery, Mosul, 136, 192 Mår Bahnam monastery, Mosul, 137, 149, 192 Mår Ishaʾya church, Mosul, 136 Mår JurjÈs church, Mosul, 136

Mar Tuma church, Mosul, 482 Mardin Christianity, 131–2, 143 Dayr al-Zaʿfaran, 43 al-Firdaws pavilion, 382, 382–4, 415–17, 416 Great Mosque, 40–1, 138 Monastery of the Virgins (Dayr al-Abkår), 131–2 residential architecture, 38, 41 St Thomas’ church, 132 måriståns see hospitals Marrakech Dar al-Hajar, 237–8, 252 al-Kutubiyya mosque, 237–8 Minbar al-Kutubiyya, 237 Mosque of ʿAli b. Yusuf, 237, 239–40 Qubbat al-Barudiyyin, 227–8, 229, 235–53, 236, 241–3, 245–9 water supply, 237, 240, 251 marriage alliances, 77–9 Mashhad, 445, 455, 466 Mashhad al-Dakka, Aleppo, 4, 30, 328 Mashhad al-Óusayn, Aleppo, 89, 441, 442, 449 mashhads see shrines Masjid al-Haram, Mecca, 249–50, 251–2 Masjid al-Khulafaʾ see Mosque of the Caliphs, Baghdad Masjid-È Óaydariyya, QazvÈn, 318 Masjid-È JåmÈ, Isfahan, 41, 138, 202, 226, 226, 325 Maslow, Boris, 239 Massignon, Louis, 404 Masʾud, Sultan, 160 Ma†bakh al-ʿAjami, Aleppo, 55, 56, 57, 68 mausoleums Arab Ata mausoleum, Tim, 202, 226 mausoleum of NËr al-DÈn, Damascus, 208, 209, 212, 213, 229 mausoleum of al-Ûåhir GhåzÈ, Aleppo, 48–9, 61 mausoleum of Zumurrud Khatun, Baghdad, 164–5, 208, 209, 214–15

517

index

patronage, 80 see also shrines Mayyafariqin, 41, 138 Mecca, 76, 140, 181, 226, 249, 251–2, 406, 466, 493 medicine, 478, 479–82, 487, 491, 493, 494 Medina, 76, 181, 195, 406 Mernissi, Fatima, 75 Merv, 54 Mervin, Sabrina, 457 Messick, Brinkley, 360 metalwork, 106, 194 Metropolitan Museum of Art (MMA), New York manuscript 40.164, 276, 278 manuscript 45.10, 272, 278 Meunié, Jacques, 239 Meyerhof, Max, 481 Michael the Syrian, 132, 181 mi˙råbs al-Aqsa mosque, Jerusalem, 17 al-Azhar mosque, Cairo, 317 khånqåh al-Farafra, Aleppo, 85 Madrasa al-Firdaws, Aleppo, 85, 88 Madrasa al-Hallåwiyya, Aleppo, 115, 116–17 Maqam Ibrahim, Aleppo citadel, 110, 111, 111, 115 Mosque al-Håkim, Cairo, 317 Mosque of NËr al-DÈn, Hama, 11 Mosque al-NËri, Mosul, 133–5, 136–7, 138, 139, 140, 146–7, 182–3, 335, 335 shrine of Imam Yahya b. alQasim, Mosul, 191, 191 shrine of Sayyida NafÈsa, Cairo, 108, 109, 109 shrine of Sayyida Ruqayya, Cairo, 108, 109 shrine of Sayyida Ruqayya, Damascus, 444 minarets Bab Sharqi minaret, Damascus, 17, 22 built by NËr al-DÈn, 17–19, 21, 22 citadel mosque, Aleppo, 47–8, 60, 100

Great Mosque of Aleppo, 9–10, 31, 32, 32–5, 34, 36, 314, 317, 320, 327–8, 329 Great Mosque of Damascus, 443 Great Mosque of Maʾarrat alNuʾman, 17, 19, 35 Great Mosque of Mardin, 40–1 Juyushi Mosque, Cairo, 48 al-Khaffafin mosque, Baghdad, 163 al-Kutubiyya mosque, Marrakech, 238 Mosque of ʿAli b. Yusuf, Marrakech, 237 Mosque al-NËri, Mosul, 133, 134, 142–3, 152, 185–6, 187 Mosque of Süleyman Pasha, 37, 40 Mosque of the Caliphs, Baghdad, 163 Qumriyya mosque, Baghdad, 163 surveillance function, 48 as symbolic of Islam, 17–19 minbars Ashkelon minbar, 108–9, 318 functions of, 12–13 Jåmiʿ al-Óanåbila minbar, Damascus, 114–15, 114–15 Minbar al-Kutubiyya, Marrakech, 237 NËr al-DÈn minbar, al-Aqsa, 13–17, 21, 110, 111–13, 112–13 NËr al-DÈn minbar, Hama, 11–13, 20, 110, 111, 112, 331, 332 political significance, 12–13 Mirdasids, 47, 95 mirror-inlaid domes, 445, 453, 472–3, 473 moats, 46, 49, 98, 161 monasteries confiscation of endowments, 181 destruction and pillaging of, 181 Mår A˙Ëdemmeh monastery, Mosul, 136, 192 Mår Bahnåm monastery, Mosul, 137, 149, 192 Monastery of the Virgins (Dayr al-Abkår), Mardin, 131–2 as places of cure, 482 see also Christian architecture

518

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Mongols, 60, 97, 100, 158, 159, 160, 173, 175, 176, 179, 266, 321, 480 monumental calligraphy Anatolia, 40, 43, 317, 333 Baghdad, 161, 173, 248, 324, 342–3 canonisation, 333–8 cursive script, 40, 43, 138, 143, 190, 248, 248, 314, 321–45, 322, 324–6, 329–39, 341–2, 344 Eastern KËfic script, 323 Egypt, 113, 314, 315–18, 328, 340–1, 342 foliated KËfic script, 315 floriated KËfic script, 140, 189, 312–21, 314, 318–20, 327, 330–1, 335, 338–40, 341–3 interconnection, 324, 331, 334, 336–8, 339, 343 Iran, 317, 318, 325–6, 335, 343 KËfic script, 7, 9, 43, 140, 141, 184, 189, 312–27, 314, 316, 318–20, 328, 330–1, 335, 338–40, 341–3 legibility, 311, 324, 328, 330, 331, 333, 334, 336–7, 343, 445, 450, 470 North Africa, 318, 324, 333, 338–40, 342 Palestine, 315, 317, 318, 334 research approaches, 259–60, 311–13 role of NËr al-DÈn, 143, 314, 321, 327–33, 335–6, 340, 342 Syria, 314, 317–18, 324, 327–34, 342 thuluth script, 11, 13–15, 138, 161, 173, 190–2, 248, 266, 323, 325, 328, 331–9, 331–40, 342 transformation from angular to cursive, 310–45 Upper Mesopotamia, 333, 335–8 see also foundation inscriptions; hadith inscriptions; historical inscriptions; Qurʾanic inscriptions Morocco, 120, 227–9, 235–53; see also North Africa; and individual locations

mosaics, 141, 384, 408, 408–9, 425, 426, 433–4 mosques Abu Dulaf mosque, Samarra, 54 Aleppo citadel, 47–8, 60, 100, 333, 334 Alhambra mosque, Granada, 100 al-Aqsa mosque, Jerusalem, 13–17, 21, 110, 111–13 al-Azhar mosque, Cairo, 314, 315, 317, 318 Bab Mardum mosque, Toledo, 240, 244 Bayn al-Qahaoui mosque, Sousse, 249 in citadels, 47–8, 60, 100, 103 conversion of churches to, 5–6, 131–2 Great Mosque of Aleppo, 5, 7, 9–10, 13, 14, 16–17, 30, 30–2, 32–5, 34, 36, 112, 314, 317, 320, 327–8, 329 Great Mosque of Cizre, 192 Great Mosque of Cordoba, 29, 30, 243, 244, 247, 248, 251 Great Mosque of Damascus, 41, 138, 141, 408, 408–9, 425, 426, 433–4, 443, 462 Great Mosque of Diyarbakir, 138, 317 Great Mosque of Dunaysir (Kiziltepe), 37, 40, 138 Great Mosque of Harran, 42, 42–3 Great Mosque of Maʾarrat alNuʾman, 17, 19, 35 Great Mosque of Mardin, 40–1, 138 Great Mosque of Mayyafariqin, 41, 138 Great Mosque of Raqqa, 135–6 Great Mosque of Samarra, 372–3 Great Mosque of Tlemcen, 248, 337, 338 Great Mosque of Veramin, 202 Jåmiʿ al-Óanåbila, Damascus, 114–15, 114–15 Juyushi Mosque, Cairo, 48 al-Khaffafin mosque, Baghdad, 163 al-Kutubiyya mosque, Marrakech, 237–8

519

index

Masjid al-Haram, Mecca, 249–50, 251–2 Masjid-È Óaydariyya, QazvÈn, 318 Masjid-È JåmÈ, Isfahan, 41, 138, 202, 226, 226, 325 Mosque of ʿAli b. Yusuf, Marrakech, 237, 239–40 Mosque al-Aqmar, Cairo, 232 Mosque al-Håkim, Cairo, 10, 314, 314, 315, 317, 319, 320 Mosque of al-Muʿayyad Shaykh, Cairo, 232, 232 Mosque of NËr al-DÈn, Hama, 11, 11–13, 19, 20, 110, 111, 140 Mosque al-NËri, Mosul, 17, 17, 19, 129–44, 134, 139, 142, 144–52, 180–6, 182–7, 335–6, 335–6 Mosque al-Qarawiyyin, Fez, 205–6, 206, 207, 229, 230, 248, 249, 318, 338–9, 338–9 Mosque of al-Íåli˙ Êalåʾiʾ, Cairo, 340 Mosque of Süleyman Pasha, Hisn Keyfa, 37, 40 Mosque of the Caliphs, Baghdad, 160, 162, 163 Mosque of ʿUmar, Aleppo, 7 Mosul Museum, 186, 189, 192 patronage, 129–30, 132, 143 Qalʿat Najm citadel, 100, 103 Qumriyya mosque, Baghdad, 163 Tinmal Mosque, 244 Mosul Bash Tabiya, 187 Christianity, 129, 131–2, 140, 143, 180, 181, 192, 482 Madrasa al-Badriyya, 187 Mår A˙Ëdemmeh monastery, 136, 192 Mår Bahnåm monastery, 137, 149, 192 Mår Ishaʾya church, 136 Mår JurjÈs church, 136 Mar Tuma church, 482 Mongol invasion, 179 Mosque al-NËri, 17, 17, 19, 129–44, 134, 139, 142, 144–52, 180–6, 182–7, 335–6, 335–6 occupation by Daesh, 186 patronage, 129–30, 132, 143, 179–95

Qara Serai, 187 shrine of Imam ʿAwn al-DÈn, 187, 191–2, 193, 210, 212, 217–18, 336–7, 336–8 shrine of Imam Yahya b. al-Qasim, 186–91, 187–92, 210, 336 trade, 97, 98 Mouchroutas dome, Istanbul, 252 Mu ʿallaqa (Zuhair b. Abi Sulma), 406 al-Muʾayyad Shaykh, 490–1 al-Muʾayyadi hospital, Cairo, 490–1 Muʾazzam (North) Gate, Baghdad, 161 al-MuʾaΩΩam ʿIsa, 80 Muhammad ibn Musa ibn Jaʿfar, 202 mu˙aqqaq script, 266, 290, 294 muqarnas ʿAbbasid Palace, Baghdad, 169–70, 171 Alhambra Palace, Granada, 210–11, 212–13 Ayyubid palace, Aleppo, 52–3, 417 BÈmåristån al-NËrÈ, Damascus, 206, 207, 207, 229, 230 Great Mosque of Aleppo minaret, 34 al-Khaffafin mosque minaret, Baghdad, 163 Madrasa of Sultan Hassan, Cairo, 232, 232, 233 Madrasa al-Zahiriyya, Aleppo, 231 mausoleum of NËr al-DÈn, Damascus, 208, 209, 212, 213, 229 mausoleum of Zumurrud Khatun, Baghdad, 164–5, 208, 209, 214–15 mirror-inlaid muqarnas, 445, 453, 472–3, 473 Mosque al-Aqmar, Cairo, 232 Mosque of al-Muʿayyad Shaykh, Cairo, 232, 232 Mosque al-NËri, Mosul, 137 Mosque al-Qarawiyyin, Fez, 205–6, 206, 207, 229, 230, 338–9

520

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

muqarnas (cont.) muqarnas columns, 208 muqarnas domes, 137, 163–5, 164, 191–2, 192–3, 201–18, 203–4, 206, 207, 209, 212–18, 224–33, 225–33, 244, 250, 252, 344, 445, 453 muqarnas façades, 232 muqarnas niches, 247 muqarnas portals, 52–3, 202, 224, 229–33, 231–3 muqarnas vaulting, 52–3, 56, 109, 113, 118, 169–70, 171, 201–7, 206, 224–33, 230–3, 338–9, 378, 380, 414, 417, 443 Qalʿat banÈ Hammåd, 202, 378 Qubbat al-Barudiyyin, Marrakech, 227–8, 229, 243, 244 Qumriyya mosque minaret, Baghdad, 163 shrine of al-Faraj ibn Barquq, Cairo, 225, 233 shrine of Imam ʿAwn al-DÈn, Mosul, 192, 193, 210, 212, 217–18 shrine of Imam Dur, Samarra, 164, 202–4, 203–4, 206, 207, 226–7, 228, 246 shrine of Imam Shafiʾi, Cairo, 231, 232 shrine of Imam Yahya b. alQasim, Mosul, 191, 192, 210 shrine of al-Najmi, Shatt al-Nil, 208, 209–10, 216 shrine of Sayyida Ruqayya, Damascus, 443, 445, 453 shrine of Shaykh ʿAbd al-Samad, Natanz, 210 shrine of the ʿAbbasid Caliphs, Cairo, 232 shrine of ʿUmar al-Suhrawardi, Baghdad, 159, 161, 164, 165 La Ziza, Sicily, 380, 414 al-Muqtadir, Caliph, 297–8, 360–2 al-Muqtafi, Caliph, 160, 162 al-Murabitun see Almoravids Mußallå al-ʿÏdayn sanctuary, Damascus, 108 Museum of Islamic and Turkish Art, Istanbul, 192 Muslim ibn Quraysh, 202, 204

al-Musta∂È, Caliph, 130, 160, 163 Mu߆afa, A˙mad, 266 al-Mustanjid, Caliph, 160 al-Mustansir, Caliph, 160, 162, 170–1, 174 al-Mustarshid, Caliph, 159 al-Mustaʾsim, Caliph, 160, 173 al-Muʾtamid, Caliph, 386 al-Mutawakkil, Caliph, 54, 371, 409–11 Muʿtazilis, 299, 300 mysticism, 76, 79–80; see also Sufism Najaf Badr al-DÈn’s burial at, 189, 191 shrine of Imam ʿAli, 189, 445, 454–5, 466, 467, 472 al-Najjar, ʿAbdallåh b. Ahmad, 115 Najm al-DÈn AyyËb, 341 Najm al-DÈn Il-Ghazi, 417 Naramsin, 55 al-Nasir, Caliph, 160, 161, 163, 164, 170–1, 174, 193, 335 naskh script, 172, 266, 279, 280–2, 284, 286–9, 289–90, 292–6, 297, 321 Nasr, S. H., 391 Natanz, 210 new ʿAbbasid KËfic script see semiKËfic script New Style script see semi-KËfic script Nilometer, Cairo, 315, 316 Nishapur, 202, 226 Nisibis, 131, 132, 482 Nizam al-Mulk, 50 nomadic tribes, 94, 97, 100; see also Bedouins norias, 389, 390, 413 Normans, 227, 252, 380–1, 414–15 North Africa architecture, 40, 202, 207, 217–18, 224, 227–9, 235–53 calligraphy, 285, 300, 318, 324, 333, 338–40, 342 Fatimid conquest of, 298 gardens, 371, 376 poetry, 411 see also individual locations

521

index

nËr (blissful radiance), 465, 466, 472–4 NËr al-DÈn, 3–21, 38, 42, 48, 55, 109–12, 118–20, 129–44, 160, 179–86, 192–5, 217–18, 314, 321, 327–33, 335–6, 340, 342, 439, 482, 493 NËr al-DÈn minbar, al-Aqsa, 13–17, 21, 110, 111–13, 112–13 NËr al-DÈn minbar, Hama, 11–13, 14, 20, 110, 111, 112, 331, 332 NußaybÈn see Nisibis oasis towns, 406 occasionalism, 211–17, 227 Orientalism, 26–7, 457 orthographic marks, 262, 270, 291, 323, 333, 355 Ottomans, 75, 77, 107, 165, 167, 168–9, 266, 487, 491–3, 494 Pact of ʿUmar, 131 palaces ʿAbbasid Palace, Baghdad, 168–71, 171 Abu Muslim palace, Merv, 54 Alhambra Palace, Granada, 52, 210–13, 369, 386, 388, 388, 393, 394–7, 395–7, 419–20, 420 Artuqid palace, Diyarbakir, 55, 99, 384, 385 Ayyubid palace, Aleppo, 47, 49–60, 52, 53, 65, 66, 69, 99–100, 102, 417 Ayyubid palace, Karak citadel, 55, 99 Ayyubid palace, Qalʿat Najm, 55, 99, 102 Ayyubid palace, Qalʾat ÍahyËn, 53, 55, 65, 99, 102, 384–5, 386 Ayyubid palace, Roda Island, 55 Balkuwara palace, Samarra, 54, 67, 371, 372, 376, 409, 410, 411 Bosra, 55, 99 in citadels, 47, 49–60, 52, 53, 65, 66, 69, 95, 99–100, 102, 412–20 courtyards, 53–8, 238, 369 Dar al-Hajar, Marrakech, 237–8, 252 dår al-imåra, Kufa, 54 Dar al-Khilafa, Baghdad, 159, 162

desert palaces, 10, 30, 370, 425–35 dimensions, 57–8, 95, 412 four-iwån plans, 53–5, 57, 99, 417 gardens, 162, 369–98, 408–20 Hammam al-Sarah, 426 Óårim, 99 Imperial Palace, Constantinople, 252 Jawsaq al-Khaqani, Samarra, 29, 30, 371, 411 Khirbat al-Mafjar, 30, 370, 371, 408, 426, 435 Lashkar-È Båzår, 55, 56, 373, 375–6, 377–8 location within cities, 47, 95 MadÈnat al-Zahråʾ, Cordoba, 247, 248, 373–5, 374–7, 387, 411–12, 412 al-Mansur palace, Bijaya, 415 Ma†bakh al-ʿAjami, Aleppo, 55, 56, 57, 68 muqarnas portals, 52–3, 57 Palace of Diocletian, 28, 30 palatial domes, 252 Parthian palace, Assur, 54, 55, 56 Pasargadae, 370 Qalʿat banÈ Hammåd, 202, 378–80, 379–80 Qasr al-ʿAshiq, Samarra, 54 Qasr al-Banat, Raqqa, 55, 68, 99 Qasr al-Haruni, Samarra, 54 Qasr al-Hayr al-Gharbi, 30 Qasr al-Jiss, Samarra, 54 Qasr al-Mubårak, Seville, 386–7, 387 Qusayr ʿAmra, Jordan, 425–35, 425, 428–30 sculptural ornament, 51–2, 57 shådirwån fountains, 56–7, 99 Shawbak, 99 Tell Harmal palace, 52 tripartite courtyard façades, 55–6, 57, 100 Umayyad palace, Amman, 54 Umayyad palaces, 10, 30, 54, 370, 425–35 Palatine Chapel, Palermo, 381, 381, 414 Palestine, 108–9, 315, 317, 318, 334 pan-Islamism, 450, 463, 470

522

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

paper manuscripts, 270–3, 276, 291, 292–4, 358–60, 362 paradise, 86–8, 369–70, 388–9, 396–7, 405, 407–9, 420–1, 424–35 Parthian architecture, 54, 55–6 Parthian palace, Assur, 54, 55, 56 Parviz II, 370 Pasargadae, 370 patronage in Aleppo, 47, 57, 79–90, 96–8, 109–10, 334 of architecture, 27, 38–9, 47, 57, 75–90, 96–8, 109, 129–30, 132, 141, 143, 158, 160, 179–95, 334, 439 of art, 75 Badr al-DÈn’s, 141, 179–80, 186–95 Badr al-JamålÈ’s, 109 in Baghdad, 158, 160, 173, 478 of calligraphy, 173, 355 in Damascus, 109–10 Îayfa KhåtËn’s, 75–90 decline of, 38–9 female patronage, 75–90 of the futuwwa, 160 of hospitals, 478, 480, 481–2 of illustrated manuscripts, 194 of madrasas, 80, 132, 482 in Mosul, 129–30, 132, 143, 179–95 al-Mustansir’s, 160 al-Nasir’s, 160 NËr al-DÈn’s, 109–10, 118–19, 129–30, 132, 143, 179–86, 192–5, 439, 482 Salå˙ al-DÈn’s, 113, 119, 482 of scholars and mystics, 79–80 of woodwork, 109–10, 113, 118–19 al-Ûåhir GhåzÈ’s, 47, 57, 81, 96–8, 334 pavilions La Cuba, Sicily, 252, 380, 414 Dar al-Khilafa, Baghdad, 162 La Favara, Sicily, 252, 380 al-Firdaws, Mardin, 382, 382–4, 415–17, 416 in gardens, 252, 370–3, 375–6, 376, 379–82, 382–4, 387–8, 392, 404, 410–18

Óayr al-Wu˙Ësh, Samarra, 410 Lashkar-È Båzår, 376 MadÈnat al-Zahråʾ, Cordoba, 375, 411–12 Qara Serai, Mosul, 187 Salon Rico, MadÈnat al-Zahråʾ, 375, 376, 411–12 La Ziza, Sicily, 252, 380–1, 380–2, 414, 415 Pentz, Peter, 98 Persia architecture, 27, 54, 56 calligraphy, 281 gardens, 369, 371, 376, 387, 388, 404, 426 ornament, 111 poetry, 421 see also Iran; and individual locations Pharaoh’s Tray fountain, Samarra, 372–3, 373 phenomenology, 458, 465, 474 pilgrimage, 76, 181, 226, 453–7, 462, 487, 493 Pinder-Wilson, R., 387–8 poetry, 3, 210–11, 213, 381–2, 393–4, 397, 404–21, 426, 434–5 pointed arches, 7, 9, 43, 161, 167, 169, 241 polylobed arches, 34 pools, 371–3, 376, 377, 379–87, 389, 393, 393–4, 409–15, 417–19, 439, 483, 484, 488–9 Porta Aurea, Palace of Diocletian, 28, 30 portals ʿAbbasid Palace, Baghdad, 169 Ayyubid palace, Aleppo, 52, 52–3, 57, 65, 100 Ayyubid palace, Qalʾat ÍahyËn, 53, 65, 100, 102 BÈmåristån al-NËrÈ, Damascus, 110, 110, 229 Dhafariya (Wastani) Gate, Baghdad, 161–2, 162 as expressions of power, 10, 162 Great Mosque of Cizre, 192 Madrasa al-Firdaws, Aleppo, 86–7 Madrasa al-Hallåwiyya, Aleppo, 328, 330

523

index

Madrasa al-Mirjaniyya, Baghdad, 176 Madrasa al-Mustansiriyya, Baghdad, 166–7, 169 Madrasa of Sultan Hassan, Cairo, 232, 232, 233 Madrasa al-Sultaniyya, Aleppo, 48 Madrasa al-Zahiriyya, Aleppo, 231 Mår A˙Ëdemmeh monastery, Mosul, 192 Mår Bahnåm monastery, Mosul, 192 Mosque of al-Muʿayyad Shaykh, Cairo, 232, 232 muqarnas portals, 52–3, 57, 202, 224, 229–33, 231–3 projecting portals, 7, 10, 167 Qastal al-Shuʾaybiyya, Aleppo, 7, 10 shrine of Imam ʿAwn al-DÈn, Mosul, 191–2, 193, 336, 336–7 porticos, 144, 171, 410, 444; see also arcades positivism, 26–7, 106–7, 262, 458 power, 10, 46–7, 50–1, 54–5, 59–60, 94, 120, 158, 162, 176, 313, 343–5, 360, 362 prayer, 19, 139–40, 141, 190, 462, 466, 468, 491 projecting portals, 7, 10, 167 propaganda, 3, 11, 42, 99, 259, 317 proportional writing (al-kha†† almansËb), 266, 294, 298, 355, 357, 360, 362 pyramidal domes, 137, 189, 210, 244 Qadi Ahmad, 173 al-Qadir, Caliph, 159, 212, 227, 250, 299–300, 325 al-Qåhir, Caliph, 360 Qalʿat banÈ Hammåd, 202, 378–80, 379–80 Qalʾat Jaʿbar, 17, 97 Qalʿat al-Mu∂Èq, 97, 98, 101 Qalʿat Najm Ayyubid palace, 55, 99, 102 citadel, 98–9, 99, 100, 101, 102, 103, 417 citadel entrance block, 98, 102

citadel mosque, 100, 103 conquest by al-Ûåhir GhåzÈ, 97 Qalʾat ÍahyËn Ayyubid palace, 53, 55, 65, 99–100, 102, 384–5, 386 citadel, 98, 99–100, 417 Qalawun, 488 Qalawun’s hospital, Cairo, 488–90, 489 Qalb Lozeh, 33, 33, 41 Qalb Meez, 33 qanåts, 251, 389 Qara Serai, Mosul, 187 Qaråmita, 298 qasidas, 406, 411, 418, 434 Qasr al-ʿAshiq, Samarra, 54 Qasr al-Banat, Raqqa, 55, 68, 99 Qasr al-Haruni, Samarra, 54 Qasr al-Hayr al-Gharbi, 30 Qasr al-Jiss, Samarra, 54 Qasr al-Mubårak, Seville, 386–7, 387 Qasr-i Shirin, 370 Qastal al-Shuʾaybiyya, Aleppo, 7–10, 8–9, 18, 20, 39, 40, 42–3, 317, 330–1 Qtaife, 493 Qubbat al-Barudiyyin, Marrakech, 227–8, 229, 235–53, 236, 241–3, 245–9 Qumm, 445, 455, 466 Qumriyya mosque, Baghdad, 163 Qurʾan canonisation, 261, 297–300, 360–2 depiction of paradise, 86–7, 388–9, 396–7, 405, 407–9, 420–1, 424–5, 429, 433–4 esoteric (bå†in) interpretations, 250, 299, 312, 321, 343 eternal and uncreated nature of, 299, 310, 355 exoteric (Ωåhir) interpretations, 143, 184, 250, 299, 312, 321, 343, 355 recensions of, 261, 275, 286–8, 291, 297–300, 360–2 Qurʾanic calligraphy Anatolia, 173 application of geometric principals, 261, 264–7, 265, 294–7, 355

524

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Qurʾanic calligraphy (cont.) Baghdad, 171–3, 259, 285, 342–3 before Ibn Muqla, 262–3, 356–7 canonisation, 266, 360, 362 colophons, 172, 274, 275, 277, 288, 291 cursive script, 172, 173, 259, 262–3, 280–301 Eastern KËfic script, 172, 269–70, 271, 290, 294 Egypt, 285, 300 floriated KËfic script, 296 frontispieces, 120, 172, 275 Ibn al-Bawwab and his influence, 172, 173, 248, 261, 277–91, 292, 297–300, 310, 324, 328, 343, 354–5, 362 Ibn Muqla and his influence, 172, 260, 261, 263–77, 285, 292, 294–8, 300, 310, 324, 354–63 interconnection, 173, 282, 324 Iran, 172, 173, 289 KËfic script, 172, 262–77, 264, 271, 273–8, 285, 291–2, 294, 296, 297, 300, 301, 342–3, 355–8, 362 legibility, 270, 276, 281, 282, 291, 299, 300, 355, 356, 360 mu˙aqqaq script, 266, 290, 294 naskh script, 172, 266, 279, 280–2, 284, 286–9, 289–90, 292–6, 297 new ʿAbbasid KËfic script see semi-KËfic script North Africa, 285, 300 patronage, 173, 355 proportional writing (al-kha†† almansËb), 266, 294, 298, 355, 357, 360 ray˙ån script, 173, 266 research approaches, 259–60, 261–2 riqåʿ script, 266 semi-KËfic script, 267–77, 273–8, 285, 291, 342, 355, 357, 358, 362 signatures, 173, 269, 272 status of calligraphers, 355 sËra headings, 266, 281, 282, 284, 285, 289, 291 tawqÈʿ script, 266, 289, 290

thuluth script, 172, 173, 266, 279, 281–5, 284–5, 289–90, 294, 324 transformation from angular to cursive, 259–301, 310 verse counts, 272, 272, 275–6, 275–6, 286–8, 290, 291, 292, 294–6, 298 Qurʾanic inscriptions cenotaph of Sayyida Få†ima, Damascus, 108 hospitals, 484 Madrasa al-Firdaws, Aleppo, 86–7 Madrasa al-Mustansiriyya, Baghdad, 167 Mosque of NËr al-DÈn, Hama, 11, 11 Mosque al-NËri, Mosul, 19, 138–42, 143, 150–1, 182, 184 NËr al-DÈn minbar, Hama, 13 shrine of Iman ʿAwn al- DÈn, Mosul, 191–2, 336–8, 337 shrine of Imam Yahya b. alQasim, Mosul, 190–1, 190–1 shrine of Sayyida Ruqayya, Damascus, 445–52, 454, 455, 469, 469 shrine of Sayyida Zaynab, Damascus, 445, 469, 470 triple shrine, Raqqa, 445 Qusayr ʿAmra, Jordan, 425–35, 425, 428–30 Qutb al-DÈn MawdËd, 19, 130, 181 al-Rå∂i, Caliph, 360 Ramla, 482 Raqqa, 17, 55, 68, 99, 135–6, 443, 445, 448, 482 Ras al-ʾAyn, 97 Rawcliffe, Carole, 487 raw∂iyyåt, 383–4 ray˙ån script, 173, 266 Raymond of Antioch, 6, 329 residential architecture, 38, 41, 84–5, 383–4, 384 residential cells, 80, 85 resthouse of Caliph al-Mutawakkil, Samarra, 54 Ribåt of Susa, 10 ribåts see khånqåhs ribbed domes, 137, 244 Richard, Yann, 453–4

525

index

riqåʿ script, 266 Riegl, Alois, 26 al-Risåla al-Qådiriyya, 299–300 Risålah fiʾl-kha†† al-mansËb (Ibn Muqla), 266 ritual, 462–74; see also ceremony Robinson, Cynthia, 404 Roda Island, 55 Rogers, J. Michael, 27 Roman civilization, 7, 10, 30, 55, 389 Rosenthal, Franz, 480, 484 Round City of Baghdad, 47, 55, 158 royal iconography, 52–3, 425, 426–9, 431–3, 435 royal manuscripts, 289–90 Rubin, Miri, 487 Ruggles, D. Fairchild, 404 Rusafa, 43, 159 Saʿdi, 421 Safavids, 195, 472 St Helena cathedral, Aleppo, 5–6; see also Madrasa al-Hallåwiyya, Aleppo Salå˙ al-DÈn (Saladin), 3, 12–13, 15–17, 78, 94–5, 96, 112, 113, 119, 160, 180, 321, 340–1, 482 Salå˙ al-DÈn II (Saladin II), 79, 83, 115 Salmån b. MaʿålÈ, 112, 113 Salon Rico, MadÈnat al-Zahråʾ, 375, 376, 411–12 salsabÈl fountains, 379–84, 381, 413–14, 415, 417, 417 salvaged materials, 27–8, 136 Samanids, 322 Samarra Abu Dulaf mosque, 54 Bab al-ʿAmma, 29, 30 Balkuwara palace, 54, 67, 371, 372, 376, 409, 410, 411 Great Mosque, 372–3 Óayr al-Wu˙Ësh, 371–2 Jawsaq al-Khaqani, 29, 30, 54, 371, 411 location of palaces, 47, 95 Pharaoh’s Tray fountain, 372–3, 373 Qasr al-ʿAshiq, 54 Qasr al-Haruni, 54

Qasr al-Jiss, 54 resthouse of Caliph alMutawakkil, 54 shrine of Imam Dur, 164, 202–4, 203–4, 206, 207, 226–7, 228, 246 Sanbak al-Badri, 190 såqiyas, 389–90, 391, 413, 414 Sarmin, 34–5, 37–8 SarËj, 97 Sassanians, 56, 176, 370, 371, 425, 426, 431, 432, 435, 480 Sauvaget, Jean, 8–9, 12, 442 Sauvaire, Henri, 441 Sayf al-Dawla, 30 Sayf al-DÈn Ghåzi I, 5, 130, 133–5, 179, 182–3 Sayf al-DÈn Ghåzi II, 130–1 Schacht, Joseph, 481 Schimmel, Annemarie, 80 Schmidt, Hansjorg, 165, 167–8 scientific manuscripts, 173–4, 357 sculpture Aleppo citadel, 51–2, 57 Talisman Gate, Baghdad, 161 Taq-e Bostan, 425, 431–2, 431–3 secular calligraphy, 262–3, 265–6, 270, 273, 292, 294, 356–9, 362 Seljuq domes, 202, 227 Seljuqs, 28, 47, 97, 120, 138, 142, 158–60, 163, 167, 175, 179, 202, 248, 250, 252, 321, 325, 327–8, 335, 342, 343, 387–8, 480 semi-KËfic script, 172, 267–77, 273–8, 285, 291, 342, 355, 357, 362 Seville, 237, 251, 386–7, 387 al-Shabushti, 491 shådirwån fountains, 56–7, 99, 378–84, 380, 413–15, 417, 421 shafåʿa (intercession), 465, 468–71 al-ShåfiʿÈ, Imåm, 113 Shafiʿis, 48, 250 Shahåda, 11, 13, 111, 141, 326, 331 Shajar al-Durr, 79, 84, 90 ShariatÈ, ʿAlÈ, 455 Shatt al-Nil, 208, 209–10, 216 Shawbak, 99 Shayzar, 97 Shiʿism, 5, 89–90, 113, 141, 159, 160, 180–1, 186, 190–5, 204, 297–9, 327–8, 439–58, 462–74

526

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

shrines Dome of the Rock, Jerusalem, 41, 235 Kazimayn shrine, Baghdad, 445, 454, 466 Maqam Ibrahim, Aleppo citadel, 47, 48, 60, 110, 111, 111, 112, 115, 331, 332 Maqam Ibrahim, Southern Aleppo, 48, 51 Mashhad al-Óusayn, Aleppo, 89, 441, 442, 449 Mashhad al-Dakka, Aleppo, 4, 30, 328 and ritual practices, 462–74 sacralisation, 441 sectarian destruction of, 195 shrine of ʿAbd al-Qadir al-Gailani, Baghdad, 159 shrine of Abu Hanifa, Baghdad, 160 shrine of al-Faraj ibn Barquq, Cairo, 225, 233 shrine of Få†ima al-MaʿsËmah, Qumm, 445, 455, 466 shrine of Imam ʿAli, Najaf, 189, 445, 454–5, 466, 467, 472 shrine of Imam ʿAwn al-DÈn, Mosul, 187, 191–2, 193, 210, 212, 217–18, 336–7, 336–8 shrine of Imam al-Dawr, Samarra, 202–4, 203–4, 206, 207, 226–7, 228, 246 shrine of Imam al-Óusayn, Karbala, 445, 454–5, 466 shrine of Imam Reza, Mashhad, 445, 455, 466 shrine of Imam Shafiʾi, Cairo, 231, 232, 340 shrine of Imam Yahya b. al-Qasim, Mosul, 186–91, 187–92, 210, 336 shrine of Muhammad al-Jawad, 160 shrine of Musa al-Kazim, Baghdad, 160 shrine of al-Najmi, Shatt al-Nil, 208, 209–10, 216 shrine of Sayyida NafÈsa, Cairo, 108, 109, 109

shrine of Sayyida Ruqayya, Cairo, 108, 109, 226, 227 shrine of Sayyida Ruqayya, Damascus, 439–58, 444, 446, 450–5, 462, 466, 469, 469–71, 471, 473, 473 shrine of Sayyida Sukaina bint ʿAlÈ, Darayya, 447, 473 shrine of Sayidda Zaynab, Damascus, 441, 443, 445, 454, 455, 462, 463–4, 466, 468, 469, 470, 471 shrine of Sayyida Zaynab, Sinjår, 137 shrine of Shaykh ʿAbd al-Samad, Natanz, 210 shrine of the ʿAbbasid Caliphs, Cairo, 232 shrine of ʿUmar al-Suhrawardi, Baghdad, 159, 161, 164, 165 triple shrine, Raqqa, 443, 445, 448 visitation, 453–7, 462, 464, 465–8, 468 see also mausoleums al-Shughr, 96, 98 Sibt Ibn al-Jawzi, 165 Sicily, 227, 252, 298, 338, 380–2, 393, 411, 414–15 signatures, 16, 34–5, 38, 105, 106, 112, 112, 115–18, 173, 231, 268, 272, 310, 335 Silvan see Mayyafariqin Sinjår, 131, 137, 143 al-Sirmani, Hasan ibn Mufarrij, 34–5 Sitt al-Shåm, 80 Sivan, Emmanuel, 3 Siyåsatnåma (Nizam al-Mulk), 50 Soucek, Priscilla, 440 Sourdel-Thomine, Janine, 317, 327 Sousse, 249 sovereignty, 4, 12–13, 17, 79, 90 Spalato, 28 Sperl, Stefan, 410 spiritual medicine, 482, 491, 492, 494 stone architecture, 27, 30–7, 229–31, 240 street medicine, 491 striped masonry, 52

527

index

stucco, 135–6, 138, 141, 152, 184, 186, 205, 206, 207, 226, 229, 232, 247, 249–50, 317, 378 Sufism, 76, 80–90, 129, 159, 160, 165, 181, 193–4, 469, 492 al-Suhrawardi, Shihab al-DÈn ʿUmar, 159, 161, 165 SuhrawardÈ MaqtËl, 89 Sulaymån b. Jandar, 94 SunnÈ Islam, 4, 48, 54, 89–90, 113, 119–20, 129, 141, 143, 159–60, 179–82, 184, 194–5, 217, 227, 236, 250–1, 301, 310, 321, 327–9, 343–4, 355, 440, 457, 462, 482 SunnÈ Revival, 90, 119–20, 143, 159, 175, 179–80, 195, 301, 310, 321, 327, 329, 343–4, 482 sËra headings, 266, 281, 282, 284, 285, 289, 291 surveillance, 48, 51 Taht al-Qalʿa, Aleppo, 48, 81 Talisman Gate, Baghdad, 161 Tall Båshir, 96 Taq-e Bostan, 425, 431–2, 431–3 Tarikh al-bimaristanat fi l-Islam, 477 tawassul (intercession), 465, 468–71 taw˙Èd, 107, 141 al-Taw˙ÈdÈ, AbË Óayyån, 264 tawqÈʿ script, 266, 289, 290 taxation, 131, 181 Tayma, 406 Tell Harmal palace, 52 terraces, 370, 371, 372, 375–6, 387, 411–12 Terrasse, Henri, 206, 239, 251 theology, 211–17, 227, 250, 299 Thévenot, Jean de, 490 thuluth script citadel mosque, Aleppo, 333, 334 Gate of the Two Lions, Aleppo, 333, 334 Great Mosque of Tlemcen, 337, 338 Madrasa al-Firdaws, Aleppo, 334, 334 Madrasa al-Hallåwiyya, Aleppo, 328

Maqam Ibrahim, Aleppo citadel, 331, 332 monumental calligraphy, 11, 13–15, 138, 161, 173, 190–2, 248, 266, 323, 325, 328, 331–9, 331–40, 342 Mosque of NËr al-DÈn, Hama, 11 Mosque al-NËri, Mosul, 138, 335–6, 335–6 Mosque al-Qarawiyyin, Fez, 338–9, 338–9 NËr al-DÈn minbar, al-Aqsa, 13, 14–15 NËr al-DÈn minbar, Hama, 331, 332 Qurʾanic calligraphy, 172, 173, 266, 279, 281–5, 284–5, 289–90, 294, 324 shrine of Imam ʿAwn al-DÈn, Mosul, 191–2, 336–7, 336–8 shrine of Imam Yahya b. alQasim, Mosul, 190, 336 Talisman Gate, Baghdad, 161 thuluth-ashʿar script, 282, 289 Tim, 202, 226 Timurids, 165, 202, 232 Tinmal, 244 titulature ʿAli b. Yusuf b. Tashufin, 248, 250 Badr al-DÈn, 192, 336 Îayfa KhåtËn, 79, 83–4 Ma˙mËd of Ghazna, 325 al-Nasir, 161 NËr al-DÈn, 3, 4–5, 6, 11, 13, 328, 329 Zangi, 4, 6, 328 Tlemcen, 248, 338 Toledo, 240, 244 Topkapı Saray Library (TKS) manuscript A3, 285, 285 manuscript EH-42, 271 manuscript HS-22, 272, 273, 274 manuscript R-10, 272, 276, 278 manuscript R-38, 270, 272, 275, 276, 276 manuscript Y-752, 272, 276, 277 trade, 97, 98, 100 trilobed arches, 241, 243, 323 tripartite courtyard façades, 55–6, 57, 100, 167

528

THE PRODUCTION OF MEANING IN ISLAMIC ARCHITECTURE

Tripoli, 5 triumphal arches, 10, 55–6 truncated cone castles, 98 Tughril Beg, 48, 159 Tughril III, 161 ÊËlËnids, 315 Tunisia, 338, 389; see also North Africa; and individual locations tunnels see underground passages Türk ve Islam Eserlerie Müzesi (TIEM) manuscript 431/2, 283, 284, 289 manuscript 449, 172, 283, 288, 288–9 manuscript 12800, 270 Turkey see Anatolia Twelver Shiʿism, 160, 194, 195, 328, 450–2, 465, 469–71 ʿUbayd b. MaʿålÈ, 113 ʿUbaydullah b. Yunis, 251 Ukhaidir, 56 al-ʿUlabi, Akram, 442, 457 ʿUmar bin ʿAbd al-ʾAzÈz, 10 ʿUmar ibn al-Kha††åb, Caliph, 7, 10, 330 ʿUmar al-Mallå, 129, 132, 180, 181 Umayyad palaces, 10, 30, 54, 370, 425–35 Umayyad poetry, 434–5 Umayyads, 10, 28, 30, 194, 373, 407–9, 425–35, 479 underground passages, 49–51, 63 University Museum, Philadelphia (UMP) manuscript NEP-27, 283, 289–90, 292–3 Uqaylids, 202 Urfa see Edessa ʿUthmån, Caliph, 297–8, 354, 355 van Berchem, Mat, 16, 321–2, 326, 327, 340 vegetal ornament, 105, 108, 109, 111, 113, 118, 120, 138, 162, 228, 244, 247 vellum manuscripts, 270–3, 276, 360, 362 Veramin, 202 vernacular architecture, 54, 344

verse counts, 272, 272, 275–6, 275–6, 286–8, 290, 291, 292, 294–6, 298 vertical format manuscripts, 272–5 Victory Monument of Hamn alRashid, Hiraqlah, 54, 67 ‘View of Baghdad’, 205 visitation (ziyåra), 453–7, 462, 464, 465–8, 468 Vitruvius, 389 vocalisation marks, 270, 355 voussoirs, 37, 51, 52, 53, 56, 162, 192 Wahhabism, 195 walåʾ (allegiance), 452, 465 walåya (devotion), 465, 467–8 al-WalÈd I, Caliph, 7, 427 al-WalÈd II, Caliph, 408, 425, 427, 433–5 walled gardens, 370, 375 waqf, 132, 135, 478, 487–8, 492–4 waqfiyyas, 487–8, 492–4 al-Wasiti, Yahya b. Mahmud, 174 water basins, 371, 372, 375, 391, 394 baths, 53, 99 canals, 251, 372, 383, 409 fountains see fountains in gardens, 369–98, 404–21 hydraulic technologies, 389–93, 390–2, 404–5, 413–14, 414 irrigation, 371, 389, 390, 413, 426 khettara water-supply system, 251 norias, 389, 390, 413 oasis towns, 406 in palaces, 56–7, 58 and poetry, 404–21 pools, 371–3, 376, 377, 379–87, 389, 393, 393–4, 409–15, 417–19, 439, 483, 484, 488–9 qanåts, 251, 389 såqiyas, 389–90, 391, 413, 414 water supply systems, 57, 237, 240, 251, 389 water symbolism, 210–11, 388–9, 418–20 Watson, Andrew, 404, 413 Watzinger, Carl, 441 Whelan, Estelle, 261 Wiet, Gaston, 96

529

index

wilåya (spiritual authority), 465, 467–8 Wilber, Donald, 404 William I, 414 William II, 252, 414 Williams, Caroline, 75, 113 windows Alhambra Palace, Granada, 212–13 arched windows, 55, 241 Ayyubid palace, Aleppo, 55 Italianate windows, 41 Maqam Ibrahim, Aleppo citadel, 110, 111, 112 in muqarnas domes, 210, 212–13 Qubbat al-Barudiyyin, Marrakech, 241 residential architecture, 38, 41 shrine of Imam ʿAwn al-DÈn, Mosul, 212 woodwork arabesque ornament, 108, 110, 111, 113, 115, 118, 120 Ayyubid period, 110, 113–15 Egyptian, 108–9, 110, 113, 119 Fatimid period, 107–9 geometric ornament, 105–20, 109–18 KËfic inscriptions, 108, 108, 111 patronage, 109–10, 113, 118–19

status of artisans, 105–6 Syrian, 105–20 vegetal ornament, 105, 108, 109, 111, 113, 118, 120 Zangid period, 109–13 Wultzinger, Karl, 441 xenodocheion see hospitals Xenophon, 370 Yaqut al-Mustaʾsimi, 173, 355 Yazidis, 194 Yemen, 360 Yusuf b. Tashufin, 237, 250, 340 al-Ûåhir GhåzÈ, 47–9, 51, 57, 78–9, 81, 89, 94–103, 111, 334 Zangi, ʿImåd al-DÈn, 4, 6, 30, 179, 328 Zangids, 16, 20, 47, 48, 77, 80, 109–13, 120, 130, 160, 175, 179–95, 227, 248, 321, 334, 343 ziyåra (visitation), 453–7, 462, 464, 465–8 La Ziza pavilion, Sicily, 252, 380–1, 380–2, 414, 415 zoological gardens, 370, 371–2 Zuhair b. Abi Sulma, 406, 410 Zumurrud Khatun, 163, 164–5